Technical Report
                              EPA 520/3-78-006
            August 1978
U.S. Environmental Protection Agency
    Office of Radiation Programs
  Criteria and Standards Division
      Washington, B.C.  20460

-------

-------
                                PREFACE


     This document contains the results of studies commissioned by the

Environmental Protection Agency and carried out by the General

Electric Company, Tempo Division, under contract to the Agency.

Funding was provided by the Office of Federal Activities under a

program to develop information required to improve the quality of

Environmental Impact Statements and simil'ar public information

materials issued by the Agency.

     The material was developed specifically in support of Federal

Radiation Protection Guidance on Dose Limits for Persons Exposed to

Transuranium Elements in the General Environment, prepared by the

Office of Radiation Programs.  It is intended to directly supplement

the guidance and provide information useful for effective planning of

remedial actions and for preparation of Environmental Impact

Statements where required.  Although written primarily from the

viewpoint of radiation protection, the scope of the document is

sufficiently broad so that it can be expected to have widespread

applicability to many other situations where cleanup of residual

environmental contamination may be required.

     Comments on this document would be appreciated.  These should be

sent to the Director, Criteria and Standards Division (ANR-460),

Office of Radiation Programs.
                                   W. D. Rowe, Ph.D.
                            Deputy Assistant Administrator
                                 for Radiation Programs

-------
                            DISCLAIMER
     This report has been reviewed by the Office of Radiation
Programs, U.S. Environmental Protection Agency, and approved for
publication.  Approval does not signify that the contents necessarily
reflect the views and policies of the U.S. Environmental Protection
Agency, nor does mention of trade names or commerical products
constitute endorsement or recommendation for use.

-------
                                    ABSTRACT
     The basic types of cleanup procedures for removing or deactivating
spilled contamination involve moving people and animals from the affected area,
scraping or grading the contaminated soil into windrows, plowing the contamina-
tion under, or digging up the contamination and hauling it away.  This report
describes and evaluates these various land-type cleanup effects in terms of
impact of the technique on the environment.  The cleanup procedures are listed
in the following table with brief definitions of proposed cleanup actions.
Conclusions about the effects of cleanup on the soil,  vegetation, and animals
in an area are summarized in the composite effects shown in the following
table as a composite ranking for all the areas studied.  The rankings are
numbered from 0 through 5 for each cleanup treatment.   Interpretations of
these rankings are as follows:

          0 - No measurable change is produced in the ecosystem.
          1 - The preferred cleanup because adverse environmental
              effects upon recovery and side effects of treatment
              are minimal0
          2 - A conditionally acceptable treatment because of sig-
              nificant impact by the treatment and/or the equipment
              upon the area.
          3 - Acceptable as a "last resort" cleanup to remove
              exceptionally hazardous material while incurring
              maximum acceptable impact.
          4 - Causes unacceptable damage but can be used as an
              interim cleanup if the injury is erased during the
              final treatment.
          5 - Not applicable to the land type for which it is proposed.

     The rankings considered the environmental insult generated during the
cleanup, the physical possibility of restoring the area to its original pro-
ductive state, side effects caused by the equipment needed to perform the
cleanup, the impact upon the environment adjacent to the cleaned up area, and
the social acceptance of the cleanup work0  Not all treatments were expected
to be evaluated with all land types; these exceptions are indicated in the
table as being outside the scope of this effort.

     The primary criterion for the evaluation of cleanup effects was the suit-
ability of the denuded surface as a medium for plant growth.  Establishment of
a vegetative cover before serious wind and water erosion despoiled the site was
considered first and, if revegetation was probable, the time to recover the
initial pre-cleanup productivity was considered next.  This is the context in
which the rankings were defined.
                                     m

-------
»
1
2
r*
w
M-
O.
0
M
O
O
•o
(»
o
H»
s
1
w



























Suburban


4>
4k
4>
M
*»
Jk
K)
4>
K>
A


H-
4V
K>
K)
A


O4
tn
»
(B
»
e>
LAND USE CLASSES
Agriculture


O4
04
04
H^
H*
t-*
>—•
1— •
h- »
-


N>
ro
K)
K>
O4


O4
4V
k—
W
e>
4V
Coastal Inter-Tidal
Marshes


04
O4
O4
O4
tn
tn
NJ
O4
O4
O4


K)
04
4V
4V
4V


O4
»
0)
B>
P
»
Tundra


tn
tn
tn
i— •
Ul
4v
N>
4V
H-
K>


to
N>
tn
tn
tn


O4
»
M
P
(0
B)
Mountain, subalpine


K>
N>
O4
K)
4V
4v
04
4V
NJ
O4


O4
M
O4
04
4V


O4
V
w
04
tn
»
Coniferous Forest


K)
to
O4
NJ
4V
4V
M
4V
-
hO


ro
O4
O4
O4
4v


04
4V
0
V
09
(U
Deciduous Forest


K>
ls>
OJ
NJ
04
O4
O4
-E»
H-
NJ


K)
O4
NJ
M
4k


O4
.U
0
»
09
(U
Prairie


tn
tn
tn
-
-
-
NJ
O4
-
-


-
K>
-
NJ
O4


O4
04
w
V
(U
ft)
5
0 Z
o o
(/)
0 H
"« •<
rt -o
m
w


ro
O4
^
£t
*>
4^
A.
Ji.
.U
Xi


ii.
4^
4^.
«k
A.


-
4^
V
V
so
w

0-1 Natural
Rehabilitation
1-1 Chemical
Stabilization
1-2 Clear Cutting
Vegetation
1-3 Stumping and
Grubbing
1-4 Scraping and Grading
(<5 cm)
1-5 Shallow Plowing
(<10 cm)
1-6 Deep Plowing
(10 to 40 cm)
1-7 Soil Cover
(<25 cm)
1-8 Soil Cover
(25 to 100 cm)
2-1 Remove Plow Layer
(10 cm)
2-2 Remove Shallow Root
Zone (<40 cm)
2-3 Remove Scraping and Grading,
Mechanically Stabilize
2-4 Remove Plow Layer (10 cm),
Mechanically Stabilize
2-5 Remove Shallow Root Zone
(<40 cm), Mechanically Stabilize
2-6 Remove Scraping and Grading,
Chemically Stabilize
2-7 Remove Plow Layer (10 cm) ,
Chemically Stabilize
2-8 Remove Shallow Root Zone
(<40 cm), Chemically Stabilize
3-1 Barriers to Exclude
People
3-2 Barriers to Exclude Large
and Small Animals
4-0 Mechanical Stabilization by
Hard Surface
5-0 Application of
Sewage Sludge
6-1 High Pressure Washing
(<3 mm)
6-2 Flooding
(3 to 30 cm)
7-0 Soil Amendments
Added

-------
                                  CONTENTS
Abstract                                                                   iii
Figures                                                                  xviii
Tables                                                                     xxi
Acknowledgments                                                           xxiv
INTRODUCTION                                                                 1
   Statement of the Problem                                                  1
   Purpose of the Report                                                     1
   Scope of the Report                                                       2
   Literature Base                                                           5
   Study Approach                                                            5
   Impact Evaluation                                                         6
   Use of the Report                                                         7
   Conclusions                                                               9

                        PART  I - NATURAL ECOSYSTEMS
CHAPTER 1-DESERT                                                         1-1
   1.1  Overview                                                          1-1
        1.1.1  Desert Climate                                              1-2
        1.1.2  Soils                                                      1-4
        1.1.3  Topographic Factors                                         1-7
        1.1.4  Vegetation                                                  1-8
               1.1.4.1  Great  Basin Desert                                 1-8
               1.1.4.2  Mojave Desert                                     1-10
               1.1.4.3  Sonoran Desert                                    1-10
               1.1.4.4  Chihuahuan  Desert                                  1-11
               1.1.4.5  Desert Grassland                                  1-11
        1.1.5  Desert Succession                                           1-12
   1.2  Natural  Perturbations                                              1-13
        1.2.1  Natural Processes                                           1-13
        1.2.2  Fire                                                       1-14
        1.2.3  Climate                                                    1-17
   1.3  Man-Made Perturbations                                            1-18
        1.3.1  Fencing                                                    1-18
        1.3.2  Overgrazing                                                1-19
        1.3.3  Mechanical  Disturbance                                     1-20
        1.3.4  Atomic Test Target Areas                                    1-23
        1.3.5  Restoration                                                1-26
   1.4  Effects  of Cleanup Procedures on Deserts                           1-26
               (Treatment  1-1)  Chemical Stabilizers                       1-27
               (Treatment  1-2)  Clear Cutting  Vegetation                   1-30
               (Treatment  1-3)  Stumping and Grubbing                      1-32
               (Treatment  1-4)  Scraping and Grading                       1-33

-------
                            CONTENTS  (continued)
               (Treatment  1-5)
               (Treatment  1-6)
               (Treatment  1-7)
               (Treatment  1-8)
               (Treatment  2-1)
               (Treatment  2-2)
               (Treatment  2-3)

               (Treatment  2-4)

               (Treatment  2-5)

               (Treatment  2-6)

               (Treatment  2-7)

               (Treatment  2-8)

               (Treatment  3-1)
               (Treatment  3-2)
               (Treatment  3-3)
               (Treatment  4-1)
               (Treatment  4-2)
               (Treatment  5-0)
               (Treatment  6-1)
               (Treatment  6-2)
               (Treatment 7-0)
   1.5  Recovery Following Cleanup
        1.5.1   Unassisted Succession
        1.5.2   Cleanup Recovery Categories
   1.6  Quantitative Assessment of Cleanup  Impacts
        1.6.1   Impact Assessment
        1.6.2   Recovery Assessment
   1.7  Conclusions
   1.8  Desert References
CHAPTER 2-PRAIRIE
   2.1  Overview
        2.1.1   Tall grass Prairie
        2.1.2   Mixedgrass Prairie
        2.1.3   Shortgrass Prairie
        2.1.4   Other Grasslands
   2.2  Natural  Perturbations
        2.2.1   Drought
        2.2.2   Flooding
        2.2.3   Fossorial Animals
   2.3  Man-Made Perturbations
        2.3.1   Close Cropping and Grazing
        2.3.2   Compaction
        2.3.3   Plowing
Shallow Plowing
Deep Plowing
Soil Cover Less than 25 cm
Soil Cover 25 to 100 cm
Remove Plow Layer
Remove Shallow Root Zone
Scrape and Grade, Mechanically
Stabilize
Remove Plow Layer, Mechanically
Stabilize
Remove Shallow Root Zone, Mechanically
Stabilize
Scrape and Grade,.Chemically
Stabilize
Remove Plow Layer, Chemically
Stabilize
Remove Shallow Root Zone, Chemically
Stabilize
Barriers to Exclude People
Exclude Large Animals
Exclude Large and Small Animals
Asphalt Hard Surface Stabilization
Concrete Hard Surface Stabilization
Application of Sewage Sludge
High Pressure Washing
Flooding
Soil Amendments Added
1-37
1-37
1-37
1-39
1-39
1-39

1-39

1-40

1-40

1-40

1-41

1-41
1-41
1-41
1-42
1-43
1-44
1-44
1-44
1-44
1-44
1-46
1-46
1-52
1-52
1-53
1-55
1-58
1-60

2-1
2-1
2-1
2-2
2-3
2-3
2-4
2-4
2-5
2-5
2-6
2-6
2-7
2-7
                                      VI

-------
                         CONTENTS (continued)
2.4
2.5
2.6
2.7
2.8
2.3.4  Natural  Succession
2.3.5  Range Improvement
       2.3.5.1   Mulches
       2.3.5.2   Seeding
       2.3.5.3   Shrub Removal
       2.3.5.4   Moisture Trapping
Effects of Cleanup Procedures  on Prairie
       (Treatment 0-1)  Natural  Rehabilitation
            (Treatment 1-1)
            (Treatment 1-2)
            (Treatment 1-3)
            (Treatment 1-4)
            (Treatment 1-5)
            (Treatment 1-6)
            (Treatment 1-7)
            (Treatment 1-8)
            (Treatment 2-1)
            (Treatment 2-2)
            (Treatment 2-3)

            (Treatment 2-4)

            (Treatment 2-5)

            (Treatment 2-6)

            (Treatment 2-7)

            (Treatment 2-8)
                        Chemical  Stabilization
                        Clear Cutting Vegetation
                        Stumping  and Grubbing
                        Scraping  and Grading
                        Shallow Plowing
                        Deep Plowing
                        Soil  Cover Less than 25 cm
                        Soil  Cover 25 to 100 cm
                        Remove Plow Layer
                        Remove Shallow Root Zone
                        Remove Scraping and Grading,
                        Mechanically Stabilize
                        Remove Plow Layer,  Mechanically
                        Stabilize
                        Remove Shallow Root Zone, Mechanically
                        Stabilize
                        Remove Scraping and Grading,
                        Chemically Stabilize
                        Remove Plow Layer,  Chemically
                        Stabilize
                        Remove Shallow Root Zone, Chemically
                        Stabilize
                        Barriers  to Exclude People
                        Exclude Large Animals
                        Exclude Large and Small Animals
                        Asphalt Hard Surface Stabilization
                        Concrete  Hard Surface Stabilization
       (Treatment 3-1)
       (Treatment 3-2)
       (Treatment 3-3)
       (Treatment 4-1)
       (Treatment 4-2)
Recovery After Cleanup
2.5.1  Irreversible Changes
2.5.2  Rates of Recovery Following Cleanup
       2.5.2.1  First Year Following Cleanup
       2.5.2.2  Fifth Year Following Cleanup
       2.5.2.3  Tenth Year Following Cleanup
       2.5.2.4  Climax
Quantitative Assessment of Cleanup Impacts
2.6.1  Impact Assessment
2.6.2  Recovery Assessment
Conclusions
Prairie References
2-7
2-9
2-9
2-9
2-10
2-10
2-10
2-11
2-11
2-12
2-12
2-13
2-13
2-14
2-14
2-14
2-14
2-15

2-15

2-17

2-17

2-17

2-18

2-18
2-18
2-18
2-19
2-19
2-20
2-21
2-21
2-21
2-22
2-26
2-27
2-28
2-30
2-30
2-31
2-34
2-35
                                  vii

-------
                            CONTENTS (continued)
CHAPTER 3-DECIDUOUS FOREST
   3.1  Overview
        3.1.1  Mixed Mesophytic Forest
        3.1.2  Oak-Hickory Forests
        3.1.3  Appalachian Oak-Chestnut Forest
        3.1.4  Oak-Hickory-Pine Forest
        3.1.5  Southern Mixed Forest
        3.1.6  Beech-Maple Forest
        3.1.7  Maple-Basswood Forest
        3.1.8  Hemlock-Hardwood Forest
        3.1.9  Theoretical Considerations
   3.2  Natural Perturbations
        3.2.1  Fire
        3.2.2  Primary Successions
   3.3  Man-Made Perturbations
        3.3.1  Surface Mining
        3.3.2  Post-Agricultural Successions
               3.3.2.1  Southeastern States
               3.3.2.2  Northeastern States
               3.3.2.3  Central States
   3.4  Cleanup Procedures, Effects on Ecosystems
               (Treatment 0-1)  Natural Rehabilitation
               (Treatment 1-1)
               (Treatment 1-2)
               (Treatment 1-3)
               (Treatment 1-4)
               (Treatment 1-5)
               (Treatment 1-6)
               (Treatment 1-7)
               (Treatment 1-8)
               (Treatment 2-1)
               (Treatment 2-2)
               (Treatment 2-3)

               (Treatment 2-4)

               (Treatment 2-5)

               (Treatment 2-6)

               (Treatment 2-7)

               (Treatment 2-8)

               (Treatment 3-1)
               (Treatment 3-2)
               (Treatment 3-3)
               (Treatment 4-1)
               (Treatment 4-2)
Chemical Stabilization
Clear Cutting Vegetation
Stumping and Grubbing
Scraping and Grading
Shallow Plowing
Deep Plowing
Soil Cover Less than 25 cm
Soil Cover 25 to 100 cm
Remove Plow Layer
Remove Shallow Root Zone
Remove Scraping and Grading,
Mechanically Stabilize
Remove Plow Layer, Mechanically
Stabilize
Remove Shallow Root Zone, Mechanically
Stabilize
Remove Scraping and Grading,
Chemically Stabilize
Remove Plow Layer, Chemically
Stabilize
Remove Shallow Root Zone, Chemically
Stabilize
Barriers to Exclude People
Exclude Large Animals
Exclude Large and Small Aninals
Asphalt Hard Surface Stabilization
Concrete Hard Surface Stabilization
3-1
3-2
3-2
3-2
3-2
3-3
3-3
3-3
3-4
3-4
3-7
3-7
3-9
3-9
3-9
3-12
3-12
3-14
3-15
3-17
3-17
3-17
3-17
3-18
3-19
3-19
3-19
3-20
3-20
3-21
3-22

3-22

3-23

3-23

3-24

3-24

3-24
3-25
3-25
3-25
3-26
3-26
                                    VTM

-------
                            CONTENTS (continued)


               (Treatment 5-0)  Application of Sewage Sludge              3-26
               (Treatment 6-1)  High Pressure Washing (<3 mm)             3-27
               (Treatment 6-2)  Flooding to 30 cm                         3-27
               (Treatment 7-0)  Soil Amendments Added                     3-27
   3.5  Recovery Following Cleanup                                        3-27
        3.5.1  Irreversible Changes                                       3-27
        3.5.2  Rates of Recovery                                          3-27
        3.5.3  Succession Stages Following Cleanup                        3-30
               3.5.3.1  First Year                                        3-30
               3.5.3.2  Fifth Year                                        3-31
               3.5.3.3  Tenth Year                                        3-31
               3.5.3.4  Fiftieth Year                                     3-34
               3.5.3.5  100 Years After Cleanup                           3-34
               3.5.3.6  Climax                                            3-35
   3.6  Quantitative Assessment of Cleanup Impacts                        3-35
        3.6.1  Impact Assessment                                          3-36
        3.6.2  Recovery Assessment                                        3-38
   3.7  Conclusions                                                       3-39
   3.8  Deciduous Forest References                                       3-41
CHAPTER 4-CONIFEROUS FORESTS                                             4-1
   4.1  Overview                                                          4-1
        4.1.1  Boreal Formation                                           4-4
        4.1.2  Rocky Mountain Forest Complex                              4-5
               4.1.2.1  Subalpine Spruce-Fir Climax                       4-8
               4.1.2.2  Douglas Fir Climax                                4-8
               4.1.2.3  Ponderosa Pine  Climax                             4-9
               4.1.2.4  Pinon-Juniper Climax                              4-10
        4.1.3  Sierra Nevada Forest Complex                               4-11
               4.1.3.1  Western Slope                                     4-11
               4.1.3.2  Eastern Slope                                     4-12
        4.1.4  Pacific Conifer Forest                                     4-13
        4.1.5  Coniferous Forests Occurring as Sub and Postclimaxes        4-14
   4.2  Natural  Perturbations                                             4-16
        4.2.1  Fire                                                       4-16
               4.2.1.1  Unassisted Recovery Sequence                      4-16
               4.2.1.2  Assisted Recovery Sequence                        4-17
        4.2.2  Other Natural  Perturbations                                4-19
   4.3  Man-Made Perturbations                                            4-19
        4.3.1  Clearcutting                                               4-20
        4.3.2  Strip Mining                                               4-20
        4.3.3  Controlled Burning                                         4-21
   4.4  Effects  of Cleanup Procedures on Coniferous Forests                4-22
               (Treatment 0-1)  Natural  Rehabilitation                    4-22
               (Treatment 1-1)  Chemical Stabilization                    4-23
               (Treatment 1-2)  Clearcutting Vegetation                    4-23
               (Treatment 1-3)  Stumping and Grubbing                     4-28
               (Treatment 1-4)  Scraping and Grading                      4-29
               (Treatment 1-5)  Shallow  Plowing                           4-29
               (Treatment 1-6)  Deep Plowing                              4-30
                                     ix

-------
                            CONTENTS  (continued)


               (Treatment  1-7)  Soil  Cover  Less than  25  cm                4-30
               (Treatment  1-8)  Soil  Cover  25  to  100  cm                   4-31
               (Treatment  2-1)  Remove  Plow Layer                        4-32
               (Treatment  2-2)  Remove  Shallow Root Zone                 4-32
               (Treatment  2-3)  Remove  Scraping and Grading,
                               Mechanically Stabilize                    4-32
               (Treatment  2-4)  Remove  Plow Layer, Mechanically
                               Stabilize                                 4-34
               (Treatment  2-5)  Remove  Shallow Root Zone, Mechanically
                               Stabilize                                 4-34
               (Treatment  2-6)  Remove  Scraping and Grading,
                               Chemically  Stabilize                      4-34
               (Treatment  2-7)  Remove  Plow Layer, Chemically
                               Stabilize                                 4-34
               (Treatment  2-8)  Remove  Shallow Root Zone, Chemically
                               Stabilize                                 4-34
               (Treatment  3-1)  Barriers to Exclude People                4-34
               (Treatment  3-2)  Exclude Large  Animals                    4-34
               (Treatment  3-3)  Exclude Large  and Small  Animals           4-36
               (Treatment  4-1)  Asphalt Hard Surface  Stabilization        4-36
               (Treatment  4-2)  Concrete Hard  Surface Stabilization       4-36
               (Treatment  5-0)  Application of Sewage Sludge              4-37
   4.5  Recovery After Cleanup                                            4-38
        4.5.1  Irreversible Changes                                       4-38
        4.5.2  Rates  of Recovery                                          4-39
               4.5.2.1  First Year                                       4-45
               4.5.2.2  Fifth Year                                       4-48
               4.5.2.3  Fiftieth  Year                                    4-49
               4.5.2.4  100 Years After Treatment                        4-50
   4.6  Quantitative  Assessment of Cleanup  Impacts                        4-52
   4.7  Conclusions                                                      4-56
   4.8  Coniferous Forest  References                                      4-58
CHAPTER 5-AEOLIAN MOUNTAIN PEAKS                                         5-1
   5.1  Overview                                                          5-1
        5.1.1  The High Mountain  Peak Environment                         5-1
        5.1.2  The Aeolian Life Zone                                       5-5
               5.1.2.1  The Nival Phase of  the Aeolian Zone               5-6
               5.1.2.2  The Aquatic Phase of the  Aeolian Zone             5-7
               5.1.2.3  The Terrestrial Phase  of  the  Aeolian  Zone         5-7
   5.2  Natural Perturbations                                             5-13
        5.2.1  Erosion and Mudflow Phenomena                              5-14
        5.2.2  Avalanches                                                  5-15
        5.2.3  Landslides                                                  5-17
        5.2.4  Overgrazing                                                5-17
        5.2.5  Recovery                                                   5-17
   5.3  Man-Made Perturbations                                            5-18
        5.3.1  Regrading and Replanting                                   5-18
        5.3.2  Alternative Techniques                                     5-19
   5.4  Effects of Cleanup Procedures on Aeolian  Mountain Peaks            5-20

-------
                     CONTENTS  (continued)
 5.4.1   Proposed  Treatment  Techniques
        (Treatment  0-1)   Natural  Rehabilitation
        (Treatment  1-1)   Chemical Stabilization
        (Treatment  1-2)   Clear  Cutting Vegetation  ("Removal")
        (Treatment  1-3)   Stumping and Grubbing
                         Scraping and Grading
                         Shallow  Plowing
                         Deep Plowing
                         Soil Cover Less than 25 cm
                         Soil Cover 25 to 100 cm
                         Remove Plow Layer
                         Remove Shallow Root Zone
                         Remove by Scraping and Grading,
                         Mechanically Stabilize
                         Remove Plow Layer, Mechanically
                         Stabilize
                         Remove Shallow Root Zone, Mechanically
                         Stabilize
                         Remove by Scraping and Grading,
                         Chemically Stabilize
                         Remove Plow Layer, Chemically
                         Stabilize
                         Remove Shallow Root Zone, Chemically
                         Stabilize
                         Barriers  to Exclude People
                         Exclude Large Animals
                         Exclude Large and Small  Animals
                        Asphalt Hard Surface Stabilization
                         Concrete  Hard Surface Stabilization
                        Application of Sewage Sludge
                        High Pressure Washing (<3 mm)
                         Flooding  to 30 cm
                    -,   Soil Amendments  Added
5.4.2  Alternative  Treatment Techniques
       (Treatment 8-1)   Snowfences  and Wind Barriers
                        Watershed Control  Devices  Constructed
                        Snow and  Ice  Additives
                        Removal of  Contaminated  Snow and Ice
             (Treatment 1-4)
             (Treatment 1-5)
             (Treatment 1-6)
             (Treatment 1-7)
             (Treatment 1-8)
             (Treatment 2-1)
             (Treatment 2-2)
             (Treatment 2-3)

             (Treatment 2-4)

             (Treatment 2-5)

             (Treatment 2-6)

             (Treatment 2-7)

             (Treatment 2-8)
               i
             (Treatment 3-1)
             (Treatment 3-2)
             (Treatment 3-3)
             (Treatment 4-1)
             (Treatment 4-2)
             (Treatment 5-0)
             (Treatment 6-1)
             (Treatment  6-2)
             (Treatment  7-0)
            (Treatment 8-2)
            (Treatment 8-3)
            (Treatment 8-4)
5.5  Recovery After Cleanup
     5.5.1  Irreversible Changes
     5.5.2  Rates of Recovery
     5.5.3  Successional Stages
            5.5.3.1  First Year
            5.5.3.2  Fifth Year
            5.5.3.3  Tenth Year
5.6  Quantitative Assessment of Cleanup Impacts
5.7  Conclusions
5.8  Aeolian Mountain Peak References
 5-20
 5-22
 5-22
 5-22
 5-22
 5-22
 5-23
 5-23
 5-23
 5-24
 5-24
 5-24

 5-24

 5-24

 5-25

 5-25

 5-25

 5-25
 5-25
 5-26
 5-26
 5-26
 5-26
 5-26
 5-26
 5-27
 5-27
 5-27
 5-27
 5-27
 5-28
 5-28
 5-29
 5-29
 5-29
 5-30
 5-30
 5-30
 5-30
 5-31
5-34
5-36
                            XI

-------
                            CONTENTS (continued)
CHAPTER 6-TUNDRA
   6.1  Overview
   6.2
   6.3
   6.4
       Geographical  Distribution
6.1.2  Environment
6.1.3  Vegetation
       6.1.3.1  Arctic Tundra
       6.1.3.2  Alpine Tundra
6.1.4  Animal Life
6.1.5  Soils
Natural Perturbations
6.2.1  Fire
6.2.2  Drought
6.2.3  Grazing
Man-Made Perturbations
Effects of Cleanup Procedures on Tundra
       (Treatment 0-1)  Natural Rehabilitation
               (Treatment 1-1)
               (Treatment 1-2)
               (Treatment 1-3)
               (Treatment 1-4)
               (Treatment 1-5)
               (Treatment 1-6)
               (Treatment 1-7)
               (Treatment 1-8)
               (Treatment 2-1)
               (Treatment 2-2)
               (Treatment 2-3)

               (Treatment 2-4)

               (Treatment 2-5)

               (Treatment 2-6)

               (Treatment 2-7)

               (Treatment 2-8)
    6.5
        (Treatment 3-1)
        (Treatment 3-2)
        (Treatment 3-3)
        (Treatment 4-1)
        (Treatment 4-2)
        (Treatment 5-0)
        (Treatment 6-1)
        (Treatment 6-2)
        (Treatment 7-0)
 Recovery After Cleanup
 6.5.1  Irreversible Changes
Chemical  Stabilization
Clearcutting Vegetation
Stumping and Grubbing
Scraping and Grading
Shallow Plowing
Deep Plowing
Soil Cover Less than 25 cm
Soil Cover to 25 to 100
Remove Plow Layer
Remove Shallow Root Zone
Remove Scraping and Grading,
Mechanically Stabilize
Remove Plow Layer, Mechanically
Stabilize
Remove Shallow Root Zone, Mechanically
Stabilize
Remove Scraping and Grading,
Chemically Stabilize
Remove Plow Layer, Chemically
Stabilize
Remove Shallow Root Zone, Chemically
Stabilize
Barriers to Exclude People
Exclude  Large Animals
Exclude  Large and  Small Animals
Asphalt  Hard Surface  Stabilization
Concrete Hard Surface Stabilization
Application of Sewage Sludge
High  Pressure Washing
Flooding to 30 cm
Soil  Amendments Added
6-1
6-1
6-2
6-2
6-3
6-4
6-10
6-11
6-11
6-16
6-16
6-17
6-17
6-17
6-19
6-20
6-20
6-21
6-21
6-22
6-23
6-23
6-23
6-23
6-23
6-23

6-24

6-24

6-24

6-24

6-24

6-24
6-24
6-24
6-25
6-25
6-25
 6-25
 6-25
 6-26
 6-26
 6-26
 6-26
                                      XI 1

-------
                            CONTENTS (continued)


        6.5.2  Rates of Recovery                                          6-26
        6.5.3  Succession Stages Following Cleanup                        6-28
               6.5.3.1  First Year                                        c oo
               6.5.3.2  Fifth Year                                        Too
               6.5.3.3  Tenth Year                                        6-28
               6.5.3.4  Climax                                            6-29
   6.6  Quantitative Assessment of Cleanup Impacts                        6~29
   6.7  Conclusions                                                       6 2g
   6.8  Tundra References                                                 6~3?

CHAPTER 7-COASTAL INTER-TIDAL MARSHES                                     7 i
   7.1  Overview                                                          7~,
        7.1.1  Occurrence of Coastal  Marshes                               7~c
        7.1.2  Related Land Types                                          75
   7.2  Natural  Perturbations                                              76
   7.3  Man-Made Perturbations                                            7^7
        7.3.1  Impact on  Baseline  Ecosystem Components                     7_7
        7.3.2  Unassisted Recovery  Sequence                                7_8
   7.4  Effects  of Cleanup Procedures on  Coastal Marshes                   7^5
               (Treatment 1-1)   Chemical  Stabilization                     7-16
               (Treatment 1-2)   Clear Cutting  Vegetation                   7_ie
               (Treatment  1-3)   Stumping  and Grubbing                      7_16
               (Treatment  1-4)   Scraping  and Grading                       7-19
               (Treatment  1-5)   Shallow Plowing                            7_19
               (Treatment  1-6)   Deep Plowing                               7_2Q
               (Treatment  1-7)   Soil Cover Less than 25 cm                 7-20
               (Treatment  1-8)   Soil Cover 25  to 100 cm                    7.20
               (Treatment  2-1)   Remove Plow Layer                          7_21
               (Treatment  2-2)   Remove Shallow Root Zone                   7^1
               (Treatment  2-3)   Remove by Scraping and Grading,
                               Mechanically Stabilize                     7-21
              (Treatment  2-4)  Remove Plow Layer, Mechanically
                               Stabilize                                  7_22
              (Treatment 2-5)  Remove Shallow Root Zone, Mechanically
                               Stabilize                                 7_22
              (Treatment 2-6)  Remove by Scraping and Grading,
                               Chemically Stabilize                      7_22
              (Treatment 2-7)  Remove Plow Layer,  Chemically
                               Stabilize                                 7_22
              (Treatment 2-8)  Remove Shallow  Root Zone, Chemically
                               Stabilize                                 7_22
              (Treatment 3-1)   Barriers  to Exclude People                7-22
              (Treatment 3-2)   Barriers  to Exclude Large Animals          7-23
              (Treatment 3-3)   Barriers  to Exclude Large and
                               Small  Animals                              7 23
              (Treatment  4-1)   Asphalt Hard Surface Stabilization         7-25
               Treatment  4-2)   Concrete  Hard Surface Stabilization        7-25
              (Treatment  5-0)   Application of  Sewage Sludge               7-25
              (Treatment  6-1)   High Pressure Washing  (<3  mm)              7-26
                                  xi

-------
                            CONTENTS  (continued)
               (Treatment 6-2)   Flooding  to 30 cm
               (Treatment 7-0)   Soil  Amendments Added
   7.5  Recovery After Cleanup
   7.6  Quantitative Assessment of Cleanup Impacts
   7.7  Conclusions
   7.8  Coastal  Inter-Tidal  Marsh  References

                        PART II - MANAGED ECOSYSTEMS

CHAPTER 8-AGRICULTURE
   8.1  Overview
   8.2  Natural  Perturbations
   8.3  Man-Made Perturbations
        8.3.1  Runoff and Fertilization
        8.3.2  Soil  Compaction
        8.3.3  Compaction and Runoff
        8.3.4  Infiltration and Crops
   8.4  Effects of Cleanup Procedures on  Agriculture
               (Treatment 0-1)
               (Treatment 1-1)
               (Treatment 1-2)
               (Treatment 1-3)
               (Treatment 1-4)
               (Treatment 1-5)
               (Treatment 1-6)
               (Treatment 1-7)
               (Treatment 1-8)
               (Treatment 2-1)
               (Treatment 2-2)
               (Treatment 2-3)

               (Treatment 2-4)

               (Treatment 2-5)

               (Treatment 2-6)

               (Treatment 2-7)

               (Treatment 2-8)

               (Treatment 3-1)
               (Treatment 3-2)
               (Treatment 3-3)

               (Treatment 4-1)
               (Treatment 4-2)
               (Treatment 5-0)
               (Treatment 6-1)
    cm
Natural Rehabilitation
Chemical Stabilization
Clearcutting Vegetation
Stumping and Grubbing
Scraping and Grading
Shallow Plowing
Deep Plowing
Soil Cover Less than 25
Soil Cover 25 to 100 cm
Remove Plow Layer
Remove Shallow Root Zone
Remove Scraping and Grading,
Mechanically Stabilize
Remove Plow Layer, Mechanically
Stabilize
Remove Shallow Root Zone, Mechanically
Stabilize
Remove Scraping and Grading,
Chemically Stabilize
Remove Plow Layer, Chemically
Stabilize
Remove Shallow Root Zone, Chemically
Stabilize
Barriers to Exclude
Barriers to Exclude
Barriers to Exclude
Small  Animals
Asphalt Hard Surface  Stabilization
Concrete Hard Surface  Stabilization
Application of Sewage  Sludge
High  Pressure Washing  (<3 mm)
People
Large Animals
Large and
                     7-26
                     7-26
                     7-26
                     7-26
                     7-33
                     7-34
8-1
8-1
8-5
8-8
8-10
8-11
8-13
8-14
8-14
8-15
8-16
8-17
8-18
8-18
8-19
8-20
8-20
8-20
8-21
8-21

8-22

8-24

8-24

8-24

8-25

8-25
8-25
8-25

8-26
8-26
8-27
8-27
8-28
                                      xiv

-------
                            CONTENTS (continued)


               (Treatment 6-2)  Flooding to 30 cm                         8-28
               (Treatment 7-0)  Soil Amendments Added                     8 28
   8.5  Recovery Following Cleanup                                        8 2g
        8.5.1  Irreversible Changes                                       o"oq
        8.5.2  Rates of Recovery                                          829
   8.6  Quantitative Assessment of Cleanup Impacts                         «%n
   8.7  Conclusions                                                       *~™
   8.8  Agriculture References                                            n"™
                                                                          o-oo
CHAPTER 9 -URBAN/SUBURBAN AREAS                                           q ,
   9.1  Overview                                                          g~
   9.2  Natural  Perturbations                                              9"i
   9.3  Man-Made Perturbations                                            9"4
   9.4  Effects  of Cleanup Procedures on  Urban/Suburban  Areas              g~-5
        9.4.1  Lawn, Plant and Soil  Area  Treatments                        9 3
               (Treatment 0-1)   Natural Rehabilitation                     g_8
               (Treatment 1-1)   Chemical  Stabilization                     9-8
               (Treatment 1-2)   Clear Cutting  Vegetation                   9_8
               (Treatment 1-3)   Stumping  and Grubbing                      9_8
               (Treatment 1-4)   Scraping  and Grading                       9.9
               (Treatment 1-5)   Shallow Plowing                            9.9
               (Treatment 1-6)   Deep Plowing                               9I9
               (Treatment 1-7)   Soil Cover  Less than 25 cm                 9.9
               (Treatment 1-8)   Soil Cover  25  cm to 100 cm                 9.9
               (Treatment 2-1)   Remove Plow Layer                          911
               (Treatment 2-2)   Remove Shallow Root Zone                   gill
               (Treatment  2-3)   Remove Scraping and Grading,
                                Mechanically Stabilize                    9_H
              (Treatment 2-4)   Remove Plow Layer, Mechanically
                                Stabilize                                 g_U
              (Treatment 2-5)   Remove Shallow Root Zone, Mechanically
                               Stabilize                                 9_12
              (Treatment 2-6)  Remove Scraping and Grading,
                               Chemically Stabilize                      9_12
              (Treatment 2-7)  Remove Plow Layer, Chemically
                               Stabilize                                  9_12
              (Treatment 2-8)  Remove Shallow  Root Zone, Chemically
                               Stabilize                                  9_12
              (Treatment 3-1)   Barriers  to Exclude  People                9 i?
              (Treatment 3-2)   Exclude  Large Animals                      9_12
              (Treatment 3-3)   Exclude  Large and  Small Animals            9-12
              (Treatment 4-1)   Asphalt  Hard Surface Stabilization         9-12
              (Treatment 4-2)   Concrete Hard Surface  Stabilization        9-13
              (Treatmentv5-0)   Application  of  Sewage  Sludge               9.13
              (Treatment 6-1)   High  Pressure Washing                      9 13
              (Treatment 6-2)   Flooding to  30  cm                          9_13
              (Treatment 7-0)   Soil  Amendments Added                      9 13
       9.4.2   Impervious, Artificial Surface Treatments                   9-13
              (Treatment 8-1)   Washing with High  Pressure Water  (>3 mm)   9-14
              (Treatment 8-2)   Vacuuming                                  9_15
                                   xv

-------
                            CONTENTS  (continued)


               (Treatment  8-3)   Sweeping                                   -J-16
               (Treatment  8-4)   Mechanized  Street  Flushing                 9-16
               (Treatment  8-5)   Surface Removal  Techniques                 9-17
               (Treatment  8-6)   Other Removal  Methods                      9-17
               (Treatment  8-7)   Containment                               9-17
   9.5   Recovery  After Cleanup                                             ^-18
   9.6   Quantitative  Assessment of Cleanup  Impacts                        »-i°
   9.7   Conclusions                                                         ~^\
   9.8   Urban/Suburban Land Areas  References                               y-^

                            PART  III -WILDLIFE

CHAPTER 10 -WILDLIFE                                                      IQ"1
   10.1 Overview                                                         j"'*
   10.2 Effects on Birds                                                 ^"°
        10.2.1 Short-Term Effects                                         f^-B
        10.2.2 Long-Term Effects                                         ^J-1U
               10.2.2.1 Deserts                                          JJJ-lj
               10.2.2.2 Prairies                                         }n\ti
               10.2.2.3 Deciduous  Forests                                in ic
               10.2.2.4 Coniferous Forests                               10-15
               10.2.2.5 Aeolian Mountain Peaks                           10-16
               10.2.2.6 Tundra Areas                                     10-1°
               10.2.2.7 Coastal Inter-Tidal Marshlands                   10-17
        10.2.3 Land Uses                                                 }°-}°
               10.2.3.1 Agricultural Areas                               1   10
               10.2.3.2 Suburban  Areas                                   10-19
   10.3 Effects on Mammals                                               1°-20
        10.3.1 Short-Term  Effects                                        »*-*»
               10.3.1.1 Large  Carnivores                                 ijj-^i
               10.3.1.2 Medium-Sized Carnivores                          10-21
               10.3.1.3 Small  Carnivores                                 10-^
               10.3.1.4 Large  Herbivores                                  JO-22
               10.3.1.5 Small  Herbivores                                  10-22
               10.3.1.6 Large  Omnivores                                   10-23
               10.3.1.7 Small  Omnivores                                   10-25
               10.3.1.8  Insectivores                                     jO-25
               10.3.1.9  Flying Mammals                                    10-Z5
         10.3.2 Long-Term  Effects                                          JO-^b
                10.3.2.1  Large  Carnivores                                  10-^o
                10.3.2.2 Medium-Sized Carnivores                          10-26
                10.3.2.3  Small  Carnivores                                  10-27
                10.3.2.4  Large  Herbivores                                     07
                10.3.2.5  Small  Herbivores                                  10-27
                10.3.2.6 Large  Omnivores                                   10-28
                10.3.2.7 Small  Omnivores                                   10-28
                10.3.2.8 Insectivores                                        on
                10.3.2.9 Flying Mammals                                    10-29
         10.3.3 Land Types                                                {0-^9
                10.3.3.1 Deserts                                          1Q-J1
                                      xvi

-------
                             CONTENTS (continued)


                10.3.3.2 Prairies                                         10-31
                10.3.3.3 Deciduous Forests                                10-32
                10.3.3.4 Coniferous Forests                               10-32
                10.3.3.5 Tundra Areas                                     10-32
                10.3.3.6 Coastal  Inter-Tidal  Marshlands                    10-33
         10.3.4 Land Uses
                10.3.4.1 Agricultural Areas                               10-33
                10.3.4.2 Suburban Areas                                    10-34
    10.4  Conclusions                                                      in 35
    10.5  Wildlife  References                                               jolag

                             PART IV  - APPENDICES
 APPENDIX A-STABILIZERS                                                   A-1
    A.I   Introduction                                                       A_j
    A.2   Types  of  Stabilizers                                               AIj
         A.2.1   Chemical                                                    A i
         A.2.2   Mechanical                                                  A"i
         A.2.3   Physical                                                    £3
         A.2.4   Chemical  with Mechanical Characteristics                    A-3
    A.3   Stabilizer  Groupings                                               A_4
    A.4   Ratings of  Stabilizer  Groupings                                    /\-5
         A.4.1   Application                                                 A  c
         A.4.2   Durability                                                  JJIg
         A.4.3   Vegetation Recovery                                         A_5
    A.5   Land Types  and  Land Use  Classes                                    A-7
    A.6   Appendix A  References                                              A_7

APPENDIX B-IMPACT  ASSESSMENT                                              B.J
    B.I   Introduction                                                       B_l
    B.2   Derivation  of Recovery Index from Successional Sere                B-l
    B.3   Ranking of  Disturbance from  Cleanup Treatments                     B-4
    B.4   Recovery Index Number Assignment                                   B-5
    B.5  Appendix B  References                                              B-ll
APPENDIX C-LAND TYPES                                                     ^

APPENDIX D-CLEANUP TREATMENTS                                             D.!

APPENDIX E-GLOSSARY
                                    xvi i

-------
                                  FIGURES
Number                                                                    Page
 1-1    The North American Desert and its  subdivisions.                    1-9
 1-2    Great Basin Desert roadway unused  for 15 years.                    1-22
 1-3    Mojave Desert roadway unused for 17 years.                         1-22
 1-4    Great Basin Desert 11 years after  blading and forming windrows.    1-24
 1-5    Mojave Desert.   An unintended water catchment basin  in a
        disturbed area.                                                    1-24
 1-6    Mojave Desert.   Larrea growing in  compacted area  alongside
        asphalt roadway.                                                  1-25
 1-7    Mojave Desert.   The effect of increased moisture  from roadside
        cut on Larrea growth.                                             1-28
 1-8    Great Basin Desert.  A 2  ha rectangular plot 11 years after
        blading.                                                           1-34
 1-9    Great Basin Desert 11 years after  blading.                         1-34
 1-10   Great Basin Desert.  Demarcation line between land bladed
        11  years  ago (right)  and  non-bladed land (left).                   1-35
 1-11   Great Basin Desert.  Area in foreground and right was bladed
        11  years  prior  to the photograph.                                  1-35
 1-12   Great Basin Desert 17 years after  application of  water to
        reduce dust followed  by blading.                                  1-36
 1-13   Great Basin Desert recovery 17 years after blading.                1-36
 1-14   Great Basin Desert 17 years after  plowing.                         1-38
 1-15   Great Basin Desert.  Foreground first treated with hot,
        rapid-cure road  oil stabilizer, then scraped.                     1-45
 1-16   Soil  17 years after application of a hot rapid-cure  road oil
        stabilizer.                                                       1-45
 1-17   Sequence  of ecologic  recovery following cleanup.                   1-47
 1-18   Recovery  sequence following shallow plowing and clearcutting
        vegetation.                                                       1-48
 1-19   Sequence  of ecologic  recovery following cleanup.                   1-49
 1-20   Sequence  of ecologic  recovery following cleanup.                   1-50
 2-1    Natural  recovery and  recovery with reseeding of prairie.           2-23
                                    xviii

-------
                             FIGURES (continued)

Number                                                                    page
 2-2    Time course of change after mowing (1-2)  and stumping  and
        grubbing (1-3) of woody component of prairie.                      2-24
 2-3    Recovery of prairie following mechanical  soil  stabilization
        and response to the erection of fencing.                           2-25
 3-1    Natural  recovery (0-1)  and generalized typical  reclamation
        of deciduous forest following cleanup.                            3-28
 3-2    Recovery of deciduous forest following clear cutting  (1-2)
        and stumping and grubbing  (1-3).                                   3-32
 3-3    Recovery of deciduous forest following hard  surface
        stabilization.                                                    3.33
 4-1    Natural  forest vegetation  of the  United States.                    4-3
 4-2    Vegetation  chart for the Santa Catalina Mountains, southeastern
        Arizona.                                                           4.7
 4-3    Sequence  of ecological  recovery following  cleanup.                 4-42
 4-4    Sequence  of ecological  recovery following  cleanup.                 4-43
 4-5    Sequence  of ecological  recovery following  cleanup.                 4-44
 4-6    Decrease  in runoff following clearcutting  coniferous forests.      4-46
 4-7    Clearcutting the whole  stand,  with  reproduction secured by
        seed disseminated from  seed  trees  located  outside the  cut
        stand.                                                             4_48
 5-1    Alpine and  aeolian regions of the world.                           5-3
 5-2    Alpine-aeolian zonation in the  eastern  Himalaya                    5-4
 5-3    Diurnal  sun/shade temperature  ranges  in air  and subsurface
        soil  at  15,500 feet.                                               5.9
 6-1    A word model  illustrating  the  sequence  of  events in the
        recovery  of damaged  arctic tundra vegetation.                      6-27
 7-1    Diagram of  the wetland  classification  hierarchy for the
        estuarine ecological  system  to  the order level.                    7-4
 7-2    A typical effects  linkage  shows the disturbance web and
        environmental  impacts of soil  removal.                             7-11
 7-3    Recovery  of coastal  inter-tidal marsh  following chemical
        soil  stabilization and  application of shallow soil  cover           7-27
 7-4    Recovery  of coastal  inter-tidal marsh  following vegetation
        removal and shallow  layer  soil  removal.                            7-28
 7-5    Response  of coastal  inter-tidal marsh to the erection of
        fencing.                                                           7-29
                                    xix

-------
                             FIGURES (continued)


Number                                                                    Page
 8-1    Indigenous vegetation zones  of the conterminous  United States.     8-3

 8-2    Surface soil  moisture-wind velocity soil  loss  factor as a
        percent of soil  loss at Garden City, Kansas.                       8-6
 8-3    Rainfall erosivity index, based on maximum 30-minute intensity,
        from an average  of annual maximums.                            .   8-7
 8-4    Idealized effective precipitation, vegetation, and sediment
        yield on west to east transect.                                   8-8
 B-l    Example of times for recovery from cleanup in  coniferous
        forest by natural and managed rehabilitation.                      B-3
                                      xx

-------
                                   TABLES

Number
 1-1    Estimates of the years to recover productivity after application
        of the various cleanup procedures.                                 1-54
 2-1    Predicted recovery index in prairie for cleanup treatments.        2-32
 3-1    Parameter trends during deciduous forest succession.              3-5
 3-2    Predicted recovery index in deciduous  forest for cleanup
        treatments.                                                        3.37
 4-1    Measurements on forest clearings.                                 4-25
 4-2    Average possible depth of compaction to 10  percent  soil  voids.     4-33
 4-3    Major forest types of the United States.                          4-40
 4-4    Approximate  age of representative coniferous forests where
        number of trees per acre is reduced by suppression.                4-49
 4-5    Height of dominants in a fully  stocked stand at 100  years -
        good to medium site fully stocked.                                 4-51
 4-6    Approximate  basal  area (m2/ha)  of representative conifer
        species—good to medium site,  fully stocked.                      4-51
 4-7    Recovery index in  coniferous  forest for various cleanup
        treatments.                                                        4.54
 5-1    Movement of  mountain detritus within measured mudflows as
        examples of  erosion.                                               5-16
 5-2    Estimates of the years to reach  recovery after various cleanup
        treatments.                                                        5.33
 6-1    Comparison of arctic and alpine  environments and vegetation.       6-12
 6-2    Representative poorly drained tundra soil.                         6-14
 6-3    Representative well  drained tundra  soil.                           6-15
 6-4    Change in conditions  caused by seismic  lines and winter  roads
        with  paired  control  plots  in  1970.                                 6-18
 7-1    Net  primary  production and plant biomass  for major ecosystems
        and  for the  earth's  surface.                                       7-2
 7-2    Effects of ditching a Delaware tidewater marsh  on the aquatic
        invertebrate  populations.                                          7-9
 7-3    Effects of denuding and/or soil  removal  on  physical  and
        chemical  characteristics  of marshlands.                            7-13
                                    xxi

-------
                              TABLES (continued)

Number                                                                    Page
 7-4    Effects of soil removal  and placement of dredge spoil  in
        marshlands.                                                       7-15
 7-5    Predicted years to reestablish precleanup marshes.                 7-31
 8-1    Relationship between annual precipitation, species  composition,
        and moisture consumption, Rocky Mountains to Missouri  transect
        of prairie.                                                       8-4
 8-2    Effectiveness of ground cover in soil  erosion loss  at  construc-
        tion sites in deciduous  forest zones.                              8-9
 8-3    Control systems and relative effectiveness.                        8-10
 8-4    Commercial crop species  in managed ecosystems.                     8-12
 8-5    Rooting depths of agricultural  crops  in  fertile,  deep, well
        drained soils.                                                    8-12
 8-6    Predicted recovery index of irrigated desert agriculture  for
        cleanup treatments.                                               8-34
 8-7    Predicted recovery index of agricultural  lands  in grasslands
        for cleanup treatments.                                            8-35
 8-8    Predicted recovery index of agricultural  lands  in forests
        for cleanup treatments.                                            8-36
 9-1    Relative rating of environmental  impact  of treatments  on  a
        scale of 0 (no impact)  to 100 (greatest  impact).                   9-19
10-1    Examples of bird species impacted by  maximum disturbance  to
        specified types and sizes of land areas.                          10-13
10-2    Examples of small  herbivorous mammals  impacted  by maximum
        disturbance to specified types  and sizes  of  land  areas.           10-24
10-3    Examples of mammals other than  small  herbivores  impacted  by
        maximum disturbances to  specified types  and  sizes of land
        areas.                                                            10-30
10-4    Relative impacts of land cleanup  treatments  on  wildlife
        following maximum possible land disturbance.                      10-36
10-5    Relative time for recovery of wildlife following  maximum
        disturbance to land and  cleanup treatment.                        10-38
 A-l    Ratings of individual stabilizers for  use in  different land
        types.                                                             A-2
 A-2    Manufacturer and application information  for  soil stabilizers.     A-8
 B-l    Hypothetical  forest sere.                                         B-2
 B-2    Hypothetical  grassland sere.                                      B-2
 B-3    Evaluation of the  effects of cleanup  procedures.                   B-6
                                   xxn

-------
                             TABLES (continued)
Number
                                                                          Page
 B-4    Years estimated for hypothetical  ecosystem to reach climax
        after the indicated cleanup.                                       B-7

 B-5    Hypothetical  coniferous  sere  as  represented by the development
        of a Spruce-hemlock forest on mineral  soil  exposed by glacial
        retreat.                                                           B-9

 B-6    Recovery  indexes  for cleanup  of  hypothetical  coniferous
        forest.                                                            B-10
                                   xxi 11'

-------
                               ACKNOWLEDGMENTS


     This report was written as a collaborative effort among the Chapter authors
and project participants.  The principal authors are listed alphabetically, in
reverse order.  They are:

          Dr. John F. Thames — University of Arizona, Tucson, Arizona
          Richard H. Rowland — GE—TEMPO, Santa Barbara, California
          Dr. James C. McBrayer — University of Michigan (now at Oak Ridge,
                                  Tenn.)
          Dr. Howard A. Hawthorne — GE—TEMPO,  Santa Barbara, California
          James C. Johnson — Moss § Johnson Associates, Springfield, Mo.
          Kenneth E. Gould — GE—TEMPO, Santa Barbara, California
          Dr. Lome G. Everett — GE—TEMPO, Santa Barbara,  California
          F. Stephen Dobson - GE -TEMPO, Santa Barbara, California (now
                              at University of Michigan)

     Project participants provided the support essential for acquiring reference
materials, supporting documents and ran down those elusive  half-documented
citations so well remembered — almost.  These supporters included Dr. Tika Verma,
Gerald Harwood, Dr. Guenton Slawson, Rosalie Oren, Gregory  Mohr, Susan Neighbors,
and Evelyn McDonald.
                                    xxiv

-------
                                  INTRODUCTION
STATEMENT OF THE PROBLEM
     In the industrialized development and commerce of today's technological
society a variety of goods are transported by land, sea, and air.  Among
these goods in transit at any particular instant are some which are toxic or
injurious to man coming into direct or indirect contact with them.  Indirect
contact includes drinking contaminated groundwater, consumption of plants
that absorbed contamination from soil, and using dairy products or meat from
contaminated animals.

     Contaminating accidents may also originate when industrial processes
malfunction.  These include explosions at chemical process plants, inadvertent
releases by plant operators, and a variety of other dumps or leaks.

     The frequency of contaminating accidents has increased yearly until they
have become an almost daily event.  In most cases, the contaminant is spread
by natural forces and human traffic during evacuation of the residents.  By
the time the contaminant is identified and the zone of contamination delineated,
the contaminant may have penetrated environmental materials and decontamina-
tion may be required.  The methods of decontamination are numerous and varied.

PURPOSE OF THE REPORT
     This report is concerned with the ecological impacts of specific cleanup
treatments on the land where they are carried out.  The report provides
guidance to rational selection among the cleanup procedures likely to be sug-
gested by State, Federal or local government officials, industrial representa-
tives of the entity responsible for the contaminating accident, concerned
citizens'  groups, or local environmental managers called in for their expertise.
The priorities of these groups are most likely to be divergent and emotionally
intense under the stress of immediate, local, present danger.

-------
     This report is intended to present the best unbiased, neutral estimate
of how local environments will be changed by specific common cleanup treat-
ments, the relative magnitude of the changes they may cause, and their dura-
tion.

     Local conditions have the greatest force in determining how interrelated
environmental actions occur.  Therefore, the cleanup treatment to be used at a
given site must be decided upon with input from local geologists, foresters,
soil scientists, agronomists, botanists, dairymen, animal husbandmen, zoologists,
or other professionals intimately acquainted with the contaminated area.

SCOPE OF THE REPORT
     This report provides two classes of generic data that may be helpful in
deciding what steps should be taken when contamination occurs.

     A chapter is devoted to each of seven land types and two managed eco-
systems in the United States.  Generic descriptions are given in Part I for
the seven land types and in Part II for the managed ecosystems of agricultural
lands and urban/suburban areas.  The vegetation of the seven land types is
described in detail with brief descriptions of their climate and upper soil
profile.   The history of ecosystem recovery from natural and man-made
catastrophes is reviewed as a guide to predicting how specific cleanup treat-
ments will affect vegetative recovery.

     The second class of information is the generic data arising from descrip-
tions of the effects each of 24 specified cleanup treatments has upon the
soils and vegetation in the affected ecosystems.  This information includes
a recovery scenario, quantitative assessment of cleanup impacts, and citing
the best and worst treatments for each land type.

     The purpose of the descriptions is to provide a general framework for
coordinating information from which to predict both the short-term effects of
cleanup and resultant long-term changes.  Different plants succeed in different
land types but all the plants lack a means of avoiding cleanup trauma.  What-
ever cleanup treatment is used will rearrange some of the soils and damage
most of the plants in a heavily contaminated area.  Land types and cleanup

-------
  treatments were indicated by the EPA to be the items of major interest, and
  specific biota of lower level interest.  Thus, defining the ways in which the
  soil-plant web subsequently develops over time after injury is the main thrust
  in the quantitative assessments.

       After several  attempts to  describe faunal displacement and recoloniza-
  tion after cleanup  in separate  ecosystems  it  became  obvious that excessive
  redundancy lay along  that  approach,  in  that animal families are adaptable to
  several  ecosystems.   To  escape  such  redundancy, the  major  faunal groups  are
  described  and  predictions  made  of  their response  in  each disturbed  ecosystem,
  and  each managed  land type, where  they  occur  naturally.  Part  III presents
  material on two divisions  of  animal  wildlife,  the birds and the  mammals.
  Animal families cross  land  types in  the  same  context that  cleanup treatments
  cross land types in Parts  I and  II of the report.  Evaluation  of the ecologi-
  cal  impact of  cleanup  treatments was confined  to mammals and birds.  Ecological
  impact is defined as the reaction of a population to disturbance of their
 habitat by cleanup.   The reactions considered were those large enough to be
 measured quantitatively under field conditions.  The  population attributes
 considered suitable  for monitoring to define  changes  in the population were
 size (in numbers of individuals) and/or  alterations  in the  numbers of species
 present in the  area  of concern.   These attributes  of  populations were taken
 as  representative  of habitat stability for  evaluating cleanup effects.

      The  critical  component in population recovery of wildlife  is the restora-
 tion  of vegetative species  suitable for  the wildlife.   The  impact assessment
 tables  presented in  the chapters are  based upon best  estimates  for vegetative
 recovery  times, or productivity  changes.

      Almost any spot in the  United  States could be involved  in  a  contaminant
 spill from some source.  The seven  natural land types chosen represent most
 of the United States.   The  land  area  of  the United States, including Alaska
 and Hawaii, totals approximately  10,900,000 km2 of which approximately 240,000
 W are measured as inland waters.  The area not covered by water (approximately
 10,660 km2) can be measured as approximately 1,068,770,000 hectares, excluding
public lands, and can be subdivided into  the land types shown below.  The

-------
percentage breakdown of the seven natural ecosystems and the two managed eco-
systems follows (note that the land-use class of ecosystems overlaps the
natural ecosystems):
  Chapter      Land Type              (percent)               hectares

1
2
3
4
5
6
7

8
9
NATURAL ECOSYSTEMS
Desert
Prairie
Deciduous Forest
Coniferous forest
Mountain
Tundra
Coastal Inter-tidal
LAND-USE CLASSES
Agriculture
Urban/ suburban

0.4
49.8
28.9
7.4
8.7
4.4
0.3

45.4
7.7

4,168,000
532,567,000
309,302,000
78,875,000
93,090,000
47,133,000
3,634,000

485,328,000
82,509,000
     The areas examined for cleanup range  from a hectare  (the  size  of  a
country home) to approximately 4  sections  of  land, which  is not  large  for  a
typical dryland grain farm but is approximately the mean  area  of a  county.
                                         2          2
In metric terms, these areas are  0.01 km  and 10 km  , respectively.  Inter-
                     2            2
mediate areas 0.10 km  and 1.0 km are  also addressed.

     The time scales for which cleanup  impacts are defined range from  short
term  (a few years) to those times when  the cleanup is no  longer  visible
 (hundreds of years).  An arbitrary cutoff  point at 100  years was imposed in
determining whether a treatment was acceptable.  If more  than  a  century is
required for recovery, the treatment is considered unacceptable.

-------
 LITERATURE BASE
      More than 400 publications are cited in the report and approximately
 3,500 others were retained for inspection and discarded.   Personal contacts
 with other professionals in the areas addressed were made freely.   The pre-
 ferred data base is included in this report.

 STUDY APPROACH
      The assembled study team was  comprised  of professional  ecologists actively
 engaged in field work in the land  areas  to which they were assigned for this
 project.   All were familiar with environmental impact assessment methodology
 and the professional  literature in their fields.

      The ecologist constructs  a mental model  of the  ecosystem  climate,  geology,
 soils,  vegetation,  and animal  forms  and  predicts  the ecological impact of
 cleanup operations.   Published observations are reviewed  to  develop an outline
 of  the  ecological  boundaries of the  ecosystem developed on the land type  being
 studied.   When published literature  was  inadequate or absent, personal
 experience was used to predict the impact  of  cleanup treatments on  the  eco-
 system  being  described.   The impact  predictions  in Sections  4 and 5  of each
 chapter represent  a blend of the published data and  personal experience.  The
 predicted  onset  of recovery appears  in Section  6  of  each  chapter and presents
 the quantitative assessment for the  impact forecasts  in Section 4.

      Contamination cleanup has been postulated in this report to occur only
 in land type  tracts that  support recognized and developed  ecosystems where
 recovery following cleanup progresses along secondary succession,  rather than
 ecosystem development by primary succession (new substrate occupied, such as
 earthquake uplift scarps).

      In considering the sequence in which cleanup treatments should be addressed,
 it became obvious that the intensity of environmental impact of a  specified
 cleanup treatment varied from land  type to land type; e.g., surface grading a
shallow soil layer on grassland was less  disruptive than cutting the same
depth and width through^ forest.  Arbitrarily, a uniform sequence  and presenta-
tion of cleanup treatments was maintained throughout  the report for the sake

-------
of continuity.  Those persons interested in following a specific treatment
through all land types are advantaged by this methodology, hopefully with
little disadvantage to those persons intent on examining ecologic impacts
within particular land types.

     Cleanup treatment impacts were ranked across all land types according to
the gross environmental disturbance they would generate.  The disturbance was
translated into "median years-to-recover," presented in relative times to
recovery (recovery indexes), and tabulated in Section 6 of each chapter.
Details of these derivations are given in Appendix B.

     The endpoint of recovery from cleanup for the prairie, deciduous forest,
and coniferous forest is ecological:  the reestablishment (succession) of the
original species of plants and animals at the pre-contamination productivity
(biomass).  Six of the chapters deal with ecosystems which do not have identi-
fiable succession and a different concept of endpoint is used in those chapters.
The four chapters of desert, mountain, tundra, and marshes use "years-to-recovery"
of their pre-contamination productivity (biomass) as the terminal point of
rehabilitation.  In the managed ecosystems of Chapters 8 and 9, the recovery
endpoints are determined by economic productivity and public acceptance, since
physical recovery can be expedited via use of machinery.  The concept of
ecological rehabilitation is inappropriate in these managed ecosystems because
they lack the attributes of succession.

IMPACT EVALUATION
     The equipment projected to be used for cleanup was limited to that operated
by small contractors or to farm machinery.  The technical evaluation steps, are
shown in Appendix B as a sequence of steps.  Where data were sparse for an eco-
system, the author integrated his field experience with published data to
construct the tables prepared for individual chapters.

     Idealized trend curves were developed for Section 5 in many ecosystem
chapters to illustrate the time course of recovery to full productivity for
some of the descriptive ecosystem parameters.  The ordinates are without units
either because they have no dimensions (e.g., homeostasis) or they are broad
generalizations over an array of heterogeneous sites.  Representative types

-------
  of  cleanup  treatments were  selected for the  illustrations to  indicate the
  variety of  parameters common to particular ecosystems.

  USE OF THE  REPORT
      The report is subdivided into 3 Parts and 5 Appendices.  In Part I seven
 natural ecosystems and some of their natural derivations are defined.  Part II
 presents managed ecosystems which are imposed on natural ecosystems and are no
 longer bound by the initial native ecosystem balances.  Part III deals with
 avian and mammalian wildlife displaced by cleanup.  Appendix A summarizes
 judgments of stabilizer chemicals and their application to various land types.
 The methodology used to derive recovery indices for those ecosystems having
 easily recognizable successions is presented in Appendix B.   Appendices  C,
 D, and E,  respectively,  give definitions  of the land types studied, cleanup
 treatments  considered,  and a glossary  of  terms.

      Each  chapter  in Parts I and  II  is  divided into seven sections.  The  first
 three sections  give ecologic data sources  used in evaluating the cleanup  impacts
 In Sections  2 and  3 the  ecological basis  is  given for the judgments utilized
 in Sections  4,  5,  6,  and 7.

      Each cleanup  treatment  is  defined  in  Appendix  D and the impact of each is
 defined in Section  4.  Section  5  gives  a brief scenario  of how the  ecosystem
 is envisioned to change  between cleanup and the time at  which  no  further
 changes are  measurable within present detection instrument capabilities.
 Quantitative assessments of  the magnitude  of  cleanup impact  are made, recovery
 time predicted, and the  effects of using the  cleanup treatment on larger areas
 evaluated.   In Section 7,  the cleanup treatments that are best suited to the
 ecosystem are named, as are  those which are inappropriate.

     Agricultural uses of ecosystems modify the basic characteristics beyond
 the capabilities of native forces to change them on  a short-term basis.  The
modified ecosystems are evaluated in Chapter 8 separately from their parent
ecosystems.   Urban/suburban landscapes  also bear few recognizable imprints of
their ecosystem origins and are described  in Chapter 9 as a unit  of managed
ecosystems.

-------
     For nonecologists who desire an accurate but direct presentation of the
essence of ecology, the paperback book, "Communities and Ecosystems," by R. H.
Whittaker, is recommended.  In general, the ecosystem is viewed as following
a definite sequence of recovery with time that is predictable, called succes-
sion.  As the cleaned up ecosystem enters into natural recovery, the initial
small, short-lived invader species is gradually replaced by long-lived and
larger species which tend to exert more and more control over their surround-
ings.  An example is the high-himidity understory of park-like dimensions
which has developed under beech-sugar maple in the deciduous forest zones.
The collective steps in recovery which follow one another are called the sere
and its use in this report is illustrated in Appendix B.  Formal classical
ecologists will doubtless object to this abbreviated presentation on the grounds
that  it is "applied ecology."

     This report was written to summarize what is known and published in tech-
nical journals, books, and reports about how cleanup affects the environment.
Unfortunately, the published data fall far, far short of being continuous in
both time and in the consecutive details of recovery.  References are cited
where information is derived from the literature; if no reference is cited,
the information is the author's own projection of the recovery after cleanup
is completed.

     The ecosystems in Part I are described in terms of recovery from cleanup
without additional interference or assistance by man.  The kinds of assistance
that are appropriate to perform on three of the ecosystems are collected into
the Agriculture chapter in Part II because these are the land types that produce
the commercial foods and fiber at the greatest rates.*  The reclamation pro-
cesses that are appropriate after plowing—irrigation, reseeding, transplant-
ing, fertilizing, contour tilling, and harvesting—are widespread management
practices of agriculture.  These are presented in Chapter 8 only to a limited
degree since this report deals with cleanup impacts—not with impact erasure.
Discussion of these management practices within ecosystem chapters has been
restricted to special cases that are unique to natural systems.
*These three land types are desert (Chapter 1), prairie (Chapter 2),  and
 deciduous forest (Chapter 3).

-------
 CONCLUSIONS
      Cursory inspection of the conclusions  sections  of  the  chapters  shows
 that  generalization  and condensation  to produce  a  single  conclusion  for  the
 report  is  not effective.   The  most disruptive  technique was  concluded  to be
 paving  over a compacted subgrade.  However, estimates of  the time for  this to
 be  weathered and  assimilated differ among ecosystems.   The general conclusion
 about ecosystems  was that  succession  was relatively unchanged by a variety of
 techniques  that assisted in early and rapid revegetation.  Mulches,  stabilizers,
 plowing, and fertilizers were  very effective in  reintroducing new growth but
 the particular ecosystem controlled which species would succeed into climax.
 Plantings of hardwoods  in  grasslands  or coniferous ecosystems would have no
 effect on the time to climax or eventual climax  species, although grasses are
most  likely  to provide  early cover in any of the widespread ecosystems.

     This report has addressed the impacts of particular cleanup techniques on
specified ecosystems but other areas of study remain.  The effectiveness of
cleanup techniques in removing or stabilizing contamination needs to be addressed
The greatest need is the development of a methodology to quantify (into monetary
terms) the information collected on environmental impact.

-------

-------
              PART I
        NATURAL ECOSYSTEMS

CHAPTER 1,  DESERTS
CHAPTER 2,  PRAIRIES
CHAPTER 3,  DECIDUOUS FORESTS
CHAPTER 4,  CONIFEROUS FORESTS
CHAPTER 5,  AEOLIAN MOUNTAIN PEAKS
CHAPTER 6,  TUNDRAS
CHAPTER 7,  COASTAL INTER-TIDAL MARSHES

-------

-------
                                 CHAPTER  1
                                   DESERT


 1.1   OVERVIEW
      There is  little accord among scientists as to what constitutes a desert.
 It is generally agreed that sparse to absent vegetation due to little rainfall
 is the distinguishing feature, but where  some scientists limit the definition
 of desert to areas receiving less than 50 mm of rainfall annually, others
 include regions with 360 mm of annual rain, or even more.  In addition, it is
 recognized that the rainfall amount alone is not the primary determinant of des-
 ert type, but  that seasonal moisture fluctuation, temperature, and edaphic fac-
 tors  are also  important.  For example, the polar regions are true deserts because
 they  lack vegetation and exhibit low levels of atmospheric water vapor.  There
 are also regions of the world which receive in excess of 300 mm of annual
 precipitation  but are dry, barren and relatively devoid of vegetation.  Such
 regions are termed edaphic deserts because the physical cause of aridity is
 in the nature  of the porous surface soil which allows moisture penetration
 at a  rate too  rapid to permit plant sustenance regardless of the amount of
 rain  that may  fall.
      The focus in this discussion is on the desert of North America, specifi-
 cally the desert of the United States.  This desert is a large irregular belt
 on the western side of the continent covering about 940,000 square kilometers
 from mid-Idaho to Northern Mexico and roughly corresponding to the rain
 shadowed intermontane region between the coast ranges and the Rocky Mountains.
 It includes major parts of the states of Arizona,  Nevada,  Utah,  and lesser
parts of California,  Colorado,  Idaho, New Mexico,  Oregon,  Texas,  and Wyoming.
     The physical geography and human utilization of the region  exhibit great
 internal variety, but its overall aridity associated with an excess of evapora-
tion over precipitation,  high percentage of sunshine,  great temperature extremes,
high diurnal  air temperature,  low atmospheric humidity,  frequent  winds and a
predominance  of eroded landscapes give it an unquestioned  unity.   In general,
                                    1-1

-------
 the region is characterized by rainfalls less than 250 mm per year.  In these
 areas the environmental gradients are altitudinal gradients, where temperature
 decreases and precipitation increases with elevation.  Desert communities occupy
 the lowest,  driest areas and juniper and pinon-juniper woodlands are on the
 higher more  mesic regions.   The eastern and northern boundaries merge into
 grassland prairie.  An intermediate land type,  the desert grassland, will also
 be briefly discussed.
      In this report,  areas  receiving 250 mm or  less of rain per year will be
 classed as desert.  The peripheral  desert grasslands receiving somewhat greater
 annual moisture will  be called arid areas.   This is not an arbitrary classifica-
 tion but rather one based on the requirement for unassisted revegetation.
 Moisture is  the dominant factor in  vegetation growth and it is generally con-
 ceded that once relatively  large areas  of vegetation have been destroyed,  about
 250 mm of rainfall per year is the  minimum needed for revegetation without
 supplemental  irrigation.59'51'6°
      In the  discussion below are considered the  climatic,  edaphic  and topographic
 features which  explain a desert,  and  the  vegetative factors which  characterize
 it.   The goal of this  summary  is to  form  a  basis  for  understanding the  effects
 produced by  natural and anthropogenic (man-made)  environmental  perturbations.

 1.1.1   Desert Climate
      Geographers  have  attempted  to classify desert  types  by climates, usually
 in  terms of precipitation and  temperature and their seasonal variation.  The
 basis  for  the most widely used classification scheme  is the  Kb'ppen  system  which
 is  based on distribution of vegetation.42   Vegetation  is  sensitive  to tempera-
 ture and precipitation  and tends to reflect  these variables  in a general way.
 Because  aridity, the distinguishing feature  of desert areas, is a measure  of
 dryness, an index based on water balance of  an area would be superior to the
 Kb'ppen system for a climatic analysis.  Meigs49 building on the indices developed
 by Thornthwate   which were  closely reflective of water balance considerations,
 formulated a classification  scheme which includes a more explicit appreciation
 of the role of evaporation,  storage of moisture in the soil and the seasonal
variability of precipitation.  The Meigs classification uses season of rain-
fall, maximum temperature of the hottest month and minimum temperature of the
coldest month to differentiate arid areas.
                                    1-2

-------
     Based on Meigs' considerations of climate, the American Desert can be
divided into four regional districts.*  The northernmost of these, the Great
Basin, has cool or cold winters, frequent snowfall and annual precipitation
ranging from 100 to 220 mm of which only about 40 percent falls in the summer
months.  In the southern part are the hot deserts:  the Mojave, with winter
rains; Sonoran, with summer and winter rains; and Chihuahuan, with summer rains.
The Chichuahuan is somewhat isolated but the other three lie in a continuous
series.
     In general, the temperature of the North American desert increases toward
the south while total precipitation decreases.  In any given latitude, precipi-
tation increases both east and west of about the 115  meridian.  The frequency
and amount of precipitation vary directly as a function of elevation with ele-
vation accounting for about 80 percent of the variance in rainfall statistics.
The southwest corner is the hottest, driest desert area,
     A curve of precipitation amount plotted against time of year for desert
areas is characteristically bimodal, with peaks in the winter and summer.  The
relative dominance Of one peak or the other has profound influence on the ecol-
ogy of the areas.   Orographic barriers are largely responsible for affecting
the seasonal distribution of rainfall in the cool Great Basin desert.   This area
lies in the zone of westerly cyclonic storms but much of the moisture in these
storms is extracted by the western boundary mountains.  These storms bring much
of the annual precipitation during the winter and early spring months.  In the
hot deserts to the south, two major climatic influences cause different sea-
sonal patterns of rainfall.  In the eastern part, the Chihuahuan and to a lesser
extent the eastern Sonoran, the major rain is during the summer.  This summer
rain results from a seasonal shift in global weather patterns which cause moist
air from the Gulf of Mexico to be drawn over the area.  Convective precipitation
originates in this moist air mass.  The western extent of this effect is about
the Arizona-California border.  To the west of this line, the Mojave receives
winter rain from frontal systems of the cyclonic storms from the west or north.
Only rarely does summer rain,  usually from moist air from the Gulf of California,
fall in this area.   ThevMojave is the driest of the areas,  with some locations
reporting only 50 mm of rain a year.  As a rule, the lower  annual average rain-
falls are associated with higher variance from the average.
                                                                  78
*The nomenclature for the desert areas is based on work by Shreve.
                                     1-3

-------
      It is important to recognize that the frequency and intensity of rain are
 important ecological variables.     A summer rain totaling 150 mm in a given
 locality might be the result  of  one or two rains during the season, perhaps with
 a duration of two to three hours for each rain.   Beatley  has shown that  in
 Mojave desert ecosystems,  heavy  (>25 mm)  rainfall triggers phenological events.
 There is a predictable relationship between early winter heavy rain and successful
 vegetative and reproductive growth of shrubs the next  spring.   Productivity of
 desert plants is  directly  related to the  distribution  and numbers  of desert
 animals.    However,  in deserts,  where small differences in the amount of  moisture
 are  of great  significance,  the numerical  relations among species are not  con-
 sistent from  year to year  or  necessarily  correlated with seed  supply.89
      The climatic division  of the American deserts into hot  and cool temperature
 regimes and into  summer, winter  and intermediate precipitation patterns is
 reflected in  the  classification  and existing distribution of desert vegetation,
 but  the recovery  potential  of an area will  be an interaction between available
 moisture  and  local soil conditions.

 1.1.2  Soils

      An ecosystem typically consists  of a  complex  mosaic  of  overlapping patterns
 that  reflect,  within the limitation  of time  and  local materials, not  only
 present but past  patterns of  distribution  of biota.  A  soil  profile  adjacent
 to a  Mesquite  (Prosopis juliflora)  or to a  creosote bush  (Larrea tridentata)
 may differ markedly from a  profile  2 meters away on barren ground.  However,
 the major  soil type of almost all desert areas is  similar.  The basic charac-
 teristics are  coarse texture;  an accumulation of carbonates within a  few feet
 of the  surface contributing to the formation of a hardpan or a caliche layer;
 and low organic matter content.
     The United States Comprehensive Soil Classification System83'84'85 classi-
 fies the soil order underlying almost all of the American deserts as Aridisols.
Soils of other orders, especially the Entisols,  are also present but less
extensive in deserts.  The Aridisols at the northern edge of the Great Basin
Desert are replaced by Mollisols.
     Shantz   has noted that regions defined as  arid by geographers cover  an
estimated 43 percent of the world's land area; however if defined on the basis
of climate, arid regions occupy only 36 percent  of the earth's surface,
                                    1-4

-------
 35 percent  if defined on the basis  of vegetation,  but  Aridisols  cover only
 about  25  percent  of the  surface.  Thus,  worldwide  regions  of soils  classified
 in the Aridisol order do not conform to  all  the  parameters of either  climatic
 or vegetation zones.
     The  Aridisols,  formed under  low rainfall, are  leached little and are  there-
 fore high in  calcium,  magnesium,  and other soluble  elements.   Desert  soils
 seem particularly subject to the  formation of  impervious surface layers—either
 by physical processes  or through  the surface development of algae and lichens —
 so that the effective  drought  is  accentuated by  excessive  runoff.   The Aridisols
 are usually dry and are  never  moist  as long as 90 consecutive  days  during  a
 period when temperature  is suitable  for  plant  growth.  The low rainfall also
 limits plant  growth so that  the soils are low  in organic matter and have low
 carbon/nitrogen ratios.  The soils are young in profile development and show
 little evidence of  leached upper  horizons.  The combined A and B horizons
 are frequently less than 30  cm thick.  As a rule the A horizons are light
 colored whereas B horizons are similar or slightly darker  in color.   The C
 horizons  are  light-colored and usually calcareous.  All horizons are  neutral
 or  mildly alkaline.
     Entisols are of such recent  development that no horizons  have  formed.  They
 are most  extensive on  steep  actively eroding slopes or alluvial fans  where the
 rate of sedimentation  is high.
     Mollisols contain a thick A  horizon high in humus and a base of  calcium or
magnesium.  In the northern  Great Basin Desert they are formed under vegetation
 consisting chiefly of grasses.  The climatic regime is moist in winter and dry
more than 60 consecutive days  in  summer.
     Much of the diversity in  soil profiles reflects their origin.   Coarser
fragments are found in soils developed nearer the base of mountains and hills,
and finer textures are found toward the middle valleys.  Clay, if present,  is
usually found in the B horizons in more level areas.  Carbonate content of
soils  is usually closely correlated with the origin of parent material, and is
highest in areas immediately adjacent to  limestone mountain ranges.   High
carbonate concentrations are almost  always correlated with the presence of a
restrictive  hardpan.  Low carbonate  soils are most often developed  on alluvium
of mixed igneous and sedimentary rocks.
                                     1-5

-------
      The particle size distribution of the soils also reflects formation of
 alluvial deposits.   Variation in texture is largely due to differences in
 distributions of sand and silt particle size fractions.  At low elevations
 there is characteristically a low clay content,  but increased clay concentra-
 tions are found on  alluvial sediments in closed  drainage basins.
      The climatic regime in which Aridisols form is characterized as one in
 which precipitation is less than evapotranspiration during most of the year and
 no water percolates through the soils.   This has two important consequences.
 First,  horizons of  soluble  salts underlying calcium carbonate and gypsum are
 normally found at the average depth of water penetration.   This depth will be
 important in  relation to the consequences  of a given depth of soil removal.
 Second,  since rainfall is seasonal,  there  will be periods  when precipitation
 is in excess  of evapotranspiration and  water will be stored in the soil  to be
 used  in  the spring  or fall  growing season.   Thus,  the time of year in which a
 postulated cleanup  takes place  relative to  the seasonal  characteristics  of
 rain  in  that  area is  important  for predicting ecosystem response.
      There is ample evidence of accelerated soil-forming processes under desert
 shrubs.   The  highest  concentrations  of  organic carbon,  organic  nitrogen52  and
 available phosphorus    occur in the  upper horizon  under  shrubs.   Salts also
 accumulate under  shrubs,  probably  through litter  decomposition  from leaf fall,
 although  Odum noted that  microbial decomposers would  be  limited by dryness.54
 (Odum speculated  that  rodent  herbivores were  important  in  this  aspect of nutrient
 cycling.)  In  the transition  zone  between the Mojave  and Great  Basin  deserts,
 the average carbon/nitrogen  ratios have been  determined  to  be about  10 while
 in the A horizon under  shrubs,  values between 12 and  15  are common  (the  range
 was 5-30).  Only  in the A horizon  did organic carbon  content exceed  1 percent.
 These low C/N  ratios could be due  to low organic matter  of desert  soils, an
 increase in nitrogen fixation,  or  a combination.66  Nitrogen fixing Azotobacters
 are usually few or lacking on desert soils.47  Other differences in soils under
 shrubs are noted:  soil pH tends to be lower, conductivity of saturation extracts
 higher, nitrates and chlorides accumulate more and exchangeable cations are
        66
 greater.
     Algae and fungi grow in and on the soils contributing to fixation of
atmospheric nitrogen and to  soil stabilization by forming crusts.21  These
                                      1-6

-------
  crusts attain a thickness of 2 to 3 millimeters and are sufficiently tough to
  permit their removal from the soil surface as a film.  They may result in
  considerable water runoff by sealing soil surfaces or they may result in
  improved water-holding capacity.   In addition to stabilizing the surface of
  cohesionless soil, algae can add  considerable quantities of nitrogen to the
  soil through their nitrogen-fixing ability.73  Soil fungi  may be necessary in
  nutrient cycling.   Symbiotic mycorrhizae,  root fungi,  associated with the white
  bursage Ambrosia and the grass  Hilaria  rigidu are  necessary for phosphorus
  absorption in soils with low available  phosphorus.97

  1.1.3  Topographic Factors

       The  effects of altitude on temperature and  precipitation  have  been de-
  termined  in  many desert  localities.  There is  an average decrease of  1°F  for
  each  100  m rise in  elevation  but the rate varies with  the  season of year  and
  with  local conditions.    South slopes at a 90° angle  to the sun receive  1 5
  times as  much heat  as level areas.  Shreve80 listed the principal features of
  altitudinal  climatic change as:
           1.   shortening the frostless season
           2.   lowering the daily curve of temperature throughout
               the frostless season
           3.   increasing the intensity and duration of all  critical
               phases of low temperature  during the  frost  season
         ' 4.   shortening the arid  fore-summer,  a  critical period
           5.   increasing rainfall  and  thus  soil moisture
           6.   decreasing evaporation.
      Beatly4  showed  the  importance  of  local topography  on vegetation distribu-
 tion.  Low temperatures  resulting from downslope  drainage of cold air  into
 closed basins coupled with differences in seasonal precipitation resulted  in
 anomalous  vegetation mosaics at the transition  between  the  Great Basin and
 Mojave deserts.

     A second effect of local topography is its relationship to rainfall   On
 flat or gently sloping surfaces,  rainfall leads to unconcentrated flow over
 large areas, with thin and-extensive sheets of moving water, known as sheetflow
This results in the removal of a  thin layer of soil  from the entire surface
Generally only the smaller soil particles are  removed,  to  be deposited  in low
                                     1-7

-------
lying areas.  Naturally rill and gully erosion occur also, and a distinctive
feature of desert areas is the erosional pattern.

1.1.4  Vegetation
     The climatic divisions of the American desert of Meigs correspond very
                                                       78
well with the four regional deserts described by Shreve   and shown in Figure 1-1
In each of these regions further subdivisions into mosaics of vegetation types
are made by plant association communities which are distinguished by distinctive
combinations of dominant species.  Desert perennial plants grow both singly
and more often in clumps, separated by bare areas of desert soil.  Often a
clump of vegetation contains more than one species of plant; as many as 10
                                                         81
different species have been recorded from a single clump.    The size and
spacing of the clumps are irregular.  Some plants, such as creosote bush (Larrea
tridentata) thrive in several desert areas but other plants are restricted almost
exclusively to a particular desert region.  These indicator plants are:
          Great Basin   —   Sagebrush (Artemisia tridentata)
          Mojave        —   Joshua Tree (Yucca brevifolia)
          Sonoran       —   Saguaro (Cereus giganteus)
          Chihuahuan    —   Agave (Agave lechuguilla)

     1.1.4.1  Great Basin Desert--
     Numerous mountain ranges are scattered over the Great Basin resulting in a
sharp vertical zonation of the plant life culminating in either a xeric or a
mesic forest cover on the range tops.   The actual desert occupies the floors of
the detrital valleys.
     The vegetation is a monotonous open shrubland with almost no cacti or large
plants.   A distinctive feature of the vegetation is the predominance of com-
munities which are very simple in composition or have as much as 95 percent of
                                        82
their stand made up of a single species.     The commonest dominants are shad-
scale (Atriplex confertifolia), sagebrush (Artemisia tridentata), rabbitbrush
(Chrysothamnus spp.), blackbush (Coleogyne ramosissima),  and winter fat
(Eurotia lanata).   Open stands of greasewood (Sarcobatus vermiculatus)  are
found on the saline flats.   Indian rice grass (Oryzopsis hymenoides),  indigo
bush (Dalea spp.),  fourwing saltbush (Atriplex canescens)  and a number  of
                                      1-8

-------
                                        I           ^ «^ ^» «^
                                                  	I	
                                                                 "I

                                                                1	1
         GREAT BASIN DESERT
Illllllllllll  MOJAVE DESERT
                                               SONORAN DESERT
                                              CHIHUAHUAN DESERT
       Figure  1-1.   The North American Desert  and  its subdivisions
                    (after Shreve78).
                                    1-9

-------
herbaceous plants and perennials are also found.  Shrub coverage averages 24
percent  in the various associations, with a range of  15 to 37 percent.
     The precipitation in several localities is higher than in any other part
of the American desert.  It ranges from 100 to 270 mm.  The winter temperatures
are  low, and frequent snowfalls occur.

     1.1.4.2  Mojave Desert—
     This is the smallest unit of the American desert and lies almost wholly in
California, with a small wedge extending into both southern Nevada and north-
western Arizona.  Its elevated margins are delineated by the range of the Joshua
Tree (Yucca brevifolia).  The bajadas are mostly dominated by creosote bush
(Larrea tridentata), but communities of four-wing saltbrush (Atriplex canescens)
and  complexes of Grayia-Lycium, Lycium shockleyii and Ambrosia dumosa are
extensive.  On the northern edge Coleogyne ramosissima and Grayia spinosa are
dominant.  Shrub coverage averages about 16 percent ranging from 7 to 23 per-
     4
cent.   More spring ephemerals are found because of the mild,  moist winters.
     Annual precipitation decreases from 125 mm on the west to less than 50 mm
on the east, and occurs during the late winter.

     1.1.4.3  Sonoran Desert—
     The subtropical Sonoran is the most varied of the American deserts and
exceeds the other three deserts in the number and variety of its life forms.
The desert lies between sea level and about 1,000 m.   Above this upper limit
are  isolated mountain masses that,  due to orographic  factors,  can receive as
much as 700 mm of precipitation per year yet be located within a short distance
of Larrea plains.
     More than half of the Sonoran Desert is dominated by plants with a stature
of less than 1.5m.   Almost pure stands of Larrea,  Ambrosia,  Lycium and Atriplex
are found.   Larrea and Ambrosia dominate the lower  portion of  the desert but
they are joined by many shrub,  tree,  and succulent  subdominants which become
more and more prominent toward higher elevations and  greater  rainfall.  Often
two species,  or sometimes three form the entire vegetation of  an area.  On
plains  and bajadas below 300 m elevation the total  perennial  flora rarely
                   82
exceeds 10 species.     The most prominent cactus is the saguaro (Cereus giganteus)
Barrel  cacti (Echinocactus spp.),  hedgehog cacti (Echinocerrus spp.),  and prickly
                                    1-10

-------
  pears (Opuntia spp.) are common.  Small legume trees such as palo verde (Cercidium
  fl°ridum) and C. microphyllunO and ironwood (Qlneya tesotal are prominent.
  Ephemerals are of two kinds, spring and summer flowering because of the bimodal
  rainfall pattern characteristic of the Sonoran.
                82
       As Shreve   notes,
            "With the prevailing simplicity of composition,  it follows that
            the entrance or exit of a single species,  or a great change in
            the relative abundance  of two or more  species,  may make a profound
            change in the physiognomy of the vegetation."
       Precipitation  ranges from almost  nothing  to  300 mm on  the eastern edge.

       1.1.4.4  Chihuahuan Desert--

       The easternmost of the American Deserts is the  Chihuanhuan Desert  in the
 United States  covering  part of New Mexico  and  Western  Texas  adjacent  to  the
 Rio Grande.  A large island of Chihuahuan  Desert  flora  occurs  along the  San
 Pedro Valley of Arizona, well  away from the main  body of the parent desert.78
      The Chihuahuan is  separated from the main part of  the American desert and
 makes no contact with other desert regions.  To the north and east it grades
 into semiarid praise*  while to the west the Continental Divide separates it
 from the Sonoran Desert.  Larrea,  common to the Mojave, Sonoran and Chihuahuan,
 is not continuous across the grassland that lies  along  the Continental Divide.
      The prevailing  plants are low shrubs,  Larrea, Acacia,  Prosopis or long  leaf
 succulents,  Agave lechuguilla,  and Hechtia.  Several  species of Yucca and in-
 frequent  cacti Ojxmtia  are also found.   Dominance  is  shared  by shrubs, succulents
 and semisucculents.   The vegetation  is  most lush  in the warm season,  and there
 are no winter and early spring  ephemerals as in the other  deserts.
     Precipitation ranges from  70  to  500 mm,  and about  80 percent  falls  between
 the middle of June and  the middle  of  September.  Winter  temperatures are  low,
 with occasional  periods  of 30 to 72 hours of freezing.

     1.1.4.5  Desert Grassland--

     This arid biome is  considered in this chapter because the  southern section
 forms a division betweenvthe Sonoran and Chihuahuan deserts and because some
 areas now classified as desert were once this grassland.  (The northern and
western edges of the  Great Basin Desert merges into the  Palouse prairie98 of
                                     1-11

-------
 western Oregon and Washington.   Recovery in this area is similar to that of the
                                                            11
 short-grass prairie detailed in Chapter 2.)   Shantz  and Zon   believed that
 precipitation greater than 300  mm per year was  required to establish and maintain
 the Aristida-Bouteloua grassland.
      The desert grassland  has undergone significant  changes in recent years.
 In some locations  grasses  have  been  supplanted  by shrubby vegetation, mesquite,
                                 19
 acacias and burroweed.   Clements  believed  that much of what is now desert was
 once grassland before overgrazing  destroyed  the grass,  but more recent studies
 emphasize the interaction  of overgrazing with climatic  change.35
      In the remaining Desert Grassland,  annual  precipitation ranges from 300  to
 400 mm,  with summer moisture slightly exceeding that  of winter.

 1.1.5   Desert Succession
     A  desert ecosystem  is  obviously  not  a static  system,  but usually tends to
 approximate a dynamic  equilibrium  or  to  fluctuate  around  a stationary state.
 An ecological climax is  an  example of that dynamic equilibrium and  the process
 by which the equilibrium is  reached is  termed succession.   In order to understand
 the effects of some  environmental  disruption, it  is clearly important to  distin-
 guish equilibrium  systems from  those  that  have  not yet  even approximated  steady
 state.   Equally important is  to  identify  differences  that  may exist  between
 exceedingly slow changes, which  are hardly distinguishable  from  the  equilibrium
 state, and  changes  that are more or less  rapid.  The  rate  of  change  is  as funda-
 mental for  the  concept of ecological  succession and development  as  is  the question
 of  whether  the  climax equilibrium actually exists.
     The  concepts of climax and  succession in desert  areas  will  be reviewed in
 the next  sections of this chapter by  identifying the  empirical observations of
 changes  that  have resulted from natural and man-caused environmental perturbations,
     A caveat  is required first however.  It is not universally  agreed  that suc-
 cession occurs  in desert environments; if by succession it  is meant that each of
 a sequence of communities alters or modifies the physical environment to be
 favorable for the next member in the sequence.  Shantz   considered succession
 to be important as a basis for recognizing and understanding plant communities.
Other arid land investigators, notably Clements19'20 and Allred and E. Clements1
 stressed plant succession.   However,  the primary interest of these writers was
                                     1-12

-------
  in arid  lands, not deserts.  Shreve79 cautioned about applying the  succession
  concept  to deserts:

           In a consideration of the dynamic aspects of the vegetation of
           the region in which the initial, sequential and final stages of
           a succession are characterized by the same species, and often by
           the same individuals, it is doubtful whether these conceptions,
           formed in regions with a very dissimilar vegetation, are of much
           real utility.
      iu. n  50
      Muller   was in general agreement and found no serai succession in recovery
 of disturbed Chihuahuan Desert.  Muller also noted that often the definition of
 climax and the serai stages leading to it are made to narrow that very little
 of the earth's surface may be said to be in climax condition.
      Some features of recovery and succession after the proposed cleanup pro-
 cedures are qualitatively similar to ecosystem recovery after natural or man-
 induced vegetation changes which have been documented for desert areas.   These
 features are presented in the next sections.

 1.2   NATURAL PERTURBATIONS

 1.2.1   Natural Processes

      Succession on naturally  disturbed desert  habitats  such as washes which
 are periodically  scoured  by floods,  actively  eroded bedrock areas, or flood
 deposition on bajadas,  has been  studied  little.  Wells"  noted that  the  charac-
 teristic  shrubs of these  natural disturbed  areas, Thamnosma,  Hymenoclea, and
 Salazaria, can be established in man-made disturbed areas and play a  role
 similar to that of pioneer successional  plants of more humid  areas.
     Campbell has described natural vegetation succession in  wind-formed hollows
 (blowouts) between mesquite dunes as following five sequential stages.15  The
 sequence  only operates during a succession of favorable,  i.e., wet, years.  (The
major mesquite growth is in the spring when other perennials are practically
dormant.)  The first stage is the Mat Stage.  If moisture has accumulated in
the blowout, species with a prostrate habit such as Tribulus become established.
 If moisture continues, either in the same or the following growing season a
Weed Stage begins, Ambrosia for example.   This stage results in a gradual
decline in sand movement as the root systems develop.   The third stage is the
                                    1-13

-------
Gutierrezia Stage, usually in the following growing season.  At this point all
of the plants of preceding stages are found.  Together they have a generalized
drought-resistant root system.  Wind and sand erosion is reduced and leveling
of the dunes along with sand deposition in the blowout area begins.  If two
or three years of above-average moisture have occurred the Sporobolus Stage
of grasses begins.  This is followed by a Bouteloua climax.  Succession to the
fourth stage requires three to four years of ample rainfall.  No estimate was
given of the time to climax.
     Muller studied succession, or rather a cyclic phenomenon, in a Larrea-
Ambrosia association in the Chihuahuan Desert.    In this area erosion removed
the climax association and all existing soil down to a clay hardpan.  Muller
noted that succession, indeed plant growth, could not occur until the clay
weathered and until a few inches of sand and gravel were washed from adjoining
plains.  The general sequence of recovery noted was shallow rooted grasses,
sparse shrubs tolerant of a slight soil covering over the clay, dense shrubs
as the soil deepened, sparse representations of Larrea, the climax association.
No time for this process was given.  At the time of climax all the species of all
of the stages were present leading Muller to reject the idea of succession in
desert scrub.

1.2.2  Fire
     Fire is known to influence desert plant distribution.  Controversy exists
regarding the role of fire in maintaining desert grasslands, that is, the des-
ert grassland may be an artificial biome (a disclimax) initiated and maintained
by man*,  If this is true,  artificial manipulation of deserts might result in
                            £ O O"7
grassland.   Numerous writers0 '   have hypothesized that grasslands are a product
of repeated fire set by primitive man.  Humphrey^ among others^ has applied
this hypothesis to desert grasslands, largely on the historical evidence that
areas now desert were formerly grasslands,  which through a combination of over-
grazing and fire suppression continued to desert climax.  The theory says grass-
lands were maintained by frequent fires which prevented establishment of seeding
shrubs.  Once the sequence of fire was broken, by overgrazing and fire suppres-
sion, seed bearing shrubs became established and a self-perpetuating, irreversible
vegetation change began.
                                    1-14

-------
      The incidence of fire in the past compared to the present is a critical
 point.   The controversy has been reviewed by Hastings and Turner35 who conclude
 after examining travel journals of 22 early explorers that fires did not occur
 frequently in the desert grassland in the past—although it is noted that fires
 may have been locally important.   They cite other factors which will be dis-
 cussed  later as cause of the decrease in size of desert grasslands.
      There is an extensive literature of the effects of burning on desert vege-
 tation.   Although these studies report intentional man-caused fire rather than
 natural  perturbations,  they will  be discussed here.   The major thrust of these
 studies  was range improvement by  using fire to remove undesirable brush and allow
 grass and palatable  forage species to grow.   (This work shows implicit belief
 in  the fire theory of desert grasslands.)
      The results appear to be contradictory.   Humphrey41 found that  burning prior
 to  the spring rains  killed burroweed (Haplopappus spp.), partially controlled
 mesquite and favored perennial  grasses.   Pickford57  in the Great  Basin desert
 found that  burning destroyed sagebrush and  allowed grasses to increase and  a
 more  recent work  indicates that  forage  species  and  grasses  are  increased
 about 70 percent three  years after burning.   Reynolds61  obtained  conflicting
 results  for a June burning of Great  Basin vegetation.   The density and produc-
 tivity of the area were  reduced and  remained  below control  areas  for  three
 growing  seasons.   The first  year productivity was  down 50  percent  and  the next
 two years down 10  percent.   The treatment did kill 90  percent  of  the  burroweed,
 half  the  cholla, and one-quarter of  the prickly pear.
      Longstreth  and  Patten    have  developed a model which  postulates  a  number
 of interacting factors that  result in  increased water  and  nitrate  loss  as
 brush is  converted into  grassland.  The loss  of soil nitrate would result in
 decreased productivity which  agrees with Reynolds  observations.   It has been
noted that  on  heavily grazed  ranges fire had  little effect on the proportion
of grasses  to  forbs,  but on  lightly grazed ranges  fire significantly reduced
                             •7/_
grasses and  increased forbs.    This also might explain some of the conflicting
results noted of the effectiveness of brush burning on range improvements.
     The effects of fire-are not permanent.   The perennial vegetation of many
desert plants resprouts from the root crown of a pre-fire plant, then grows at
                                    1-15

-------
the rate normal for the plant species.  The time for complete recovery is
unknown; however 40 to 50 years, about 1/2 of the life span of desert plants, may
be a good estimate.  '    Range reclamation experts recommend three to five-
year intervals between burnings to minimize undesirable species, but no indica-
tion is given of the amount of recovery to the pre-burn condition during this
time.  Beat ley  measured vegetation parameters on adjacent burned and unburned
plots in Grayia-Coleogyne and Coleogyne associations in the Mojave-Great Basin
transition.  The plots had been burned 13 and about 22 years prior to her study.
The burned areas were quite distinctive and easily seen in the vegetation mosaic.
The cover of herbaceous perennials in the burned area was only 20 percent that
of unburned plots; however, winter annuals increased by a factor of 2.5 and
grass (Bromus) was increased by a factor of 3 in the burned plots.
     Pechanec recommended burning of sagebrush range in the late summer or early
fall after the seeds of perennial grasses had matured.    The spring was recom-
mended by Semple as the best time to burn, for then the soil is damp and losses
                                              69
of organic matter were not likely to be great.    The principle appears to be
avoidance of detrimental effects on the perennial seed.
     Piemeisel  '   made detailed studies of the early stages of secondary
plant succession after fire in the Great Basin desert.  He notes that the original
sagebrush dominated cover was exceedingly complex.   The community was composed
of perennials with bulbs, tuberous roots, and rhizomes and perennials forming
mats, rosettes and tussocks.  Most growth is made between October and July;  the
time of autumn rains and the beginning of the next  hot summer.   Growth is inter-
rupted by winter cold and terminated by summer heat.   The annuals grow in spring
and summer.  The successional order is first Russian Thistle (Salsola spp.),  then
mustards (Sophia parviflora or Norta altissima), then downy chess (Bromus
tectorum).   These successional changes usually take place within five or six
years.   Grazing by livestock or rodents may result  in holding the vegetation
at one of these lower stages and retarding or preventing the reestablishment of
perennials.
                                    1-16

-------
 1.2.3  C1imate

      Long  term climatic fluctuations which influence vegetation are well docu-
 mented.  For example,  a synchronous change from juniper and pine woodland to
 desert grassland occurred about  8000 years ago in the Mojave,  Sonoran and
 Chihuahuan deserts,  probably associated  with changes in global atmospheric circula
 tion patterns influenced by the  melting  of continental glaciers.92   Shorter term
 and more recent  fluctuations are more pertinent to predicting  desert ecosystem
 response to cleanup  operations since rainfall,  the major variable in climatic
 change,  is also  the  critical determinant  of both desert vegetation  community
 type and speed of recovery.
      Studies  of  tree rings  have  documented 25-  to 60-year periods of moist cool
 climates interspersed  with  dry warm climates  in the southwestern deserts  since
                  9 f~\  7 X
 the 15th century.  '     These studies  show climatic anomalies  to be  of non-
 uniform  geographic extent.   For  example,  in the early to  mid 1600's  the Mojave
 and large  portions of  the Great  Basin  deserts  experienced a climate  warmer
 and drier  than usual while most  of  the Sonoran  and  eastern Great  Basin were
 normal,  and the  Chihuahuan was cooler  and  moister  than  usual.27
      Recognition  of  the  existence of these  climatic  fluctuations  helps to  inter-
 pret  apparent  historical vegetation changes and  their causes.  A  number of
 authors  have noted that  the  change  from desert  grassland  to desert in the  late
 1800's coincided with  a marked increase in man's use of grassland as cattle range,
 and have speculated a  cause  and  effect relationship between overgrazing and the
 decrease in grass cover and  increase in desert  shrub vegetation.14'32  Hastings
 and Turner   documented the  influence of climatic change on this massive shift
 in regional vegetation and concluded that the new ecosystem was a response to
the unique combination of climatic and cultural stress.  Climatic change reduced
 soil moisture to a level below that required to establish new grass.  Over-
grazing contributed to the imbalance between infiltration and runoff through
compaction  of the soil and removal of existing grass.
                                      1-17

-------
 1.3  MAN-MADE  PERTURBATIONS

 1.3.1  Fencing
     Studies of evidence of vegetation change resulting from the exclusion of
 livestock from formerly heavily grazed land are abundant.  Shreve recorded
                                                      77                       81
 changes in Sonoran vegetation 22 years after exclusion   and again at 30 years.
 He noted a general increase in plant population, that is, the number of species
 remained rather constant over the years but the total number of plants increased.
 The increase in the first 22 years was greatly exceeded in the next 8 years.  No
                                              Q
 evidence of succession was found.  Blydenstein  followed Shreve's work at 50
 years.  He compared the original enclosures with nearby control sites.  In both
 the protected and unprotected sites the same species grew and in the same order
 of abundance.  The most notable change was an overall increase in density of
 vegetation in the enclosed areas.  Grasses and shrubs were beginning to replace
 Larrea.
     Gardner,30 reporting after 30 years of protection, found grass density about
 twice as great in the enclosed area.  A more recent study of only 10 years'
        14
 duration   noted no vegetation change.  The authors speculated that soil compac-
 tion by grazing animals and/or erosion might account for the absence of effect.
                  o "7 r\
     Other authors '    believe that it is soil erosion more than any other
 feature that is responsible for the many instances of slow recovery following
 livestock exclusion.   When there has been little or no erosion and the original
 plant species have not been entirely removed,  protection from grazing may lead
 to a rapid change in species composition and an increase in palatable (to cattle)
        34
 species.
     To test the effect of rodents on range quality Brown   established plots
 that were grazed by cattle and rodents, by rodents only, and ungrazed by both
 cattle and rodents.  He found that mesquite and Opuntia increased for all treat-
ments but the poisonous shrub burroweed (Haplopappus)  decreased under total
                  53
 exclusion.   Morris   found that excluding rodents increased the density of
native perennial grasses.   A more recent study on 0.3 ha plots indicates that
 seed-eating rodents and ants are in direct competition for the available food
 supply.  The exclusion of either taxon results in an increase in numbers and
                                    1-18

-------
 biomass of the other than results in a compensatory impact on plant population.
 When both fauna populations are excluded a 2- to 5-fold increase in plant density
 is noted.-^
      It has been speculated that the existence of a fence leads to increased
 animal utilization along the fence line, resulting in soil compaction and vege-
 tation removal.   Hansen33 showed that cattle do not graze more closely and tram-
 ple the vegetation near a fence than they did a hundred feet  or more away.
      Fences can  act as  protective as well as excluding landscape features.   Elk
 in Washington  have been restricted to the Cascade Mountains  in historic times.
 However,  they  recently  moved onto the ERDA Hanford research  site which is fenced
 and not subjected to livestock  grazing.

 1.3.2  Overgrazing
      The  effects  of cattle  grazing on desert grasslands have  been extensively
 studied,  but it  is not  always clear whether the  observed effect,  characteristically
 an increase in density  in unpalatable shrubs and  a decrease in perennial  grass,
 were  due  directly to grazing or  to associated  factors,  e.g.,  soil  compaction by
 the grazing animal.   Most of the  literature deals  with  the role of livestock in
 promoting shrub  invasion  of  grassland.   Hastings  and  Turner35  have summarized
 these  studies and  classified them  to  emphasize  four ways  in which  domestic
 grazing animals have been thought  to  contribute to  the  spread  of  woody  plants.
 The animals may act as  disseminators  of  seeds, may reduce  competition  by
 selectively removing one  species  which inhibits growth  of  another,  might  serve
 as a  fire suppression by  removing  the grass  fuel  from an area,  or  might reduce
 the moisture content of the  upper  layer  of  the  soil by  compaction.
                          18
     Christian and  Slatyer   have  suggested  a sequence  of events which  follows
overgrazing and leads to  irreversible (within human memory) change  in vegeta-
tion.  First, a depletion of perennial grass occurs,  followed  by wind erosion
and a reduction of  surface litter.  The result is  increased rainwater runoff,
decreased water penetration, and higher soil temperatures.  All of these factors
create a more arid microclimate which inhibits re-establishment of the original
vegetation.
     The effect of grazing on lands that  are now desert is less well studied
than the effect on desert grasslands.  Total primary productivity is low in
                                      1-19

-------
 desert areas and the grazing pressure on the most palatable species is high.*
 However  a characteristic of shrubs that persist in arid lands is adaptation to
        48
 stress.    Rapid regrowth is a characteristic that enables shrubs to dominate
 desert areas.  A number of shrubs are known to increase primary productivity
 and produce more new shoot growth when defoliated or closely cropped.100  Over-
 grazing can also lead to a change in species composition and introduction of
 non-native species.   Pechanec56 reported overgrazing of sagebrush (Artemisia
 tridentata) in the Great Basin desert has led to the introduction of cheat-
 grass (Bromus tectorum)  at the expense of native perennial  grasses.   If more
 species are removed  than introduced,  ecosystem stability will  be  lessened.
            48
      McKell   has  noted  that  the  worth of any shrub  species can only be deter-
 mined by assessing  its relation to  the major ecosystem  and  the extent  to which
 user  groups depend on it.   Thus changes  in  species composition are not  necessarily
 detrimental but  must  be  evaluated in  relation to land use.   For example,  the salt
 desert  shrub  community in  the  Great  Basin desert  moves  toward  dominance by  black-
 sage  (Artemisia  nova) when  protected  from grazing, while intense  winter grazing
 promotes  an increase  in  shadscale (Atriplex  confertifolia).38   This  is  not  an
                                 48            ~	
 undesirable change for a rancher.

 1.3.3  Mechanical Disturbance
     The  effects on vegetation  of compaction  of non-agricultural desert  soils by
 vehicles  or animals,  and by land alteration through blading or plowing, are
 generally specified by empirical observation  rather than experimental measure-
 ment.

     Almost any mechanical disturbance of the soil results in compaction since
no soil is free from susceptibility to compaction.  The magnitude of the compac-
tion affects the rate of vegetation  recovery.  Compaction invariably results
in long lasting visible effects.  Jeep and tank tracks from desert training
early in World War II are prominent  in the Mojave, and Indian trails  and wagon
 The definition of "most  palatable"  is  obscure  since relative  shrub abundance
 and local  environment  may modify animal  preferences.   The  phenomenon of "scarcity
 improves palatability" is well  documented  in range  management.48
                                    1-20

-------
 tracks of the last century can still be seen.86  Although compaction slows
 recovery, natural processes eventually mitigate the effect.  For example,
 although the streets of 19th century mining towns are still clearly visible,
 vegetation succession is occurring.
      The primary effects of compaction are a greatly reduced soil pore space and
 increased bulk density which alter the relative amount of soil components present
 in a given volume of soil.   These processes lead to a decrease in water infiltra-
 tion, a decrease in water storage capacity and restricted air movement.  This
 damage to soil structure severely impairs plant growth through decreased soil
 moisture and by limiting root penetration.  A reduction in pore space from 55
 to 34 percent has been shown to eliminate all plant growth.29
      The secondary effects  of compaction are changes in microclimate,  since soil
 bulk density and thermal capacity are increased while diffusivity is almost
 unaffected.   Soil temperature changes of -7°C have  been recorded  in compacted
 vehicle tracks (vegetation  removal  increased temperature by almost  the  same
 amount  indicating a  dynamic  interaction  between the effects of these processes).45
 Changes in microclimate  can  affect  the germination  of seeds.
      Few data are available  on  recovery  of compacted desert  areas.   Wells"
 showed  possible  succession  in compacted  areas.   Gardner30 and  Buffington14
 speculated that  compaction,  and subsequent erosion  were  major  factors in  inhibit-
 ing  regrowth  in  disturbed areas.  Figures  1-2  and 1-3  show  the  persistence  of
 compacted areas  in the Great  Basin  and Mojave  deserts.
      Recovery of  plowed  and  furrowed  desert  land is  somewhat better  documented
                             fir                                               *
 however.  Rickard and Shields  'measured the  early  stages of recovery on  2.5  ha
 plots denuded of vegetation by removing the  surface  7-30  cm of  soil  with  a  bull-
 dozer.  After two years  no shrubs had reappeared.  Annuals were less abundant
 than on adjacent undisturbed areas.  This  was probably related  to the removal
 of seeds with the surface dirt.  Since annual growth is highly  correlated with
 rainfall and  seeds can lay dormant until  climatic conditions are optimum for
 growth, the surface might contain the seeds from the previous three or four years'
production.   Bunchgrass  (Qryzopsis)  was present only when the roots of the pre-
disturbance plant had not been completely removed.   The authors concluded that
mechanical soil and plant removal  were more deleterious than plant destruction
by atomic tests.

-------
                             •K- •* <£?'• »•*£.•' iBJ^C ^L*-" •*-**ia,'T^^
                             5^*^^j£'5Wfc
                             •tfBfcSr ^-*^ At^t*tfc4^
                             •Mk    %<'*'!S3iSr     .Vi«rn^t
                                  v jKiiifct'*
Figure 1-2.  Great Basin Desert  roadway unused for 15 years.  The  effects
             of compaction  are evident.  (Photo courtesy of UCLA.)

     Figure 1-3.  Mojave Desert roadway  unused  for 17 years.  (Photo
                  courtesy of UCLA.)
                                   1-22

-------
       Abandoned farm land in desert areas generally follows the pattern of
  invasion by Russian Thistle (Salsola)  during the first growing season, then
  the replacement of Salsola by grass,  beginning in the second growing season,
  and Anally the gradual ^establishment of the original brush vegetation.59'
  Gardner   shows that  thirty years  of  protection were  not  enough to erase
  the visible effects of plowing.  Since  soil  is known  to be compacted by improper
  tillage,    compaction  is  probably  the effect  noted.   A disc harrow will compact
  soil  at shallow depths (7-10 mm) while  heavy  machinery and trucks  will  drastically
  compact the upper  12-15 mm.  Plowing often compacts soil  at the  plow depth.
       Mechanical  land alteration can also be beneficial  to  desert vegetation
  recovery.   Brown and Everson11 found that primary productivity on  a  ten year
  old furrow  in the Sonoran desert was two and  one half  times  greater  than on the
  intra-furrow areas.  The furrow impedes runoff, increases  infiltration  and
  reduces evaporation.   Figure 1-4 shows this effect.  These  factors combine to
  increase available soil moisture and increase  localized plant growth.  This
  technique of land surface manipulation is used to assist the recovery of mine
  spoil banks where different surface topographies-deep chiseling, offset
  listering,  gouging and basin construction-are routinely used.90  It is also
 employed to contour water catchment basins  to assist  agricultural production
 in desert areas.    Occasionally a  catchment  basin is  inadvertently created.95
 Figure 1-5  shows an unplanned  catchment  basin  and  indicates how plant growth
 is enhanced by  the  slight  extra water.   Figure 1-6 indicates that the increased
 water  from  roadway  runoff  can counteract  the  effects of compaction  in desert
 soils.
1.3.4  Atomic Test Target Areas
     Atomic test and associated construction activities at the Nevada Test Site
have produced massive, and often-repeated, environmental perturbations.  A
typical surface or low altitude atmospheric detonation would destroy, by blast
and thermal radiation, the vegetation in a circular area of 2-3 km2/ An under-
ground detonation could cover 3-4 km2 of desert with dirt ejected from the crater
In addition, large areas would be cleared of vegetation or leveled to accommodate
diagnostic test equipment.
                                    1-23

-------
                     , V<--t- „.»'»•>
                        «"

Figure  1-4.
              ~x' «?7v;V-~ •; / •'i^.*^. V" •'  - •
           Great Basin Desert 11 years .after blading  and forming windrows
           The fence at right encloses an area which  has excluded grazing
           animals for 11 years.  (Photo courtesy of  UCLA.)
 ^&^*&«3s£ ••   v  ~  '*:-,~ •  •
 %C$£tg£^,:' v:.-:i ;:^..v'--r;^v
         Ife2**c'\3ft-,: 'C: • ^i^ "^\
                                                               - ' *"»>
 ^-A- •> "jw-fc.^
 £? -> .^ .-•* iTTu . - i
Figure  1-5.  Mojave  Desert.  An unintended water  catchment basin  in a dis-
            turbed  area.  This basin concentrated  the sparse moisture to
            allow plants growing  within it to become better established
            than in the surrounding area.  (Photo  courtesy of UCLA.)
                                 1-24

-------
       Figure 1-6.
                                           9r°Win?  1n comPact*d area  alongside
                                                              -
                                                        ,s.s
                    from  highway,  is  present.   (Photo  courtesy  of UCLA.)
      Three circular zones of damage have been identified around an atmospheric
 test burst point.    The first is an area about 1 km in diameter where all
 surface vegetation has been eliminated by blast or fire.  The inner 0 2 to 0 5
 km portion has had the seed containing layer of soil removed by blast winds
 The second is an area extending about  1 km beyond the first zone where the
 perennia! brush is broken but grass survives relatively unharmed.   Finally
 there is an area of shrub survival.  This last  area  may respond  to ionizing
 radiation effects in the following  growing season.62'63
      The immediate injury is  attributable to mechanical and thermal  effects-74
 however,  recovery may  be  influenced  by  ionizing  radiation.   Radiation  has  been
 suggested  as  "stimulating- the  growth of  annuals,3 or conversely as  causing
 no  observable  effect.

     The pattern  of revegetation is well  documented3'67'74'75-76 and the early
 stages proceed faster than recovery on bladed areas.76  During the first year
 the inner portion of the barren zone is sparsely populated by annuals, evidently
 from seeds introduced naturally.  In the outer portion of this zone the ground
cover by annuals exceeds the total  cover of nearby undisturbed areas, possibly
                                     1-25

-------
because of the lack of competition with perennials for the available moisture.
As the spring annuals die, the summer maturing Salsola dominates.  Beyond the
perimeter of denudation recovery occurs through the crown sprouting of the
shrub species.  In the second to fourth year the species composition of
annuals changes.  Grasses begin to dominate and the area resembles the grass-
land scars of natural burn areas.  Sal.tbrush (Atriplex canescens), a characteris-
tic species of roadsides, sandy washes and other disturbed areas, becomes
abundant.  The short term changes in vegetation are analogous to the succession
                 99                        74
observed by Wells   in a deserted townsite.    After twenty years, summer and
winter annuals were found to be abundant, but there was virtually no recovery
of the original perennial shrubs.  The few shrubs which persisted were
                         67
apparently crown sprouts.
     Little work has been done to document the vegetation recovery on crater
ejecta.   After nine years, Romney   reported that Salsola and summer and winter
annuals were abundant except where the ejecta was "too thick," and speculated
that the soil moisture infiltration rate on the loose dirt was too great to
support vegetation.

1.3.5  Restoration
     Attempts have been made to transplant seedlings and rooted cuttings of
shrubs to speed the process of revegetation.  '    It was concluded that without
supplemental irrigation transplanting perennial vegetation was a futile effort.
Plant response and climatological behavior indicate that at least two and perhaps
three consecutive years of favorable soil moisture are necessary for sage
                                         96
(Artemisia tridentata) seedling survival.    In one experiment where additional
moisture was supplied to each transplanted plant, small animals destroyed all
unprotected plants within a few weeks.    Construction of topographic features
which concentrate available natural moisture appears to be the most practical
method of speeding recovery.

1.4  EFFECTS OF CLEANUP PROCEDURES ON DESERTS
     A number of cleanup techniques are available which vary in cost and efficiency
This section will list some of these techniques in the order specified by the EPA
and will:   (1)  explain the technique; (2) discuss its effect on the local ecology;
(3) indicate the closest analogy to vegetation succession and recovery from some
natural  or man-induced environmental perturbation;  and (4)  discuss the effect of
increasing treatment area.

                                     1-26

-------
       It is assumed that the methods used in construction industry and in farming
  to protect the surrounding areas from harm, especially from erosion or the de-
  position of eroded material, will be used.   This will generally require the
  construction of silt dams and diversions.

       (1-1)   Chemical  Stabilisers

       The general  purpose of chemical stabilizers is to bond the surface  soil
  particles  together to form a protective  layer  and isolate the  subsoil  from  erosion
  due to  wind  or rain.   The  procedure  often accompanies construction  or  road  build-
  ing activities.   Chemical  stabilizers  could be used to immobilize a  contaminant
  for a short  (one  to two  years)  period  with  a single application, for longer time
  periods  with periodic  application, or  they  could  be used  in conjunction with
  other cleanup  techniques  (e.g., scraping) to increase the effectiveness of  the
  technique by bonding  the contaminant to easily removed clumps,  or to ameliorate
  the erosion effects after the cleanup, especially on  steeply sloping areas.
  Vegetation is normally removed prior to treatment in  construction activities-
  however, most of the chemicals, especially the polymer emulsions or resins  could
  be applied over vegetation if it was necessary to quickly immobilize the contamina-
  •f~ -\ c-\-n
 tion.
      Appendix A lists the chemical stabilizers preferred for desert areas.  They
 include oil and latex polymers, polyacylamides (PPA),  polyvinyl alcohol resin
 adhesives and other resins and polymers.   These preferred treatments generally
 have an effective life of less than one year.   Other  chemical  agents such as salt
 sodium hydroxide,  sodium carbonate,  or  iron chloride,  also cause aggregation of
 soil particles.   The ecological effect  of  these deflocculants  is dependent on the
 soil ion  exchange  capacity and is  highly area  specific,  thus will  not  be  dis-
 cussed here.   Attempts to  stabilize  Pu  contaminated soils at the Nevada Test  Site
 with deflocculants  have  been reported.95
      The  soil  bonding  produced  by  the chemical  stabilizers will  reduce  moisture
 infiltration  and increase  rain  runoff.  The chemical surface will  also  form a
mechanical barrier to  seed germination.   Some types of treatment will result  in
absorption of  solar energy and  an  increase in soil  temperature.  The increase may
be great enough to inhibit seed germination as the  chemical surface begins to
degrade; however,  this is unlikely since desert plants are tolerant to high surface
temperatures.
                                    1-27

-------
     The ecological effects result from changes in available soil moisture.  Water
runoff from the impervious surface will increase vegetation growth on areas
immediately surrounding the treated area, much like vegetation growth is enhanced
next to desert roads (Figure 1-7).  Care must be taken that this runoff does not
erode the soil from adjoining areas.
     If vegetation is not removed prior to treatment, bunchgrass and perennial
growth will continue during the next growing season, either by normal branch
sprouting or, if the treatment kills the above-ground portion of the plant, by
root crown sprouting.  Since moisture will be able to enter the soil at the
junction of the plant trunk and the earth, increased growth should occur since
much of the available moisture will be channeled to the shrub,  and the treat-
ment will eliminate competition from nearby ephemerals while reducing soil
moisture evaporation.  Ephemeral vegetation and annual grasses will not appear
until the surface begins to degrade.   This might be the next growing season or
the year following, depending on the time of year of treatment  application.  In
    Figure  1-7.  Mojave Desert.  The effect of increased moisture from
                roadside cut on Larrea growth.   (Photo courtesy of UCLA.)
                                    1-28

-------
  time, about a year depending on the chemical selected, cracks will develop in
  the surface.  Moisture will concentrate in these cracks and allow germination of
  seeds resulting in localized vigorous growth which will accelerate the breakdown
  of the surface.  Moisture concentration at the cracks may have a deleterious effect
  If heavy rain occurs prior to the breakup of the crust and the establishment of
  annuals,  erosion can begin under the crack area.  The probable results would
  include faster breakup of the crust, gully erosion in the treatment  area possibly
  exposing  hardpan and a change in the drainage pattern.
       Although the effective duration of most  chemical treatments is  about a  year
  clumps of stabilized  soil,  5-20  cm in diameter  will  persist  for  longer periods
  of time.   Recovery of the area,  compared to  surrounding areas, depends on re-
  covery of the  ephemerals  and grasses,  which  in  turn  is  related to  the  breakdown
  of the clumps  of stabilized  material.   This  should occur within  three  to  five
  years  after  treatment.

       If the  vegetation is removed  prior  to treatment, recovery of bunchgrass and
  perennial  shrubs will proceed from the underground portion of the perennial
  vegetation as the  surface degrades.  The area will recover full biological
  productivity within three to five years except where the soil has been compacted
  by the brush removal or chemical  application equipment.  The treatment area will
  be visible until the perennial vegetation achieves full stature-about twenty
  to thirty years.  The compaction  effects of the heavy equipment will  be visible
 for a longer time.

      If the chemical layer is not toxic to animals it will  not  inhibit the
 burrowing  activities of rodents;  moreover,  the burrowing might  bring  to the
 surface the material the chemicals  were intended to bind.   If the area is
 relatively small,  i.e.,  1  ha,  the rodents will probably  not be  displaced but
 will  forage for food outside  the  treatment  zone.   If  the treated  area  is larger
 than  the animals^normal forage range,  about 50  m for pocket  mice to  100 m for
 Kangaroo Rats,  '    a migration to other  areas can be  expected.  The effect
 of  this migration on the surrounding  area would  be difficult, if not impossible
 to  measure.   It  is  generally  hypothesized that desert  rodent  species diversity '
 is  directly related  to precipitation, through the relationship between precipita-
 tion and primary production.12  It  is doubtful whether the surrounding habitat
 could support the additional population.  The weaker members would die off
As vegetation reappears in  the treated areas,  animals  will  return.  Desert rodents
                                    1-29

-------
of different species harvest different sizes of seeds and forage in different
              12
microhabitats.    Since there is little evidence of sequential desert vegeta-
tion succession, there should be no change in animal species diversity as the
treated area is recolonized.  If vegetation is removed however, changes in rodent
diversity might be expected until the area is well along toward full recovery.
This point is discussed later.
     Progressively greater care must be exercised for the larger treatment areas
to handle rain runoff.  If the largest areas are treated on a routine basis over
a number of years, ground water recharge will be inhibited.  The effects must
be estimated for each individual case.

     (1-2)  Clear Cutting Vegetation
     Vegetative cover can intercept part of the deposit of an airborne contaminant,
and removing vegetation might remove a considerable portion of the contaminating
material while also inhibiting transfer and possible concentrations through food
chains.
     A number of methods are available for vegetation removal.  Hand hoeing and
shovel cutting would be applicable for the smallest area, or for steep hillsides;
however, mechanical techniques would be required for larger areas.   These mechanical
techniques include cabling and anchor chaining and are common in range improve-
              31
ment programs.
     In cabling, a 45 to 60 m long 3.75 cable is dragged between two tractors
traveling on parallel courses.  The cable breaks off or uproots brush.  Cabling
can be used where the brush breaks easily and is not willowy.   The technique
results in little effect on perennial bunchgrass and usually leaves the
A soil horizon and humus at the base of the brush intact,  an important considera-
tion for revegetation.
     Anchor chaining consists of dragging a heavy (up to 50 kg per link)  chain
through the vegetation to break or uproot plants.   The general procedure is much
like cabling but the ground is more disturbed,  the underbush humus  is usually
scattered,  and soil stabilizing algae clumps are broken.   Chaining  will clear
brush up to the size of small trees.   The spacing  of the parallel  tractors,
and thus the spacing of the compacted areas,  is dependent upon density of
                                    1-30

-------
 vegetation,  weight  and length of chain,  size of tractor,  bite of tractor tread,
 and ground slope.
      The effects  of hand  removal would be relatively benign.   Most perennial
 vegetation would  resprout in the following growing  season.   The seeds  of annual
 vegetation would  not be disturbed.
      Soil  compaction by the  mechanical equipment  would  be the major deleterious
 effect  of  mechanical removal.   Broken  perennials  would  resprout from underground
 plant parts.   The summer  annual  forb Russian Thistle (Salsola)  would invade the
 area, especially  the chained area,  the following  summer since Salsola  only
 germinates where  the soil  surface has  been scarified.94  Salsola roots  cannot
 penetrate  compacted soil  and would  thus  be restricted to  areas  between  the
 tractor  paths.  Winter  annual  grasses  (Bromus)  and  the  Salsola  might delay re-
 vegetation of  the areas by native species  by their  effective  use of available
 soil  moisture.  Salsola can  thrive  at  below normal  desert rainfall  levels;93
 with  normal rain annual grasses, followed  by native  ephemerals,  would become
 established.   Recovery would be  more rapid  than if  the  seed containing  layer of
 soil  were  removed.
      Erosion by wind would be  a  major  problem until  the annual  vegetation or
 algal mat  was  reestablished, .or  until  natural soil compaction by rainfall occurred
 This problem is largely non-ecological however, since the main  effect of eolian
 erosion  is the removal of  the  A  horizon of  soil, which  virtually does not exist
 in desert  soils.  The major  danger would be  due to winds spreading  the con-
 taminant beyond the treatment  area.   Dust from the area should  produce no lasting
                                                     •-)
 effects on the surrounding vegetation;  however, 10 km   denuded  of vegetation
 could produce a highly visible and undesirable dust cloud.
     Vegetation removal is likely to affect microclimate.   The  low-level wind
 speed profile will be changed  due to changes  in aerodynamic roughness.44  This
change will lead to increased  evaporation, increased soil  temperature and de-
creased humidity.   These will affect both the revegetation of the treatment area
                                      2 "^
and the plant growth outside the area.
     Temporary food  reduction would  be  the major effect  on animals.  This effect
                       V                                                         '
like chemical stabilization,  will increase with increasing treatment area.
                                    1-31

-------
      (1-3)   Stumping  and Grubbing
      This procedure would produce  a drastic  perturbation on the desert  area,
 and would probably only be used in conjunction with another cleanup technique if
 it were desired to restrict the regrowth of  perennial  vegetation.
      Desert  plants have extensive  root  systems.   The ratio  of root  to top  of
 plant varies from 0.5 to more  than 4, and the  majority of plants have a greater
                                 96
 biomass below ground  than above.     The structural  roots systems range  from
 massive tap  roots and secondary root systems to  filamentous systems.  The  depth
 of roots somewhat depends on the depth  of hardpan or caliche deposits,  but unless
 these deposits are shallow most roots will extend to a depth of 1 m.  Creosote
 bush roots are known  to extend 2 m.
      Roots can be removed by hand.   However, mechanical  methods would probably
 be employed.   A backhoe or crawler type tractor  with a "hula dozer"  blade  can be
 used to remove root systems.
      The major ecological  problems are  again associated  with soil compaction  by
 the  equipment.   However,  the earth disturbed while  removing  the roots could
 cover the annual-seed  containing layer  of soil and  slow  the  initial  stages of
 natural  recovery to two  or  five years.   If caliche  were  brought to the  surface,
 it could form  a  pavement-like  crust  that  might further restrict  plant growth.
      The treatment will  destroy rodent  burrows and  probably  result in immediate
 death to a great  portion  of  the local rodent population.  Long  term effects
 will  be  related  to the destruction of the food supply.    Beatly   has shown  that
 in Southern Nevada the dominance order  of some heteromyids changes following
 brush  removal  or  ground  clearing,  i.e.,  replacement of Dipodomys microps by the
 smaller  D. merriami.  She suggested an  animal vulnerability  to  elimination by
 predators of the  larger  species where there are few shrubs in the environment.
     The  environmental effects will scale with the  size of the area:  the greater
 area disturbed, the greater the effect.
     Areas of the Great Basin and Chihuahuan deserts which receive more  than 250
mm of annual  precipitation could be reseeded with grass to result in desert grass-
 land;  however, the prognosis for recovery within a human generation in  other
desert areas  is poor.
                                     1-32

-------
      (1-4)   Scraping and Grading
      The upper layer of soil to about 5 cm depth along with the contaminant would
 be removed by road grader,  land scraper,  bulldozer or vacuum cleaner.   This is a
 drastic perturbation on the desert environment since it removes the small amount
 of A horizon soil that exists under brush and also removes the seed containing
 layer of soil.   Wind and water erosion potential would be maximized and a further
 deleterious effect would be soil compaction by the earthmoving equipment.
      Under most  circumstances the above ground portions of perennial vegetation
 would be removed with the soil layer.   However,  it has been suggested  that leaving
 200-800 shrubs per ha would speed recovery by providing seed for natural  re-seed-
     95
 ing.     Desert perennial plant density normally ranges from 400 to  3000 plants
 per ha.     Thus,  between one half and  one  quarter of the perennial  shrubs would
 remain if this suggestion is followed.
      If both soil  and plants are removed,  some resprouting of shrubs could be
 expected; however,  the treatment would  remove all  traces of bunch grass.   The
 revegetation sequence of Salsola,  annual grasses,  native annuals, and  the
 gradual  reintroduction of perennial  species  as outlined in Section  III  would
 occur.   Recovery would be slow.   Animal recovery would  be  expected  to  follow
 plant  recovery.

      If  some  shrubs  are  allowed  to remain, the effect on the  animals would  not
 be  as  drastic, since  the  brush would supply  animal cover and  food and  provide
 seed  for  continued revegetation.  The humus  under  the plant could contain  seeds
 of  ephemerals and would partially preserve the fertile  shrub  site in which
 seedlings can become  established.
      In addition to the potential off-site problems that are associated with
 erosion,  some provision must be made either  to remove or to store the contaminated
 soil.  This normally will require a tract of  land  in addition to the treatment
area.
     Figures 1-8  through 1-13 show the recovery that might be expected  eleven
and seventeen years after scraping a desert area.
                                     1-33

-------
   l^^^^^^pS;^----'-^"^p^^^^p^^^


   ^^P^--^-'-:^^^^^^^®



   ^"•r V-v . r."i'"-7*u!3PB!'..*.:->4**»HlL- .••^WjiC.,r4.-.-A- ^wSBUBHw.'^.,.^B^^LflHHBBiHt ~~ 5BS

                                    S*:^&^^i^.«=--'
•*?:>•  	

*gs*--/
:^i - -•-•.'
Figure 1-8.   Great Basin Desert.  A 2 ha rectangular  plot 11 years  after

             blading.  Many grasses are present.   The poisonous shrub

             Halogeton can be seen in the lower part  of the photograph.

             (Photo courtesy of UCLA.)
i' %r:v»;T**e^Sg»SI"3
w^4** rf.*^i^iff^^
                  - -,^^c-rvv.'>v-jt*fe-£-••"*&'$.*&****
                  •*V*^-f%*»^-a'i^»r*J^i V^ 'Jit ..•>•:*•,-• -"iivi^.^A
                  .<•••... - .-•:- • .5 •„ :^*..?**i. • T--*- x^fjj.. '• - '^Q-i ^••^*i*,-.'


                  :-7+^3^&-?Pz--&*£&

                  ^••^/;:- >-:^::,^|:^^¥Q
Figure  1-9.  Great Basin  Desert 11 years after  blading

            of UCLA. )
          VAi


(Photo courtesy
                                1-34

-------
                          -ȣ,

            Great Basin Desert.  Demarcation line between land bladed
            11 years ago (right) and non-bladed land  (left).  The bare
            areas may relate to depth of blading.  (Photo courtesy
            of UCLA.)
Figure 1-11.  Great Basin Desert.   Area in foreground and right was
              bladed 11 years prior to the photograph.  (Photo
              courtesy of UCLA.)
                               1-35

-------
  Figure 1-12.
  Great Basin Desert 17 years after application of water
  to reduce dust followed by blading.  (Photo courtesy
  0T UL/LA. )
Figure 1-13.
Great Basin Desert recovery 17 years after blading.   The
cover in the foreground bladed area is about 25% of the
undisturbed area.  It is unknown if the plant grew from
seeds or crown shoots.   (Photo courtesy of UCLA )
                                1-36

-------
      (1-5)  Shallow Plowing
      It is assumed this technique plows the soil to a maximum depth of 10 cm
 between the perennial shrubs.  It disturbs both annuals and bunch grasses but
 not most of the shrubs.
      The ecological impact would be less than for surface scraping.  The shrubs
 would provide food and cover for animals and a seed reservoir for new growth.
 Little humus would be lost and the seed-containing layers of soil would not be
 removed.  Scarification of the soil would enhance Salsola growth for the first
 two or three years.  This growth would have little effect on germination of
 annuals since Salsola is a summer germinating plant while other annual  forbs
 and grasses are spring and fall  germinating.   The potential for wind erosion
 would be increased until ground  cover by annuals is established.   The potential
 for water erosion  could either be increased or decreased depending on the plowing
 technique.
      The visual effect,  e.g.,  low furrows,  would be visible for many years.   There
 would be little effect on animal  food supply  or population density.

      (1-6)   Deep Plowing
      Plows  capable of plowing  as  deep as  100  cm are available.   This technique
 would result  in a  complete destruction  of the  local  ecosystem.   Succession  would
 start from  bare ground.   The  relatively thin  soil  under  perennial  plant  clumps
 would be destroyed and most of the  seed-containing  soil  would be  turned  under.
 Animal  life  would  be  displaced or destroyed.
      Larger  areas  would  undergo initial  stages  of  recovery more  slowly than
 smaller  areas unless  some  form of revegetation  were  attempted.
      Figure  1-14 shows the recovery to  be expected  from deep plowing.

      (1-7)  Soil Cover Less than 25 cm
      The purpose of this technique would be to bury and immobilize the contaminant.
 It is a relatively drastic treatment.   It would be difficult, if not impossible,
to accomplish without affecting the hydrological drainage system outside the
treatment area.  As new drainage patterns are established and old ones abandoned,
the vegetation of  individual areas would change, although the overall effect on
the region might remain unchanged.
                                    1-37

-------

    Figure 1-14.
Great Basin Desert 17 years after plowing.  The area in
foreground was plowed to a depth of abou? 30 cm.  Cove?
Koto                     "                '
                                                                          the
                                                                          on
      If the soil were placed without extensive compaction,  some shrubs would
 survive since a few desert shrubs are evidently adapted to  periodic silting by
 erosional processes.   Few data are available on the plants  that could survive-
 it is assumed those found in washes or arroyos such as  Hymenoclea or Thanmosma
 could withstand limited  soil cover and grow.   However,  it is  epxected that  most
 shrubs  and  all  bunch  grasses and  annuals  would be  incapable of growth or  germina.
 tion.

      Recovery would be essentially the  same  as starting  from  bare  ground.   Un-
 less  some type  of seeding program  were  instituted,  the growth  of annuals  on  the
 larger-sized treatment areas  would  not  occur until  winds deposited  seed.
     Compaction density of the fill will be an  important variable  for regulatii*
 the time of the initial stages of recovery.  If no compaction occurs the  situa-°
 tion could approximate that around nuclear craters where the water infiltration
 rate is so great that germination cannot occur.  Alternatively, compacting to
a density greater than the original soil would also inhibit  plant growth
                                     1-38

-------
       (1-8)  Soil Cover 25 to 100 em
       This technique would result in the complete loss of the treatment area
 ecosystem as well as off-site changes induced by changing surface drainage
 patterns.  Recovery of the treatment area would be from bare ground.

       (2-1)  Remove Plow Layer
      This cleanup procedure increases the severity of method 1-4, Surface Scrap-
 ing and Grading, by removing slightly more soil.  When the top 10 cm of soil are
 removed all annual seed,  humus,  bunch grass and the meristematic tissue of
 most shrubs would also be removed.   A few shrubs would resprout from the roots,
 but most would not.   Recovery would essentially start with bare earth.   In some
 desert locations extensive deposits of caliche are  found at  plow depth.  If
 these are exposed the existing slow rate of recovery would be retarded.

      (2-2  Remove Shallow Root Zone
           This  method would  remove  the top 40 cm of soil from the contaminated
 area.   This technique would  remove  both  the A and B soil horizons and expose
 the calcarous C horizon.   The  result  would be complete destruction of the  local
 ecosystem.   Recovery  would begin  from  bare ground.
      In any of  the surface alteration  procedures, i.e.,  scraping,  grubbing,
 plowing,  or  soil  removal,  loss of vegetation  and  the  exposure of  the  soil  to
 winds  and rain  will lead  to  erosion problems.  The  magnitude of these problems
 is  discussed in Chapter 2, Prairies, and Chapter  8, Agriculture.

      (2-3)  Scrape and Grade., Mechanically Stabilize
     This  technique would  use mechanical methods  to provide stability to the
 freshly scraped  soil to retard or prevent  soil erosion.  The method would be
 expected  to lessen the impact due to erosional deposition on surrounding areas.
However, none of the possible mechanical stabilization processes would
appreciably speed natural^recovery,  and some of the possible processes would
slow it.

     Two categories of mechanical stabilization are considered here:  compac-
tion and manufactured materials such as nettings, meshes, or plastic films.
                                    1-39

-------
      Soil can be mechanically compacted by bulldozer,  rollers,  or vibratory
 processes to a density that rejects water.   The degree of compaction is largely
 a function of the amount of clay in the soil.   The compacted soil can resist
 wind and water erosion for decades.  Eventually natural processes allow ephemerals
 and grasses to become established.   This vegetation helps loosen the soil  so that
 natural  revegetation  processes can  occur.   If  the  soil has a very high clay content,
 as is the case for dry lake playas, revegetation might require  centuries.
      Although the surrounding area  is protected from material eroded from  the
 treatment area,  the increased water runoff  during  a rain,  unless controlled,
 would cause severe erosion problems on the  adjacent land.
      Manufactured material,  like netting, placed over  the  scraped area would
 provide  protection against water erosion, and  limited  protection to  winds.
 The ecological  effect  would  be substantially the same   as  if the mechanical
 stabilization were not  done.
      Plastic  film could also  provide  a protective  layer.   If exposed to  air  the
 film would  last  3 months  to  2  years,  depending  on  the  composition of the film.
 Films covered by earth  could  last much longer;  however,  winds could  quickly
 remove the  cover.   As the  film degraded, normal  restoration  would occur.

      (2-4)  Remove Plow Layer, Mechanically Stabilize
      A severe environmental disruption.  The effect  would  be  substantially the
 same  as  described  above, expect  the compaction  of  the  clay C  horizon  found in
many  desert areas  would form a  long lasting barrier  to  recovery.

      (2-5)  Remove Shallow Root Zone, Mechanically  Stabilize
     A severe environmental disruption.  The effect would be  substantially the
same as described  above.

     (2-6)  Scrape and Grade,,  Chemieally Stabilize
     This composite method, like the mechanically stabilized methods described
above, would depend on soil removal  to reduce contamination and  a stabilizer* to
both fix any remaining contaminant in place and to  retard soil erosion.  Any
cleanup procedure that removes soil  results in  loss of the local desert eco-
system.  The chemical  (and mechanical) stabilizers  protect the surrounding

*Appendix A lists the  stabilizers.  It is assumed here  that one  of the preferred
 stabilizers of Appendix A, as listed in Treatment  1-1  above, would be used.

                                     1-40

-------
  areas from losses due to erosion,  but at the cost of slower recovery at the
  treatment area.

       The effects  of scraping and grading were described in treatment 1-4 above.
  The addition of a chemical  stabilizer would  retard the initial  stages of re-
  covery until the  stabilized crust  was degraded.   Shallow scraping  will  leave  some
  perennial  brush roots  capable of resprouting,  but forbs and grasses  cannot
  germinate  until the  surface develops  cracks  which allow wind borne seeds to be
  deposited.

       Since the stabilizer will prohibit  water  infiltration,  provision must  be
  made  for mitigating  the  effects  of runoff on the  surrounding areas.

       (2-7)  Remove Plow Layer, Chemically Stabilize
       Soil removal is a drastic environmental perturbation.   The chemical
  stabilizer would retard the  initial stages of recovery one to two years.

       (2-8) Remove Shallow Root Zone,  Chemically Stabilize
      The ecological effects are substantially the same as described above.

      (3-1)  Barriers to Exclude People
      This treatment would depend  on isolating the contaminated area from human
 contact.  This will only be  effective  if the  contaminant cannot  enter the food
 web  leading to man.   The most probable barrier would be chain link  fencing six
 to ten feet high.   Such fencing can be expected to be useful for 25 to 30
 years.

     As long  as provision is made for  animal  passage the barrier would  have  no
 effect on  local ecology.   Some cultural,  e.g.,  archaelogical,  features would
 also be protected  by  the  barrier.

     (3-2)  Exclude Large Animals

     Fences can be designed  to  inhibit animal movement.   If cattle and large
game animals  are excluded from the contaminated area, one direct link in man's
food web will  be blocked.

     Studies,  described in preceding sections, have shown that cattle exclusion
does  not result in changes in vegetation species composition  (at least over the
                                     1-41

-------
 50-year measurement  period),  but that  the number of individual  plants,  and thus
 the total  productivity of the fenced area,  increases markedly.   An increase in
 primary productivity should lead to  an increase  in the number of small  herbi-
 vores.
      A  fence  might produce deleterious effects on large animals in two  ways.
 First,  it  might  exclude animals  from natural  water places.   Second,  it  could
 impede  a natural  migration path.   Lack of a secondary watering  place would prove
 fatal.   If other  migration routes were available they would  probably be used.

      (3-3)  Exclude  Large  and Small  Animals
      This  procedure  would  inhibit the  transfer of the contaminant  through  the
 food  web.   Barriers  can be made  to stop both  the immigration and emigration of
 animals.   It  is assumed that  large grazing  animals,  e.g., cattle,  burros,
 antelope,  deer, would  be removed  from  the contaminated  area.   It would  be  possible
                                                          o
 to  exclude birds  from  the  smallest treatment  zone (0.01  km )  and it  might  be
 worth the  effort  if  a  particular  microenvironment,  e.g., one containing  an
 endangered plant  species,  were to be preserved,  but  it  would be  impractical to
 exclude  birds from larger  areas.  The  actual  isolation  of the food web  leading
 to  man would require analysis of  the local conditions and extent of  man's use  of
 local birds.
      It  might also be  possible to remove  small animals  from  areas  smaller than
           2
 about 1  km , but  it  is  doubtful  if attempts at either direct physical removal
 or  removal  by poisoning  from  larger  areas would  be effective.
      If  small seed-eating  animals are removed it  is postulated that  a compensa-
 tory  increase in  seed  eating  insect  populations  will result  in about the same
 utilization of the local plant resource.15
     The effect of inhibiting small  animal movement, rather  than removal, is
 less clear and is undoubtedly related to the size of the area fenced.  The inter-
 action between plants, herbivores, and carnivores plays an important role in
 the local ecology.  Plants could exist in very small enclosed spaces.  Herbi-
vores and carnivores  require areas at least as large as their foraging range.
                            2
Only the largest area, 10 km , may be large enough to support a population of
carnivores  in a closed system, based  on ecologies existing on small isolated
                                     1-42

-------
   desert  areas  such  as  islands.   It  is not  clear  what  the  effect  of  smaller
   fenced  areas  would be.  One study,  using the vole Microtus.  found that  0 8 ha
   fenced  plots  evidently destroyed the species' natural population control and
   regulatory mechanism.  The  fenced  population expanded until it  was about three
   times as large as  control populations on adjacent sites.  The result was
   overgrazing and destruction of the habitat, followed by  starvation induced
   population decline.    No evidence of this phenomenon was observed on 8 ha
  Plots in the Mojave desert25'96 although herbivore population densities within
  the fenced areas differed from adjacent control  densities in non-uniform ways.
       As long as the vegetation is not  degraded  or destroyed, recolonization  of
  the areas after removal  of the fence would be relatively rapid.   The abundance
  of small mammals  respond  quickly to periods of good  vegetation growth.   They
  have been reported  to  increase their densities at  least  five-fold within one
  breeding season.    French25 has  shown  that  desert rodents  in  the Mojave Desert
  disperse to great distances  both  as young  animals and as  adults  during  the non-
  breeding season.

      Continued maintenance of  isolated animal populations could  affect  the gene
  composition of the  group.  The snlallest groups would  be affected the most.

      (4-1)  Asphalt Hard Surface Stabilization
      In this procedure asphalt  or hot rapid cure  road oil would be applied  over
 the contamination to immobilize it.   The vegetation would be removed prior  to
 the asphalt treatment,  but might remain for the  road  oil  application.
      The technique limits  wind  erosion;  however,  drainage and runoff of rain-
 water to adjoining areas would  be  a  problem.
      In  the  asphalt  treatment the  ecology of the  treatment area would  be de
 stroyed.   Some  perennial plants would survive the road oil,  but the  area would  be
 virtually useless  to support  annual  vegetation or animals.   It  is expected that
 vegetation at the edge of  the area would exhibit  enhanced  growth  due to  addi-
 tional available moisture.

     The asphalt and. road oil treatments will affect local microclimate   De-
 creased surface roughness will result in changes   in the wind profile and will
decrease humidity.   The local heat balance will  be changed since an essentially
spongy surface of low heat  conductivity will be turned into an impermeable
                                    1-43

-------
layer with high capacity for absorbing and conducting heat.  This will raise local
air temperatures.  In calm weather conditions an inversion in the nocturnal
atmosphere can be created that will affect surface wind flow and temperatures
over a relatively large area.  This "heat island" effect has been noted in
                        2 44
areas as small as 0.1 km .    The lowered humidity and increased temperature
will deleteriously affect the small annuals and ephemerals more than the
perennials.
     No recovery can begin until the surface begins to degrade.  This should occur
in about three to five years for the road oil and five to ten years for asphalt.
(However, asphalt can be placed over a compacted base at a thickness which will
survive fifty years or more.)
     As cracks develop in the surface both moisture and wind borne seeds will be
concentrated.  The resulting growth will help break up the surface.  As more
of the surface degrades, annuals, then perennial plants, will become established.
                                                      2
The largest area considered for this treatment,  0.1 km  should not reduce infiltra-
tion and recharge to the extent that ground water would be affected.  However,
the water table under large paved desert areas,  such as airports, has been shown
to move closer to the surface in response to reduced surface evaporation.
Figures 1-15 and 1-16 show recovery from road oil stabilization.

     (4-2)   Concrete  Hard Surface Stabilization
     This treatment  is outside the  scope of work.

     (5-0)   Application  of Sewage Sludge
     This treatment  is outside the  scope of work.

     (6-1)   High Pressure Washing
     This treatment  is outside the  scope  of work,

     (6-2)  Flooding
     This treatment  is outside the  scope  of work.

     (7-0)  Soil Amendments Added
     This treatment  is outside the  scope  of work.
                                    1-44

-------
Figure 1-15.  Great Basin Desert.   Foreground first treated  with  hot
             rapid-cure road oil  stabilizer, then scraped.   Photo 17
             years after disturbance.   (Photo courtesy of UCLA.)
                                            *	 .»-•
                                                             I?**?. .Sfiiy:
                                                           :  ^fi^-Sii
                                                           S-&. jJsSLtvl t^-r i«9
                                                - Y-^H1
                                                ^I:^^f^§^^^?
                                                ^^^^^S^^l^
                                                ;'i^i^^^-Sr
                                                 '•^S^^??^-;-^^:.
                                                      "*ijv
                                             ,-; '^ -../•.:^4';'s^
                                             :. \  ^jF^-^iY.-'S:
Figure 1-16.
                                        »^*-
                                                          . >-«••-• ••-.
          Soil 17 years after application of a hot rapid-cure  road oil
          stabilizer.  The soil  pieces vary from 1 to 5 cm thick.
          Vegetation is growing  in  the area.  (Photo courtesy  of UCLA  )
                              1-45

-------
 1.5  RECOVERY FOLLOWING CLEANUP
      Earlier  in  this  chapter  it was  noted  that  the  concept  of  a  single  stable
 climax  vegetation  in  each  desert type  was  not valid.   A  number of  different
 stable  ecosystems  may be possible depending  largely on long term climatic
         35
 factors.    Further,  existing ecosystems are dynamic  and changing  with  time.
 Thus, recovery is  best defined in terms relative  to surrounding  areas rather than
 to  some standard climax ecosystem, or  to some measured pre-treatment condition.
 It  has  also been shown that there are  no irreversible changes; given enough
 time  the desert  will  adjust to any insult.   Treatments which lead  to massive
 erosion and land form change  will be permanently  altered visibly although  the
 change  will not  affect the equilibrium ecosystem.
      The endpoint  for natural  secondary succession  and unassisted  recovery follow-
 ing application  of one of  the  cleanup  techniques  can  be  specified  in a  number of
 ways  depending largely on  the  effort undertaken to  measure  the completeness of
 recovery.
      These endpoints  include:
      (1)  No measured difference,  other than normal variance based on measure-
          ment technique and  ecological variability,  in  selected ecosystem com-
          ponents  between  the  treatment area and  adjoining  undisturbed  areas.
      (2)  No visible  difference  between treated and untreated areas apparent
          to a desert  ecologist,  without benefit  of field measurements.
      (3)  No visible  difference  to the layman.
 In this chapter endpoint 3 is  specified as defining recovery to climax  condition.
 Figures 1-17 through  1-20 give the recovery sequence  for several of the cleanup
 techniques between the  end of  cleanup  and climax  as defined by endpoint 3.  See
 Appendix B for applications.

 1.5.1   Unassisted  Succession
     The vegetation stages leading to  the defined recovery, based on information
presented in Sections  1.2 and  1.3 of this chapter are  summarized below.*
 These stages may not be serai in that the vegetation which forms the pioneer
 community is present at climax—indeed, the same plants may be present.
                                    1-46

-------
                                                                HYPOTHETICAL DESERT
 -n

 to
 -s
 O>
 GO
 .a
 c
 o
 o
 -h

 a>
 o
 o

 o
ta
 _j.
 o

 -s
 fD
 O
 O
 <
 0)
tQ


O
8
                               7
       ERODABILITY
                                  BIOMASS
                                                        PRIMARY PROD.
                              I
                                                                                     COVER
                                                                                                          SP. DIVERSITY

                                                                                                          (VEGETATION)

-------
                                                                           HYPOTHETICAL DESERT
 i
-^
00
~n
UD
5 *
n> -
i — >
i
00
Z3 fD
CL 0
O
o <
— ' fD
fD -5
-5 o
O CO °
c n> X
S£ 8













	 N




1
\
)




	 ^^x>>^
^*^>





>






0
3D
>
bo
bo


|

••^
PERENNIALS




^




-.
cn
V)
>
t~
r~
O
W PLOWING


   O5
 rs is
UD O
   o>
 <
 CD -h
to o
               IQ


                00
                ir
                CU
               TD


               O
                            ERODABILITY
                                                        BIOMASS
                                                                             PRIMARY PROD.
                                                                                                          COVER
                                                                                                                               SP. DIVERSITY

                                                                                                                               (VEGETATION)

-------
                                                               HYPOTHETICAL DESERT
VO
        CO
        c.
        -$
        n>
        vo
        co
        n>
 n>
 3
 O
 n>

 o
 -h

 ft)
 o
 o

 o
CO

 o


 ft)
 o
 o
        o


        o
             p
             ll
     ?!
     ii
             053
             3-
-------
2 3 to 25  SOIL REMOVAL, MECHANICAL STABILIZATION3     4-1,4-2  HARD SURFACE STABILIZATION
                                                    00
                                                     . uj

                                                    &>
                            PERENNIAL
                                                                                   PERENNIALS
                          10
                                     100 200
                                                                          10
                                                                                      100 200
             Includes Scraping and Gradmg-Mechanically stabilize, Remove plow layer- Mechanically stabilize,

             Remove shallow root zone-Mechanically stabilize





     Figure 1-20.    Sequence of ecologic  recovery  following  cleanup.
                                           1-50

-------
 It  is  assumed that  secondary  succession begins with bare ground, which contains
 no  particular deleterious edaphic features such as surface clay hardpan or
 caliche.
     Bare ground, if  scarified, is invaded by Russian Thistle  (Salsola) the
 first  summer following cleanup.  '   *  *    During the first year pioneer annuals
 and grasses appear  if adequate moisture is available, otherwise an additional year
 or  two may elapse.  The timing of this initial annual colonization is also de-
                                                                       2
 pendent on the size of the affected  area.  Areas larger than about 3 km  will
 require an additional year.    The introduction of non-native annual grasses is
 at  the expense of native perennial grass.    The result of this initial stage
 in  secondary succession is an annual grassland which can persist for decades,
 or  possibly hold at grassland stage  by grazing and range management techniques.
 During this stage the density of grass and ephemerals on the treated area is two
 to  three times that on adjacent areas  although the total productivity of the
 area is lowered   largely because older perennial shrubs produce the greatest
 biomass.    There may be a shift in  the relative species composition of rodents,
 which is restored as perennial shrubs reappear.
     Starting in the  second year—again,  assuming adequate moisture—pioneer
 shrubs characteristic of disturbed sites, e.g., Atriplex, Thamnosma, Salazaria,
                      59 99                                          	
 and Hymenoclea appear.  '
     In the third year perennial shrubs which will make up the climax community
 begin to sprout.   Perennial vegetation requires a number of consecutive years
 of  above normal soil moisture and some relief from small animal grazing to become
 established.     Individual plants reach maturity in about 30 to 40 years.   '
 The area can be expected to have a perennial  shrub density of about 20 percent
 that of surrounding areas in 13 years  and about 25 percent in 17 or 20 years.   '
 Complete recovery of the area should occur in about 80 years.
     Manipulation of the landscape  to create  pockets which trap moisture may be
used to speed the intermediate stages of  succession,  i.e.,  the early perennial
            96
 shrub stage.     Enhancements of shrub density by factors of 2 within these
 features have been noted   but there is a corresponding lessening of density on
nearby plots  due  to restricted soil  moisture  so the net density for the area is
about the same.   The landscape manipulation would speed the restoration of the
                                      1-51

-------
original rodent species composition by providing earlier food and cover for the
displaced species.

1.5.2  Cleanup Recovery Categories
     From the preceding summary of a hypothetical desert secondary succession
and the description of the cleanup techniques presented in Section 1.4, it is
clear that there are major divisions in impact among those cleanup techniques
that delay recovery, those that result in bare ground with no vestige of or-
ganic matter, and those where factors which relate to bare ground recovery are
ameliorated.  In the first category it is important to note that eventually
the bare ground condition is approximated; therefore the time difference in
recovery among all the procedures that remove and stabilize the soil is actually
the time from the application of the treatment until the bare ground approxi-
mation is achieved.   The consequences of this outlook are significant, for it
implies that once the seed containing layer of soil and the roots of perennial
plants are removed,  the time to recovery is relatively independent of treatment
type.   In the last category, recovery from a given cleanup procedure is di-
rectly related to the difference between the disruption produced by the cleanup
procedure and the bare ground state.

1.6  QUANTITATIVE ASSESSMENT OF CLEANUP IMPACTS
     The rationale underlying assessment of the ecological impact of various
cleanup techniques is to be able to relate the cleanup procedure to the potential
hazard of a contaminant so that the environmental costs associated with the
cleanup can be weighed against the benefits of hazard reduction.  Since these
cleanup techniques will have different levels of efficiency for different types
of contaminants,  some measure of the impact of each procedure is necessary for
this evaluation to be made.
     A number of measures are available to estimate the rate of ecological
recovery.  They are usually in terms of changes in energy cycles, nutrient
cycles, food chains, density patterns in time and space, and various control
mechanisms„  Unfortunately, the interrelationships among most of these in
desert ecosystems are Largely unknown.

-------
      In  this  section,  time  of  recovery  to  the  climax  desert  community  is  taken
 as  a  surrogate  for  ecological  impact  quantification.   Implicit  in  this measure
 is  the assumption that recovery  time  integrates  the effects  of  changing primary
 productivity, species  composition of  both  plants and  animals and chemical and
 energy cycling  in the  environment.  Any other  measure of  the differential impacts
 among the  treatment  types would  scale about  the  same  as time to recovery  and
 would produce the same ordering  of  impacts.
      Unfortunately,  the measure  selected cannot  be precise.   First, deserts are
 not homogeneous areas.  Local  eadphic and  climatic factors interact with  local
 variations  in plant  and animal species  to  influence recovery rates.  Second,
 recovery is very sensitive  to moisture, especially in the early stages of
 pioneer  vegetation growth.  This means  that  the  temporal distribution of  rainfall,
 as well  as  its magnitude, is an  important  variable.   These climatic factors tend
 to have  a  greater variance  from  the mean as  the  mean  decreases.  Third, the
 magnitude  of impact  depends on the  severity  of the cleanup in any  given situa-
 tion, for  example, shallow  plowing was  shown in  Section 1.5  to  be  relatively benign,
 If, however, the perennial  vegetation is plowed  (a circumstance which might be
 dictated by local conditions) and if  the plowing occurred at  such  a time that
 the loosened seed containing surface  soil  were blown  away by winds, recovery
 would require 40 to  80  years rather than the 12  years  estimated.
     Table 1-1 lists the median time  to complete recovery for each cleanup
 procedure for various  sized areas.  There  is no  clear  relationship between size
 of area and impact listed in the literature; however,   the impact is expected to
 increase as the area increases.

 1.6.1  Impact Assessment
     For all size areas there is a relationship between severity of impact and
magnitude of surface disruption.   Soil removal, no matter how slight,  constitutes
a major perturbation to the ecosystem.  Unfortunately, most procedures require
 soil removal.
     The real  choices are:   first, do nothing other than erect a barrier;
second,  stabilize or turn  under part of the contaminant; third,  remove con-
taminated vegetation; fourth,  remove contaminated soil.  Choice 1,  2,  or 3
would allow time to  evaluate expected effects in detail for a particular loca-
tion,  but would  be of limited usefulness against a hazardous  contaminant that

-------
           Table 1-1.   Estimates  of the years to recover productivity after  application  of the
                        various  cleanup procedures.
Treatment
0-1 Natural Rehabilitation
1-1 Chemical Stabilization
1-2 Remove Vegetation by Clear Cutting
1-3 Remove Stumps and Grub Roots
1-4 Surface Scraping and Grading (<5 cm)
1-5 Shallow Plowing (<10 cm)
1-6 Deep Plowing (10 to 40 cm)
1-7 Soil Cover (<-25 cm)
1-8 Soil Burial (25 to 100 cm)
2-1 Remove Plow Layer Soil (10 cm)
2-2 Remove Shallow Root Zone Soil (<40 cm)
2-3 Scrape Surface, Remove (<5 cm), Mechanically Stabilize
2-4 Remove Plow Layer (10 cm), Mechanically Stabilize
2-5 Remove Shallow Root Zone (<40 cm), Mechanically Stabilize
2-6 Scrape Surface, Remove (<5 cm), Chemically Stabilize
2-7 Remove Plow Layer (10 cm), Chemically Stabilize
2-8 Remove Shallow Root Zone (40 cm), Chemically Stabilize
3-1 Barriers to Exclude People
3-2 Barriers to Exclude Large Animals
3-3 Barriers to Exclude Large and Small Animals
4-1 Asphalt Hard-Surface Stabilization
4-2 Concrete Hard-Surface Stabilization
5-0 Sewage Sludge Application
6-1 Washing (-3 mm)
6-2 Flooding (3 to 30 cm)
7-0 Soil Amendments Added
Area Disturbed
(km?)
0.01
80
4
40
85
80
12
85
80
85
85
80
95
9.5
95
85
85
85
0
0
0
95
95
*
*
*
*
0.1
80
4
40
85
80
12
85
80
85
85
80
95
95
95
85
85
85
0
0
0
95
95
*
*
*
*
1.0
80
4
40
85
80
12
85
80
85
85
80
100
100
100
85
85
85
o
o
0
NA
NA
*
*
*
*
10.0
85
6
40
90
85
15
90
85
90
90
85
105
105
105
90
90
90
o
o
0
NA
NA
*
*
*
*
Assumptions


70 years for chaining







Assume compaction
as mechanical
stabilizer





Assume 15 years to degrade hard surface.
Assume 15 years to degrade hard surface.




 *0utside scope of work.
NA - not appropriate for this land type.

-------
 could enter man's food chain.  The real decision may be between erecting, main-
 taining, and monitoring a barrier until the contaminant is no longer hazardous,
 or performing a cleanup procedure whose effects cannot be eradicated naturally
 for 80 to 100 years and where ameliorative reclamation may be impossible.  In
 one respect this simplifies the cleanup decision since it implies that the only
 variable which needs to be considered in desert environments is the efficiency
 of the cleanup procedure.
      In summary, cleanup techniques can be grouped into two major categories
 of severity of impact, each category being further subdivided into groups of
 slightly differing degrees of impact.  In the first category, that of less-
 severe impact, the group of treatments having the least impact is the erec-
 tion of barriers (3-1  through 3-3).   The next lowest impact group includes
 those procedures which are less extreme in the degree of bare ground state
 produced.   In this group are chemical stabilization (1-1),  shallow plowing
 (1-5),  and  clear cutting (1-2).   In each of these treatments, the seed-con-
 taining layer of soil  is not removed, and perennial plants  achieve relatively
 rapid recovery.   Ephemeral and annual grass growth is either enhanced or  un-
 affected.
      In  the  second category,  that  of  much greater impact, the first  group
 includes  scraping  and  grading  (1-4),  soil  cover  less  than 25 cm  (1-7),  and
 plow layer removal  (2-1),  as  these procedures  result  in  a bare ground state.
 Of very  similar  impact,  though slightly  greater,  are  stumping and  grubbing (1-3),
 shallow  root  zone  removal  (2-2), deep plowing  (1-6),  soil cover  greater than  25
 cm  (1-8), and  soil  removal  in  combination  with chemical  stabilization  (2-6
 through  2-8).  The  last  group, that with  the most  severe impact,  includes  the
 treatments that  combine  soil removal  with  mechanical  stabilization  (2-3 through
 2-5), and hard-surface stabilization  (4-1  and 4-2), as the resulting  soil  compac-
 tion would further delay forbs and grasses  establishment and  perennial vegeta-
 tion germination.

 1.6.2  Recovery Assessment
     As discussed in Section 1.5,  there are major divisions among cleanup  tech-
niques with regard to time to recovery.  In general, those treatments involving
 less perturbation to the seed-containing layer of soil will  result in shorter
recovery times than the treatments which effect the bare ground condition.
                                       1-55

-------
     Areas with barriers (3-1 to 3-3) will not experience succession in the
sense of the other cleanup technologies.  The effects of barriers on vegetation
are somewhat dependent on the extent of pre-treatment overgrazing but are gen-
                                                                     o 77 o-i
erally considered to be favorable by increasing the density of plants »»
and reducing animal-caused compaction.  >    The longer an area is protected
                                                                          77 5? 1
from grazing pressure, the greater the total number of plants and biomass.  »
However, exclusion of large animals might result in changed plant species compo-
sition leading to an increase in poisonous species, " and containment of rodent
populations might lead to massive density fluctuations and a reduction of vege-
tation biomass.43  For these treatments, the size of area would have little
effect; however rodent population fluctuation would not be expected for the
largest area.
     Chemical stabilization (1-1), shallow plowing (1-5), and clear cutting
vegetation (1-2) can be considered a family of treatments with sequentially
increasing recovery time related to the amount of initial environmental disrup-
tion.  In each treatment the seed-containing layer of soil is not removed, and
perennial plants (or their roots) remain to form the basis for relatively rapid
recovery.  Chemical stabilization affects annual production for one or two grow-
ing seasons but should not otherwise alter recovery.  Shallow plowing allows
most of the perennial shrubs to remain.  Scarification of the surface will allow
Salsola to become established and reduce erosion, but once the soil is naturally
compacted to pre-treatment density and local erosion removes the plow furrows,
recovery will be complete.  The slightly greater soil moisture found in the
plow furrows should enhance ephemeral and annual grass growth; however, even
this impact will be unmeasurable within 10 to 12 years.  With both chemical
stabilization and shallow plowing, it is expected that the recovery time would
                                                 2
increase slightly as the area increases over 1 km .
     Clear cutting vegetation allows perennial plants to resprout from existing
roots, thus time to total recovery is shortened and is equal to the time required
for a perennial shrub to reach mature stature — about 40 years.  The ephemerals
and grasses should not be affected by this procedure.  Shrubs have been shown
to increase productivity when closely cropped.     Hand cutting would not result
in the surface scarification necessary for Salsola germination.  Chaining would
disturb some of the seed-containing soil layer and slow the early phases of
                                      1-56

-------
 recovery, but the slowest portion of recovery, regrowth of the perennial vege-
 tation would not be affected and recovery within 40 years can be expected.  The
 size of area would have no effect on recovery time with this treatment.
      The procedures of scraping and grading (1-4), removal of plow layer (2-1),
 and soil cover less than 25 cm (1-7) result in an initial bare ground state and
 exhibit the same successional pattern and timing as natural revegetation.  Full
 recovery is expected in about 80 years.   The size of area is expected to have
 a slight effect on recovery time as the  area size increases beyond 1 km2.
      Four of the techniques,  stumping and grubbing (1-3), remove shallow root
 zone (202),  deep plowing (1-6),  and soil cover greater than 25 cm (1-8)
 result  in similar recovery patterns.  Each treatment is capable of removing
 or destroying all vegetation  and both the A and B soil horizons,  therefore total
 recovery is  retarded somewhat by the presumed  existence of hardpan on the sur-
 face.   The hypothetical  succession  stages and  their initiation are about the
 same as for  natural  revegetation; however,  at  any given equal  time plant density
 would be slightly less  for these  treatments  than  for natural  succession.  It is
 estimated  that  recovery  would occur  in approximately 85 years.   The size of area
 is expected  to  have  a slight  effect  on recovery time as the  area  size increases
           2
 beyond  1  km  .
      Procedures  which use  stabilization  in  conjunction with  soil  removal (2-3
 through  2-8)  result  in slightly  longer recovery times  than the  preceding group,
 but  the  recovery  is not  qualitatively different.  The  chemically-stabilized com-
 bination treatments  (2-6 through 2-8) result in a delay of the  initial invasion
 phase for  a year  or two, and  climax  should be achieved in 80 to 85  years.   The
 size of area  is expected to have a slight effect on  recovery time as  the area
 size increases beyond 1 km  .  The mechanically-stabilized combination treat-
ments (2-3 through 2-5), exemplified by compaction, would delay the establish-
ment of the pioneer forbs and grasses for 2 to  10 years and would lengthen  the
time until perennial vegetation germinated to about  15  years.  About 90  to  95
years would be required for full recovery.  It is expected that increasing  area
size would increase the time of recovery.   Hard-surface stabilization is an
extension of mechanical  compaction;  here, however, no pioneer species can become
established until the hard surface begins to degrade.   From that point however,
recovery will occur in about 80 years.
                                       1-57

-------
     In summary, to a large extent, recovery time is relatively independent of
cleanup technique when the seed-containing soil layer and vegetation roots are
removed.  The most drastic treatment adds only about 15 percent to the overall
time of recovery, compared to natural revegetation.

1.7  CONCLUSIONS
     One important topic has not been explicitly addressed.  That is, what is
the goal of restoration?  The goal might be to restore the area to the pre-
disturbance ecological community.  It might be to establish a similar, though
not exact, community, for example, a pure stand of creosote bush in place of
mixed perennial species.  It might be to change the land use from desert to
desert grassland for grazing or even to agricultural land.
     Two important factors dominate desert revegetation and ecological recovery
of disturbed areas.  First, moisture is the critical determinant both of the
speed of recovery and the relative dominance of perennial shrubs or perennial
grasses.  Recovery is usually impeded by the generally low infiltration rate
of desert soils so that much of the precipitation is lost as runoff.  Second,
there is no serai succession; the perennial plants first established form
the stable climax community.  A mature stable community is normally achieved in
about 80 years after a disturbance.  Superimposed on these factors is evidence
that natural vegetation changes occur in response to long term climatic change,
and there is historic evidence that vegetation changes occurred at a time scale
that is comparable to that of unassisted revegetation.  Successive dry years
reduce perennial grass cover, while a series of wet years has the opposite ef-
fect.  A goal of returning a desert to a pre-cleanup condition may be impossible
if, during the restoration period, climatic conditions which existed prior to
the cleanup change.  A new stable communuty may exist, but it might be different
from surrounding communities.
     Of the treatments evaluated, several would require thoughtful considera-
tion prior to their use based on their deleterious effect on recovery of the
ecosystem.  These are:  hard-surface stabilization (4-1, 4-2), soil removal
followed by chemical stabilization (2-6 through 2-8), removal of the shallow
root zone (2-2), soil cover greater than 25 cm (1-8), deep plowing (1-6),
stumping and grubbing (1-3), removal of the plow layer (2-1), soil cover less
than 25 cm (1-7), and scraping and grading (1-4).  As discussed in Section 1.6,
                                       1-58

-------
the common factor in all of these treatments is the effecting of a bare ground
condition; once this happens, recovery time is greatly increased over the time
involved for treatments that do not create this condition.
     Preferred treatments, therefore, in terms of ecosystem recovery, would be:
erection of barriers (3-1 through 3-3), chemical stabilization (1-1), shallow
plowing (1-5), and clear cutting vegetation (1-2).  Unfortunately, for many
postulated materials requiring cleanup, the ecologically preferred treatment
may not be the most effective for hazard reduction.
                                      1-59

-------
1.8  DESERT REFERENCES

1.   Allred, B. W. and E. S. Clemments, Dynamics of Vegetation.  Selections
     from-the Writings of Frederic E. Clemments.  New York: H. H. Wilson, 1949.

2.   Allred, D. M. and D. E. Beck, Range of Movements and Dispersal of Some
     Rodents at the Nevada Test Site, J. Mamm. 44: 190-200, 1963.

3.   Beatley, J. C. Ecology of the Nevada Test Site. IV. Effects of the Sedan
     Detonation on Desert Shrub Vegetation in Northeastern Yucca Flat, 1962-65.
     USAEC Report UCLA 12-571, 1965.

               , Effects of Rainfall and Temperature on the Distribution and
     Behavior of Larrea tridentata (creosote bush) in the Mojave Desert of
     Nevada.  Ecology 55:245-261, 1974.

5.   	, Phenological Events and Their Environmental Triggers in Mojave
     Desert Ecosystems, Ecology 55:856-863, 1974.

6.   	, Environments of Kangaroo Rats (Dipodomys} and Effects of En-
     vironmental Change on Populations in Southern Nevada, J. Mammology 57:
     67-73, 1976.

7.   	, Rainfall and Fluctuating Plant Populations in Relation to
     Distribution and Numbers of Desert Rodents in Southern Nevada, Oecologia 24:
     21-42, 1976.

8.   Blydenstein, J., C. R. Hungerford, G. I.  Day and R. R. Humphrey, Effect of
     Domestic Livestock Exclusion on Vegetation in the Sonoran Desert, Ecology 38
     522-526, 1957.

9.   Branscomb, B.  L., Shrub Invasion of a Southern New Mexico Desert Grassland
     Range. J. Range Management 11:129-32.

10.  Brown, A. L.,  Shrub Invasion of Southern Arizona Desert Grassland.  J. Range
     Management 3(3):172-177, 1950.

11.  	, and A. C.  Everson,  Longevity of Ripped Furrows in Southern
     Arizona Desert Grassland.  J. Range Management 5:415-419, 1952.

12.  Brown, J. H.,  Geographical Ecology of Desert Rodents.  In (Cody, M.  L., and
     J.  M.  Diamond, eds.) Ecology and Evolution of Communities, Cambridge,
     Belknap Press, 1975.

13.  	, and D. W.  Davidson,  Competition Between Seed-Eating Rodents
     and Ants in Desert Ecosystems.   Science 196:880-882, 1977.

14.  Buffington, L. C. and C. H.  Berbel,  Vegetational Changes on a Semi-Desert
     Grassland Range from 1858  to 1963.  Ecol.  Monograph 35:139-164, 1965.
                                     1-60

-------
 15.  Campbell, R. S., Vegetation Succession in the Prosopis Sand Dunes of
      Southern New Mexico.  Ecology 10:392-398, 1929.

 16.  Champlin, J. B. F., Ecological Studies Along Transmission Lines  in South-
      western United States. J. of Env. Sci. p 11-16, Nov./Dec., 1973.

 17.  Chen, R. M. and A. E. Chen, The Primary Productivity of a Desert Shrub
      (Larrea tridentata) Community.  Ecol. Monog. 35:355-375, 1975.

 18.  Christian, C. S. and R.  0. Slatger, Some Observations on Vegetation Changes
      and Water Relationships in Arid Areas.  Arid Zone Research XI, Climatology
      and Microclimatology, Paris, Unesco. p 156-158, 1958.

 19.  Clements,  F. E., The Origin of the Desert Climax and Climate.   Essumpin
      Geobotany-In Honor of William Albert Setchell.  Berkeley,  U.  of California
      Press, 1936.

 20'  	>  Plant Succession—An Analysis of the Development  of Vegetation
      Carnegie Inst.  Wash.  Publ. No.  242,  1916.

 21.  Darrell,  L.  W.,  and L. M.  Shields,  Characteristics of Soil  Algae Relating
      to Crust  Formation.   Trans.  Am.  Microscopical  Soc.  80:73-79,  1961.

 22.  Evenari, M.,  L.  Shannor  and  N.  Tudmor,  The Negev.   London,  Oxford University
      Press, 1971.                                                                3

 23.  Fourhier D'Albe,  E. M.,  The  Modification  of  Microclimates.  Arid Zone
      Research,  Climatology, Paris UNESCO,  1958.

 24.  French, N. R., B. G. Maza, H.  0. Hill,  A. P. Aschwanden, and H.  W Kaaz
      A  Population  Study  of  Irradiated Desert Rodents.   Ecological Monographs'44:
 25.
                , T. Y. Tagami and P. Hayden, Dispersal  in a  Population of
     Desert Rodents.  J. Mamm. 49:272-280,  1968.
26.  Fritts, H. C., Tree-Ring Evidence for Climatic Changes  in Western North
     America, Monthly Weather Review 93(7):421-443, 1965.

27•  	> Tree Rings: A Record of Climate Past.  Environmental Data Ser-
     vices, Department of Comm., monthly publication, July 1977.

28t  	' Tree Rings and Climate.   London, Academic Press, 567 pp, 1976.

29.  Fuller, W. A., Management of Southwest Desert Soils.  Tucson, Arizona
     University of Arizona Press, 1975.                                   '

30.  Gardner, J.  L., Effects of Thirty Years of Protection from Grazing in Desert
     Grassland.  Ecology 31 (1):  44-50,  1950.

31.  Giles, R.  H.  Jr.,  Wildlife Management Techniques.  The Wildlife Society
     Washington,  D.C.,  1969.
                                     1-61

-------
32.  Glendening, G. E., Some Quantitative Data on the Increase of Mesquite and
     Cactus on a Desert Grassland Range in Southern Arizona.  Ecology 33:319-
     328, 1952.

33.  Hanson, H. C., Intensity of Grazing in Relation to Proximity to Isolation
     Transects.  Ecology 10: 343-346, 1929.

34.  Haskell, H. S., Successional Trends on a Conservatively Grazed Desert Grass-
     land Range.  J. Amer. Soc. Agron. 37(12):978-990, 1945.

35.  Hastings, J. R., and R. M. Turner, The Changing Mile: An Ecological Study
     of Vegetation Change with Time in the Lower Mile of an Arid and Semiarid
     Region, Tucson, University of Arizona Press, 1965.

36.  Hervey, D. F., Reaction of a California Annual-Plant Community to Fire.
     J. Range Management 2(3):116-121, 1949.

37.  Hills, E. S., Arid Lands and Human Problems.  In (E. S. Hill, ed.) Arid
     Lands:  A Geographical Appraisal, London: Methuen and Company, 1966.

38.  Holmgren, R. and S. Hutchings, In (McKell et al., eds.) Useful Wildland
     Shrubs—Their Biology and Utilization.  U.S. Forest Service, General Tech-
     nical Report INT-1, 1972.

39.  Hull, A. E., Jr., and R. C. Holmgren, Seeding Souther Idaho Rangelands.
     U.S. Forest Service Research Paper INT-10, 1964.

40.  Humphrey, R. R., The Desert Grassland, a History of Vegetational Change and
     an Analysis of Causes.  The Botanical Review 24:193-252, 1958.

41.  	, Fire as a Means of Controlling Velvet Mesquite, Burroweed and
     Cholla on, ... , Southern Arizona Ranges.  J. Range Management 2(4):175-182,
     1949.

42.  Kb'ppen, W. Die Klimate der Erde, Berlin, 1931.

43.  Krebs, C. J., M. S. Gaines, B. L. Keller, J. H.  Myers, R.  H. Tamarian, Popula-
     tion Cycles in Small Rodents.   Science, 179:35-41,  1973.

44.  Landsberg, H. E., Man-Made Climatic Changes.  Science 170:1265-1274, 1970.

45.  Liddle, M. S. and K. G. Mavre, The Microclimate of Sand Dune Tracks: The
     Relative Contribution of Vegetation Removal and Soil Compression.   J.  Appl.
     Ecol. 11(3):1057-1068, 1974.

46.  Longstreth, D. J. and D. T. Patten,  Conversion of Chaparral to Grass in
     Central Arizona:  Effects on Selected Ions in Watershed Runoff. Am.  Mid.
     Nat. 93(l):25-34, 1975.

47.  Martin, W. P., Distribution of Activity of Azotobaoter in  the .Range and
     Cultivated Soils of Arizona.   Arizona Agricultural  Experiment Station,
     Technical Bulletin 83:332-364, 1940.
                                     1-62

-------
 48.  McKell, C. M., Shrubs—A Neglected  Resource  of Arid  Lands.   Science  187-
     803-809,  1975.

 49.  Meigs,  P., World Distribution of Arid and  Semi-Arid  Homoclimates.  Reviews
     of Research on Arid  Zone Hydrology,  Paris, UNESCO, 1953.

 50.  Muller, C. H., Plant Succession in  the Larrea-Flourens-ia Climax.   Ecology
     21:206-212, 1940.

 51.  National  Academy of  Sciences, Natural Academy of Engineering, Rehabilitation
     Potential of  Western Coal Lands. Environmental Studies Board, Cambridge,
     Massachusetts; Ballinger Publishing, 1974.

 52.  Nishita,  H.,  and R. M. Haug, Distribution  of Different Forms of Nitrogen
     in Some Desert Soils.  Soil Science  116:51-58, 1973.

 53.  Norris, J. J., the Effect of Rodents, Rabbits and Cattle on Two Vegetation
     Types in  Semi-Desert Range Land.  New Mexico Agricultural Experiment Station
     Bulletin  No.  353, 1950.

 54.  Odum, E.  P.,  Fundamentals of Ecology.  Third Edition, Philadelphia: W. B.
     Saunders  Company, 1971.

 55.  Parker, K. W., G. Stewart, A. P. Plummer, J. H. Robertson, A. C. Hill, Jr.
     Controlling Sagebrush on Rangelands.  U.S. Department of Agriculture Farmers
     Bulletin  No.  2027, 1954.

 56.  Pechanec, J. C., G. Stewart, and J.  P. Blaisdell.  Sagebrush Burning:  Good
     and Bad.  U.S. Dept. of Agriculture  Farmers  Bulletin 1948 (rev.),  1954.

 57.  Pickford, G. D., The Influence of Continued  Heavy Grazing and of Promiscuous
     Burning on Spring-Fall Ranges in Utah.  Ecology 13(2):159-171, 1932.

 58.  Piemeisel, R.L, Causes Affecting Change and  Rate of Change in a Vegetation
     of Annuals in Idaho.  Ecology 32(l):53-72, 1951.

 59.  	,  Changes in Weedy Plant Cover on  Cleared Sagebrush Land and Their
     Probable Causes.   U.S. Department of Agriculture Technical Bulletin 654,
     1938.

 60.  Plummer, A.  P., D.  R. Christenson,  and S. B.  Monsen,  Restoring Big Game
     Range in Utah.  Utah Division of Fish and Game Publication No.  68-3, 1968.

 61.  Reynolds,  H.  G.,  and J. W.  Bohning, Effects  of Burning on a Desert Grass-
     Shrub Range  in Southern Arizona.  Ecology 37 (4):769-777,  1956.

 62.  Rhoads,  W. A., A.  D.  Kantz,  and H.  L. Ragsdale,  Ecological and Environ-
     mental Effects from Local  Fallout  from Schooner.   2.  The Beta and Gamma
     Radiation Effects from Close-in Fallout.   U.S.  Government Report PNE-529,
     1972.

63.  	,  R. B.  Platt,  and R.  A.  Harvey,   Radiosensitivity of Certain
     Perennial  Shrub Species Based on a Study of the Nuclear Excavation Experi-
     ment, Palanguin,  With Other  Observations of Effects on  the Vegetation.   U.S.
     Government Report  CEX-68.4,  1969.

                                     1-63

-------
64.  Richard, W. H.,  J. D. Hedlund, R. E. Fitzner, Elk in the Shrub-Steppe Region
     of Washington: An Authentic Record.  Science 196:1009-1010, 1977.

65.  Richard, W. H. and L. M. Shields, An Early Stage in the Plant Recoloniza-
     tion of a Nuclear Target Area.  Radiation Botany 3:41-43, 1963.

66.  Romney, E. M., V. Q.  Hale, A. Wallace, 0. R. Lunt,  J. D. Childress, H. Kaaz,
     G. V. Alexander, J. E. Kinnear, and T. L. Ackerman, Some Characteristics of
     Soil and Perennial Vegetation in Northern Mojave Desert Areas of the Nevada
     Test Site.  USAEC Report UCLA 12-916, 1973.

67.  	, A. Wallace, J. D. Childress, Revegetation Problems Following
     Nuclear Testing Activities at the Nevada Test Site.  In (Nelson, D. J. ed.)
     Radionuclides in Ecosystems, Proc. Third National Symposium on Radioecology.
     USAEC Publication CONF. 710501 pp 015-1022, 1973.

68.  Sauer, C. 0., The Agency of Man on the Earth.  In Man's Role in Changing
     the Face of the Earth, Chicago: University of Chicago Press, 1956.

69.  Semple, A. T. Improving the World's Grasslands, FAO Agricultural Studies
     No. 16, 1951.

70.  Shantz, H. L. ,  History and Problems of Arid Lands Development.  In (G. F.
     White, ed.) The Future of Arid Lands.  Am. Assoc. Adv. Sci. 43, 1956.

71.  Shantz, H. L., The Relation of Plant Ecology to Human Welfare, Ecol. Monog.
     10(3):311-342, 1940.

72.  Shantz, H. L. and R.  Zon, Natural Vegetation.  U.S. Department of Agriculture,
     Atlas of American Agriculture Part I Section E, 1924.

73.  Shields, L. M.,  C. Mitchell and F. Druuet, Alga- and Lichen-Stabilized
     Surface Crusts as Soil Nitrogen Sources.  Am. J. of Botany 44:484-498, 1957.

74.  	, and P. V. Wells, Effects of Nuclear Testing on Desert Vegetation.
     Science 135:38-40, 1962.

75.  	, and 	, Recovery of Vegetation on Atomic Target Areas
     at the Nevada Test Site.  In (Schultz, V., and A. W. Klement, Jr., eds.)
     Radioecology, Reinhold Publ. Company, N.Y., N.Y., pp 307-310, 1963.

76.  	, 	, and W. H. Rickard, Vegetational Recovery on Atomic
     Target Areas in Nevada. Ecology 44:697-705, 1963.

77.  Shreve, F., Changes in Desert Vegetation.  Ecology 10:364-373, 1929.

78.  	, The Desert Vegetation of North America. Botanical Review 8(4):
     195-246, 1942.

79.  	, Ecological Aspects of the Deserts of California.  Ecology 6(2):
     93-103, 1925. '

80.  	, The Vegetation of a Desert Mountain Range as Conditioned by
     Climatic Factors.  Carnegie Inst. Wash.  Publ. No. 217, 1915.


                                      1-64

-------
 81.  	, and A.  L. Hiclley, Thirty Years of Change  in Desert Vegetation.
     Ecology  18:463-478,  1937.

 82.  	, and I.  L. Wiggins, Vegetation and Flora of the Sonoran Desert.
     Palo Alto, Stanford  University Press, 1964.

 83.  Soil Survey Staff,  Soil Classification, a Comprehensive System — 7th
     Approximation.  U.S. Department of Agriculture, 1960.

 84.  	, Supplement to Soil Classification, a Comprehensive System —
     7th Approximation.   U.S. Department of Agriculture, 1964.

 85.  	, Supplement to Soil Classification, a Comprehensive System —
     7th Approximation.   U.S. Department of Agriculture, 1967.

 86.  Stebbins,  R. C. and  N. W. Cohen, Off-Road Menace—A Survey of Damage in
     California.  Sierra  Club Bulletin, July/August 1976, pp 33-37.

 87.  Stewart, O.C., Fire  as the First Great Force Employed by Man.  In Man's
     Role in Changing the Face of the Earth.  Chicago: University of Chicago
     Press, 1956.

 88.  Tenney, H.W., Variations of Atmospheric Temperature with Altitude in the
     United States.  Electrical Engineering 60(5):230-232, 1941.

 89.  Tevis, L. Jr., A Population of Desert Ephemerals Germinated by Less than
     One Inch of Rain.   Ecology 39:688-695, 1958.

 90.  Thames, J. L., ed., Reclamation and Use of Disturbed Land in the Southwest.
     Tucson, Arizona.   University of Arizona Press, 1977.

 91.  Thornthwaite, W.,  An Approach Towards a Rational Classification of Climate.
     Geog.  Rev., 138:55-94, 1948.

 92.  Van Devender, T.  R., Holocene Woodlands in the Southwestern Deserts.
     Science 198:189-192, 1977.

 93.  Wallace,  A. E., Water Use in a Glasshouse by Salsola Kali Grown at Different
     Soil Temperatures  and at Limiting Soil Moisture.   Soil Science 110:146-149,
     1970.

94.  	, W. A.  Rhoads,  and E.  F.  Frolich, Germination Behavior of Salsola
     as Influenced by Temperature, Moisture,  Depth of Planting,  and Gamma
     Irradiation.   Agronomy J.  60:76-78,  1968.

95.  	, and E.  M. Romney, Feasibility and Alternate Procedures for
     Decontamination and Post-Treatment Management of Pu-Contaminated Areas in
     Nevada.  In The Radioecology of Plutonium and Other Transuranics in Desert
     Environments,  U.S.  Energy Research and Development Administration, NVO-153,
     1975.
                                      1-65

-------
96.  	,  and 	,  Radioecology and Ecophysiology of Desert Plants
     at the Nevada Test Site.  USAEC Publication TID-25954, 1972.

97.  	,  E. F. Frolich, and G. V. Alexander, Effect of Steam Sterilization
     of Soil on Two Desert Plant  Species.  Plant and Soil 39: 453-456, 1973.

98.  Weaver, J. E. and F. E. Clements, Plant Ecology.  New York, McGraw-Hill,
     2nd Ed., 1938.

99.  Wells, P. V., Succession in  Desert Vegetation on Streets of a Nevada Ghost
     Town.  Science 134:670-671,  1961.

100. Willard, E.E. and C.M. McKell, Simulated Grazing Management Systems in
     Relation to Short Growth Responses.  J. Range Management 26:171-174, 1973.
                                      1-66

-------
                                   CHAPTER  2
                                    PRAIRIE
2.1  OVERVIEW
     Before overgrazing and cultivation of the 1870's broke the cover, the
native grasslands covered between 2.6 and 2.9 million square kilometers of the
                         40
contiguous United States,   or nearly one-third of the land surface.  As a
general rule, grasslands occurred in areas having between 25 and 75 cm annual
              22
precipitation.    Less precipitation usually resulted in desert and more
resulted in forest; exceptions occurred, however.  Tallgrass prairie extended
as far east as Ohio (up to 100 cm mean annual precipitation),  at least partially
the result of recurring fires, and grasses formed the understory of desert shrub
communities having as little as 18 cm precipitation per year.

     Grassland communities,  as the name implies, are dominated by grasses and
typically contain little woody vegetation.  The grassland growth form is a
single, uniform layer of vegetation which does not provide for the variety of
                                                         48
niches found in more diverse communities, such as forests.     Consequently,
the diversity of birds and insects, which can take advantage of vertical strati-
fication, is reduced.   Grazing mammals predominate and range from large, running
herbivores, such as bison, to a considerable diversity of herbivorous and
granivorous rodents.

     The term prairie is used in this report to mean midcontinent, perennial
grassland.  Kiichler   has identified approximately 15 grassland vegetation types
within this region.  There is a general increase in cover from west to east
which roughly corresponds to rainfall distribution.

2.1.1  Tall grass Prairie
     Tallgrass prairie consists of grasses 150 to 250 cm in height with roots
                                     22
extending to depths of 180 cm or more.     The dominant Andropogon-Panicum-
Sorghastrum association was  continuous in the 18th century from Canada to
                                      2-1

-------
Oklahoma with tallgrass islands in Wisconsin and east to Ohio, Kentucky and
Tennessee.    The island prairies were undoubtedly successional stages, for fire
prevention in those areas invariably resulted in an increase in woody vegetation
Because of its great agricultural value, little tallgrass prairie remains.  One
remaining tract, described by Risser,^! will serve to illustrate a typical
association of climatic, soil, and biological characteristics.  Mean annual
precipitation was 93 cm with 63 cm falling during the 205 day growing season.
Mean January temperature was 2.8°C and mean July temperature was 27.8°C.  Soils
were Brunizems with the A horizon extending to 40 cm and the B horizon to over
100 cm.  The list of common plants included the tallgrasses and forbs but did
not mention woody plants.  The list of common animals was comprised primarily
of herbivorous rodents and insects.

2.1.2  Mixedqrass Prairie
     The mixed or midgrass prairies are found to the west of the tallgrass
prairies in zones of generally lower precipitation.  Midgrass prairies average
60 to 120 cm in height.  They occur from Canada to northern Texas and the cli-
matic differences yield a variety of plant associations over the range.  Mid-
grasses predominate but both shortgrasses and tallgrasses may be important
                  19
components.  Lewis   described midgrass prairie in western South Dakota.  The
midgrasses Agropyron smithii and Stipa virdula predominated under relief from
grazing but there was an understory of shortgrasses, especially Bouteloua
gracilis and Buchloe dactyloides.   Increased grazing pressure led to a reduc-
tion of midgrasses and the formation of shortgrass sod.   Mean annual precipi-
tation was 38 cm of which 30 cm fell during a 126-day growing season.  Ten
year (1958-1967) average temperatures were -1°C (January) and 24°C (July).
Annual potential evaporation was high,  about 140 cm from a free standing water
surface.

                                       39
    In a mixed grass prairie in Kansas,   tallgrasses (Andropogon gerardi,
Sorghastrum nutans), midgrasses (Andropogon scoparius),  and shortgrasses
(Bouteloua curtipendula, B.  gracilis, and B.  hirsuta)  all occurred in ungrazed
areas.   Grazing increased the frequency of shortgrasses.  Annual precipitation
was 58 cm with a 165-day growing season annually.
                                     2-2

-------
 2.1.3  Shortgrass Prairie
      Shortgrass prairies average 15 to 45 cm in height and occur from Canada
 to southeast New Mexico and west Texas.   A northern shortgrass prairie described
 by Jameson   had Bouteloua gracilis and Buchloe dactyloides as dominant plant
 species.   Annual precipitation ranged from 25 to 38 cm with 80 percent occurring
 during the summer.   Mean January temperature was 4°C and mean July temperature
 was 29°C.   Soils (A + B horizons)  ranged from 35 to 76 cm deep.   Huddleston14
 described  a shortgrass prairie near Amarillo, Texas.   Bouteloua gracilis pre-
 dominated  with various other shortgrasses present.   Prickly pear cactus (Opuntia
 polycantha)  was abundant with kochia (Kochia scoparia)  occurring in disturbed
 areas.  Mean annual  precipitation  was 53 cm but pan evaporation was nearly
 225 cm per year.

 2.1.4  Other Grasslands
      Outside of the  midcontinent plains,  major  grasslands  occurred  in  the  in-
 terior of  California,  as understory to conifers or  as  openings  in the  western
 mountains,  and as understory to  arid-land shrubs in the  interior northwest
 and southwest.

      Kuchler   considered  the  native  vegetation of  the "California  steppe" to
 be  perennial  grasses  of the  genus  Stipa.   Today,  these grasslands plus  portions
 of  the  surrounding,  former oak woodlands    are  dominated by  introduced  annual
 grasses, including Avena,  Bromus,  and Festuca?0  Climatic  data for  the  California
 Agricultural  Experiment  Station's  Hopland  Field  Station indicate great  varia-
 bility  from  year to  year.    Mean  annual  precipitation for a ten-year period
 was  94  cm, but  the range was 63 to more than  150  cm.  Summer droughts average
 92  days in length and there  are an average  70 days  with frost each  year.
 Although the  five unfavorable months divide the  year into two separate  growing
 seasons, the  dominance  of annuals provides  rapid  response to community  perturba-
 tion.

     The palouse prairie  (shrub steppe) of the Columbia plateau is dominated
 by Artemesia  tridentata, a shrub, and Agropyron spicatum, a bunchgrass.  The
 area is characterized by hot, dry summers and cool,  moist winters.29  Annual
precipitation is 25 cm or less.  Desert grasslands are superficially similar,
                                     2-3

-------
 being associations of arid-land shrubs and grass understories.  Desert grass-
                                 1 O
 land near Las Cruces, New Mexico1^ is characterized by a black gramma climax
 (Bouteloua eriopoda) with various combinations of desert shrubs, including
 creosote bush (Larrea divaricata)  and honey mesquite (Prosopis juliflora).9
 Mean annual precipitation is 23 cm with 55% coming between July 1 and
 September 30.  Characteristic of desert regions, however, is wide variation
 and unpredictability of annual rains.21  Several successive dry years reduces
 perennial grass cover while a series of wet years can have the opposite effect.
 Pan evaporation is approximately 236 cm annually and mean maximum temperatures
 range from 13°C in January to 36°C in June.

      The vegetation of mountain grasslands varies with numerous factors,  such
 as latitude,  altitude,  slope and aspect.   Van Dyne40  estimated a total area of
           2       ..       17
 267,731  km  from Kiichler's   data,  including  that which occurs as understory
                                       20 6
 to ponderosa pine.   Site descriptions  '   indicate annual  precipitation  in
 the range of 33  cm and  mean annual  temperatures  in the  range of 7°C.   Soils
 (A + B horizons)  may be fairly deep (60 to 110 cm)  with the grasses  being
 rooted to the full  depth.   Agropyron  and Festuca are  the dominant grasses but
 many other genera also  are widespread.

      Many of  the  proposed  cleanup procedures  are qualitatively similar to pro-
 cesses,  either natural  or  man-induced,  which  have gone  on previously  in grass-
 lands.   These have  been  divided  into  natural  and man-made perturbations,  those
 activities  which  might  be  considered  disruptive,  and  range  improvement  measures.

 2.2  NATURAL  PERTURBATIONS

 2.2.1  Drought
     The distribution of plant communities  in  the central United  States in
general  conforms  to the distribution  of rainfall.  Any  long-term  changes  in the
distributional pattern of rainfall would be expected to produce a general shift
in the distribution of plant communities.   There  is evidence that such  shifts
did, in  fact, occur in the past.    A broad prairie "peninsula" extended  into
Ohio during a warm, dry period following the last glaciation, but that prairie
was reduced to islands surrounded by forests by the time the first white  settlers
reached the region.  Drought of a lesser extent is characteristic of the prairies
                                     2-4

-------
 and, if of sufficient duration, affects the vegetational distribution.
 Buffington and Herbel5  reported that in 1858 over 90 percent of the Jornada
 Experimental Range in New Mexico was "good grass" while by 1963 less than 25
 percent was.  Located in a Bouteloua-Hilaria-Larrea association,17 a slight
 shift in climate, with continued grazing, resulted in reduced grass cover and
 an increase in desert shrubs.12

      The drought of 1934-1936 permitted the invasion of shortgrass and midgrass
 species into the tallgrass prairie zone.   Weaver and Bruner40 reported on an
 Andropogon-Poa pasture which was converted to Sporobolus and Bouteloua.  Only
 8  years were needed for recovery,  however,  once the drought had ended.

 2.2.2  Flooding

      The effects  of flooding  are exactly  opposite to  those  of drought.   Periodic
 floods,  as  well  as  ground  water percolation,  produced flood plain  forests well
 into  the shortgrass prairies  of Montana and  Colorado.17

      Scifres  and  Mutz    reported on  an area  of  coastal rangeland in Texas which
 undergoes periodic  flooding.   The native vegetation is mesquite-acacia  savanna
 in which the  dominant  grass is Setaria.  Low-lying areas  (lagunas) may  become
 inundated with fresh water for as long as 5 to  10 years as  a  result of  tropical
 storms.  As the lagunas dry,  a mesic sere is established which results  in mid-
 grass prairie, often dominated by spike dropseed  (Sporobolus  contractus).
 Grazing capacity  is doubled,  relative to pre-flood stages, but long-term
 grazing results in reversion  to  dominance by Setaria.

 2.2.3  Fossorial  Animals

     Fossorial (burrowing)  animals are characteristic of grasslands.   They include
ants and termites, as well  as the more conspicuous vertebrates.  They all have
the net short term effect of  setting back  the development of both grassland
plant communities and soils.   This is accomplished through both the direct
consumption of vegetation and the burial  of vegetation with soil brought up
from depths of two or three meters or more.16
                                      2-5

-------
     Abandoned prairie dog towns have been valuable in the study of local succes-
                                          23
sional patterns (seres).  Osborn and Allan   reported on succession following
the decline of a prairie dog town in the Wichita Mountains of Oklahoma.  As
the colony declined over a 9-year period, it progressively abandoned the outer
perimeters of the town until there remained an 8-step representation of the
local sere in more or less concentric rings.  In the central, active portion of
the town, there remained barren soil with a few forbs.  On that portion which had
been abandoned the previous year grew an annual species of threeawn (Aristida).
The next stage was also annual threeawn and the next a mixture of annual three-
awn and forbs, indicating that the local sere requires three or more years in
the annual grass stage.  The next stage contained perennial grasses (Chloris,
Sporobolus) as well as threeawn.  That was followed by a shortgrass stage domi-
nated by blue gramma (Bouteloua gracilis) and buffalo grass (Buchloe dactyloides)
Next came what the authors called a subclimax midgrass stage, characterized by
silver bluestem (Andropogon saccharoides) and sideoats gamma (Bouteloua curti-
pendula).  The climax vegetation surrounding the abandoned prairie dog town was
a tallgrass prairie dominated by Panicum, Andropogon and Leptoloma species.

2.3  MAN-MADE PERTURBATIONS

2.3.1   Close Cropping and Grazing
     The nonselective removal of grasses, such as by mowing, produces results
which differ considerably from selective removal through grazing.  When species
composition and growing season biomass were studied in an Andropogon-Panicum-
                                                    24
Sorghastrum tallgrass prairie near Norman, Oklahoma,   it was found that denuda-
tion (close cropping and removal of mulch) had no significant effect on species
composition when compared to the control.  Net primary production (NPP, biomass
produced during a single growing season) was greater in the denuded plots than
in controls, however.  Slightly different results were reported by Weaver and
       42
Rowland   for tallgrass prairie in Nebraska.  They compared the prairie growing
on an abandoned dirt road with adjacent grazed or mowed fields.  Again they
found that yields (NPP) were 25 percent to 50 percent less in the climax prairie
growing in the abandoned roadway than they were in the grazed or mowed fields.
However, they found a relative increase in the frequency of occurrence of
                                      2-6

-------
  Andropogon in  the  roadway.   Since  mowing  is  nonselective,  they attributed the
  increase  in Andropogon  to  accumulation  of natural  mulch  (see  Section 2.3).

       Grazing animals  tend  to  select  the plant  species  they consider  to  be most
  palatable  with  the result  being that  intense grazing usually  produces a shift
  in  species  abundance.   In  mixed grass prairies, grazing  can result in shortgrass
  species ^replacing midgrass species,19'38  and in a  general  decrease in canopy
  height.     In more arid regions, intense  grazing pressure  favors the establish-
  ment  of desert  shrubs  and can interfere  with reclamation  attempts.13

  2.3.2  Compaction

      The short-term effects of compaction are all negative.  The primary effect
  is to increase bulk density of the soil (weight per unit volume) which decreases
  infiltration of water, reduces water storage capacity,  reduces aeration, inhibits
 root penetration and  restricts movement of soil animals.26  These effects are
 most severe in  moderately wet soils and in those with a higher clay,  but they
 can be significant  in  coarse-textured,  arid-land soils  as well?

      The effects of compaction are  not  irreversible,  however.   Stipa-bunchgrass
 stands may be found growing on the  streets of a Nevada  ghost town44 and  climax
 tallgrass  prairie was  reported to have  been established 15  years after the
 abandonment  of  a dirt  road  in  Nebraska.42   Factors  affecting the rate of recovery
 from compaction  include  soil  type,  climate and  presence or  absence  of surface
 sealants.

 2.3.3   Plowing

     The effects  of plowing are in many  ways similar to the activity  of  prairie
 dogs.  Vegetation is killed and soil is  turned over.  The major  difference lies
 in the fact  that  deep,  organic-poor soil is not deposited on the surface by
plowing.  Most of our understanding of plant succession has come from the  study
of abandoned agricultural land, most of which had been plowed for one or more
years.

2-3-4  Natural Succession
     Succession in tallgrass prairie can be extremely rapid, especially if a
suitable seed source is readily available.   Rice and Penfound27 reported  an
                                     2-7

-------
experiment carried out in an Andropogon-Sorghastrum-Panicum prairie in Oklahoma.
In the first year after plowing, Ambrosia psilostachya (a ragweed) was the most
important dominant.  By the second year, the dominants in the control had
replaced the ragweed.  The authors concluded that "succession back to the
prairie stage was, therefore, extremely rapid under the conditions of the
experiment."

     The fruits of grasses tend to be windborne and the fruits of pioneer species
tend to be carried further than those of climax species.   Consequently, increasing
the size of the disturbance tends to lengthen the time required for succession.
             28
Rice, et al.,   indicated that the fruits of triple awn (Aristida oligantha)
are carried great distances and triple awn is typically the first invader in
disturbed tallgrass prairie.  Little bluestem (Andropogon scoparius) fruits
are windborne only about 2 meters and thus may take several generations to
reach the center of a large old field.
                                                                           25
     A typical tallgrass prairie succession was given by Perino and Risser.
Following abandonment of a cultivated field in Oklahoma,  a weed stage is
established which persists for two or three years.  This pioneer stage is charac-
terized by Ambrosia, Sorghum, Bromus tectorum, and Rumex.  The weed stage is
followed by an annual grass stage lasting 9 to 13 years and is dominated by
Aristida oligantha, Aster sp., Ambrosia sp., and Andropogon saccharoides.
A perennial bunchgrass stage dominated by Andropogon scoparius follows and can
last 30 years or more.  Tallgrass prairie climax becomes established after 40
to 50 years and consists of Andropogon gerardi, A. Scoparius, and Panicum virgatum
The weed stage is characterized by a predominance of live vegetation (51%) which
declines to 26 percent at climax.  Standing dead increases from 28 percent in the
weed stage to 32.5 percent in the bunchgrass stage and declines to 17 percent
in climax.  Litter increases throughout the sere from 21 percent to 57 percent.
Areal cover declines from the weed stage to the annual grass stage but increases
through the rest of the sere.

-------
  2.3.5  Range Improvement

       2.3.5.1  Mulches--
       Mulches increase  soil  moisture through increasing infiltration of precipi-
  tation  and  reducing  runoff  and  evaporation.   Soil  moistures  and  temperatures  are
  stabilized  and  the conditions for  germination of grass seeds are improved.38
  The  effects of  natural  and  synthetic mulches  are similar,  but effectiveness may
  vary with the grass ^species,    growth  stage of the  grasses,42  and  local  climate.3
  In desert grassland,    an increase  in  soil moisture of 20  percent,  due  to mulch,
  increased forage production by  50 percent.  In climax  tallgrass  prairie, however,
  naturally-occurring mulch was said  to  have reduced yields  by 25  percent  to 50
  percent as  compared to mowed or grazed fields where mulch  did not accumulate.42
  In addition, application of mulch can reduce the frequency of less desirable
  species, such as ragweed, by improving the conditions  for  emergence or growth
 of the desired grass species.  }1

      2.3.5.2  Seeding--
      Seeding of desired grass  species offers several advantages over dependence
 on natural  recovery.   First, it  is  often possible  to bypass succession and seed
 directly to  climax.   Riegel, et  al.  30  seeded  with  climax species in west-central
 Kansas and noted no  differences  in  yields over a  17-year period when compared
 to native grassland.   In other cases,  selective seeding can actually improve
 the range quality.  Rumsey33 reported  forage yield  increases  of up  to 300 per-
 cent  in  7-8  years  in  an  Agropyron/Artemesia  association.   Seeding is not a  simple
 proposition,  however.   Bement, et al.,2  were  more  successful with  introduced
 species  than they were with  native  grammas,  whereas  Terwillinger  and Jensen
 had their greatest success with  the  grammas  and other native  species.   Seeding
 success  or failure depends upon  a number  of  factors, many of  which are  beyond
 direct control.  Terwillinger and Jensen37 found seeding in shortgrass prairie
 to be  successful only when the precipitation to evaporation ratio was 0.3 or
 greater.  Once the seed has germinated, sufficient ground cover must be achieved
 before a periodic dry year to avoid  wind  erosion around the roots.   In drier
 areas, e.g.,  desert grassland and shrub steppe, probability of seeding success
 declines unless extraordinary measures are taken,  such  as water entrapment,
 reduction of competition from woody vegetation, delay of grazing,  mulching! and
rigid  seeding standards.
                                     2-9

-------
     2.3.5.3  Shrub Removal--
     In 1858, over 90 percent of the Jornada Experimental Range in New Mexico
(Desert Grassland) was in good grassland range and by 1963 that figure had
declined to less than 25 percent.    Climatic change may have been partly
responsible but grazing pressure on the grasses certainly gave a competitive
advantage to the shrubs.  Range restoration   included plowing to 9 inches
to kill the shrubs, seeding, and moisture trapping.  In addition to range
restoration, new range has been created from shrubland by clearing of shrubs
                                     50
and seeding with appropriate grasses.

     2.3.5.4  Moisture Trapping--
     Twenty-five centimeters annual precipitation is generally considered to be
the minimum required for successful revegetation of disturbed lands; less
                                                                     g
requires intensive management including irrigation and fertilization.    Re-
clamation and improvement of desert grasslands is further impeded by generally
low infiltration rates so that much of the precipitation received is lost as
runoff.  Seeding success and production of herbage have been improved through
the construction of pits which trap runoff and allow a greater time for infiltra-
tion.  In southern Arizona where annual rainfall averages 15 to 20 cm, moisture
penetration averaged twice as deep and herbage production was five times greater
when pits were employed as compared to adjacent flats.    In a Utah salt desert,
             43
Wein and West   found that moisture entrapment was essential to survival of
crested wheatgrass (Agropyron cristatum) seedlings even if there had been suf-
ficient rainfall to produce germination without entrapment.
2.4  EFFECTS OF CLEANUP PROCEDURES ON PRAIRIE
     This section describes the effects that imposition of certain contamina-
tion cleanup treatments produce on prairie ecosystems.  Qualitative descrip-
tions of the effects caused by each cleanup treatment are related in this sec-
tion; the quantitative assessment of the impacts of cleanup on prairie ecosys-
tems is given in Section 2.6 with estimates of the relative times for a return
to the native ecosystem.  Recovery time projections are based upon the presump-
tion that, if man were restrained from interfering, plant succession would go
                                       2-10

-------
 forward under the vegetative potential at the site and climate to recreate
 a prairie.  Reclamation procedures and other "management" of prairie sites
 is discussed in Chapter 8.
      Three groupings of the cleanup procedures can be seen in this section.
 First is cleanup that provides temporary holding of a contaminant in place.
 Second are procedures that remove the physical contaminant and stabilize the
 site until renovation occurs,  and third are alternatives to holding and
 removing the contaminant.

      (0-1) Natural Rehabilitation
      No significant impacts are expected to remain more  than two  or three years
 after manual cleanup.   Annual  grasses will  occupy the shoveled and scooped-out
 excavations during the  next growing season  and that may  be the entire sere.
 No "area"  effect  is anticipated.   The extrapolation is one of a succession of
 "pock-marks."

      (1-1)   Chemioal Stabilization
      When  vegetative cover  is  removed,  soils become subject  to  accelerated
 erosion  by both wind and water.   The  various stabilization practices  may  tend
 to  retard  the initial stages of revegetation.  Chemical  stabilizers form  a less
 permeable  surface  layer which reduces infiltration  of water  and may also  prove
 to  be a  physical barrier to seed  germination.  Further, dark-colored  emulsions
 may increase soil  temperatures  above  the range of tolerance  for germinating
 seeds.   With time, however, cracks  tend to develop  in stabilized areas which
 both allow  seeds to germinate and  channel runoff to seeds sprouting in the
 cracks.  As natural rainfall decreases, cracks in soil sealants can provide
 micro-habitats which support relatively vigorous growth.   In prairies, however,
 sufficient moisture to permit revegetation is usually available and stabiliza-
 tion techniques which retard revegetation must be viewed as negative, long-term
 impacts.
     The properties of two groups of soil additives, 1)  chemical stabilizers
 and 2) chemical stabilizers with mechanical properties,  are given  in Appendix
A,  Stabilizers.   Eleven of the chemical stabilizers are rated "preferred" for
                                     2-11

-------
prairie use, 10 of these require post-application site renovation for plant
regrowth; 7 of the 25 "preferred" chemical stabilizers with mechanical proper-
ties permit immediate natural regrowth of vegetation.   All eight of the pre-
ferred stabilizers are rated to break down in a year or less after application.

      (1-2)  Clear Cutting Vegetation
     Clear cutting of grasses  (mowing) and removal is a common practice with
well-documented results (Section 2.3).  Since the process is nonselective,
there are virtually no effects on species composition and, in the short term,
productivity is actually increased.  On shrub-free grasslands, effects of
mowing and removal would probably not be discernible with a single growing
season.  The presence of woody shrubs, however, increases the impact of clear
cutting.  Woody plants have their meristematic (growth) tissues located at the
tips of branches as opposed to at ground level in grasses.  Clear cutting thus
removes the primary growth tissues resulting in longer recovery times.  Most
woody plants do have the ability to form stem tissue at the roots (root-sprouting)
so that clear cutting may have no profound effect on long-term community compo-
sition.  A one-time clear-cut would allow woody plants to resprout using food
reserves stored in the roots but, as farmers well know, repeated cuttings
deplete stored reserves and eliminate the woody component.

     (1-3) Stumping and Grubbing
     "Stumping and grubbing" of grasses, that is, the removal of a major portion
of the root system, is not possible, owing to the diffuse nature of their root
systems.  The procedure will prevent root sprouting of shrubs and will prolong
the period needed for complete recovery.  Since reestablishment of shrubs would
then depend on reseeding,  the size of the area treated would be proportional
to recovery rate.  A secondary impact would be the effect of the treatment on
the area's attractiveness to wildlife.  Grasses, generally, produce a single
vertical stratum which limits the number of niches available to bird species.
As shrubs and trees are added to grasslands, the number of strata increases
resulting in increases in bird species diversity.  Removal of shrubs, trees,
or both could produce losses of bird species lasting decades or more and could
                                      2-12

-------
 possibly lead to the eradication of species already faced with limited habitat.
 A detailed discussion is deferred to Chapter 10.
      Positive results of shrub removal would include improvement of pasture
 for cattle.  Following the discussion above, it is apparent that improvements
 would persist longer as size and severity of the treatment are increased.
 Further, seeding success of shrubs would initially be reduced by the presence
 of the grasses but, once established,  shrubs are successful competitors of
 grasses.  The presence of cattle yields a favorable competitive balance to
 shrubs since cattle will selectively remove grasses,  when they are available.

      (1-4)  Scraping and Grading
      Surface scraping and grading of the upper  5 cm of grass roots and soil
 could  reasonably be  expected to have only  a moderate  effect  on  grassland eco-
 systems.   In  the prairies,  ground  irregularities would preserve the  meristematic
 (growth) tissues of  a portion of the grass  plants  allowing  relatively  rapid
 recovery.   Soil compaction, if  done with crawler tractors,  could possibly have
 a greater  impact than the blading  itself.   In mountain grasslands, shrub steppe
 and desert  grasslands, blading  would remove  the woody component yielding a
 significant change in community  structure.   Release from competition could
 result in an  increase in grass  cover in arid zones with a corresponding
 increase in animals utilizing grasses.  The number of niches would be decreased,
 however, with a resultant displacement of the birds and insects which require
 vertical heterogeneity in a plant community.  Taking the extreme examples,
 0.01 km  and  10 km  of blading,  the lesser disruption would produce effects
probably too slight to measure while the larger could have significant effects.
The significant effects could be both detrimental,  such as long-term loss of
bird species, and beneficial,  such as range improvement for domestic livestock.

     (1-5)  Shallow Plowing
     Shallow plowing (<10 cm)  will not penetrate the sod layer of most grasses
and would probably do no more than break off woody plants, allowing them to
resprout.  It is not likely to be used as a clean-up procedure and if used  is
not likely to have much impact.
                                    2-13

-------
      (1-6)  Deep Plowing
      Deep plowing  (up  to 40  cm) would result  in  complete  removal  of  the  eco-
 system with succession starting from bare  ground.  This is  the perturbation
 that  farming represents and  is the best  known  event  leading to secondary
 succession  (Section 2.3).

      (1-7)  Soil Cover Less  Than 25 cm
      A number of factors can affect the  outcome  of adding soil cover to  the
 community.  With shallow coverage  (<25 cm), quality  of soil  applied, actual
 depth, and plant community would all affect impact level.   The more  shallow
 the soil cover the less disturbance would  occur, generally.   In a mixed
 plant community, woody vegetation would  tend to  be disturbed less than grasses.
 In either case, grasses recover more quickly.  The addition  of porous topsoil
 would have less detrimental  effect than  adding subsoil, since the roots  might
 still receive sufficient aeration.

      (1-8)  Soil Cover 25 to 100 em
     There is probably no difference between adding  25 cm or less soil and up
                                                         2
 to 100 cm.  Even the smallest area to be covered, 0.01 km ,  would require 2500
 3
 m  of soil for coverage to 25 cm.  Expediency would  undoubtedly result in
 coverage with any soil available with the  result that effectively all of the
 vegetation would be killed.  Thus, burial  would  result in a  successional start-
 ing point similar to scraping and grading  (1-4), plowing  (1-5, 1-6)  or soil
 removal (2-1,  2-2).

      (2-1)  Remove Plow Layer
     Soil removal is a radical perturbation to an ecosystem.  Removal of the
plow layer (M5 cm) would have an increasingly serious impact with decreasing
mean annual precipitation.    In tallgrass prairie, the effect could be no more
 serious than plowing,   for the A soil horizon extends well below that depth
 (Section 2.1).   In shortgrass prairie,  removal of the plowed layer could result
 in virtual loss of the A and B soil horizons with long-term  impairment of the
production capability of the soil.
                                      2-14

-------
      (2-2)  Remove Shallow Root Zone
      Removal of the shallow root zone in tall- and midgrass prairies is
 almost a contradiction of terms due to the depths to which grasses are rooted
 (Section 2.1).  Shallower soils, such as some shortgrass prairies, desert grass-
 lands and mountain grasslands, could be removed down to C-horizon material if
 30 cm of soil or more were removed.  This would result in revegetation problems
 analogous to deep cuts along roadways or spoil-bank reclamation.   Comparable
 removals in tallgrass prairie might shorten the projected cultural lifetime
 (i.e.,  number of centuries which the land might be tilled successfully)  but
 the short-term (e.g.,  decades) effect might be indistinguishable  from the effects
 of plow-layer removal.
      In either case,  as with scraping and grading (1-4)  and plowing (1-5,  1-6),
 the complete removal  of vegetation  and exposure of soil  would  lead to predictable
 erosional  losses.   Water erosion potential  is  obviously  greater under the  higher
 precipitations and storm-intensities found  in  the more eastern grasslands.   These
 losses  should not  exceed normal  agricultural  losses on a scale of 0.01 to  0 1
   2                                               9               i
 km  and, depending on  locality,  possibly over  1  km .   At the 10 km  scale,
 however, significant  increases in erosional  losses may be anticipated, for  it
 would be hard to  imagine landscape  units  of  that  magnitude which  did  not  include
 major drainages.   Thus,  complete removal  of  the  soil or  its  retaining vegeta-
 tion would  effectively  prohibit  sedimentation  from occurring prior to entry
 into major  drains.
     Potential  losses from wind  erosion  increase  both  with increases  in the
 size of the  barren  area  and with the  reciprocal of  the square  root  (1//F) of
                      49
 surface soil moisture.    Again  losses would be comparable to  normal  agricultural
 losses when  the area disturbed was similar in size to agricultural  fields.
 Plowed fields are typically broken up with trees planted for windbreaks,  the
 removal of which would increase  soil losses exponentially since the sheltered
 zone is calculated as ten times the height of the trees.
     (2-3)  Remove Scraping and Grading*  Mechanically Stabilize
     Removing less than 5 cm of soil impacts prairie ecosystems less than
removing a shallow plow layer (2-1).  The mulch layer would be taken, with
                                      2-15

-------
a reduced impact on seed stores, soil tilth, and fertility, but wind erosion
effects could be approximately equal.  Wind erosion on bared, prairie soil is
precipitation dependent.  For prairie with low precipitation  (shortgrass) the
length of unobstructed wind run and soil particle aggregation at the exposed
surface become critical factors that determine the site wind erosion potential
in dry seasons of the year.  In tallgrass prairie climates, soil losses by
wind erosion are normally minimal.
     Stabilization is attempted to retard erosion, a desired result, but
carries with it a slowing of plant establishment, an undesired result.  Mechanical
and chemical stabilization perform differently.  Mechanical stabilization
compacts the soil while chemical stabilization binds soil particles together.
Either is accomplished by a number of methods or chemical agents, all with
qualitatively the same results relative to plant establishment and growth.
The properties of stabilizers are evaluated in Appendix A, Stabilizers.
     Few mechanical stabilizers are rated "preferred" for prairie soils in
Appendix A, including deflocculants, clays, and membranes.  Natural and
synthetic rubbers, polyethylene and vinyl membranes, paraffin wax, plus asphalt
emulsions exhaust that list, and only the asphalt emulsion permits regrowth
of plants the next season.
     One popular method of mechanical stabilization depends on soil clay and
physical compaction to make an impervious surface.  This may compress soil
sufficiently in making the physical barrier that it prevents subsequent seed
wetting and seedling emergence and revegetation.  Soil moisture input through
the compacted layer is decreased by two conditions.  Permeability to surface
water is seriously impaired by pore space reduction and a shortening of
infiltration opportunity occurs because of faster runoff from the smoothed
surfaces.   Runoff can readily become 70% of annual precipitation.   The
effects on revegetation, in principle,  are similar among deflocculants,  added
clays,  and direct surface compaction.   Heavily treated areas are toxic to
plants but are likely to be breached in a few months by resident small rodents.
The composite effects of surface soil  compaction are intermittantly visible
in reduced vegetative growth for decades afterwards (Section III).
     The use of meshes and mulches to  reduce erosion are described in Sections 4.4
and 8.4 of Chapters 4 and 8.
                                     2-16

-------
      (2-4)   Remove  Plow Layer,  Mechanically Stabilize
      Revegetation potential  at  the  cleanup  site  will  be  partially reduced
 compared  to  plowing (1-5,  1-6),  by  loss  of  seed  sources,  mulch layer  and
 the  upper fertile A-horizon.  The significant  revegetation  impact is  caused
 from compaction  by  stabilization and  its effect  on  water infiltration and
 seed emergence.  The  smoothed surface allows rapid  runoff with the possibility
 of water  concentration  leading  to piping and gully  development on long over-
 land runs.   The  high  organic matter will make  stabilization more  difficult.
 Physical  retardation  of seedling emergence  limits early  revegetation  to shrub
 crown sprouts.   Gradually  frost  heaving,  aided by localized punctures of  the
 surface by shrubs and small  rodents,  will be effective in opening the surface
 to serious erosion; runoff water erosion in tallgrass and eastern midgrasses
 and  wind  erosion at western mixed grass  and shortgrass sites.   This treatment
 becomes a degraded  version of 2-1.  Chemical stabilization is  preferable  in
 the  eastern  prairie.

      (2-5) Remove  Shallow Root  Zone,  Mechanically Stabilize
      Treatment 2-2  has  defined the  ecological effects of  removing 40  cm of
 prairie soils and the mechanical effect  of  stabilizing was covered  in  Treat-
 ments 2-3 and 2-4.   This cleanup has  severe  side effects  on shortgrass  soil
 profiles  because most of the B horizon will  be removed and with it much of the
 fertility needed for  revegetation.  On these westside soils renovation  is
 required  before  succession toward grasses will begin.  This is next to  the
 most  severe  treatment on prairie.

      (2-6) Remove  Scraping and Grading,  Chemically Stabilize
      This composite treatment appears similar to treatment 2-3 but is much
 less  drastic.  In some  respects  it has an impact equivalent to windrowing of
 the  surface  which might be done  in  1-4.  While there is a loss of surface
 seed  stores,  the potential for erosion is reduced by chemical stabilization
while revegetation is started.   Breakdown of chemical stabilizers occurs  in
the next growing season if one of the preferred treatments from 1-1 is applied.
Revegetation  should have produced a  crop  of annuals  in the grasses and crown-
 sprouting of shrubs should be well underway.  Because these stabilizers are
 limited to relatively small areas (Appendix A)  natural seeding should suffice.
                                      2-17

-------
     (2-7)  Remove Plow Layer,  Chemically Stabilize
     The effects of removing 10 cm of surface sod present some problems to
cleanup crews.   Using a turf-cutter would be the most likely procedure but
its decontamination effectiveness would be limited by micro-depressions and
contamination inside burrow tunnels.  Plowing was described in Treatment 1-5
and those comments apply.   The  net effect of removing the plow layer to 10 cm
would be quite similar to  Treatment 2-6 just described and have the same
impact on soil fertility.   The impacts would be greater in shallow soils where
the plow layer is deeper than the A horizon.  The prospects for revegetation
would be impaired to a minor degree.

      (2-8) Remove Shallow Root Zone, Chemically Stabilize
     This is a radical treatment because of the removal of the fertile, plant-
supporting medium on many sites and exposure of the C horizon in shortgrass
prairie.  The application of chemical  stabilizer to small areas as wind-erosion
protection can be tolerated but revegetation potential seems similar to 1-6,
starting from bare ground.  Severe wind erosion should be expected in 2 or 3
years.

      (3-1)  Barriers to Exclude People
     Fencing of a contaminated area to restrict movement of man and other
animals has been proposed as an alternative to decontamination, at least temporarily,
Generally speaking, barriers can prevent direct contamination but can be ineffec-
tive in preventing contamination by way of the food web.

      (3-2  Exclude Large Animals)
     Fences designed to exclude cattle and other large grazing mammals will
block one direct link in the food web  leading to man.  The release from grazing
pressure will not necessarily be a desirable impact on the grassland ecosystem,
for the resulting increase in standing dead organic matter and mulch will have
a tendency to reduce primary production.  Large exclusion areas would prove to
be an economic burden to the local population and possibly a detriment to
                                       2-18

-------
 migrating wildlife.   Otherwise,  the species diversity and stability of the
 ecosystems would not be significantly affected and will  not be treated
 further.
      (3-3)   Exclude  Large and Small Animals
      The  effect  of barriers  to exclude  small  mammals  (rodents,  etc.)  will  be
 dependent,  at  least  in  part,  on the size  of the  area  fenced.   Small  areas   on
                     2
 the  order of 0.01  km ,  may prove to be  too  small  to support viable populations.
 The  resultant  release from rodent,  as well  as large mammal, grazing  pressure
 could possibly begin to affect  the  rate of  return of  nutrients  to the  soil  and
 overall community  stability.   Such  a small  area  is not  likely to be  unique  and,
 if not, the  loss cannot  be considered to  be significant.   Large areas, on the
 order of  10  km , would  contain  self-supporting populations of small  mammals
 and  would be largely unaffected  by  barrier  erection.  Gene flow and  genetic
 plasticity could become  affected, in time, but insufficient data exist to
 evaluate this  effect.
     Neither birds nor insects can be effectively controlled by barriers.  In
the case of upland game birds, such as ring-neck pheasants, feeding in the
contaminated area could lead to direct contamination of man.  However, insects
are sufficiently small and far removed from the human food web, so unless the
contaminant is extremely mobile, contamination by way of insects is unlikely,
although some insect vectored contaminents might prove toxic to migratory birds


     (4-1)   Asphalt Hard-Surface Stabilization
     Decomposition of 3-inch-thick, unsupported hot mix asphalt requires more
than 10 years in the midgrass prairie on sandy soils when trampling by cattle
is prevented.  In the meantime, reinvasion by stands of grasses is blocked
until the asphalt fractures or the surface cracks.   The retardation of revegeta-
tion is absolute but if soil moisture conditions were right at the time of
surfacing cold weather frost heaving can fracture the asphalt  enough to allow
seed penetration, and germination and growth in the next season.
                                       2-19

-------
     Where shrubs intermingle with grasses the likelihood of puncture by root-
or crown-sprouts is high.  Holes made by the sprouts and the new shrubs act as
collection points for debris, weed and grass seeds, and soil.  Seeds germinating
in the drift piles speed breakup and weathering of the asphalt around the
punctures.  Those local species that normally reproduce from seed should remain
at these breaks.
     Soil compaction by the tank trucks and heavy trailers needed for asphalt
surfacing would be evident more than a century later as demonstrated by bison
wallows and trails.  Hydrological impacts generated by the hard surface would
equal those described for Treatments 2-3, 2-4, and 2-5, but the site prepara-
tion impacts are greater than those treatments cause.
     The gross effect of the asphalt hard surfacing is to delay for a decade or
more, beyond the delay caused by Treatment 2-5, the entry of the site into the
prairie sere.  The treatment might be tolerated in shortgrass prairie as a
secure, temporary cover to minimize redistribution of contamination, the
certainty of breakup in a few years makes that use highly questionable.   The
potential for gully creation in the relatively mesic tallgrass prairie precludes
use there anytime.
     (4-2)   Concrete Hard-Surface Stabilization
     Covering a site with concrete deals the ultimate impact among those cleanup
treatments proposed.  All the disadvantages of asphalt surfacing plus the several
times greater durability of concrete chunks and blocks make later uses contingent
on removal of the concrete.  This would be exceedingly expensive in either
transportation costs or onsite burial.
                                      2-20

-------
  2.5  RECOVERY AFTER CLEANUP

  2.5.1   Irreversible Changes

      No  irreversible  changes  are  anticipated with  scraping  and  grading, plowing,
  vegetation removal  or flooding.   Each  falls within the range of analogous
  perturbations  (discussed in Sections 2.2 and 2.3 for which  successional informa-
  tion exists.  Whether or not  irreversible changes  (i.e., failure to reestablish
  a similar community within  25 years or more) would occur as the result of soil
  removal or burial would depend upon local conditions and the extent of per-
  turbation.  In most cases, no irreversible changes would be anticipated for
  removal of the plow layer of soil or if suitable soil were used for burial.
  In more extreme cases, such as on shallow mountain soils, in the most arid
 margins of grassland distribution or where burial was with particularly low
 quality material, extraordinary reclamation procedures might be required for
 successful reestablishment of vegetation.   These might include the land manage-
 ment practices (Chapter 8)  of water entrapment,  irrigation or addition of soil
 amendments.   It should be noted,  however,  that  reestablishment  of  vegetation
 on a site does not  imply lack  of  irreversible  changes,  nor does  irreversible
 change  imply  complete  loss  of  a resource.   The  change may lower  productivity
 and  prevent recovery of climax by  establishing  an entirely different  flora than
 the  displaced  grasses.

 2.5.2   Rates of  Recovery  Following Cleanup
     Rates of  recovery are  affected by  several factors, including severity of
 perturbation,  proximity to  seed source, local climate and soil conditions.   In
 general, recovery will  be most rapid the less severe the perturbation, the
 smaller the area involved and the  greater the normal rainfall.   In the dis-
 cussion that follows,  certain caveats are appropriate.  Conclusions will be
 drawn for major plant  associations such as tallgrass prairie, which may cover
 on the order of 10  km  .  The data supporting those conclusions are drawn from
 studies which represent small samples from within the larger association and
typically do not begin to cover the range of possible seres.  In a  few instances,
appropriate seres simply have not  been documented and it has been necessary to
extrapolate from the most similar  situations that could be identified.   It  is,
                                       2-21

-------
therefore, assumed that this analysis will be useful in formulating policy
but that field-level decisions will require site-specific information
     Idealized trend curves were developed as guides to illustrate the time
course of recovery to full productivity for some of the prairie ecosystem
parameters.  The ordinates are without units either because they have no
dimensions (e.g., homostasis) or they are broad generalizations over the
array of heterogeneous prairie sites.  Representative cleanup treatments were
selected for the illustrations.   Assisted and unassisted recovery are shown in
Figure 2-1; recoveries of grasses and woody elements are illustrated in
Figure 2-2, and in Figure 2-3 the maximum contrasts in impact effects are shown,

     2.5.2.1  First Year  Following Cleanup--
     With at least two cleanup procedures, mowing and fencing (clear cutting
and barriers), prairies should attain complete recovery within one year.
        24
Penfound   reported that denudation resulted in no significant change in
species composition and actually resulted in an increase in biomass.  Since
"stumping and grubbing" are inappropriate for prairies, denudation would
consist only of mowing and removal of biomass.  Complete recovery can, there-
fore, be assumed after one year.  Size of area mowed would have no effect.
     If a suitable seedbed remains after cleanup, prairies may be seeded
directly to climax,   thus providing for near recovery the first year.  This
would include scraping and grading, plowing, and most instances of soil re-
moval where some A horizon is left as a seed bed.  The addition of nitrogen
                                                     32
fertilizer can aid establishment of perennial grasses   especially if suffi-
                          9
cient moisture is present.   Size of area is unimportant.
     Assuming a suitable seedbed remains but with reliance on natural revegeta-
                                                                           00
tion, then the size of the area disturbed becomes important.  Rice, et al., °
reported that the seeds of a climax species were windborne only about 2 m while
the seeds of an early invader were carried a great distance.  Thus, assuming
seed sources completely surround the disturbed area and no differences exist
due to prevailing wind direction, in one year seeds of the climax species might
                            2
reach 7 percent of a 0.01 km  circular plot (the configuration for maximum
potential invasion under the stated conditions).  The area invaded declines to
                                       2-22

-------
                                                                        HYPOTHETICAL PR AW IE
ro
 i
ro
CA>

-n
(O
o>
•^ w
I— •
o z
n- o
TJ C
32
-s*
-*• -5
CD CD
• O
1 -
•s
	
,
1
— A
1
— 1
1
1
1
I



i
~\~
\
N
\
N
\
\
\
\
\
1


, 	 i
I
V
N
\
1— 1
1
	 1

J
1
i
\
V
— \
N
N
N
\
\
\
\
\
\
t
1

1
1
\
\
\
\
\
\
\
N
\
\
1
/
QL

-J
n>
o
o
<
fD
           -J
           fD
           to
           n>
                            ERODABILITY
                                                    HOMEOSTASIS
                                                                           PRIMARY PROD.
                                                                                                       COVER
                                                                                                                           SPECIES DIVERSITY

—

1


1
1
1


f







~ r

^N
\
\
»
1
1
1
1
1
                                                                                                                                                 -<


                                                                                                                                                 >
                                                                                                                                                o
                                                                                                                                                O

                                                                                                                                                m
                                                                                                                                                3}
                                                                                                                                    V)
                                                                                                                                    m
                                                                                                                                    m
                                                                                                                                    O

                                                                                                                                    Z
                                                                                                                                    O

-------
         12 CLEAR CUTTING
                                           1-3 STUMPING & GRUBBING OF
                                              WOODY COMPONENT

       I
    \L
                                          I     /

                                 YEARS
                                                          5-10
Figure 2-2.  Time course of change  after  mowing (1-2) and stumping
             and grubbing  (1-3)  of  woody  component of prairie.
                               2-24

-------
      23
      24 MECHANICAL SOIL STABILIZATION
                                               3-1 BARRIERS (LARGE MAMMALS)
                                       o
                                       V)
                                       UJ

                                       u
                                       UJ


      •
                                      CO
                                      O
                                      UJ


                                      O
                                      I
                                      00
                                      <
                                      Q
                                      O
                                      CC
                                      UJ
                                    YEARS
Figure 2-3.   Recovery of  prairie following mechanical  soil  stabili-
              zation and response to the  erection of  fencing.
                                  2-25

-------
                                                  2      9           2
2 percent, 0.7 percent,  and 0.2 percent for 0.1 km ,  1 km",  and 10 km  circular
plots, respectively.
     In areas not reached by seeds of the climax species, first-year invaders
are annual grasses and "weeds."4   In tallgrass prairie, first-year invaders
include Ambrosia psilostachya27 and Aristida sp23   Russian thistle (Salsola
Xali) is a first-year invader of mixed prairie.16  Launchbaugh18 reported first-
year basal area cover of 20 percent to 60 percent in Texas mixed prairie with
the annual grass Panicum texanum providing 36 percent of the cover.  Other
annuals provided another 10 to 15 percent cover with the reminder due to non-
climax perennials.  First-year invaders of shortgrass prairie include Salsola
kali, Amaranthus retroflexus, and Chenopodium album.
      The  annual weed  stage persists for more than a year, typically.  In an
Oklahoma  old-field studied by  Perino and Risser25 a weed  stage comprised of
Ambrosia,  Sorghum, Bromus, and Rumex persisted  2 to 3 years.  Characteristic of
the weed  stage  is the relatively  large proportion of  community biomass  con-
tained  in the living  vegetation component, 51  percent as  compared  to  26 percent
 in the  tallgrass prairie climax for the  area.
      To summarize first-year  conditions  in prairie natural  successions, there
 is little ground  cover  and that which  is present  is comprised  of plants which
 die  at  the end of the growing season.
      If the cleanup  procedure has left the  surface with a less than suitable
 seedbed,  such as  might  be  the case when soil  is added to the surface, then it
 is entirely possible that  no  significant invasion will  occur during the first
 year.  Stabilization would be required to  prevent erosion and would result in
 still further delays in vegetation establishment.   Improvement of the seedbed,
 such as addition of fertilizer and mulching if precipitation averages less
 than about 7 cm 9 can lead to seeding success.  If the imported soil is of a
 quality to encourage recovery, then it may also contain weed seeds which may
                                                  9
 initially out-compete the local,  desired species.

      2.5.2.2  Fifth  Year  Following Cleanup--
      Weaver  and Bruner41 reported that 8 "normal" years, including 3 consecu-
 tive "good"  years restored a tallgrass climax which had been converted to
                                       L-26

-------
  midgrass by the drought of 1934-1936.  It is assumed that 5 normal years fold-
  ing a particularly wet year, would operate on a similar time scale.
       Under the natural revegetation of circular plots, and assuming the 2 m
  movement of climax species seeds suggested by Rice et al.28  continues each
  year,  nearly one-third of a 0.01 km2 disturbed area would be in climax vege-
  tation.   Only about 1 percent of a 10 km2 disturbed area would be in climax.
      Areas too far from seed  sources to contain  climax species would,  after
  5  years,  remain in an earlier successional  stage.   Osborn and  Allan23 identi-
  fied 7 serai  stages within a  9-year succession on  an  abandoned prairie dog
  town.  The  midpoint,  after about 5  years, was characterized  by a  mixture
  of annual  grasses  (Aristida sp.)  and  subclimax perennial  grasses  (Chloris and
  Sporobolus) .   In another abandoned  prairie dog town,16  subclimax  perennial
  grasses dominated  after 7  years.  Each  area was a  tallgrass  prairie climax.
      After  S years  following plowing of midgrass prairie,    it  too was a
 mixture of  annual grasses  (Aristida] and subclimax perennial grasses
           3nd   droDoon).  Climax grasses had begun to appear  but contributed
 an insignificant amount of the total cover, only 0.2 percent.  Total ground
 cover (basal area) itself, was less than half that of the regional climax.
      Thus, cleanup procedures which leave a suitable seedbed would most
 likely remain in subclimax grassland following S years of natural succession
 Burial with soil could yield  similar results if the quality of the soil added
 is similar.   Reliance on  natural  succession on poorer quality soil material
 would probably result in  slower rates of succession with annual  grasses and
 forbs possibly predominating  after  5 years.

      2.5.2.3   Tenth Year  Following  Cleanup--
      For each  of the  cleanup procedures  which  leaves  a plowed-field-like
 starting point, all prairies should  be in subclimax grassland after 10  years.
Tallgrass prairie regions may still  be dominated by midgrasses and mixed
grass or shortgrass prairies may be  dominated by bunchgrasses.  The 9-year-old
successional stage in the abandoned prairie dog town23 was subclimax mid-
grasses,  silver bluestem (Andropogon saccharoide.O and sideoats gamma (Bouteloua
                                     2-27

-------
curtipendula).   Tallgrass prairie was the regional climax.  A. saccharoides
still comprised 84 percent of cover after 14 years in Texas tallgrass prairie
                              18
and 31 percent after 20 years.    Western wheatgrass (Agropyron smithii),
needle-and-thread (Stipa comata)  and feather bunchgrass (Stipa viridula)
dominated abandoned fields in western North Dakota within 10 years.    The
authors stated that these midgrasses were characterized by low total density
but that they nevertheless produced a relatively high yield of good quality
forage and were valuable hay land.
                                            4
     Not all authors agree, however.  Booth   stated that annual grasses still
predominated after 10 years in southeastern Kansas and east-central Oklahoma.
     Few differences between cleanup procedures would be discernible after 10
years of prairie succession.  Prairie soils are sufficiently deep that scraping
and grading, deep or shallow plowing or removal of soil up to 30 cm would leave
a sufficient seedbed to produce comparable stands within a decade.  If burial
had been with particularly low quality soil, and no other treatments had been
performed, then succession would lag sufficiently that prairies might remain
in the annual grass or grass/forb stage after a decade.

     2.5.2.4  Climax--
     Tallgrass prairie apparently can return to climax very quickly if the
disturbed area is small and climax species are nearby.  Weaver and Rowland^
reported that Andropogon-Sporobolus-Panicum prairie was established on a dirt
                                                        25
road only 15 years after abandonment.  Perino and Risser   reported a somewhat
longer sere.  They described a 4-stage sere starting with (1) a weed stage
(Ambrosia, Sorghum, Bromus, Rumex) lasting 2 to 3 years, (2) an annual grass
stage (Aristida oligantha, Andropogon saccaroides) lasting 9 to 13 years, and
(3) a perennial bunchgrass stage (Andropogon scoparius) lasting 30 years or
more.  Equilibration or climax would not occur, therefore, for 40 to 50 years
following disturbance.  Climax was Andropogon gerardi, A.  scoparius and
Panicum virgatum.
             7
     Costello   described succession in northeastern Colorado mixed grass
prairie as (1)  an annual forb stage (Salsola kali, Amaranthus retroflexus,  Cheno-
podium album) lasting 2 to 5 years, (2) a forb stage lasting 3 to 6 years,
(3) a short-lived perennial grass stage (Schedonnardus paniculatus, Hordeum
                                      2-28

-------
 jubatum> Sporobolus cryptandrus) lasting 4 to 10 years and  (4) an Aristida
 Stage  (Aristida longiseta and A. purpurea) lasting 10 to 20 years.  Transi-
 tion to climax was estimated to take another 10 to 20 years.  Thus, the sere
 could be as short as 15 to as long as 60 years with drought cycles'retarding
 succession and wet years accelerating succession.
                 18
      Launchbaugh   described the sere-of the San Antonio Prairie in Texas.
 Following cultivation,  an annual grass stage was established the first year
 with Panicum texanum being the dominant.   Annuals gave 82 percent to 85 per-
 cent of the cover and subclimax perennials gave  the rest.  No climax species
 were present the first  year and cover (basal area)  was low 20 to 60 percent.
 After 5 years,  annuals  (Aristida supp.)  contributed only 46 percent of the
 cover and subclimax perennials had  increased to  47  percent (Paspalum and
 Andropogon).   Climax species were represented (0.2%)  and  basal area had in-
 creased to  2.1  percent.   After 14 years,  annuals were only 7.2 percent of  the
 plants  present  and  subclimax perennials  (Andropogon saccharoides)  had  increased
 to  84 percent.   Climax  species  (Sporobolus  and Stipa)  had increased  to 5 per-
 cent and basal  area  had  increased to  2.7  percent.   After  20  years,  annuals were
 3.5  percent  and  subclimax  perennials  31 percent.  Climax  species  had increased
 to  60 percent, most  of which  were Andropogon  scoparius.   Basal  area  was  up
 to  3.3  percent.  No  estimate  of  the additional time required to reach  climax
 was  given.  A climax community was described, however, and the  climax  species
 (Andropogon and Bouteloua) were  94 percent of the vegetation and basal area
 was  up  to 4.8 percent.  Both  annuals and subclimax perennials were represented
 in the climax community, 1.3 percent and 1.1 percent respectively.   Whitman,
 et al.,   estimated 40 to 60 years or more were required for succession to
 climax in North Dakota mixed grass prairie.
     A shortgrass prairie in Colorado was still  in the annual grass  (Aristida) '
 stage 21 years after abandonment.  Shantz35 estimated to reach climax in short-
grass (Bouteloua-Buchloe)  prairie it would take  20 to 50 years following plowing,
                                    2-29

-------
2.6  QUANTITATIVE ASSESSMENT OF CLEANUP IMPACTS
     The boundless forests of the eastern U.S. are matched in diversity by
the limitless span of grasslands reaching west up the Rockies from the eastern
forest's edge.  On high western plains removal of 40 cm of soil (2-2) leaves
an alkaline, frequently saline excavation in North Dakota, with 300 to 500 mm
annual precipitation for short and mixed grasses to revegetate the Colorado
plains on intermittantly exposed parent material (A through upper B_ horizons
removed).  On the excavated eastern prairie there may be no noticeable reduc-
tion in soil profile properties, plus 1800 mm of precipitation falls under
which tall grass revegetation will progress.  Assessments of cleanup impacts
and subsequent time to recover productivity are reduced to generalities that
address this range of site diversity.  Over their summer rain, wind, and
unshaded range, the grasses did maintain, with rare exceptions, an almost
permanently dry subsoil below the root zone.  This land has been appropriated
throughout for agriculture.  Recovery is treated in this chapter in the context
.of principles governing an unmanaged ecosystem; agriculture is examined in
Chapter 8.

2.6.1  Impact Assessment
     Mowing prairies (1-2) has no long-term effect.   The slight increase in
impact rating with increased area relates to reduced food availability to
herbivores and reduced cover for wildlife.  Scraping and grading (1-3), and
shallow plowing (1-5) have limited impact when localized but represent an
extension of the trend mentioned for mowing.  Deep plowing (1-6) requires
reinvasion of species from outside of the disturbed area, hence the greater
increase in impact level as area increases.  Soil burial (1-7, 1-8) requires
plant reinvasion and also has greater impact as area increases.  But since the
transported soil added is assumed to provide a seedbed inferior to the soil
covered up, impact is greater than for other treatments requiring reinvasion,
such as deep plowing.  Soil removal  (2-1, 2-2) requires reinvasion also and
since deep plowing may leave seeds to germinate, the removal of seeds with
the soil results in a greater impact.  One intuitively feels that removal of
up to 40 cm of topsoil should be a greater impact than removal of 10 cm.  In
deep prairie soils, however, it is not entirely clear that it  is.  Hence, there
is just a slight increase in impact  level.  Barriers to exclude people would have
                                      2-30

-------
 essentially no impact on prairies.  Trees might invade the eastern prairie
 remnants if man-made fires were excluded, but that is not considered a negative
 impact.  Exclusion of animals by barriers (3-2, 3-3)  should have little long-
 term impact, since the grasses to support grazers would remain.   Animal popula-
 tions are expected to quickly recover when permitted  back in.   Paving represents
 the ultimate loss with the area paved being the only  qualifier.

 2.6.2  Recovery Assessment
      Factors which affect the rate of recovery in prairies have  been discussed
 in Sections IV and V.   Because of the interactions of those factors, quantifica-
 tion of the time  required for recovery (Table 2-1) cannot be precise.   For
 example,  scraping and  grading (1-4)  of grassland may  not  result  in  death of
 significant numbers of grass  plants  and recovery may  be extremely rapid,  two
 or three  years.   However,  the potential exists for all  of the  grass plants to be
 killed  with the subsequent  recovery  taking  as long as recovery from plowing,
 up to 60  years.   Nor can  a  superficially more-destructive,  cleanup  procedure
 be counted  upon to require  longer  recovery  times than a less destructive  one;
 hence,  the  overlapping  ranges  in  time required to  return  to  full  productivity.
      Impact  is also expected  to increase  as area increases,  but with the
 exception of more  rapid revegetation  of smaller  areas,  no  clear relationship
 between area and  time-to-climax is apparent.   Succession  on homogeneously
                            2
 barren  areas of up  to  10 km  has not  been documented  and  extrapolation  from
 smaller areas is dangerous.  The dividing point  in time to recover  seems to
 be between areas  smaller than  0.1 km   and larger than 0.01 km2 where area
 effects become discontinuous.  The discontinuities come with disturbance of
 the seedbed and the general east to west decline in annual precipitation.  The
 impacts increase and the time to recover climax  increases along the east to
west precipitation  gradient.
     Recovery from  toxic constituents  in chemical stabilizers  (1-1)  used for
temporary "paste down" is expected in two or three years.   Mowing (1-2) of
prairie leaves minimum impact and recovery would be expected within the next
growing season.
                                       2-31

-------
                                  Table  2-1
Predicted  recovery
(see Appendix  3).
index  in  prairie  for  cleanup treatments
Treatment
0-1 Natural Rehabilitation

1-1 Chemical Stabilization
1-2 Remove Vegetation by Clear Cutting
1-3 Remove Stumps and Grub Roots
1-4 Surface Scraping and Grading (<5 cm)
1-5 Shallow Plowing (<10 cm)
1-6 Deep Plowing (10 to 40 cm)
1-7 Soil Cover (<25 cm)
1-8 Soil Burial (25 to 100 cm)
2-1 Remove Plow Layer Soil (10 cm)
2-2 Remove Shallow Root Zone Soil (<40 cm)
2-3 Scrape Surface, Remove (<5 cm), Mechanically Stabilize
2-4 Remove Plow Layer (10 cm), Mechanically Stabilize
2-5 Remove Shallow Root Zone (<40 cm), Mechanically Stabilize
2-6 Scrape Surface, Remove (<5 cm), Chemically Stabilize
2-7 Remove Plow Layer (10 cm), Chemically Stabilize
2-8 Remove Shallow Root Zone (40 cm), Chemically Stabilize
3-1 Barriers to Exclude People
3-2 Barriers to Exclude Large Animals
3-3 Barriers to Exclude Large and Small Animals
4-1 Asphalt Hard-Surface Stabilization
4-2 Concrete Hard-Surface Stabilization
5-0 Sewage Sludge Application
6-1 Washing (<3 mm)
6-2 Flooding (3 to 30 cm)
7-0 Soil Amendments Added
Area Disturbed
( km2 )
0.01
0

0
0
—
1
1
2
4
4
2
4
2
3
4
1
2
4
0
0
0
+
+
*
*
*
*
0.1
0

0
0
—
2
4
4
5
5
4
5
3
5
5
2
4
5
0
0
0
+
+
*
*
*
*
1.0
0

0
0
—
2
4
4
5
5
4
5
3
5
5
2
4
5
0
0
0
*
*
*
*
*
*
10.0
0

0
0
	
2
4
4
5
5
4
5
3
5
5
2
4
5
0
0
0
*
*
*
*
*
*
Assumptions
Effects gone first or second year following
manual cleanup.
Effects gone first or second year.

Not feasible here






















rv>

U)
          *0utside scope of work.
          +Not recommended for this  land type.
          The index numbers show the approximate  number of transition stages to reach pre-cleanup conditions.
          methods of indexing in Appendix B, Tables B-2, B-5, B-6 and the related text.
                                               Appparent climax in  ^80 years;

-------
      In considering treatments  1-4 through 2-2.  those procedures which left
 conditions similar to post-agricultural  seedbeds (1-4,  1-5,  1-6, 2-1)  were
 considered to have "typical" successions with the expected upper limit equivalent
 for all treatments and the  expected lower limit  being varied according to per-
 ceived  potential  for minimal impact.   Those which were  considered to yield impact
 in  excess  of  agriculture  (1-7,  1-8,  2-2)  were assigned  potentially longer
 recovery periods.
      The various  mechanical  and chemical  stabilizers  used  following cleanup
 (2-3  through  2-8)  may have  significant impacts on early succession but are not
 thought to be particularly  significant in determining the  ultimate length
 of  the  sere.   That is,  the maximum time  to  recovery may result  from more  basic
 factors such  as soil  erosion and minimums in  precipitation cycles.   The
 initial  effects of stabilization may become lost  before climax  vegetation is
 established.   This  is  not to recommend for  or against stabilization.   Stabiliza-
 tion  decisions should  be made on a case-specific  basis  and based on local
 range management,  soil  science, and road  construction expertise  and experience.
      Treatments 3-1 and 3-2  are essentially "no action" options  and, conse-
 quently, have no recovery time.  This does  not imply  that  there  will be no
 effect  on  ecosystems,  however (see Section  2.4).
     Hard-surface  stabilization  (4-1, 4-2)  represents the  greatest  potential
 impact with the longest recovery times even on the smaller areas  to  0.10  km2.
 Its use  is not recommended,  therefore, and  it is  classed as "prohibited"  on
prairie.
      In general,  the smaller the area and the less radical the cleanup treat-
ment the quicker the recovery becomes.   The variations  in  soils,  climate, and
local mixtures of plant and animal species make the return to full productivity
a highly site-specific process.
                                     2-33

-------
2.7   CONCLUSIONS

     Among the several "preferred" cleanup treatments, in most instances, the
"no action" options (3-1, 3-2) are first choice.  Their advantage is in allow-
ing time for careful consideration of possible courses of action with respect
to specific units of landscape that are contaminated.  Presumably, it would
be possible to conduct field-plot trials to determine what combination of
alternatives would be the most desirable decontamination procedures.  Whether
the barriers should exclude only people or all animals cannot be completely
defined as a generality.  As long as food chains leading to man are isolated
from the contamination, there should be no concern about non-game wildlife
crossing the barriers.  Mowing (1-2) represents the second level of ecological
preference.
     It is recognized that health and safety considerations may require immediate
and effective action.   In such cases, preservation of a seedbed through plowing
(1-5, 1-6) is preferred, especially if the area can be reseeded instead of
relying on natural recovery.  Full recovery with reseeding may require fewer
than five years as opposed to up to sixty to eighty years for bare ground
after soil removal.
     Soil removal (2-1, 2-2) or deep soil cover (1-7, 1-8) should be considered
last-resort options, when followed by reclamation to reduce recovery times.
The choice between these latter options should be made by local professionals
in soils, range management, and highway construction.
     The "prohibited" cleanup treatments in prairie are the variations of hard-
surface stabilization, with asphalt (4-1) or with concrete (4-2).  The 30-year
delays caused by their natural breakup require that they be removed bodily
eventually and reclamation procedures instituted.   Other contamination cleanup
methods are available which are only moderately damaging to this land type.
                                     2-34

-------
  2.8  PRAIRIE  REFERENCES


  1.    Bement,  R.  E.,  D.  F.  Hervey,  A.  C.  Everson,  and L.  0.  Hylton, Jr.   Use
       of Asphalt-Emulsion Mulches to Hasten Grass-Seedling Establishment.
       Journal  of  Range Management,  14:102-109,  1961.

  2.    Bement,  R.  E.,  R.  D.  Barmington,  A.  C.  Everson, L.  0.  Hylton, Jr.,  and
       E.  E.  Remmanga.  Seeding of Abandoned Croplands in  the Central Great
       Plains.   Journal of Range Management, 18:53-58, 1965.

  3.    Beutner,  E.  L.  and D.  Anderson.   The Effect  of  Surface Mulches on  Water
       Conservation and Forage  Production  in Some Semidesert  Grassland Soils,
       Amer.  Soc.  Agron.  J.  35:393-400,  1943.

  4.    Booth, W. E.  Revegetation of Abandoned Fields  in Kansas  and  Oklahoma
       Amer.  J.  Bot. 28:415-422,  1941.

  5.    Buffington,  L.  C.  and  C.  H. Herbel.   Vegetational Changes on  a Semi-desert
       Grassland Range  from  1858  to  1963,  Ecol. Monogr. 35:139-164,  1965.

  6.    Collins, D.   Comprehensive Network  Site Description, Bridger,  US/IBP
       Grassland Biome  Tech  Rep.  No.  38, Colorado State University,  Ft. Collins
       1970,  10 pp.

  7.    Costello, D.  F.  Natural  Revegetation of Abandoned  Plowed Land in the Mixed
       Prairie Association of Northeastern  Colorado, Ecology  25:312-326, 1944.

  8.    Davidson, E.  and M. Fox.   Effects of Off-Road Motorcycle  Activity on Mojave
       Desert Vegetation  and Soil, Madrono  22:381-412, 1974.

  9.    Environmental Studies Board.   Rehabilitation Potential of Western Coal  Lands,
       Cambridge, Mass.,  Ballinger,  1974, 198 pp.

10.    Gleason, H.  A. and A.  Cronquist. The  Natural Geography of Plants.
       Columbia University Press, New York,  1964.

11.    Heady, H. F.  Comprehensive Network  Site Description, HOPLAND, US/IBP
       Grassland Biome Tech.  Rep. No. 42, Colorado State University,  Ft. Collins
       1970, 11 pp.

12.   Herbel, C.  H. and R. D. Pieper.  Comprehensive Network Site Description,
      JORNADA,  US/IBP Grassland  Biome Tech. Rep.  No.  43,  Colorado State University
      Ft. Collins, 1970,  21  pp.
                                      2-35

-------
13.  Herbel, C. H., G. H. Abernathy, C. C. Yarbrough, and D. K. Gardner.  Root-
     plowing and Seeding Arid Rangelands in the Southwest.  J. of Range Manage-
     ment 26:193-197, 1973.

14.  Huddleston, E. W.  Comprehensive Network Site Description, PANTEX, US/IBP
     Grassland Biome Tech. Rep. No. 45, Colorado State University, Ft. Collins,
     1970, 12 pp.

15.  Jameson, D. A.  General Description of the Pawnee Site, US/IBP Grassland
     Biome Tech Rep. No. 1, Colorado State University, Ft. Collins, 1969,
     32 pp.

16.  Koford, C. B.  Prairie Dogs, White faces, and Blue Grams.  Wildlife Monogr.
     3:1-78, 1958.

17.  Kiichler, A. W.  Potential Natural Vegetation of the Conterminous United
     States, Amer. Geog. Soc., New York, Special Publ. No. 36, col. map,
     95x149 cm, and manual to accompany the map, 1964, 116 pp.

18.  Launchbaugh, J. L.  Vegetational Changes in the San Antonio Prairie Asso-
     ciated with Grazing, Retirement from Grazing, and Abandonment from Cultiva-
     tion, Ecol. Monogr. 25:39-57, 1955.

19.  Lewis,  J. K.  Comprehensive Network Site Description, COTTONWOOD, US/IBP
     Grassland Biome Tech. Rep. No. 39, Colorado State University, Ft. Collins,
     1970, 25 pp.

20.  Morris, M. S.  Comprehensive Network Site Description, BISON, US/IBP
     Grassland Biome Tech. Rep. No. 37, Colorado State University, Ft. Collins,
     1970, 23 pp.

21.  Noy-Meir, I.  Desert Ecosystems:  Environment and Producers, Annual Rev.
     Ecol. System 4:25-51, 1973.

22.  Odum, E. P.  Fundamentals of Ecology, Saunders, Philadelphia, 1971.

23.  Osborn, B. and P. F. Allan.  Vegetation of an Abandoned Prairie-Dog Town
     in Tall Grass Prairie, Ecology 30:322-332.

24.  Penfound, W. T.  Effects of Denudation on the Productivity of Grassland,
     Ecology 45: 838-845, 1964.

25.  Perino, J. V. and P. G. Risser.  Some Aspects of Structure and Function
     in Oklahoma Old-Field Succession, B. Tor. Bot. C. 99:233-   , 1972.

26.  Reynolds, H. G. and F. E. Packer.  Effects of Trampling on Soil and Vege-
     tation, U.S. Dept. Agr. Misc. Publ. 940:117-122, 1963.

27.  Rice, E. L. and W. T. Penfound.  Plant Succession and Yield of Living Plant
     Material in a Plowed Prairie in Central Oklahoma, Ecology 35:174-180, 1954.
                                    2-36

-------
  28                                 °
       ment 16:60-63,  1963.7      Uld beedlnS> J- of Range Manage-


  31.
 32<
                                                                           Central
                    -„  _.,,  xxcuiovacn  nj.gnveid,  tcology 44:576-579  1963

 33.
 34.

             	 """ "coastirn   TTT  buccession  Following Extended
                          coastal Rangeland, J.  Range Management  28:279-282,  1975



                                                                     Colorado,
      Ranieland   ,  p     „  "'  Cable'  Urg6r PltS Aid "eseeding of Semi-desert
      Rangeland,  J.  Range Management 23:333-335, 1970.


 37.   Terwillinger,  C., Jr.  and  J.  E.  Jensen.   Analysis of Range Reseedine

      Results, Springfield  Land  Utilization  Project,  Colorado A^r  Exp  Sta

      Colorado State University,  Ft. Collins,  General  Ser.  Paper 666^95^>,"




 38.   Tomanek, G  W.  Dynamics of Mulch  Layer  in  Grassland  Ecosystems,  p 225-240

      in R. L. Dix and R. G. Beidleman  (ed.).   The  Grassland  Ecosystem-  A Pre

      liminary Synthesis, Range  Sci. Dept. Sci. Ser. No.  2, Colorado  State
     University, Ft. Collins, 1969.                        i-oioraao  state



39.  Tomanek, G. W.  Comprehensive Network Site  Description,  HAYS  US/IBP

     Grassland Bioae Tech. Rep. No. 41, Colorado State'universjy,' Ft^Collins,
                                                Analysis of Structure and Func-
                  x^,i^ ^^^cejiib, Colorado state University, Ft  Collins
                  ,  622 pp.                                               '
                                      2-37

-------
41.  Weaver, J. E. and W. E.  Bruner.  A Seven-year Quantitative  Study  of  Suc-
     cession in a Grassland,  Ecol. Monogr.  15:297-319,  1945.

42.  Weaver, J. E., and N. W. Rowland.  Effects of Excessive  Natural Mulch  on
     Development, Yield, and  Structure of Natural Grassland,  Bot. Gaz.  114:
     1-19,  1952.

43.  Wein,  R. W. and N. E. West.  Seedling  Survival on  Erosion Control  Treat-
     ments  in a Salt Desert Area, J. Range  Management 24:352-357, 1971.

44.  Wells, P. V. Succession  in Desert Vegetation on Streets  of  a Nevada  Ghost
     Town,  Science 134:670-671, 1961.
                                                      o
45.  White, K. L. Old-Field Succession on Hastings Reservation,  California,
     Ecology 47:865-868, 1966.

46.  Whiteman, W. C.  Comprehensive Network Site Description, DICKINSON,  US/IBP
     Grassland Biome Tech. Rep. No. 40, Colorado State  University, Ft.  Collins,
     1970,  15 pp.

47.  Whiteman, W. C., H. T. Hanson, and G.  Loder.  Natural Revegetation of
     Abandoned Fields in Western North Dakota, North Dakota Agr. Exp.  Sta.  Bull
     321, 1943, 18 pp.

48.  Whittaker, R. H.  Communities and Ecosystems, Macmillan, New York, 1975.

49.  Woodruff, N. P. and F. H. Siddoway.  "A Wind Erosion Equation, "Soil Sci.
     Soc. mer. Proceedings 29:602-608, 1965.

50.  Young, J. A. and R. A. Evans.  Downy Brome—Intruder in  the Plant  Success-
     sion of Big Sagebrush Communities in the Great Basin, J. Range Management
     26:410-415, 1973.
                                       2-38

-------
                                    CHAPTER 3
                                DECIDUOUS FOREST
  3.1   OVERVIEW
      Temperate deciduous  forest  is a vegetation  type  in which broad-leaved
 trees, the dominant growth form, exhibit an annual  cycle of  shedding  their
 leaves and becoming dormant as day length decreases and developing new  leaves
 and becoming photosynthetically active as day length  increases.  Deciduous
 forest is found in parts of some 33 states in the eastern half of the United
 States, including the flood-plain forests which extend, finger-like, deeply
 into the prairies. 6  Deciduous forest is bordered by prairie on its western
 edge,  which extends from Canada to Texas.  The border (ecotone)  between decidu-
 ous forest and prairie is typically savanna,  grassland with scattered trees,
 comprised of species contributed from each of the two major vegetations (e.g.
 Andropogon-guercus).   Moisture limits the western extension of deciduous forest,
 The northern boundary  is with  the Canadian  boreal forest  with a  broad ecotone
 of deciduous  trees and conifers extending across  the great  lakes region into
 England.   The eastern  and southern  borders  are  effectively  the Atlantic Ocean
 and Gulf  of Mexico,  respectively.

     Regionally, precipitation  ranges from approximately  76 cm to  152  cm,
 annually,  and  is evenly  distributed throughout the year.36  Locally,  such
 factors as soil type,  aspect, and topography may  exert a greater influence on
 forest type or distribution than gross precipitation.8

     The deciduous forest of the eastern United States is by no means homoge-
neous.   Kuchler   mapped 14 major forest associations, Braun8 recognized nine.
A great deal of variability is contained within either system under native,
undisturbed conditions.  Anthropogenic (man-made)  changes in vegetation have
greatly increased the variability over native  conditions.
                                     3-1

-------
3.1.1  Mixed Mesophytic Forest
     Mixed mesophytic forest is the prevalent forest of the unglaciated
Appalachian Plateau Region of Pennsylvania, West Virginia, Ohio, Kentucky and
Tennessee.  Diversity characterizes the mixed mesophytic forest, both in flora
and fauna.  The floral dominance is shared by maple  (Acer saccharum), buckeye
(Aesculus octandra), beech (Fagus grandifolia), yellow poplar  (Lirodendron
tulipifera), white oak (Quercus alba), northern red  oak (Q. rubra) and basswood
                             •jfi
(Tilia heterophyla).  Kuchler   listed 26 other tree species as major compon-
ents.

3.1.2  Oak-Hickory Forests
     A wide-spread forest type in the central United States, oak-hickory forest
occurs in the drier regions of the deciduous forest  biome.  Lying west of the
mixed mesophytic forests, it extends in a discontinuous distribution from Michi-
gan to Alabama in the east, to Oklahoma and Texas to the southwest and as
lowland forest into Nebraska and Iowa.  Dominant trees are various oaks  (Querus
alba, Q. rubra, Q. velutina) and hickories  (Carya cordiformis, C. ovata) with
American elm (Ulmus americana), basswood  (Tilia americana), black walnut (Juglans
nlgra) and white ash (Fraxinus americana) being important components.

3.1.3  Appalachian Oak-Chestnut Forest
     Appalachian oak-chestnut forest includes the Ridge and Valley Province
and the Blue Ridge Province in the south  and widens  to include much  of Ohio,
Pennsylvania and New Jersey, and the coastal plain north to southern Maine.
Since the loss of American chestnut  (Castanea dentata) to an introduced fungal
disease, the dominants are white oak (Quercus alba)  and northern red oak
(Q. rubra).  Maples (Acer rubrum, A. saccharum) and  hickories  (Carya cordifor-
mis, C. glabra, C. tomentosa) are important, as are  black birch  (Betula lenta),
beech  (Fagus grandifolia) and yellow poplar (Lidrodendron tulipifera).

3.1.4  Oak-Hickory-Pine Forest
     Characteristic of the piedmont, this forest type extends west to Texas.
Dominants include various hickories  (Carya spp.), shortleaf pine  (Pinus echinata),
loblolly pine  (Pinus taeda), white oak (Quercus alba) and post oak  (Q. stellata).
                      O fl                                            -^——————
A rich forest  (Kuchler   listed 19 other  components), the climatic climax is

                                     3-2

-------
  probably oak-hickory8 but much of the region has been dominated by second-growth
  pine forests since pre-settlement times.  The erodable, red and yellow clay
  soils of the region are well-known.

  3.1.5  Southern Mixed Forest

       This forest type occupies the coastal plain from South Carolina to Louisiana
  and  into Texas.   Much of the region  is covered with pine,  especially loblolly
  (Pinus  taeda)  which some authors36 have considered to be an edaphic (soil
  derived)  climax.   Experimental data  have shown,  however, that  the  pines yield
  to hardwoods when  fire  is  excluded.   Kuchler26 listed the  dominants as  beech
  (Fagus  grandifolia),  sweet  gum (Liquidambar styraciflua),  magnolia (Magnolia
  grandiflora),  slash  pine (Pinus elliottii),  loblolly pine  (P.  taeda), white  oak
  (Quercus  alba)  and  laural  oak  (Q.  laurifolia).   Rraun8  considered  the forest
  to be transitional to subtropical  evergreen  forest  with  northern species  drop-
  ping out  in central  Florida  and being  replaced with  more southerly species.

  3.1.6  Beech-Maple Forest

      The beech-maple forest  is found from Indiana to New York, including parts
 of Michigan.   Mature beech-maple forest presents almost  a park-like appearance
 with  large trees forming a light-extinguishing canopy such that little under-
 growth can survive.   Beech (Faus grandifolia) and sugar maple  (Acer saccharum)
 are  the  dominants with Ohio buckeye (Aesculus glabra). shagbark hickory (Carya
 ovata),  white ash (Fraxinus americana). black walnut (juglans nigra).  yellow
 P°plar (lirodendron tulipifera),  wild black cherry (Prunus  serotina).  red oak
 (Quercus rubra),  basswood (Tilia americana),  American elm (Ulmus  americana).
 and slippery  elm (U.  rubra)  being  locally important.

 3.1.7  Maple-Basswood Forest

     Maple-basswood  is a  limited forest type  in northern  Illinois,  northeastern
 Iowa, Wisconsin,  and  Minnesota.  Sugar  maple  (Acer  saccharum) and basswood
 (Tilia americana) are the dominants with  northern red oak (Quercus  rubra)  usu-
ally being present.   Other important species  include  box  elder  (Acer negundo),
bitternut  hickory (Carya cordiformis),  red ash  (Fraxinus  pennsylvanical. Hop-'
hornbeam  (Qstrya virginiana), bur oak (Quercus macrocarpa), American elm  (Ulmus
americana) and slippery elm (U. rubra).
                                     3-3

-------
3.1.8  Hemlock-Hardwood Forest
     An ecotone and not a distinct forest type,  the hemlock-hardwood forest
stretches from Wisconsin to Maine with islands at higher elevations of the
Appalachians south to Tennessee and North Carolina.  It merits consideration
on size alone.   The dominant trees are beech (Fagus grandifolia),  sugar maple
(Acer saccharum),  yellow birch (Betula allegheniensis)  and hemlock (Tsuga
canadensis).

     The eight forest types described represent what is considered to be the
potential vegetation of significantly large portions of their regions.  None
of them was ever homogeneous over their ranges and that is certainly true today.
The vegetation of a region has always been a dynamic concept with climate, fire,
                                                        14
and anthropogenic influences producing continuous change

     Vegetational development on a particular site has varied with the influ-
ences producing change   and as new perturbations occur, new developmental
                                   47
sequences and end points may result  .  Nevertheless, ecosystem development,
either on new substrate (primary succession) or recovery following disturbance
(secondary succession), follows a course which is highly predictable in gener-
ality, if not in specific cases.  The generalities offer a point of reference
and will be discussed first.

3.1.9  Theoretical Considerations
     As a forest ecosystem developes, changes take place which convert a com-
munity comprised primarily of small, short-lived, opportunistic species having
little effect on their environment to a community comprised of organisms which
tend to be larger, long lived, and which often exert great control over their
                                                              35
own environment.  Some of those trends were summarized by Odum   from which
Table 3-1 is adapted.

     A great deal of evidence has been accumulated which supports the generali-
zations in Table 3-1.  That biomass accumulates is obvious to anyone who com-
                                                  46
pares a forest with an abandoned field.  Whittaker   described a succession on
Long Island, New York  (oak-hickory-pine in the system used here) in which
                                2
biomass was approximately 1 kg/m  in the earliest stages and increased to
                      2
between 30 and 40 kg/m  in the mature forest.  Biomass increased sigmoidally
(Figure 3-2) with the result that increasing metabolic demands reduced the
                                     3-4

-------
         Table 3-1.  Parameter trends during deciduous forest
                     succession (adapted from Odum35).*
ECOSYSTEM ATTRIBUTE EARLY SUCCESSION
Biomass
Gross production/respiration
Gross product! on/biomass
Net exosystem production
low
high
high
high
MATURE FOREST
high
low
low
low
 Food  chains

 Species  diversity •
 Stratification
 Size  of  dominants
 Life  cycles
 Niche relations

 Nutrient cycling
 Detritus

Overall  homeostasis
Stability
 simple  linear,
 mainly  grazing
      low
      little
      small
short, simple
Broad, oppor-
tunistic
     open
little,
unimportant
poorly developed
     low
 complex,  inter-
 connected, detritus-
 based
       high
       much
       large
 long,  complex
 narrow, specialized
       closed
 important

we 11-developed
      high
 *Terms in this table that are not in Glossary should  follow
  definitions  given  by Odum.3°
                                 3-5

-------
energy available for incorporation into new biomass.  Consequently, yield
declines with ecosystem development.

     With development in forest ecosystems, an organic layer accumulates on
the forest floor.  Prior to the accumulation of this detritus, animal food-
chains are limited to those organisms which can consume the living vegetation
(herbivores and their predators).  As the detritus increases, a complex of
microflora and small cryptozoa  (literally "hidden animals") begin to exploit
it as a resource base.  In a beech forest in Denmark, Bornebusch  found an
                           2
average of 4424 cryptozoa/m  with a biomass of 37.8 g.

     There is consensus that species diversity increases during succession
although mature ecosystems tend to be less diverse than younger, transitional
             •Z f.
stages.  Odum   gave an example of a forest which attained maximum diversity
at an age of almost 160 years with diversity declining beyond that.  Whittaker
gave as an example a forest which attained maximum diversity quickly, in about
8-10 years, and declined gradually as the ecosystem developed.  It is accepted
that maximum diversity is attained during the transition from one more-or-less
discrete successional stage to another when species belonging to each state
coexist in the same area.   Consequently, in the example given by Odum   maximum
diversity apparently occurred during the transition to a mature forest, the
usual case, and in the example given by Whittaker   maximum diversity occurred
during the transition from the herb to shrub stages.

     Stratification,  layering in the forest, is a correlate of species diver-
sity.   The generalized sequence of herb to shrub to tree dominance is, in
effect, the adding of layers which add species to the community.   Bird species
reflect the effect of increasing stratification perhaps most clearly.   Johnston
        21
and Odum   reported two breeding bird species in a three-year-old field having
a grass-forb dominance,  13 species for the local oak-hickory-pine climax.   The
size of dominants and their life history complexity also are correlated with
increased diversity and stratification as succession proceeds to climax.
Bazzaz  reported that Solanum carolinense,  Ambrosia artemisiifolia,  and
Digitaria sanguinales,  all low-growing annuals (plants which complete their
life cycles in one season),  were established in a cornfield before the first
winter after abandonment.   Climax for the area was oak-hickory with Quercus
velutina being the major dominant.
                                     3-6

-------
            Early  successional  stages  are  said  to  have  open  nutrient  cycles  in
  which  the  nutrient  elements are primarily  extrabiotic (outside  of organisms
  and  their  products).  Mature  stages  are  said  to  have  closed  cycles  in which
  detritus serves  as  a  critical  link between uptake and loss of nutrients from
  vegetation.  Vitousek and Reiners45  examined  this hypothesis and found that
  for  elements which  have atmospheric  inputs to ecosystems,  outputs from the
  ecosystem  in the form of stream discharges were  low during the  rapid growth of
  early  succession and increased with maturity to  where  element input and output
  were approximately  equal.   Thus, it may be concluded  that uptake in mature
  forests is in equilibrium with release of elements from organic matter.   Thus,
  constancy characterizes mature ecosystems, and represents a measure of sta-
  bility and homeostatic control.

      Two categories of perturbations have occurred during the history of
 deciduous forests which may yield insight into the impacts of spill  cleanup.
 Natural perturbations  and  anthropogenic (man-induced)  perturbations  are  not
 always  distinct from each  other,  however.  An  example  of where  they  overlap is
 fire.  Fires  have occurred naturally throughout  the  history of  deciduous
 forests   but,  observing the  changes  that fire brought to deciduous  forest
 ecosystems  or their  utility as hunting tools,  the American  Indians would
 deliberately  set  fires.12   Thus,  inclusion  of  fire as  a "natural" perturbation
 is  somewhat arbitrary.
 3.2   NATURAL  PERTURBATIONS

 3.2.1   Fire
      The role of fires in ecosystems has recently been summarized.24   Lightning
 has been the primary source of naturally occurring fires although man  has been
 setting fires in North America for an estimated 20,000 years.23  The primary
 difference between natural or lightning fires and set fires is the seasonality.
 Electrical storms are essentially summertime phenomena, occurring when overall
 combustibility is typically low.   Set fires can occur at any time and  thus are
 potentially more destructive.   It is probable that the grass fields of central
 Kentucky described by Clark12 were maintained by frequent fires set by Indians.

     Early succession following complete forest removal by fire in Finland44
was qualitatively similar to succession following other massive perturbations.
                                     3-7

-------
The pioneer invaders were reselected, annual species with windborne seeds,
such as dandelion  (Taraxacum officinale) hawksweed  (Hieracium spp.), and fire-
weed (Chamaenerion angustifolium).  Other early invaders, such as raspberry
(Rubus idaeus), were carried by birds.  An herb stage was evident after 3 years
and grasses were dominant by the  sixth year.

     Henry and Swan   reconstructed a 300-year sere  (successional sequence) in
New Hampshire from evidence contained in the soil and trees growing on the site,
From charcoal and partially burned wood found in soil and from the ages of the
oldest trees in the forest, they  concluded that a massive fire had destroyed
the forest in about 1665.  They were able to identify eastern white pine (Pinus
strobus), white oak (Quercus alba), hemlock (Tsuga canadensis), spruce (Picea
sp.)> red maple (Acer rubrum), and aspen (Populus sp.) as components of the
                                                    9f\
destroyed forest, species found in the region today.    Herbaceous or short-
lived early invaders left no records, but the ages of extant trees allowed
them to reconstruct the succession of tree species.  White pine (Pinus strobus)
was the first tree invader, coming in between 1665 and 1687 and producing a
distinct pine forest stage.   Hemlock (Tsuga canadensis) invaded between 1684
and 1780, paper birch (Betula papyrifera) in 1742, 1770 and 1795, red oak
(Quercus rubra) between 1790 and  1860 and beech (Fagus grandifolia) began
entering slowly in 1815.  Thus, it took 150 years for the dominants of the new
climax forest to appear and possibly another 50 to 100 years for them to estab-
lish dominance.  Consequently, in New Hampshire, it may take up to 250 years
for a mature forest to reestablish itself following a massive perturbation,
such as complete consumption by fire.  Furthermore, minor perturbations,
such as wind-felling of trees during some 250 years, can open patches of forest
at different successional stages, and consequently different environmental con-
ditions, which permits the invasion of still further tree species which will
determine the ultimate composition of the forest.   Therefore, specific predic-
tions cannot be made about the ultimate consequences of a particular gross
perturbation.

     Other fires,  which remove the understory of forests without seriously
injuring the dominant trees, have no correlate among the proposed clean-up
procedures and will not be treated.
                                     3-8

-------
 3.2.2  Primary Successions

      Primary successions occur where no ecosystem has previously existed, such
 as on newly-created volcanic land.   Land masses within the eastern deciduous
 forest region have been remarkably stable for thousands of years and,  conse-
 quently,  primary successions are poorly understood.   Surface mining using over-
 burden for reclamation, may produce a substrate which can lead to essentially
 a primary succession.   Costly mandatory rehabilitation bonds,  for contouring
 spoil banks to avoid instability and slope failure,  exert economic pressure
 for "natural recovery."  Information on the probable years for primary succes-
 sion, on  surface-mine  spoil banks,  becomes vital  to  predict the years  to com-
 plete rehabilitation.

      A case history of primary succession was developed by Morrison and
 Yarranton  '    for sand dunes.   They have identified three stages in primary
 succession.   The  first stage or colonizing stage  lasts up to 1600 years and is
 characterized by  beachgrass (Ammophilia breviligulata),  balsam poplar  (Populus
 balsamifera),  willow (Salix glaucophylloides),  and sand cherry (Prunus pumila).
 The  colonizing stage is followed by a transition  stage lasting another 1300
 years and characterized by  scattered trees without canopy closure.   The typical
 species are  hybrids  of red  and  black oaks  (Quercus rubra x velutina),  fragrant
 sumac (Rhus  aromatica),  choke  cherry (Prunus  virginiana),  a sedge (Carex lan-
 guinosa)  and  snowberry (Symphoricarpos  racemosus).   That is followed by a
 persistent  stage  lasting at  least 1900  years  and  comprised of  white  oak (Quercus
 alba.), New Jersey  tea  (Ceanothus  americanus), black  cherry (Prunus  serotina)
 and  teaberry  (Gaultheria procumbens).

      The number of species  apparently increases to a stability point approxi-
 mately 2000 years  after the  dune  is  formed while  frequency continues to  increase
 throughout the 4800 years, which  is  too  long to wait.

 3.3  MAN-MADE  PERTURBATIONS

 3.3.1  Surface Mining
     Although not nearly as extensive as clearing for agriculture, surface
mining represents a massive perturbation which is both highly visible and
difficult  to ameliorate.  The mine spoil which typically finds its way to the
                                     3-9

-------
surface is analogous to a new land surface and,  consequently,  natural succes-
sions on mine spoil somewhat resembles primary successions.   Since no
4800-year-old mine spoils have been identified,  there is no evidence that the
two are directly parallel.  A further complication is the acidity problem
shared by the central and Appalachian coal fields.  In these areas, iron disul-
fide is exposed to air, oxidizes and produces sulfuric acid.  Spoil pH is often
too low to support plant life and the low pH tends to increase the solubility
of a number of metal ions which reach toxic levels.   None of the proposed
cleanup procedures is likely to expose appreciable quantities of these acid-
forming disulfides.  Once the pH of surface mine spoils is adjusted, the plant
substrate is possibly quite similar to the product of cleanups, particularly
on shallow soils.   Therefore, examination of the revegetation history of mine
spoils may provide a model of the consequences of some cleanup procedures.

     Surface mining in the eastern United States, especially in the rugged
topography of Appalachia, has created problems of water runoff, erosion, and
sedimentation which have led to extensive studies and actions directed toward
stabilizing spoil surfaces through soil amendments and revegetation.  Conse-
quently, there exists at least some data both on natural revegetation and
reclamation.  In general, low pH inhibits both reclamation and natural recovery,
If pH is not limiting, planting can greatly accelerate early recovery.  One
study estimated that replanted spoil was as much as 10 years ahead of untreated
                          1 8
spoils after only 3 years.

     Untreated spoil often does revegetate naturally, however.  One of the
                                                         29
earliest reports of natural revegetation was by McDougall   who reported on
revegetation of surface-mined bottomlands in Illinois.  He concluded that
proximity to a seed source was an overriding factor in the invasion rate of
species.  He described a sere which began with annual weeds, such as knotweeds
(Polygonum aviculare and P. persicaria) with occasionally barnyard grass
(Echinochloa crusgalli) and giant ragweed (Ambrosia trifida) in furrows.  The
drier ridges contained sweet clover (Melilotus alba), aster (Aster ericodes),
sunflower (Helianthus hirsutus and H.  dec apet a1us), ragweed (Ambrosia artemi-
siifolia), evening primrose  (Oenothera biennis), black mustard (Brassica
nigra) and prickley lettuce  (Lactuca scanola).  Pioneer woody species included
willow  (alix nigra) and cottonwood (Populus deltoides), if a seed source were
available.  They were followed by seedlings of the forest-stage species:

                                     3-10

-------
 silver maple (Acer saccharinum), elm (Ulmus americana) and sycamore (Platanus
 occidentalis).   He stated that a 25-year-old strip mine was covered with "well-
 established bottomland forest."

      Croxton   reported on revegetation of strip-mined lands in the same region
 of Illinois and found evidence of the variability of results which have come
 to characterize such areas.   He attributed the rapid recovery of the site
 studied earlier by McDougall to the deposition of river alluvium over the spoil
 and the creation of unusually favorable conditions.   The sere he reported was
 essentially the same as that reported by McDougall,  allowing for availability
 of various weed seeds.   He indicated that recovery within 25 years was unusually
 rapid and that  areas with pH below 4.5  failed to revegetate.   He noted that the
 old-field shrub stage of succession, which is characteristic of deciduous
 forests,  was poorly represented on mine spoil.

      A distinct shrub stage  was reported on mine spoil in southern Indiana.11
 Beginning approximately 7 years after abandonment of the spoils,  the stage
 contained sumac (Rhus sp.),  honeysuckle (Lonicera japonica),  trumpet vine
 (Campsis  radicans)  and  poison ivy  (Rhus radicans).   Included  in  the shrub
 stage were saplings  of  cottonwood  (Populus  deltoides),  river  birch (Betula
 nigra), persimmon  (Diospyros  virginiana),  sycamore  (Platanus  occidentalis),
 sassafras  (Sassafras  albidum),  slippery elm (Ulmus rubra),  black  cherry
 (Prunus serotina),  flowering  dogwood (Cornus  florida)  and shingle  oak  (Quercus
 imbricaria).  By 11  years  after abandonment,  a tree  layer had developed  and
 by  21  years  pin oak  (Quercus  palustris)  and persimmon  were  the dominant  species.

      The  conclusions  to be drawn from these studies  of early  succession  on
 strip-mine  spoils are that natural revegetation  can  occur  if  soil  conditions
 permit and  that the rate and  pattern of  recovery  is  highly variable.  The
 variability  is primarily the  result  of the  complexity  of  the plant  growth
 medium resulting from mining  and the variability of  the seed source  in the
 surrounding  area.
     Revegetation of strip mine spoil has emphasized establishment of ground
 cover and the prevention of erosion.   Species planted, such as sericea les-
pedeza, have been selected for their ability to quickly establish ground cover
with the expectation that they will eventually be replaced by natural inva-
 sions of local species.   For the most part, workers have been satisfied with
                                     3-11

-------
 achieving  their  primary  objective.   Long-term studies  which  seek  to  evaluate
 the  return of  such  disturbed  lands  to  pre-mining  ecosystems  are virtually non-
 existent,  although  "full-growth"  of "high-grade"  forests  has been achieved
 within  50  to 70  years.

 3.3.2   Post-Agricultural Successions
     Secondary succession following the abandonment of agricultural  land  is
 the  best-known example of ecosystem development in deciduous forests.  Clearing
 of unproductive  land, changing markets, and  loss  of fertility has led to  aban-
 donment of land  throughout the deciduous forest.  That abandonment began  almost
 as soon as settlement began and continues into the present.   Consequently,
 virtually  all aspects 6f succession have received at least some study.  For
 convenience, a regional approach  will  be taken.

     3.3.2.1  Southeastern States--
     Succession  on  the Georgia piedmont has  been  studied  by  Nicholson and
     32  33
 Monk '    who identified 5 major  stages.  Their "early forb  stage" occupied
 the  first  10 years  following abandonment and  was  dominated by perennial grasses
 (Andropogon scoparius, Aristida purpurascens, Eragrostis  sp.) and forbs (Allium
 vineale, Erigeron canadensis, Haplopappus devaricatus, Cassia nictitans,  Heter-
 otheca subaxillaris, Aster pilosus, Solidago  sp., Lespedeza  sp.)-  Chickasaw
 plum (Prunus angustifolia) and brambles (Rubus spp.) formed  a shrub  layer along
 with tree  seedlings (Pinus taeda, P. echinata, Diospyros  virginiana, Celtis
 occidentalis, Rhus  copallina) and vines (Lonicera japonica,  Smilax bona-nox,
 Wisteria sinensis,  Campsis radicans).  This  is followed by what they called
 a "broomsedge stage" in which Andropogon scoparius increases  its  cover and
 community  richness  declines.   The broomsedge  stage lasted from the tenth  to
 the  twentieth year  following abandoment.  The old-field shrub stage replaced
 the broomsedge stage and persisted up to approximately 30 years after aban-
 donment.   Loblolly pine (Pinus taeda) and sweet gum (Liquidambar  styraciflua)
 increase as shading, soil moisture and soil organic matter increase.   Species
 richness increases  in all strata except the herb  layer, which declines due to
 shading and litter accumulation.   The end of the stage is identified by canopy
 closure (approximately 30 years)  and the reduction of the herb stratum to  about
percent of the open-field level.
                                     3-12

-------
      Canopy closure* signals the start of the pine stage in which  loblolly
 pine is the dominant.  The stage lasts approximately 70 years  (30  to 100 years
 post-abandonment) and there is a slight decline in total above-ground biomass
 for the first 60 years.  Woody species increase in the understory  as broad-
 leaved deciduous species invade.   Maximum richness occurs as the species from
 early succession overlap with the tree species of the later climax and subcli-
 max forests,  around 70 years post-abandonment.  Richness declines  as hardwoods
 replace the pines (100 to 125 years) to increase again around 150 years as new
 hardwood species appear.   The hardwood stage lasts from 100 years,  post aban-
 donment, to 200 years or more.
                       21
      Johnston and Odum   reported on breeding bird populations of the Georgia-
 piedmont sere.   They found that both density and diversity increased with suc-
 cession up to the pine stage  where  each declined sharply.   Only one species,
 the pine warbler (Dendroica pinus),  used  the pine  overstory,  but other  species
 came in as the  hardwood understory  developed.   Approximately  12 species  with
 140 breeding  pairs/40 ha  (100 acres)  were present  before  canopy closure  and
 this dropped  to  10  and  17  at  20 years.  The  maximum attained  was 20 species
 and 240 breeding pairs/40  ha  in the  transition  forest with  23  and 220 in  the
 deciduous  forest stage.  Only the cardinal  (Richmondena cardinalis) was present
 in  all  stages with most bird species  being stage-specific.
     On  the North Carolina piedmont,6  short-leaf pine  (Pinus echinata) domin-
 ates the pine stage with up to  7400  stems/ha  (3000 stems/acre).   The first
 oaks  (Quercus spp.) appear at about  20 years with the first hickories (Carya
 spp.) only a few years later.  The sequence during the first 3 years was given
 by  Keever.    Crabgrass (Digitaria sanguinalis) is able to invade plowed fields
 in  the fall of the year of abandonment because of its drought resistance and
 ability to germinate under cool soil temperatures.  Horseweed  (Leptilon cana-
 dense) may germinate at about the same time but grows to maturity the follow-
 ing summer.  Horseweed and ragweed (Ambrosia elatiorl  dominate one-year-old
 fields with germination of new horseweed seeds being inhibited by the decay
products of tbe parent plants.   An aster (Aster pilosus)  appears the first
year and becomes a dominant the  second, along with ragweed and buttonweed
'See Glossary for definitions of terms.
                                     3-13

-------
 (Diodia  teres).   A  few  broomsedge  (Andropogon  virginicus)  plants  appear the
 plants appear  the first year  but require  2  years  to  set  seed.   They become
 dominants by the  third  year.
      On  the Virginia piedmont, Virginia pine  (Pinus  virginiana) dominates  the
 pine  stage producing a  closed canopy within 10 to 20 years.  They mature within
 50  to 70 years.
      Further inland, the old-field  stage  in middle Tennessee is characterized
 by  elm (Ulmus  americana) and  hackberry  (Celtis occidentalis) with an under-
 story of broomsedge  (Andropogon virginicus), goldenrod  (Solidago  altissima),
 aster (Aster pilosus) and panic-grass  (Panicum sp.),39   Sassafras (Sassafras
 albidum) may be a component of old-fields and,  because of  allelopathy,  may
 persist  in pure clones  into the mature forest.
      3.3.2.2   Northeastern States-
      Long-term succession in  New England has been  described earlier.     Bard2
 described secondary succession on the New Jersey  piedmont.  The first-year
 dominants were ragweed  (Ambrosia artemesiifolia)  and  evening primrose  (Oeno-
 thera parviflora) with  later  dominants represented.   They were replaced in the
 second year by a  goldenrod (Solidago nemoralis).   S.  nemoralis remained impor-
 tant  for 15 years when  it was replaced by S^. juncea which persisted  for another
 45 years or more.  By the fifth year, an aster  (Aster ericoides)  became the
 dominant herb and remained important for another  20 or more years.   Fifteen-
 year-old fields were dominated by broomsedge (Andropogon scoparius with A.
 virginicus)  which remained important for another 45 years.  Eastern  red cedar
 (Juniperus virginiana)   is among the first woody invaders (2- to 3-year-old
 fields)  and is the major arborescent species for more than 60 years.  Dominant
 shrubs are brambles (Rubus flagellaris) which reach their peak 25 years  after
 abandonment and poison  ivy (Rhus radicans) which is most abundant 60 years
 after abandonment.  The trees of the climax oak-hickory forest enter before
 the red cedar is mature and are well established in the understory by the
                                                 20
 sixtieth  year after abandonment.   Recent evidence   indicates that allelopathy
 is important to the early replacement of species.   Aster pilosus,  a  second-
 stage dominant, was  shown to produce a chemical which inhibited germination
of first-stage dominants.
                                     3-14

-------
       There  is  a  succession  of animals  which follows  the plant  succession.
  Kircher   showed that  bird  diversity increased  with  each serai stage and that
  there was a gradual  shift during  succession from  species which used  the area
  for  feeding only to  species which nested  in the area.   Pearson37  reported  that
  small mammals  increased with  early  succession but declined  as  the forest be-
  came  established.
      At least  occasionally, the orderly replacement  of  species  becomes  delayed
  when one or more  species establishes a stable,  persistent community  prior  to
  the expression of climax.  Examples include34 speckled  alder (Alnus  rugosa)
  for 20 years or more, witch-hazel (Hamamelis virginiana)  for 40 years, nanny-
 berr>r (Viburnum lentago) for 45 years,  and meadowsweet  (Spiraea latifolia).
 highbush blueberry (Vaccinium corymbosum),  lowbush blueberry (V.  angustifoliunQ .
 arrowwood (Viburnum recognitum).  sweetfern  (Comptonia peregrina)  and common
 juniper (J"niperus communis) for 47 years.  Thus,  the typical succession cannot
 be expected  in  every instance.  Nor is  it possible to predict the frequency of
 occurrence  or duration  of stable  shrub  stages.
      3.3.2.3 Central States--
      Early  succession in southern  Illinois was  described by Bazzaz.3   Follow-
 ing the  harvest of corn in August, cover  was provided by weeds  already estab-
 lished,  including Solanum carolinense,  Ambrosia  artemisiifolia  and Digitaria
 sanguinalis.  During  the first growing  season following  abandonment,  annuals
 (plants which complete  their life  cycle in a single season)  were the  dominant
 vegetation and  included Alliumvineale,  Digitaria sanguinalis. Geranium maculatum
 and Aster pilosus.  Sawbrier (Smilax glauca)  was the  first woody plant to be-
 come established.  Total ground cover was  approximately  65 percent.
     Perennials (plants which  live for  several years) were the  dominant  vege-
 tation by the second  year of abandonment.   Erigeron annuus flowered in early
 summer, when it was the conspicuous  dominant, and Aster  pilosus was most  con-
 spicuous later  in the summer.  Ground cover  attained  approximately 115 percent.
     By the third year,  erosion had reduced ground cover  to approximately 85
percent,  but the species replacement pattern continued.  Herbaceous vegetation
had begun to develop strata with goldenrod  (Solidago nemoralis)  dominating
the upper stratum  and panic  grass   (Panicum dichotomum) and three-awn  (Aristida
                                     3-15

-------
 dichotoma)  codominant  in the  lower  stratum.   Old-field  shrubs  began  invading,
 including black  raspberries  (Rubus  occidentalis)  and persimmon  (Diospyros
 virginiana).
      Four-year-old fields had produced a stable herbaceous  community, dominated
 by broomsedge  (Andropogon virginicus).  The developing  shrub layer contained
 persimmon,  sassafras (Sassafras albidum), elm  (Ulmus alata), eastern red cedar
 (Juniperus  virginiana) and sumac  (Rhus spp.).  Cover (all strata  summed) had
 increased to 119 percent, with 8 percent being contributed  by  shrubs.  Vegeta-
 tion  of  10- and  15-year-old fields  remained similar in  composition with the
 shrubs growing to a cover value of  approximately  16 percent.   By  the 25th year,
 the older sassafrass and persimmons had produced  a tree  layer with the composi-
 tion  of  the shrub and  herb layers as before.
      By  40  years after abandonment, the tree  layer had  reached a  basal area of
 about 15.6  and the shrub layer had  increased  to 21 percent.  Broom sedge still dom-
 inated the  herb  layer.   Elm and sumac remained the dominants in the shrub layer
 but a few seedlings of the climax oaks (Quercus imbricaria  and Q. velutina) had
 appeared.  A total of  22 species occurred in the  tree layer with  sassafras and
 persimmon remaining the dominants.
     The rate of invasion of the climax oaks was  considered to be dependent
 on the proximity to mature forest, since the seeds were dispersed by squirrels
 which prefer the forest edge.  Diversity increased throughout the succession
                                4
 to a maximum in the forest stage  but may be low when allelopathic species are
 present.   Environmental heterogeneity produces a variety of seres within a
 given area.
     The animals which occupy serai stages change with the vegetation.   Sly
 reported that deer mice (Pereomyscus maniculatus)  occurred in the early,  grass
 stage of successions while the white-footed mouse (P.  leucopus) occurred in
the later,  forest stages.   The two species overlapped during the old-field or
grass-shrub stages.   Serai stage is not the only determinant of acceptability
to wildlife, however.   Beckwith  concluded that the size of a tract of other-
wise suitable vegetation would influence whether or not game animals  would be
found there.
                                     3-16

-------
   3.4   CLEAN-UP  PROCEDURES,  EFFECTS  ON  ECOSYSTEMS
        This section describes the effects that the proposed cleanup methods can
   be predicted to have on deciduous  forest ecosystems.  The description is quali-
   tative for each treatment's impact in the ecosystem.  A quantitative assessment
   of all the cleanup methods is deferred to Section 3.6, after the impact is
   described and the likely time course of recovery is given (Section 3.5).  The
   cleanup treatments range in severity from a light spray, of "tacky" chemicals
   over the vegetation and soils, to severe impacts caused by removing vegetation
  and topsoil,  mixing the subsoil with caustic chemicals and compacting the mix-
  ture with cementing agents.

       (0-1)  Natural  Rehabilitation
       For  manual  cleanup of  contamination, no significant  impact  on  the prairie
  is  anticipated  to be evident after  2 or  3 years.

       (1-1)  Chemical Stabilization
       Properties of the  two  general  classes of chemical stabilizers are given
  in Appendix A, Stabilizers.  Twenty-seven are rated "preferred"  for reducing
  soil  erosion under normal rainfall  regimes in the deciduous forest.  Six allow
  seed  to germinate in place but they break down in a year.  Four  of the six
  chemical stabilizers are safe to handle without special equipment, in machine
  or hand sprayers.   Significant leaf fall from shrubs and trees is caused by
 the four stabilizers but little damage  to roots is expected.
      The application of stabilizers  that  kill vegetation is acceptable when
 vegetation to  which  they are applied must be  removed  in some of the  cleanup
 methods described in  subsequent  discussion  (treatment  1-3).   In general   the
 plant-damaging stabilizers leave a more weather-resistant  and  erosion-inhibiting
 coating over soil surfaces.

      (1-2)  Clear Cutting Vegetation

     Clear cutting deciduous forest  is used in forest management.   Clear cutting
or block cutting permits reproduction from nearby forest trees, but public ac-
ceptance has been a major obstacle to the practice.
     Clearcut  forests revegetate in two ways.   First,  many species have the
ability to produce sprouts from the stumps and roots of the former trees.
                                      3-17

-------
Since the remaining root network is essentially that of a mature tree, sprouts
typically show exceedingly rapid growth.   A coppice forest, which clear-cutting
produces, can produce canopy closure in a decade or two.   Since growth is from
the remains of the former forest, species composition is  typically quite similar,
In addition to sprouts, the opening of a forest by clear  cutting allows for
invasion of pioneer species and an abbreviated version of old-field succession.
The resultant, short-lived heterogeneity is attractive to game species.  Those
pioneer species which grow rapidly enough and can survive canopy closure add
to the diversity of the resulting forest.  Consequently,  coppice forests may
differ in that respect from the original forest.

     (1-3)  Stumping and Grubbing
     Removal of stumps and roots blocks the formation of  a coppice forest and
results in longer recovery times.  A natural succession would be similar to
post-agricultural recovery and include the typical annual-perennial-shrub-tree
sequence of dominants.  The rate at which these stages replace each other will
depend on what seeds were present in the soil prior to cleanup, the size of the
affected area and erosion.  This treatment will differ from post-agricultural
land in that the seeds of species present when the treatment began can be ex-
pected to remain behind.  The time-lag before the climax  species appear as
seedlings would thus be reduced to about a year and the total ecosystem de-
velopment reduced accordingly.  Thus, the up to 250 years required for reaching
climax following agriculture might be reduced to 100 years or less.  Assuming
that the forest removed was not climax but an earlier successional stage, the
time required for recovery to that point would likewise be shortened, with
the exception of early shrub stages.  Since woody shrubs  appear very early in
succession, typically the first or second year, and require several years to
attain dominance, no shortening of that interval would be expected.

     The size of the affected area influences the invasion rate from outside
                              2
the area.  Small (i.e., 0.1 km ) areas have large perimeter to area ratios
                                      2
compared to large areas (i.e., 10.0 km ).  Since squirrels prefer forest
     3
edge,  acorns could be expected to be buried over the smaller areas more quickly
than the larger one with a proportional recovery of oaks.
                                     3-18

-------
       Uncontrolled erosion can act as a recurring perturbation which continually
  sets back recovery in the cleaned-up area to an earlier stage, and can expand
  the perturbation beyond the spill zone.  Thus, presence of erosion will result
  in longer recovery times and larger perturbations.

       (1-4)  Scraping and Grading

       In a forest,  scraping and grading requires that stumps first be removed
  In addition  to the effects of stumping and grubbing (1-3),  scraping and grading
  (1-4)  would  remove the surface organic matter that remained.   That would result
  in higher soil temperatures,  decreased infiltration of water,  increased runoff
  and erosion  and nutrient  depletion.  As a  result,  recovery  times  could  reason-
  ably be expected to  be longer  than with stumping and grubbing  (1-3)  alone,  but
  not as  long  as required for post-agricultural  recovery since some seeds would
  remain  in  the  soil.  The  pattern of recovery would  be  expected to remain the
  same with  the  size of  the  area affecting the rate  as well.  It has been found
  in  strip-mine  reclamations that scraping and grading  (1-4) delays  recovery
  relative to a  rough surface, primarily because of a  reduction in  infiltration
  and an  increased risk of erosion.

      (1-5)  Shallcw Plaaina

      Shallow plowing (<10 cm)  following stumping and grubbing would probably
 have no observable  effect  on ecosystem recovery beyond that  of vegetation re-
 moval.   The surface organic matter would be incorporated into the soil  and
 seeds redistributed,  but  the long-term  effects would not likely be discerible
 The effect would be less  than  soil  removal  because  the seeds  and  organic matter
 would remain  and because the resulting  pattern  of ridges and  furrows  would  trap
 water and decrease  erosion.

      (1-6)  Deep  Plowing

     As  recently  as the early 20th century, farmers  in  the deciduous  forest
 zone cleared and  planted "new ground."  Trees were  killed by girdling and then
 felled and  usually burned.  The farmers then plowed  their fields,  although
 cutting  through the buried roots was punishing to both mule and man.  They
plowed partly out of tradition and partly to keep down weeds, that is, to pre-
vent natural succession.  Deep plowing as a cleanup procedure will do the same
except being done only once the effect  will  be shorter-lived.   Thus, deep plow-
mg on deep soils may delay early succession a year  or two.   Deep plowing on
                                     3-19

-------
shallow, hill soils can be expected to have a longer-lasting effect, if it can
be accomplished at all.  In either case, the effect will be less than grading
or soil removal, since the quality of the seed bed remaining should be superior
to that left by soil removal and the ridges and valleys will retain precipita-
tion.  The succession should be no slower than post-agricultural successions
on similar soils.

     (1-7)  Soil Cover Less than 25 cm
     The addition of uncontaminated soil to a forest floor has several potential
effects, not the least of which is the disruption of vegetation by the applica-
tion of the soil.  One acre-furrow-slice (the soil of one acre to a depth of 6
                                        10                             2
inches) weighs 2,000,000 pounds or more.    Therefore, covering 0.01 km  with
                                             f\             f\              ^
25 cm of soil would require at least 3.7 x 10  kg (8.2 x 10  Ibs or 4 x 10
                                9               2
tons) of soil and up to 3.7 x 10  kg for 10.0 km .  Such large masses of soil
require heavy earth-moving equipment which cannot be operated in a forest with-
out first removing or knocking down the trees.  It must be assumed, therefore,
that any soil additions will be done after the forest is removed.
     Twenty-five centimeters of soil or less would not cover the fallen boles
and stumps of mature deciduous trees, leaving the possibility that at least
some of the trees might resprout from unkilled tissues.  The severity of the
impact then would lie somewhere between that of a clearcut forest, assuming
a greater kill rate, and surface alterations.  Herbaceous material would be
buried below an effective recovery depth and the seeds present on the forest
floor would be too deep to germinate.  Consequently, recovery would consist
of a combination of resprouting of the former trees and invasion of pioneer
species, including the seeds carried in the uncontaminated soil.  As such, re-
covery rate would depend upon the area disturbed, which affects the invasion
rate of trees, the kinds and amounts of seeds contained in the imported soil,
the seed-bed quality represented by the imported soil and sprouting rate of
the fallen trees.  All of these interact to produce a range of variables which
greatly diminishes predictability.

     (1-8)  Soil Cover 25 to 100 cm
     Covering with additional soil primarily magnifies the problems identified
in the previous section.  As depth of added soil increases, the probability of
resprouting of fallen trees decreases.  A full 100 cm of soil would cover most
                                     3-20

-------
 of the fallen boles and many of the uprooted stumps.  Resprouting cannot be
 completely ruled out as a recovery factor, however.
      Since noncontaminated soil must be excavated elsewhere, it is safe to
 assume that seed-bed quality will decline with the area and depth to be covered.
 That is, constraints on disturbing noncontaminated areas will require the de-
 contamination crews to dig deeper at the borrow site as volume demand goes up
 rather than to increase the area disburbed.   In most regions, that translates
 as a lower quality seedbed.   Thus, recovery time may be expected to be extended
 as area treated and burial depth are increased.  Quality of the seedbed would
 be the primary factor affecting early recovery with the size of the area dis-
 turbed affecting later rates  of invasion by trees.
      The level of impact would  be expected to be greater than that of vegeta-
 tion removal,  surface alteration,  plowing, and possibly soil removal.   The one
 ameliorating  factor would be  resprouting of fallen  trees,  if it were to occur.
      The environmental  costs  of disturbance  to the  borrow  sites must  be factored
 into any cost-benefit  analysis  of burial procedures.   For  even  the smallest  area
 considered  (0.01  km ),  2500 m5  soil would be  required  for  burial to  25  cm  and
         3
 10,000  m soil would be required  for  burial  to  100  cm.   Because of the  desire
 to minimize the area affected,  the impact  on  the  borrow  site would be greater
 than  that of any  of the cleanup procedures considered, with  the exception  of
 hard-surface stabilization.

      (2-1)  Remove  Plow Layer
      The  effects of removing the plow layer,  approximately 15 cm (6 inches)
 of soil,  depend upon topography, soil type, and what other cleanup procedures
 are done  in conjunction.  Deciduous trees, growing in forests,  have roots in
 contact with the surface organic layer.  This is where the active exchange of
 nutrients occurs.  Assuming it  is possible to remove the top 15 cm (6 inches)
 of soil without first removing  the trees,  it is probable that no tree would
 survive the cleanup.  While the presence of standing, dead trees would not be
 expected to interfere with recovery,  they would also not expedite it, leaving
no justification for not removing them.
     Assuming removal of the plow layer to follow stumping  and grubbing, the
pattern or recovery would be that of,  at best, a post-agricultural  succession
or, at worst,  a strip-mine spoil.   All plants would  have to invade  from outside
                                     3-21

-------
the area making the size of the affected area important.  If the cleanup were
to occur on nearly level topography having well-developed soils, the remaining
seed bed could be of high quality.  In steep terrain, however, forest soils
typically have a very thin (less than 15 cm) layer of topsoil which would be
removed entirely.  The resulting seed bed would resemble mine spoil and eco-
system recovery could be expected to be accordingly slow.  Subsequent chemical
and most types of mechanical subsoil stabilization have little added impact
but would minimize erosion in the short term recovery.

     (2-2)  Remove Shallow Root Zone
     In deciduous forest, shallow root zone can be interpreted to be a plow
layer or less.  At any rate, soil removal is also seed removal and the early
colonizations must be from outside of the affected area.  Increasing the depth
of soil removal lowers the probability of leaving a suitable seed bed and in-
creases the time necessary for recovery.   Whereas canopy closures can occur
                                                 32 33
within 30 years in post-agricultural successions,  '    such rapid recovery is
unusual following surface mining   which is analogous to soil removal on shallow
soils.  Use of chemical and mechanical stabilizers to prevent erosion would
mimic post-agricultural recovery.

     (2-3)  Remove Scrap-ing and Grading3  Mechanieally Stabilize
     As the sizes of stabilized areas increase more internal structures are
needed for overland water control.  These include dikes, berms, drainage
ditches, shaped waterways and overfall structures on steeper-grade channels
which are all beyond the scope of this presentation.
     The specific consequences of scraping and grading (1-4) were presented
earlier and apply to the pre-stabilized deciduous forest.   Denuded areas
depend on early reseeding by native plants.  A smoothed and compacted surface
having reduced water permeability is the medium represented after erosion
control by stabilization.
     Mechanical stabilization may involve compaction, meshes, mulches, defloc-
culants, and clays.   The effects of mechanical soil compactors remain evident
for many years.  Complete recovery will seldom occur by 100 years.  Simple
ground cover by plant invaders will take up to five years  and succession may
be delayed 25 or 100 years compared to simple clear-cutting.  Heavily compacted

                                    3-22

-------
  areas are subject to repeated regressive cycles following erosive denuding of
  large open areas.   Infiltration  can readily be reduced to an inch per hour or
  less  and offsite runoff is  a  serious problem at even  one hectare of clearing
  and surfacing.

       Non-compacting  stabilization  by meshes and mulches  is  indicated for  the
  cleared,  deciduous forest and is the method rated  in  Table  3-2  (see  Section
  3.6).  This assumes  that mesh is held down  by  pegs  (non-metallic  if  mOwed) on
  steep  slopes and anchored to  prevent  wind detaching or under-running  sections.
  Mulches  are anchored by tractor disc  or cleats  in areas  free of  significant
  water  erosion.

      Deflocculants and clays  have characteristics that produce impacts inter-
 mediate among soil compactors and the various mulches.  They are subject to
 frost  heaving damage and are relatively short-lived since seeds can collect
 and germinate in the breaks.  Deflocculants depress the early growth of plants
 by salt effects in germination and rooting compared to clays.
     Significant damage by small  rodents is probable in a few months for any
 of the chemical and mechanical stabilizers used after  soil removal.

      (2-4)   Remove  Plow Layer., Mechanically  Stabilize
     The  ecological  impacts  were  described in scraping and grading (1-4) and
 the effects of  stabilizing the exposed soil  mechanically  were defined in
 treatment 2-3.   In  general,  natural  recovery will be delayed compared to re-
 moving  5  cm or  less of  soil, by removal  of more seed sources and  mechanical
 stabilization will  seriously retard  revegetation  compared  to clear cutting.

     (2-5)  Remove Shallow Root Zone,  Mechanically Stabilize
     This treatment has effects almost identical  to plow  layer removal  (2-1).
Soil fertility is reduced by deeper cuts and  the potential for natural revege-
tation   shifts to  favor invaders instead of local pioneer species.   The subsoils
will tend to be more erosive than surface layers due to reduced organic matter
and so  require heavier applications of stabilizers.   Higher clay content will
increase density under compaction  and exacerbate water  erosion.   Non-compacting
stabilizers are much superior in this case.
                                    3-23

-------
      (2-6)  Remove Scraping and Grading, Chemically Stabilize
      The physical effects and the ecosystem impact of removal of scraping and
 grading  (2-6) were described under treatments  1-4 and 2-3.  The effects of
 chemical stabilizers are less durable than those of mechanically stabilized
 soils and require prompt revegetation, usually by herbs and grasses in the
 clearings made.  Either class of stabilizer reduces surface erosion, increases
 runoff, and has specific effects on water infiltration according to the spe-
 cific stabilizer and its limits of use  (Appendix A, Stabilizers).
      Chemical stabilizers for soil stabilization control soil erosion while
 seedlings become established then the "preferred" stabilizers break down.
 Overland runoff is reduced by these compounds compared to soil compaction
 treatments.  Stabilization with chemicals will reduce erosion of the exposed
 soil  horizons and should allow early revegetation of native pioneer species.

      (2-7)  Remove Plow Layer, Chemically Stabilize
      The effects of removal of the plow layer  (2-1) are combined with those of
 treatment 2-6.  The deeper soil cut reduces the seeded zone to those hidden by
 small animals and exposes the infertile subsoil as the plant growth medium.
 It presents a hostile environment to transplants and invader species and a
 lengthening of the time for recovery.   Compared to the time required for re-
 covery to climax the delay is probably negligible in the long term.  Chemical
 stabilizers would be the method of choice for revegetating the cleanup zone.

      (2-8)  Remove Shallow Root Zone,  Chemically Stabilize
      The effects of removing the surface soil to a depth of 40 cm increase the
problems of revegetating in the infertile soil over treatment 2-7 and eliminate
revegetation by local pioneer species.   The limited potential for recovery
favors exotic invaders.   Low organic matter increases runoff to a serious ero-
sion  threat and heavy applications of chemical stabilizers are called for,
including those that are not acceptable after other cleanup (Appendix A,
Stabilizers).   Chemical  stabilizers of three to five year durability are
highly toxic and rehabilitation is required before revegetation will succeed.
This  is one of the four  deciduous forest cleanup treatments that require in-
tervention by reclaimers before it can start recovery towards the original
productivity.   If not reclaimed the area will be lost to erosion processes.

                                    3-24

-------
       (3-1)  Barriers to Exolude People
       Natural ecosystems can be benefited by excluding humans from the area
  without any impact on the plants,  soils, and wildlife.

       (3-2)  Exclude Large Animals
       Barriers  which restrict  large animals  only will  have a greater effect on
  the  animal populations  themselves  and  not on ecosystem  structure  and function.
  Large animals  were once nearly exterminated from deciduous  forests  but  several
  species  are now recovering.   White-tailed deer  (Odocoileus  virginianus)  feed
  in old  fields  and  have  increased their numbers  as agricultural  land has  been
  abandoned.   The fencing-off of small tracts (0.01 km2)  would not  represent  a
  significant  proportion  of  available habitat,  but the maximum tract  (10.0 km2)
  could seriously limit the  habitat  of a local  population.  Even  so,  the effect
  would be insignificant  on  a regional or  state basis.

       (3-3)  Exolude Large and Small Animals
      The erection of barriers to prohibit animal crossings affects only ground-
 dwelling vertebrates without impacting soil  and plants.   Birds, insects and
 squirrels are not likely to be restricted by any barrier which meets engineer-
 ing and  cost acceptability.  Nor are small  mammals and reptiles likely to be
 successfully removed with the enclosure of the area.  Thus,  species composition
 of an area would  be virtually unaffected by  the erection of barriers.  Restrict-
 ing movement of animals  into or out of  the area would  restrict  gene flow but
 only  the smallest  area of concern would likely be impacted and  then only if
 the barrier remained intact for several decades.   The  effect on large mammals
 is  treated  in the  next paragraphs.
      Black  bear (Ursus americanus)  are  also  recovering in  the deciduous  forest
 region.   Omnivorous  animals which thrive  on  the  absence  of human contact  are
 showing  the greatest  recovery  in back country forests of the eastern mountains.
 One of the  healthier  populations exists in the Great Smokey Mountains National
 Park,  an  area of 1865  km .  The maximum area considered, 10.0 km2, is only
 0.5 percent  of  that area, which could not be  construed to be significant.
 Fringe populations might be impacted more heavily, however.
     The eastern cougar  (Felis concolor cougarl is also recovering from former
population declines and is being sighted with increasing frequency.  Home
                                    3-25

-------
ranges are not well-known, but  it appears certain that  a viable population
requires areas many times  larger than the largest cleanup area considered.
Thus, habitat reduction  is not  expected to be a factor.  Potentially more im-
portant is the effect that closure might have on migratory routes.  Large
predators must travel great distances hunting prey or risk depletion of their
food source.  In the eastern United States, that usually means passing through
frequent narrow corridors between heavily populated areas.  Closure of those
corridors could seriously restrict the home range of cougars and affect their
recovery.  Impact on the species could be serious.

     (4-1)  Asphalt Hard-Surface Stabilization
     Paving a spill area with asphalt represents the most severe impact con-
sidered so far.  Any plant invasion or animal utilization of the area would be
delayed until the asphalt begins to decompose, possibly a decade after treat-
ment.  Establishment of a community would require fairly complete decomposi-
tion of the asphalt which would require further decades.  The impact of an
                            2
asphalt surface over 0.01 km  might be acceptable if the area originally was
of low value ecologically, but not if the area has the potential for produc-
tion of a rich community.  It is not possible to consider acceptable a paving
               2
cover of 0.1  km  or more.  Even the most disruptive of the high priority tech-
niques, soil removal or burial, would have significantly lower impact.

     (4-2)   Concrete Hard-Surface Stabilization
     The effects of paving with concrete are qualitatively the same as with
asphalt and the primary evaluation is the same.   Since concrete decomposes
more slowly than asphalt, the effects would be longer lived and thus less
acceptable than those of asphalt.

     (5-0)   Application of Sewage Sludge
     Application of sewage sludge can replace organic matter and plant  nutri-
ents lost in vegetation removal and help balance soil pH for better nutrient
availability.  * '   '     The result is more rapid  surface stabilization  and
encouragement of plant  growth.   The amendment is not  without problems,  however.
Municipal sewage commonly contains  heavy metals  which can  be toxic,  particu-
larly to higher organisms.   A potential  exists for impacts  on wildlife  and
food-chains  leading to  man.   Secondly,  a potential  exists  for saturation  of

                                    3-26

-------
  the soil  cation exchange complex with excess ions  leaching into streams.   This
  problem has  been demonstrated for liquid effluent38  but  could become a problem
  with sludge,  as well.   On graded land sludge could reduce runoff and erosion
  by  increasing infiltration,  which would  speed recovery.

       (6-1)  High Pressure Washing (<3 mm)
       Outside  the scope  of work of this study.

       (6-2)  Flooding to  SO cm
       Outside  the  scope of  work of this study.

       (7-0)  Soil Amendments Added
      Outside  the  scope of work of this study.

 3.5  RECOVERY FOLLOWING CLEANUP

 3.5.1  Irreversible Changes
      In the context of geological time, no changes  are irreversible.   Hard
 surface stabilization would be expected to produce  changes which would persist
 long enough to be considered as irreversible in a practical sense.  All other
 treatments fall  within  the range  of experience in the deciduous forest region
 for  which  revegetation  has occurred.

 3.5.2  Rates of  Recovery

      The deciduous forest  biome is probably  the most  easily revegetated of the
 various native terrestrial  land types  considered.   This results  primarily
 from the fact  that rainfall  is both abundant  and  well-distributed through  the
 year.  Thus, given a  suitable seedbed,  plants  face  a  minimum of  factors  limit-
 ing  establishment.  If theseedbed  is not  suitable,  the abundant  rainfall and
 moderate temperatures provide for  more  rapid weathering of  the substrate and
 creation of seedbed.  Smith, et al.,43  actually found increased  levels of N,
 P, K, Mg, and  Ca  in leaves  of sassafras grown on a  100-year-old  iron-mine spoil
 in West Virginia, when compared to contiguous soils.
     Figure 3-1 illustrates a series of characteristics during recovery from a
major disruption, without reclamation  (A)  and with reclamation (B).   The

                                    3-27

-------
                                                                   B
               01 NATURAL RECOVERY



             HERB  |  SHRUB  |  TREE STAGES
GENERALIZED TYPICAL RECLAMATION
                                                •Z. LU
                                                Q O
                                                 . uj
   CO
   LU
   oc
   o
   u.

   CO


   O


   o

   o
   LU
   Q
  o
            k
                                              CO
                                              <
                                              a
                                              o
                                              oc
                                              UJ
            to 1
                       10
                                 100 200
            NO DATA


       COMPOSITE FEASIBLE
                                                                  10
                                                                            100 200
        •Clear Cutting (1-2), Stumping and Grubbing (1-3), Plowing (1-5, 1-6), Shallow Soil Burial (1-7)



Figure 3-1.   Natural  recovery (0-1) and  generalized  typical reclamation of

               deciduous forest following  cleanup (after  Whittaker4^).
                                       3-28

-------
  processes are generalized and thus do not relate to a specific location.  The
  major disruptions include severe vegetation removal, surface alteration, soil
  removal,  plowing and soil fin.   The time units are approximate with variability
  resulting from local differences within treatments and between treatment differ-
  ences.  The ordinate CY-axis)  is without units either because it  is  dimension-
  less (e.g.,  homeostasis)  or site specific (e.g.,  primary productivity,  species
  diversity).   See Appendix B for  further use.
       Species  diversity  increases from  the bare-ground  stage  to some  maximum
  occurring  during a transition  from one major  stage  in  succession  to  another.
  In  the example shown, the maximum occurs  during transition from the  herb  stage
  to  the shrub  stage,  approximately 15 years after  the sere begins. In other suc-
  cessions,   the maximum diversity  might  occur at the  transition  from the shrub
  to  tree stage (approximately SO  years)  or as the  climax forest replaces a sub-
  climax forest (perhaps 150 years).  Reclamations  seed a mixture of plants,
 particularly legumes such as sericea lespedeza and black locust.  Seeding'
 creates a  diversity which may be greater than early serai stages but which,
 through competition,  do not yield to the waves of early invaders typical of
 early successions.  Thus,  species diversity may remain low for a number of
 years.  The highly diverse old-field  stage may be essentially bypassed with
 the  result  that  the maximum diversity obtained may be less than the  maximum
 attained by a natural sere.   The  end  point is  the  same,  however,  and  a climax
 forest resulting  from a  reclamation-based sere is  presumed to be similar  to
 the  climax  resulting  from  a  natural sere.
      Cover  increases  gradually  during natural  successions, becoming more  rapid
 as trees replace  the  earlier  stages.  Mature forests with  their distinct  layers
 produce the greatest  cover of the sere.   Reclamations are  similar  except that
 greater initial cover is possible by seeding, but  cover may not increase until
 trees  are able to  overcome the early cover species.
     Primary productivity  generally increases with community age, reaching its
maximum in   the mature forest.  The rate of change may not be constant, however
but depend  upon the rate of species replacement.  Thus, if a persistent serai  '
stage is established,  the primary production rate may be essentially constant
for several  years.
                                    3-29

-------
      In general, homeostasis increases with community development with the
 climax forest representing the most stable state (generalized typical reclama-
 tion curve).  In some instances, shrub stages attain a high state of homeostasis
 (natural recovery curve) and persist for several decades.
      Erodability increases sharply with a perturbation but declines as vegeta-
 tive cover protects the soil.  During natural successions, erodability remains
 higher in early stages than the later stages, although the most rapid decline
 occurs during the first few years.   The main point  of reclamation is to quickly
 reduce erosion.   By seeding directly to a grass-shrub stage, the first several
 years of succession are compressed  into 2 or 3.   Consequently,  erodability
 drops sharply as cover is established and declines  in accordance with increased
 cover thereafter.

 3.5.3  Succession  Stages Following  Cleanup

      3.5.3.1  First Year--
      During  the  first  year  following treatment,  the  clearcut forests would begin
 to produce sprouts  from the remaining  stumps  (Figure 3-2A).   The spring  flowering
 herbs  of  the former forest  should reappear as  usual.   The opening of the  forest
 would  also permit the  invasion  of pioneer  plant  species  from outside with rag-
 weeds  (Ambrosia  spp.)  and other annuals being  among  the  first invaders.   Peren-
 nial grasses,  such  as  broomsedge (Andropogon virginicus),  and even woody  plants
 may also  invade  but  typically do not become conspicuous.   With the exception  of
 the sprouts,  areas  which have been stumped and grubbed  (Figure 3-2B),  scraped,
 graded, plowed,  or  had  soil added or removed should  experience a  similar  first
 year.  Cover  would  remain low on all treatments but  especially on those leaving
 the poorest  seedbed, such as removal of the shallow  root  zone or burial with
 up to 100 cm  soil.  No plant invasion could be expected on soils with hard sur-
 face stabilization.
     Animal use of first year successions would be limited.  Insect herbivores
 could be expected and, consequently, avian insectivores.  The low productivity
could not support a large consumer  biomass, however.   Cover on first year fields
is too low to support resident bird  populations and  the small mammals found
there are transients, as well.  Large mammals would  not find the area attrac-
tive.
                                    3-30

-------
       Areas around which barriers were erected should not be visibly different
  after one year.  Soils and plant cover have no visible impact following erec-
  tion of barriers (Figure 3-3), or after flooding with a few era of water.
       On denuded areas receiving sewage sludge, first year germination success
  and growth would be expected to exceed that of bare ground.  The greater cover
  would support a larger consumer bioraass but no vertebrates would be expected
  to become residents during the first year.
       3.5.3.2   Fifth Year-
       Five years following  cleanup,  most  procedures  should  be  represented by a
  stable  herbaceous coramunity.   In  central  Tennessee,39  similar  aged  fields  were
  dominated by  broomsedge  (Andropogon  virginicusl and  aster  (Aster  pilosusl.   In
  southern  Illinois,   broomsedge  was also dominant.  Woody vegetation invades by
  the  fifth  year  and  includes persimmon  CDiospyros virginianal.  sassafras  (Sassa-
  fras albidum).  elm  (Ulmus alataj . eastern red cedar  (Juniperus virginiana^!^
  sumac (Rhus spp.).   Ground cover is high, about 119 percent in the Illinois
  example, but the woody vegetation contributes only a small fraction of the  cover.
      Where disturbance was less than that represented by post-agricultural
  successions, such as clear cutting,  recovery would be more advanced.   Root and
  stump sprouting could produce a dense copse woodland in five years which might
 prove to be attractive to deer and other forest-edge wildlife.   Herbaceous
 groundcover should remain dense because canopy closure  would not be expected
 to occur within  five years.   If the  treatment had  been  to  simply erect  barriers
 no visible differences would  be expected  between treated and untreated  areas.
      Soil  stabilization by  compaction,  binding agents,  or flocculating  agents
 would probably no  longer  be distinguishable  from non-stabilized cleanups  which
 left  bare  ground.  Hard surface  stabilizations would  not be expected to have
 broken down or decomposed in five years and,  consequently, there would be no
 recovery  (Figure 3-3).
      3.5.3.3  Tenth Year--
     For the cleanups which left bare ground, the tenth year would be quite
 similar to the fifth  year, differing primarily in the growth of the woody
vegetation.  More individuals of the same shrub species  might also be expected
The characteristic old-field/shrub  stage is quite attractive to wildlife,  how-
ever.   The clear-cut  forest  might achieve canopy closure by  the tenth  year
                                    3-31

-------
                                                                HYPOTHETICAL DECIDUOUS FOREST
co
i
CO
                 8
           CJ
            1
           ro
 »-» fD
 I  O
 ro o

   fD
 f\i ^c
 =3 <<
 o.
   o
 U) -h
 d-
 c= a.
 3 fD

"2.2.    g
 13' a!    8
tQ C
   O

 13 t/1
 Q.
   -h
CQ O
 -s -s

 O" to
 CT r+
        IQ O
           —•   
-------
CO
I

CO
CO
             n>

             CO

             CO
          fa  ro
          -5  o
          Q. o
             <
          CO  fO
 -b
 O* O
 O -*>
 fD
   CL
 to n>
 rt- O
 G» -i.
 cr Q.
 -«• c
 — • o
 -••c
 N to
 Oi
 r+ -*>
 -i. o
 O -S
 =5 0)
•  to
            -b
            O
                          EROOABILITY
                                                         HYPOTHETICAL DECIDUOUS FOREST


                                          HOMEOSTASIS            PRIMARY PROD.
                                                                                                         COVER
                 8 —
c          v
                                                                                                                     SP. DIVERSITY

                                                                                                                     (VEGETATION)
                                                                                                                               6
                                                                                                                               x
                                                                                                                               >
                                                                                                                               3}
                                                                                                                               O
                                                                                                                               CO
                                                                                                                                                  o
                                                                                                                                                  m
                                                                                                                                                  CO
                                                                                                                                                  N
            o

-------
 if  sprouting had been vigorous., and, if so, the herb stratum would be declining.
 Hard  surface stabilizations, particularly asphalt, could be starting to break
 up  with pioneer species starting to invade cracks.  Reclaimed  land should have
 a mixture of trees, shrubs, and herbaceous vegetation which is attractive to
 wildlife.
      3.5.3.4   Fiftieth Year—
      In southern Illinois,  the local sere lacks a pine-forest stage.  Conse-
 quently, 50 years of recovery would not likely produce a closed canopy.  After
 30  years, 22 species of trees were present but so was a perennial grass (Andro-
 pogon virginicus) layer, which indicates a high light level at the herb stratum.
 Seedlings of the climax oaks (Quercus imbricaria and Q. velutina) had begun to
 appear, indicating that a mature forest was less than 100 years away.
      In the southeastern states, a subclimax pine forest typically replaces
 the shrub stage and can produce canopy closure within 30 years from the begin-
                   33
 ning  of succession.    When canopy closure by conifers occurs, light to the
 forest floor is greatly diminished and the herb layer is largely eliminated.
 Consumers do not utilize pine forests to any great extent and, consequently,
 species diversity tends to be low.
     Stumping and grubbing, surface alterations, soil removals, plowing, stabi-
 lizations, and soil fills should all be as described above.  Clear-cut forests
 would likely have produced a closed-canopy, deciduous forest and be more ad-
 vanced than the bare-ground treatments.   Hard surface stabilizations would not
 be sufficiently decomposed to permit normal community development to progress
 although individual plants might be doing well.   Reclaimed land could be ex-
 pected to be in closed-canopy forest  but not in climax.
     3.5.3.5  100 Years After Cleanup--
     Extrapolating from conditions at 50 years,  forests lacking a pine stage
 should be approaching climax,  although full growth and development would not
 be expected.   In forests with a pine stage, 100 years marks the transition
 from pine to deciduous forest.   In many seres,  this is the point of maximum
 diversity and productivity, both declining into the climax.   This point may
 contain the greatest density of breeding birds,  if not the greatest diversity.21
This peak in diversity,  density,  and biomass  extends  even to the populations
                                   28
of decomposers in the forest floor.

                                    3-34

-------
       One-hundred-year-old reclamations are not known but it seems reasonable
  to assume that the accelerated development during the first, decade would no
  longer be evident  and reclamations  would be similar to natural  revegetations.
  Neither are  100-year-old,  abandoned pavements  known.   It is assumed that decay
  of either asphalt  or  concrete  would be fairly  complete after a  century  and  re-
  vegetation would be proceeding.   Vegetational  composition  and developmental
  stage  is  not predictable,  however,  because of  the spatial  and temporal  heter-
  ogeneity  which  would  result  from  the breakup pattern  of the  pavements.   Where
  diverse microsites occur  on  mine  spoils,  the revegetational  patterns do  not
  conform to theoretical seres.9
      3.5.3.6  Climax--
      In deciduous forest,  climax  is more  often a  theoretical  concept than an
  observable fact.  Using the term  loosely  to mean  forest dominated by a species
  complex similar to the climatic potential, whether or not mature, climax forest
 might be expected to occur 150 to 200 years following the disturbance.   Assuming
 that the^forest disturbed was climax, the new climax may or may not be the same.
 Woodwell   has  suggested that man-made (anthropogenic) influences are changing
 the basic physical  and chemical properties of the earth and may interfere with
 ecosystem development  as we have understood it.  If so, it is understood that
 these influences,  such as acid  precipitation,  will retard rather than accelerate
 ecosystem recovery.

      Deciduous  forest  is  an extremely diverse  biome with numerous local  seres
 leading to a  few climax types.   In generality,  natural successions from  bare
 ground  go  through a  series of defined  stages:   annual  weed,  perennial herbs,
 shrubs,  trees.   Often,  one or more subclimax forest stages  precede climax
 forest.  Occasionally,  metastable  subclimax communities  apparently resist
 invasion by trees,34 possibly through allelopathy.16   In  other cases, periodic  .
 perturbations, such as  fire,  lead  to subclimax  stages.36  All  of  these extend
 the  full recovery time  of  ecosystems.

 3.6  QUANTITATIVE ASSESSMENT OF CLEANUP IMPACTS
     Of the cleanup procedures specified,  stumping and grubbing (1-3), scraping
and grading (1-4), soil removal  (2-1  to 2-8), plowing  (1-5, 1-6), and cover
with uncontaminated soil (1-7, 1-8) are similar in that the rate and pattern
of recovery will be similar,  ni ffpr^nr^c  ,,v,i ^      u
                              uitrerences  which  may be apparent during the
                                     3-35

-------
first few years should no longer be apparent after a decade or so.   Less
severe disturbances lead to more rapid recovery, more severe ones recover
more slowly.

     Because of favorable moisture and temperature conditions, deciduous
forests are reclaimed relatively easily.   The growth form of mature communities
and their homeostatic mechanisms, prevent revegetation of climax species, di-
rectly.  Trees of the climax forest typically do not do well on freshly re-
claimed surfaces and the forest floor, which retains nutrients and retards
erosion, requires years to develop.  Consequently, reclamations can only
shorten recovery time by about a decade,  which represents only about 5 percent
of the time required to replace a "climax" forest.  However, reclamation can
provide early vegetative cover over the entire area immediately which amelio-
rates erosive degradation in the cleanup area and gullying outside the spill
zone.

     Early stages of natural successions are low in diversity and productivity.
Wildlife is not found on them.  Reclamations attempt to seed directly to the
grass-shrub stage which can establish valuable wildlife habitat within 5 years.
Consequently, the benefits gained from reclamations go beyond simply reducing
erosion or slightly shortening the sere.

3.6.1  Impact Assessment
     Cleanup procedures in deciduous forest are indexed (See Appendix B) in
Table 3-2 as impacts on ecosystems, with 0 equal to no impact and NA equal to
complete loss of the ecosystem.  The index estimates include consideration
of the physical, biological and sociological impacts of the treatments on the
nearby vicinity of the cleanup area.  There are several caveats which must be
observed.  First, the numerical rankings are highly subjective.  Secondly, as
explained below, most impact rankings for a given procedure in a given eco-
system increase as the area affected increases.  It is also likely that the
confidence decreases as the area increases.  Given enough time, tens of years
to centuries depending on the treatment, recovery to a condition paralleling
that of surrounding untreated areas would occur due to natural recovery.
Thus the index values in Table 3-2 will tend toward 0 with time.  Since the
relative impact assessment is time dependent, the time selected for comparison
was the first growing season after cleanup treatment.

                                      3-36

-------
                     Table 3-2.
                          Predicted recovery  index  in  deciduous forest  for cleanup treatments
                          (see Appendix  3).
Treatment
0-1 Natural Rehabilitation
1-1 Chemical Stabilization
1-2 Remove Vegetation by Clear Cutting
1-3 Remove Stumps and Grub Roots
1-4 Surface Scraping and Grading (<5 cm)
1-5 Shallow Plowing (--10 cm)
1-6 Deep Plowing (10 to 40 cm)
1-7 Soil Cover (• 25 cm)
1-8 Soil Burial (25 to 100 cm)

2-1 Remove Plow Layer Soil (10 cm)
2-2 Remove Shallow Root Zone Soil (--40 cm)
2-3 Scrape Surface, Remove (<5 cm), Mechanically Stabilize
2-4 Remove Plow Layer (10 cm), Mechanically Stabilize
2-5 Remove Shallow Root Zone (<40 cm), Mechanically Stabilize
2-6 Scrape Surface, Remove (<5 cm), Chemically Stabilize
2-7 Remove Plow Layer (10 cm), Chemically Stabilize
2-8 Remove Shallow Root Zone (40 cm), Chemically Stabilize
3-1 Barriers to Exclude People
3-2 Barriers to Exclude Large Animals
3-3 Barriers to Exclude Large and Small Animals
4-1 Asphalt Hard-Surface Stabilization
4-2 Concrete Hard-Surface Stabilization
5-0 Sewage Sludge Application
6-1 Washing (-3 mm)
6-2 Flooding (3 to 30 cm)
7-0 Soil Amendments Added
Area Disturbed
(kV)
0.01
0
0
2
4
4
4
4

*
5
ft
4
5
3
4
f
0
0
0
NA
MA
0
*
*
*
0.1
0
0
2
4
4
4
4
5
**
5
a
4
5
*
3
4
t>
0
0
0
NA
NA
0
*
*
*
1.0
0
0
2
4
4
4
4
5
a
5
tt
4
5
f
3
4
0
0
0
*
*
0
*
*
*
10.0
0
0
2
4
4
4
4
5
tt
5
«
4
3
4
tt
0
0
0
*
*
0
*
*
*
Assumptions
0 = no significant impact. Manual cleanup done
Surface applications by spray, broadcast
Climax ^100 years by resprouting
Seed sources intact, climax 100 years (N) +
Plants removed first
Plants removed first
Plants removed first
Plants removed, climax can be >100 years
Can require rehabilitation, then -100 years
Plants removed first, climax >100 years (N+l)+
Can require rehabilitation, then .100 years
Vegetation removed first, climax near 100 years
Plants removed first, climax -100 yrs;rehab --10 km2
Rehabilitation required
Plants removed first; early regrowth
Normal climax 100 years
Rehabilitation probable, then 100 years
No climax retrogression
No climax retrogression
No climax retrogression
Excessive recovery delay (N+2)
Excessive recovery delay (N+2)
Vegetation remains in place.




GO
I
OJ
-t-Normal climax in .100 years; methods of indexing in Appendix B,  symbols defined (5,
                                                                            i», NA) and Guide Numbers (N,  N+l, N+2)  in Table B-5.

-------
     The rationale for a particular index value is discussed in Section 3.4
and summarized here.   Intermediate impact rankings are explained in Appendix
B.  Impact of clear cutting (1-2) on deciduous forest is related to the time
required for succession, hence the possibility of root sprouting lowers the
impact relative to stumping and grubbing (1-4) .   Impacts obviously increase
with area.  Scraping, burial,  plowing and soil removal were all considered to
be impacts in addition to forest removal, for trees and shrubs would generally
have to be removed before the other procedures could be applied.  Scraping and
grading (1-4) are not considered to be any more disruptive than stumping and
grubbing  (1-3) since both preparation and the results are similar.  Shallow
plowing (1-5), over a large area, would be a slightly greater impact since
herbaceous plants and seeds would be buried.  Deep plowing (1-6), if possible,
on many forest soils would bring lower quality subsoil to the surface and re-
sult in a degraded seed bed.  Removal of the top 10 cm (2-4) would remove local
seeds and require plant invasion from outside of the area.  Hence, resulting
in a greater impact and an increasing impact with areal increase.  Thinner
topsoils would also be removed.  Removal of up to 40 cm would remove virtually
all forest topsoils and leave poor seed beds.  Similar arguments apply to
burials.  Barriers to exclude people (3-1) would have no impact while barriers
to exclude animals (3-2, 2-3)  might have slight impact over large areas.  The
impacts to animal populations themselves might be greater.  Hard-surface sta-
bilization (4-1, 4-2) is effectively a complete, permanent loss.

3.6.2  Recovery assessment
     Deciduous trees require decades to reach canopy height.  Consequently,
any area where trees are removed will not be soon restored to its former state.
Furthermore, climax species have habitat requirements which differ significantly
from conditions prevailing on bare ground.  Climax species are not early in-
vaders and often cannot even be planted on barren sites.  Table 3-2 is based
on estimates of recovery times following various treatments and gives Recovery
Indices.  The ranges of time for expected recovery are quite large because
the factors which affect them  (e.g., soil factors, precipitation cycles, etc.)
have potentially greater impact on recovery times than do the differences
between the impacts of various cleanup treatments.  Consequently, while deep
plowing (1-6) has a greater perceived impact than shallow plowing (1-5), for
                                    3-38

-------
 example, the seres are sufficiently long and local variability is sufficiently
 great that climax recovery times cannot be effectively differentiated.

      Minimum time to recovery of climax appears to be around a century.  That
 is the time required to pass through the various serai stages and create micro-
 climatic conditions favorable to climax species.  The exception is where the
 trees were clearcut but the forest was otherwise not disturbed (1-2).   Sprout-
 ing may result in more rapid recovery, but even that is limited by the growth
 rate of the trees and the degree of invasion by non-climax species.   Treatments
 1-3, 1-4,  1-5, and 1-6 all require a full  sere and the local variability is
 probably greater than the differences  between them.   Shallow soil cover (1-7)
 and shallow soil removal  (2-1)  may result  in conditions similar to the above
 (optimally) or could possibly be more  severe perturbations.   If unassisted,
 natural  revegetation is possible,  time to  climax should be similar to  the above,
 If not,  reclamation may be indicated.

      Reclamation might  be necessary where  depth of soil  applied is great  (1-8)
 or the depth  of  soil  removed  is  great  (2-2),  even  if stabilizers  are used (2-5,
 2-8).   If not  reclaimed,  decades  could be  added to the recovery times.  With
 reclamation,  time to  climax should  be  within  the  range of  natural  seres on
 more favorable sites  (1-3,  1-4,  1-5, 1-6).   In  general,  stabilizers  (2-3, 2-4,
 2-5,  2-6, 2-7, 2-8) would be  expected  to greatly  influence early  succession
 but  their effects  would not be expected to be discernible  by  the  time  climax
 is approached.

     Hard-surface  stabilization  (4-1,  4-2) would significantly  retard the onset
 of succession and  lengthen the potential minimum time  for  recovery.  It is  not
 clear if any effect would be discernible on a sere requiring  200 or more years.

     Barriers  (3-1, 3-2,  3-3)  do not disturb vegetation in the  system and have
no recovery time associated with them.   Wildlife is discussed in Part III.

3.7  CONCLUSIONS
     The preferred treatment is the "no action" option of erecting barriers
 (3-1, 3-2,  3-3).   The advantage is that it  allows time for careful evaluation
                                    3-39

-------
of what further steps might really be required.   Particularly where large areas
might be contaminated, the effectiveness and impact of other treatments could
be evaluated experimentally.

     Hard-surface stabilization offers no apparent benefits and should not be
considered.

     As a generality, there are no significant advantages or disadvantages
obviously associated with any of the other treatments.   The decision to actively
reclaim or not will be influenced by the area disturbed and location.   A small
        2
(0.01 km ), isolated area might not need to be reclaimed while a similar area
                                                                              2
having significant population exposure probably would need to be.   Any 10.0 km
area disturbed would probably require reclamation of some kind.   Beyond these
considerations, local experience with road and other construction will influ-
ence decisions on need for stabilizers and/or reclamation.
                                    3-40

-------
  3.8  DECIDUOUS FOREST REFERENCES



  1.  Anonymous.  Reclamation of Strip Mined  Land.   In- J  W   Lewis  fed  )
          Illinois Blue Book 1971-1972, State of  Illinois, Springfield, ' 1972 .
          pp . 2- 16 .


  2.  Bard, G. E.  Secondary Succession on the Piedmont of New Jersey    Ecol
          Monogr. 22:195-215, 1952.


  3.  Bazzaz,  F. A.   Succession on Abandoned Fields  in the Shawnee Hills
          Southern Illinois, Ecology 49:924-936, 1968.


  4.  Bazzaz,  F. A.   Plant Species Diversity in Old-Field Successional Eco-
          systems in Southern Illinois, Ecology 56:485-488, 1975.


  5.  Beckwith,  S.  A.   Ecological Succession on Abandoned Farm Lands and its
          Relationship to  Wildlife Management, Ecol.  Monogr.  24:349-376, 1954.


  6.  Billings,  W.  D.   The Structure and Development  of Old-Field Shortleaf
          Pine Stands  and  Certain Associated Physical Properties of the Soil
          Ecol.  Monogr.  8:437-499,  1938.


  7.   Bornebusch, C. H.  The Fauna  of Forest Soil,  Forst.  ForsVaes,  Danm.
                  j  j. y o u •
  8.   Braun,  E.  L.   Deciduous  Forests  of  Eastern  North America,  Blakiston
          Philadelphia,  1950.   596  pp.


  9.   Brewer  R. and E.  D. Triner.  Vegetational  Features  of  Some  Strip-Mined
          Land in Perry  County,  Illinois, Trans.  Illinois  Acad.  Sci.  48:73-84,
          •1. */ O \j •


10.   Buckman, H. 0. and N. C.  Brady.  The Nature and  Properties of Soils
         Macmillan, New York,  1969.   653 pp.


11.   Byrnes, W.  R. and J. H. Miller.  Natural Revegetation and Cast  Overburden
         Properties of Surface-Mined Coal Lands in Southern  Indiana.  In-
         Ecology and Reclamation of Devastated Land, Gordon  and Breach Sci
         Publ.,  New York, 1973.  pp. 285-305.


12.  Clark,  T.  D.   Kentucky, Land of Contrast.  Harper and Row, New  York
         1968.   304 pp.                                                   *


13.  Croxton  W   C.  Revegetation of Illinois Coal  Stripped  Lands.   Ecology 9:
         i Di> — 1/«>,  i y 
-------
 14.   Curtis,  J.  T.   The Vegetation of Wisconsin.   Univ.  Wisconsin Press,
          Madison,  1959.   657  pp.

 15.   Dickerson,  J.  A.  and W.  E.  Sopper.   The  Effect  of Irrigation with Munici-
          pal  Sewage Effluent  and  Sludge  on  Selected  Trees,  Grasses and Legumes
          Planted in Bituminous  Strip  Mine Spoil.   Penn.  State University
          Research  Briefs 7:1-4,  1973.

 16.   Gant,  R.  E. and E.  E.  C. Clebsch.   The Allelopathic Influences of
          Sassafras  albidum in Old-Field  Succession in  Tennessee.  Ecology
          56:604-615, 1975.

 17.   Henry, J. D.  and  J.  M. A.  Swan.   Reconstructing Forest  History from  Live
          and  Dead  Plant  Material  — an Approach to the Study of  Forest Succes-
          sion  in Southwest  New  Hampshire.   Ecology 55:772-783, 1974.

 18.   Holland,  R. F.  Wildlife Benefits from Strip-Mine Reclamation.   In:  Ecology
          and  Reclamation of Devastated Land,  Gordon  and  Breach Sci.  Publ    New
          York, 1973.

 19.   Hosner, J.  F.  and  D.  L. Graney.   The Relative Growth of Three Forest
          Tree  Species  on Soils Associated with Different Successional  Stages
          in Virginia.  Adm. Midi.  Nat. 84:418-427, 1970.

 20.   Jackson,  J. R.  and  R.  W. Willemsen.  Allelopathy  in the First  Stages
          of Secondary  Succession on the  Piedmont of  New  Jersey.   Am  j' Bot
          63:1015-1023,  1976.

 21.   Johnston, D. W. and  E. P. Odum.   Breading Bird  Populations  in  Relation
          to Plant Succession on the Piedmont of Georgia.  Ecology 37-50-62
          1956.

 22.   Keever, G.  Causes  of  Succession on Old-Fields  of the Piedmont, North
          Carolina.  Ecol. Monogr.  20:229-250, 1950.

 23.   Komarek, E.  V.  Effects of Fire on Temperate  Forests and Related  Eco-
          systems:  Southeastern United States.  In:  T.  T. Kozlowski, and
          C. E. Ahlgren  (eds.), Fire and Ecosystems,  Academic Press, 1974
          542 pp.

 24.   Kozlowski, T.  T. and C. E.  Ahlgren  (eds.).   Fire and Ecosystems, New
         York, Academic Press, 1974.  542 pp.

 25.   Kricher, J.  C.  Summer Bird Species Diversity in Relation to Secondary
         Succession on the New Jersey Piedmont.   Am.  Midi.  Nat 89-121-137
          1973.

26.  Kuchler, A.  W.  Potential Natural Vegetation  of  the Conterminous United
         States.  Amer.  Geog. Soc., New York: Special  Publ.  No. 36, 1964.

27.  Lejcher,  T.  R.  and S. H.  Kunkle.   Restoration of Acid Spoil  Banks with
         Treated  Sewage Sludge.   In: Recycling Treated  Municipal  Wastewater
         and Sludge Through Forest and Cropland.  Penn.  State University Press,
         University Park, 1973.
                                     3-42

-------
28.  McBrayer, J. F., J. M. Ferris, L. J. Metz, C. S. Gist, B. W. Cornaby,
         Y. Kitazawa, T. Kitazawa, J. G. Wern, G. W. Krantz, and H. Jensen.
         Decomposer  Invertebrate Populations in U.S. Forest Biomes.  Peda-
         biologia 17:89-96, 1977.

29.  McDougall, W. B.  Forests and Soils of Vermillion County, Illinois, with
         Special Reference to the "Striplands."  Ecology 6:372-379, 1925.

30.  Morrison, R. G. and G. A. Yarranton.  Diversity, Richness, and Evenness
         During a Primary Sand Dune Succession at Grand Bend, Ontario.  Can.
         J. Bot. 51:2401-2411, 1973.

31.  Morrison, R. G. and G. A. Yarranton.  Vegetational Heterogeneity During
         a Primary Sand Dune Succession.  Can. J. Bot. 52:397-410, 1974.

32.  Nicholson, S. A. and C. D. Monk.  Plant Species Diversity in Old-Field
         Succession on the Georgia Piedmont.  Ecology 55: 1075-1085, 1974.

33.  Nicholson, S. A. and C. D. Monk.  Changes in Several Community Charac-
         teristics Associated with Forest Formation in Secondary Succession.
         Am. Midi. Nat. 93: 302-310, 1975.

34.  Niering, W. A. and R. H. Goodwin.  Creation of Relatively Stable Shrublands
         with Herbicides: Arresting "Succession" on Rights-of-Way and Pasture-
         land.  Ecology 55:784-795, 1974.

35.  Odum, E. P.  The Strategy of Ecosystem Development.  Science 164:262-270,
         1969.

36.  Odum, E. P.  Fundamentals of Ecology.  Philadelphia, Saunders, 1971.  574 pp

37.  Pearson, P. G.  Small Mammals and Old-Field Succession on the Piedmont of
         New Jersey.  Ecology 40:249-255, 1959.

38.  Perkins, M. A., C. R. Goldman, and R. L. Leonard.  Residual Nutrient
         Discharge in Streamwaters Influenced by Sewage Effluent Spraying.
         Ecology 56:453-460, 1975.

39.  Quarterman, E.  Early Plant Succession on Abandoned Cropland in the
         Central Basin of Tennessee.  Ecology 38:300-309, 1957.

40.  Riley, C. V.  Ecosystem Development on Coal Surfaced-Mined Lands, 1918-75.
         In: J. Cairns and E. E.  Herricks (eds.) Recovery and Restoration of
         Damaged Ecosystems.  Univ. Press of Va., Charlottesville, 1977.
         pp. 303-346.

41.  Scanlon, D. H., C. Duggan, and S. D. Bean.  Evaluation of Municipal
         Compost for Strip Mine Reclamation.   Compost. Sci.  14:4-8, 1973.

42.  Sly, G. R.  Small Mammal Succession on Strip-Mined Land in Vigo County,
         Indiana. Am. Midi.  Nat.  95:267, 1976.

43.  Smith, R. M.,  E. H.  Tyron, and E. H. Tyner.  Soil Development on Mine
         Spoil.  Bull.  604,  W.  Va.  Univ. Agr. Expr.  Stn., Mortantown,  1971.  •>


                                    3-43

-------
44.  Viro, P. J.  Effects of Forest Fire on Soil.  In: T. T. Kozlowski and
         C. E. Ahlgren (eds.), Fire and Ecosystems, Academic Press, New York
         1974.  pp. 7-45.

45.  Vitousek, P. M. and W. A. Reiners.  Ecosystem Succession and Nutrient
         Retention:  A Hypothesis.  BioScience 25:376-381, 1975.

46.  Whittaker, R. H.  Communities and Ecosystems.  New York, Macmillan
         1975.  385 pp.

47.  Woodwell, G. M.  Success, Succession, and Adam Smith.  BioScience 24-
         81-87, .1974.
                                    3-44

-------
                                    CHAPTER 4
                               CONIFEROUS FORESTS
 4.1   OVERVIEW
      The coniferous forests of the United States occur in all sections of the
 Nation except for the Great Plains.   In the eastern portion of the country,  the
 distribution of conifers is considerably more continuous  and less complex than
 in the mountain and far western states.   In the  East,  zonation tends  to be pri-
 marily a function of latitude,  while  in the West,  altitude and slope  aspect
 become major factors.   In the  East, it  is likely that  cleanup procedures would
 be limited  to a single forest  type, but  in the West, one  or more  zones may be
 involved if a spill occurs in  a mountainous area.   It  is  not possible to con-
 sider the coniferous forests apart from mountainous regions,  so both  will  be
 discussed here,  as  will  the interface between the  coniferous and  tundra zones.
 The primary purpose of the first section of this Chapter  is to provide a rela-
 tively detailed overview of the major types of coniferous  forest  in the United
 States with respect to their distribution and major species associations.

      Costing    has  reviewed the climax vegetation  types of North  America and
 notes  that  there  is considerable variation  among vegetation maps  with regard
 to detail due  to  disagreement on the interpretations of climax relationships.
 Examples  of such  vegetation maps may be  found58'63*29.  More  recent maps have
                         42            38           ^1
 been published  by Little    and  Kuchler   .   Costing   believes  that studies of
 local  investigations of  vegetation are necessary for a comprehensive  understand-
 ing of the  bases  for the differing interpretations  found on vegetation maps.
 This statement  is no doubt  also  applicable  to most  descriptions of the distri-
 bution  of coniferous forest types, particularly  those found  in the more diverse
 mountainous areas of the country.  It is also important to note that  forest
 types  change in many areas  due  to shifts in land use and in natural succession
 processes, dating older maps.   For example, the harvesting of Douglas fir has
 resulted in the replacement of  large stands of that tree by western hemlock and
various hardwoods with a resultant net increase of western hardwood forest types
                                      4-1

-------
of nearly two million acres in the period of 1962-1970   .  The concept of
vegetational climax formations and their associations permits classification
on a regional basis; it assumes an acceptance of monoclimax theory which im-
plies the existence of pre and postclimax communities.   While this system  is
dated, it has been widely used in various forms in the  literature on  North
American vegetation and it will provide a basis for the  following description
of the coniferous forests of the United States.

     The coniferous forest has been divided into four climax formations in
North America:  Boreal, Subalpine, Montane, and Pacific  Coastal.  There are
also large areas of coniferous forest that are considered to be sub or post-
climax by this system.  Costing51 has characterized the  nature and distribu-
tion of these formations in some detail and the following descriptions are
largely adapted from his work.  Another system of nomenclature refers by the
major species associations and/or their geographical location:  northern conif-
erous forest (spruce—fir); northwestern coniferous forest (cedar—hemlock,
western larch—western white pine, Pacific Douglas fir—redwood); western  pine
forest (ponderosa pine—sugar pine, ponderosa—Douglas fir, lodgepole pine);
northeastern pine forest (Jack, red and white pines); and southeastern pine
forest (longleaf, loblolly, and slash pines). (See Figure 4-1.)

     The Boreal is the most continuous of the major forest formations with
respect to distribution.  In the western United States,   the various coniferous
formations appear in vegetation zones which occur as a function of altitude
and size of a mountain mass, the latter affecting the quantity of precipitation
which will occur on a given mountain range.  At a given  latitude, an isolated
mountain peak may not support stands of species having high moisture require-
ments that may be satisfied on a nearby mountain range of lower elevation but
greater mass.  Due to the prevailing winds from the west, significant vegeta-
tion differences occur between the western and eastern slopes of the mountain
ranges.   The character of the communities within a given zone also differs among
mountain ranges.   The Rocky Mountain and Sierra Nevada forest complexes differ
significantly and are discussed separately here.   The zonal climaxes occurring
within these complexes include:  Tundra, Subalpine,  Montane,  and Foothill  (not
discussed).
                                      4-2

-------
                                            t^-s^ 	~~~r-
                                              »   ^-_. '


                                                        & >-^   \   1 i^-^wftg&gm

                                                            *i%^=~:_=. ^ ™ ^^ ">" l • JfAvir1-LJ^ £?W~^;
                                                              I  K v>^^-/;i?Fi|i^S
                                                             "A^Vj/'l?  ^"" *
-------
4.1.1  Boreal Formation
     The Boreal formation, sometimes referred to as "taiga" in its northern
areas, forms a broad band across the North American continent, lying to the
south of the Tundra formation; it occurs as far south as the New England
states.  The climate in the Boreal forest region is almost as severe as that
of the Tundra, being characterized by a cool, short growing season and moder-
ate precipitation that is heaviest near the East Coast (approximately 40 inches)
and lightest in the western region and the Alaskan interior (10 to 12 inches).
The topography characteristic of the Boreal formation is largely the result of
glaciation and abounds with lakes, bogs, and muskegs.  Thin residual soils
overlay rock masses, or deep moraine and outwash occur along the southern boun-
dary of the formation.

     The Boreal formation is characterized by climax growth of white spruce
(Picea glauca) and balsam fir (Abies balsamea); this climax is best developed
along the St. Lawrence River Valley.  White spruce and balsam fir are found
scattered throughout the range and they are associated with paper birch (Betula
papyrifera) which may occur following fires in pure stands as a subclimax.
Black spruce  (Picea mariana) is frequently dominant north and west of Lake Su-
perior, and it is frequently found associated with tamarac (Larix laricina) as
climax species along the northern and far western transitions in high rocky
areas.  Further to the south, both species become associated with bogs.  To
the west, paper birch and Jack pine  (Pinus banksiana) occur as climax species,
and are subclimax nearer the center of the Boreal formation.  Jack pine and
black spruce form a subclimax in the western part of the range and Jack pine
also occurs extensively in areas of sand plains and gravel soils.

     The northern transition of the Boreal formation into the Tundra is irregu-
lar with the coniferous forest extending far into the Tundra region in sheltered
valleys while the Tundra is found on high ridges extending into the forested
                                       O£
area.  Indications were found by Griggs  , who studied the reasons underlying
the positioning of the Arctic timberline, which suggested that the timber line
in Alaska is advancing northward while it is retreating in portions of Canada.
The southern transition grades into grassland in the west and into deciduous
forest in the east.  Also, in Alaska, the Boreal coniferous forest merges with
the northwestern coastal forest.  The southern Boreal transition zone is marked
by pure stands of white pine  (Pinus strobus) from New England to Minnesota and

                                      4-4

-------
 and red pine (Pinus resinosa) in the Lake States; Jack pine also occurs in
 less favorable sites in the Lake States area.  In the eastern transition,
 spruce and fir may mix with hardwoods or grow in alternating stands.  In the
 Rockies,  Picea glauca is associated with Abies lasiocarpa.  The northern con-
 iferous forest extends southward on the higher Appalachian mountains as far
 as the Great Smoky Mountains of North Carolina.  There,  the growth form and
 species associations are similar to those found throughout the Boreal forma-
 tion,  but red spruce (Picea rubens) tends to replace white spruce from New
 Brunswick southward into New England.   Further south,  balsam fir is replaced
 by Fraser fir (Abies fraseri).   Therefore,  the dominant  species in the south-
 ern Appalachians are the ecological equivalents of those found elsewhere in
 the Boreal formation,  but they  may be distinguished taxonomically.   As the
 latitude  decreases,  the spruce-fir forest is limited to  increasingly higher
 elevations;  this association may be found virtually anywhere in the northern
 portions  of the  Boreal  formation,  but  only above approximately 5,000 feet in
 the Great Smoky  Mountains.   Spruce tends  not to occur  at the highest altitudes
 in the southern  range  and,  in its  absence,  fir becomes dominant51.

 4.1.2   Rocky Mountain  Forest Complex
     The  Rocky Mountain forest  complex  contains several  distinguishable  vege-
 tation zones that  extend as  far north as  Alberta and southward into northern
 Mexico, terminating  at  the  southern end of  the Sierra  Madre  range;  this  zona-
 tion extends westward to the eastern foothills of the  Sierra Nevada range  and
 the eastern  slopes of the Cascades, it  also  extends  eastward into the  Black
 Hills.  In addition  to  the major Rocky  Mountain Ranges,  the  complex encompasses
 the basin  and range  physiographic  province with its  numerous north-south ranges
 separated by desert  basins.   The zones  that  occur are  a  discontinuous  alpine
 zone on the  higher peaks, subalpine, montane and  woodland forest.   These zones
 are discontinuous in their distribution from range to  range  and the  full com-
plement of zones is not  necessarily present  in  each  of the ranges.   For  exam-
ple, in the northern portion of the complex, the  upper zones exist at  lower
altitudes and the lower  zones may not be present; each zone occurs at  increas-
 ing elevation as the latitude decreases.  As previously discussed, there are
significant differences  in zonation between north and south exposures and there
are marked differences in vegetation between the  eastern and western slopes of
the individual ranges within the system and for the entire system as a whole.

                                      4-5

-------
In mountain systems, narrow valleys tend to allow the dominants of a given
zone to extend downward into a lower zone while high dry ridges permit upward
finger-like projections of the dominants into continuous higher zones.  Drain-
age of cold from higher elevations can significantly disrupt zonal distribu-
tion patterns. *•  Whittaker and Niering°  have detailed the effects of alti-
tude and slope aspect upon vegetation in the Santa Catalina Mountains of
southern Arizona.   (See Figure 4-2.)  A review and synthesis of the factors
which produce and control the zonation of the vegetation in the Rocky Mountains
by Daubenmire   may be of particular interest in the determination of the prob-
ability of reestablishing forest growth after implementation of cleanup procedures
     The zonal climaxes which occur within the Rocky Mountain coniferous forest
complex have been grouped by Costing   as follows:
          Alpine zone
              Tundra climax
          Subalpine zone
              Engelmann spruce—Alpine fir climax
          Montain zone
              Douglas fir climax
              Ponderosa pine climax
          Foothills (Woodland zone)
              Pinon —Juniper climax
              Oak—Mountain Mahogany climax
     The vertical range of full development of each of the zones is approxi-
mately 2,000 feet,  with the foothill zone narrowing and disappearing at the
higher latitudes and each of the other zones being found at progressively
lower altitudes.  In the transitions between the zones, the species character-
istic of each zone become restricted to favorable habitats.  Near the upper
boundary, these species will be more common on ridges and south-facing slopes
where less moisture and higher temperatures occur; whereas in the lower tran-
sition, the species characteristic of a given zone will tend to occur in the
relatively cool, moist sites.  Under these conditions, species of one zone
may extend into another zone.  Precipitation tends to increase with altitude;
a linear increase of 4.94 per 1,000 feet of elevation was found between the
                                                         45            51
sagebrush and alpine zones on the Wasatch Plateau of Utah   and Costing   notes
that similar observations have been made elsewhere.
                                      4-6

-------
                 3000
                 2500
                 2000
                 1500
                1000-
                      10,OCK
                       900:
                       8000
                       7000
                       6000
                       5000
                      4000
r~ rv>7.x'.\' ^'^^^tfQiir^^^^ijS^^^^i^^^^. Engdmonn spruce
/.;: SUBALPINE FOREST^^..^^^l^^lS^^^:^^---

                                                          Jg£5 PiNH-pAK: "WOODLAND
                                                          :'~23™ .-I :?:•>.#£**.'£ v'..--:- *~*fe*
                                                                            ,   ---Emory ook
                        Chihuahuo pine    ' '"
                            Walnut" - -- ^ -"
                              .;-"/  Cypress
                            •^:T/^-rp^V^:$lp^r^:--V^^«
                            *fe3^Wl«a^S^Bfi«ffiiS5a^teiSfe^            rut»-^ ~
                               -Y«g«»»«   .. «• it«=p^a^p^.rn^ • ^a^By^s^^-* ..^. g g*.. .>
                   Elevation            .     _;
                                      Lower Slopes	
                              Draws        Open Slopes —
                                       NE     N     NW
                          Moist
                                       W     SE    S     SW,

                                      -hart for the  Santa Catalina Mountains, southeastern Ari-
                                  above 9000 ft is for the nearby Pinaleno Mountains.)  Four hun-
                                  mDleS WPr
-------
     4.1.2.1  Subalpine Spruce-Fir Climax --
     Engelmann spruce (Picea engelmannii) and alpine fir  (Abies lasiocarpa)
are dominant in the subalpine zone.  The spruce tends to be larger, more
abundant, and longer-lived than the fir.  Variations in the dominant species
are found in different regions.  For example, in Arizona and New Mexico,
Abies lasiocarpa var. arizonica is as important as A. lasiocarpa is in other
ranges, while mountain hemlock (Tsuga mertensiana) frequently occurs in Mon-
tana and northern Idaho; still further north, Picea glauca and A. lasiocarpa
may be associated as the transition with the northern conifer forest is ap-
proached.  There is more variability among the subordinate species found in
the Rockies than among the dominants.  In the subalpine climax zone, lodge-
pole pine (Pinus contorta var. Murrayana), aspen  (Populus tremuloides), and
Douglas fir (Pseudotsuga menziesii) occur as climax stands following burns;
restoration of the original climax stand is quite slow.  The same situation
may obtain following cleanup procedures.  Lodgepole pine does not occur in
the southern Rockies, and in other areas, aspen frequently predominates over
the pine.  Aspen has the ability to regenerate from sprouts and this may pro-
vide a significant advantage over the pine after a burn.  This factor should
also be considered with respect to forest regeneration after cleanup.

     The subalpine forest grades gently into alpine tundra in the transition
zone with a gradual lessening of the density of tree stands.  Near the timber
line, dwarfing and distortion of the trees occur, producing a tree form re-
ferred to as "Kummholz."  The timber line region is characterized by certain
tree species that are not adapted to conditions in the tundra zone and which
are unsuccessful competitors in the climax forest at lower elevations.  How-
ever, these species do occur on dry and windy ridges.  This is the niche oc-
cupied by the bristlecone pine (Pinus aristata) in the southern Rockies, limber
pine (P. flexilis) in the central Rockies, whitebark pine (P.  albicaulis), and
alpine larch (Larix lyallii) in the northern Rockies, with the exception of
the far northern areas where lodgepole pine is found at the timber line.

     4.1.2.2  Douglas Fir Climax --
     The Douglas fir climax zone lies at altitudes just below the subalpine
spruce-fir zone.  Douglas fir characteristically occurs in such density that
few subordinate species are present.  There are north-south differences in
the Douglas fir forest.   In the south, white fir (Abies concolor) and blue
                                      4-8

-------
 spruce  (Picea pungens) are present, but they are restricted to relatively
 small damp sites.  In the north, the subclimax species present differ with
 respect to the direction of the slope face and the relationship of the site
 to the continental divide.  Grand fir (Abies grandis) is found west of the
 divide and characteristically occurs on the western slopes.  East of the
 divide Picea glauca,  a representative of the northern conifer forest, exhib-
 its codominance with Douglas fir,  and is found southward through the montane
 zone into the Black Hills.

      On exposed ridges,  open stands of pine which include species character-
 istic of timber line  are found both in the montane and subalpine zones.   The
 species found differ  with the region:   P.  fliexelis var.  reflexa is found in
 the south;  P. aristata occurs in northern  Arizona,  southern Utah and Colo-
 rado;  and P.  flexilis occurs northward to  the point where it is replaced by
 P.  albicaulis.

      When stands  of Douglas  fir climax forest are burned,  lodgepole pine or
 aspen  replace it,  exhibiting the same  relationships as discussed  for the sub-
 alpine zone.

      4.1.2.3   Ponderosa  Pine Climax  --
     Just below the altitude of the  Douglas fir belt  is the  Ponderosa pine
 (Pinus  ponderosa)  zone.   This  species  and  its close relatives occur in rela-
 tively  open forest  that becomes  increasingly  like savannah as the altitude
 decreases.  Due to  the open  structure of the  growth, grasses form a ground
 cover.  The Ponderosa climax  is  separated  from the  Douglas fir by a rather
 broad transition where both  species  coexist in a shared dominance relation-
 ship.  The term Ponderosa climax is  something of a  misnomer as that species
 is truly dominant only in the northern Rockies west of the continental divide.
 In other regions, Ponderosa is replaced by or found in association with closely
 related varieties, having similar ecological characteristics such as P. pon-
 derosa var- scopulorum on eastern slopes in the north; throughout the zone
 southward, P.  ponderosa var.  arizonica, P.  leiphylla var.  chichuahana, and
 P.  contorta var. latifolia, all occur in the southern Rockies.   Along streams
 and drainage lines, narrow leak cottonwood  (Populus augustifolia)  will form
postclimax stands with P.  acuminata and P.  sargentii.   Aspen (P. tremuloides)
and box elder  (Acernegundo) may also occur  in moist areas.

                                      4-9

-------
      Older Ponderosa trees are somewhat resistant to fire and are killed only
 by severe burns.  Pondersoa seedlings are destroyed and saplings tend to oc-
 cur in the understory in groupings of even age, representing growth of several
 consecutive fire-free years.   Severe fires may result in stands of lodgepole
 pine,  while lumbering and overgrazing may result in dense growth of scrub
 species found in the oak-mountain mahogany zone.

      4.1.2.4  Pinon-Juniper Climax --
      The lowest of the Rocky  Mountain coniferous zones is characterized by the
 pinon-juniper climax.   Typically,  this is a zone of open vegetation with small
 trees  from 10 to 30 feet in height.   This vegetation zone also occurs on many
 of the low ranges in the Great Basin as the only zone present.   The pinon-
 juniper climax is typical of  the  intermountain region and forms a distinct
                            84
 zone  in the southern Rockies   .   The appearance and characteristics of the
 vegetation within the  pinon-juniper  zone are relatively constant,  although
 it includes several  species with  restricted ranges  and marked regional dif-
 ferences  with respect  to taxonomic and sociologic considerations occur.   Among
 the junipers  present in  the zone  are Juniperus scopulorum, J.  osteosperma,
 J. monosperma,  J.  occidentalis, J. deppeana (this list  is  not  all-inclusive);
 the pinons, or nut pines,  are  variously referred to  as  species  or  varieties
 and include Pinus  cembroides  (P.  edulis,  P.  monophylla,  and  P.  quadrifolia).

     This  climax  zone  extends  from northern  Mexico along the western  slope
 of the  Rockies  to  the  Snake River  in  Idaho  and into  southern Alberta  with the
 piiion while the desert species of  Juniper are  replaced by J. occidentalis or
 J. scopulorum.  On the eastern slope  of  the  Rockies, the zone extends  as far
 as Fort Collins, Colorado, at which point pifion  disappears and J.  scopulorum
 continues northward  through Wyoming,   frequently  associated with  sagebrush.
 In the northern Sierras, pinon-juniper  is not  present with Jeffrey pine  (Pinus
Jeffreyi) grading  directly into Artemisia and  Pushia; it does exist on the  low
eastern slopes of  the southern Sierras but tends to occur in mixed growth with
other species.  However, piiion-juniper does occur on every westernmost range
and mountain of the Great Basin, and  frequently is separated from the Sierra
Nevada range by only a single  valley.
                                    4-10

-------
 4.1.3  Sierra Nevada  Forest Complex
      The following discussion of the Sierra Nevada Forest Complex will in-
 clude the southern Cascade Mountains and the Sierra Nevada which extend from
 Oregon southward along the California-Nevada border, including the innermost
 ranges of the coastal mountain system.  The western slope of the Sierras is
 gradual while the eastern slope is quite abrupt, dropping from peaks of more
 than 14,000 feet to the floor of the Great Basin at an elevation of approxi-
 mately 4,000 feet.  Due to this slope difference, the western slope encom-
 passes a much greater land area than does the eastern.

      4.1.3.1.  Western Slope --
      Due to the north-south axis of the Sierras, and the prevailing westerly
 winds,  the eastern slopes of the range are much drier than the western.   At
 the base of the western slopes,  10 to 15 inches of precipitation fall annually
 with a long,  unbroken summer dry season.  As the altitude increases,  the  dry
 season shortens,  temperatures  fall,  and precipitation increases with  a greater
 proportion falling as snow.   Approximately 80  to 85 percent  of the  precipita-
 tion occurs  in the winter,  with  35  to 75 feet  of snow falling  in the  subalpine
 zone.   The heaviest  total  precipitation is  reported to  occur between  5,000  and
 7,000  feet    where the mixed coniferous forest  of the montane  zone  is found.
 The subalpine  zone coincides with  that  zone having the  greatest  snowfall,
 and possibly  this  is the  zone of  greatest  precipitation  since  snow gauges
 which are  exposed  to  high  winds  tend  to give low readings; spring snow sur-
 veys indicate  a greater water content of the accumulated  snow  than can be ac-
 counted for by reported precipitation^.  This  fact may have implications for
 cleanup procedures.

     Subalpine Zone-The subalpine zone  is  usually found  in a 1,000-foot ver-
 tical range between 6,500 and 9,500 feet depending upon the latitude.  Typ-
 ically, this zone experiences cool, short growing seasons with wet winters
 and  dry summers.

     The major climax  species found in the  subalpine zone in the Sierra Nevadas
 is red fir (Abies magnifica), which comprises 80 to 90 percent of the forest
 tree growth, occurring in dense stands.   Other species are associated, but of
minor importance in the climax growth.  Western white pine (Pinus monticola)
 is usually present as a minor constituent.  Lodgepole pine (Pinus contorta)

                                    4-11

-------
 is often present,  at the edges of meadows,  but in a successional role.  Ir-
 regular growth of mountain hemlock (Tsuga mertensiana)  and white fir is
 found in the zone,  with the former tending to occur at  the higher elevations
 of the zone and the latter at lower elevations.   Shrubs and herbs are found
 throughout the zone and lichens occur on the trees.

      Timber line is often found at lower elevations in  the Sierras than in
 the Rockies,  ranging from 7,000 to 10,000 feet depending upon the latitude;
 this is due to the  more severe environmental conditions occurring in the
 Sierras.   Pinus albicaulis,  P.  flexilis,  and P.  balfouriana are character-
 istic of timber line growth in the Sierras65.   Juniperus occidental is is also
 common at timber line on exposed,  rocky slopes;  it  is also found on  the west-
 ern slopes at lower altitudes.   Generally,  red fir  growth does not extend
 to timber line,  but instead grades into a relatively narrow band of  Pinus
 contorta - Tsuga mertensiana dominance.   In this  zone,  P.  contorta assumes
 a  climax role,  although it  is successional  to  Abies magnifica at lower alti-
 tudes;  hemlock also has climax  characteristics at the upper limits of the
 subalpine zone.

      Montane  Zone—The  montane  zone occurs  between  2,000 and  6,000 feet  in
 the  Cascades,  between 4,000  and  7,000 feet  in  the central  Sierras, and be-
 tween 5,000 and  8,000 feet or more in the south.  Five  or  six principal
 species have  climax characteristics in  this  zone and tend  to  occur in any
 combination at any  altitude,  although the upper and  lower  margins  of  the
 zone  are  likely  to  have  consistent vegetational differences.   In the  upper
 part  of the zone, white  fir  is usually  the  important dominant, where  it  is
 sometimes  found  in  pure  stands;  its numbers decrease at  lower  elevations.
 Important  species at lower elevations include  sugar pine  (P.  1ambertinana),
 Jeffrey pine,  Ponderosa  pine, Douglas fir, and incense  cedar  which tends  to
predominate on the most  favorable sites.  Douglas fir is more  frequent in  the
north; sugar pine and Jeffrey pine are more abundant at higher elevations,
the latter being the most drought-resistant of the major species of the zone.

     4.1.3.2  Eastern Slope --
     The eastern slope of the Sierra Nevada range has the same vegetational
zones as those found on the western slopes.   However, the abruptness of the
eastern slope and its drier climate result in less favorable conditions and

                                    4-12

-------
 some differences in the distribution of vegetation from that found on the
 western side.  The red fir forest is found in restricted areas such as the
 Carson Range east of Lake Tahoe and in localized areas in the northern
 Sierras.   The subalpine zone is generally represented by patches of lodge-
 pole pine and those species of pine found at timber line on the western side.
 The montane zone extends to high altitudes with poorly developed vegetation.
 In this zone, Jeffrey pine is the most important species and it occurs in
 open stands which differ in appearance from those on the western slope.
 Piiion-juniper forest is restricted to a few eastward-extending high spurs
 of the range, although it is well developed on the next range to the east
 of the Sierras

 4.1.4  Pacific Conifer Forest
      The  coniferous forests of the Pacific coastal region parallel  the coast
 from the  timber line in Alaska southward to central California.   Mountain
 ranges of various altitudes are found within the range of this forest  type.
 The entire region from north to south is characterized by a mild climate
 which is  tempered by the presence of the Pacific Ocean.   Subzero temperatures
 are rare  on the Alaskan coastline and from Oregon southward,  frosts  are rare.
 Precipitation ranges from 30 to 150  inches and  the humidity tends to remain
 high.   In the south,  rainless  summers tend to be compensated  for by  fog.
 Snowfall  increases  with latitude  where  50  to 60  feet may  occur in some of
 the higher mountain ranges  in  the north.

      The  character  of  the coastal  forest is primarily  montane  throughout  its
 altitude  range  from 0  to  5,000  feet.   In the northern  portion  of the Cascades
 in  the Washington-British Columbia border  region,  some subalpine forest oc-
 curs,  but  it contains  dominant species derived  from the  Rocky Mountains
 (Abies lasiocarpa),  the Sierra  Nevadas  (Tsuga mertensiana) and the coastal
 forest  (Abies amabilis  and A. procera).  Tundra  exists on some of the higher
mountain peaks.

     The best development of coastal forest species occurs in the general
vicinity of Puget Sound, with the most luxuriant vegetation occurring in the
rain forest of the Olympic Peninsula where the ranges of all of the major
species overlap and most of the trees attain their maximum size.  The climax
                                    4-13

-------
dominants are western hemlock  (Tsuga heterophylla),  western  arborvitae
 (Thuja plicata) and grand fir  (Abies grandis).  Douglas  fir  is  the most
abundant and widespread  species;  it tends to  be found  in the drier sites.
It  is intolerant of shade and  becomes the major dominant after  fires  and
is  therefore subclimactic in its  relation to  the forest.

     North of Puget Sound, sitka  spruce  (Picea sitchensis) becomes increas-
ingly abundant and is a  climax dominant  along with Tsuga heterophylla and
T.  mertensiana in the northern extension of the forest.   In  Alaska, where
the Boreal and Coastal forests interface with the tundra, P.  sitchensis and
P.  glauca occur at timber line.   Southward from Puget  Sound,  Douglas  fir
becomes increasingly important and Port  Orford cedar (Chamaecyparis lawsoniana)
becomes an added climax  species.  Along  the coast, redwood (Sequoia semper-
virens) replaces Sitka spruce.  The redwoods are generally limited to the
fog belt and their range extends  irregularly for a number of miles south of
Monterey, California.  The species of the coastal forest  extend eastward
along the Washington-British Columbia border, expanding  to the north  and to
the south on the western slop  of  the Rockies.  Species found in this  exten-
sion include Tsuga heterophylla,  Thuja plicata, and Pseudotsuga mensiesii
and occupy a zonal position between the  Douglas fir and  spruce fir zones of
the Rockies.  Western larch, white pine, and grand fir also  occur in  this
extension.  Daubenmire  noted  that this  extension follows a  well developed
storm track and is frequented  by  winds which blow eastward from the coast
51
4.1.5  Coniferous Forests Occurring as
       Sub and Postalimaxes
     Extensive coniferous forests occur in the eastern portion of the Nation
as subclimaxes which are maintained by fires or swamps; in some regions these
forests are not considered to be climaxes.  The coastal plain from New Jersey
to Florida and along the Gulf States into Texas is a relatively flat area at
low altitude.  The soil is generally sandy and tends to have poor drainage
with many swamps.  Raised areas are usually dry for a portion of the year.
Dry seasons are accompanied by low water tables and relatively frequent fires,
In New Jersey, pitch pine forests occur and loblolly and slash pine are found
in the states further south.   These pine forests are generally open with
grass floors and are referred to as savannahs.  Frequent grass fires permit
survival of the relatively fire-resistant conifers at the expense of oak and

                                    4-14

-------
 hickory, which would replace the conifers if the area were given prolonged
                     8112
 protection from fire  '   .   Some consider these areas to be a "pyric climax."
 Whatever terminology is  used, the presence of frequent fires in such areas
 must be considered in the reestablishment of the original growth that pre-
 ceded cleanup procedures.

      Coniferous forest may  also be encountered within those regions generally
 dominated by hardwoods.   Also,  an association of hemlock and hardwoods occurs
 in the transition between the northern coniferous forest and the deciduous
 forest.   Hemlock (Tsuga  canadensis)  frequently occurs with beech and sugar
 maple and less frequently with  birch,  white  pine,  and other hardwood species
 from northwestern Minnesota through  the Lake States into Nova Scotia.   Many
 pure stands  of pines  (P.  strobus and P.  resinosa)  have been destroyed  by fire
 and lumbering.   White pine  is found  within the hardwood climax forests of the
 region,  but  Costing    believes  that  these long-lived trees should  be consid-
 ered as  relicts,  although they  do reproduce  in open areas in the hardwood for-
 est.   As such,  their  replacement following cleanup  procedures may  prove to be
 difficult  and  might be considered to be  optional, depending upon local  circum-
 stances.   Tamarack, black spruce,  and  white  cedar  (Thuja occidentalis)  occur
 in  postclimax  stands  and  occupy bogs throughout  the region.   Aspen and  pine
 also  occur in  burned  or logged  out areas.

      Pitch pine  (P. regida),  Virginia  pine (P. viginiana)  and  shortleaf pine
 (P.  echinata)  occur from  southern New  England  to Georgia  in  that region fre-
 quented  by an  association of  oak  and chestnut, with  the  latter two  species
 becoming locally abundant in  the  southern part of the range.   In the south,
 the occurrence tends to be  limited to  higher altitudes, but  in the north,  it
 occurs at lower elevation and occurs as far east as  Long  Island.  Virginia
pine, shortleaf pine and  loblolly pine also occur on abandoned land throughout
the region of the oak-hickory association on the Piedmont Plateau and on the
coastal plain of the Atlantic and Gulf States.  Virginia pine predominates
in the northern Piedmont  with shortleaf and loblolly pine preceding climax
growth in the south,  usually occurring in pure stands.
                                    4-15

-------
 4.2   NATURAL  PERTURBATIONS
 4.2.1   Fire
      4.2.1.1   Unassisted  Recovery  Sequence  --
      Fossil records  indicate  that  fire has  probably  always  been  a  part  of  the
 environment of forests on the surface of  the earth.  Most of  the species of
 conifers have  a number of fire-adapted regenerative  qualities which  range  from
 a need  for mineral seed beds  to a  requirement  for high  temperatures  to  permit
 the opening of serotinous cones.   Thick bark,  adventitious  buds, and coppice
 sprouting also enhance survival of fire.  Most coniferous species  are also
 tolerant of exposure to sunlight and open space   .   It  seems  likely  that many
 of these fire-adaptive qualities will serve to assist the recovery of the
 coniferous forest from cleanup procedures,  if  those  procedures are signifi-
 cantly  more disruptive of the forest than would be a severe fire.  However,
 it is important to note that  these characteristics are  not  universal  among
 coniferous trees.

     On a study plot in the Sierra Nevada mountains, which  was established
 after a severe fire in 1960,  it was found that natural  revegetation  (un-
 assisted) by Jeffrey pine seedlings greatly exceeded that by  fir seedlings8.
 This situation was reversed in data obtained from ten unburned control  plots
 where establishment and growth of  fir seedlings exceeded the  pine  seedlings.
 The pine seedlings were found to have high  survival  rates and demonstrated
 an exponential rate of growth.  Some of the trees had attained heights  of
 250 cm  at fourteen years  of age (having germinated the  summer of the  burn).
 It was  also noted that pine seedling recruitment occurred at  approximately
 identical rates on burned and unburned plots.

     The sequence of events following a severe forest fire  will differ  for
 each particular forest and location within  that forest.   It is difficult to
 generalize in  this regard for all  of the  coniferous forests of North America.
 At the  site of the Conner Ridge test plots,  the natural  postfire reforestation
 pattern appears to be:   pine  > pine + fir > fir.    However,  the stage of pure
 fir forest is  seldom reached  in the region due to interference by fire,  and
           o
 Bock et al.   note that it is difficult for this fire-regulated successional
pattern to be reconciled  in terms of classical ecological climax community
               8              8
classifications .   Bock et al.  suggest that Jeffrey pine is particularly
adapted to the fire environment,  since the observed sustained exponential

                                    4-16

-------
 growth pattern would not be expected in the irregular climatic conditions
 that have obtained in recent years in that area.

      Following a forest fire, the successional pattern is reset.  Fungi and
                                     67
 other microorganisms invade the area  , and the composition of wildlife popu-
 lations  is  altered due to the drastic alterations in habitats produced by
 the fire.  Lyon and Stickney   have examined the vegetal succession that fol-
 lows large fires in the northern Rocky Mountains.  They found that succession
 in the forest is not an autogenic process in which initial serai plants modify
 the site to their own exclusion and then subsequently permit the establishment
 of interseral and eventually climax species.   Instead,  they note that succes-
 sion is a sequential development of vegetation in which the more rapidly ma-
 turing and often shade intolerant plants assume initial dominance and then in
 turn are dominated by taller,  slower growing,  and often more shade tolerant
 species.   They obtained data from three large  wildfire  burns that indicated
 that a high percentage of  plant  species on a  site at  the time of the  fire
 survive and reestablish themselves on the  burned area.   The majority  of recog-
 nizable survival  adaptations are on-site plant  parts  and seeds or fruits,  and
 these  provide the major source of early serai vegetation.   They noted an ex-
 ception to this  rule  in a  few species  that have both  an  on-site survival mech-
 anism  and airborne  seeds.   They  found  that all  of the dominants of early suc-
 cession will  be  established in the initial postfire growing  season, whether
 or  not  they are derived  from on-  or off-site sources.   Lyon  and Stickney46
 believe that  reliable  estimates  of the  likely composition of the  early  suc-
 cessional  forest  can be  derived  from the composition  of the  prefire or  the
 first postfire year communities.   Further  predictions may be possible based
 upon the  survival strategies and dominance potentials of the various plant
 species in the community which will indicate the probable structural config-
 urations of the early successional period.

     The findings of Lyon and Stickney46 suggest that in many areas, the
probable course of unassisted recovery from a cleanup operation in a conif-
erous forest might be predicted on the basis of local data on fire recovery.

     4.2.1.2  Assisted Recovery Sequence --
            64
     Shearer   has compared the  effects of unassisted fire recovery with as-
sisted recovery in coniferous forests in the Rocky Mountains.  He notes that

                                    4-17

-------
  historically coniferous  forests  of the  northern Rockies  have been returned
  to  early  stages  of  succession  by fire39.   This  is  largely a consequence of
  heavy  accumulations of litter  and other flammable  debris accumulated  over
  many years.   Pioneer species such as western  larch (Larix occidentalis)  and
  lodgepole pine  (Pinus contorta)  require an early successional  stage,  and
  they rarely  become  established in the absence of a major disturbance  of for-
  est.   Frequently, land managers  utilize mechanical scarification  on slopes
  of  less than 30 percent  in order to enhance the  formation of a seed bed.
  Shearer   describes  experiments  designed to obtain quantitative data  con-
  cerning the  effects  of varying intensities of burning on natural  regeneration
  in western larch/Douglas fir forest.  It was found that  early establishment
  of coniferous forest regeneration following prescribed broadcast  burning of
  clearcuts was influenced by the  interaction of several factors, the most im-
 portant of which were (1) seed bed condition, (2) seed supply, and (3) en-
 vironmental  characteristics affecting seedling survival.

      Shearer   noted that even  under the most extreme conditions  of fire, the
 soil temperature remained at  66°C at a  depth of  29  mm while the surface temp-
 erature was  260°C;  therefore, the underground systems of  the plants receive
 relatively little damage, and most species  were  able  to reestablish them-
 selves  by  vegetative reproduction.

      It was  observed that most  of the seeds that  fell on  the clearcuts came
 from adjacent growth, although  some were from  unidentified sources, perhaps
 miles distant.  Most of the natural  seeding occurred  during periods of tur-
 bulence associated with frontal weather  systems.  The actual  distribution of
 the  seeds  on  the  large clearcuts  decreased  rapidly  with increased  distance
 from the edge of  the adjacent timber stand  surrounding the burn, but there
 was a low, relatively constant concentration of seeds  throughout the central
 section of the burned area.   Possibly a higher proportion of the  total  seed
 in the  central area  was derived from unidentified distant  sources, rather than
 from the adjacent timber  stand.

     Germination of  seeds was adequate on treated and untreated seedbeds,
but was greater on bare mineral  soil (burned or scarified) than on scorched
or unburned duff.   It was found that germination  began sooner on clearcut test
beds possibly due to higher soil temperatures than obtained beneath residual
timber.   It was also found that  the establishment of seedlings could be increased

                                    4-18

-------
 by burning the residual duff on clearcuts.  Burned clearcuts facing from
 300 degrees to 90 degrees azimuth (west-northwest to east) regenerated well.
 Clearcuts with the least amount of regrowth were judged to have poor seed
 sources.  Regeneration was reduced on less favorable aspects where high sur-
 face temperatures and low soil water limited seedling establishment.  Smaller
 burned areas having three or four adjacent timber edges tended to have in-
 creased regeneration due to a better seed supply and possibly better shading
 from the adjacent forest.

      Shearer   also notes  that successful regeneration was so strongly asso-
 ciated with the type of habitat.   Growth was particularly successful under
 shade cast by either living or dead  trees on burned seedbed.   Growth was
 sparse on poorly burned seedbeds  and on hot open slopes with  little or no
 shade.   In general,  moist  habitats regenerated well,  while drier areas were
 understocked.   Large clearcuts received insufficient  seed fall  and the upper
 soil  was soon depleted of  available  water by heavy vegetative competition.

      Shearer    concluded that  clearcutting coupled with prescribed broadcast
 burning on north- and east-facing  slopes usually leaves  the cutover area re-
 ceptive to natural  regeneration if sufficient  seed is present.   On hot,  dry,
 and steep south-  and west-facing  slopes,  shelterwood or very  small  clearcuts
 (group  selections)  offer a  better  chance  for regeneration  by  leaving a  greater
 seed  source and better protection  for the  site.  Clearcuts  without  seedbed
 preparation, regardless  of  the  aspect,  restock  slowly.

 4.2.2   Other Natural  Perturbations
     Other than fire,  major natural perturbations to coniferous  forests in-
 clude earthquakes, land  and snow slides and flooding.  The  severity of the
 damage  depends upon the  intensity of the initial disturbance and the location
 of the  affected forest; prospects for recovery from these events are subject
 to the  same variables.  In  some  instances, permanent alteration of the land-
 scape will result with attendant changes in the biota of the area.

 4.3  MAN-MADE PERTURBATIONS
     Probably the most obvious of the man-made perturbations to the conifer-
ous forest of the United States are logging and land clearing  for agricul-
tural  purposes,  construction,  and  mining.  Only in the cases of logging and,

                                   4-19

-------
recently, strip mining is reforestation contemplated after the clearing or
clearcutting of the land.  Clearcutting for the purpose of harvesting usually
does not entail the processes of stumping, grubbing, and soil importation
that would probably occur in the case of a cleanup operation.  Clearing for
agriculture is generally intended to be a permanent treatment.  Therefore,
neither of these examples is strictly analogous to the sequence of events
that might be involved in the cleanup process.  In some respects, the restora-
tion of a forest after a strip mining operation would resemble the cleanup
operation, but there is little historic precedence of sufficient age for this
to provide prognostic information.

4.3.1  Clearcutting
     In addition to the loss of vegetation and disruption of animal habitats,
clearcutting has a number of additional effects.  The ecological balance of
the forest flora and fauna is affected at all biological levels.  For exam-
ple, the loss of habitats from the destruction of the trees may change the
bird population in the clearcut area, which in turn may affect the insect
population.  The slash remaining after clearcutting will offer new habitats
for new species, and there may be increased damage to young trees in the
regenerating area from diseases and pests  .  Water quality in a forested
area is generally decreased after clearcutting due to increased sediment in
         52          30
the water  .  Hibbert   has reviewed the historical effects of forest treat-
ment on water yield.  Generally, he found that removal of forest increases
water yield, but he notes that the results of individual treatments were
highly variable and were, for the most part, unpredictable.  Hibbert also
found evidence that the increase in streamflow was proportional to the amount
of reduction of the forest.   Surface runoff and streamflow were also shown to
                                                                         •7 7
be increased in the Hubbard Brook experiment.   (See discussion by Horwitz  .)
Soil erosion is greatly increased in clearcut mountainous areas.   Disruption
of the soil and increased insolation alter the microbiological populations
after clearcutting.  Soil nutrients are leached from the soil in clearcut
     9
areas .

4.3.2  Strip Mining
     Strip mining operations in forested areas represent a widespread source
of severe perturbation.   Only in recent years have there been significant

                                    4-20

-------
 attempts at reclamation of mined areas, and the technology for doing so is
                                                   48
 still in a developmental stage.  May and Striffler   have addressed the water-
 shed aspects of stabilization and restoration of strip mined lands.  They
 found that erosion and sedimentation from disturbed watersheds are attribu-
 table to three main sources:  slides from sidecast overburden, haul roads,
 and the mined area itself.   Proper spoil placement to ensure bank stability
 by drainage control will serve to minimize disturbance.   Studies of replant-
 ing of strip mined areas may be found in Boyce and Neebe10 and Sindelar,
 Atkinson,  Majarus and Proctor  .   Many of the efforts at replanting of strip
 mined lands have not attempted to reproduce the original forest,  but rather
 to stabilize the land with  rapidly growing (often imported)  species.

 4.3.3  Controlled Burning
      Fire  has been used as  a tool in the forest for centuries.   In is  com-
 monly used now in forest management  as a means of decreasing fire potential,
 for removal of debris after logging  operations,  etc.   Stark67 has studied  the
 effects  of controlled burning of  forest understory material  upon  soil  nutri-
                     68
 ent content.   Steele   has  examined  various  aspects of smoke management  from
 such fires  in relation to atmospheric  conditions.   A discussion of the envi-
 ronmental  aspects  of prescribed burning may  be found  in  Fillmore  et  al.20.

      In  the consideration of methods for reforestation after cleanup proce-
 dures, assisting recovery through the  planting of  seedlings  seems essential.
      74
 Tinus    has noted  that  in order to survive,  the  species  and  seed  source  used
 must be  adapted to the  particular site,  the  seedlings  must be in  the proper
 physiological  state  to  meet  the new environment, and root contact  with the
 soil must be  quickly  established.  When  compared with  outdoor grown bare root
 seedlings,   containerized  seedlings can be more precisely controlled and will
 not  lose root contact with the soil in the planting process.  It may also be
 of  interest to consider mychorrhizal inoculation of the soil in conjunction
 with replanting, if the soil  is imported and alien to the forest which is
 destroyed in cleanup procedures.  Mychorrhizal inoculation in afforestation
                           49
has been .reviewed by Milola   .
                                    4-21

-------
4.4   EFFECTS OF CLEANUP  PROCEDURES ON  CONIFEROUS  FORESTS
      This section will discuss the effect on coniferous forest  ecosystems of
a number of possible cleanup procedures.

      (0-1)  Natural Rehabilitation
      The coniferous forest  is not always pure but frequently occurs  in mixed
stands with varying proportions of hardwood pioneers which become  established
mostly on small bare patches, brush patches, or other non-grassy patches in
the disturbed area.  An  example of primary succession following complete de-
struction of a conifer forest is that  of the six  stages of a red spruce forest
on mineral soil   .  The  successional stages are:
          1.  Moss meadow
          2.  Aster-fireweed meadow
          3.  Hair grass-sedge meadow
          4.  Wetlow-birch  thicket
          5.  Aspen forest
          6.  Spruce forest.

      In the north, gray  and paper birch, pin and black cherry,  and big tooth
and trembling aspen are  the typical hardwood pioneers.  In the  south, sweet
gum,  red maple, and many other hardwoods become established following clear-
cutting.

      Along the east slope of the Rockies, severely disturbed areas,  such as
those resulting from fires, tend to be colonized first with lichens, moss,
and fireweed.  Lodgepole pine reseeds  concurrently, either in pure pine stands
or in association with Englemann spruce in clearings at the upper edge of the
coniferous forest.  However, cleared areas may be characterized by such harsh
climatic conditions near the ground that conifers cannot reseed in the area.
Many  mountain meadows of the western American ranges may have originated in
this  manner.

      If natural rehabilitation is to be relied upon for the clearcut areas,
the major source of seed will be the surrounding forest.  Areas with dimen-
sions greater than five  tree heights (of the surrounding forest) will not
receive adequate coverage.  If the treatment is carried out on a fire species
such  as lodgepole pine or Jack pine, cones will not open and reseeding may
not occur.
                                    4-22

-------
       Although the site may be replanted with coniferous species,  hardwoods
  may soon invade,  and subsequent possibilities for variations in the pattern
  of succession of  various  sites then become too numerous for generalizations.

       (1-1)   Chemical Stabilization
       Appendix A lists the  characteristics  of 23 "chemical"  and  "chemical  with
  mechanical  characteristics"  stabilizers  which are preferred for coniferous
  forests.  These soil  sealers  and  stabilizers mainly  include cementitious  cal-
  cium  compounds and mixtures,  organic bituminous products, waterproofing agents,
  and emulsified binders.

       The use  of emulsifiers is  a  temporary treatment for stabilizing a soil
 until vegetation is established.   Its effective  life is not much beyond 2
 years.  It would only be useful on sloping areas.  It could, to some extent
 on most treated sites, enhance  succession and the development of wildlife
 habitat by retarding erosion until plants became established.   It would not
 adversely affect any site over and above the primary treatment.

      Deflocculants such as salt, sodium hydroxide, and commercial  materials
 such as Calgon, when present on the soil exchange complex in sufficient quan-
 tity can virtually render certain soils impervious to water.  The  degree to
 which  this  can be  accomplished depends  upon soil texture,  structure, and the
 type and concentrations of ions already present in the  soil.  However,  it  is
 assumed the  soils  of  the treated areas  would  be appreciably  affected by the
 chemical treatment.   Accordingly,  infiltration would  be  reduced, and surface
 runoff increased.   Thus,  less  water would be  available for vegetation that
 might  invade  or be  planted  on  the  site.   Erosion on-site and sedimentation
 off-site would be  increased as a result of  both  increased runoff and the de-
 flocculated  soil which would go  into  suspension  more  readily.

     (1-2)  Cleareutting Vegetation
     Clearcutting of coniferous  forest has been  suggested as a possible tech-
nique  for cleanup of spilled hazardous materials.  Precedent for this proce-
dure and many associated problems may be found in the commercial harvesting
of Douglas fir in the Pacific Northwest, western white pine  in northern Idaho,
Jack pine in the Lake States, loblolly pine in the south, and lodgepole pine
in the Rocky Mountains.  Revegetation in these cases may be achieved through
                                    4-23

-------
growth of previously established seedlings, natural seeding prior to cutting,
artificial seeding after cutting and plant programs.  However, it is assumed
that in the case of clearcutting for purposes of spill cleanup, the clearcut
areas will be replanted or reseeded.

     Technically, artificial rehabilitation is applicable to any kind of for-
est.  It contrasts sharply with the seemingly closely related method of clear-
cutting combined with rehabilitation by natural means which has much greater
limitations and gives satisfactory results only under favorable conditions.
                                                                  19
     Some of the general effects of clearcutting can be summarized   as
follows:
              Clearcutting eliminates the danger of wind damage or
              disease infection to residual trees in the cutover area,
              but increases the risk of windthrow or heat damage to
              trees bordering on the cutover area.
              Clearcutting improves forage for game animals such as
              deer, elk,  and provides habitats for certain animals that
              were not present before logging.  However,  streams may be
              adversely affected by clearcutting with increased temper-
              atures, debris jams, and sedimentation, all of which are
              detrimental to fish populations.  The procedure will also
              reduce habitats of some of the smaller animals such as
              woodpeckers,  and tree squirrels.
              Water yields during low-flow periods  will  be increased
              by clearcutting; snow catch may also  be increased.   In
              swampy areas,  the water table may be  elevated by the
              procedure.
              Clearcutting will permit the use of genetically improved
              tree planting stock, but growing conditions will be also
              enhanced for many unwanted brush species,  which will com-
              pete with the young conifer seedlings.
              Good growing conditions for shade-intolerant tree species
              such as Douglas fir and noble fir will  be  created by clear-
              cutting but seedlings will be exposed to injury from cli-
              matic extremes.
                                    4-24

-------
     Effects of Clearcutting on Climate  and Hydrology—Small  clearings within
 forest stands can  create extreme  effects on the microclimate,  and  larger
 clearings significantly affect hydrological factors.  As an illustration of
 these effects, Table 4-1 was taken from  Geiger:25

                Table 4-1.  Measurements on forest clearings.
Measurement item
Size index D:H*
Outgoing radiation
of open land)
Diameter D (meters)
0 12 22 24 38 47 87

(percent
Rain (percent of open land)
Temperature (excess
land midday July)
over open
0 0.46 0.85 0.93
0 11 31 33
87
0 0.7 2.0 2.0
1.47
52
105
5.2
1.82 3.36
66 87
102
5.4 4.1
 *D = diameter of the clearing, H = height of forest trees, and tree height
  averages approximately 20 meters.
     (a)  Temperature:  Temperature excesses (temperature increases over those
found in the forest surrounding the clearcut) increase with the size of cleared
openings in the forest up to a size index of about 1.8 because of solar heat-
ing.  However, beyond this size temperature again decreases because of wind
action.  It has been found that size indices less than 1.25 can provide com-
plete frost protection while frost damage will be considerable at indices of
1.50 or greater.  In colder regions, air temperatures decrease linearly to
about 5°C below the temperature obtained in a closed stand as the diameter
of the clearing increases from 0 tc 80 meters.   Extrapolation of the data in
Table 4-1 indicates that the excesses would be minimal (except for the dif-
ferences in cooling by transpiration) in clearings 180 meters in diameter.

     Soil temperatures within clearings will increase by 2°C to 3°C over
those temperatures found in soil beneath a closed stand.   In colder climates,
spring snow thaws in clearings may occur as much as 4 weeks earlier than in
closed stands.
                                    4-25

-------
     (b)  Radiation:  Incidented radiation will, or course, be greater in
clearings.  Albedos will be higher in the clearings and net radiation in
clearings may vary between 70 and 90 percent of that in the forest stand.

     (c)  Soil Moisture:  In general, the forest bordering a clearing will
deplete soil moisture in the clearing to the extent of their root systems
during the summer, while soil moisture within the clearing beyond the forest
root zone will remain high throughout the year.    Soil moisture within the
clearings in mixed conifer forests can be expected to be 5 to 40 percent
above that of the surrounding forest. '   '   '

     (d)  Precipitation:  Although small clearings will, because of wind
eddys, tend to increase in rainfall, their greatest effect is upon snow
catch.  During periods of snowfall in small clearings, incoming snow at one
time or another comes from all directions.  The ultimate pattern of snow
accumulation, however, will most strongly reflect the average velocity and
direction of air flow.  The zone of maximum snow accumulation will generally
                                   24 29
be near the center of the clearing.  '    Snow catch in small clearings
within coniferous forests can be expected to be 20 to 40 percent higher than
                             24 54
that found within the forest.  '    It should also be pointed out that the
total quantity of snow in the forest area will not be increased.  Clearings
                                                                   24
will have the effect of reducing snow catch downwind in the forest.

     (e)  Runoff:  Numerous experiments have demonstrated that clearing a
coniferous forest will increase runoff   because of reduced transpiration
and interception.  The amount of additional runoff will vary with the climate
of the area, the forest type, soils, slope and aspect.  The increase will be
greatest during the first year or two after cutting.  As the area is invaded
by herbaceous vegetation and broad leaf species or as planted stock develops,
the effects of clearcutting will decrease  (see Section 4.5).

     Erosion and sedimentation will not be greatly affected if the forest
floor is not severely disturbed by the clearcutting process.

     Effects of Clearcutting upon Wildlife—The impacts of clearcutting upon
the existing and potential wildlife habitats wihtin coniferous forests may
be categorized as:  direct alterations to the habitat including obvious damage
                                    4-26

-------
to the structural and functional aspects of the area; and indirect altera-
tions and changes within the habitat and its surrounding area through inter-
ferences with the processes normally provided within the larger natural bio-
logical system of the forest.  While the true quantitative and qualitative
impacts on wildlife generated by clearcutting in terms of ecological recov-
ery and stabilization are subject to debate, some generalizations may be made
          1.   The structural alterations resulting from clearcutting
              obviously influence the mature trees.   In the moist cool
              transcontinental coniferous forest biome between the
              45th and 57th latitudes (north),  the coniferous forests
              modify and temper the environment such that the lesser
              plant species and codominant species are dependent upon
              them.  The removal of the  dominant species is equivalent
              to forcing plant succession back  to a more primitive state
              near the pioneer or early  second  stage of succession.
          2.   Habitats are defined by the plant community both struc-
              turally and functionally;  direct  utilization of the three-
              dimensional space formerly occupied by the clearcut forest
              for reproduction,  food and shelter is  no longer possible.
          3.   The ground plane,  depending upon  its condition after treat-
              ment will  redevelop a plant cover of some type.   Erosion,
              drying  and heating,  and general physical exposure will
              make this  a harsh environment  for rapid development.
          4.   Plant succession will  in most  cases start anew unless  the
              area is  planted.   Migration of animal  and plant  species
              from adjacent  areas  will participate in setting  the char-
              acter of redevelopment.  If the image  and statistical
              character  of this  area is  to be similar to  its  surround-
              ings, it must  be supported  by  management  activities that
              recognize  the  needs  of the  area in  terms  of  specific habi-
              tat  requirements.
          5.   Reutilization  of the area by the  native  mammal,  bird, rep-
              tile, and  insect populations is a function of how well
              and  how  fast the former habitat begins  to simulate  the
                                   4-27

-------
              the functions of the surrounding area for the support
              of wildlife, and of the relative adaptability of those
              species that are able to find niches within the treated
              area.

     The reestablishment of plant stages represents a new-found supply of
food and shelter for the wildlife.  Competition and simultaneous resource
use will prevent this space from developing immediately into its former
character.

     To bring treated areas back to their former wildlife habitat requires
judicious replanting and management.   How well and how much these practices
are employed will greatly influence the degree and amount of recovery.

     (1-3)  Stumping and Grubbing
     The removal of stumps and roots  creates depressions and scarifies the
treatment area.   The net effect of this treatment would be modification of
the effects of simple clearcutting in several ways:
          1.  Surface runoff would either remain unchanged or possibly
              be decreased.   The additional water storage in depressions
              created by the treatment would tend to offset  the addi-
             tional runoff  that would be caused by  antecedent soil
              moisture.
          2.  The surface runoff would carry greater sediment  loads
              because of the  additional  scarification  caused in the
              cutting process.
          3.  The depressions  and scars produced  by  the  treatment  would
              occur  at a slower  rate  after  use of clearcutting  as  a
              treatment  procedure followed  by stumping and grubbing,
              than in the case of simple  clearcutting  for harvesting
              due to the absence  of roots and  stumps for coppice
              stumping.
          5.   Small  mammal burrows would be  destroyed and/or damaged.
                                   4-28

-------
      (1-4)   Scraping and Grading
      Scraping  of coniferous  forest  areas  would require prior clearcutting
 and  removal  of stumps from the  treatment  area.   The  litter layer and or-
 ganic soil  layers would  be removed  in the process  of scraping.   Accordingly,
 infiltration rates would be  greatly reduced,  surface runoff would increase,
 and  detention  and retention  storage of water  would be reduced,  resulting  in
 unchecked soil erosion.   Further, the residual  seed  pool  in the soil would
 be greatly diminished and nutrient  levels drastically lowered.   Depletion
 of populations of soil-building microfauna and  flora would also take place.

      Grading of coniferous forest would result  in  an amplification of all of
 the  effects  of scraping  treatment;  erosion in particular  would  be increased
 since smooth grading  would eliminate  detention  and retention storage.   The
 grading equipment  would  also tend to  compact  the soil,  thereby  increasing
 runoff.  Runoff efficiencies, even  on slopes  as  low  as  2  percent,  might be
                                     13
 expected to  exceed 60 to  70 percent.

      (1-5)   Shallow Plowing
      This treatment would  in most instances require  some  smoothing of the
 site  following  stumping and grubbing.   Over the long  term,  the  effects of
 shallow plowing  would  probably  not  differ  greatly  from  simple clearcutting.
 Over  the short  term,  this  treatment would  be  superior to  stumping  and  grub-
 bing.  Essentially, on areas with soils stone-free enough  for plowing, and
 of adequate  depth,  the area would be  tilled as an  agricultural  field.  Sur-
 face  organic matter would be incorporated  into the soil, and the  soil  loos-
 ened, thus creating a good seed bed,  and favorable conditions for  root
development.

     The infiltration potential would be increased by the  loosened soil re-
 sulting from shallow plowing and be enhanced by the creation of a uniform
depression storage for water.  Assuming that the area will be quickly reveg-
etated, the potential for surface runoff and erosion would be minimized.
Such areas would not be unlike abandoned fields.  The overall effects upon
infiltration created by the shallow plowing would be greater than those re-
sulting from simple clearcutting but not excessively so.

     On steep shallow soils which are underlain by shallow bedrock the plow-
ing treatment could create a potential hazard for land slips and slides.  By
                                    4-29

-------
 retaining  water  and  with  no  vegetation  to  remove  the  excess  water,  the  soil
 could  become  saturated  and slip.   This  would probably be  true  of soils  with
 low  shearing  indices, such as  those  developed  from  shale  materials.

      (1-6)  Deep Plowing
      It  is  assumed that this treatment  would be accomplished either  with  rip-
 pers,  root  plows or  counter-disk plows, all of which  can  cultivate  to a depth
 of 12  to 18 inches.  The  deepest depth  would be no  greater than  24  inches,
 unless specialized equipment were  used.  Generally, plowing  will  have simi-
 lar  effects to those produced  by shallow plowing.   The treatment  would  of
 necessity be  restricted to the deeper soils.   Infiltration would  be  improved
 over shallow  plowing, and a  greater  kill of herbaceous species would be
 achieved.   The natural succession  from  coppice sprouting  and residual seeds
 in the soil would be delayed for longer periods than  in the  case  of  simple
 clearcutting.  The hazard of slips and  slides  would be increased  in  moun-
 tainous areas.   Small animal habitat, burrows, runs,  etc., would  be  largely
 destroyed.

     Plow pans are common on old oil field soils.  They hinder deep  infil-
 tration, increase interflow    and  impair root  development.   Fragipans and
 hardpans on these often previously eroded soils are also  at  shallow  depths.
 Deep plowing  would break  the pans  if present.  In addition,  for those soils
 where nutrients have leached to lower levels,  deep plowing would  implement
 turnover and mixing of the soil.  Growth and development  of  invading species
 or planted  species would  be  enhanced.

     (1-7)  Soil Cover less  than 25 cm
     Soil fill to even these shallow depths would require large quantities
of soil material and the use of heavy equipment for hauling  and spreading.
The forest floor would be buried,  effectively  smothering much of the life
that  it fosters and the biological/physical interactions that occur in and
near the surface of coniferous forest soils.   Small animal populations would
suffer loss of habitat and the food supply found upon the forest floor.

     Surface runoff will increase  and subsequent erosion of the loose dress-
ing material would be severe.  On  steeply sloping land,  off-site sedimenta-
tion  rates would be high.   The effects of the treatment  would be comparable
                                    4-30

-------
                                     14 23 35
 to the construction of logging roads  '   '    but  more severe since larger
 expanses  of area would be involved,  and  there would be no provisions for
 compaction and drainage as in the case of road construction.

      The  low growing plants of the forest floor would be  buried and killed.
 The roots of the taller,  herbaceous  plants  would  also be  buried and the  less
 hardy of  these should be  expected to die due to lack of soil  aeration.

      Since the soil-covering material  must  be imported, it  is likely that it
 will  not  have come  from a coniferous forest,  and  that it  will probably be a
 mixture of topsoil  and subsoil,  or be  alluvial  material and,  as such, it  is
 likely that  it will  contain low levels of nutrients.   In  any  event,  the ma-
 terial will  be inferior to the native  soil  underneath.  It  will be loose  and
 could provide a good physical  seed bed for  germination and  establishment.
 The residual  seed pool  and nutrient  level will  be  low.  There is also the
 possibility  of importing  plant and animal species  with the  material  that  would
 compete with  the natural  species.  The quality  of  the imported  soil  material
 would be  the  primary factor affecting  early recovery  by natural  means.   If
 the area  is  artificially  rehabilitated,  there should  be no  appreciable delay
 beyond that  induced  by  simple  clearcutting,  unless  the covering  material  is
 of  very poor  quality.

      (1-8)  Soil Cover  25  to 100  om
      The use  of deep  soil  burial  would as a  treatment  tend  to magnify the
 effects mentioned under shallow soil burial.  In addition,  nearly  all of  the
 lower growing  herbaceous  species  would be killed,  and  the root  systems of
 larger broad  leaf species  would be effectively buried and smothered.  This
 soil material  would  contain lower proportions of seeds and  organic matter.
 The potentials for erosion on-site and sedimentation off-site would be in-
 creased because of the greater volume of loose material.

     The destructive effect on small animal habitats would be complete.   De-
velopment  of herbaceous and broad leaf species would be delayed, particularly
on the larger treated areas; this would affect both large  and small animals
for an extended period.
                                    4-31

-------
      (2-1)   Remove  Plow Layer
      Logically,  removal of the  "plow layer" (uppermost  10 cm of soil)  would
 follow stumping  and grubbing.   In  old forest  soils  this would remove humus
 layers and  most  or  all  of  the A horizon.   Many coniferous stands,  particularly
 in  the East and  Midwest, are  fire  or old  field subclimax forests.   Loblolly,
 shortleaf and  slash pine exist  in  the South as the  result of fire  or previous
 clearing and cultivation as do  Jack  pine  in the Midwest and  lodgepole  pine  in
 the West.   The forests  of  white pine in the Northeast and Virginia pine  in
 the Southeast  are classic  examples of old field subclimaxes  which  developed
 on  soils with  only  B or C  horizons remaining.   Removal  of an additional  six
 inches would,  except in soils with deep C horizons  such as those found in
 the eastern and  southern coastal plains,  essentially "sterilize" the areas
 for long periods.   This would be particularly  true  of coniferous forests
 located on  their characteristically  shallow and/or  rocky soils  in  areas of
 steep  terrain.   Surface runoff  and consequential erosion would  be  accelerated.
 The reestablishment  of  the forest, either naturally or  by planting,  would be
 impaired and in  some cases impossible.

     The effects of  this treatment would  be less drastic on  deep soils.  The
 deeper layers  of many coniferous forest soils  have  a higher  nutrient content
 than the surface material.  If  the forest  is artificially rehabilitated on
 these  treated  areas, production should not be  greatly impaired.

     (2-2)  Remove Shallow Root Zone
     It is  assumed that the depth of this  zone  is somewhat greater than that
 of  the plow layer.   Essentially, the effects of the removal  of the described
 plow layer  above would  be applicable to removal of  the Shallow Root  Zone.
 Both treatments would not leave suitable  seed beds  for natural rehabilitation
 and would delay natural succession.

     (2-3)   Remove Scraping and Grading,  Mechanically Stabilize
     The effects of scraping and grading have been described in treatment 1-4
above.   Mechanical treatment in addition to the scraping and grading is as-
sumed to involve compacting of the area with mechanical  equipment  (e.g.,
rollers, vibrators,  rams, etc.).  Thus, it would first  be necessary to clear-
cut the trees,  remove slash,  stump and grub roots,  and  smooth the area.

                                    4-32

-------
      The  smoothing  treatments  could  involve  grading,  scraping,  soil  removal
 and  burial  with  imported  soil.   In any  case,  the  degree  of  compaction  will
 be affected more by the nature of the soil material,  the soil moisture con-
 tent,  and the  compaction  treatment than by the particular smoothing  treat-
 ment.   Therefore, it  is assumed that those areas  to be mechanically  treated
 will  essentially exhibit  similar enviornmental effects and  recovery  rates,
 regardless  of  the smoothing treatment used.

      The  average depths of compaction by various  types of machines are given
 in Table  4-2.  The  actual depth for a specific site will  vary with soil char-
 acteristics.   Generally,  at optimum moisture  contents, compaction potential
 will  increase  in the  order:  heavy clay, silty clay,  sandy  clay, sand,  and
 gravel-sand-clay.
             Table  4-2.   Average possible depth of compaction to
                          10 percent soil voids (from  Armstrong4).

                 Equipment                     Depth Compacted (in.)
          2 3/4-ton smooth-wheel roller             5
          8-ton  smooth-wheel roller                 6
          2 1/2-ton vibrating  roller                5
          2-ton  vibrating plate compactor          12
          D-8  crawler tractor                       6
          Ram  (portable)                            6
     Footed rollers, either sheeps foot or taper foot, comprise a special
class of compactor for cohesive soils.  At least 64 passes are required to
completely cover a given area.  Depth of compaction can reach 18 inches or
more.

     The effects on microclimate of mechanical treatment will not differ
from those of clearcut areas.   Wildlife should find little interest in the
treated areas since it will not support appreciable amounts of vegetation
for several years.
                                    4-33

-------
      (2-4)  Remove  Plow Layer,  Mechanically Stabilize
      The  effects  of this treatment  are  essentially the same as those de-
 scribed in  treatment 2-3 above.

      (2-5)  Remove  Shallow Root Zone, Mechanically Stabilize
      The  effects  of this treatment  are  essentially the same as those de-
 scribed in  treatment 2-3 above.

      (2-6)  Remove  Scraping and Grading,  Chemically Stabilize
      The  result of  the  addition of  a chemical  stabilizer  to stumped  and
 grubbed land which  has  been smoothed would  depend  on the  type  of  chemical
 treatment.  Chemical  emulsifiers will counteract the effects of scraping
 and grading by stabilizing  the  soil bed until  vegetation  is established.
 They  could  speed  succession by  retarding  erosion until  plants  became es-
 tablished.   Chemical deflocculants will  intensify  the  effects of the me-
 chanical  disturbance.   They will reduce infiltration and  increase surface
 runoff.   Less water  will be available for vegetation that might invade the
 site, thus  succession will  be retarded.

      (2-7)  Remove Plow Layer,,  Chemically Stabilize
      The  effects  of  this treatment are essentially  the  same as those  de-
 scribed in treatment  2-6 above.

      (2-8)  Remove Shallow  Root Zone, Chemically Stabilize
      The  effects  of this treatment are essentially the  same as  those  de-
 scribed in treatment 2-6 above.

      (3-1)  Barriers to Exclude People
      The  exclusion of humans would have no negative  effects on  the natural
processes within  the ex.closure, and could be beneficial, especially in areas
of high population density.

      (3-2)  Exlude Large Animals
     Barriers that exclude  large browsers, grazers, and predators will affect
the succession within the exclosure.  With the exclusion of browsing and
grazing animals,  a greater density and diversity of herbaceous  species will

                                    4-34

-------
 develop whether or not the coniferous forest  cover  is removed.   If  the  trees
 are removed, the early stages of succession will be extended and the develop-
 ment of the climax forest will be prolonged.

      The presence of more herbaceous plant species results in making avail-
 able more energy to the small herbivores.  In addition, herbivore population
 should increase due to the absence of some of their predators.  Small preda-
 tors not limited by the exclosure (slumber weasel, etc.) will increase due to
 the additional food supply and absence of competitive predators.

      The elimination of the soil compaction produced by the larger mammals
 may favor the rehabilitation of the forest floor and the reestablishment of
 its ecological  processes.

      The large  biota mammal  habitat  would be  minimally influenced by small
 exclosures.   Large  exclosures (10  km2)  in areas  of limited forest could have
 serious negative  impacts upon the  local  area.  Home  ranges of the larger ani-
 mals  (deer,  elk,  bear,  cougar,  etc.)  are invariably  larger than  the  largest
 treated areas proposed.  In  the  densely  populated  forest areas of the eastern
 United  States,  and  to a lesser  extent in the more  sparsely populated western
 states,  large exclosures could  interfere with  animal movement.

     Barriers that  exclude large browsers, grazers,  and  large predators  will
 ultimately influence the dynamics of succession  within the  exclosure.  An
 immediate reduction in browsing and grazing will favor the  development of
 plant species subject to these processes.  These plant species, which usually
 include the grasses, forbes,  shrubs and palatable tree seedlings, will exper-
 ience a measurable increase in growth, an increased density and therefore
 higher measurable productivity (accumulated biomass).  These  (r-strategy)*
 plant species tend to dominate the earlier stages of plant succession when
moisture, available nutrients and light quality are favorable.  Their removal
favors those less palatable tree species  (k-strategy) typical of later succes-
sional stages.   Therefore  browsing  and grazing can foreshorten plant  succession
and create conditions which favor faster rates of development toward  climax.
'Using the terminology of growth equations where r is the intrinsic rate of
 increase, and k is the upper asymptote or equilibrium population size,   r
 selection predominates in early colonization with k-selection prevailing
 as more and more species and individuals attempt to colonize.5^
                                    4-35

-------
 If  trees  are  removed,  the  early  stages  of  succession  will be  extended, ulti-
 mately  prolonging  the  development  of  a  climax  forest.

      (3-3)  Exclude Large  and Small Animals
      This treatment would  benefit  birds and tree-dwelling animals because of
 the resultant additional availability of protection and food.  The effects on
 the density and diversity  of herbaceous vegetation within the enclosure and
 the subsequent effects on  microclimate, soil compaction and succession would
 be  amplified  over  those resulting  from the exclusion  of large animals.

      It is also expected that any  small animals trapped within the exclosure
 would benefit, particularly if a source of water were available.  Gene flow
 would be restricted, but this effect  would not be significant unless the ex-
 closure remained effective for many reproductive cycles.

      (4-1)  Asphalt Hard Surface Stabilization
     Paving a contaminated area represents a severe impact.  Asphalt may be
 laid either between the trees or after trees have been removed and the site
 smoothed.  Asphalt laid between the trees will result in the gradual (1 to 3
 years) death of the trees  in the treated area.  Coniferous trees have rela-
 tively shallow roots and depend on the infiltration of surface water for
 moisture and nutrients.  Paving the surface will block infiltration, and
 runoff efficiency will be  90 to 98 percent of the rainfall.  Plant or animal
 invasion of the area would be delayed until the asphalt begins to decompose.
 The existence of decaying  trees on the site would speed the process of as-
 phalt breakup by mechanically disturbing the surface as the trees fell.

     (4-2)  Concrete Hard Surface Stabilization
     The effects of this treatment are very severe and are essentially the
 same as described in treatment 4-1 above.   Infiltration of surface water
would be stopped.   Runoff would be virtually 100 percent of rainfall leading
 to extensive off-site erosion.   Establishment  of a community would require
 complete breakup of the concrete, a process that would required decades to
centuries.
                                    4-36

-------
      (5-0)  Application  of Sewage Sludpe
      Sewage Sludge Application Without Clearcutting-The  effects upon  '.Conif-
 erous forests and their  associated wildlife of applying sewage  sludge  to  the
 forest soils are dependent upon the characteristics of individual wastewaters.
 Chemical contaminants in wastewater may be classified as  either inorganic
 (heavy metals) or organic including PCBs, pesticides and  other synthetic ma-
 terials.  Pathogenic biological pollutants include a wide variety of bacter*
 ial,  protozoan, parasitic helminth and viral forms.  Wastewater treatment
processes can provide significant reductions in the levels of these contami-
nants, but conventional treatment processes are not adequate to remove certain
chemical and biological pollutants.

      In general,  based upon studies  in Douglas fir stands17'55'69 the appli-
cation of sewage  sludge may be predicted to have  the following effects:
          1.   An  increase in the  rate  of forest floor decomposition
          2.   An  increase in the  rate  of tree  and ground  cover growth
          3.   A possible  initial  decrease  in surface water infiltration
              in  the  short  run due to  the  presence of  fine sediment  in
              the sludge, but  a net long-term  effect of increased infil-
              tration due to enhanced  ground cover growth
         4.  A decrease  in anaerobiosis in the upper  soil profile
         5.  An increase in coliform  bacteria which will  move through
             the soil profile with the soil solution
         6.  An increase in the flow  of nitrates  through  the soil sys-
             tem accompanied by the stripping of  cations from the soil
             complex
         7.  An increase in carbon dioxide production resulting from
             organic decomposition processes in the soil
         8.  An increase in cationic  losses due to increased mobile
             anion  pools
         9.  An increase in heavy metals in soils and animal tissues
             (see also  Peterson et al.53)
        10.   An increase  in microbial  aerosols, depending  upon the
            method of  application of  the  sewage  sludge.
                                  4-37

-------
      Phosphorus may also  increase, particularly  in those  soils having high
 exchange  capacities; however,  it  is most  likely  that  such accumulation will
                                          27  7R
 occur primarily in the upper soil  layers.   '     Soluble salts of  soil ex-
                                                                         32
 tracts, notably chlorides, sulfates and calcium, will also be increased.
 In general, animal viruses have not been  detected following sewage  sludge
   ...   _   43,69
 application.   '

      Accumulation of heavy metals  in plants  grown on  sludge-amended soil has
 been  found to  be greater  than  that found  in  plants grown  on untreated soil,
 although  toxic levels have not been detected.    Similarly, dangerous levels
 of heavy  metals in the tissues of birds and  small herbivorous mammals sampled
                                               3 18 21
 in sludge-amended areas have not been detected.  '  '

      It may also be expected that populations of larger browsing mammals
 would increase due to the availability of additional forage.  The effects
 of sewage sludge application upon invertebrates will vary from species to
 species.  Earthworms and  other invertebrates will probably tend to  increase
 due to the higher nutrient levels provided by the applied effluent.  If the
 area  is heavily irrigated, the diversity and numbers of most soil microar-
 thropods  will decrease.    Song birds may increase because of greater food
 availability.

     Sewage Sludge Application Combined With Clearcutting—The combined ef-
 fects of  clearcutting and sewage sludge application on soils will be similar
 to those produced by sludge treatment alone.  In addition, the growth of
 ground cover, particularly that of broad leaf species will increase.  Ac-
 cordingly, the return to a coniferous forest will be delayed or possibly
 prevented entirely unless the area is artificially rehabilitated.  The ef-
 fects on wildlife will be essentially those discussed in the clearcutting
 treatment 1-2 above,  but with the additional enhancement of food supply.

 4.5  RECOVERY AFTER CLEANUP
 4.5.1  Irreversible Changes
     Over the very long run (thousands of years)  the consequences of all  of
 the treatments discussed here are theoretically reversible,  assuming that no
major climatic changes occur and that the requisite sequence of events and
 conditions obtain for the successional reestablishment of the original

                                    4-38

-------
  coniferous forest.   Over a period of a hundred years or less,  hard surface
  stabilization,  particularly in the case of concrete treatment,  may be con-
  sidered as virtually irreversible.  Other treatments such as removal of the
  plow layer,  if  done on a shallow soil on a steep hillside,  might  also be ir-
  reversible if no  supplemental  remedial treatments are applied.

  4.5.2   Rates  of Recovery
      The perturbations  discussed  in  this chapter on  cleanup procedures  do  not
  have a  great  deal of precedence.   The only common analogous situations  are
  clearcutting, fire, surface mining,  and  abandoned fields.   Even under these
  situations natural  successions will  not  necessarily  lead to the same  sub-
  climax  or climax coniferous forest that  existed  prior to the disturbance.
  For example, Douglas fir in the west  seldom reestablishes itself naturally
  following clearcutting.  Longleaf pine,  an important species in the south,
  does not always reseed itself on burned  areas.   Loblolly pine, an old field
  species, may not become reestablished on adjacent old fields that are first
  invaded by hardwood species and then protected from fire.

      If it  is desirable to have the treated area return to its former pro-
 ductivity as soon  as possible,  then artificial planting is the restorative
 method  of choice.   Seedlings of nearly all of the important  conifer species
 are grown in State or Federal nurseries.   The largest treatment  area consid-
 ered,  10 km ,  would  require only about 250,000 seedlings,  well within the
 production  capabilities of  any  large  nursery.   In northwestern Mississippi,
 a planting  schedule  of  50,000,000  loblolly pines  annually was maintained for
 many years.

      Table 4-3 summarizes the successional  associations of some major  conif-
 erous forest types.  Recovery sequences over future time  are shown  graphically
 to demonstrate the duration of treatment  effects,  assuming no retrogression
 occurs.  The endpoint for the curves  in Figures 4-3 through  4-5 is restoration
 of productivity to the extent that a  layman would view it as similar to the
productivity of undisturbed areas.  Representative cleanup treatments were
 illustrated in the figures and the application of that information in evalu-
ation of the impact of cleanup on coniferous forest is demonstrated in some
detail in Appendix B.
                                    4-39

-------
                              Table 4-3.  Major forest types of the United States.
-p»
o
Type Name
White Pine

Shortleaf pine


Virginia pine

Loblolly pine-
shortleaf pine
Loblolly pine

Longleaf pine-
slash pine
Douglas fir
(Pacific, 80°6;
interior, 50"0)





Englemann spruce-
subalpine fir



Region
Appalachian

Appalachian §
Piedmont

Appalachian £,
Piedmont
South

South, Coastal
Plain
South, Coastal
Plain
Pacific Coast f,
Rocky Mts. into
Canada, f, south
to Mexico




Rocky Mts. f,
Cascades into
Canada


Principal Associate Species
Yellow-poplar, chestnut, hemlock,
red oak, white oak, scarlet oak
White oak, southern red oak, post
oak

Shortleaf pine, chestnut oak,
white oak, red oak
Persimmon, blackgum, post oak,
red oak
Sweetgum, southern red oak,
blackgum, laurel oak
Water oak, laurel oak, post oak,
1 ive oak
Most western trees native to
local ity






Most western trees native to
locality except "xerophytes"
such as ponderosa pine and
jun ipers

Successional Stage
Long-lived, temporary
on abandoned lands
Climax on certain
areas, old-field type
on others
Temporary, old-field

Subcl imax

Subcl imax

Subc 1 imax

Generally fire type
except in drier
habitats where it
becomes climax




Cl imax , although
type may follow
directly after log-
ging or fire without
intervening stages
Climax Association
Northern hardwoods

Oak -pine-hickory


Oak-pine-hickory

Pine-hardwoods

Pine-hardwoods

Pine-hardwoods

Climax only on re-
latively warm, dry
habitats. Western
redcedar -western
hemlock & spruce
fir succeed it on
cooler, moister
habitats
Spruce-fir





-------
                                                             Table  4-3.   (.Continued)
               Type Name
    Reg ion
                                                       Principal Associate Species
                                                           Successional  Stage
                                                                                                                 Climax Association
Larch-Douglas fir   Washington f,  Ore-
                    gon  east  of
                    Cascades,  &
                    north Idaho f,
                    western Montana,
                    north into
                    Canada
            Lodgcpole pine
            Ponderosa  pine
-p.
i
            Redwood  (20°0)
           Sitka spruce
           Western Hemlock
           (80%)
 Rocky  Mts.  to  Pac-
 ific Const  £  into
 Canada
 Pacific Coast  to
 Black Hills and
 Nebraska
Northern Calif.
f» southern Oregon
North Pacific-
Coast to Alaska
Pacific Coast to
Alaska $ northern
Rocky Mts.
                                                    Most local  trees
                                                    Most  local  trees  but  frequently
                                                    occurs  in  nearly  pure stands
                                       Nearly pure  stands occur over  large
                                       areas, but on moister  sites, most
                                       native trees  intermingle. Douglas
                                       fir and white fir are  most common
                                       associates

                                       Douglas fir, western hemlock,
                                       western redcedar, and  Port-
                                       Orford-cedar

                                       Western hemlock, Douglas fir,
                                       Pacific silver fir, and western
                                       redcedar

                                       All local  species except the
                                       "xerophytes" such as ponderosa
                                       pine
                                                                                          F:ire type
                                                                              Fire  type  but very
                                                                              stable  over  large
                                                                              areas  in Rocky Mts,
Climax on warmest £
driest habitats.
Temporary on cooler,
moister habitats
Climax, but will
also follow fire or
cutt ing

Fire type, but very
stable
Climax, but may also
directly follow fire
or logging
                                                                                  Douglas  fir,  west-
                                                                                  ern  hemlock-west-
                                                                                  ern  redcedar,  or
                                                                                  spruce-fir
 Douglas  fir,  west
 ern  hemlock-west-
 ern  redcedar, or
 spruce-fir

 Ponderosa pine
 within that vege-
 tational zone.
 Hlsewhere, genera-
 lly  Douglas fir
                                                                                                                Western redcedar
                                                                                                                western hemlock
Western redcedar-
western hemlock

-------
CO
UJ
cc
O
u.

CO
D
O
CC
UJ
O
u
o
I-
LU
x
h-
o
a.
            0-1 NATURAL REHABILITATION

               HERB (SHRUB I TREE STAGES
ARTIFICIAL REHABILITATION'
                                  100 200
                                                         lo 1         10         100 200

       "Clear Cutting (1-2), Stumping and Grading (13), Plowing (15, 16), Shallow Sod Burial (17), Flooding (62)




  Figure 4-3.   Sequence  of  ecological  recovery  following  cleanup.
                                       4-42

-------
                                                 HYPOTHETICAL CONIFEROUS FOREST
-pi


CO
 10
 C
 -s
 n>
 GO
 n>
 .o
 c
 n>
 is
 o
 n>

 o
 -h

 fD
 O
 O

 O
 to
 _j.
 o
 £U
 -J
 n>
 o
 o
 <
 fD
CQ

o
c
•a
8
      EROOABILITY
                             HOMEOSTASIS
                                                            PRIMARY PROD.
                                                                                      COVER
                                                                                                 SP. DIVERSITY

                                                                                                 (VEGETATION)

-------
              20
          cc
          O
          Z
          O
          QC
          O
          z
          D
          QC
          Q
          HI
          CO
          <
          LU
          QC
          O
          Z
15
10
               02468
                              YEARS AFTER CLEARCUT

    Figure 4-6.  Decrease  in runoff following  clearcutting  coniferous  forests
                 (after Swane and Mines??).

imported soil material, depending upon depression storage and  infiltration
characteristics, but it will carry heavier sediment loads than expected  from
the simple clearcut treatment.

     The effects of chemical and mechanical stabilizers combined with  other
treatments would be greatest for the soil removal, soil burial  and scraped
treatments, and least for the stumping and grubbing and simple  clearcut  treat-
ment.  Recovery of the area would be delayed until leaching occurred and soil
structure could re-form.  Treatments which left the most organic matter  would
recover more rapidly, but succession on the site should be expected to be
delayed at least 5 years by the treatment.

     With the exception of the chemical emulsions, the application of  stabi-
lizers will require:  (1) breaking up and tilth forming; (2) application and
mixing in of the stabilizer; (3) application of water to, or its part removed
from, the soil to achieve moisture control; and (4)  compaction.  Thus, the
treated area must first be clearcut,  stumped and grubbed, slash removed,
                                     4-46

-------
 smoothed and plowed.   The stabilizing treatment would show effects far in
 excess of those between the additional treatments of topsoil removal and
 burial with imported  soil.   It is therefore assumed that all treatments (e.g.,
 stumping and grubbing,  grading,  smoothing,  burial,  soil removal and plowing)
 will have very similar effects and responses after stabilization.

      Generally,  the effects on wildlife succession and hydrology will be
 greater than those of simple mechanical stabilization.  Succession and wild-
 life habitat will  be  retarded for at  least  an additional 5 to 10 years,  and
 runoff efficiency  will  increase  to above 70 percent.   However,  erosion and
 sedimentation will be considerably less than for the  mechanical treatments,
 but  structures to  handle the excess runoff  will  have  to be installed to  mini-
 mize offsite damage from runoff.

      Complete recovery  of a site  subjected  to a  combined treatment,  in some
 cases,  may take  centuries.   Soil  with high  proportions of montmorillonite  clay
 and  soil  in areas  subject to heavy frosts will  loosen within a  few years.   Soils
 in the  sub-humid areas  where rainfall  is high and plant invasion occurs  at
 higher  rates may require up to 5  years  for  the establishment of a  complete
 ground  cover of  pioneer species.   It  is assumed  that  the area would  not  be
 suitable  for planting.   Thus, depending upon  location,  soil  characteristics
 and  degree  of compaction, succession  will be  delayed  over  that  which might
 occur on  the simple clearcut  area  by  25 to  100 years.   Of  course,  heavily
 compacted areas  would be  very prone to  retrogression.

     This tendency would  be  amplified by the  increased  surface  runoff  and
 erosion that  would result from treatment.  A  compacted  area,  even  if  compacted
byasheeps  foot  roller, and devoid of vegetation and  depression  storage, will
present little friction to flowing water.   Infiltration, particularly  in soils
with clay and  silt components, will be  reduced at least  to 1  inch per hour or
 less.   It is  expected that runoff  efficiency will be on  the order of 70 per-
cent or more.

     The damage from offsite runoff, even from areas of  1 hectare,  could be
considerable and would require the construction of diversions, lined channels,
or dams for water storage and desilting.
                                    4-47

-------
      4.5.2.2  Fifth Year —
      Three years after treatment,  assuming successful rehabilitation of the
 conifer,  if erosion has not been severe and if the site is of good to medium
 quality,  the forest growing conifer species (e.g., loblolly,  short leaf,
 slash,  etc.) will have attained heights of 20 to 60 inches.11  The slower
 growing species,  such as Douglas fir,  will have reached 15 to 20 inches.70
 At  ten  years,  heights of the faster growing species will be on the order  of
 20  to 30  feet,  and those of the slower species will have attained heights of
             28 71
 4 to  5  feet.   '     In the moist temperate  zones,  crowns will  have started to
 close and stratification will  have begun.   In colder regions, there will  still
 be  heavy  competition between hardwood  brush and coniferous species.

      If natural  rehabilitation (0-1) were  relied upon and had been success-
 ful the density  of the new  reproduction might appear as shown in Figure 4-7,
 to  dense  near  the seed source  and  sparse or absent at greater distances.

      The  plowed  areas  will  probably exhibit the best reproduction followed
 in  order  by  stumping  and grubbing  (1-3), simple clearcut (1-2),  soil  burial
 (1-7),  scraping  and  grading (1-4),  and  soil removal  (2-1 through 2-8).  In
 stands  with  associated   hardwoods,  the  effects  of sludge treatment may  be
 evident in a greater development of hardwood  competition.
                             YOUNG TIMBER
YOUNG TIMBER

       Figure 4-7.
                                     TIMBER OF
                                     SEED-BEARING AGE
                        TIMBER OF SEED-BEARING AGE
                                                   WIND AT TIME
                                                   OF SEED FALL
Clearcutting the whole sta,nd, with reproduction
secured by seed disseminated from seed trees lo-
cated outside the cut stand.  The density of the
reproduction 5 years after the cutting is indi-
cated by the dots (from Smith66).
                                     4-48

-------
      The  effects  of  forest  removal  on  surface  runoff will have  diminished
 as  indicated  in Figure 4-4.   Runoff from  the concrete  and asphalt  areas  will
 not be  appreciably diminished since these materials will not have  deteriorated
 greatly.

      In those forests with  associated  hardwoods, the crown cover of hardwoods
 within  clearcut areas will  have developed in some stands to the point of clo-
 sure  and  will provide cover and food for  wildlife.  Use of the  area will be
 enhanced  by the border created by the  clearing.

      Within the exclosure,  competition of hardwood and herbaceous  species may
 have  delayed conifer growth.

      4.5.2.3  Fiftieth Year --
      On clearcut areas (1-2)  with or without supplemental treatment (e.g.,
 stumping  and grubbing, plowing, etc.) which were successfully restocked,
 either naturally or artificially, complete crown cover will have been ob-
 tained and the herbaceous and hardwood understory will have been suppressed.
 Stratification of the forest  into dominants, codominants,  intermediates, and
 suppressed trees will have developed.  The number of trees per unit area,
 except in the boreal  forests,  where such processes occur at slower rates,
will have decreased due to suppression by larger trees as  indicated for rep-
resentative tree age  in conifer forests in Table 4-4.

       Table 4-4.   Approximate age of representative coniferous
                   forests where number of trees per acre  is  re-
                   duced  by suppression.
                                        Approximate age (years)
       Eastern Forests                 ~~~	•
         White - red Jack pine               20 to 30
         Loblolly-shortleaf pine             20 to 25
         Spruce-fir                          30 to 40
       Western Forests
         Douglas fir
         Ponderosa  pine
         Hemlock-sitka spruce
                                     4-49

-------
      The effects of the clearing treatments on wildlife, runoff  and  erosion
 will not be evident.  The wildlife habitat will have been essentially  stabi-
 lized to that of the surrounding forest.  It is possible, but not  likely, that
 the exclosures could remain 100 percent effective, even if maintained, over
 this long a period.

      Asphalt treated surfaces (4-1), if not maintained, will appear as do
 abandoned World War II air strips today, heavily cracked and buckled and
 invaded by grasses and pioneer species.  The concrete treated areas  (4-2)
 will, of course,  be weathered but still intact except for plant invasions in
 cracks or expansion joints.   Offsite scouring and gullying by concentrated
 runoff from the areas will still be active,  assuming that any structures in-
 stalled to prevent this have not been maintained.   The areas treated by
 sludge (5-0)  will show no significant effects of treatment  by this point
 in time.

      It is  not possible to make  general predictions for those sites that may
 have retrogressed as  a result  of the treatment  applied.   The effects could
 range from  exposure of bedrock to the development  of a  permanent  meadow,  de-
 pending upon  the  particular  site and treatment.

      4.5.2.4   100 Years  After  Treatment --
      In all cases except the boreal  forest, assuming  that artificial  rehabi-
 litation or successful natural rehabilitation of clearcut areas has occurred,
 the  new coniferous  forest  will be approaching maturity.  The  forests  of the
 southeast will have passed the point  of commercial  maturity.  The dominant
 trees will have attained heights equivalent to the  representative data  in
 Table 4-5.  Basal area development will  be on the order of data presented
 in Table 4-6.

     On areas treated by scraping, grading, topsoil removal and burial, a
maximum depth of litter will have accumulated on the forest floor.  An  F
horizon will have begun to develop,  but otherwise the rest of the soil pro-
file will show no appreciable development.  The soils of the other treatments
will show no significant change.

     Wildlife habitat  100 years after treatment will not be discernible from
the surrounding habitat,  assuming that no additional perturbations have occurred
                                   4-50

-------
        Table 4-5.  Height of dominants in a  fully stocked stand at
                   100 years- good to medium site fully stocked.
                                              Height (feet)
       Eastern Forests
           White - red Jack pine
           Loblolly-shortleaf pine
           Spruce-fir
       Western Forest
           Douglas fir
           Ponderosa pine
           Hemlock-sitka spruce
120 to 140
 90 to 100
 40 to 120
Table 4-6.   Approximate basal  area (m2/ha) of representative conifer species -
            good to medium site, fully stocked (after Forbes22).
Eastern Forests
Age
(yrs)
20
30
40
50
60
70
80
90
100
110
120
130
140
150
Eastern
Loblolly Slash white
pine pine pine
29 32
35 36 41
37 37 48
38.5 37.5 53
39.5 38 55.5
57.5
59
60.5
61
61.5
62



Red
spruce


29.5
44
38.5
49
53.5
57
58.5
59.5




Jack
pine
29
32
33
33
32.5
31.7
29







Western Forests
Ponderosa
pine
19
35
45
48.5
49
49
49
49
49





Douglas
fir
21
32
40.5
47
52
56
59.5
62.5
65
67
69
71
73
74.5
Sitka spruce
and
western
hemlock
25.5
43
50.5
56.5
62
66
69
72
74
76
77.5
78.5
80
81
                                    4-51

-------
to either the naturally rehabilitated or treated forest areas.  However, it
is unlikely that the exclosures will have been maintained for this long a
period.  Thus, evidence of the exclosures and their effects would not be
apparent.

     There is no precedent for predicting the effects on succession of the
concrete and asphalt treatments.  It is reasonable that natural succession
will have been delayed for at least 50 years by the asphalt treatment and
more than 100 years for the concrete treatment.  Even after such a long
period, the spatial patterns of the process would be controlled by the
breakup pattern of the covering.

     The succession processes on the clearcut treatments will be approach-
ing maturity.  In areas where hardwoods are the climax species, broad leaves
will begin to dominate.  For other climaxes (Ponderosa pine in Arizona, for
example) the area will support a pure (unmixed) mature stand.  However, the
surrounding stand will have become over-mature and have begun to retrogress.

4.6  QUANTITATIVE ASSESSMENT OF CLEANUP IMPACTS
     The relative severity of the different cleanup treatments on various
environmental parameters is shown in Figures 4-3 and 4-5 and Appendix B,
Table B-3.  It is apparent that the most severe cleanup techniques are hard
surface stabilization  (4-0) and soil removal (2-1 through 2-8), while the
least severe treatments are exclusion by barrier (3-1 and 3-2) and applica-
tion of sludge (5-0).   This reflects the unfortunate fact that there is an
inverse relationship between a treatment's effectiveness in contaminant de-
mobilization and its destructive environmental impact.
             40
     Lawrence   has described the hypothetical development of a spruce-
hemlock forest following glacial retreat.   He estimated that climax could
be reached within 130 to 200 years,  at a median of 170 years.  Other authors
agree that the duration of sere from bare mineral soil to climax coniferous
forest is about 200 years.   Using this generalization,  and noting the dura-
tion of successional periods in the development of mature forest given by
         40
Lawrence,   a series of index numbers from zero to nine (not measurable to
irreversible) can be imposed on the set of successional periods.  These index
numbers can then be applied to the cleanup treatments performed over different

                                    4-52

-------
  areas,  the  index numbers  representing  the  approximate point  in  the  sere  to
  which the cleanup  treatment retrogresses the  climax  forest.  The result  serves
  as  an indication of the relative  severity  of  the  treatments.  Table 4-7
  presents this analysis for different sizes of disturbed areas and the details
  are given in Appendix B.

      As expected from consideration of Table  4-7, the most severe treatments
  over all areas are mechanical soil stabilization  after soil removal  (2-3 to
  2-5) and hard surface stabilization (4-0).  In consideration of treatments
  2-3 through 2-5, it must be realized that erosion of the land after  scraping
  and compaction is a very important factor in the  indicated severity  of these
  treatments.   With treatment 2-3, the removal of only the upper five  centi-
 meters of soil will assure that some roots remain, from which vegetation
 will emerge to check erosion during the first season after treatment.  How-
 ever,  as deeper soil layers are removed through treatments 2-4 and 2-5,  lit-
 tle or no rhizomal  material will remain to sprout erosion-inhibiting vegeta-
 tion.   Because of the  compaction of the soil,  seeds  blown in from surrounding
 areas  will  not find germination sites  until rills and gullies have started
 to form, particularly  on  sites with greater slope.  In almost all  conceivable
 situations,  human intervention,  in the  form of gully control  and artificial
 seeding  and  planting,  will be  required  to  ensure  recovery of  the coniferous
 association  within  half a  millenium.

     Hard surface stabilization,  treatment  4-0, also  represents  serious de-
 structive impact  to the coniferous forest.  Asphalt will  decompose at rates
 determined by  climate, slope aspect, asphalt composition,  and vigor  of me-
 chanical pavement destruction  by vegetation and frost.  However, concrete,
 due  to its inorganic nature and resultant resistance  to decomposition, will
 have consistently higher durability than asphalt  in various biomes.   If the
 earth is compacted  or any  exotic subgrade is used  below the concrete, vege-
 tative recovery will be further inhibited to the point where human interven-
 tion will be required to ensure recovery of the coniferous association.  Small
 areas of high-level contamination may be served best through this treatment,
 since it is the most effective and provides the greatest public safety.
However,  for areas larger than 0.1 km2,  this treatment can be considered
unacceptable.
                                     4-53

-------
               Table 4-7.
Recovery index in coniferous  forest for various cleanup treatments
(see Appendix 3).
Treatment
0-1 Natural Rehabilitation
1-1 Chemical Stabilization
1-2 Remove Vegetation by Clear Cutting
1-3 Remove Stumps and Grub Roots
1-4 Surface Scraping and Grading (<-5 cm)
1-5 Shallow Plowing (-.10 cm)
1-6 Deep Plowing (10 to 40 cm)
1-7 Soil Cover (- 25 cm)
1-8 Soil Burial (25 to 100 cm)
2-1 Remove Plow Layer Soil (10 cm)
2-2 Remove Shallow Root Zone Soil (-40 cm)
2-3 Scrape Surface, Remove (-5 cm), Mechanically Stabilize
2-4 Remove Plow Layer (10 cm), Mechanically Stabilize
2-5 Remove Shallow Root Zone (-,40 cm), Mechanically Stabilize
2-6 Scrape Surface, Remove (--5 cm), Chemically Stabilize
2-7 Remove Plow Layer (10 cm), Chemically Stabilize
2-3 Remove Shallow Root Zone (40 cm), Chemically Stabilize
3-1 Barriers to Exclude People
3-2 Barriers to Exclude Large Animals
3-3 Barriers to Exclude Large and Small Animals
4-1 Asphalt Hard-Surface Stabilization
4-2 Concrete Hard-Surface Stabilization
5-0 Sewage Sludge Application
6-1 Washing (--3 mm)
6-2 Flooding (3 to 30 cm)
7-0 Soil Amendments Added
Area Disturbed
(kii2)
0.01
4
2
2
2
2
2
4
4
4
5
7
8
8
4
4
4
0
0
0
7
8
0
*
*
*
0.1
4
2
2
2
2
2
4
4
4
5
7
8
8
4
4
4
0
0
0
7
8
0
*
*
*
1.0
5
3
2
2
2
2
4
4
4
6
7
8
8
4
4
4
0
0
0
7
8
0
*
*
*
10.0
5
3
2
2
3
2
5
5
6
7
7
8
8
5
5
5
0
0
0
7
8
0
*
*
*
Assumptions
No seed reserve, average seed bed
Seed reserve, average seed bed
Seed reserve, good seed bed
Seed reserve, gooc seed bed
Some seed, poor seed bed, erosion
Some reserve, good seed bed
No seed, average seed bed, erosion
No seed, average seed bed, erosion
No seed, poor seed bed, erosion
No seed reserve, very poor compacted seed bed
No seed reserve, very poor compacted seed bed
No seed reserve, very poor compacted seed bed
No seed reserve, very poot compacted seed bed
No seed reserve, very poor compacted seed bed
No seed reserve, very poor compacted seed bed
Climax is initial stage
Climax is initial stage
Climax is initial stage

At least 50 years required for concrete breakup




  -     -        , .

Greater impact shown by larger number (see Appendix B)

-------
       Other  treatments  have  serious  impacts  when  applied  over large  areas.
  Removal  of  the  plow  layer  (treatment  2-1) or  shallow  root  zone  (treatment
  2-2)  has serious  effects over  the larger areas.   First,  because  the best
  soil  is  removed by the treatment, new topsoil must be formed before most
  species  can return.  Second, the effects of erosion will be  much more  severe
  over  the large  areas than over the  small areas.   That  is,  over areas of from
  10 to 100 km2 erosion will proceed  almost as rapidly  as  soil  formation and
  vegetation recovery, while over areas of 0.01 to  1 km2 vegetative recovery
  will occur much faster than erosion.  This  is because the  turbulent currents
  which produce accelerated water erosion and the high  surface wind velocities
  which produce accelerated wind erosion rarely occur over small areas, while
  they are common over large, barren tracts.   An exception would be the case
  of a small area of high slope,  where water erosion would be quite severe.

      Finally,  chemical  soil stabilization (treatment 1-1) and soil cover
  (treatments  1-7 and 1-8)  would  have  serious  impacts at the scale of 100 km2.
 In the case  of chemical stabilization this  is  because, over such a large
 area,  erosion  would probably overwhelm the  stabilization  effect  of the  com-
 pound,  thus  pushing vegetative  recovery  to or  beyond  the  maximum natural
 recovery  span.   The importance  of  erosion over larger  areas is also  the
 reason why soil  cover is  considered  a  drastic  treatment at  the scale of
 100 km2.   Natural  reseeding  and sprouting from stumps  and rhizomes simply
 will not  occur fast enough at this areal scale  to  counter the destructive
 impacts of erosion and mass  wasting.   The vulnerability to  erosion from these
 treatments also  makes them undesirable for contaminant  demobilization;  it
 makes  little sense to use a  treatment  which  is ineffective  and causes great
 environmental damage.

     Three treatments are benign,  or perhaps beneficial, to coniferous for-
 est:  barriers (treatments 3-1 and 3-2) and application of  sewage sludge
 (treatment 5-0).   Sludge application may be effective in demobilizing a con-
 taminant,  and therefore  is a preferred treatment,  especially in sensitive
 areas where low-level contamination has occurred;  however, the possibility
of contaminant  entry into the food  chain  must be evaluated prior  to its  use.

     The next group of relatively benign  treatments includes clearcutting
 (treatment 1-2),  stumping  and grubbing  (treatment  1-3), and  shallow and  deep
                                    4-55

-------
plowing (treatments 1-5 and 1-6).   Deep plowing and stumping and grubbing have
greater impact at the scale of 100 km^ for the same reason as chemical stabil-
ization, namely that colonizing vegetation is usually overwhelmed by erosion
in the first few seasons after treatment.  Surface scraping and grading (treat-
ment 1-4)  has greater impact at the 10- and lOO-km^ scales than over smaller
areas because the treatment accelerates erosion while, through topsoil removal,
intrinsically decreasing the rate of revegetation.

     Finally, the group of treatments which involve chemical stabilization and
soil removal, treatments 2-6 and 2-8, have serious, though not drastic, impacts
on coniferous forest associations.  Quantitatively and qualitatively, their
short-term effects and effectiveness are identical to their counterparts
(treatments 2-3 through 2-5) which involve mechanical, rather than chemical,
stabilization.  The difference is that most chemical stabilizers do not com-
pact the soil to produce an environment hostile to seedling growth, but they
do effectively retard erosion while vegetation establishes itself.  By the
time the stabilizer breaks down, in one to five growing seasons, vegetation
will be firmly established and will continue to retard erosion as the chemical
stabilizers decompose.  This is true, in most cases, even over large areas,
although more durable stabilizers should be chosen to give plants more than
one season to become established over large areas.

4.7  CONCLUSIONS
     The treatments which will have the least, and approximately equal, effect
will be exclusion by barrier (treatment 3-1 and 3-2) and sewage sludge appli-
cation  (treatment 5-0).  These treatments may have a beneficial effect on the
development and perpetuation of the climax association.  Among the treatments
which require the removal of the forest, recovery would be most rapid with
clearcutting  (treatment 1-2) and shallow plowing  (treatment 1-5), with stump-
ing and grubbing (treatment 1-3) and deep plowing  (treatment 1-6) having
slightly longer recovery periods only at the lOO-km^ scale.
     All of these treatments would be acceptable.  In cases other than low-
level contamination, restriction by barriers should be used only in remote
areas, because barriers provide no significant, immediate demobilization to
a contaminant.
                                    4-56

-------
     Treatments  involving mechanical compaction  (treatments  2-3 through  2-5)
and paving  (treatment 4-0) would have  severe impacts over all areas.  These
treatments  should be prohibited at all scales greater than 1 km2, and should
be used over smaller areas only in the case of contamination severe enough
to prohibit the use of less damaging, though highly effective, techniques
such as those involving chemical stabilization (treatments 1-1, 2-6, 2-7, and
2-8).  Soil cover (treatments 1-7 and 1-8) should be used only for contami-
nants which would naturally degrade before natural erosion removes the soil
cover.   The exact determination of the proper treatment must be made consid-
ering the quantity and quality of released contamination, as well as the
proximity of the site to human populations and activities.
                                    4-57

-------
4.8  CONIFEROUS FOREST REFERENCES
     1.   Anderson,  H.  W.,  and C.  H.  Gleason.  Logging Effects on Snow, Soil
             Moisture, and Water Losses Western Snow Conference, Proceedings
             27:57-65, 1959.

     2.   Anonymous,  Forest Cover Types of North America.  Report of the Committee
             on Forest Types, Society of American Foresters, 1954.

     3.   Anthony,  Robert A.,  and C.  W.  Wood.  Effects on Municipal  Wastewater
             Irrigation on Wildlife  Habitat.  Symposium on Municipal Wastewater
             and Sludge Recycling on Forest Land and Disturbed Land.  Penn.
             State  University,  School of Forest Resources, Philadelphia,  Penn
             1977.

     4.   Armstrong,  C.  F.   Soil Mechanics in Road Construction.   E.  Arnold,
             London, England,  1961.

     5.   Avery,  C. C.,  and L.  J.  Fritschen.   Hydrologic and Energy  Budgets of
             Stocked and Non-stocked Douglas-fir Sites  as Calculated by Meteo-
             rological  Methods,  1971.

     6.   Berndt, H. W.  Some  Influences  of Timber Cutting on Snow Accumulation
             in  the Colorado  Front Range.   Rocky Mountain Forest and Range
             Experiment Station,  Research Notes 58,  1961.   3 pp.

     7.   Bock, J. H.,  and  C.  E. Bock.  Natural  Reforestation in  the  Northern
             Sierra Nevada—Donner Ridge Burn.   Proc. Annual Tall Timbers  Fire
             Ecology Conference 9:119-126,  1969.

     8.   Bock, J. H., C. E. Bock, and V.  E.  Hawthorne.   Further  Studies of
             Natural Reforestation in the  Donner Ridge  Burn.   Proc.  Annual
             Tall Timbers  Fire  Ecology Conference  14:195-200,  1974.

     9.   Bormann, Frank, Likens,  Fisher,  and Pierce.  Nutrient Loss  Accelerated
             by Clearcutting of a Forest  Ecosystem.   Science 159:882-884,  1968.

   10.   Boyce, S. J.,   and D.  J. Neebe.  Trees  for Planting  on Strip Mined Land
             in Illinois.  USFS Div.  of  Mgmt. Techn.  Paper No. 164,  1959.
             33 pp.

   11.  Campbell, T. E., and W. F. Mann, Jr.   Regenerating  Loblolly Pine  by
             Direct Seeding, Natural Seeding, and Planting.  U. S. Department
             of Agriculture Forest Service Research Paper  SO-84,  1973.

   12.  Chapman, H.  H.  Is the Longleaf Type a Climax?  Ecology  13:328-34,  1932.

   13.  Cluff, B.   Engineering Aspects of Water Harvesting  Research.  Univ. of
            Arizona Proceedings of the Water Harvesting Symposium,   1975.

   14.  Copeland,  0.  L.  Forest Service Research in Erosion Control.  Vol.  12(1)-
             75-79 (reprint!,  1969.
                                     4-58

-------
  15.  Daubenmire, R. F.  Vegetational  Zonation  in  the  Rocky Mountains.
          Botanical Review 9:325-93,  1943.

  16.  Dindal, D. L.  Effects of Municipal Wastewater Irrigation  on  Community
          Ecology of Soil Invertebrates.  Symposium on Municipal  Wastewater
          and Sludge Recycling on Forest Land and  Disturbed Land.   Penn
          State University, School of  Forest Resources,  Philadelphia  Penn
          1977.

  17.  Edmonds, R. L.  Microbiological  Characteristics of Dewatered  Sludge
          Following Application to Forest Soils and Clearcut Areas.  Symposium
          on Municipal Wastewater and  Sludge Recycling on Forest  Land and Dis-
          turbed land.   Penn.  State University, School of Forest  Resources
          Philadelphia, Penn.,  1977.

 18.  Ellertson,  R., and G.  R.  Geoffrey.  On Uplake of Redwinged  Blackbirds
          Nesting on Sludge-Treated Soils.   Symposium on Municipal Waste-
          water and Sludge  Recycling on Forest Land and Disturbed Land,  Penn.
          State University,  School of Forest Resources, Philadelphia  Penn
          1977.

 19.   EPA.   Processes,  Procedures  and Methods to Control  Pollution Resulting
          from Silvicultural Activities.  EPA 430/9.73-010 Office of Air and
          Water Programs,  1973.  pp.  91.

 20.   Fillmore, W.  J.,  D.  I. Aldrich,  J. S.  Barrows,  R.  B.  Perry, and B.  F.
          Wake.   Is  Prescribed  Burning Compatible  with  Environmental Quality?
          In:   Proc. Tall Timbers  Fire Ecology Conference 14:627-44, 1974.

 21.   Fitzgerald, R.  Recovery  and Utilization  of  Strip-Mined Land by Appli-
          cation  of  Anaerobically  Digested Sludge  and  Livestock Grazing.
          Symposium  on  Municipal Wastewater  and  Sludge  Recycling  on  Forest
          Land  and Disturbed Land,  Penn.  State  University,  School  of Forest
          Resources, Philadelphia,  Penn., 1977.

 22.   Forbes, R. D.  (editor).  Forestry Handbook.   Ronald  Press Co.,  New  York,
          1961.

 23.   Fredrickson, R. L.  Sedimentation After Logging Road Construction in a
         Small Western Oregon Watershed.  USDA Misc. Publ.  970,  No.  8, 1965.

 24.  Gary, H. L.   Snow Accumulation and Snowmelt as Influenced by a Small
         Clearing in a Lodgepole  Pine  Forest.  Water Resources Research  10T2V
         348-353, 1974.                                      ~           w^j.

25.  Geiger, R.  The Climate Near the  Ground.  Oxford University  Press
         1961.  pp.  611.

26.  Griggs, R. F.   The Edge of the Forest in Alaska and the Reasons for its
         Position.   Ecology,  15:80-96,  1934.
                               4-59

-------
 26.  Harris, A. R.  Physical and Chemical  Changes  in  Forested  Michigan  Sand
          Soils with Effluent and Sludge Fertilization.   Symposium  on  Munici-
          pal Wastewater and Sludge Recycling on  Forest  Land  and  Disturbed
          Land.  Penn. State University, School of  Forest Resources, Phila-
          delphia, Penn., 1977.

 28.  Harris, A. S.  Clearcutting Reforestation and Stand Development.   Journal
          of Forestry 6(330), 1974.

 29.  Haupt, H. J.   Relation of Wind Exposure and Forest Cutting  to Changes
          in Snow Accumulation.   In:  International Symposia  on the Role of
          Snow and Ice in Hydrology, Banff, Alberta, 1972.

 30.  Hibbert,  R.   Forest Treatment  Effects on Water Yield.    In:  International
          Symposium on Forest Hydrology.  (Sopper and Lull,  ed.)  Pergamon
          Press, New York,  1967.   pp.  813.

 31.  Hoover, M. D.  Effect  of Removal of Forest Vegetation  Upon Water Yields.
          American  Geophysical Union,  Transactions 25 969-975, 1944.

 32.  Hornbeck,  J.  W.,  M.  T.  Katerba,  and R. S.  Pierce.  Effects of Sludge
          Application  in  a Northern  Hardwood Forest  on Soil  Water Chemistry
          and California  Population.   Symposium  on Municipal Wastewater and
          Sludge Recycling on Forest Land and Disturbed Land.   Penn. State
          University,  School  of Forest Resources,  Philadelphia,  Penn.,  1977.

 33.   Horwitz,  E. C. J.   Clearcutting:   A View from  the Top.  Washington, D.C.:
          Acropolis Books  Ltd., 1974.

 34.   Johnston,  R.  S.   Soil Water  Depletion  by Lodgepole Pine  on Glacial  Till,
          U.S.  Forest Service, Intermountain Forest  and Range  Experiment  Sta-'
          tion,  Ogden,  Utah,  Research  Note  INT-199,  1975.  8 pp.

 35.   Kidd,  W. J.,  Jr.  Soil  Erosion Control Structures on Skid  Trails.  USFS
          Research  Paper  INT-1, 1963.   8 pp.

 36.   Kittredge, J.  Forests  and Water Aspects Which Have Received Little
          Attention.  J. For., 34:417-19,  1936.

 37.   Komarek, E. V., Sr.  Fire Ecology  Review.  Proc.  Annual  Tall Timbers
          Fire Ecology Conf.   14:201-16,  1974.

 38.   Kuchler, A. W.  Potential Natural  Vegetation of the  Coterminous United
         States.  American Geographical Society Special  Publication No.  36.
         New York:  American Geographical Society,  1964.  144 pp.

 39.   Larsen, J'. A.   Fires and Forest Succession in  the Bitterroot Mountains
         of Northern Idaho.   Ecology 10:67-76, 1929.

40.   Lawrence, D.  B.  The Glaciers and Vegetation in Southeastern Alaska.
         American Scientist,  Summer, 1958.  pp.  89-122.
                              4-60

-------
  41.   Lee,  C.  H.   Total  Evaporation for Sierra Nevada Watersheds by the
          Method  of Precipitation and Runoff Differences.   Trans  Amer
          Geophys.  Union,  Pt.  I.   (Cited from Costing,  1956), 1941, pP.'sO-66.

  42.   Little,  L.,  Jr.  Atlas  of United States Trees:   Volume 1.  Conifers and
          Important Hardwoods.   U.  S.  Dept.  of Agriculture.   Misc  Publica-
          tions No.  1146  (undated).

  43.   Lui-Hing, C.,  and S. J. Sedita.   Viral  and  Bacterial  Levels Resulting
          from the  Land Application  of Digested Sludge.  Symposium on  Munici-
          pal Wastewater and Sludge  Recycling on  Forest  Land and Disturbed
          Land, Penn. State University,  School  of Forest Resources,  Philadel-
          phia, Penn., 1977.

 44.  Lull,  H. W.   Soil Compaction on  Forest  and  Range Land.  U.S  Forest
          Service,  USDA Misc. Pub. #768,  1959.

 45.  Lull,  H. W., and L.  Ellison.   Precipitation  in Relation to Altitude in
          Central Utah.   Ecology 31:479-484,  1950.

 46.  Lyon,  L. J., and P.  F.  Stickney.  Early Vegetal Succession  Following
          Large Northern Rocky Mountain Wildfires.  Proc. Tall Timbers Fire
          Ecology Conference 14:355-75, 1974.

 47.  Martinelli,  M., Jr.   Watershed Management in the Rocky Mountain Alpine
          and Subalpine  Zones.   Rocky Mountain Forest and Range Experiment
          Station Research Note RM-36, 1964.   7 pp.

 48.   May, R.  F.,  and W.  D.  Striffler.   Watershed  Aspects of Stabilization
          and Restoration  of  Strip-Mined Areas.   In:   International Symposium
          on Forest Hydrology.   Ed.  Sopper and Lull.   Pergamon Press, New
          York,  1967.  pp.  663-71.

 49.   Mikola,  P.   Mychorrhizal  Inoculation in Afforestation.   In:  International
          Review of Forestry  Research.   Ed. J.  RomBerger  and P.  Mikola.   Volume
          3,  New York, Academic  Press,  1970.   pp.  123-196.

 50.   Odum.   Fundamentals of  Ecology,  Saunders,  Philadelphia,  1971.

 51.   Costing, H. J.  The Study  of Plant  Communities:  An Introduction  to
         Plant Ecology.   Second Edition.   San Francisco; W.  H. Freeman  and
         Company,  1956.   440 pp.

52.   Packer, P. E.  Forest Treatment Effects  on Water Quality.   In-  Interna-
         tional Symposium on Forest Hydrology.  Ed. Sopper and Lull   Per-
         gamon Press, New York, 1967.  pp. 687-99.

53.  Peterson, J.  R., R.  I. Pietz, and C. Lui-Hing.  Crop Production on
         Illinois Coal Mine Spoils Amended with Digested Sewage  Sludge
         Symposium on Municipal Wastewater and Sludge Recycling on Forest
         Land and Disturbed Land,  Penn. State University, School  of Forest
         Resources, Philadelphia,  Penn., 1977.
                                   4-61

-------
  54.  Rich,  L. R., and J.  R.  Thompson.   Watershed  Management  in  Arizona's
          Mixed Conifer Forests:  The Status  of Our  Knowledge.   U.S.  Forest
          Service, Rocky Mountain Forest  and  Range Experiment Station,  Ft.
          Collins, Colorado.  Research Paper  RM-130,  1974.   15 p.

  55.  Riekenk, H., and R.  J.  Zasoski.  Effects  of  Dewatered Sludge Applica-
          tion to a Douglas Fir Forest Soil on  the Ground Water  and Soil
          Solution Compositions.  Symposium on  Municipal Wastewater and
          Sludge Recycling on Forest Land and Disturbed Land, Penn. State
          University, School  of Forest Resources,  Philadelphia,  Penn.,  1977.

 56.  Roth,  P. L., and G.  T. Weaver.  Survival, growth and heavy metal uptake
          of Woody Plant Species on Sludge-Treated Spoils in  the Palzo Mine.
          Symposium on Municipal Wastewater and Sludge Recycling on Forest
          Land and Disturbed Land,  Penn.  State University, School of Forest
          Resources,  Philadelphia,  Penn., 1977.

 57.  Russel, W.  L.,  and J. L. Thames.   Water Use and Management of Ponderosa
          Pine in Northern Arizona.   Progressive Agriculture in Arizona 20f5)•
          10-11,  SWRA W70-01829, 1968.

 58.  Shantz, H.  L.,  and R. Zon.   The Physical Basis  of Agriculture:   Natural
          Vegetation.   In:   Atlas  of American Agriculture  (Pt. 1, Sect.  E).
          U.S.  Dept.  of Agriculture,  Washington, D.C., 1924.   29*pp.

 59.  Sharpe,  G.  W.,  C.  W.  Hendee,  and S.  W.  Allen.  Introduction to  Forestry,
          4th ed.  McGraw  Hill Book  Company,  New York, 1976.

 60.   Shearer,  R.  C.   Early Establishment of  Conifers Following Prescribed
          Broadcast Burning in Western Larch/Douglas  Fir Forests.  Proc.  Tall
          Timbers  Fire  Ecology Conf.  14:481-99,  1974.

 61.   Shapherd, J.  The Forest Killers:   The  Destruction of the American
          Wilderness.  Weybright and  Talley,  New York, 1975.

 62.   Shigo,  A. L.  Successions  of Organisms  in  Discoloration  and Decay  of
          Wood.  Intl. Review  of Forestry  Research, Vol. 2.  Ed.  J. A.
          Romberger and P.  Mikola.  New York:  Academic Press,  1967.

 63.   Shreve, F.  A Map of  the Vegetation  of the United States.   Geog  Rev
          3:119-125, 1917.

 64.   Sindelar, W., R. Atkinson, M. Majarus, and K. Proctor.   Surface-Mined
          Land Reclamation  Research at Colstrip, Montana.  Res. Report No. 69
         Montana AES Bozeman, Mont., 1974.  98 pp..                       '    '

 65.  Smiley, F. J.  A Report Upon the Boreal Flora of the Sierra Nevada of
         California.   University of California Publication in Botany 9  1921
         423 pp.                                                      " '  "    *

66.  Smith,  D. M.   The Practice of Silviculture.  John Wiley  § Sons  Inc
         New York, 1962.   578 pp.                                  -»     •»
                                 4-62

-------
67.   Stark, N.   Fuel Reduction—Nutrient Status and Cycling Relationships
         Associated with Understory Burning in Larch/Douglas Fir Stands.
         In:  Proc. Tall Timbers Fire Ecology Conference.  14:573-96.

68.   Steele, R.   Smoke Considerations Associated with Understory Burning
         Associated with Larch/Douglas Fir.  In:  Proc. Tall Timber Fire
         Ecology Conference, Missoula, Montana, 1974.

69.   Stendnick,  J. D., and Woodridge.  Effects of Liquid Digested Sludge
         Irrigation on a Second Growth Douglas Fir Stand.  Symposium on
         Municipal Wastewater and Sludge Recycling on Forest Land and Dis-
         turbed Land, Penn. State University, School of Forest Resources,
         Philadelphia, Penn.

70.   Strothmann, R. 0.  Douglas-Fir Survival and Growth in Response to Spring
         Planting dta and Depth.  USDA Forest Service Research Note PSW-228,
         1971.

71.   Strothmann, R. 0.  Douglas-Fir Seedlings Planted by Four Methods...
         Results After 10 Years.  USDA Forest Service Research Note PSW-310,
         1976.

72.   Swane, W.  T., and N. H. Mines.  Conversion of Hardwood Covered Water-
         shed to White Pine Reduces Water Yield.  In Water Resources Research,
         4:947-954,  1968.

73.   Thames, J. L., and S. J. Ursic.  Runoff as a Function of Moisture-
         Storage Capacity.  Journal of Geophysical Research, 1960.

74.   Tunus, R.  W.  Characteristics of Seedlings with High Survival Potential.
         In:  Proc. of the North American Containerized Forest Tree Seedling
         Symposium.  Great Plains Agricultural Council Publication No.  68,
         August 1974.  pp. 276-82.

75.   Tourney, J. W., and C. F. Korstian.  Foundations of Silviculture.   John
         Wiley § Sons, Inc., New York, Chapman  $ Hall  ltd., London,  1947.
         467 pp.

76.   Troendle, C.  A.  A Comparison of Soil-Moisture  Loss From Forested  and
         Clearcut  Areas in West Virginia.   U.S. Forest Service,  Research
         Note NE-120, 1970.  8 pp.

77.   U.S.  Forest Service.  The Outlook for  Timber in the United  States.
         USFS Forest  Resource Report No. 20,  1974.

78.   Urie,  D. H.   Nutrient Recycling  in  Forests Under  Effluent and Sludge
         Treatments  in Michigan.  Symposium on  Municipal Wastewater  and
         Sludge Recycling  of Forest  Land and  Disturbed Land, Penn. State
         University,  School of  Forest Resources, Philadelphia. Penn.,  1977.

79.   Wali,  M. K.,  and P. G. Freeman.  Ecology of Some Mined Areas in North
         Dakota.   In:  Some Environmental Aspects of Surface Mining  in  North
         Dakota, North Dakota Geological Survey Educational Series,  ed.
         M. K. Wali,  Grand  Forks, North  Dakota, 1973.
                                  4-63

-------
 80.  Weaver, J.  E.,  and  F.  E.  Clements.   Plant  Ecology,   McGraw Hill  2nd
         Edition,  1938.  pp.  601.


 81.  Wells, B. W.  Plant Communities of  the  Coastal  Plain of North  Carolina
         and Their Successional Relations.   Ecology,  9:230-242,  1928.

 82.  Whittaker,  R. H.  Communities and Ecosystems.  Macmillan  Publishing
         Co., Inc.,  1970.   385 pp.

 83.  Whittaker, R. H., and W. A. Niering.  Vegetation of  the Santa  Catalina
         Mountains, Arizona.   (II) A Gradient Analysis of the  South Slope
         Ecology 46:429-452, 1965.

84.  Woodbury,  A. N.   Distribution of Pigmy Conifers in Utah $ Northeastern
         Arizona.  Ecology 35:473-489,  1947.
                                 4-64

-------
                                 CHAPTER 5
                           AEOLIAN MOUNTAIN PEAKS
5.1  OVERVIEW
     The discussion in this chapter will largely be restricted to high mountain
regions and peaks that lie above the tundra life zone, but much of the material
may be applicable to desert mountains of much lower elevation and the barren
peaks and outcroppings of forested mountain ranges.  Generally, the environ-
ments to be considered herein consist of steep slopes and barren rock; depend-
ing upon the altitude and latitude, ice and snow and harsh meteorological
conditions may be present, unmitigated by the protection of vegetative cover.
Depending upon geographic location, tundra, desert and various forest life
zones surround the mountain peaks at lower elevations, frequently in close
proximity.  Because of the fact that the surrounding life zones are a priori
"downstream," it is likely that cleanup operations may ultimately involve more
than the primary site of a spill, if such is a mountain peak.  If containment
or stabilization of a spill on a mountain top is not effected by the cleanup
procedures implemented, the discussions in the reports for other life zones
will become applicable at lower elevations.  It should also be noted at the
outset of this discussion that probably few, if any, of the cleanup procedures
covered in this series of reports will be either effective or logistically
possible on many mountain peaks due to the severity of the physiographic and
meteorological components of the environment.

5.1.1  The High Mountain Peak Environment
     The vegetational zonation found in mountain ranges varies largely with
the latitude, altitude, slope aspect, mass of the mountain range, prevailing
wind patterns and the location of the range with respect to land and oceanic
masses.  These factors are discussed in some detail in the coniferous forest
chapter.
                                     5-1

-------
     Life zones in the uppermost mountain regions of the world have been
                      43
described by Whittaker   as arctic-alpine desert and semidesert.  The semi-
desert occurs in arid regions above the timberline.  In temperate zones, sparse
communities of low shrubs and/or cushion plants may occur.  In the arctic, the
semiarid areas are generally considered to be part of the tundra zone, but with
respect to physiognomy, plant adaptations, soil characteristics and sparseness
of animal life, they are desert-like.   Arctic-alpine deserts tend to have very
low plant cover, dominated by snow, ice or rock which may support growth of
bacteria, algae and lichens.  Vascular plants may occur in irregularly scattered
pockets having less extreme microclimates.  These deserts occur in Greenland,
Antarctica, and above timberline on high mountains all over the world.  Alpine
deserts occur in the lower alpine zone in arid mountain ranges.  At lower
latitudes, permanent ice and snow are  not necessarily present, and the presence
and nature of vegetation will vary with the amount of moisture present.
         TO
     Swan   notes that the alpine regions of the world are delimited by timber-
lines extending from latitudes 72°N to 56°S with a maximum altitudinal extension
to approximately 14,000 ft.  Figure 5-1 depicts the worldwide distribution of
timberlines and the arctic and aeolian zones.  The upper edge of the alpine
zone is usually defined as the upper limit of the range of vascular plants.
Here too occurs an upward transition from autotrophic to heterotrophic life
forms.  The characteristics of this boundary identify three basic types of alpine
regions:  high-latitude, low-latitude  and a depressed equatorial alpine region.
In the low latitude alpine environment, vascular plants are found above 20,000
ft. (6,100 M) and a long growing season is typical.
     Above the alpine zone lies the aeolian zone which Swan   has divided into
terrestrial, nival and aquatic phases.  In the aeolian zone, autotrophic algae,
lichens, mosses and a variety of heterotrophic animals and plants occur, but
vascular plants are entirely absent except in isolated instances where micro-
habitats exist which provide more favorable circumstances.  The aeolian zone
extends in altitude to include the summit area of Mount Everest where bacteria,
fungi and yeasts are found.  Figure 5-2 illustrates alpine-aeolian zonation in
the eastern Himalaya.
                                      5-2

-------
O
en
QJ
 03 CO
   CO
-o  c
 C  03
 03  2
   00
 0)
 c  E
•r-  O
 a. s-
  i
LO
                                     CO

                                     LD

-------
en
 i
                                     UD
                                     C.
                                     -5
                                     rt>

                                     (Jl
                                      I
                                     ro
                                  3 -a
 a»  a>
^c   i
 CU  CU
    O)
x— O
 -h — '
 -$  -"•
 o  &
 3  3

 CO N
 s:  o
 OJ  13
 13  DJ
  con-
  CO-'-
— o
 •   3
                                     3

                                     r+
                                     3"
                                     CD

                                     o>

                                     00

                                     fD
                                     -5

-------
      Most  of the  work  on  high-mountain  ecology has  been  done  in  Europe  and  in
 North America;  a  considerable  amount  of work  has  also  been  done  in  the  Soviet
             •70
 Union.   Swan   notes that unifying  concepts need  to be developed in order to
 link the ecological characteristics of  the widely spread treeless mountain  zones
 throughout the  world that are  commonly  lumped under the  category of "alpine."
 For  example,  in the eastern  Himalaya, the life found above  the level  of the
 highest  vascular  plants forms  a  distinct category.

 5.1.2  The Aeolian Life  Zone
      In  the  aeolian zone,  nutrients are supplied  by the  atmosphere  (carried in
                38
 by wind).  Swan   has  distinguished between the aeolian  and alpine  regions  by
 contrasting  the sources of nutrition:   alpine biota are  dependent upon  local
 autotrophic  flowering  plants,  whereas aeolian organisms  are basically supplied
 by air-transported nutrients that are rarely  of local  origin.  Hutchinson19 con-
 siders the aeolian region  to be  roughly equivalent  to  the hyperallobiosphere
 or that  portion of the eubiosphere  at high altitudes where heterotrophic
 organisms  are dominant.
      The aeolian and alpine regions may be defined  geographically as  well as
 nutritionally and  include  some portions of the high  alpine zone  and the nival
                                        23      38
 zone  described  by European researchers.     Swan   believes that  the aeolian
 region is  most  easily recognized where  established  or  semipermanent organisms
 exist in the absence of flowering plants.  Vascular plants may be absent at
 lower altitudes in particular  localities or seasons, and when the ground is
 covered with snow and organisms dependent upon snow and wide-blown nutrients,
 the distinction may become blurred.   Swan   suggests that the picture may be
 simplified if the aeolian region is arbitrarily defined as being near or above
 the upper  limits of flowering plants.
     Nutritional patterns in the aeolian zone are still not completely under-
 stood.  Animals are sometimes found at altitudes far above the highest vascular
plants; in such instances, wind-blown nutrition is ascribed.  Also,  insects  and
crustaceans occur in temporary glacial pools,  and it may be inferred that those
organic nutrients present are probably of atmospheric origin,  falling on old
snow fields and eventually being released from melting  ice.   Arthropods  and
birds that  feed upon wind-blown insects  are  obviously aeolian, but  they  may  be
                                       5-5

-------
 transient  visitors  to  the  zone.   In  many,  perhaps  most,  instances  the  type  and
 quality of nutrients involved  in  the support  of  aeolian  life  and the degree
 to  which animals  and plants  are dependent  upon atmospheric  nutrition are  un-
 known.   Gregory  has  reviewed various  organic particles carried by the atmo-
 sphere  such as  pollen,  bacteria,  fungi  and insects.   Luty and Hoshaw22 have
 examined airborne dissemination of green algae.  Swan58  has reviewed the  ecology
 of  the  three phases of  the aeolian zone  (nival,  aquatic  and terrestrial); a
 brief adaptation  thereof follows.

     5.1.2.1  The Nival Phase  of  the Aeolian  Zone--
     Descending air over snow  fields may have the  effect  of concentrating
 particles  suspended in  the atmosphere; hence, analysis of snow fields  may pro-
 vide an  exaggerated impression of the distribution of airborne debris.   Wind-
 blown particles tend to blow on and  off of most  surfaces, but remain upon snow
      30
 Osburn   studied  the phenomenon of the accumulation of radioactive fallout
 contamination of  snow;   he found that the particles concentrate in a snow patch
 as  it melts, presumably as a function of adsorption of the particles to surfaces
 within the  snow.  Interestingly,  radioactive debris was held to a large extent
 within a shrinking snow patch and was released in reduced quantities in the melt
 water.  He  also noted that radioactive particles tend to concentrate in soil
 depressions beneath melting snow.   These observations may be of potential  signifi
 cance in the design of alternative cleanup procedures for implementation in the
 high mountain environment.
     Autotrophic green  and bluegreen algae may be considered to be of an aeolian
 nature in the high mountain environment in that nitrogen and other nutrients must
 be derived  from the atmosphere.  Bettler  has suggested the existence of a  food
 chain consisting of pollen > bacteria > algae > annelids for the green  snow worm
Mesenchytraeus solieugus that occurs  on a British Columbian  glacier.  These
aeolian annelids are found in the  mountains of the  northwestern United  States,
Western Canada and Alaska.
     There  has been  some work suggesting the possibility of  organic precipitates
and proteinoid materials of oceanic  origin  being  deposited upon high mountains,
but Swan   believes  that further  study is necessary on this  topic.   Accumulations
of dead insects are  occasionally  found at high elevations, and these are
scavenged by certain species  of birds.   As  previously mentioned,  Osburn50  found
                                     5-6

-------
 that  radioactive  particles  concentrate  in  soil depressions beneath melting
 snow;  it  is  also  possible that  this  is  true  for organic materials.

      5.1.2.2  The Aquatic Phase of the  Aeolian Zone—
      The  organisms  that  inhabit the  pools  and torrents of the aquatic phase
 of  the  aeolian  zone are  considered to be aeolian by virtue of the fact that
 the organic  materials upon  which they subsist are assumed to be originally of
 atmospheric  derivation.  Torrential  fauna  include turbellarians, crustaceans,
 hydracarinids,  and  various  insects that inhabit areas primarily or solely near
 the sources  of  torrents  or  those whose  primary nutritional source is closely
 linked  with  organic materials released  from  snow or ice.  Large populations of
 fairy shrimp  (Branchinecia  sp.) are  found  in temporary glacial pools at altitudes
 up  to 19,000  ft.  in the  eastern Himalaya.  Examination of their intestinal
 contents  has  revealed mostly fine granular glacial dust mixed with traces of an
 unidentifiable  organic material; algae  have not been found.  That algae are found
 in  more permanent ponds  suggests the existence of nutrients derived from auto-
 trophic plants, but if,  in  fact, the primary algae are green algae, the presence
 of  nitrogenous  material  derived from the organic debris of snow may be inferred.
    •70
 Swan    believes that many lakes classified as alpine are, in fact, aeolian.

     5.1.2.3  The Terrestrial Phase of  the Aeolian Zone--
     The  organic and other particulate matter that is deposited upon high
 altitude  snow and ice surfaces also deposits and collects upon snowless areas,,
 the particles being trapped within cracks and crevices where air turbulence is
 reduced and the particulate contents of the air can fall out.  Such accumula-
 tions of  the terrestrial phase are frequently obvious and largely account for
 collected dust under rocks on mountainsides.   Rock slopes free of snow harbor
more scattered and thicker accumulations of debris than snow-covered slopes;
 such concentrations serve to sustain aeolian organisms.   It is likely that such
 sites will also trap spilled materials that are not  blown off rocky mountain
top surfaces.
     Terrestrial aeolian animals are frequently found at lower altitudes on rock-
covered glaciers.   Various spiders,  mites,  and insects have been found in this
environment in the Himalaya.  Mites,  collembolans  and flies live in the absence
of plants on the snow free slopes  on the sides of  glaciers;  their nutrition is
aeolian in nature.  Salticid spiders (Europhrys sp.)  have been found  at 22,000
                                     5-7

-------
 ft.  (6700 m)  in the Himalaya and are presumed to feed upon collembola, anthomyid
 flies and possibly wind-blown insects.  An example of this aeolian food  source
 may  be seen on the peak of Citlaltepetl (Pico de Orizaba) in Mexico where July
 winds are from the east with a sweeping uplift of air from the tropical  valleys
 of Vera Cruz  onto the mountain peak; many insects are transported to the peak
 by these winds.  The lizard Sceloporus microlepidotus Wiegmann is relatively
 abundant on rocky slopes above timberline   and is found at 15,800 ft. of
 elevation (700 ft. above the highest flowering plants).  The lizard colonies
 are  visited by ravens (Corvus corax sinuatus Wagler) and sparrow hawks (Falco
 sparverius sparverius Linn.).  Salamanders (Pseudoeurycea gadovii) were  collected
                                                                         TO
 at 15,000 ft. and rattlesnakes (Crotalus triseriatus) at 14,850 ft.  Swan
 suggests that all of these organisms exist, at least partially upon the  aeolian
 food chain.   Lizards have also been found at extreme elevations in Peru  and
     i 38
 Nepal.
     The upper limit of the terrestrial phase of the aeolian region is uncertain.
     •70
 Swan   notes that sun/shade temperature differentials of 43°C (78°F) are found be-
 tween shade and sun near the ground at altitudes of 15,500 ft.  He suggests that
 it is at least plausible for such differentials to exist at higher altitudes,
 perhaps even near the summit of Mount Everest, in micro habitats sheltered from
 high wind velocities.  Assuming a lapse rate of approximately 3°F per thousand
 foot increase in elevation, minimum shade temperatures near the summit of
Mount Everest would approximate -9°F (-23°C), and maximum shade temperatures
would approach 20°F (-7°C).  Based upon data from free-air temperatures  and
 comparative temperatures between snowless valleys and snow-covered slopes,
     38
 Swan   suggests that actual shade temperatures near the summit of Mount  Everest
are  somewhat colder,  but even if shade temperatures approximate -30°F (-33°C),
 it is conceivable that on calm days, surface sun temperatures may approach
 50°F (10°C).   The fact that icicles exist  where snow overlies dark vertical
rocks at extreme altitudes  indicates the occurrence of above freezing tempera-
tures.   Figure 5-3 indicates the sun/shade temperature ranges occurring  at
 15,500 ft.  of elevation for soil and air in a valley in eastern Nepal.
         38
     Swan   lists a number of bacteria, fungi and yeasts that were collected
by the 1963 American Mount  Everest expedition.  He suggests that periodic equable
temperatures and water are  available at high altitudes;  the fact that the
organisms obtained on Mount Everest represent soil microflora rather than a
                                      5-8

-------
en
i
                                    -5
                                                      10°
                                                            15
                                                                  20°    25'    30°
                                                                    CENTIGRADE
                                                                                    35"
                                                                                               45*
                                         Temperature profile of the soil and lower air:  May 23, 1954, 15,500 ft (4724 m),
                              Barun Valley, East Nepal.  Temperatures were obtained on a level surface near prostrate plants
                              (Potentilla sp., Sibbaldia sp., etc.). The soil was sandy but brown with humus. May 23 repre-
                              sented a relatively warm day.  The effect of the sun was apparent at 9 A.M. Clouds intervened
                              prior to 3 P.M.,  and full shade occurred before 5 P.M.  Wind velocity  was estimated at less than
                              10 mph. Temperatures at  1/16 inch below the surface were obtained by placing the thermometer
                              on its side just under the surface so that a minimum  thickness of soil intervened between  the
                              thermometer and direct sunlight.  The black-bulb thermometer was an ordinary thermometer
                              with the bulb painted with India ink.
                             Figure  5-3.   Diurnal  sun/shade temperature  ranges in  air  and  subsurface
                                             soil  at  15,500 feet  (from  Swan38).

-------
 miscellany of  air-borne  organisms  supports  the  thesis  that  living  organisms
 do  exist  and grow on  very  high mountains.
      Lichens are  widespread  in the  alpine zone  and  are  often  found at  higher
 altitudes than are local vascular plants, but they  are  relatively  scarce  in
 extremely arid environments  or on  surfaces  recently uncovered by permanent
 snow.   In eastern Nepal, certain vascular plants that  can obtain water from the
 soil  exist  at  higher  elevations than do the noticeable  lichens.  That  lichens
 can grow  upon  glass and  smooth rock surfaces without leaving  any trace of etch-
 ing is  suggestive of  their aeolian nutritional  status.   Lichens are known to be
 sensitive to smog and to concentrate radioactive fallout.31'38
      Although  lichens are  scarce at the high elevations  of  the aeolian mountain
 zone, they  do  constitute the most visible of the plant  life forms  in this zone;
 therefore,  it  is  of interest to consider their  ecology  in greater  detail.  Hale18
 states  that lichens occur  in virtually every pioneer terrestrial habitat  from
 antarctic and  arctic to tropical areas.  Their  ability  to adapt to xeric  en-
 vironments permits them to dominate in certain  habitats  that  are devoid of
 other plant life  forms.  Generally, lichens have a  high  light requirement and
 they  are  succeeded by bryophytes and higher plants  where they are  shaded  and other
 conditions permit  such successional growth.   However, in habitats  where condi-
 tions are relatively stable as in the Arctic and Antarctic, in deserts, and rock
 outcrops, lichen  communities can persist for centuries.  Hale notes that  various
                                                            ? i
 lichens tend to be specific for certain rock types.   Mattick   states  that lichens
 have been found as high as 6200 m (20,341 ft.).   Gams14 has reported lichens at
 7000 m  (22,965 ft.) on K-2 in Karakonm.  Kappen   states that lichens  are often
 the predominant form of vegetation in mountains of northern regions such  as
 Scandanavia, but many species demonstrate reduced vitality above 5200 m (17,000
 ft.).
          •70
     Swan   ascribes the scarcity of lichens at high elevations to lack of surface
 water from snowmelt, or possibly to snowblast.   Below 5750 m  (18,864 ft.), sub-
 surface liquid water from snowmelt permits the development of small "oases" of
                             o
vascular plants in the scree.   Here (in what  is,  strictly speaking in view of
our working definition of the aeolian zone,  an upward extension of tundra),
 lichens are found entangled with small  herbaceous  plants.  The relative aridity
of the area confines the vascular plants to  portions of scree slopes wit\  under-
                                                 o
ground meltwater sources, or to  rock-base niches.    Lichens, lacking a vascular
                                      5-10

-------
 system and roots,  are  unable  to  utilize  subsurface  water,  and  therefore  are  limited
 by the degree  of aridity  of the  particular  mountain environment.20
          •70
      Swan   has  noted  that studies  of  barren  areas  of  the  world, notably those of
 the antarctic, give  little consideration  to airborne nutrition.  Llano21  has
 discussed the  contribution of wind-blown  guano  to populations  of collembola  and
 mites found  within a few  hundred miles of the South Pole.
      In order  to provide  a basis for estimation of  time to recovery from  a clean-
 up operation,  it is  of interest  to  briefly  consider the nature of lichen  growth.
 Armstrong has reviewed studies  on  the growth rates of lichens.  Lichen  colonies
 do not  have  a  steady growth rate; initial growth is slow, until a diameter of
 approximately  1-1.5  cm is reached at which  time the rate increases and becomes
 nearly  linear until  a  diameter of 12 cm is  attained and then the rate slows, be-
 coming  non-linear.   Studies under more ideal growth conditions than those obtain-
 ing  in  the aeolian mountain peak environment indicate thallus  (plant body) dia-
 meter growth rates (in the rapid linear growth  phase) ranging from 1.5 to 3.0 mm
 per  year  for the lichen Parmelia grabratula.  A study of maximum thallus dia-
 meter versus thallus age  indicates that nearly  25 years are required for the
 lichen  to attain a diameter of one centimeter,  then  (at an increased rate of
 growth) a diameter of  7 cm may be reached after fifty years.   Armstrong2 suggests
 that there is considerable variation in the growth  rate on an individual plant
 and  that  the rate  is quite sensitive to environmental factors.   Hale18 notes
 that while lichens with diameters of 30-40 cm have  been reported in the Arctic
 and  are reputed  to be  several  thousand years old,  only direct and continuous
 observation can  establish the  true age of a lichen thallus.  Hale lists growth
 rates for several species; one is reported to have an annual  radial  growth
 rate of 25 mm and another ranges from 11  to 90 mm.   Growth rates for a given
 species increase  with the temperature of  the local  climate.  Hale believes that
 the average life  span of a given lichen thallus may range from  20 to 40 years,
 depending upon species  and conditions.   In the foliose  lichen Parmelia con-
 spersa, growth slows  after a  thallus diameter of 9-10 cm is reached,  following
which the center  decays and is overgrown  by crustose lichens.   These crustose
 lichens also overgrow freshly  exposed rock surfaces  where the centers  of the
thalli have fallen away.   Hale is of the  opinion that such  "recycling" of a
 lichen community  can  continue  until  the micro-environment  is  so altered  that
                                      5-11

-------
 other plant forms succeed the lichens.   Since the last glaciation of North
 America and Europe some 10,000 years ago,  such recycling of lichen communities
 may have occurred hundreds or thousands of times.
      There are several methods of measuring and expressing lichen growth rates,
 none of which are entirely satisfactory.   Armstrong2 suggests that when the
 problems of such measurements are solved,  the science of lichen ecology will  be
 able to progress rapidly.   If the diameter and radial growth rate can be accurately
 determined, it becomes possible to estimate the age  of a colony,  but,  according
 to Armstrong,  there  are currently several  problems with such determinations.
 This "state of the art" probably partially explains  the absence of growth data
 on lichens in  the aeolian  mountain zone.
      Bailey  states  that there have  been few studies in which the germination  and
 development of lichen  propagules (bodies containing  the algal  and/or  fungal
 spores  which are disseminated by wind,  water,  animals,  etc.),  on  substrates
 conducive  to lichen  growth.   Lichens are structures  composed of both  fungi  and
 algae in a symbiotic relationship.   They produce  propagules  containing  either
 or both symbiotics,  but only  the latter type  can  reproduce the  lichen  structure.
 Propagules are dispersed by wind and Bailey cites studies  indicating  that the
 propagule  content  of the wind is proportional  to  the  regional  abundance  of  lichens.
     9
 Bond found  that  glacial ice  in  Colorado contains arctic-alpine lichen  spores
 that are presumably  aeolian in derivation.  Once  a spore has been  deposited on
 a  suitable substrate,  it must  germinate and then  attach  itself  to  the substrate
 by the  formation  of  a  "holdfast."  Thallus  fragments  of preexisting plants can
 also  become  established on a  new site by the  formation of a  holdfast.  Different
 species  of lichens vary in the types of spores that are formed.
      Bailey  notes that little is known of the nature of colonization of newly
 exposed  surfaces, partially due  to the relative scarcity of  such areas.  It was
 found that the  island of Krakatau was colonized some  3 years after the volcanic
 eruption,  which removed all plant cover, by blue-green algae; lichens followed
 at a  later stage.  Studies on Surtsey, a volcanic island which appeared off the
 southwest  coast of Iceland in 1963,  indicated the presence of three species of
 lichens by 1970 and 11  species by 1973.   Four of the species were widespread
and evenly distributed  over the  island.   Aeolian transport is ascribed to these
botanic colonists.  The lichens first became established in and around holes in
the lava,  suggesting  that surface texture  was important in providing a "foothold"
for the holdfasts of  the plants.
                                      5-12

-------
      Based upon  the  foregoing discussions,  it would  appear  that  lichens  are
 able  to  colonize a new  area within a  10 year period  of  the  first  exposure of
 the substrate  surface,  but development of  the plants is  a slow process,  requir-
 ing 50 years to  attain  a diameter of  7 cm  and perhaps 80 to 100 years  to "mature"
 to a  diameter  of 12  cm.  These estimates are based upon  the results of studies
 in more  temperate environments than that which obtains  in the aeolian  mountain
 peak  zone; here,  conditions are more  severe, and the time to recovery  may be
 markedly increased.
      The time  required  for recolonization  of a treated area will  depend  upon the
 extent and nature of the treatment, and the abundance of lichens  surrounding the
 treatment area.   If  viable fragments  of lichen thalli survive treatment  which
 includes the removal of the lichens,  the rate of recolonization may be relatively
 rapid and independent of the size of  the treatment area.  If no viable thalli
 fragments or propagules survive a treatment, colonization will have to come from
 outside, and will probably be slower  for a  large treatment  area than for a
 small area.  It  should  be emphasized  that  this brief review only  skimmed the
 surface of lichen biology and ecology, and  that many important aspects of their
 developmental  patterns  were not discussed.  Further, no  data were found  for
 environmental  conditions comparable to the  aeolian mountain zone.
      Swan suggests that the reestablishment of life  forms in areas uncovered
 by permanent snow may require considerable  time and  that aeolian  forms are
 successional to  alpine  flora and fauna in environments capable of supporting
 alpine life forms.

 5.2   NATURAL PERTURBATIONS
      Only aeolian life, as discussed  earlier in this chapter, will be  considered
here.  There is  little precedence for intentional perturbation of high mountain
peaks and relatively little literature concerning either natural or intentional
perturbations.  Also, in the discussions of other life zones, steep slopes per
 se were considered only incidentally.
      The lack  of vegetation and the steep slopes characteristic of high mountain
peaks make them particularly vulnerable to the erosive forces of earthquakes,
wind, water, ice and associated debris.  Meltwaters  and rainfall wash  soil away,
 leaving rills  behind.  The steepness and shape of a  slope strongly influence
 the velocity and destructive force of water runoff.  High mountains tend to have
                                       5-13

-------
 steep slopes with an average gradient of 25° or more.1  The gradient is a func-
 tion of various factors,  including the type of rock present, the age of the
 mountains,  their tectonics,  climate,  etc.   Soil destruction also is a function
 of the shape of the slope,  the relative positions of sections with different
 gradients  and the nature  of the soil  formation; for example, podzolic soils,
 formerly under forest,  are  particularly vulnerable to erosion.1   Animal grazing
 activities  also serve to  erode mountain slopes as well as  enhance conditions
 for water  and wind erosion.
      The hydrographic networks of  mountainous  regions are  complicated systems
 composed of rivers and  streams of  various  sizes,  gullies and ravine type
 gorges,  all of which usually terminate  in  larger  mountain  rivers.   Al'benskii
 and Nikitin  have discussed  these  systems  with respect to  various  aspects  of
 mountain system erosion.  They note that the conditions of headwaters of rivers
 serve  to determine the  role  of rainfall  in feeding  the rivers.   In high mountains,
 the role of rainfall  is negligible; snow is the major source of  runoff.  Ground-
 water  provides  approximately one third  to  one  half  of the  water  of mountain
 rivers.  Rivers  originating  from glaciers  and  snowfields have fairly  regular
 rates  of flow,  whereas  those  originating below the  snowline  frequently  run dry
 in  the summer.   Where forestation  is  light  or  absent,  the  rate of  runoff in-
 creases  rapidly  and  the rate  of decrease is only  slightly  less;  heavily  forested
 areas usually do  not have sharp runoff peaks.1

 5.2.1  Erosion and Mudflow Phenomena
     Al'benskii  and Nikitin   observe  that  the  greatest force of  linear  erosion
 in  mountains is  exerted by small rivers and streams,  and temporary  torrents,
 particularly in periods subject to strong  showers and/or intensive  snowmelt.
 Such torrents are  known as mudflows (or mud-stone flows) and are loaded  with
 silt, sand, debris, stones and other  solid material.  The  character of the
 flow is  influenced by the degree of solid material  loading;  if there  is  little
 solid material present  (a fluid or turbulent mudflow), the flow resembles a
 freshet, but if the water is heavily  loaded (a cohesive or structural mudflow),
 the movement of the flow may be greatly slowed.
     Cohesive mudflows have a high specific gravity and tremendous destructive
force derived from flow pressure and shocks from stones carried by the flow.
Such flows  tend to congeal,  forming a wave-like elevation when their movement
stops.   Fluid flows tend to be contained within the action section of the

                                      5-14

-------
stream bed, depositing and, to some extent, sorting their solid contents,  Mud-
flows may also be classified according to their origin, e.g., fed by slope
erosion and weathering products, fed from linear sources, etc.  In arid mountains,
the drainage areas of mudflows may contain great reserves of solid material and
strong showers may produce highly destructive mudflows.
     Mudflows may reach volumes exceeding those of freshets by a factor of 30
or more.  Table 5-1 provides data from mudflows in the Soviet Union.   Mudflows
have been known to carry rocks of 100 cubic meter volume for considerable dis-
tances.  Mudflows are characterized by episodic activity, being common in some
regions and irregular in others.  Mudflow velocities vary from 3-7 m per second
and tend to proceed in waves of varying height and frequency.  Waves are caused
by obstructions in the path of flow.  Erosion by mudflow is a widespread phe-
                                                                   1       32
nomenon on mountains, but flows are rare in densely forested areas.   Owens
has studied the morphology of alpine mudflows.
     Should an uncontained or unstabilized spill be washed from the primary site,
the inclusion of radioactive material in a mudflow is a possibility.

5.2.2  Snow Avalanches
     Al'benskii and Nikitin  note that snow avalanches are most frequent in
high mountains, but they do occur occasionally on medium and low ranges.  Ava-
lanches tend to form where large masses of snow accumulate on steep slopes or
along ravines.  Three types of avalanches may be distinguished on the basis of
their formation:  surface, bottom and jumping.  Surface avalanches are actually
snowslides coming off an entire surface of a steep slope; they are common on
smooth grassy slopes and are more frequent on slopes facing the sun.  Bottom or
trough avalanches form in ravines and hollows and slide in the shape of a talus
fan.  Jumping avalanches differ in that they form in ravines or other surfaces
terminating in high escarpments from which the sliding snow "jumps."  Avalanches
are capable of producing severe destruction, averaging 50-60,000 tons in weight
and slides of 1.5 to 2 million cubic meters have been observed.  Avalanches
tend to flow from slopes of 30° gradient or more, although some have been ob-
served on slopes of 15-18°.  Generally, avalanches do not form in densely
                          34
forested stands.  Schaerer   discusses the effects of terrain and vegetation
upon avalanche sites.  An excellent reference to the physics, ecololy and
                                                                       33
management research of avalanches may be found in Perla and Martinelli.
                                       5-15

-------
                            Table 5-1.   Movement of mountain detritus within measured mudflows
                                        as examples of erosion (from Al'benskii and Nikitin1).
Name of Stream
Malaya Almaatinka (Kazakh SSR, 1921)
Kish-chai (Azerbaidzhan SSR, 1936)
Gedar (Armenian SSR, 1946)
Approximate
drainage area
of the mud-
flow, km?
60
25
30
Solid
material
transported
1000 m3
3,000
2,175
500
Material
removed ,
average in m^
Per 1 km2
50,000
87,000
17,000
Content
of detritus
by weight,
%
37.0
68.2
49.0
en

i—»
cr>

-------
 5.2.3   Landslides
                29
     O'Loughlin   has  studied  factors  affecting  the  incidence  of  landslides.
 He  found  that  a combination of steep slopes,  heavy seasonal  rainfall  and  an
 impermeable  till substratum predisposes  a  site to catastrophic failure.   The
 occurrence of  springs  at  the heads of  landslide  scars  suggests the  importance
 of  groundwater in  the  release  of materials  for slides.
     Dingwall    has found that  erosive overland  flow of debris and  fine sedi-
 ments from bare debris surfaces is greatest during summer months.   This process
 is  limited:  to those months when the debris is snowfree and unfrozen; to  those
 surfaces  unprotected by vegetation and coarse debris; and to finer  fractions of
 the debris.

 5.2.4   Overgrazing
     Overgrazing is a cause of  soil destruction on mountain slopes; while there
 is no forage in the aeolian zone, the topic is included here as a perturbation
 affecting high mountain slopes.  Animal grazing trails criscross in increasing
 density with time,  resulting in vegetation removal and soil destruction.
 Continued use  of the trails inhibits regrowth of vegetation.  Where the trails
 merge,  sheet erosion can reach proportions of as much as 100 cubic meters of
 soil per hectare per year.

 5.2,5   Recovery
     Recovery,  either unassisted or assisted,  from perturbations in the mountain-
 top environment does not have the same  meaning as in  the sense of the discussions
 covering other  life zones.  In cases where perturbations such as rock, mud,
 snow,  or ice slides erode barren mountain peaks or slopes,  devoid of higher
forms  of vegetation,  the term "recovery"  seems to have little meaning.  In lower
 life zones,  recovery refers  to restoration of the original  life forms such as
forests or grasslands,  but in the aeolian environment where the only life  forms
are microorganisms, lichens,  insects, and a few incidental  animals,  there  is
 little to restore.   Erosion,  either violent or gradual,  must be considered as
a natural  process of wearing  down mountain ranges.  At  lower elevations  where
 life zones other than aeolian are present,  the term recovery regains its signifi-
cance  and the reader is referred to the appropriate chapters of this series.
                                     5-17

-------
      However, while recovery is not an applicable terra, prevention or reduction
 of the  severity of certain perturbations of mountain environments may be both
 possible and desirable.  With respect to transuranic spills and cleanup, certain
 preventative measures may aid to reduce spread of the spilled material.  This
 concept will be discussed at greater length in a later section.  It is of in-
 terest here, however, to note that some success has been encountered in the re-
 duction of the incidence and/or severity of avalanches, erosion and mud flows
 through afforestation.  Afforestation is certainly not a short-term procedure,
 but it may have some applicability with respect to treatment of transuranic
 spills.  Discussions of afforestation at lower elevations in the aforementioned
                                                  1               24          4
 context may be found in:  Al'benskii and Nikitin;  Margaropoulos;   Aulitzky;
 Martinelli;   and De Quervian and Der Gand.
      Al'benskii and Nikitin  discuss the topic of afforestation as a possible
                                                     4
 means of reducing mudflows and avalanches.  Aulitzky  notes that in many cases,
 timberlines lie below altitudes where tree growth is climatically possible and
 he discusses the possibility of afforestation above present timberlines.  Mar-
         js
 tinelli   lists possible methods of snowpack and avalanche control;  weather
 modification by cloud seeding; intentional avalanching for the purpose of stor-
 ing snow in shaded valleys at high elevations; reshaping of natural terrain to
 improve show trapping efficiency and capacity; control of snowmelt by addition
 of materials to the snow surface; and the erection of snow fences and other
                                                                    12
 barriers to increase snow accumulation in desired areas.  Dortignac   discusses
 the possibility of increasing the melt rate of glaciers and packed snow by
 scattering dark materials over the surface.

 5.3  MAN-MADE PERTURBATIONS
      No precedents were found in the literature for intentional perturbations
 of mountain peaks in the aeolian life zone; however, perturbations of mountain
 peaks and slopes at lower elevations are common in the form of plowing and re-
 contouring for strip mining and agricultural purposes; these perturbations may
 occur in desert mountains which have certain similarities to the aeolian zone
 with respect to the absence of vegetation.

5.3.1   Regrading  and  Replanting
                      28
    May and  Striffler    discuss  regrading  and  revegetation  of mountain watersheds
                                        5-18

-------
 sustaining severe mechanical and hydrological disturbances due to strip mining
 operations.   Al'benskii and Nikitin  discuss increased erosion due to agricul-
 tural efforts on mountain slopes,  and methods of afforestation of these damaged
 slopes by terracing and replanting.

 5.3.2  Alternative Techniques
      The aeolian mountaintop environment  presents unique logistical  problems  for
 cleanup that  are not present in  other terrestrial environments.   As  discussed
 earlier, the  potential  for immediate  and  possibly widespread  dissemination  of
 spilled material by wind and water is great.   Possibly,  the first problem to  be
 addressed will  be that  of containment rather than cleanup.  As noted in the fore-
 going section,  most of  the treatments considered for  other  terrestrial  environ-
 ments will not  be applicable to  mountain  peak spills.   They might, however,
 become applicable if an uncontained spill  spreads;  the contamination of streams,
 lakes,  and mudflows presents a definite hazard.   The  derivation  of possible
 substitute treatments is beyond  the scope  of this report, but  certain observa-
 tions are pertinent.
      It  is important  to note that  wind is  an  important  component  of  the high
 mountain macroenvironment  in the aeolian  zone.   Life  forms  that  are  adapted to
 an  existence  based  upon the  providence of  the  windfall most likely will  be more
 abundant  in sheltered microhabitats where  windborne debris  can settle out of
 suspension.   This precipitation process will  include  atmospheric  contaminants.
 Osburn    has  noted  that  lichens collected  in  areas adjacent to the Greenland
 "Broken  Arrow"  incident  contained  higher than  background contamination.
      Various  aspects of the  physics and management of  snow  and ice are  discussed
 in  Martinelli,    Dyunin, 3,  Gary,15 Shiotani and  Arai,35 Tabler,39'40 Tabler and
         41                          33
 Schmidt,   and  Perla and Martinelli.    An important  factor in the treatment of
 a spill  in the  aeolian  zone  will be whether the  spill occurs on a  snow-covered
 surface,  soil surface,  or rock surface.  The observation of Osburn30 that snow
 adsorbs and concentrates particulate matter can be of extreme importance for the
 design of effective containment and cleanup procedures.  The combined actions of
 water and gravity, which present the  immediate danger of spread of the  spill to
 lower levels on the mountain, might also make possible the utilization of the
properties of an affected watershed to concentrate and contain a spill.
                                       5-19

-------
      Similarly,  the  effects of wind barriers,  serving to concentrate organic
 and  particulate  matter  in the aeolian  environment, might also provide additional
 possibilities  for treatment.  Devised  techniques embodying these and similar
 factors might  also be environmentally  compatible and offer comparatively short
 recovery times.
 Gold.
     The possibility of controlling snow surface temperatures by dusting with
     01 s
     44
                                             3              17
various materials has been studied by Arnold,  Grove et al.,    and Williams and
      It  is  likely that there will be a degree of similarity between the responses
                                                                   37
of the aeolian and alpine environments to cleanup procedures.  Swan   primarily
distinguishes between the biological components of the two zones on the basis
of nutrition; with respect to the physical phase of the environment, there probably
will be a number of similarities in the response of terrain in the two zones to
cleanup operations.  It should be realized that mountains, particularly the
peaks, present a dynamically unstable condition rather than the dynamic equilibria
presented by biological systems.  Geologically speaking, all mountains are doomed
to destruction by erosion; therefore, "recovery" of the physical component of the
mountain ecosystem does not have the same meaning that it does in more horizontal
situations.  At best, "recovery" will imply a return to the pre-treatment dynamic
state, or rate of erosion.

5.4  EFFECTS OF CLEANUP PROCEDURES ON AEOLIAN MOUNTAIN PEAKS

5.4.1  Proposed Treatment Techniques
     As discussed previously in the section on recovery from natural perturba-
tions, the high mountain peak environment presents a totally different situation
from those discussed in reports on other life zones.  First, the very nature of
what is meant by the term "recovery" is open to question.  There are few life
froms present to be significantly affected by possible cleanup procedures..
There are no flowering plants in the aeolian life zone; life forms present on
the high peaks are limited to bacteria, fungi, algae, lichens, insects, nema-
todes, and a few species of birds and reptiles.  Second, the nature of the moun-
tain peak terrain, combined with the lack of flowering plants will make questions
concerning recovery from most of the treatments considered in other reports
moot:  there is virtually no plant life to "recover" and such treatmentfs as

                                      5-20

-------
vegetation removal,  surface alteration,  soil removal, plowing,  soil  fill,
sludge application and flooding are either not applicable or, if  logistically
possible, impractical.   It appears likely that alternative and  innovative
treatment procedures will be necessary to treat primary spill sites, while
conventional treatments  may be necessary at points below the spill if it is not
contained.
     While it is true that there is little variety or abundance of life in the
aeolian environment, the ecosystem that exists may be considered  quite fragile.
As previously discussed, it has been observed that lichens are  slow  to become
established on high mountain surfaces that have been uncovered  from  permanent
snow.  Those life forms  that have become established in the aeolian  system are
dependent upon accumulations of nutrients that are ultimately derived from the
winds; it is likely that the time required for such accumulations to form is
rather long.  There are very little data on this subject.
     An important consideration in the design of cleanup techniques  for the
mountain peak environment is the accessibility of this zone to men and equip-
ment.  The operation of helicopters may not be feasible at higher elevations;
ground access to many sites by heavy equipment may be impossible, and the severity
of mountain physiognomy may preclude its ure.  Equipment used in cleanup opera-
tions may be restricted to that which can be packed in or parachuted.  Air removal
of contaminated materials may be the only feasible method.  For these reasons,
methods based upon (or partially utilizing) the management and manipulation of
natural features and phenomena (see discussion of the Nival Phase of the Aeolian
Zone) at the site, such as snow cover and water runoff,  may be desirable from
the standpoints of logistics,  economics and ecology.   The impact of devices to
manage wind, snow drift and water runoff would be variable with their design,
construction and placement.   It is expected that their major impact would be upon
the abiotic phase of the environment (erosion,  etc.)  and that the impact of
such devices upon the life forms of the aeolian zone would be minimal.
     The following discussion  of suggested cleanup techniques will be generally
limited to the aeolian life zone.   Some of the  comments  might also have limited
application to steep mountain  slopes and peaks  in lower  life zones or in a
desert environment.
                                     5-21

-------
     (0-1)   Natural  Rehabilitation
     Little has been published on the topic of natural rehabilitation of per-
turbations to the aeolian mountain  peak environment, making the estimation of
recovery time from a spill cleanup  operation rather difficult.   "Natural re-
habilitation" would  depend largely  upon the rates of two processes:   dispersion
of spilled material  by wind and water and the rate of decay of spilled material
that becomes trapped in rock crevices, snow and ice.  These processes could be
quite rapid or very  slow, depending upon the properties of the spilled material
and climatic conditions at the site.

     (1-1)   Chemical Stabilization
     Soil stabilization by the addition of chemical stabilizers to the surface
is probably not a viable treatment  in the aeolian mountaintop environment.

     (1-2)   Clear Cutting Vegetation ("Removal")
     This technique, applicable in  many life zones, will be generally inapplicable
in the aeolian environment.  With the exception of lichens, fungi, and certain
algal species, there is no vegetation present.  The removal of lichens from rock
surfaces or algae from cliff faces, snow, or ponds probably will have little
lasting impact upon  the ecosystem,  with the exception that lichens probably
will be very slow to become reestablished.  It is unlikely that the removal of
lichens will have measurable effects upon temperature, radiation, soil moisture,
precipitation, or surface runoff.  Removal of algae from snow or ice may increase
surface albedo, thereby decreasing  surface temperatures and runoff;  this treat-
ment might possibly  affect the microclimate.  The possibility of controlling snow
surface temperatures by dusting with various materials has been studied.  Removal
of algae from cliff  surfaces can increase surface erosion.
     (1-3)   Stumping and Grubbing
     Not applicable to this land type.

     (1-4)   Scraping and Grading
     Whether surface alteration techniques, such as scraping and grading with
"air-dropped" equipment, will be logistically possible on mountain peaks in the
aeolian environment is highly questionable.  Surfaces are likely to be snow, ice,

                                      5-22

-------
 rock, and debris with little soil accumulation on steep slopes; some surfaces
 might be almost vertical.  High winds, precipitation, or dense cloud cover might
 be present.  The potential for an avalanche must be considered in some areas.
 Under these conditions,  manual labor might be the only effective resource,
 as was largely the case  in the Greenland "Broken Arrow" incident.
      Removal of surface  snow and ice can be expected to produce minimal lasting
 impact.   Surface scraping and grading of rock, soil, or debris will tend to
 accelerate erosive action, the severity of which will depend upon  the composi-
 tion and slope of the surface.   If the treatment increases the smoothness of the
 surface,  the potential for a snow avalanche will be increased.  Severe altera-
 tion of  mountain peaks will produce a strong visual or aesthetic impact.   Post-
 treatment measures might  be required to reduce the  potential for erosion and
 avalanches.   Methods  of afforestation and terracing might  be considered for these
 purposes.   Depending  upon the location of the spill,  climatic conditions might
 permit some  degree of afforestation or vegetation above  the  timberline.   Without
 such measures,  recovery,  or return  to the pre-treatment  conditions,  could be
 quite slow.   In fact,  the net effect of the treatment  might  be to  potentiate the
 pre-treatment  erosive  condition.  Erosive processes  could  be accelerated  and
 recovery  might  never  occur.

      (1-5)  Shallow Plowing
      Not  applicable to this  land type.

      (1-6)  Deep Plowing
      Not applicable to this  land type.

      (1-7)  Soil Cover Less than 25 om
     Application of soil  as a coverup type of treatment would not seem to be
applicable in the mountain environment because of steep slopes and rapid erosion.
Construction of mechanical retaining structures probably would be required.  This
treatment might be possible behind glacial moraines  and in small valleys but
contamination of ground water and subsequent flow to streams would  appear to
render such treatments ineffective.   Because of the  absence of plant cover growth
in the aeolian environment, stabilization of such soil fill would be quite diffi-
cult.
                                       5-23

-------
      (1-8)  Soil Cover 25 to 100 cm
      Not  applicable  to this land type.

      (2-1)  Remove Plow Layer
      Removal of soil  (if present) will increase exfoliation and erosion of the
underlying rock surfaces by making them accessible to temperature extremes and
alternation between  freezing and thawing.  As discussed in Section 5.1.2.3, the
daytime sun and shade temperature differential on a mountain peak can exceed
70°F.  Snow removal  would be expected to produce similar effects.  Reference to
the particulate adsorptive properties of snow was made earlier in Section 5.1.2.1

      (2-2)  Remove Shallow Root Zone
      A "root zone" in the sense used in other ecological zones does not exist in
the aeolian mountain peak environment.  The comments in 2-1 will apply to
soil  removal to depths usually thought of as the root zone.

      (.2-3)  Remove by Scraping and Grading3  Mechanically Stabilize
      Removal of soil and rock by scraping and grading, if logistically possible,
would probably serve to increase erosive processes.  Mechanical stabilization in
the sense of soil compaction would probably not be applicable to many sites in
this  zone.  Partial  stabilization by methods analogous to those used in trail
maintenance may be possible, but their long-term effectiveness is questionable.

      (2-4)  Remove Plow Layer^  Mechanically Stabilize
     The mechanisms for thermal fractures of newly exposed surface rock were
described earlier in treatment 2-1.   For this land type, the feasible mechanical
stabilization would be the temporary system of rock-weighted mesh and drilled
pins  in lead anchors or the use of "rock-bolts."  Mechanical stabilization of
soil patches by a soil compaction process would be quickly exfoliated by frost
heaving and cannot be recommended here.   Large areas would be prohibitive to
cleanup,  and "recovery" in this context  would be an indefinable term.
                                       5-24

-------
      (.2-5)  Remove Shallow Root  Zone, Mechanically Stabilize
      This combined  treatment  is  inappropriate for the aeolian mountain peak
 environment.

      (2-6)  Remove  by  Scraping and Grading3  Chemically Stabilize
      Soil removal followed by chemical  stabilization  would  mitigate  some  of the
 consequences  of  rill and  gravitational  creep erosion  caused by slope steepness
 until  the chemicals break down under  ultra  light  exposure.  This breakdown  is an
 indeterminate result and  highly  dependent upon  the cumulative duration of snow
 and  ice  cover, and particularly  dependent upon  slope  aspect.
     The removal of spilled material  by scraping  and  grading  of hard surfaces
 is subject to those points discussed  in treatments 2-1  through 2-3.   The  effec-
 tiveness and  consequences of  chemical stabilization would depend upon the nature
 of the site and  the properties of  the stabilizing material.   The use of these
 methods  in areas subject  to severe  erosive  forces is  questionable.

     (2-7)  Remove Plow Layer, Chemically Stabilize
     This treatment is inappropriate for the  same reasons as  treatment  1-4.
 Chemical  stabilization would  be  less effective than weighted  mesh upon  those
 patches  where  some soil has collected.

     (2-8)  Remove Shallow Root Zone,  Chemically Stabilize
     This treatment is not appropriate.

     (3-1)  Barriers to Exclude People
     If the supporting posts can be bolted into rock to support the  ice during
 snow melt, this might  be the most realistic  temporary procedure while cleanup
 is carried out.  Access barriers might be necessary at greater distances from
the primary spill site than for other  life zones.   Because of  the absence of
protective vegetative  cover,  rapid transport of spilled material by water and
dissemination  by wind  can  occur immediately  after the  spill; the spread of con-
tamination to  locations remote from the  primary site  is a definite possibility.
                                      5-25

-------
      Due  to  the  high  danger  of  spreading  of  spilled  material  by  wind  (in  the
 absence of protective plant  cover)  and  water to  sites  downslope  and/or  possibly
 remote from  the  primary  location,  erection of barriers some distance  from the  spill
 may be necessary.   It may  be necessary  to restrict access  to  lakes  and  streams at
 lower elevations in the  watershed.  Valleys  at the base of a  mountain peak spill
 site may  be  subject to wind  contamination.   Assessment of  the dangers of  such
 spread will  have to depend upon  local evaluation of  the site  and consideration
 of  the properties of  the spilled materials.
      (3-2)   Exclude Large  Animals
      The  erection of  animal  access  barriers  might or might not be necessary,
 depending upon the  site  location and altitude.  Barriers excluding  large  animals
 probably  would not  be necessary for spills in the aeolian  environment since it is
 far  above their  normal range.  It  is unlikely that animals will  interfere  with
 the  recovery of  a treated  area, however, they can act  as a vector for direct
 dissemination of radioactive materials.

      (3-3)  Exclude Large and Small Animals
      Small animals  in  the  lower portion of the aeolian zone are generally
 limited to a few reptiles  and insects; these are discussed in the first part of
 this  chapter.  Birds,  which  are occasional visitors to this life zone (also
 discussed  earlier in  this  report),  can be problematical.

      (4-1)  Asphalt Hard-Surface Stabilization
      Outside the scope of work of this study.

      (4-2)  ConoTete Hard-Surface Stabilization
      Outside the scope of work of this study.

      (5-0) Application of Sewage Sludge
      Outside the scope of work of this study.

      (6-1) High-Pressure Washing (<3  mm)
     Not  applicable to this land type  due to logistics and recontamination
problems.

                                     5-26

-------
      (6-2)  Flooding to SO cm
      Not  applicable to this  land  type.

      (7-0)  Soil Amendments Added
      Outside the scope of work of this study.

 5.4.2  Alternative Treatment Techniques
      Severe climate is characteristic of this  zone with high velocity winds
 and  severely eroding steep slopes at those  seasons free water occurs.  Snow
 blasting  can redistribute contaminants impacted on high vertical  surfaces
 without interference but horizontal movement can be influenced by diverting
 melt  water with structures, and snow may be managed or directed while in
 motion .
     (8-1)  Snowfenees_and Wind^
     These structures can be of temporary benefit in reducing snow accumula-
tion at one small area by holding it in another site.  The wind shadow extends
leeward 10-15 times the height of permeable barriers.  By rule of thumb, a
solid wall 0.75 meters high protects the adjacent surface from snowblast and
wind erosion to 10 meters downwind and a snowfence 1.2 meters high and of
about 50% frontal area casts a wind shadow 12 meters downwind.  The most
effect is obtained with a rough edge.  Wind velocity is least near the ground
at 3 to 5 barrier heights on the downwind side of both solid and penetrable
barriers.  Effectiveness of the barriers and snowfences seems independent
of wind speeds, at even high velocity if the barrier is vertical.   The principal-
problem with barriers is that of determining the perpendicular to the direction
from which the prevailing high winds scour a given area consistently.

     (8-2)   Watershed Control Devices Constructed
     The usual operations of shaping by clearing (1-2), scraping and grading
(1-4, 2-3,  2-6), and revegetation are not viable in this ecosystem as ways of
controlling movement of contamination in water.   Streams are torrential and
mudflows scour deeply in above-freezing weather.  The high kinetic energy
                                    5-27

-------
 requires that manual construction methods or small air-freighted equipment
 of limited capability be used for the diversion structures.  These can include
 stone riprap and diversion dikes and stone gabions across rills and smaller
 channels.   Contoured terraces,  stone walls and check dams, and rock outfalls
 may be feasible as aids in containment when the contaminated area is snow-free.
 Grade stabilizing structures such as chutes and stone drop spillways can be
 supplemented by stone levees to modify smaller stream channels and reduce
 erosive cutting into contaminated "soils" and snowbanks.   None of these appear
 to be effective controls in areas with landslides,  avalanches, and even severe
 snow blast.

      (8-3)   Snow  and loe Additives
      The application of dusts to snow and ice to  retard or enhance melting
 can be useful.   Experience with dusting  charcoal  dusts and lamp black  powders
 over snow  and ice showed  significant  increases  in melt rates.   That observation,
 when taken with the  observation that  deposits of  solid particulates are captured
 while melting occurs,  suggests  a unique  procedure for reducing the volume of
 snow to be cleaned up.   The  transportation  requirements for dusting material
 are modest enough to be handled by backpackers, small fixed-wing  aircraft,
 or  even avalanche-prevention artillery rifles.  Subsequent snow cover  cancels
 the effectiveness of the dusting, but  it  can  be renewed again  repeatedly.   On
 the other hand, the  potential of reducing melt rates  by foamed polyurethane
 insulation blankets  seems  not to have  been  tested experimentally  in snowfields
 as  a  containment  technique.  In  thoery it would be an effective method  for
 isolating small areas of contamination from heat  and  winds.

      (8-4)   Removal of Contaminated Snow and loe
      The logistics of travel in  the area seem the most likely  source of environ-
mental  damage.  Melting ice and  packed snow requires heat,  but given air  support,
 fuel  tanks can be  supplied and the contamination concentrated  by melting  the
 snow  and ice carrier.  The characteristics of the contamination would control
the selection of resin column adsorbers, or using alcohol  gels or thixotropic
mixtures for containerizing the  contamination prior to transportation.   Reflec-
tive  surface solar stills would also be effective as concentrators where sunlight
is available because of the high light intensity at high elevations.
                                     5-28

-------
5.5  RECOVERY AFTER CLEANUP
     The applicability of the term "recovery" to both the biotic and abiotic
components of the aeolian zone and mountaintop environment was discussed at
the beginning of Section 5.4 on cleanup procedures.  The following discussion
is subject to the same reservations.

5.5.1   Irreversible Changes
     Any treatment that drastically alters the shape or surface of the mountain
peaks or slopes or greatly increases the rate of erosion in the barren ecosystem
under discussion, can for all practical purposes be considered as irreversible.
Destruction of habitats or extremely slow-growing life forms as lichens might
be considered to be irreversible from the standpoint of the human lifespan.  There
is little data on the growth rate of these organisms in the aeolian environment.

5.5.2  Rates of Recovery
     There is insufficient data available from cleanup procedures to provide a
meaningful discussion of recovery rates of the aeolian mountaintop environment.
The abiotic components of the ecosystem might not recover, per se, but will
instead establish new rates of erosion which incorporate changes brought about by
the cleanup procedures.  The degree of the effect will depend largely upon the
severity of the severity of the treatment implemented.
     With respect to the biotic components of the aeolian system, it can be
inferred from observations on the establishment of lichens in areas uncovered from
permanent snow that recovery for these organisms will be extremely slow.  Algae,
if removed from cliff surfaces, or with snowpack, probably will be able to be-
come reestablished in a relatively short period of time.  If food sources for
zoological components of the aeolian environment are removed from the spill
area, it can be expected that their reestablishment (dependent upon buildup
of concentrations of organic materials of aeolian derivation)  will be slow in
some cases and rapid in others.  Lizards that depend upon windblown insects
might be able to become reestablished in a treated area within a few generations;
aeolian insects, dependent upon collected organic materials in cracks and crevices,
might require many years to become reestablished.  There is insufficient data to
speculate upon the recovery rate of snowworm populations that  might be depleted
or eliminated with the removal of contaminated snow during the cleanup process.
                                      5-29

-------
 5.5.3   Successional  Stages
     The  "recovery"  events  that  occur  within  arbitrary  "successional  stages"  will
 vary greatly depending  upon the  location,  situation,  season  of  the  spill,  and the
 treatment  applied.   The discussion  here  will  apply only to the  biotic phase of the
 environment.   It  is  emphasized that  there  is  little data or  precedence  for these
 estimates,  and  that  they are essentially educated guesses.   Deterioration  can be
 expected  for the  abiotic phase of the  system; the rate  will  depend  upon the loca-
 tion of the spill and the nature of  the  treatment.  Since lichens are the  most
 visible biotic  component of the  aeolian  mountain environment, their recovery
 will be assumed to essentially signal  the  "recovery"  of the  biotic  component.

     5.5.3.1  First  Year--
     In the first year,  regeneration of  disturbed, removed,  or  destroyed microbial
 forms,  including algae  on cliffs and snow  fields, probably will be  well underway.
 It is assumed that these organisms are reasonably well  adapted  to the environment
 and  enjoy relatively short  generation  times.  Higher  forms of animal life  that
 are  mobile  and can subsist  on windblown  insects might have begun to reinvade
 the  treated area.

     5.5.3.2  Fifth Year--
     By the  end of the  fifth year, the microbial lifeforms will be well established
 and  the higher animal life  forms probably will have recolonized the niches able to
 provide sufficient food.  If algal forms were removed from cliff faces during
 the  treatment, they also probably will be reestablished, although the thickness
 of their "mat" probably  will not be fully developed.

     5.5.3.3  Tenth Year--
     By the end of the tenth year,  there should be sufficient accumulation of
windblown organic debris  in cracks and crevices to permit survival of small
populations of those species of spiders,  mites,  etc.,  that  utilize this nutri-
tional  source.  Algal mats should be well developed on cliff  faces  (if there  is
a favorable habitat  in the cleanup area).  Vertebrate  populations (lizards)
probably will have attained equilibrium by this time.   With the exception  of  the
slow-developing lichens, the biological climax populations  of the aeolian  life
                                      5-30

-------
zone should be well on the way to recovery in the treatment zone, assuming that
the treatment did not leave residual chemicals or mechanical structures.  By the
tenth year, small lichen colonies may be reestablished over the treatment area,
if lichen growth preexisted the treatment and lichen growth was completely removed
during the treatment.
     Speculation beyond this time period seems meaningless because, in the absence
of protective vegetation sufficient to provide stable microhabitats, changes in
climatic patterns might occur that are great enough in magnitude to alter the
environment and its biota.  If lichen growth preexisted the treatment, it should
be reasonably recovered by the 100th (or perhaps by the 50th) year, subject to
the factors discussed earlier in this section.

5,6  QUANTITATIVE ASSESSMENT OF CLEANUP IMPACTS
     Environmental forces are asserted differentially in the aeolian mountain
from "normal" zones, and the effects of recovery from cleanup are evaluated
differently.   Lack of soil cover throughout and the absence of flowering
plants lends a restricted meaning to recovery.  In the following discussion,
recovery means that the rate of erosion of surface features has been slowed to
the rate of loss that existed before contamination.  The eroding surfaces
may be snow fields, snow banks, aeolian mineral drifts from rock, ice features,
rock faces, or other surfaces, including patches of frozen and unfrozen "soils."
Life forms that exist year 'round are sophisticated, such as lichens and
snow worms or primitive bacteria and algae.  Transient populations of insects
attract scavenger birds temporarily into the aeolian peaks.
     Wind-drifted mineral deposits collected around rocks and ice features
and on and in snowbanks form the sites for "soil" accumulation.  Soils and
vegetation in the normal sense are absent.   In their place are patches of
rock-weathering products combined with wind-drifted solids to form mineral
pockets of a few square meters in area.   Continuous areas of soil patch as
                2                                                              2
large as 0.01 km  are rare, and for discussion of larger areas to 0.1 and 10  km
of cleanup, the effects must be imagined as imposed upon an increasing number of
patches scattered around one or several  mountain peaks.   Consequently, the
impacts from the effects of cleanup are those registered upon small mineral
patches whether it is one patch or large numbers of "soil" patches among
bare rock and snow and ice fields.
                                     5-31

-------
      The majority of the cleanup treatments evaluated in the report are in-
 appropriate in this zone, especially those which involve the redistribution
 of soils (1-3, 1-6, 1-7, and especially 1-8).
      Use of stabilizers is judged ineffective without there being precedent to
 judge from.   This area is considered significantly different from spoil banks
 at lower elevations.   At those sites particle size ranges exist,  and the surface
 dries off after precipitation.  In the  mountain,  piles of aeolian debris are
 intermittantly soaked or ice sheathed where chemical  stabilizers  are unlikely
 to stick, and it is not likely that sublimation is an effective drying agent.
 Single-grain coarser material  is filled with rock fragments or  covered with
 rock  too large to hold in place with sprayed solution.   No  effective use can
 be projected for treatments  1-1,  and 2-3  through  2-8.   Thermal  problems with
 chemical stabilizers  are extreme at these  heights.
      Five groups  of time to  recover are shown in  Table  5-2.   These  rank
 as groups from most destructive but acceptable  at  some  sites  (1-4,  2-1,  and
 2-2);  through  less  destructive,  if considered to  apply  only to  lichens  (1-2
 and 1-5); then,  potentially  destructive, but  not  required to  be (8-2).   These
 are the  cleanup  techniques that  can take a  century  or more  to repair.
      Impacts only fractionally as  retrogressive are imposed by  Treatments 8-1
 and 8-3  and mainly  on  solid  precipitation.   Removal of  contaminated  ice  or
 snow  is  ranked  in this  same  time  span.  The non-impact  cleanups are  the
 barriers  (3-1,  3-2, and  3-3).
      In  Table  5-2 the operations of clear cutting  (1-2) and shallow  plowing  (1-5)
 are restricted to scraping lichens  from their support.  In this snow-blasted
 and windy zone, ten years may pass  before recolonization occurs, and one or
 several  centuries are likely to pass before a full-standing crop is  established
 again.   This is a drastic cleanup and has the longest recovery time of the
 likely treatments on vegetated sites.  Similar effects and recovery are
 attributed to treatment  1-5 when applied to lichens.
     The most drastic cleanup after-effects upon the environment derive from
removal of the aeolian "soils" in treatments 1-4, 2-1, and 2-2.   Replacement
with additional aeolian debris to an equivalent depth is postulated to take
one to several centuries.  If treatment  8-2 were to be constructed of these
                                     5-32

-------
                   Table  5-2.  Estimates of  the years to reach recovery after  various  cleanup  treatments.
Treatment
0-1 Natural Rehabilitation
1-1 Chemical Stabilization
1-2 Remove Vegetation by Clear Cutting
1-3 Remove Stumps and Grub Roots
1-4 Surface Scraping and Grading (<5 cm)
1-5 Shallow Plowing (<10 cm)
1-6 Deep Plowing (10 to 40 cm)
1-7 Soil Cover (<25 cm)
1-8 Soil Burial (25 to 100 cm)
2-1 Remove Plow Layer Soil (10 cm)
2-2 Remove Shallow Root Zone Soil (<40 cm)
2-3 Scrape Surface, Remove (<5 cm), Mechanically Stabilize
2-4 Remove Plow Layer (10 cm), Mechanically Stabilize
2-5 Remove Shallow Root Zone (<40 cm), Mechanically Stabilize
2-6 Scrape Surface, Remove (<5 cm), Chemically Stabilize
2-7 Remove Plow Layer (10 cm), Chemically Stabilize
2-8 Remove Shallow Root Zone (40 cm), Chemically Stabilize
3-1 Barriers to Exclude People
3-2 Barriers to Exclude Large Animals
3-3 Barriers to Exclude Large and Small Animals
4-1 Asphalt Hard-Surface Stabilization
4-2 Concrete Hard-Surface Stabilization
5-0 Sewage Sludge Application
6-1 Washing (<3 mm)
6-2 Flooding (3 to 30 cm)
7-0 Soil Amendments Added
8-1 Snow Fences and Wind Barriers
8-2 Watershed Control Devices Constructed
8-3 Snow and Ice Melt Additions
8-4 Removal of Contaminated Snow and Ice
Area Disturbed
(kn2)
0.01
---
NA
30-100
NA
100+
30-100
NA
NA
NA
100+
100+
NA
NA
NA
NA
NA
NA
0-1
0-1
0-1
*
*
*
0-5
*
*
0-10
1-100
0-10
0-5
0.1
---
NA
30-100
NA
100+
30-100
NA
NA
NA
100+
100+
NA
NA
NA
NA
NA
NA
0-1
0-1
0-1
*
*
*
0-5
*
*
0-10
1-100
0-10
0-5
1.0
—
NA
30-100
NA
100+
30-100
NA
NA
NA
100+
100+
NA
NA
NA
NA
NA
NA
0-1
0-1
0-1
*
*
*
NA
*
*
0-10
1-100
0-10
0-5
10.0
...
NA
30-100
NA
100+
30-100
NA
NA
NA
100+
100+
NA
NA
NA
NA
NA
NA
0-1
0-1
0-1
*
*
*
NA
*
*
0-10
1-100
0-10
0-5
Assumptions
Insufficient data to estimate.

Applies to lichens only.

Soil replaced only by wind & sedimentation action.
Lichens only.



Soil replaced only by wind & sedimentation action.
Soil replaced only by wind & sedimentation action.






No significant effect.
No significant effect.
No significant effect.






Minimum damage.
Variable impact from specific structures placed.
Relatively minor impact but additive specific.
Site dependent impact depends on removal logistics.
00
CO
        *0utside scope of work.
        NA - not appropriate in  this land type.

-------
 "soil" materials it would be rated equally damaging.   The recommended con-
 struction materials are rock fragments for the structures proposed to divert
 snow melt.
      Stone filled diversions created  from tatus (8-2)  are expected to be
 restored in a few years by natural rock-falls.   Except for the  potential for
 misuse of soils,  this  technique  would be  rated as  little  impact,  as are
 Treatments 8-1  and 8-3.   Installation of  fences (8-1)  to  create drifts and
 melting of snow and ice (8-3)  are  rated benign as  is  removal  of contaminated
 snow and ice (8-4).  High pressure hosing down with water (6-1)  has little
 impact if the logistics of obtaining  the  required  water is met  by an environ-
 mentally benign way.
      The procedures  judged least destructive  are shown in Table 5-2 as
 barriers,  mainly barriers to keep  people  out  (3-1) and a  few  transient animals
 that may pass through  (3-2,  3-3).   Damage during barrier  installation is pre-
 sumed minimized and  remedied by  solid precipitation cover in  a  few months.
      The effect of increasing  the  area of cleanup  can  only be visualized as
 including  additional numbers of  discrete  and  overall  cleanup  sites because
 of  the terrain  being limited in  area.   The  impacts and the times  to recover
 are area independent due  to  frequent  transitions from  ice to  rock  to snow
 cover.

 5.7  CONCLUSIONS
      Because  of the complete lack  of vegetative cover  and  the extreme  severity of
physiographic and  climatic factors  in  the aeolian mountain peak environment,
most  of  the conventional  cleanup procedures considered  for other terrestrial life
zones  and environments appear to be either  inapplicable or logistically  impracticable,
and probably  would be minimally effective.  For these reasons, and  the fact that
precedents for  perturbations of this environment and recovery therefrom  are ex-
tremely  rare  in the literature, accurate quantitative estimation of  the  impacts
of these treatments would be difficult.  It might appear that alternative treat-
ments that are applicable to the aeolian mountain peak  situation should be
devised as a  separate study, and their impacts then considered.   Possible alterna-
tive avenues of approach have been mentioned in this report.  With respect tc
"conventional" cleanup  procedures,  those involving  surface alteration and hard-
                                     5-34

-------
surface stabilization appear to have the most lasting impacts.  The severity of
impact and the amount of time required for "recovery" varys depending upon the
location and precise situation of the spill and the exact nature of the treat-
ment imposed.
     It is suggested that alternative techniques for immediate containment
and subsequent cleanup which are suited to the aeolian mountain peak situation
be devised.  These techniques would be more effective and less detrimental than
those methods devised for other environments.
     The preferred treatments are those involving barriers (3-1, 3-2, 3-3) and
removing contaminated snow or ice (8-4).   The least suitable treatments are
those which strip the lichens from surfaces, and defined this way are Treat-
ments 1-2 and 1-5, although they are not  physically the most destructive.
                                    5-35

-------
 5.8  AEOLIAN MOUNTAIN PEAK REFERENCES
 1.  Al'benskii, A. V. and P. D. Nikitin.  Handbook of Afforestation and Soil
         Melioration. 3rd ed. (Gosudarstvennoe Izdatel'stvo Sel'skokhozyais-
         tvennoi Literatury, Moscow: 1956.) Translated from Russian by Israel
         Program for Scientific Translations. Jerusalem, 1967.

 2.  Armstrong, R.  A.  Studies on the Growth Rates of Lichens.  In: Lichenology:
         Progress and Problems.   D. H. Brown, D.  L. Hallsworth and R. H. Bailey,
         eds. London: Academic Press, 1976.

 3.  Arnold, K. C.   An Investigation into Methods of Accelerating the Melting
         Ice and Snow by Artificial Dusting.  In:  Geology of the Arctic.
         G. D.  Raasch, ed.  University of Toronto Press. 11:198-1013, 1961.

 4.  Aulitzky,  H.  Forest Hydrology Research in Austria.  In: Forest Hydrology.
         W. E.  Sopper and H. W.  Lull eds. Oxford: Permagon Press, 1967.

 5.  Aulitzky,  H.  Significance of Small Climate Differences for the Proper
         Afforestation of Highlands in Austria.  W. E. sopper and H. W. Lull
         eds.  Oxford: Permagon Press, 1967.

 6.  Bailey, R. H.   Ecological Aspects of Dispersal and Establishment in Lichens
         In: Lichenology: Progress and Problems.  D. H. Brown, D. L. Hallsworth
         and R. H.  Bailey, eds.  London:  Academic Press, 1976.

 7.  Bettler, G. T.  The Plant Ecology of the Drakensberg Range. Nat'1 Museum
         Ann.,  v. 3, pp.  511-565, 1964.

 8.  Billings,  D. W.  Arctic and Alpine Vegetation: Plant Adaptations to Cold
         Summer Climates. In: Arctic and Alpine Environments. J. D. Ives and
         R. G.  Barry, eds. London: Metheuen ^ Co., Ltd., pp. 403-443, 1974.

 9.  Bond,  E. K.  Plant Disseminules in Wind Blown Debris from a Glacier in
         Colorado.   Arctic Alpine Res. 1, 165-169, 1969.

10.  DeQuervain, M. R. and H. R. Der Gand.  Distribution of Snow Deposit in a
         Test Area for Alpine Reforestation.  In: Forest Hydrology.  W. E.
         Sopper and H. W. Lull eds. Oxford: Permagon Press, 1967.

11.  Dingwall,  P. R.  Erosion by Overland Flow on an Alpine Debris Slope. In:
         Mountain Geomorphology. Slaymaker, 0.  and McPherson, H. J., eds.
         Vancouver: Tantalus Research, 1972.

12.  Dortignac, E.  J.  Forest Water Yield Management Opportunities. In: Forest
         Hydrology, W. E. Sopper and H.  W. Lull,  eds.  Oxford: Permagon Press,
         1967.
                                    5-36

-------
 13.   Dyunin,  A.  K.   Fundamentals  of the Mechanics  of Snow Storms.  In:  Physics
          of Snow and Ice.  International Conference on Low Temperature  Science
          Proceedings.  Sapparo,  Japan,  pp.  1065-1073,  August  1966.

 14.   Gams,  H.   Die  Herkunft  der Hochalpinen  Moose  und Flechten.  Jahrb.  Ver.
          Schutre Alpenpflanz. Munchen  25,  1-11,  1960.

 15.   Gary,  H.  L.  Airflow  Patterns  and Snow  Accumulation  in  a  Forest Clearing.
          West.  Snow Conference  Proc. 43:106-113, 1975.

 16.   Gregory,  P.  H.  The Microbiology  of the Atmosphere.   New  York: Wiley
          (Interscience), 1961.

 17.   Grove,  C.  S.,  Jr., S. T. Grove, and A.  R. Aidun.   The Theory  and  Use of
          Aqueous  Foams  for Protection  of Ice Surfaces.  In:  Ice  and Snow  Pro-
          perties, Processes  and Applications. W. D.  Kingery, ed. Cambridge-
          M.I.T.  Press,  pp. 666-684,  1963.

 18.   Hale, M.  E.  Jr.  The Biology of Lichens. London:  Edward Arnold (Publishers!
          Ltd.,  1967.  176 pp.

 19.   Hutchinson,  G.  E.  The  Ecological  Theater and the  Evolutionary Plan.  New
          Haven: Yale University Press,  1965.

 20.   Kappen, L.   Response to Extreme Environments.   In: The Lichens. V. Ahmndjian
          and M. E. Hale, eds. pp.  289-400.    New York: Academic Press, 1973.  697 pp

 21.   Llano, G. A.  The Terrestrial Life  of the Antarctic.  Sci. Am. 297-212-230
          1962.

 22.   Luty, E. T.  and R. W. Hoshaw.  Airborne Algae of  the Tucson and Santa
         Catalina Mountain Areas.   Journal  of the Arizona Academy of Science
         pp. 179-182, 1967.

 23.  Mani, M. S.  Introduction to  High Altitude Entomology. London: Methuen
         1962.

24.  Margaropoulos,  P.   Woody Revegetation  as a Pioneer Action Towards Restor-
         ing of Totally Eroded Alopes in Mountainous Watersheds.  In:  Forest
         Hydrology.  W.  E.  Sopper and H. W.  Lull,  eds. Oxford: Permagon Press
         1967.

25.  Martinelli, M.  Jr.   Snow Fences for Influencing Snow  Accumulation.  In:
         The Role of Snow and Ice  in Hydrology.  Symp.  on Measurement  and Fore-
         casting Proceedings. Banff, Alberta. 1972. Handout,  pp.  1394-1398.

26.  Martinelli, M.  Jr.   Possibilities  of Snowpack  Management in  Alpine Areas.
         In:  Forest  Hydrology.   W.  E.  Sopper  and  H. W.  Lull,  eds. Oxford:
         Permagon Press, 1967.

27.  Mattick,  F.   Die Flechten als  Ausserste  Vorposten des Lebans in Gebirge.
         Montagne e  Uomini, 2, 494-6,  1950.
                                     5-37

-------
28.  May, R. F. and W. D. Striffler.  Watershed Aspects  of  Stabilization  and
         Restoration of Strip-Mined Areas.   In: Forest Hydrology. W. E. Sopper
         and H. W. Lull, eds. Oxford: Permagon Press, 1967.

29.  O'Loughlin, C. L.  A Preliminary Study  of Landslides in the Coast Mountains
         of Southwestern British Columbia. In: Mountain  Geomorphology.  0. Slay-
         maker and H. J. McPherson, eds. Vancouver: Tantalus Research Ltd.

30.  Osburn, W. S.  The Dynamics of Fallout  Distribution in a Colorado Alpine
         Tundra Snow Accumulation Ecosystem.  In: Radioecology. V. Schultz and
         A. W. Klement eds. N.Y.: Reinhold and Am. Inst. Biol. Sci. Publ., 1963.

31.  Osburn, W. S. Jr.  The Impact of Technology: Large  Scale Examples. In:
         Arctic and Alpine Environments.  Ives, J. D., ed.  London: Methuen,
         1974.

32.  Owens, I. F.  Morphological Characteristics of Alpine Mudflows in the
         Nigel Pass Area. In: Mountain Geomorphology. 0. Slaymaker and H. J.
         McPherson, eds. Vancouver: Tantalus Research Ltd., 1972.

33.  Perla, B. I. and M. Martinelli, Jr.  Avalanche Handbook.  USDA Forest
         Service. Agriculture Handbook 489,  1976.  238 pp.

34.  Schaerer, P. A.   Terrain and Vegetation of Snow Avalanche Sites at Rogers
         Pass, British Columbia.  In: Mountain Geomorphology.  0. Slaymaker
         and H. J. McPherson, eds. Vancouver: Tantalus Research Ltd., 1972.

35.  Shiotani, M. and H. Arai.  On the Vertical Distribution of Blowing Snow.
         In: The Physics of Snow and Ice. pp. 1075-1083, 1967.

36.  Swan, L.  W.  Some Environmental Conditions Influencing Life at High Alti-
         tudes. Ecology 33:109-111, 1952.

37.  Swan, L.  W.  Aeolian Zone.  Science 140:77-78, 1963.

38.  Swan, L.  W.  Alpine and Aeolian Regions of the World. In: Arctic and Alpine
         Environments.  H.  E. Wright, Jr. and W.  H.  Osburn, eds. Indiana
         University Press,  1967.

39.  Tabler, R. D.  Predicting Profiles of Snowdrifts in Topographic Catchments.
         West. Snow Conference Proc. 43:87-97, 1975.

40.  Tabler, R. D.  Evaporation  Losses of Windblown Snow, and the Potential for
         Recovery. West.  Snow Conference Proc. 41:75-79, 1973.

41.  Tabler, R. D. and R.  A.  Schmidt.  Weather Conditions that Determine Snow
         Transport Distances at  a Site in Wyoming. In: The Role of Snow and Ice
         in Hydrology, pp.  118-126, 1973.

42.  Wangersky, P. J.   The  Organic Chemistry of Sea  Water.  Am. Scientist.  53:
         358-374,  1965.
                                     5-38

-------
43.  Whittaker, Robert A.  Communities and Ecosystems. Macmillan Publishing
         Co.,  Inc. 1970.  385 pp.

44.  Williams, G. P. and L. W. Gold.  The Use of Dust to Advance the Breakup
         of Ice on Lakes and Rivers.  East. Snow Conference Proc. 1963. pp 31-60
         (Also available as: Tech. Paper 165. Division of Building Res. Nat. Res,
         Council of Canada, Ottawa.)

45.  Wilson, A. T.  Organic Nitrogen in New Zealand Snows. Nature 183:318-319,
         1959.

46.  Wilson, A. T.  Surface on the Ocean as a Source of Airborne Nitrogenous
         Material and Other Plant Nutrients.  Nature. 184:99-101, 1959.
                                     5-39

-------

-------
                                   CHAPTER 6
                                    TUNDRA

6.1  OVERVIEW
     Tundra, which in Russian-Finnish usage referred to the treeless zone of
northern Eurasia, especially Finland, was subsequently applied to its counter-
part in the Arctic and Alpine North America.  The Soil Conservation Society of
America defines tundra as "the treeless land in arctic and alpine regions;
varying from bare area to various types of vegetation consisting of grasses,
sedges, forbs, dwarf shrubs, mosses and lichens."  It is generally considered
that the tundra vegetational life-form represents the integrated effect of
primary climatic parameters.  From a practical point of view, tundra should be
defined by combinations of physiognomy and environment.    In this definition
the tundras are the grassless arctic plains, the vegetation of which may form
varied and often complex patterns of dominance by dwarf-shrubs, sedges and
grasses, mosses, and lichens.  The tundras of North America and Eurasia are
quite similar; their principal herbivorous mammals include the musk ox, cari-
bou, arctic hare, and lemming.  Longspurs, plovers, snow bunting, and horned
larks are characteristic birds; reptiles and amphibians are few or absent, as
in alpine grasslands.  In many tundras the deeper layers of soil are perma-
nently frozen, and only the surface soil is thawed and becomes biologically
active during the summer.  The semi-arid areas of the Arctic are generally
considered as part of the tundra.
     Dagon  stated that the variety of definitions and the resultant misunder-
standing accompanying the application of the term tundra have been far too
numerous.  He lists numerous definitions for the term and points out that there
is a great deal of ambiguousness associated with the use of the word tundra.
For the purpose of this report, the term tundra defined on the basis of its
physiognomy and environment is more appropriate.
                                      6-1

-------
6.1.1  Geographical Distribution
     Tundra lies between the northern limit of trees and the area of perpetual
ice and snow in the far north, or above timber line in high mountains.  In
North America, it forms a broad band completely across the continent, and it
also occupies the narrow low coastal area around most of the periphery of
Greenland.  It occurs on mountains as far south as Mexico if their elevation
is sufficient to produce a timber line.  Thus it is limited in its northern
or upward extent by ice and bounded on the southern or lower margin by boreal
or subalpine conifer forest.  The tundra is essentially an arctic grassland.
It is the cold polar area beyond the treeline in the northern circumpolar lands
of North America and Eurasia.  The tundra vegetation and environment continues
southward along mountain alignments in western North America and Siberia.  The
term tundra is further subdivided into Arctic and alpine.  Alpine and Arctic
tundra are discussed in detail in the vegetation section (6.1.3) of this
chapter.

6.1.2  Environment
     Low temperatures and a short growing season (about 60 days) are the main
characteristics of tundra environment.  The ground remains frozen except for
the upper few inches during the growing season.  The dominant factors which
                            14
shape the tundra environment   are:  the character of the solar energy input,
the nature of the immediate and adjacent surfaces, weather systems, and topog-
raphy.  The annual and daily cycling of solar energy received on a unit surface
in the arctic tundra is quite different from that experienced at lower lati-
tudes.  In general, tundra summer is limited to the three months of June, July,
and August; winter occupies the remainder of the year.  Daily air temperatures
are generally below 10°C in summer but may rise occasionally to 20°C during
brief sunny spells.  The mean daily range of temperature in summer is about 6°C,
The coldest months are December through April; the mean daily temperature in
tundra areas in February is about -30°C and the mean daily range varies from
about 10 degrees in continental areas to about 6 degrees elsewhere.  Extreme
low temperatures of about -50°C occur in the exposed tundra.  Topography and
vegetative cover are probably important factors in the temperature regimes of
tundra environments.
                                      6-2

-------
     The tundra is often termed a "polar desert."  Annual precipitation in this
area is indeed light.  However, thawing of the active layer each summer causes
vast tracts of wet lands.  By and large the annual rainfall is equal to the
water equivalent of snowfall.  Normally there is adequate precipitation for
vegetative growth; deficiencies in moisture supply become critical only briefly
during the occasional summer droughts.  Droughts have been sufficiently severe
to permit serious tundra fires to occur.  Low evapotranspiration rates and the
presence of permafrost result in a large percentage of the precipitation run-
ning off into the tundra's poorly developed streams and stagnant ponds.  It is
difficult to overemphasize the importance of the snow cover on the tundra
during the eight to ten months of the year.  Snow is essential to the survival
of many plants and animals.  In addition to being a source of water, snow acts
as a protective blanket to many plants and animals.
     Wind plays a dominant role in the tundra.  Strong winds are usually the
most critical.
     Both plant and animal life exist precariously, mainly by taking refuge in
a particular microclimate environment and thereby escaping the rigors of gen-
eral climate.  The microclimatic environment is in a delicate state of balance,
and alteration of this balance either naturally, or unnaturally, such as through
vegetative modification, may have serious consequences.  For example, the alter-
ation of albedo can upset the energy balance at the soil surface and result in
thermokarst conditions.  The removal of tall vegetation will result in increased
winds and shallower snow cover which may also significantly change the thermal
regime near and in the ground.

6.1.3  Vegetation
     Tundra vegetation is low, dwarfed, and often matlike, and includes a high
proportion of grasses and sedges.  Even the woody plants, including the willows
and birches, are usually prostrate.   The herbs are mostly perennial and of a
rosette type, producing relatively large flowers, often with conspicuous colors.
Mosses and lichens may grow anywhere and in favorable habitats form a thick
carpet with the low herbs.  The number of species is small compared with floras
of temperate climates; and, even within the tundra, the number decreases north-
ward.  Most of the genera and many of the species are to be found throughout
the Northern Hemisphere wherever tundra occurs.

                                      6-3

-------
      The uniformity  of  the  flora  is undoubtedly  related  to  the  peculiarities
 of  environment.  The growing  season is  short  and its  temperatures  are  relatively
 low.   Permafrost in  the ground  and its  depth  of  thawing  in  summer  is of great
 importance  as  is the extreme  instability of soil, related to  the frost action.19
 Light is continuous  throughout  the growing season in  the arctic tundra, and is
 intense and high in  the ultraviolet rays in alpine tundra.  Local  marked dif-
 ferences in vegetation  are  commonly related to minor  variations in topography
 and differences they produce  in drainage and  retention of snow.
      The tundra ecosystem can be  divided into (1) Arctic Tundra, and (2) Alpine
 Tundra.

      6.1.3.1  Arctic Tundra--
      The circum-artic tundra can  be regarded  as  comprising  a  simple formation-
 type  composed of Eurasian and North American  formations.  Even  within  the same
 formation,  however,  in  spite of superficial similarities, considerable con-
 trasts are  evident in both  life-form and general appearance.  Although flower-
 ing herbaceous plants,  particularly grasses and  sedges, are almost universal
 throughout  all communities, dwarf shrubs, mosses and  lichens  compete successfully
 in most places.  Furthermore, there is  a great contrast between the almost
 closed-cover tundra  communities near the forest margins and the scattered,
 sparse communities on almost bare ground in the most  inhospitable parts of
 north Greenland and Spitzbergen.  It is clear also that not all the extensive
 tundra communities are  to be regarded as true climatic climax.
      In the climatically more favored areas of the tundra, the  true climatic
 climax communities are probably those in which grasses and sedges are dominant
 with a substantial understory of  lichens and mosses.  Communities of this
 nature are found on undulating areas with a considerable range of slope,  aspect
 and lithology.  The dominants in  the American formation are various species of
 sedge  (Carex spp.), cotton grass  (Eriophorum  spp.),  and woodrush (Luzula spp.),
 along with species of many genera of grasses  such as the bents  (Agrostis  spp.),
 foxtail grasses (Alopecurus spp.), fescues (Festuca spp.), and timothy (Phleum
 spp.) which are also of common occurrence in middle latitudes.  The commonest
 associated lichens belong to the genera Stereocaulon,  Alectoria, Cetraria,  and
 Cladonia,  while mosses of the genus Polytrichum are most frequently encountered.
Along with these are a great variety of plants with showy flowers belonging to

                                      6-4

-------
genera which are also distributed throughout the woodlands and meadows of mid-
dle latitudes such as the anemones (Anemone spp.), marsh marigolds  (Caltha spp.),
buttercups (Ranunculus spp.), avens (Beum spp.), cinquefoils (Potentilia spp.),
campions (Lychnis spp.), catchflies (Silene spp.), and primroses (Primula spp.)-
These are accompanied by flowers which are more typically associated with high
latitudes and high altitudes such as the mountain avens (Dryas spp.), saxifrages
(Saxifraga spp.), and gentians (Gentiana spp.).  A varied cover of predominantly
herbaceous vegetation thus forms an almost complete turf over the ground.
     Over very large areas, however, the grasses and sedges are much less im-
portant and dominance is assumed by foliose lichens and mosses, particularly
the former.  This cover of cryptogams, if ungrazed, may be anything from four
to ten inches high and, in the more southern parts of the tundra, may be almost
continuous.  Further north communities similar to the two already described
are much more discontinuous and a great deal of bare ground is found between
individual plants or plant aggregates.  This type of vegetation has been most
vividly described in the Taimyr Peninsula by Middendorf.  Here, in many places,
the ground is only about half covered by plants, permafrost is not more than
three inches beneath the surface in many places, and a varied patchwork of com-
munities is found.  In places mosses (Polytrichum spp.) dominate; elsewhere
sedges or lichens (Cladonia spp.).  Throughout all these communities, however,
there is an admixture of flowering plants, creeping shrubs such as the dwarf
willow (Salix lapponum) and xerophytic, heath-like shrubs such as the crowberry
(Empetrum nigrum).  Even at the height of summer, when many species are in
flower, these communities give the landscape a drab or yellowish aspect.  This
is partly due to the colors of the mosses and lichens, but partly to the fact
that many of the grasses and other flowering plants, as an adaptation to reduce
transpiration,  retain the dead remains of last year's growth.  Ultimately,
around the very fringes of ice caps and on some exposed sea coasts, the vegeta-
tion dwindles and only an occasional tussock of sedge or cushion-shaped plant
is to be found.
     Dwarf shrubs occur sporadically throughout the communities already de-
scribed but locally shrubs become dominant.   Usually, however,  the shrubs,  which
are subordinate elements in predominantly herbaceous communities, belong to
different species from those which form arctic heath and arctic scrub communi-
ties.  In arctic heath the species are low growing, the Canadian communities

                                      6-5

-------
being dominated by small  shrubs such as  the arctic bell-heather  (Cassiope
tetragona), crowberry  (Empetrum nigrum), and alpine bearberry  (Arctostaphylos
alpina).   In arctic scrub, on the other  hand, quite a busy growth  is attained
which may  reach a height  of several feet.  It is dominated by  broad-leaved
shrubs  such as the narrow-leaved Labrador tea (Ledum palustre),  Lapland rose-
bay  (Rhododendron lapponicum), alder (Alnus sinuata), scrub birch  (Betula
glandulosa), and numerous species of willow (Salix spp.).  Many  communities are
found containing both these types of shrub, however.
     Usually the arctic shrub communities are found on sheltered slopes, par-
ticularly  where the soil  thaws unusually deeply in summer and  where it is rich
and moist.  It seems probable that they  constitute a zone of ecotone between
the more typical tundra communities on the interfluves and the outliers of the
forest  on  the lowest slopes.  There is some evidence that, where trees are
nearby, invasion takes place sporadically in a series of favorable years only
to be rebuffed in a subsequent unfavorable period.
     Finally, there is another suite of  communities, covering  extensive areas
in the  tundra, which cannot be regarded  as climatic climax; these are the tundra
bogs and moors.  They are very similar in appearance and species-content to the
bogs or "muskegs" which occupy ill-drained localities within the general area of
the boreal forest.  Very often in both Europe and America, they are dominated
by cotton grass (Eriophorum vaginatum)  along with many other species of sedge
and some grasses.   Usually also various species of moss, particularly bog moss
(Sphagnum  spp.), compose a sub-dominant understory.  The surface of these bogs
is usually "tussocky" and the tussocks of cotton grass and moss are frequently
invaded by dwarf willows.  Indeed, at an advanced stage of development, when
the peat has achieved considerable thickness,  the vegetation may become quite
luxuriant, similar in many ways to that of the better-drained arctic heaths.
These "heath moors," as they have been called, ensure a copious supply of
available moisture throughout the growing season and,  at the same time, thaw
out to a sufficient depth to accommodate the rooting systems of shrubs.  Apart
from the heathy shrubs, some broad-leaved,  deciduous shrubs also invade this
kind of surface.   In Canada, for instance,  several species of birch (Betula)
and many of willow (Salix) are of frequent  occurrence.   It seems likely that
soil conditions here would be quite suitable for those species of tree which
regularly invade drying-out muskegs within  the boreal  forest areas—species

                                      6-6

-------
like the black spruce CPicea mariana)  and the Canadian larch or tamarack
CLarix laricina).   Indeed,  if these species of tree were able to advance un-
interruptedly, along a solid front, across bogs in the southern parts of the
Canadian tundra,  they would probably demonstrate their ability to survive the
climatic rigors there.  Individual tree seeds which have been carried by the
wind have been observed to germinate here quite satisfactorily but have sub-
sequently been incapable of surviving the desiccating blizzards of winter.
Isolated saplings cannot survive but if they had the partial shelter of nearby
parent trees, they would probably be able to do so; the forest would thus en-
croach gradually.  This cannot take place, however, since the deep peat areas
are  separated from the forest edge by less favorable types of terrain.
     The intense physiological drought of  the  winter months  should be  stressed
as  a limiting factor  in the  tundra far more  than mere  low temperatures.   If the
air  always remained  still, many  of the northern trees  could  probably withstand
even lower temperatures than they do.  Strong  winds,  even at very low  tempera-
tures,  are able  to  abstract  water quite  quickly from  living  plant tissues;  if
this proceeds beyond a  critical  point, death ensues,  since  no  replacements  are
obtainable from  the completely  frozen soil.   The  plants of  the tundra  minimize
the danger of this  in several distinct ways.  On  the  most  exposed areas,  all
are very low-growing; many die  down completely in winter and survive as under-
ground rhizomes, root stocks, conns and  bulbs.  Those herbaceous plants which
 do maintain  organs  above  ground level in winter do so in the form of very tight-
 packed rosettes  or  cushions; many species of saxifrage do this very success-
 fully.  As a group, the lichens are noted for their ability to survive in an
 almost desiccated state for protracted periods.  It is understandable therefore
 that they should form such an important element in the vegetation of many of
 the bleakest and baldest arctic interfiles.  The arctic shrubs  are also low-
 growing except in the most  sheltered valleys.  They have adopted different de-
 vices to prevent over-rapid  transpiration.  Many of the broad-leaved  ones  are
 deciduous (Salix, Alnus, Betula) while  the  heaths all  have  the  characteristic
 small  leaves with rolled-back margins.
      The shortness  of  the growing season  also imposes  great restrictions on
 plant  growth.   It  is mainly because  of  this factor that the flowering herbs of
 the tundra  fall  into two  classes; there are those  which flower at  the beginning
 of the  growing  season  and  those which flower towards  the  end  of it.   The
                                        6-7

-------
early-flowering species are typically plants with substantial underground food-
storage organs; they use the stored energy from the preceding year to flower
and produce seed and then utilize the remainder of the growing season to build
up food for the following year.  The late-flowering species use the first part
of the season to build up a store of energy and then use most of it in produc-
ing seed; they then pass the winter as inconspicuous underground organs or, if
annuals, survive merely as seeds.
     The factors of winter frost, winter winds, short growing season and perma-
frost are responsible, collectively or separately, for precluding trees from
the tundra.  Because of contrasts in relief and aspect the relative importance
of these different factors varies greatly and an evaluation of the exact reasons
for the absence of trees in a particular place is usually obscure.  However,
the complex inter-digitation of the southern edge of the tundra with the north-
ern edge of the forest indicates that, generally speaking, wind force along
with shallowness of unfrozen soil are the operative factors in the southern
parts of the tundra.  Here, forests occupy the valleys while the interfluves
are tundra.  Wind force is obviously important on these interfluves but,
whether forest would extend further up the slopes if the depth of the thawed
layer in summer were deeper is not clear.  The existence of permafrost, as
such, does not preclude forest since vast areas of the northernmost parts of
the boreal forest are underlain by it both in Siberia and Canada; northern
species such as the Dahurian larch (Larix dahurica) and Siberian dwarf-pine
(Pinus pumila) are adapted to this soil condition in that they develop a root
system which is entirely spreading with no tap root whatsoever.  In the north-
ern archipelago of Canada and the arctic peninsulas and islands of the U.S.S.R.,
as well as in the Spitzbergen and the north of Greenland, it appears probable
that trees cannot possibly regenerate regardless of wind and soil depth.  Though
increase in the length of the season with continuous daylight does, to a cer-
tain extent, compensate for increased latitude, it is doubtful if the growing
season here is sufficiently long for trees.  Even if trees could survive the
winter, it is very doubtful if they would be able to flower and produce seed.
In these fringing areas, so near to perennial ice, there is no recognizable
frost-free season.   Everywhere in the tundra the mean length of the frost-free
season is less than fifty days but here, in spite of continuous daylight,  air
frosts occur frequently in all the summer months.  Returning to the southern
                                      6-8

-------
fringes of the tundra, however, it is doubtful whether the length of the frost-
free season has any bearing on the position of the limits of the forest.  Prob-
ably the length of the frost-free season is sufficient here to permit forest
regeneration but wind speed in winter and extreme shallowness of soil in summer
prevent the spread of trees.
     Fluctuations in the position of the northern edge of the boreal forest
have occurred during the past few millennia.  Locally, in European Russia, a
distinctive fringing zone of "tundra moor" has been described.  Here, hummocks
of bog moss (Sphagnum spp.) peat, each several yards in diameter, with narrow,
marshy channels between them, cover the landscape.  The mosses on the hummocks
are often  still growing but the cores of the hummocks are permanently frozen.
This permanently frozen peat contains the  stumps of both birches and conifers,
many of which are in position of growth.   Clearly these areas were forested
not so very long ago but the reason for forest recession is  not altogether
clear.  Possibly a deterioration  in climate was responsible; possibly,  the
activities of reindeer herders along  the former forest margins  caused the sap-
lings  to  be grazed off and  thus prevented  regeneration; possibly both climatic
change and human activity played  some part in  the changes  that  occurred.
     Although  the  flora of  the tundra is fairly well  known,  its communities  and
their  successional relationships  have not  been sufficiently studied.  In con-
trast  with temperate  vegetation,  many species  may occur  in any  type  of  habitat,
and several  that  appear to  be  climax  may also  be  pioneers  in the newest of
habitats.  Even climax  is  not  agreed upon, possibly because observations have
been made in widely  separated  areas.  Variations  do exist, as has been  shown
 for eastern Canada.   It  is also  possible that the climax criteria used  for
 temperate climates are  not always applicable in  the arctic.  Interpreted in
 terms of Greenland vegetation, Cassiope heath appears to be climax, and a Sedge-
 Dryas dominated community, of equal extent but on drier sites,  is preclimax.
 Two subclimaxes are frequent.   Any habitat with sufficient moisture, whether it
 be pond margin, seepage area,  or boggy ground, eventually is covered with a
 thick moss mat supporting several herbs, of which cotton grass (Eriophorum spp.)
 is most conspicuous.  Xerarch succession on rock exposures eventually results
 in a  lichen-moss mat, which may continue almost indefinitely.
       Important climax dominants are  Cassiope tetragona, one or more species of
 Vaccinium, Archtostaphyloas alpina,  Empetrum nigrum, Andromeda polifolia,

                                       6-9

-------
Ledum palustre, Phododendron lapponicum, and species of Betula and Salix.
These and other species occur in varying combinations and degrees of importance
     Practically all habitats support some of the many species of Carex, of
which the commonest include C. capillaris, C. nardina, and C. rupestris.  The
preclimax sedge community invariably includes Elyna bellardi in abundance.
Some grow mats, some are in clumps, but all are dwarfed.  The same can be said
for the grasses, which, although relatively abundant and widespread, are re-
stricted to a few genera, of which Festuca and Poa are especially well repre-
sented.  Many of the conspicuous herbs previously mentioned are included in the
numerous species of one of the following genera:  Saxifraga, Potentilla, Ranun-
culus, Draba, Cerastium, Silene, Lychnis, Stellaria, and Pedicularis.  Conspic-
uous and widespread species typical of tundra are Oxyria digyna, Papaver spp.,
Dryas octopetala, and Epilobium latifolium.

     6.1.3.2  Alpine Tundra-
     Mountains high enough to have timber line support tundra, whose upward
                                        *
extent is limited by the snow line.  In the east, as a consequence, tundra is
found only on a few high peaks in New England.  Farther south, the Appalachians
are not of sufficient height to support tundra.  That on Mt. Washington is
representative of the type and is essentially similar to the not far distant
arctic vegetation.
     Alpine tundra in the western mountains mostly lies far to the south of the
arctic and, consequently, is found at high altitudes only.  In the Canadian
mountains, it is found as low as 6,000 feet, but southward its lowest elevation
steadily increases some 360 feet per degree of latitude to 30° north latitude,
and then declines very gradually to the equator.  In the central Rocky Moun-
tains, tundra is well developed between 11,000 and 14,000 feet.   In the Sierra
Nevada, where the snow line is lower, tundra lies mostly between 10,500 and
13,000 feet.  In general, it is lower on coastal than interior ranges and on
the wetter sides of mountains.
     When climate changed and terminated the glacial period, vegetation similar
to modern tundra must have followed the ice as it receded northward.   This left
only these high peaks and ridges where tundra could survive as relicts.   The
relict vegetation obviously belongs to the Tundra Formation because of the
                                     6-10

-------
growth form and the duplication of characteristic genera as well as many species.
The greater importance of grasses and the presence of numerous endemics in the
western mountains suggest that these alpine tundras might be classed as asso-
ciations of the Tundra Formation.
     Webber23 compared the composition and production of the vegetation at
Point Barrow (Alaska) and Niwot Ridge (Colorado).  The apline tundra vegetation
on Niwot Ridge is twice as productive as arctic  tundra (Point Barrow) but only
because of higher incoming solar radiation (Table 6-1).

6.1.4  Animal Life
     The animal  life  in the tundra biome consists of caribou, musk ox, arctic
hare, lemming, longspurs, plovers, snow bunting, and horned  larks.  There are
very few,  if any, reptiles and amphibians  in the tundra  environment.  Though
vertebrate animals  inhabiting arctic and alpine  tundra environments are faced
with many  similar environmental problems,  such  as  low temperatures and scarce
cover,  important differences between arctic and  alpine environments exist in
day-length and solar  radiation patterns, topography  and  moisture,  and other
physical  factors of the environment.  These differences  in environment are
reflected  in the distribution of animal  life  in the  arctic and  alpine  tundra.
The  differences  occurring between arctic and  alpine  faunas are  due not only
 to real differences in the habitats  available  as a result  of physical  environ-
mental  differences, but also  to  the  much patchier "island" distribution  of
 alpine  habitats  in contrast  to  those of the more continuous circumpolar  arctic.

 6.1.5  Soils
      Low temperatures and, in most places, the presence of an impermeable
 perennially frozen substratum,  are the dominant factors affecting the develop-
 ment of soils of the arctic and alpine tundra.  The cold environment modifies
 soil-forming processes to such an extent that it overshadows, and may even
 completely obliterate, the effects of relief and time on soil characteristics.
 Tundra soils are wet nearly continuously during the thaw period.  In general,
 the tundra soils can be divided into two groups:  (1) poorly drained soils that
 are usually saturated, and (2) well drained soils.
                                      6-11

-------
   Table 6-1.   Comparison of arctic and alpine environments  and vegetation.
Component
Latitude
Altitude (meters)
Average July solar radiation
(cal/cm2/min)
Maximum photoperiod
Maximum air temperature (°C)
Maximum soil temperature (°C)
Annual mean precipitation (mm)
Permafrost
Average length of growing period
(days)
Number of common vascular plants
Average area! vascular production
Arctic tundra
71°N
7
0.30
84 days
3.9
2.5
107
Universal
55
40
100
Alpine tundra
40°N
3749
0.56
15 hours
8.5
13.3
1021
Sporadic
90
100
200
  (grams/m2/yr)

Average ratio of above- to
  below-ground biomass

Average areal net photosynthetic
  efficiency (percent)
1:8


0.5
1:12


 0.5
                                    6-12

-------
     Poorly drained mineral soils of the tundra all have essentially the same
sequence of horizons.  They consist, typically, of a mat of organic materials
at the surface; a grey, greyish brown, or bluish mineral layer which may be
mottled in whole or in part with colors ranging from olive or olive brown to
reddish brown and which may contain patches or streaks of organic matter; and
a grey perennially frozen layer which may have patches of organic matter in its
                                                                          10
upper part.  The description in Table 6-2 of a soil in northwestern Alaska
is representative.   (Color and Munsell color notations are for moist conditions
unless otherwise noted.  Depths are measured up and down from the surface of
the mineral soil.)
     Well drained soils occur throughout tundra areas, but are of relatively
small total extent except in arid regions near the northern limits of vegeta-
tion.  South of those regions they occur almost exclusively in stony, gravelly,
or coarse sandy materials on steep slopes, high narrow ridges, escarpment edges,
stabilized dunes and beach ridges, and elevated portions of flood plains.  The
combination of coarse-grained materials and exceptionally good surface drainage
is, with rare exceptions, an essential condition for their existence.  Either
bedrock or dry permafrost underlies these soils; where they are developed in
thick unconsolidated materials,  solid permafrost may occur at greater depths.
     The well drained  soils differ  from most  soils of the tundra  in  that excess
moisture is able  to  escape.  As  a result, oxidizing rather than reducing
processes  are dominant, and soluble products  of weathering are redistributed or
removed from the  soil.  The rate of weathering, decomposition or  organic mate-
rials,  and movement  of materials (that  is,  the intensity of  soil  development)
depend  on  temperature  and  the  amount  of moisture  available.   In the  cold and
relatively arid  arctic environment, these processes go  on at  a much  slower  rate
than  in forested  areas to  the  south.  As a  consequence, the  soils that  are
formed  may have  the  same  sequence of  horizons as  in comparable well  drained
soils of  taiga areas,  but  have much thinner horizons.   In nearly  all cases,
physical  and  chemical  properties of the  soils reflect those  of the parent
materials.
      The  well  drained  soils vary in the kinds of  horizons  that develop,  and in
 the thickness  and degree  of development of  the horizons.   The soil described in
                                     18
Table 6-3, from southwestern  Alaska,    is  representative  of  the most extensive
 of these  soils.   Colors  are  for moist conditions.

                                      6-13

-------
                                  Table 6-2.  Representative poorly drained tundra soil.
        Horizon    Depth  (cm)
                       Color
                 Consistency
          Oil


          012



          B21g
cr>
i
          B22g
          Clf
          C2f
18-8       Black (7.5YR2/1, wet) to dark
           brown (10YR4/3, squeezed)

 8-0       Very dark grey (10YR3/1, wet)
           to dark brown (10YR3/3,
           squeezed)

 0-3       Dark grey (N 3/) silt loam,
           with common large distinct
           olive brown (2.5Y 4/4), light
           olive brown (2.5Y 5/4), and
           grey (N 5/) mottling or
           staining

 3-23      Dark grey (N 4/-3/) silt loam;
           massive, with tendency toward
           platiness
23-28      Frozen very dark grey (N 3/)
           silt loam
28-46      Frozen dark grey (N 4/) silt
           loam
Partially disintegrated, finely fibrous peat.
Partially disintegrated, finely fibrous peat,
with inclusions of olive yellow and olive
brown silty material.

Nonsticky and nonplastic; few roots.
Nonsticky and nonplastic; some small  pebbles;
abrupt smooth lower boundary.  The lower 3 cm
of this horizon has erratic streaks of fine
soft concretionary material intermixed with
partially disintegrated organic material.

Includes partially disintegrated, finely fi-
brous, very dark greyish brown organic
material; some small  rock fragments.

Includes dark brown (7.SYR 3/4) partially  dis-
integrated, finely fibrous organic material;
ice constitutes about 75 percent of volume.

-------
                                  Table 6-3.  Representative well  drained tundra soil.
         Hori zon
         Depth
                   Color
                     Consistency
CT>
I
01


02


All



A12
                    5-2


                    2-0


                    0-8
        Dark reddish  brown  (5YR 2/2)
        Dark brown  (7.5YR 3/2)
8-15    Dark brown (7.5YR 3/2)
                   15-30    Brown (10YR 4/3)
                    30-71    Brown  (10YR 5/3)
Mat of loose litter, roots of shrubs and grasses,  and
partially decayed organic materials.

Well decomposed organic matter; many roots;  strongly
acid; abrupt smooth lower boundary.

Very gravelly silt loam; weak very fine granular
structure; very friable, smeary when rubbed; many
roots; very strongly acid; abrupt wavy boundary.

Very gravelly silt loam; weak very fine granular
structure; friable; roots common; very strongly acid;
clear wavy boundary.

Very gravelly silt loam; weak very thin platy struc-
ture; friable; roots common; many fine tubular pores;
dark brown (7.5YR 3/2) krotovinas up to 8 cm in diam-
eter; few roots; very strongly acid; gradual boundary

Very gravelly silt loam; weak thin and medium platy
structure, breaking to very fine angular blocks;
friable; no roots below 50 cm; strongly acid.

-------
     The principal horizons in this soil are the dark upper horizon and the
yellower subsoil horizon, which grades into the shaly parent material.  The
soil is acid and low in exchangeable bases throughout.  Chemical analyses in-
dicate the A12 horizon is slightly higher in free iron than both the All hori-
zon above it and the B horizon below it.  A similar distribution of iron has
                                       9 5
been noted in other soils of this kind, '  indicating that there is a tendency
in these soils toward the development of an iron-rich subsurface horizon.  The
process of podsolization, in which iron, aluminum, and organic carbon are moved
in solution from the upper mineral horizon and reprecipitated in a lower hori-
zon, is common in soils of cool, moist, forested regions.  In a few places in
tundra regions just north of the treeline where large quantities of water move
through coarse-grained soils, as in micro-depressions in uplands or protected
areas where snow accumulates, the process is also intensive.  There the soils
have bleached upper horizons over thin horizons of iron accumulation.   Soils
of this kind are more common in tundra areas immediately above treeline in sub-
arctic regions.
     Soils developed in materials derived from basic rocks, such as basalt,
limestone, and calcareous shales, are similar in appearance to the acid soils
described above, but pH values and base saturation are high.  In some places,
calcium carbonate has accumulated in the lower part of the soil.
     With increasing distance from the treeline, the amount of organic matter
added to the soil decreases and, because of lower temperatures and lower pre-
cipitation, chemical and biological processes such as oxidation and huraifica-
tion proceed at a slower rate.  Well drained soils in the colder tundra areas
may have only very thin or no dark upper horizons and thin brown B horizons.
In the far north, where precipitation rates are very low and the vegetative
cover is sparse, well drained soils that have no discernible horizon develop-
ment are dominant.  Other soils with no horizon development are the young sandy
                                                 17
soils of active dunes bordering some major rivers   and the very gravelly soils
of flood plains, low terraces, and steep colluvial slopes.
6.2  NATURAL PERTURBATIONS
6.2.1  Fire
     Little is known of the role of fires in tundra ecosystem, yet they do oc-
cur.  Fire removes most of the insulating layer of live vegetation and litter,
                                     6-16

-------
damages or destroys the peat, and greatly changes the albedo of the surface.
The result is a potential rapid increase in thermal erosion with increased
runoff.  A general account of vegetation recovery following a major fire in
the tundra is given by Cody.6  Nine years after the fire, grasses, sedges,
and forbs recovered fairly well but lichens and mosses showed little recovery.
With lichens playing an important role in the diet of caribou and reindeer, it
is evident that burned areas will be lost to grazing for many years.

6.2.2  Drought
     The tundra ecosystem is subjected to frequent droughts.  Droughts have
been sufficiently  severe to permit serious tundra fires  to occur.  Any long
term drought patterns are expected to affect the structure and productivity of
an ecosystem.

6.2.3  Grazing
     Overgrazing  by  the  caribou  and reindeer is  quite  common  in  the  tundra eco-
system.  These  animals  lead  to selective removal of  lichens through  grazing.
Vegetation composition  and density are  affected  considerably.   In the  tundra,
intense  grazing activities tend  to result in more  woody  and hardy shrubs.

6.3  MAN-MADE  PERTURBATIONS
     Disturbances to tundra  vegetation  due to  activities of man derive from
both winter and summer roads,  pipelines, seismic lines,  drill sites, crude oil
 spillage,  tundra fires,  living sites,  and waste  disposal.
     The discovery of oil on July 18,  1968 near  Prudhoe  Bay on Alaska's arctic
 coastal  plain (the North Slope)  and  subsequent flurry of oil  and gas explora-
 tion and development have already had a pronounced impact on the tundra environ-
 ment over a large area.  In addition to the direct effects of oil development
 on the North Slope, there will be subsidiary effects in other parts of arctic
 tundra.   Klein12 described the impact of oil development in Alaska.  Detailed
 descriptions of human impact of tundra  are also given by Watson, Bayfield, and
      22               21
 Moyes   and by Walker.
      Bliss and Wein3 examined 25 seismic trail  sites in the Low Arctic.   By
 using control plots in adjacent vegetation, they were able to characterize

                                      6-17

-------
the disturbance (Table 6-4).  They concluded that wherever viable roots and
rhizomes of shrubs, sedges, or grasses remained in the disturbed substrate, a
30 to 50 percent recovery in natural plant cover will occur in 3 to 5 years.
Subsidence will occur in areas of high ice content.  To date, there is little
evidence that immediate reseeding can prevent ground ice melt and subsidence.
Less detrimental effects from surface rutting and scraping were measured in
the High Arctic.
         Table 6-4.  Change in conditions caused by seismic lines and
                     winter roads with paired control plots in 1970
                     (adapted from Bliss and Wein^).
Community type
Shrub-sedge-heath
hummock
Hummocks
Wet sedge meadows
Spruce-alder
Upland winter road
Year
disturbed
1965
1967
1969
1969
1969
Increased
active
layer (cm)
26
8
1
10
7
Increased
bare soil
(percent)
63
25
5
98
100
Decreased
plant cover
(percent)
65
34
7
169
99
     Revegetative studies are underway in several locations.  Bliss and Wein
reported trails with 16 species seeded in 1970.  At the rate of 20 kg per ha,
Festuca rubra, Poa pratensis, Poa compressa, and Phleum alpinum have the best
results in establishment and winter survival on mineral and peat soils.
Alopecurus pratensis grew better on mineral soil.  Fertilizer treatment with
100 kg per ha actual N and P^O^ doubled culm height over nitrogen fertilizer
alone.  The authors recommend sowing seeds with snowmelt in the spring or be-
fore snowfall in the autumn.
     Experiments with oil spillage on terrestrial vegetation are few.  Bliss
        3
and Wein  reported that oil poured directly on the foliage killed those leaves
in contact with the oil in a few days.  Some regrowth was evident in late sum-
mer from June treatment, but none from July treatment.  Greater damage was
                                     6-18

-------
evident in wet sedge and cotton grass tussock communities than in shrub com-
munities.  Microbial activity in the soil may increase tenfold, but the impli-
cations of this activity for the nutritional status of the community are not
yet established.  McCown and Brown15 applied oil to the soil surface (avoiding
foliage contact) and observed little immediate damage.  Shoot growth response
of Dupontia through the summer was not suppressed, except at the heaviest
treatment of 12 liters per square meter, but root biomass seems to have been
reduced.  Observations in the second growing season showed reductions in stems
per plot and in growth per stem.  Carex aquatilis and Eriophorum angustifolium
seemed  less sensitive than Dupontia.  To the extent that effects of oil spillage
are manifest through limited root growth, vegetative response may not be appar-
ent in  aboveground  tissue for one or more years.
      A  study of the effect of an air cushion vehicle  (SK-5)  on  tundra vegeta-
tion1'16 was measured quantitatively.  The  results  clearly  show that the exhaust
air blows  the  litter and  loosens organic matter,  but principal  damage was
caused  by  dragging  of the vehicle  skirt  on  the  surface.  Although  such  effects
change  the albedo,  the  direct visible  damage  by the air  cushion vehicle after
 25 passes  was  often less  than  that  of  a single  weasel  pass.
      In general,  any  disturbance  which eliminates or greatly reduces plant
 growth, vegetative cover,  or the  organic litter or peat  layer by any  cause will
 increase thaw depth as  a result of increased solar heat  flux.   Greater under-
 standing of response  to disturbance in different landscape units is needed for
 regulating land use and for providing design criteria for land vehicles.

 6.4  EFFECTS OF CLEANUP PROCEDURES ON TUNDRA
      The concept of the fragility of the tundra has probably been too uncriti-
 cally  accepted as  a convenient shibboleth by conservationists, in order to
 totally exclude resource development from the tundra.  Undoubtedly, wet tundra,
 which  may contain  frozen water up to five or six times in excess of dry soil
 weight within the  upper 40 meters, is fragile.  This problem is rare in alpine
 tundra.   Degree of disturbance potential could be mapped on a  regional scale
 according to water content, grain  size and vegetation physiognomy.  However,
 there  are other reasons behind the concept of  the  fragile  tundra that  relate
 specifically to plant  adaptation and  community structure and primary productivity
                                       6-19

-------
                                                                    2
considered both as daily rates and total season's production.  Bliss  has argued
that the very limited number of plant species in the Arctic is a real indica-
tion of fragility of the ecosystem since adverse impact on a very few species
may damage the whole:  the potential for recovery may be proportional to plant
                              o
community variability.  Dunbar  emphasizes that a proper consideration of
"fragility" hinges on one's definition of stability.  A steady-state situation,
which may pertain to arctic lakes and perhaps also to tundra vegetation, is
vulnerable to excessive disturbance since the system cannot adapt.  Another
type of stability is represented by a system which can absorb disturbance and
return to its previous state.  Tundra animal populations undergo drastic fluc-
tuations but recover, perhaps as a result of their ability to draw on outside
areas.  Dunbar concludes that simplicity of an ecosystem, coupled with slow
growth rates, leads to vulnerability to disturbance which can be tolerated only
if the spatial scale is large (as in the Arctic) and the time scale is suffi-
ciently long.

     (0-1)  Natural Rehabilitation
     Natural rehabilitation after a contamination of land in the tundra environ-
ment is suggested as a high priority treatment for cleanup.  Ecosystem recovery
tends to follow ecologically normal and predictable successional trends, i.e.,
patterns following damage from most postulated contaminants are similar to those
following fire, logging, or shipping of top soil and other natural catastrophic
              2
events.  Bliss  reported that nine years after a fire, grasses, sedges and forbs
recovered fairly well but lichens and mosses showed little recovery.  It is as-
sumed that in case of a contamination of an area in the tundra environment,
ecological recovery will follow a different pattern.  The area would still have
its litter and organic cover and will not be subjected to a deeper thaw during
summer.  The vegetation in a tundra ecosystem has a higher below-ground biomass
production and would provide better soil binding against water erosion during
the summer thaws.

     (1-1)  Chemical Stabilization
     The cleanup by chemical stabilization is to reduce the infiltration rate
and thereby reduce the mobility of contaminants into the vadose zone.  The
chemical stabilization will not be practical in the tundra environment.

                                     6-20

-------
     (1-2)   Clearcuttinq Vegetation
     Removal of woody shrubs by clearcutting can be used as a possible technique
for cleanup.  Along the Alaska pipeline route,  the tundra vegetation was removed
by clearcutting of woody plant species and shrubs.  In arctic and alpine envi-
ronments the plants are small, close to the ground, and often widely separated
by bare soil or rock.  Unlike the situation within a forest, the modification
of microclimate by vegetation is minimal and the physical environment dominates
the vegetation.  In such open and windy places, the effects of microenvironment
are pronounced.  Even a few centimeters difference in microtopography makes a
marked difference in soil temperature, depth of thaw, wind effects, snow drift-
ing and resultant protection to vegetation.  The microtopographic effects may
be caused by a rock, a soil polygon rim or by another plant itself.  The re-
moval of larger vegetation plants would result  in  an uneven distribution of
snow and thus differential distribution of soil moisture, heat exchange, and
protection  from windblast.
     It is  assumed  that the clearcut  areas will not be  disturbed as far as the
litter and  organic  horizons of  the  tundra soils,  and  that  the clearcut  areas
will be replanted  or reseeded.
     General  effects of clearcutting  are described in the  coniferous  forest
land type  (see Chapter 4).  Clearcutting of  the woody shrubs  in  a  small  area
within  the  tundra  ecosystem can have  extreme effects  on the microclimate and
the hydrological  factors,  since so much of  the  alpine or arctic  landscape  is
characterized by  scattered woody shrubs and forbs.  Removal of  the vegetation
will affect the snow and  soil moisture distribution due to microtopographical
changes  created by the removal of larger  shrubs.   Soil temperature within  the
clearing  will increase and there would probably be a deeper and earlier thaw
during  the  summer.   The effects of clearcutting on heat balance, soil moisture,
precipitation, runoff, and on wildlife have been described in the coniferous
 forest  land type  (Chapter 4).  These effects of clearcutting in tundra environ-
ment would be somewhat different in magnitude  but there are no data available
 on these effects  of clearcutting in a tundra environment.

      (1-3)  Stumping and Grubbing
      Removal of roots and stumps in  a tundra environment would cause considerable
 disturbances.  The practice would  create depressions and other microtopographical
                                       6-21

-------
 disturbances  in  the  treatment  area.  As mentioned  earlier,  the  effects  of micro-
 topographical  changes  in  a  tundra  environment  are  quite drastic and would af-
 fect  snow distribution, heat exchange, depth of  thaw,  soil  moisture,  soil
 temperature,  and surface  runoff.   Surface  litter and the organic layers  on  the
 surface would  also be  disturbed.
      Due to the  drastic nature of  disturbances such as increased potential
 hazards for soil  erosion  and sediment losses and deeper thawing and freezing
 that  would be  caused by a stumping and grubbing  operation,  the  practice  of
 vegetation removal by  stumping and grubbing should not be used  as a contaminant
 cleanup technique in the  tundra environment.

      (1-4)  Scraping and  Grading
      The surface  alteration through scraping and grading in the tundra environ-
ment  would require some prior  clearcutting and removal of stumps  from the treat-
ment  area.  Surface litter and some organic matter would be removed in the
process of scraping.   Grading  processes may even expose some mineral soil hori-
 zons.  The insulative  effects  of vegetation, surface litter and organic matter
would be destroyed and therefore the heatflow would not be  insulated.  Soil
thawing and freezing would be  relatively faster  and deeper.
      Infiltration rates would  decrease with an increase in  surface runoff and
water erosion.  The frequent freeze-thaw cycles  at the soil surface coupled
with the common presence  of permafrost not far beneath the  surface results  in
physical frost-thrusting  of soil, rocks and plants wherever water is abundant.
Surface scraping and grading would enhance such  soil frost  activity which
produces various kinds of polygons, stone nets,  and solifluction  terraces.
These disturbances would make  the edaphic environment unstable.    Such wide-
spread soil frost activity,  combined with the frequent presence of permafrost
not far below the soil surface, would have a significant effect on getting the
vegetation reestablished  in the treatment area.  Vegetational patterns in
arctic and alpine tundra are strongly shaped by  such soil frost activity.11
     Other effects of surface alteration in the tundra environment would be
similar to those described in the coniferous forest land type (see Chapter 4).
                                     6-22

-------
     (1-5).   Shallow Plowing
     Shallow plowing operations for contaminant cleanup would not be practical
for the tundra land type.

     (1-6)   Deep Plowing
     Deep plowing operations for contaminant cleanup would not be practical for
the tundra land type.

     (1-7)  Soil Cover Less Than 25 cm
     Top soil dressing the treatment area would cover up the  surface  litter and
organic matter horizons.  Ecological impacts of the  treatment would vary  accord-
ing to the depth of top  soil application.  The treated  areas  should be  reseeded
or replanted  for getting vegetation reestablished.
     Beneficial effects  of top soil dressing would be  in terms  of better  in-
sulation against heatflow, thus slowing soil thawing and freezing.   Infiltration
rates  would  also increase.   However,  lack  of vegetative cover would make  the
soil surface vulnerable  to the soil  losses due to wind and water erosion.

      (1-8)   Soil Cover to  25 to 100
      Same as for  treatment 1-7.

      (2-1)   Remove Plow Layer
      Removal of surface soil  (10 cm)  from a contamination area in the tundra
 environment would be similar  to scraping and grading practices.  Surface litter
 and organic matter  layers would be disturbed.  Ecological impacts of this opera-
 tion would be similar to those described under Scraping and  Grading  (1-4).

       (2-2)  Remove  Shallow Root lone
       Removal of shallow root  zone soil (<40 cm)  would  also be  similar  to the
 operations  for surface  alterations.  The  ecological disturbance  and  impact are
 very  similar to those described under  Scraping and  Grading  (1-4).
                                       6-23

-------
      (2-3)  Remove  Scraping  and Grading, Mechanically  Stabilize
      Surface  scraping  and  removal  of a shallow layer (<5 cm)  from a contami-
 nated area  in the tundra environment would be  impractical.   Tundra soils have
 thick layers  of  litter and organic horizons on the  surface.   Moreover,  the re-
 moval of  a  shallow  layer from  soil surface would  disturb the  area and would
 result in loss of insulative layer against heat flow.   The  operation would
 cause increased  frost  action and unstable  surface conditions.   Exposed  surface
 would not be  suitable  for  mechanical stabilization.

      (2-4)  Remove  Plow Layer, Mechanically Stabilize
      The  technique  would not be  applicable for the  same reasons described above
 (2-3).

      (2-5)  Remove  Shallow Root  Zone,, Mechanically Stabilize
      Would  not be practical.

      (2-6)  Remove  Scraping  and  Grading, Chemically  Stabilize
      Chemical stabilization  is not  practical in the  tundra  environment.

      (2-7)  Remove  Plow  Layer, Chemically  Stabilize
      Chemical stabilization  would  not be practical.

      (2-8)  Remove  Shallow Root  Zone, Chemically  Stabilize
      Chemical stabilization  would  not be practical.

      (3-2)  Barriers to Exclude People
      Exclusion of people from the  treatment area  would  not affect  the ecology
of tundra.

      (3-2)  Exclude Large Animals
     Exclusion of large herbivores  such as reindeer, caribou or musk oxen
would reduce local grazing pressure and result  in increased surface biomass,
increased insulation by organic material and a gradual  change in permafrost

                                     6-24

-------
depth.  Since large herbivores can cause a decrease in lichens and in increase
in woody shrubs, continued exclusion might lead to a shift in relative abundance
of some species, but otherwise would not affect the ecology of the area.
     Herding arctic animals follow set migration paths; however, unless the
exclusion fence blocked a natural barrier such as a mountain pass or river ford,
even the largest fenced area should have little effect on gross animal movement.

      (3-3)  Exclude Large and Small Animals
     Exclusion  of the small and large animals from the treatment area would
help  the process of natural restoration.  There would be no disturbances caused
by browsers and grazers.  Large animals like reindeer, caribou and musk oxen,
and small animals like  lemmings, have quite a grazing  impact  on the  ecology of
the tundra environment.  The other  impacts are due to  compaction, surface  soil
disturbances and physical damage to the vegetation.

      (4-1)  Asphalt Hard Surface Stabilization
      This treatment is  outside  the  scope  of work  of  this  study.   However,  ex-
perience  in  construction  of  arctic  roadways and air  strips  suggests  that  changes
 in  the insulation properties  of the surface and the  resultant change in the
 depth of  permafrost would be  the major  effect.  Lachenbruch has detailed  the
 procedures  for calculating the  proper thickness of gravel fill for  different
 surface materials  and climatic  conditions that will  minimize the effect on the
 permafrost  layer.

      (4-2)   Concrete  Hard Surface  Stabilization
      This treatment is outside  the scope of work of this study.

      (5-0)   Application of Sewage Sludge
      This treatment is outside the scope of work of this study.

       (6-1)   Hiah Pressure Washing
      This treatment is outside the scope of work of this study.
                                      6-25

-------
      (6-2)  Flooding  to  30 cm
      This treatment is outside  the  scope of work of  this  study.

      (7-0)  Soil Amendments Added
      This treatment is outside  the  scope of work of  this  study.

 6.5   RECOVERY AFTER CLEANUP
 6.5.1   Irreversible Changes
      The cleanup techniques described earlier for the tundra environment would
 not have irreversible impact on the ecology.  However, the harsh nature of the
 tundra  environment, relatively  short growing season  and low temperatures during
 the growing season would make the vegetative regeneration, ecological succession
 and animal recovery very slow processes.

 6.5.2   Rates of Recovery
     The sensitive nature of the tundra ecology to disturbances has led to the
 concept of "fragile" tundra.  Figure 6-1 illustrates the  series of events which
 follow  the damage of tundra vegetation.  The altered plant cover reduces both
 surface insulation and albedo, and more heat is absorbed  by the soil with the
 result  that there is a greater depth of thaw and both subsidence and erosion
may ensue.  Such disturbance will ultimately recover due  to the natural homeo-
 static mechanisms in ecosystems which bring about new equilibrium.  Unfortu-
nately  little is known about the time factor in these recoveries; it may range
from a few to hundreds of years.  In Figure 6-1, well drained sites where the
ground contains little ice show slight changes and recovery to a new, little
altered equilibrium is rapid.   However, on wet sites containing much ground ice
greater damage is caused, and recovery will be slow and may result in a different
vegetation when equilibrium is regained.
     One of the greatest concerns of ecologists is the degree to which tundra
vegetation can be disturbed and yet reestablish a turf within three to five
years prior to thermokarst development.   According to Brown et al.,  even the
replacement of mulch results in nearly as deep a thaw as does the removal of
total plant cover,  61  cm versus 75. cm,  respectively.   The control plots averaged
                                     6-26

-------
cr>
i
ro
SURFACE STABILIZATION
    AT NEW LEVEL
         NATURAL INCREASE IN PLANT COVER,
                INCREASED ALBEDO,
              REDUCED SUMMER THAW
               POTENTIAL SUBSIDENCE
                 ANDTHERMOKARST
                                                   NATURAL PLANT-COVERED SURFACE
                                                        SURFACE DISTURBANCE:
                                                           TRACTOR TRACKS
                                                            SEISMIC LINES
                                                                ETC.
                                                         LOSS OF PL ANT COVER
 REDUCED ALBEDO,
INCREASED THERMAL
      INPUT
                                           INCREASED DEPTH OF
                                              ACTIVE LAYER
                                              PERMAFROST
                                                                                  SURFACE
                                                                                 STABILIZED
                                                                              NATURAL INCREASE
                                                                               IN PLANT COVER
                                   LITTLE SUBSIDENCE
                                   AND THERMOKARST
             Figure 6-1.  A word model illustrating the sequence of events  in the recovery of damaged
                          arctic tundra vegetation  (from Bliss^).

-------
 32 cm for maximum  summer thaw.  The changed albedo is probably the major factor
 even though an insulative  layer was returned in the one treatment.
     The rates of  vegetative recovery and seedling establishment in the tundra
 environment are very slow, thus increasing the potential for permafrost melt
 and thermokarst development.

 6.5.3  Succession  Stages Following Cleanup
     6.5.3.1  First Year—
     During the first growing season after the cleanup the ecological succes-
 sion would depend  upon the nature of treatment and the degree of disturbances
 associated with it.  In the natural restoration areas there would be no signi-
 ficant disturbances in the surface litter and organic horizons.  Some of the
 woody shrubs which have quite an extensive and well insulated rooting system
 will start sprouting.  Grasses, sedges and forbs would also regenerate fairly
 well.  Lichens and mosses would not show much of regeneration.  Cleanup pro-
 cedures such as stumping and grubbing, surface alterations, and removal of sur-
 face soil layers would cause major surface disturbance and thus the treated
 area would be subjected to severe frost action.

     6.5.3.2  Fifth Year—
     In the natural regeneration area, and in the areas which were fenced out
 to keep the people or large and small animals out, the surface disturbances
 and frost action were minimum.  The vegetative recovery for grasses, woody
 shrubs and forbs should be quite normal.   Again the lichens and mosses would
 have a relatively  slow recovery.  The seeds from the nearby sources, and sprout-
 ing of shrubs and grasses,  which have an extensive rooting system,  would get
 the vegetative establishment started.   In the areas where surface litter and
 organic horizons were removed or disturbed,  there would be a considerable in-
 crease in the frost action and microtopographic changes.   Erosion rates would
be quite high.  There would be very poor vegetative recovery and succession.

     6.5.3.3  Tenth Year-
     Vegetative succession after ten years from cleanup treatment of an area in
the tundra environment  would have a near normal vegetative cover recovery

                                     6-28

-------
except for lichens and mosses.  This is based on an assumption that there were
no major disturbances to the surface litter and the organic horizons.   The
situation would be different in areas where the cleanup operation disturbed the
insulating layers of litter and organic matter and changed the albedo con-
siderably.  The vegetative establishment would have a very slow start in the
disturbed areas.  However, during the first ten years after the treatment, most
of the frost action induced disturbances and land subsidence would have already
taken place and natural or man induced revegetation would start taking place
on a new surface.

     6.5.3.4  Climax--
     It would be very difficult to evaluate the post cleanup treatment effects
on vegetative recovery and succession to a climax stage.  Differences in eco-
logical conditions after fifty or 100 years after different cleanup treatment
are not expected.  In terms of geological time, there would be no differences
due to different cleanup treatments reflected in the climax vegetation.

6.6  QUANTITATIVE ASSESSMENT OF CLEANUP IMPACTS
     The preceeding sections of this chapter have clearly shown that any cleanup
procedure  (except the benign exclusion by barrier) involves major perturbation
of the tundra biome that will result in long lasting environmental effects.
Quantitative assessment of the ecological differences among the various pro-
cedures is meaningless since any procedure which would  result  in removing  a
hazardous material will result in approximately the  same ecosystem desruption.

6.7  CONCLUSIONS
     Cleanup techniques result in changes  in albedo, in microclimate,  in  thermo-
equilibrium, and  in  the relationships  between unfrozen  ground  and permafrost.
To a  large  extent homeostasis  is maintained by  the mosses  and  lichens.   During
the short summer  growing  period when the  sun  is the  warmest,  there  is  an
abundance of standing melt water on the tundra.  Moss  has  a very  high  evapo-
transpiration  rate  that tends  to keep  the air  layers closest  to  the  ground
cool.   The lichens  prevent  the direct  rays  of  the  sun  from striking  the ground
surface.   The  interaction of cool air,  cool  water  and  no  direct  sunlight permits
                                      6-29

-------
only a minimum melting of the permafrost and leads to the formation of a shal-
low active area where plant growth takes place.
     If the cover material of the tundra is disturbed in any way the active
layer becomes thicker while the permafrost table is lowered.  The thicker
active layer collects additional water from surrounding areas.  This additional
water leads to formation of frost heave structures, thermokarsts and soil move-
ment through solifluction.
     All cleanup techniques alter the cover material and are therefore delete-
rious to the tundra.
                                     6-30

-------
6.8  TUNDRA REFERENCES


 1.   Abale, G.,  W.  H.  Parrott,  and D.  M.  Atwood.   Effects of SK-5 Air Cushion
     Vehicle Operations on Organic Terrain.   CRREL Report, 1972.  141 pp.

 2.   Bliss, L.  C.   Primary Production  within Arctic Tundra Ecosystems.  In:
     Productivity and Conservation in  Northern Circumpolar Lands, 77-84.
     IUNC.  Pub.  16. Merger, Switzerland,  1970.

 3.   Bliss, L.  C.,  and R.  W.  Wein.  Plant Community Response to Disturbances in
     the Western Canadian Arctic.  Can. J. Bot.,  50:1097-1109, 1972.

 4.   Brown, J.,  W.  E.  Richard,  and E.  Vietor.  The Effect of Disturbance on
     Permafrost Terrain.  CRREL Special Report 138, 1969.  13 pp.

 5.   Brown, J.,  and J. C.  F.  Tedrow.  Soils of the Northern Brooks Range, Alaska,
     4: Well-Drained Soils of the Glaciated Valleys.  Soil Sci. 97:187-195,  1964.

 6.   Cody,  W. J.  Reindeer Range Survey.   Plant Res. Inst., Canada Dept.  of
     Agric., Ottawa, 1964.  17 pp.

 7.   Dagon, R.  R.  Tundra - A Definition and Structural Description.  Polar
     Notes. 6:22-34, 1966.

 8.   Dunbar, M.  J.   Stability and Fragibility in Arctic Ecosystems.  Arctic,
     26:179-185, 1973.

 9.   Hill,  D. E., and J. C. F. Tedrow.  Weathering and Soil Formation in the
     Arctic Environment.  Amer. J. Sci.  259:84-101, 1965.

10.  Holowaychuk, N., J. H. Petro, H.  R.  Finney, R. S. Farnham, and P. L. Gersper.
     Soils of Ogotoruk Creek Watershed.  In: Environment of the Cape Thompson
     Region Alaska  (N. J. Wilimovsky,  and J. N. Wolfe, eds.),  1966. pp. 221-273.


11.  Johnson, P. L. and W. D. Billings.  The Alpine Vegetation of  the Beartooth
     Plateau in Relation to Cryopedogenic Process  and Patterns.  Ecological
     Monographs, 32:105-135, 1962.

12.  Klein, D.  R.   The  Impact of  Oil  Development in Alaska.   In: Productivity and
     Conservation  in Northern Circumpolar Lands, 209-242.  IUCN.  Pub.  16. Merges,
     Switzerland,  1970.

13.  McCown, B. H.  and J.  Brown.  Ecological  Effects of Oil Spills  and Seepages  in
     Cold-Dominated Environments.  In:  1971 Tundra  Biome  Symposium  Proceedings,
     61-65,  1971.
                                     6-31

-------
 14.  McKay, G. A., B. F. Findlay, and H. A. Thompson.  A Climatic Perspective of
     Tundra Areas. In: Productivity and Conservation  in Northern Circumpolar Lands,
     10-33.   IUCN Pub. 16. Merges, Switzerland,  1970.

 15.  Osburn,  W. S. Jr.  Radioecology.  In: Arctic and Alpine Environments  (J. D.
     Ives and R. G. Barry, Eds.), 875-903. Methuen, London, 1974.

 16.  Richard, W. E.  Ecological Evaluation of Air Cushion Vehicle Tests on Arctic
     Terrain.  CRREL Tech. Report, 1971.  30 pp.

 17.  Rickert, D. A. and J. C. F. Tedrow.  Pedological Investigations on Some
     Aeolian  Deposits of Northern Alaska.  Soil  Sci.  104:250-262, 1967.

 18.  Rieger,  S.  Dark Well-Drained Soils of Tundra Regions in Western Alaska.
     J. Soil  Sci. 17:264-273, 1966.

 19.  Sigafoos, R. S.   Frost Action as a Primary  Physical Factor in Tundra Plant
     Communities.  Ecology.  33:480-487, 1952.

 20.  Vuilleumir, F.  Insular Biogeography in Continental Regions.  The Northern
     Andes of South America.   Amer. Nat. 104:373-388, 1970.

 21.  Walker, J.  The Influence of Man on Vegetation at Churchill.  In:  Pro-
     ductivity and Conservation in Northern Circumpolar Lands, 266-269.  IUCN.
     Pub.  16.  Merges, Switzerland, 1970.

 22.  Watson, A., N. Bayfield, and S.  M.  Moyes.  Research on Human Pressures in
     Scottish Mountain Tundra, Soils,  and Animals.  In:  Productivity and Con-
     servation in Northern Circumpolar Lands,  256-265. IUCN.  Pub. 16. Morges,
     Switzerland, 1970.

23.  Webber, P.  J.   Tundra Primary Productivity.  In:  Arctic and Alpine Environ-
     ments (J. E. Ives and R. G. Barry,  eds.), 445-473,  Methuen, London, 1974.

24.  Whittaker,  R.  H.   Communities and Ecosystems.  Macmillan Publishing Company,
     New York, 1975.   385  pp.
                                   6-32

-------
                                 CHAPTER 7
                        COASTAL INTER-TIDAL MARSHES

7.1  OVERVIEW
     Heavy vegetational development takes place where the soil is waterlogged
or covered by shallow standing water for all or most of the year.  Where this
vegetation is dominated by grasses, reeds, sedges, and other non-woody types
the development is referred to as a marsh.  If the vegetation is largely
bushes and trees, it is called a swamp.  Coastal marshes and swamps represent
some of the highest zones of primary productivity in the world.  Some marshes
                                                            2
exhibit a net primary productivity of greater than 3,000 g/m /yr (Table 7-1).
In coastal marshes, however, less than 10 percent of the net production of the
ecosystem is consumed by grazing herbivores, mostly insects in this case, and
at least 90 percent follows the detritus path of energy flow.  The bulk of
the animals in the coastal marsh such as shellfish, snails, and small crabs,
seem to obtain their energy directly or indirectly from detritus.
     The five mechanisms and conditions that maintain this biological energy
flow at rates far superior to the adjacent ecosystems are:  (1) Tidal action
promotes a rapid circulation of nutrients and food, and aids in the rapid re-
moval of the waste products of metabolism.   (2) A diversity of plant species
and life forms provides a continuous photosynthetic carpet despite variable
physical conditions.  The three major  life forms of autotrophs that work to-
gether to maintain a high gross production rate are:   (a) phytoplankton; (b)
benthic microflora—algae living in and on mud, sand, rocks, or other hard
surfaces, and bodies or shells of animals;  (c)  large attached plants—the sea-
weeds, submerged eel grasses, and emergent marsh grasses.   (3) A coastal marsh
is often an efficient nutrient trap that  is  partly physical  (differences in
salinities cause vertical as well as horizontal stratification of water masses)
and partly biological.   (4) A year-round  primary production by a succession  of
"crops," even  in northern regions.   (5) Close contact  between  autotrophic and
heterotrophic  layers.
                                     7-1

-------
      Table  7-1.   Net  primary  production  and  plant  biomass  for major  ecosystems
                       and  for  the  earth's  surface  (from Whittakerl2).
NET PRIMARY
PRODUCTIVITY,
AREA*
10« knv


2
2
20
18
12
7
15
9
8
18
24
14
149
332
27
2
361
510
PER UNIT AREA t
1 dry g/m2/yr
normal
range
100-1,500
800-4,000
1,000-5,000
600-2,500
400-2,000
200-1,200
200-2,000
150-1,500
10-400
10-250
0-10
100-4,000

2-400
200-600
500-4,000




mean
500
2,000
2,000
1,300
800
600
700
500
140
70
3
650
730
125
350
2,000
155
320
WORLD NET
PRIMARY
PRODUCTION **
109 dry tons/yr


1.0
4.0
40.0
23.4
9.6
4.2
10.5
4.5
1.1
1.3
0.07
9.1
109.
41.5
9.5
4.0
55.
164.
BIOMASS PER
UNIT
dry
normal
range
0-0.1
3-50
6-80
6-200
6-40
2-20
0.2-15
0.2-5
0.1-3
0.1-4
0-0.2
0.4-12

0-0.005
0.001-0.04
0.04-4


AREA t
kg/m2

mean
0.02
12
45
30
20
6
4
1.5
0.6
0.7
0.02
1
12.5
0.003
0.01
1
0.009
3.6
WORLD
BIOMASS * *
109 dry tons


0.04
24
900
540
240
42
60
14
5
13
0.5
14
1,852.
1.0
0.3
2.0
3.3
1,855.
  Lake and  stream
  Swamp and  marsh
  Tropical forest
  Temperate forest
  Boreal forest
  Woodland and shrubland
  Savanna
  Temperate grassland
  Tundra and alpine
  Desert scrub
  Extreme desert, rock and ice
  Agricultural land
  Total land
  Open ocean
  Continental shelf
  Attached algae and estuaries
  Total ocean
  Total for  earth

^ Square kilometers X 0.3861 = square miles.
t Grams per square meter X 0.01  = t/ha, X 0.1 = dz/ha or m eentn/ha (metric centers, 100 kg, per  hectare, 104 square meters),  X 10 = kg/ha,
I* 8.92 = Ibs/acre.
j** Metric tons (106 g) X 1.1023 = English short tons.
j t Kilograms per square meter X 100 = dz/ha, X 10 = t/ha, X 8922 = Ibs/acre, X 4.461 = English short tons per acre.

-------
     Coastal marshes develop on relatively flat terrain between the limits of
normal high and low tide of protected bays, estuaries,  and lagoons.  They are
dominated by a few species of tall emergent reed-like vegetation which are
tolerant of rhythmic submergence and saline conditions.  The water may be nearly
fresh or nearly marine or highly variable in salinity,  and flushing occurs with
each tidal cycle.  Within saltwater marshes there is a gradual build-up of organic
peat deposited by the vegetation itself.  Highly dendritic tidal creeks dissect
the marshes and serve as avenues of entrance and egress of the tidal waters
which alternately flood and drain the marshes.  Because of their adaptations to
the intertidal zone, salt marshes are highly sensitive to even minor change
in water levels.
     The coastal marsh ecosystem lies from a few feet above sea level to a
few feet below it.  Hence, it is subject to the ebb and flow sweeping action of
tidal currents, and all must be tolerant of some salinity change.  All trap
suspended nutrients by slowing down the water currents, and they all provide
shelter and food for a variety of small brackish water and marine animals.  These
are among the most productive ecosystems of the world with annual production
rates running around five tons per acre.  Much of this plant production becomes
available as organic detritus which provides the chief food base for the coastal
fish and shellfish populations of commercial importance.  Without these important
production and nursery areas, our coastal seafood resources would suffer severe
decline.
     Although tolerant of short-term inundations with fresher or more saline
waters and even short-term exposure to the air, these systems cannot tolerate
long-term changes in these environmental factors.  Drying of the habitat or
major intrusion of fresh- or saltwater has been shown to change the composi-
tion of the dominant vegetation with long-term erosion of the productivity of
these systems.
     The marsh vegetation is dominated by several species of the tall Spartina
grass and to lesser extents by other emergent species such as Distichlis, Juncus,
and Salicornia.  Around the bases of these plants and on the surfaces of old
leaves grow a variety of filamentous algae including blue-green, brown, and red
algal types.  On the mud flats between the bases of the plants grow a variety of
                                    7-3

-------
 ECOLOGICAL

  SYSTEM
                             ESTUARINE  SYSTEM
ECOLOGICAL

SUBSYSTEM
         Subtidal
                     Inter tidal
             (nonvege fated)      (vegetated
                                   (nonvegetated)
                                         (vegetated)
CLASS
Bottom  Reef
SUBCLASS
                            Submergent
                              Bed
Floating
 Leaved
                                   Reef    Flat   Beach/Bar  Rocky  Emergent  Shrub   Forested
                                                         Shore  Wetland   Wetland  Wetland
                                                           I        I
                                                           I        I
                                                      O  n Z1
                                                                                 I      «/>   3   5    ™

                                                                                1     UJU
ORDER
                                                                  |o    292920  ?o
                                                                  3 *£    <» *£  S'fi  «*£  <> •£
                                                                  ij_    *J«  &j_.i?  ti?
         Figure 7-1.  Diagram of the wetland classification hierarchy for the
                      estuarine ecological  system  to  the order  level  (from U.S
                      Department of Interior9).

-------
diatoms and blue-green algae.  Plant production of the marsh is dependent upon
all three groups of producers, the tall emergent species, the filamentous attached
forms, and the mud flat inhabitants.
     Only grasshoppers and a few birds such as seaside sparrows may be found among
the tall Spartina grass, but the water and mud flats are teeming with animal
life.  This includes snails, mussels, and oysters, as well as a variety of worms,
crabs, shrimp, and small fishes.  Many of the crabs, shrimp, and fishes are juve-
niles of species which support the commercial catch as adults.  Large numbers
of shore and wading birds forage in these marshes at low tide.
     Studies have shown that there is a regular export of decomposing organic
matter through the tidal creeks to the estuaries, but the major export occurs
when storms inundate the marshes with high water and flush out great quantities
of organic matter to the estuaries and the continental shelves.

7.1.1    Occurrence of  Coastal  Marshes
     Coastal marshes (or intertidal emergent wetlands) are recognized as an eco-
logical class within the Estuarine System of a wetland classification (Figure
7-1).  There are nearly 70 million acres of natural wetlands in the continental
United States of which coastal marshes total about  17.5 million acres.

7.1.2    Related Land Types
     Coastal marshes as an ecosystem type are very  similar to coastal swamps
and  grass flats.  All  are emergent wetland ecosystems  (Figure 7-1).
     A coastal  swamp  is an  association of mangrove  trees  which  grows  in  the  sea
or in the water of sheltered bays and  estuaries  of  subtropical  regions.  The
average  water depth may be  from a few  inches  to  about  four  feet.   Extensive  prop
root systems  reduce the action of waves  and tidal currents  and  produce a charac-
teristic set  of internal  environmental conditions.   Water flow  is  greatly  reduced.
Sedimentation  is  high,  and  much organic  matter  accumulates.   Oxygen  levels are
 low, and the  environment  is in many respects  similar to  that of anaerobic  sub-
merged  soils.   Strong gradients in  oxygen  content and other factors  exist  from
the periphery to  the  interior of  the swamp.   In some cases,  saltwater swamps
 grade inland  into saltwater marshes.
                                      7-5

-------
      Coastal swamps are dominated by the low, bush-like red. black, and white
 mangrove trees.  A few other shrubs and vines may also be present.  The exten-
 sive root systems developed by the mangroves provide surfaces for attachment of
 filamentous algae, and the surface muds may support large and productive diatom
 floras.  Large numbers of oysters are often found attached to the mangrove roots;
 a variety of small crabs, shrimp, and fishes feed on the organic forage around
 the roots and mud flats at low tide; and birds nest in the branches of the man-
 groves.

      Grass flats or submarine meadows consist of a few species of grasses which
 are tolerant of continual submergence in salt and brackish waters.   They are
 normally found from the low water line to a depth of about three feet,  but they
 may extend considerably deeper in very clear waters.   Although seldom found in
 very strong current,  they are most luxuriant where there  is moderate flushing.
 The long flat  blades  of the dense beds protect  the bottom from erosion,  and
 extensive deposition  creates  a substratum of finely particulate,  high organic
 muds.   Sufficient  water penetrates the beds to  maintain high oxygen levels in
 the water above  the bottom.
      The grass  flats  are dominated by eelgrass  (Zostera spp)  in  northern  lati-
 tudes  or by turtle grass (Thalassia spp)  or manatee  grass  (Cymodocea spp)  in
 the more tropical  areas.   The long grass  blades  are  often  clothed with a  layer
 of  attached  filamentous algae which  produce organic matter  and which also  act
 as  brushes  to remove  suspended matter  from  the  flowing water  above.   Many  small
 animals  live among the  stems  and  roots of the grass beds, and larger fishes and
 birds  forage there.

 7.2    NATURAL PERTURBATIONS
     Coastal marshes are generally considered to represent a  serai stage in the
 succession of water to dry land.   In areas  with relatively stable climatic
 conditions,  succession  is extremely slow and the marsh maintains the same appearance
 year after year.   In other areas, violent climatic fluctuations cause coastal
marshes to revert to an open water phase in some years and the concept of succes-
 sion in the normal sense of the word has little meaning.   Fire has been shown to
play an important role in maintaining coastal marshes.   Coastal marshes can
recover from fires in a couple of years if the fire does  not burn down to the
bare soil and thereby permit excessive erosion.
                                    7-6

-------
     Large storm systems (e.g.,  hurricanes)  can cause extensive damage to coas-
tal wetlands due to erosion and excessive freshwater runoff.   Erosion physically
disrupts the marsh while the freshwater inflow may temporarily upset biota
adapted to marine or estuarine conditions.  These effects are generally short
term and marsh recovery would be expected within a few years.  Studies of natu-
ral perturbation show effects on primary and secondary productivity but these
perturbations are well within the range of homeostatis within the marsh ecosystem.

7.3   MAN-MADE PERTURBATIONS

7.3.1  Impact on Baseline Ecosystem Components
     The coastal marsh is a semi-aquatic system in equilibrium with the prevail-
ing climatic, hydrographic, geological, and biological forces of the coast.
Even slight modification in the level of the water table or the rate of surface
freshwater flow greatly modifies the biological characteristics of the system.
Although the coastal marshes vary greatly in detail, a more or less typical
marsh has freshwater vegetation at the landward side, saltwater vegetation at
the marsh is drained by highly dendritic tidal creeks which empty into the bay
or estuary.  Freshwater entering along the upper edges of the marsh drain across
the  surface  and  enter  the tidal creeks.
     Many of the marshes of  the Atlantic  and Gulf  coasts have  undergone  great
attrition  in recent  years, primarily as  a result of  levee and  canal  construction.
                                                                                 **
A levee  placed  across  the upper end of a coastal marsh has the following  primary
effects:
      -  cuts off all  distributaries feeding the marsh
      —  prevents freshwater  flooding
      —  prevents annual flushing
      -  prevents annual renewal  of sediments  and nutrients
      -  ends formation of  new marshes.
      Canals which lace the coastal marshes  for navigation, pipelines,  or mosquito
 control have the following primary effects:
      —  intercept and carry off freshwater  drainage
      -  block freshwater from flowing  across  the portion of  the marsh that
         is seaward of the  first canal
      -  rapidly carry off  freshwater to the bay or estuary

                                      7-7

-------
      -  lower the water table
      -  permit saltwater intrusion well into the marsh proper.

  7.3.2   Unassisted Recovery Sequence
      The biological consequences are clear.  On the Atlantic coast where sub-
 sidence rates are fairly slow, the marsh vegetation gives way to dry land vege-
 tation with accompanying changes in the animal populations.  Bourn and Cottam2
 reported on a detailed 10-year study of the fate of a drained coastal marsh in
 Delaware.   Prior to canalization,  90 percent of the marsh vegetation was salt-
 marsh cordgrass (Spartina alterniflora) with smaller amounts of other marsh
 species, especially at higher  elevations.   A few open water areas supported
 luxuriant  growths of  widgeongrass  (Ruppia  maritima)  and other submerged aquatic
 species.

      Ten years after  ditching  had  taken place the wetland plants had been reduced
 to small groups in the remaining low spots and  along canal margins.   Groundsel-
 bush  (Baccharis halimifolia) dominated  the plant community which now was made up
 largely  of dry land species  such as  asters,  goldenrods,  terrestrial  grasses,  and
 young  trees (pine,  juniper,  sweetgum, maple,  and hawthorn).   Aquatic animal
 populations of the  ditched areas had  been  greatly reduced in  areal extent and in
 density, even  in  the  wetland habitat  which still  remained (Table 7-2).   The
 density  of the total  invertebrate population  was  reduced  from 39 to  97  percent
 in.the various samples, and the  mollusks and  crustaceans,  which  make up important
 food items  for many fishes and shore birds, were  reduced  32 to 100 percent.   Open
 aquatic  areas, which  formerly  supported widgeongrass and  other important  duck foods,
 had been reduced to mud flats and dry land.   Thus, the wetland habitat,  important
 in the production of  fishes, shellfishes,  ducks,  and wading birds, had  given way
 to land  with its low  wildlife values.

     On  a  subsiding coast, such  as occurs  in  southern Louisiana,  elimination  of
 the normal  freshwater  and sediment input upsets the  land-water equilibrium, and
 the subsiding  marsh tends to become an open water area.   This tendency  is intensi-
 fied by  canals which  drain the marshes, enhancing compaction.  These canals tend
 to grow  wider  as a result of marginal subsidence, wave erosion,  and  disturbance
 from boat traffic.  Plant production by marsh grasses of  the Gulf coast is very
 high,  exceeding 10 tons per acre per year and a great deal of additional plant
production occurs in the marshes due to attached algae, mud flat diatoms, and

                                      7-8

-------
  Table 7-2.  Effects of ditching a Delaware tidewater marsh on the aquatic
              invertebrate populations.  Vegetational zones are characterized
              by the dominant plant species.  Six-foot square quadrants were
              sampled for comparison of invertebrate density in the drained
              and undrained sections of the marsh, and they represent three
              consecutive years of sampling during the months of April-December
              (from U.S. Environmental Protection Agency11).
                                                     Percent reduction of the
                                                     invertebrate populations
                                 Feet above mean
                                sea level (for the      Total       Mollusks and
  Vegetation Zones              undisturbed marsh)  invertebrates   crustaceans
Saltmarsh cordgrass
(Spartina alterniflora)
Saltgrass
(Distich! is spicata)
Saltmeadow cordgrass
(Spartina patens)
Saltmarsh bulrush
(Scirpus robestus)

1.88-2.93
2.35-2.90
2.58-3.32
2.75
39-82
64-88
41-97
50-97
32-95
82-94
55-100
58-98
phytoplankton in the shallow waters.   A large fraction of this organic matter
is exported through tidal creeks to nearby bays and estuaries.  When the marsh
becomes an open water area, however,  production is apparently reduced, and
instead of exporting organic matter,  the area becomes a nutrient sink.  Birds
and mammals no longer find food and refuge among the marsh grasses, and canals
create migrational barriers to terrestrial and semi-terrestrial animals which
utilize the marsh.  Complete shifts in vegetation accompany increased salinity
and subsidence.

     Saltwater intrusion increases the salinity of the marshes, eliminating
the broad mixing zone so important as nursery grounds for juvenile fishes,
shrimp, and crabs.  In deeper channels where reducing conditions prevail,
large quantities of -hydrogen sulfide are produced which are toxic to the marsh
       O
grasses  and to the aquatic animals.   Acid conditions of the canals may also
result in release of heavy metals from the sediments.  As a result of habitat
loss, decreased food supply, increased salinity, and increased hydrogen sulfide,
                                      7-9

-------
 populations of aquatic animals are adversely affected.  Moore and Trent7 compared
 oyster production in natural and canalized marshes.  They found that in the
 altered marsh the set of young oysters was reduced by over 90 percent, juvenile
 growth was slowed, average length was reduced by 36 percent, weight was reduced
 by 27 percent, and mortality was increased by 39 percent.  As a result of these
 and other studies, St. Amant (unpublished) has concluded that lack of freshwater
 has drastically modified the ecology of coastal marshes and severely damaged
 production of valuable oysters,  shrimp, fur animals,  and waterfowl.
      In areas of eroding shorelines where dredged spoil has been used to sup-
 port vegetation,  both seeding and transplanting methods have been used to re-
 develop the flora.  Both methods have been highly successful and after two grow-
 ing seasons,  little or no difference existed in appearance and primary produc-
 tivity between these new marshes and long established natural marshes.
      The methods  of soil removal are dependent on the depth and characteristics
 of the unstable material as  well as upon  the nature of the underlying stable
 substrate (Figure 7-2).   Where depths of  unstable material do not exceed about 10
 feet a dragline,  shovel,  or  dredge  may be used to excavate the unstable material
 to sufficient  width.   The excavated material handled  by dragline or shovel  is
 usually side  cast beyond the  excavation,  forming  a spoil  bank,  and it must  be
 removed a sufficient  distance to prevent  lateral  displacement back into the
 excavation.
      Where deeper unstable deposits are encountered (about 10 to 25 feet),  dis-
 placement methods are commonly used.   In  such methods  the fill  is advanced  by
 end-dumping and placement with a bulldozer in a V-shape (i.e.,  highest  along  the
 central  crest).   Fill height  is  increased  until the load  is  sufficient  to produce
 failure  in the underlying unstable  materials,  displacing  them laterally.  Dis-
 placement  may be  accelerated  by  jetting with water  prior  to,  during,  and after
 displacement of the fill.  As the fill settles, additional material must be
 placed to maintain the grade.  The weight of the fill  will cause lateral compres-
 sion in the displaced materials which may result in settling and horizontal
movement of the shoulders for several years.
     Displacement of unstable materials under the fill may also be accomplished
by blasting.  In the "underfill"  technique the surface layers are broken with
equipment or light charges,  the  fill is placed, and explosives are then positioned
through the fill in jet holes or  casings.   Usually one to three rows of explosives

                                     7-10

-------
       MOTIVATION
                    DREDGING PROCESS
Figure 7-2.  A typical  effects  linkage shows the disturbance web and environmental  impacts
             of soil  removal  (from U.S. Department of  Interior^0).

-------
are placed along the center line about midway between the fill bottom and the top
of the underlying solid substratum.  In addition, at each edge of the fill two
or three rows of charges are placed 4 to 5 feet below the surface.  The explo-
sion displaces the soft materials, creating a cavity under the fill which then
settles rapidly.  In the "toe-shooting" technique the soft material is displaced
by blasting ahead of the advancing fill.  Added fill material is pushed into the
cavity left by the blast, and the fill is advanced with a V-point which displaces
the soft material and develops a wave in front of the fill.  In the process, the
front face of the fill is overburdened, and the explosive charges are placed
around the toes of the fill near the bottom of the soft material.
     The cutting and digging action of the dredging operation breaks through
the thin oxidized layer of the submerged soil and exposes the deeper unoxidized
layer.  Furthermore, most of the sediments placed in suspension are removed from
this layer and, hence, are in the chemically reduced state.  Such materials have
very high chemical and biological oxygen demands.  Frankenberg and Westerfield
calculated that some dredge spoils require 535 times their own volume of oxygen
for complete oxidation, and Brown and Clark  reported oxygen levels near dredges
18-83 percent below normal.  Both the sedimentary particles and the interstitial
waters released contain immediately toxic materials such as hydrogen sulfide,
methane, and a variety of organic acids, ketones, aldehydes, etc., as well as
heavy metals and pesticides which exhibit persistent toxic effects.  Turbidity,
per se, reduces light penetration and interferes with photosynthetic production
of oxygen, and it tends to elevate water temperatures.  Eventually the suspended
material settles to the bottom either near the dredging site or far downstream.
Thus, there is a redistribution of sediments together with whatever nutrients
and chemical pollutants which they may contain, and this may result in modified
bottom topography and altered patterns of water circulation.  Such sedimentation
problems are greatly accentuated when dredge spoil is placed back into the water.
General and immediate effects of dredging and spoil placement are listed in
Table 7-3.
     In coastal marshes the matter is complex because of the adjacency of salt-
water and the natural process of coastal subsidence.  Coastal marshes, in the
undisturbed state, represent vast drainage systems.  Freshwater enters on one
side, generally from the floodwaters of annual river overflow, and it gradually
works down to the coast in a broad, flat surface sheet.  Occasional creeks and
bayous aid in the runoff.  Near the estuaries the marshes are dissected by

                                     7-12

-------
Table 7-3.  Effects of denuding and/or soil removal on physical and
            chemical characteristics of marshlands (modified from
            U.S. Environmental Protection Agency!!).
Physical and chemical effects
Loss of natural vegetation
Loss of topsoil
Increased surface runoff
Lowering of water table
Increased erosion
Leaching of soil minerals
Violent fluctuation in stream flow
Violent fluctuation in water levels
Increased downstream flooding
Increased sediment load
Increased bottom sedimentation
Loss of wetland habitat
Reduction in habitat diversity
Increased turbidity
Changes in water temperature
Changes in pH
Changes in chemical composition
Increased oxygen demand
Denuding and soil removal
Clearing
and
grubbing
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Soil
removal
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Rock
excavation









X
X
X
X
X
X
X
X
X
Subgrade
stabilization


X



X
X








X

Long
term
effects
X
X
X
X


X
X
X


X
X



X


-------
dendritic tidal creeks.  Thus, through the marsh there is naturally a gradual
salinity gradient from freshwater to the more brackish waters of the estuary.
Marshes overlie deep layers of unconsolidated river deposits which gradually
undergo compaction as the water is squeezed out.  Such marshes would gradually
subside, but sediment input through river overflow and buildup of organic matter
through plant growth counteracts this tendency and maintains the delicate land-
sea, and freshwater-saltwater balances.
     Soil removal (including denuding) from marshes accelerates the rate of
freshwater runoff, and it may lower the water table of the soil, drying out the
higher areas of the marsh.  Artificially denuded areas do not correspond to nat-
ural coastal meandering tidal streams.  They may, however, erode to form open
canals.  Once opened, such canals tend to widen due to tidal and other natural
action or due to the effects of boat traffic.  Land loss from canal erosion has
reached serious proportions in Louisiana and elsewhere.
     In addition to draining away the freshwater, the canals offer paths for
saltwater penetration of the marshlands, and this is especially prominent in
the deeper canals.  Since rivers no longer are permitted to flood the upper
reaches of the marshes, they are now deprived of both the annual freshwater
and the annual sediment load.  Thus, as compaction and subsidence proceed, and
as saltwater penetrates through the canals, the effect of saltwater is being
felt further and further inland.  Vehicular traffic over the marshlands (mud-
boats, marsh buggies, and heavy equipment) associated with restoration activities
accentuates this problem.

     Marsh canals have very high contents of organic matter and high oxygen
demands.  Yet water  circulation is often poor, and this leads to reducing or
near reducing conditions, especially  in the bottom water.  Saltwater is rich in
sulfates, and when the sulfates enter  the reducing conditions, they are con-
verted to sulfides,  which are very potent biotoxins.  Precipitated iron sulfide
is a common marsh deposit.
     Spoil banks are often cast up alongside the canals creating a surface dam
effect.  Such banks  impound waters on  both sides and seriously interfere with
normal surface drainage patterns.  The  spoil banks directly cover vast acreages
of marshland, and erosion from the spoil banks tends to drain back into the
canal, on one side,  and into  the marshland, on the other.  Since the sediment

                                     7-14

-------
 itself  is mostly  in the chemically reduced  state,  it  tends  to  lower  the  oxygen
 concentration of  the  canal waters when  it flows back.   Erosion of  spoil  banks
 and  shallowing  of canals require redredging in a never-ending  cycle.   Effects
 of soil  removal and spoil placement  in  marshlands  are  listed in Table  7-4.
          Table 7-4.  Effects of soil removal and placement of dredge
                      spoil in marshlands (modified from U.S. Environ-
                      mental Protection Agencyll).
              Interference with surface drainage patterns
                  Acceleration of surface drainage by canals
                  Damming of surface drainage by spoil banks
                  General acceleration of freshwater runoff

              Loss of marshland habitat
                  Loss due to canalization
                  Loss due to water table lowering
                  Loss due to erosion and widening of canals
                  Loss due to spoil coverage
                  Loss due to acceleration of marsh subsidence

              Acceleration of saltwater penetration

              Conversion of sulfates (of saltwater) to sulfides in the
                  canals and precipitation of iron sulfide in the
                  canals

              Erosion of spoil banks and distribution of chemically
                  reduced sediment into canals and open marsh
7.4  EFFECTS OF CLEAN-UP PROCEDURES
     ON COASTAL MARSHES

    Clean-up procedures considered for marsh land types include temporary con-
tainment, contaminant removal and barrier exclusion.  The following subsections
discuss the effects of these treatments on the marsh ecosystem.
    The first techniques considered are those which may be applicable for pre-
venting further spread of contaminants across the marsh until contaminant removal
                                     7-15

-------
 by clean-up can be accomplished.  The various alternative methods are considered
 to be appropriate for coastal marshes if they might prevent further redistri-
 bution.  The techniques discussed in this subsection include chemical stabili-
 zation, clear cutting, stumping and grubbing, plowing, and application of un-
 contaminated soil cover.
     (1-1)   Chemical Stabilization
     Chemical stabilizers  are materials employed to bind soil particles together
 as an aid to preventing erosion.  Chemicals may also be applied to mulch mater-
 ials to act as a binding  agent.
     The effect of chemical stabilizers is usually temporary and in the water-
 logged soils of the marsh environment may be very transitory.   Because of this
 characteristic, chemical  stabilization may not be an appropriate treatment for
 large areas.   However, for small plots,  this approach may be of some use.   Since
 chemical application is commonly a part of revegetation programs,  little eco-
 logical impact would be expected from application in marsh areas.
     Certain chemical stabilizers are designed to reduce the rate of infiltration.
 In a coastal  marsh environment, with the frequent flooding associated with high
 tides,  infiltration reduction has little meaning.   Use of this  treatment is
 therefore  inappropriate for this land type.
     (1-2)   Clear Cutting  Vegetation
     Clear  cutting would be an appropriate treatment for relatively upland  por-
 tions of coastal  marshes  which support significant populations  of  trees  and
 to large emergent stands  such as found in mangrove swamps.   Since  root systems
 would not  be  removed,  erosion potential  will  clearly  be  appreciably diminished
 relative to root  layer removal  techniques  discussed in  subsequent  sections.
 Root  and/or stump sprouting  is  likely  to  occur and  recovery would  thus be ac-
 celerated  relative  to  total  clearing treatments.

      (1-2)  Stumping and Grubbing
     Stumping and grubbing involves the removal of all surface vegetation and
major root systems.  Disposal of vegetation may be by natural decomposition or
by burning, depending on the volume of organic material and the properties of
the contaminant.
                                     7-16

-------
     The clearing of the  land removes the vegetative cover and permits the rain-
 fall to strike the bare land surface.  These processes lead to increased surface
 runoff and severe erosion.  The effects will be accentuated in rainy weather.
 In dry weather considerable quantities of soil in relatively upland areas may be
 raised as dust clods which will be transported at a later date when the rains
 fall.  Runoff and erosion will add a great deal of soil solids to the marsh re-
 sulting in greater water  turbidity and increased sedimentation in tidal creeks.

    Denuded areas have been shown to lose large quantities of dissolved min-
 erals,  particularly sodium, potassium, calcium, magnesium, nitrates, and phos-
phates.  In some cases increased groundwater and springflow has been noted
 immediately following removal of phreatophytes (which normally pump water up
 through the roots to be lost via transpiration),  but the springflow may even-
tually diminish as the water table is lowered through lack of recharge.
    If the brush cleared from the land is burned in the flood plain the ashes,
which are highly alkaline, may enter tidal creeks and cause an immediate in-
crease  in the pH of the water (in one study the pH jumped almost immediately
from 7.8 to 11.3 and remained high for some time).   In addition,  heat from
the fire can elevate the marsh water temperature  quickly and keep it high for
some hours.
    Denuding equipment in operation as well as spills in maintenance yards can
result  in the passage of petroleum products into  the water courses.
    The net result may be summarized in the following points:
        — loss of habitat from devegetation of the  area
        — loss of land fertility from surface erosion and subsurface flow
        — increased erosion from denuding site activities
        — lowered groundwater level from devegetation
        — greatly increased fluctuation in water  level  due to  faster runoff
          following rains and decreased flow during dry periods because of
          loss of groundwater
        — greatly increased marsh sediment load due to  erosion and
          runoff
        — greatly increased marsh turbidity due to  erosion and runoff
                                     7-17

-------
        — modified chemical composition of the water due to increased
          sedimentation and runoff, turbidity, leaching of soil
          nutrients.
    As a result of these influences, the coastal marsh will undergo a number
of changes.  The violent fluctuation in water level will result in greater
freshwater flow rate during wet weather.  The creek beds may be cut deeper,
the banks will be undercut, and the open water sections will be widened.  Rif-
fles may disappear and pool areas fill.  Marsh areas may be swept clear by
flood waters.
    Increased sediment loads clog the interstices of riffles, fill the pools,
and cover the bottom generally with a layer of inorganic silt.  Bottom habitat
diversity would be significantly diminished.  Accompanying the increased run-
off, there would be an increase in water turbidity.  This lowers the light
penetration of the water, increases oxygen demand, and modifies the chemical
characteristics of the water in other ways.  Loss of vegetative cover and in-
crease in turbidity both serve to elevate the temperature of the water (as
much as 10°F).5
    During dry weather freshwater flow may slack off or it may cease entirely,
since the marsh now receives less groundwater inflow than before.   As pointed
out by Bayly and Williams,   land clearing may so alter the local hydrological
regime that formerly perennial streams may approach or become intermittent.
Since open water pools tend to be reduced or lost, the aquatic habitat may be-
come severely restricted or dried up between floods.  Any water that remains is
                                                          •»•
now subjected to more rapid and estreme temperature fluctuation in response to
prevailing atmospheric conditions.
    The long-term results will depend greatly upon local circumstances but,
in general,  they would include the following:
        — permanent loss of natural land habitat — If the topsoil  is
          eroded away,  a native ecosystem will not return for many years
        — increased surface runoff and reduced groundwater flow — The
          denuded surfaces  will continue to yield rapid and complete runoff
       — persistent  chemical  changes — The  high  level  of  sedimentation
         and  turbidity  may eventually  taper  off.
                                     7-18

-------
    Recovery after stumping and grubbing would be dependent upon several fac-
tors including
        — size of area denuded (the magnitude of the effects outlined
          above would be increased with increasing area)
        — location of disturbed area relative to open water (areas
          adjacent to open water would be subject to runoff effects
          as well as the erosional influence of tidal currents and
          wave action)
        — weather following cleanup (heavy rains would magnify runoff
          impacts and extended dry periods would delay recovery by
          possibly drying exposed muds).

    (1-4)   Scraping and Grading
    Surface alteration such as scraping and grading of a marsh area would
have environmental consequences similar to those discussed above for vegeta-
tion removal.  However, more complete removal of the surface layer organic
matter would result with this type of treatment.  In addition, epibenthic
and burrowing animals would be more directly affected by surface alteration.
    Because of the relatively flat characteristics of marsh lands, grading
for containment would not entail leveling as may be the case in more upland
ecosystems.  Grading in coastal marshlands would probably include banking
of materials so that drainage patterns would be from the edges of the af-
fected area inward (the reverse of the normal drainage pattern).  This could
alter the fresh - salt water gradients within the marsh and ultimately modify
the biological character of the marsh.  However, the slight gradients in-
volved here would undoubtedly be overcome by tidal and wave action and flood
events and eventually more-or-less normal gradients would prevail and con-
tainment would be temporary.

    (1-5)   Shallow Plowing
    Plowing is not felt to be an appropriate treatment for a coastal marsh.
Bottom sediments in marshes are continually in a state of flux in response
to tides, waves and runoff events.  Plowing would be not only destructive to
the marsh habitat but would also probably enhance contaminant mobility by dis-
rupting the hydrologic regime and creating new flow routes.
                                     7-19

-------
       (1-6)    Deep Plowing
       See treatment  1-5; this technique  is not appropriate  for this  land type.

       (1-7)    Soil Cover Less than  25  om
       Because of the scouring action of  tides, waves and flood events, thin soil
  cover may be only a very temporary containment treatment  for this  land type.
  Thus its use in marsh systems may be limited.
      The effects of this treatment are in two areas-the area covered and the
  source of the cover material.   Naturally the magnitude of effects, and hence
  the restrictions to recovery are a direct function of the area treated.
      In the covered  area,  the habitat  of epibenthic, burrowing and semi-
  terrestrial  animals would  be destroyed.  Also young stands of marsh grasses
  would be buried and would  not  be expected to survive.   Larger grasses, bushes
  and trees  would not be buried;  however  machinery used  to  dump the cover  soil
  may be very  destructive.

      Recovery would  be  more  rapid than if  the covering  process was preceded
  by  vegetation removal, as root  or  stem  sprouting  may not be prevented with the
  shallow cover.  Vegetation  recovery would be fairly rapid  with  this treatment
  due to  the shallowness of the cover which would not bury larger plants and
  which would  probably soon erode.

      (1-8)   Soil Cover 25 to 100 cm
      If  the uncontaminated soil  fill is  up to  100  cm in depth, then  the marsh
has  simply been filled  in and it  no longer can be  referred  to as a coastal marsh
The  addition of 100 cm of soil may cause the  land  level to  excbed  the tidal
ranges and, consequently, the saltwater  intrusion  which is  characteristic of a
coastal marsh will cease to occur.  The marsh may:   (1)  cease to have a hydric
soil and begin to behave as a predominantly non-hydric type; (2)  change from a
vegetation cover of hydrophytes to predominantly mesophytes or xerophytes;
(3)  change from land  that is flooded at some time of the year to land that is
infrequently or never flooded during years of normal precipitation.

    Treatments considered next are those which remove contaminated soil
layers.  Mechanical and chemical stabilization after soil removal is also
discussed.
                                     7-20

-------
    (2-1)    Remove Plow Layer
    The plow layer is defined here as the soil layer from the surface to a
depth of about 10 centimeters.  Removal would be accomplished by hand shovel-
ling or larger digging machinery.  The unstable marsh soils would necessitate
working from platforms, barges or other support structures.
    The ecological effects of soil removal in the marsh would be similar to
those discussed for surface alteration (scraping and grading) and vegetation
removal.  Some inference can also be drawn from the earlier discussion of
dredging effects  (Section III).  Major influences include
        — acceleration of surface runoff
        — alteration of drainage patterns
        — loss of habitat for wildlife
        — increased erosion
        — loss of organic matter and nutrients in surface
          soil layers.
Recovery would be similar to that for scraping and grading but would be
slower because more of the soil-organic layer is removed and elevations are
somewhat altered thus changing the tidal flooding depth.

    (2-2)   Remove Shallow Root Zone
    In a coastal marsh, the root zone is relatively shallow and therefore,
the effect of soil removal to the depth of a plow layer and/or to the shallow
root zone can be treated the same.  See treatment 2-1.

    (2-3)   Remove by Scraping and Grading3
            Mechanically Stabilize
    The effects of scraping and grading are discussed in treatment 1-4.
This treatment, combined with "mechanical stabilization" includes removal of
contaminated soils (less than 5 jm in this case) and compaction of subsurface
soils or surfacing with impervious materials to seal the subsurface against in-
filtration.  As previously noted  (treatment 1-1) infiltration control in a marsh
area which is at'least periodically flooded has little meaning.  Hence this
combined treatment method is inappropriate since mechanical stabilization is
inappropriate.
                                      7-21

-------
     (2-4)   Remove Plow Layer,  Meohanioallii Stabilize
     This combined treatment is  inappropriate for the marshland type; see treat
 ment 2-3 and 2-1  for further discussion.

     (2-5)   Remove Shallow Root Zone,  Mechanically Stabilize
     This combined treatment is  inappropriate for the marshland type; see treat-
 ments  2-3 and 2-2 for further discussion.

     (2-6)   Remove by Scraping  and Grading,
             Chemically Stabilize
     The  effects of this  combined treatment  are  essentially those discussed in
 treatments 2-3 and 1-1.   The chemical  stabilization method employed would
 have to  be instituted for erosion control rather than infiltration  control for
 the  reasons  previously given.   The erosion  control  provided would be expected
 to be  transient but might be sufficient to  somewhat mitigate runoff impacts
 resulting in a more rapid recovery than scraping and grading alone.

     (2-7)    Remove Plow  Layer,  Chemically Stabilize
     This  combined  treatment  would result in  effects similar to  those discussed
 for  plow  layer removal alone (treatment 2-1).   However, as  discussed above,
 erosional  problems  would  be  mitigated  temporarily and  recovery  would be  some-
 what more  rapid than  that  without  stabilization.

     (2-8)    Remove  Shallow Root Zone.,  Chemically Stabilize
     The effects of  this treatment  are  discussed  in  treatments 2-2 and 2-7.
     Barriers to exclude people  and to  exclude both  large and small animals  are
 discussed  in the following treatments.

     (3-2)   Barriers to Exclude  People
     It is assumed that exclusion would be accompanied through the use of
barbed wire fencing or some  sort of mesh material such as chain  link or Cy-
clone fencing, hardware cloth, or poultry wire.
     If a natural  state is deemed beneficial, then nothing but benefit can
derive from the exclusion of humans.  Benefits due to increased "naturalism"
will accrue in direct proportion to the intensity of previous human use or
 intrusion.  Numbers of species and their diversity will certainly increase as
 shy  species return to the marsh; this may be especially true in the  case of

                                   7-22

-------
migratory  game  fowl  if hunting  is  terminated.  Vegetative  cover,  and  hence
primary  and  secondary productivity, will  increase  if  these parameters had been
below maximum levels due  to human  trampling.  However, note  that  secondary
productivity is not  likely to increase  greatly due to the  lack  of macrophyte-
consuming  herbivores which frequent salt  marshes.   Finally,  erodibility would
likely remain unchanged or perhaps even decrease;  indeed,  a  negative  erodibility
may  exist  for marshes since they are  areas  of deposition.

     Humans  often find cause for interfering in coastal marshes,  especially
 when they live near such areas.   Natural marshes  are doomed to self-destruc-
 tion due to sedimentation and consequent evolution (succession)  into  an up-
 land habitat.   Part of this evolution  is often the buildup of a  bar at the
 tidal inlet(s), hence,  the tidal  flushing of the marsh is greatly reduced or
 eliminated.  As this happens, conditions may become anaerobic as water temper-
 ature and salinity rises and, therefore, marsh flora and  fauna die and decay.
 Such conditions create a great nuisance near points of human activity, and
 they provide a haven for insects  such  as gnats and mosquitoes, further aggra-
 vating the  condition of local human populations.  For exactly these reasons,
 and also to maintain a maximum level of productivity, marsh channels  are often
 dredged regularly, especially at  the tidal inlet(s).4  Such dredging  also en-
 hances the  flood control value of marshlands by providing open channels for
 runoff during precipitation events.   Thus, exclusion of human activity may not
 be taken to include channel dredging,  especially  if there is significant human
 activity in the vicinity of the contaminated area.  It should be noted that, if
 periodic dredging is performed, the dredged materials must be checked for con-
 taminant level and, if found to contain unacceptable levels of contamination,
 must be dealt  with accordingly.

     (3-2)    Barriers to Exclude Large Animals
     Not applicable to this land type.

     (3-3)    Barriers to Exclude Large and Small Animals
     It is  assumed that  exclusion would be accomplished through the use of
fine-meshed materials,  such as screen with mesh diameter less than 1.0 cm,  or
through the use of solid materials, for example,  brick and cinder block.
                                     7-23

-------
      Animals such as mice, voles, raccoons, rabbits, possums, skunks, weasels,
  and coyotes could be excluded by this treatment.4  Mammals rarely live within
  a coastal marsh; generally they reside in nearby upland habitat and use the
  marsh only for forage,  consuming seeds,  some plants,  insects,  and other small
  animals.   Further,  such mammals are noted to be "represented by relatively
  small populations."  Hence,  exclusion of these creatures would have very little
  effect on the marsh ecosystem;  a slight  drop in species diversity and in the
  total number of species occurring in the marsh would  be the only noteworthy
  effects.
      However, if a large area,  e.g.  1 to  10 km2,  were  fenced off,  the result-
  ing pressure on surrounding  lands for forage and prey could cause significant
  harm to these lands.  This also could cause effects on nearby  human popula-
  tions ranging from  a  minor nuisance to a significant  economic  hardship,  for
  example,  if the search  for food led the  displaced animals  into human dwellings
  or  into agricultural  lands,  of  if predators were forced to prey upon domestic
  animals.

      Birds are the most  important  small animals in a coastal marshland from  the
perspective of contaminant mobility.   Birds  forage on seeds, benthic  invertebrates,
open  channel  vertebrates, and insects;  some  of these prey species may concentrate
a contaminant through direct ingestion  (e.g., worms and filter feeders), or
through  their own food source which has already ingested a contaminant and per-
haps  concentrated it to  some degree.  Hence the birds may ingest significant
quantities of a contaminant and may then pass it through their digestive system
and deposit it offsite in feces or concentrate the contaminant in their own tis-
sues and then die or be preyed upon offsite.  The former seems more probable.
The same set of circumstances may also apply to fish and insects, in which cases
the latter event is more likely.

   Therefore it may be pointless to prohibit only horizontal crossings at the
 land surface; some sort of barrier to movements above  and below the surface
 may also be desirable.  The effects of such barriers would be considerable,
 compounding greatly as the size of the area of exclusion increases.
                                     7-24

-------
    First, although net secondary productivity may change very little over a
long period of time, large short-term fluctuations may be expected as predator
species and prey species adjust to sudden partitioning from each other.  It
is especially the prey species which may explode in population with a sudden
halt to predation.  "Boom and bust" cycles of prey populations could be ex-
pected as the affected species increase their numbers, outstrip the available
food supply, and die back in catastrophic instances, whereupon the cycle is
renewed.  It^can be seen that, within this cycle, vast effects on the food
sources of the former prey species can be expected.
    Additional effects can be anticipated from the elimination of certain
species which spawn in the marshland.  These effects may range from the triv-
ial to the extreme depending on the size of the area fenced from the intru-
sion, the availability of alternative breeding sites, and the importance of
the species whose spawning grounds have been contaminated.
    Additional pressure on surrounding lands from avian predators and for-
agers displaced from the contaminated areas can be highly significant if sur-
rounding marshland is burdened beyond its capacity to produce food for the
displaced birds.  This could occur, for example, if half of a 20 km2 marsh
were fenced and screened,   Thus, wildly fluctuating populations of predator,
prey, and forage species may occur offsite in areas unaffected by contamina-
tion.
    Finally, diversion of migratory waterfowl from contaminated marshlands
can not only have serious effects on surrounding areas, but may also irrevoc-
ably reduce the populations of sensitive species already pressured by general
habitat reduction, diversion from nesting or wintering sites by human activi-
ties other than habitat reduction, and hunting.   Such effects would be en-
tirely site-specific and cannot be addressed further for the general case.

    (4-1)   Asphalt Hard Surface Stabilization
    This treatment is outside the scope of work.

    (4-2)   Concrete Hard Surface Stabilization
    This treatment is outside the scope of work.

     (5-0)   Application of Sewage Sludge
    This treatment is outside the scope of work.

                                    7-25

-------
       (6-1)  High Presure Washing (<3 cm)
      This treatment is outside the  scope of work.
       (6-2)   Flooding to 30 am
      This treatment is outside the  scope of work.
      (7-0)   Soil Amendments Added
      This treatment is outside the  scope of work.

  7.5    RECOVERY AFTER  CLEANUP
      Recovery of  coastal marsh ecosystems following clean-up operations  is de-
  pendent  upon several  factors  beyond the characteristics of the clean-up oper-
  ation itself.  Included in  these factors are
          - the size of the area affected by the clean-up operation
         - location of the disturbed area relative to open water  (areas
           adjacent to open water are subject to wave and tidal effects
           as well as  fresh water runoff)
         - weather conditions following  cleanup (heavy rains would
           increase runoff erosion problems)
 The sequence of events leading to "complete" recovery in a coastal marsh sys-
 tem is not a multistep successional scheme  such as is the case with develop-
 ment of the classic climax forest.   Recovery in the marsh system would be
 essentially directly from disturbed state (e.g.  denuded)  to  the original
 vegetation type  (e.g., Spartina).   Graphs portraying  a  very general  recovery
 sequence  are provided  in  Figures  7-3 through 7-5  (see Appendix B).
     The rate of recovery  is  largely determined by  the erosional processes .
 which characterize  the marsh system and  which may  be  intensified  by  cleanup.
 The growth rate of  vegetation  is probably secondary to  the physical  processes
 determining  soil  stability and substrate composition.   With the normal high
 productivity rates  characteristic of the marsh land type, once erosion is con-
 trolled and  revegetation initiated, recovery would be expected to proceed quite
 rapidly.

 7.6    QUANTITATIVE ASSESSMENT  OF CLEAN-UP IMPACTS
    Clean-up procedures in coastal marshes are scaled in Table 7-5 with re-
gard to impact on this ecosystem.   The scaling provides a relative view of
the length of time (in growing seasons)  required for complete recovery to the
                                    7-26

-------
                                                         HYPOTHETICAL COASTAL INTER-TIDAL MARSH
   to
   c
    I
    CO
 CO ?0
 o n>
 -j- o
 — • o
   <
 to n>
 c+ -*s
 Q) <<
 cr
 N O
 Ql O
 rf OJ
 _i. to

 O r+
 rs a*
r-l
\
\
\
\
\
\
\
1
1
1
_
\
\
\
— \
1












t.l g
• r-
\ 5
\ w
r-
	 N
H
O
Z
O
i
— 1

cu n>
-a -s
•a  i
o o.
O> Q>
c+ —•

o' 3
rs o>
   -s
O to
to tQ
o
-•. o
—' 3-
   a>
o 3
o -••
< o
n> cu
-s —•
                      ERODABILITY

                                              HOMEOSTASIS
                                                                       PRIMARY PROD.
                                                                                                   COVER
SP. DIVERSITY

(VEGETATION)
^ r-ji
\
\
\
\
\
\
\
i
1
1
1
_ __J
\
\
^
— i
i


i
i
i

i
\
\
\
— \
\
i
i

i
•
i

\
\
\
— \
i
i
i
i
i
i
c
— I 	 z
!
\ 1
\ z
— \ 5
tm
O
1 8
1
i =
— i ^"


-------
        CQ
         C

         o>
HYPOTHETICAL COASTAL INTER TIDAL MARSH
00
1
0*
n> n>
3 0
O 0
B> n>
3 O
a. -+,
CO O
3- O
(U O>
— « CO
O O» O

— J -••
«< s
o> o>
1
CO <-* •<
O -•• ni
±; °- 5
0> 3
3 0»
O -J
< CO
O» 3- rf*
• -h
O
nl
o
_j.
3 o>
CO

<
CD
CO
(D
OJ
M^
^j. — •
!J. °
o
^ i
V
***.^
_ \
1
1
1
__ 1








ERODABILITY

.
1
1
\
\
»

\
\
\
X
\
_ \
\
I
., 	 1
1
\
^^
	 "*\



	







HOMEOSTASIS

, 	 J
1
|
V
x
	 X



\
\
\
— \
\

^v~^
***«*!»
\
— 1
1
1
1
|


1





PRIMARY PROD.

T 	 J
f
I
\

\

	 X
X
\
\

1
1
1

^
^»«*.^
""•••
—










COVER




^
\
X
X
	 X
X
X
\
\

1



'N
\
/
/
1

1





















J
r 	 J
\ m

^%^ ^
*^\ >
\ ±
	 • 0
a
m
O
^
—






SP. DIVERSITY
(VEGETATION)

T 	
/
\ M
X "•
v i-
v 3D
	 N. m
Nv 2
\ i
\ >
\ n
\
— I
•

-------
                   ACCESS BARRIERS
        00
        V)
        O
        UJ
                          5

                        YEARS
                                   10
Figure 7-5.   Response of coastal inter-tidal
              marsh  to the erection of fencing
                        7-29

-------
 "natural" state after clean-up operations are terminated.  It should be
 stressed that these rankings are highly subjective in nature.  The specula-
 tive nature of these results is also emphasized by the location and meteor-
 ological factors which will influence recovery.
     The confidence with which this scaling is provided is somewhat inversely
 related to the size of the area affected.   This results from the uncertainties
 related to recovery processes and the potential for external influences on
 the marsh ecosystem subsequent to clean-up operations.  Hence the scaling
 here is much more relative than absolute  (more numerical than quantitative).
     The rationale for developing Table 7-5 is provided in Section 7.4.   General
conclusions drawn from Section 7.4 and Table 7.5 are presented in the following
section.
                                                            2
     Examination of Table 7-5 indicates that for the 0.01 km  area, the treat-
 ments may be grouped into three basic categories with regard to level of ef-
 fect and length of time to recovery.
     The lowest impact group includes  construction of barriers (3-1,  3-2),
 chemical stabilization alone (1-1),  clear  cutting (1-2)  and soil covering
 (1-7,  1-8).   The common characteristic of  this group of treatments is gen-
 erally a lack of disturbance of the  root and "native soil" structure. Or-
 ganic muds and root systems may mitigate erosion,  which is an important fac-
 tor in recovery of the coastal  marsh.   Naturally,  some disruption will  occur
 from the means and equipment involved  in the clean-up operations.  Hence,  bar-
 rier construction (3-1,  3-2)  will  have the least impact  and soil covering
 (1-7,  1-8)  the most within this relatively low impact group of treatments.
     Deep soil  coverings  (1-8) may elevate  the marsh to a level where the
 natural flooding cycles  are eliminated.  Recovery in such areas  would not
 lead to a marsh habitat,  but creation  of a new terrestrial habitat type with-
 in the coastal marsh.
     The group  of treatments which  have an  intermediate level  of  effect  in-
 cludes surface alteration (scraping and grading)  alone (1-4)  or  accompanied by
 chemical stabilization (2-6); stump and root removal  (1-3); and  removal of
 the plow layer followed  by chemical stabilization  (2-7).   These  clean-up
 treatments are characterized by a  more severe disruption  of the  native  soil
                                    7-30

-------
Table 7-5.   Predicted years to reestablish precleanup marshes.
Treatment
0-1 Natural Rehabilitation
1-1 Chemical Stabilization
1-2 Remove Vegetation by Clear Cutting
1-3 Remove Stumps and Grub Roots
1-4 Surface Scraping and Grading (<5 cm)
1-5 Shallow Plowing (<10 cm)
1-6 Deep Plowing (10 to 40 cm)
1-7 Soil Cover (<25 cm)
1-8 Soil Burial (25 to 100 cm)
2-1 Remove Plow Layer Soil (10 cm)
2-2 Remove Shallow Root Zone Soil (<40 cm)
2-3 Scrape Surface, Remove (<5 cm), Mechanically Stabilize
2-4 Remove Plow Layer (10 cm), Mechanically Stabilize
2-5 Remove Shallow Root Zone (<40 cm), Mechanically Stabilize
2-6 Scrape Surface, Remove (<5 cm), Chemically Stabilize
2-7 Remove Plow Layer (10 cm), Chemically Stabilize
2-8 Remove Shallow Root Zone (40 cm), Chemically Stabilize
3-1 Barriers to Exclude People
3-2 Barriers to Exclude Large Animals
3-3 Barriers to Exclude Large and Small Animals
4-1 Asphalt Hard-Surface Stabilization
4-2 Concrete Hard-Surface Stabilization
5-0 Sewage Sludge Application
6-1 Washing (^3 mm)
6-2 Flooding (3 to 30 cm)
7-0 Soil Amendments Added
Area Disturbed
(km2)
0.01
0
o
5
10
10
NA
NA

5
12
15
NA
NA
NA
7
10
12


0

*
*

*
*
0.1
0
o
5
10
10
NA
"

7
10
15
20
NA
NA
NA
8
12
15


0

*
*

*
*
1.0
0
MA
10
1 5
15
NA

12
20
20
25
NA
NA
NA
15
20
25

0
0
0

it
*

*
*
10.0
0

15
25
NA

18
30
30
35
NA
NA
NA
15
30
35

0
0
0

*
*

*
*
Assumptions


In upland or mangrove areas


Marsh elimination possible
Mechanical stabilization
inappropriate for
marshland type
Chemical stabilization
may not be appropriate
for large areas









"Outside scope of work. 	 • 	 '
NA - not appropriate for this land type.

-------
structure than would be experienced with the lowest impact group.  Thus the
potential for erosion is more substantial and recovery would be delayed as
a result.
    In addition to increasing erosion, removal of roots and disruption or re-
moval of surface soils would remove the plant and soil materials which would
most enhance rapid habitat recovery.  Root sprouting, which would occur after
clearcutting (1-2) would be eliminated and revegetation would occur via root
systems in adjacent areas and seeds dispersed by wind and flooding of the area.
Revegetation by seeds would be hindered to some extent by surface disruption
and soil removal.
    The most deleterious set of clean-up treatments include removal of the
plow layer (2-1) and removal of the shallow root zone (2-2).  The processes
of increased erosion and delayed revegetation, which are discussed above, are
enlarged by these treatments.
    As previously stated, the level of effect associated with any of the
treatments for the marsh land type is area dependent.  That is, the effects
generally increase with size of the area affected.  This dependence can be
clearly seen from the importance of erosion in determining recovery and the
importance of neighboring areas for revegetation.  With the smaller sized
clean-up areas, the surrounding area would act as a buffer for both tidal
and freshwater flooding events.  In addition, seeding and root sprouting from
adjacent areas would be more effective with smaller cleared zones.  Larger
clean-up sites would benefit less from these erosion buffer and revegetation
processes, and hence slower natural recovery would be expected.
    Treatments which do not involve disturbing the spill area (i.e. natural
rehabilitation  (0-1) and barrier construction (3-1, 3-2) would not have area
dependent impacts.  However, some increased stress in adjoining areas may re-
sult with increasing size of barrier enclosures.
    As discussed in the preceding sections several of the alternative clean-
up treatments may effectively destroy the marsh land type.  In these instances,
recovery would not be to the original habitat form.  Such treatments include
2-2, removal of the shallow root zone (less than 40 cm), and 1-8, deep soil
covering  (up to 100 cm).  The flooding cycle may be disrupted to the extent that
the former treatment results in an open water area and the latter in a more
                                     7-32

-------
 dryland environment.  Because of these consequences, these treatment types may
 not be reasonable with regard to recovery considerations.   The effects of
 other treatments are probably not irreversible although subsequent erosional
 damage may result in an extended recovery period,  particularly for larger
 treatment  areas.

 7.7   CONCLUSIONS
      Of the clean-up treatments  considered,  several  were concluded to  be  inap-
 propriate  for  the coastal marshland  type.  Shallow and  deep plowing is not  ap-
 propriate  for  containment of  spill materials.   Mobility would  be  enhanced rather
 than retarded  by this action.  Also, mechanical stabilization  is  inappropriate
 as  a mechanism for sealing  the marsh soils against infiltration.   This is because
 infiltration control has no clear meaning for  a frequently  flooded environment
 such as a  marsh.   For similar  reasons,  some  chemical  stabilizers  are inappropri-
 ate treatments in coastal marshes.   However, for binding soil  materials together
 to  mitigate erosion,  chemical  treatments may be useful  as a temporary  measure,
 particularly for  rather small  disturbed areas.

     Soil removal  methods were  concluded to be  the  most  deleterious  to marsh
 recovery.   Effects  of such  treatment increases  with the  depth  of  removal.
 Deep  soil  coverings  may have similar effects.   In  fact,  appreciable  eleva-
 tion alterations  in  either  direction may effectively destroy the marsh  hab-
 itat; recovery may be to more  terrestrial or to open-water  communities.   Chem-
 ical stabilization may mitigate erosion and thus aid recovery  for all soil re-
moval techniques  from scraping and grading to removal of the shallow root
 zone.  Vegetation removal was  felt to be less destructive than soil removal
treatments since the substrate materials which supported "climax" vegetation
may remain largely undisturbed.  Clear cutting is certainly less detrimental
than removal of all vegetation and roots.
    Barriers were felt to have little overall effect  on the coastal marsh
although certain stresses in neighboring areas may result.
                                   7-33

-------
 7.8   COASTAL INTER-TIDAL MARSH REFERENCES


 1.  Bayly, I. A. E. and W. D. Williams.  Inland Waters and Their Ecology.
     Longman, Australia, 1973, 316 pp.

 2.  Bourn, W. S. and C. Cottam.  Some Biological Effects of Ditching Tide-
     water Marshes, Res. Rep. 19, U.S. Fish and Wildlife Service, 1950, 30 pp.

 3.  Brown, C. L. and R. Clark.  Observations of Dredging and Dissolved Oxygen
     in a Tidal Waterway, Water Resour. Res. 4(6):1381-1384, 1968.

 4.  California Department of Fish and Game.  The Natural Resources of Capinteria
     Marsh and Recommendations for Use and Development, Coastal Wetlands Series
     No.  13, 1976.

 5.  Chapman, D.  W. Effects of Logging Upon Fish Resources of the West Coast,
     J. Forest.,  60:533-537,  1962.

 6.  Frankenberg, D. and C. W. Westerfield.   Oxygen Demand and Oxygen Depletion
     Capacity of Sediments from Wassaw Sound, Georgia, Bull. Georgia Acad. Sci.,
     1969.

 7.  Moore, D. and L. Trent.   Setting, Growth and Mortality of Crassostrea
     virginica in a Natural Marsh and a Marsh Altered by Housing Development,
     Proc.  Nat'l. Shellfish Assoc.  61:51-58, 1941.

 8.  Smith, W. G. Spartina "Die-back" in Louisiana Marshlands, Coastal Studies
     Bull.  Special Sea Grant  Issue,  5:89-96, 1970.

 9.  U.S.  Dept. of Interior.   Interim Classification of Wetlands and Aquatic
     Habitats of the United States,  U.S. Fish and Wildlife Service,  1976,  109 pp.

10.  U.S.  Dept. of Interior.   A Study of the Disposal of Effluent from a Large
     Desalinization Plant,  Office of Saline  Water R§D Prog.  Rep. No. 316,  1968.

11.  U.S.  Environmental Protection  Agency.  Impacts of Construction Activities
     in Wetlands  of the United States, Rep.  No.  EPA-600/3-76-045, 1976, 392 pp.

12.  Whittaker, R. H.  Communities  and Ecosystems.   The Macmillan Company,
     London, 1970.
                                     7-34

-------
              PART  II
        MANAGED ECOSYSTEMS

CHAPTER 8,  AGRICULTURAL LAND AREAS
CHAPTER 9,  URBAN/SUBURBAN LAND AREAS

-------

-------
                                  CHAPTER 8
                                 AGRICULTURE
 8.1   OVERVIEW
      Lands  used  for  the  production  of renewable  resources  and  products  con-
 sumed as  food, fiber,  timber,  and feeds  for  livestock  are  by definition agri-
 cultural  lands.   Crops raised  for livestock  feeds  are  discussed  as  forage,
 permanent pasture, rangeland and the  native  pastures of woodlots.   Plant spe-
 cies  discussed range from  local area  natives to  patented specialty  cultivar
 crops.
      In the broad sense, agricultural  land could include tundra  crops of li-
 chens consumed by reindeer herds and  the alpine  meadows grazed by rodent her-
 bivores.  A narrower view  is held in  this chapter.  Agricultural lands  are
 considered  to be  those managed to produce crops  at commercially  significant
 rates for food, feeds, and fibers.
      In this  chapter the impacts by cleanup  are  taken  as short-term disturb-
 ances because the vegetative component can be replaced precisely—for a  price.
 Mature deciduous  trees as  large as 20 to 23  cm in diameter and 8 meters  tall
 can be dug  up, moved,  and  replanted in normal landscaping.10 Sodding or  reseed-
 ing is likewise economically determined as are most decisions  about rehabili-
 tation of agricultural lands after cleanup of some contaminant.  As a routine
practice  in the deciduous forest,  a farm can be  cleared of standing crops,
 whether small grains or timbered woodlot, surface-mined for coal, the spoils
 graded to the original contour, limed, fertilized,  reseeded, and be back  in
 full  production in 2 years.  Equipment used  in surface mining  is different
 from  earthmoving  equipment for local road building, light building construc-
tion  and  farming.  In general,  surface-mining earthmoving equipment is orders
of magnitude larger than usual  construction equipment.
     Agricultural land is estimated to be from 526  million hectares to 930
million hectares  in the United  States5 but a variety of figures are used for
agriculture depending on the author's interest.   The percentages of use are

                                     8-1

-------
  approximately  25  percent  for  grazing  in  livestock production  and  20 percent
  for  crop production.

      This chapter concentrates on farming  in three of the natural ecosystems.
  These are the  hot desert  of the southwestern United States, where irrigation
  supplies almost all agricultural water supply, the adjoining  midcontinental
  grasslands, and the formerly hardwood forest areas.  These are located as
  defined in Figure 8-1.  Because it lies  in the intermediate position between
  the other two  sources of  agricultural land, and because almost the entire
  grassland has been appropriated for agriculture, the prairie  is described in
  some detail with its crops.  The irrigated desert is confined to gently slop-
  ing or flat terrain and the actual commercial use is a small portion of the
 desert terrain.  Deciduous hardwoods occupy rolling to steeply sloping terrain
 but agriculture is practiced on the flatter slopes which,  as in desert, are
 only a fraction of the total area.
      Agriculturally important, the midcontinental grasslands in Figure 8-1
 reach from  the  30th parallel in the south to the north beyond  the  Canadian
 border.   Bordered  on the west  by the Rocky Mountains,  they rise along  the
 105th meridian  from the central Mexican  border  into  Wyoming, turning north-
 west ly across 8 degrees of longitude before meeting  the  Canadian border.   The
 eastern  grasslands boundary lies  along the  96th meridian with  a major  incur-
 sion,  generally northeast  on a concave arc,  from the Red River into south  cen-
 tral  Ohio.  The east  grassland boundary keeps west of  the Mississippi  River
 to  the Missouri River,  crosses  both  in a  few miles and arches  north of  the
 Ohio  River into central  Ohio,  north  across  Ohio,  and turns generally west  to
 recross the Mississippi  and parallel it to  the  northwest, crossing into Canada
 on  the 96th meridian.
     Topography of the agricultural prairie ranges from the western short
 grass high terraces and  alluvial fans  across the plains to the steep hills and
 lakes on the eastern fringes of Illinois  and west Ohio.  The wide,  level to
 gently rolling plains with tree-clad streambanks in the prairie midsection
 were the first to fall totally to the plow as croplands.   Hard red winter
 wheats grow in the southern prairie and grade into hard spring wheats in upper
Nebraska north to the Canadian border.13   Annual precipitation in the winter
wheat belt ranges from 350 mm in eastern  Colorado, Wyoming,  and Montana to

                                      8-2

-------
oo
i
oo
                                 \-
IRRIGATED AGRICULTURE
AND DRY FARMING
                                                                        RAINFED AGRICULTURE
                                                                                                   EASTERN HARDWOODS
                                                                                                 RIVER BOTTOM FOREST
                                  DESERT SHRUB   DESERT     SAVftNNA
                                            GRASSLAND  IMESOUITE. MCTD
                               SHORT GRASS  TALL GRASS
                            INTERIOR
                            DESERTS
  TO
SEMI-ARID
BASINS
MAJOR
GRASSLANDS
CENTRAL AND
NORTHERN
DECIDUOUS
FORESTS
                   Figure  8-1.   Indigenous vegetation  zones  of the conterminous  United States
                                   (adapted  from  Shantz and Zon15).

-------
 900  mm in  eastern  Kansas  and  Oklahoma,  with  warmer  summers  and  longer  growing
 seasons than  the spring wheat belt.   Precipitation  in  the spring  wheat belt
 ranges from 300 mm in northern Montana  to  600 mm  in western Minnesota  and
 eastern Dakotas.   Generally,  the  separation  of the  winter and spring red
 wheats is  the isotherm of 18°C for the  three summer months  of growth into
 ripeness.
     Precipitation regimes for prairie  are shown  in Chapter 2,  Part I,  and
 these  are  basic figures around which  agricultural needs are managed.   In the
 west side  near the  Little  Powder  River  in northeast Wyoming, annual rainfall
 is 400 mm  with an  evaporation potential for  1500 mm in an area  where wheat
 and oat  agriculture displaced mixed prairie  grassland and artemesia species,
 but the  primary use became cattle-grazing because of the combination of lim-
 ited rainfall and deficiencies in soil phosphorus and nitrogen.
     The physical characteristics of prairie soils reflect  the  environmental
 regimes  of the grasslands as shown in Table  8-1.   On the Missouri side, the
 soils  are deep and the rainfall is sufficient for tall grass.   The Colorado
 side of  the prairies have permanently dry subsoil and short grass.  Between
 these boundaries lies the minimum line of annual  runoff shown in Figure 8-1
 as the 2-cm precipitation isogram.  The small grains straddle this precipita-
 tion feature and their presence results in the designation of "breadbasket."
Table 8-1.
Relationship between annual precipitation, species composition,
and moisture consumption, Rocky Mountains to Missouri transect
of prairie (adapted from Jennyi2).
Indigenous
Plant
Community
Grama grass
Buffalo grass
Wire grass
Bluestem grass
Bunch grass
Bluestem grass
Sod grass
Depth to
permanently
dry zone
(cm)
30-36
60-80
85-115
absent
Approximate
annual
pptn
(cm)
<45
45-55
55-70
70-100
Plant
height
(cm)
10-35
45
75
110
Region of
observations
Eastern Colorado
Western Kansas
Central Kansas
Eastern Kansas and
Western Missouri
                                    8-4

-------
 8.2   NATURAL  PERTURBATIONS
      The  significant agricultural event  in physical and  economic terms  is the
                                  18
 loss  of soil  by  erosion processes.    These  losses are generated by two dif-
 ferent agencies, wind and water runoff.  West of the  2-cm  isogram  (Figure 8-1)
 erosion is driven by wind.  East of the  2-cm isogram  water runoff  is the
 principal erosive force.
      Publications dealing with erosion are numbered in the thousands.  One
 agency field  office has collected 2,500  reports, bulletins, and books on wind
 erosion alone.   This author estimates runoff erosion publications number
 more  than twice  as many as cited for wind erosion.  A simplification is called
 for.
      The national wind erosion area is defined in Figure 8-2, while water
 erosion sites are shown in Figure 8-3.
      Wind erosion data and its erosive forces are used in general predictions
                                   3
 of soil loss  as  shown in handbooks.    However, these  must be used with cau-
 tion.  Attempts  to estimate wind erosion by  land types across the spectrum of
 suggested cleanup techniques faces a basic inadequacy in the primary data.
 The data on wind speed and direction are generally collected at airport loca-
 tions and a few  electricity-generating plants.  These locations are represen-
 tative of flat terrain rather than rolling,  or hilly  or mountainous regions.
 The erosive power of wind may be seriously underestimated by using published
                                                               o
 data  to predict wind erosion damage for sites as small as 10 km  and smaller
 and a few kilometers distant,  since the extreme wind  speed is pertinent.
 Horizontal extrapolation of wind speed for offsite predictions faces direc-
 tional shear, turbulent discontinuities, and orographic deflection of unknown
 magnitudes.
     Wind-energy distribution maps are being undertaken at this time and may
 alleviate a degree of the risk in making horizontal extrapolations from the
 site of measurement to the cleanup site.  These maps will be useful in flat
 terrain under overcast conditions.   They are not available for generic use
unless the cleanup treatments  are restricted to the vicinity of airports.
     Probably the best return  in deciding what sites should be chosen as can-
didates for cleanup would come from  installing remote wind recording equipment

                                     8-5

-------
                   erosion
          Soil moisture-Wind velocity factor
              Very low   Qj 0-10%
              Low
              Intermediate f "JJ 26

              High      HH81-150%
              Very high   QU > 150%
Figure  8-2.  Surface soil  moisture-wind velocity  soil  loss
              factor as a  percent  of soil  loss at  Garden City,
              Kansas (marked bv X)  (adapted  from Agricultural
              Research Service^).
                              8-6

-------
           Figure  8-3.   Rainfall  erosivity  index,  based  on maximum
                        30-minute intensity,  from  an  average of
                        annual maximums  (from Wischmeier^).

 and precipitation recorders at the sites.  From onsite  data general regres-
 sions would be possible with the weather station network data.  Without onsite
 data the predictions are worthless since they can be highly misrepresentative
 of the site.
     Rainfall intensity is the factor detaching soil fragments and allowing
 runoff to carry them away.  The data summarized in Figure 8-3 represent a
 portion of the work done by Wischmeier and his colleagues; .One of their sum-
 maries  has been widely used in erosion predictions.   Two reviews of erosion
 dislocation mechanisms and of detachment by wind9 and water16 lay the basis
 for the widely used "Universal Rainfall-Erosion Equation."  This is based on
 the kinetic energy of the rainfall and the maximum 30-minute  intensity of
 the storm.   It must be used by trained staff for particular  sites.
     Plant  growth, precipitation and  climate are interrelated  variables that
determine the erosive force exerted upon soil from which sediments  are derived.

                                     8-7

-------
A general integration for all these factors is shown in Figure 8-4.  That
precipitation regime transects the natural ecosystems upon which the agri-
culture discussed in this chapter draws.  These are the natural forces that
test the wisdom in selection of cleanup treatment, when the choice is field-
tested at a site.



Q
_J
UJ
h-
z
UJ
O
UJ
m
>ivnNN>
^








350

280
cc
UJ
K-
LJJ
5
O
u 210
cc
o
00
CC
UJ
n
§ 140
0
o
cc
1-
UJ
s
70
o
—

Shrub 1

/ \
/ \
/ *
\
1 \
/ \

i ^^*" —
1 "" 	 	
/ 	 	 	
/
_ i
/,,,,,.
0 25 51 76 102 127 152
CENTIMETERS
EFFECTIVE PRECIPITATION
Figure 8-4.  Idealized effective precipitation, vegetation, and sediment yield
             on west to east transect (adapted from Bennett and Donahue^).

8.3  MAN-MADE PERTURBATIONS
     All agriculture is a man-made perturbation that functions as a production
system because of the energy added to change it from a natural ecosystem.  In
                                     8-8

-------
 the context of this chapter the alleviation of natural perturbations on
 cleanup areas is addressed.

      The ideal preventative of wind and water erosion is a layer of wetted,
 attached mulch.   For semi-arid lands the procedures were derived after the
 enormous dust storms of the late 1930s.  The amount of mulch needed varies
 with soil texture and species of crop.3  A summary of application rates for
 deciduous forest agriculture is given in Table 8-2.  As shown there seedling
 plants hold soil losses to those comparable to tonnage quantities of straw
 and stover (fodder)  residues.   Plant residues and seedling grasses are almost
 universally used for wind and water erosion control.

           Table  8-2.   Effectiveness of ground cover on soil  erosion
                       loss at construction sites  in deciduous forest
                       zones.
           Ground  cover
           Established  seedlings
             Permanent  grasses
             Ryegrass (perennial)
             Ryegrass (annual)
             Small grain
             Millet and sudangrass
             Field bromegrass
          Grass sod
          Hay (5 tons per ha)
          Small  grain straw (5 tons per ha)
          Corn residues (10 tons per ha)
          Wood chips (10 tons per ha)
          Asphalt emulsion (310 gal per ha)
Soil loss relative
 to bare surfaces
     (percent)
         1
         5
        10
         5
         5
         3
         1
         2
         2
         2
         6
         2
          (Adapted from Hotes et al.11
     Large amounts of surface runoff are generated by acreages of cropped
soils.   Need for runoff control is intensified where the soil is bared and
                                    8-9

-------
 multiple treatments are rated for runoff control  in Table 8-3.   These are
 the kinds of combined treatments  that  are evaluated in Section  8.6,  Quanti-
 tative Assessments.

            Table 8-3.  Control systems and relative effectiveness.
                              Components               Percent  effectiveness
        1.        Seed,  fertilizer,  straw mulch  plus             91
                 erosion  structures,  normal
                 Same as  (1)  except chemical  (12                90
                 months protection) replaces  straw
                 Same as  (1)  except chemical  straw              91
                 tack replaces asphalt
        2.        Seed,  fertilizer,  straw mulch  with             90
                 diversion berms plus sediment  basins
        3.        Seed,  fertilizer,  straw mulch; down-           96
                 stream sediment basin using  flocculants
                 Same as  (3)  without  straw mulch                94
        4.        Chemical (12 months  protection) sedi-          94
                 ment basin using flocculants
                 Same as  (4)  with seed, fertilizer              96
     (Adapted from Hotes et al.11)


8.3.1  Runoff and Fertilization

     Runoff does more than produce sedimentation.  It can strip away organic
matter and mulches.  The plant nutrients lost in this way are some of those
recommended for reseeded areas in Table 8-3.  These nutrient losses were
studied quantitatively on field plots in 1917 and included nitrogen, phos-
phorus, calcium, and sulfur.   A recent review on phosphorus14 losses from
vegetables fertilized with manure placed losses at 70 kg/ha per year.14
This is much more than many farm programs call for in fertilization regimes.
Seasonal variations are pronounced with three identified loss periods.4  Most
of the nitrogen and phosphorus loss occurred when the plants were newly estab-
lished and correspond to plowing and soil cover treatments evaluated in Section
8.6.

                                    8-10

-------
      The  24 cleanup treatments are potential treatments for a variety of
  soils  and crops grown in the desert, grassland, or forest.  Representative
  groupings of species for different purposes are shown in Table 8-4 with the
  environment in which they are being produced.  A number of the contaminant
  removal techniques include soil removal to a specified depth and these depths
  can be compared to the optimum root zone shown in Table 8-5 that each species
  will permeate in deep soils.

  8.3.2  Soil  Compaction
      The average density of weathered rock fragments in soil is approximately
  2.5 times the density of water.   Taking the density of air as close to zero,
 compared to  rock,  and a mixture of 50 percent air  and 50 percent  rock,  the
 average density of the mixture would be expected to be about 1.25 times  that
 of water.   For most soils with rock minerals as their principal  ingredient,
 the density  is near 1.3.   In these soils there  is  a little more rock  than
 air if  the density is  judged from this  discussion.
      Our hypothetical  soil  with  50 percent  air  would  be  a  mass of small  rock
 fragments, with  holes  among  the  solid particles for the  air.  The rock frag-
 ments would not  float  because  they are  about  2.5 times heavier than water.
 After a rain part  of the  air space is filled with water; in  fact,the  escape
 of air  can be  seen where  water is  held  on soil  as in  basin  irrigation.   In
 most soils the water wets the weathered  rock particles and  lubricates them
 with the result that they slide across one another much easier wet than when
 dry.

     The objective  in compacting soils is to squeeze  the particles close to-
 gether  so  the volume occupied by the soil is smaller, and the density is more
 like rock  than air.  When the density is near 1.5 to  1.7 plant roots do not
penetrate  and plant growth is halted.  The water sprayed onto soil during
compaction is intended to lubricate the particles so they move across one
another under pressure from equipment.   A list of equipment and  the depths
to which it compacts the root zone is given in Table 4-2 of Chapter 4.  These
are the depths to which the soil  must be broken apart  for crop production if
the soils are intended for agriculture.
                                    8-11

-------
             Table  8-4.   Commercial  crop  species  in managed  ecosystems.
 Ecosystem
    Forage/pasture
Row crops/field crops
    Orchards
 Desert       alfalfa/ladino clover
 Grasslands   sweet clover/fescues
             alfalfa/wheatgrasses
                    home grass

 Deciduous    clovers/1espedosa
  forests    timothy/bluegrass
                       sugar beets/dry farmed     citrus
                       dry beans/dry farmed       apricot
                       vegetables/dry farmed      peach

                       corn       /winter wheat   household
                       soybeans   /spring wheat     plants
                       sorghum    /barley
                       sugar beets/rye

                       corn       /spring wheat   apple
                       soybeans   /oats           pear
                       tobacco                    peach
                       vegetables
Table 8-5.
Rooting depths of agricultural crops in fertile, deep, well drained
soils.*
Use/species
            Rooting depth
               observed
               (meters)
      Use/species
Rooting depth
  observed
  (meters)
Forage
alfalfa
sweetclover
clover
timothy

Pasture
ladino clover
fescue
wheat grass
brome grass
1 espedosa
blue grass
Row Crops
cotton
sugar beets
dry beans
corn

3-4
3
1-2
1


0.5
1-2
3-4
1-2
1
0.5

1.5-3
1-2
1.6-1.8
2
Field Crops
wheat
barley
rye
oats
flax
Orchards
citrus
peach
apricot
apple
pear

Row Crops
sorghum
soybeans
tobacco
vegetables

2-3
2-3
2-3
2
2

2-3
2
5
8-10
2


2-3
1.5-2
1-1.5
1-4
*Examples are irrigated desert and tall-grass prairie.
                                    8-12

-------
 8.3.3  Compaction and Runoff
      Any description of hydrologic effects originating from clearing large
 tracts of cultivated land and then resurfacing with an impervious layer de-
 pends on having generic predictors available.  The net hydrologic effects
 from resurfacing do have an area-dependent generic factor.  Intuitively, the
 surface-storage of runoff is related to precipitation intensity and to local
 topographic features.  Resurfacing a flat, downwardly concave area creates
 runoff storage capacity that may enhance groundwater supplies locally,  but
 resurfacing the rolling topography that formerly supported deciduous forest
 has variable outcomes that are directly dependent on the area involved.
      Flood water runoff on the northeastern belt of the former deciduous
 forests of Ohio was  most highly correlated with watershed area.   However,
 there were two distinctly different area-dependent watershed classes,  so
 different  that square kilometers of area drained was  used as the  criterion
 for dividing the classes.   The zone separating the two  flood classes was
            <~\
 50 to 75 km .   Runoff volume correlated best  with drainage area size, rain-
 fall intensity,  and  soil  infiltration  rate.   These three  variables were  sig-
 nificantly  different  from  random correlates  (probability  of random correla-
 tion being  less  than  1 percent  for  all  three  variables  taken together or
 individually).

      The effects of compacting  soil on  local  runoff from  micro-areas are
 well  known.  Examples are  streets that  require  storm drains and diversion
 structures,  parking lots, building  sites, and construction  areas in general.
 Bar-Kochba and Simon  rated  infiltration rate into surface  soil as the most
 important non-climatic variable  in their eastern Ohio study.  Reduction of
 surface water infiltration rates has the obvious effect of  increasing runoff
 and reducing groundwater recharge rates.  This effect is  less damaging upon
 10-km  areas in humid precipitation regions than in semi-arid zones where
 agriculture depends on pumped water for crop production.  Thus, the effect
 of a compacted area in restricting groundwater recharge increases  as the hypo-
 thetical compacted area is moved westerly along a decreasing precipitation
gradient.
                                    8-13

-------
8.3.4   Infiltration and Crops
     In an intensively studied area, eastern Ohio, the average water infil-
tration rate for grassed areas was more than twice the intake rate of adja-
cent row-cropped fields.   A soil-series-dependent influence on intake rate
was measurable but the relative infiltration rates were consistent among the
                                                      2
several soil series within the approximately 40,000-km  study area.  The
higher infiltration rate suggests why runoff is reduced and groundwater re-
charge  increased by grass cover in deciduous forest agriculture.  Grass is
the preferred cover for drainage waterways because the stems reduce runoff
flow velocity and the abundant fibrous roots keep soil aggregates in place
against detachment.

8.4  EFFECTS OF CLEANUP PROCEDURES ON AGRICULTURE
     Qualitative descriptions of the effects that each of the different
cleanup treatments have on agricultural land are presented in this section.
The recovery of longer-lived agricultural crops to their precleanup produc-
tivity is briefly described in Section 8.5.  Quantitative assessment of the
impacts of cleanup treatments on agricultural systems is given in Section 8.6
in terms of the estimated times required for return to commercial productivity,
     The cleanup procedures described in this section have essentially three
different kinds of functions:  (1) treatments temporarily holding the con-
taminant in place, (2) treatments physically removing contaminants which
depend upon stabilization after the cleanup to prevent wind or rainfall ero-
sion until rehabilitation occurs, and (3) restriction of access into the
area by humans and animals.
     The projections made in this section are predicated on the information
which was available at the time the literature survey was completed.  Serious
revisions of these recommendations or rejections of the projected consequences
of the cleanup in agricultural lands as more information is obtained appear
unlikely.   However, a number of publications will soon be available which are
indirectly related to cleanup techniques in agricultural areas.  Surface min-
ing in the Western states has resulted in a great many techniques being per-
formed for rehabilitation and reclamation of semi-arid and arid lands.   These
techniques are experimental  at present and are being evaluated for their

                                    8-14

-------
  effectiveness  at  the  different  sites  involved  in  surface  mining.   In a  few
  years  these  experimental  methods  will  have  been published in  journals which
  carry  reports  of  experimental research,  and subsequently  they will  be in-
  corporated into books  and reports  that will be of use  to  individuals  con-
  cerned with  agricultural  land cleanup after a contaminating event.  An  exam-
  ple of the kind of material to  be  anticipated from surface mining rehabili-
  tation is the  statement in the  recently  released  operating handbook published
  by Coal Age.   This handbook stated in its Foreword:
           This Coal Age operating handbook  of underground mining
           is the first of a series of books planned to cover key
           areas of present day  coal mining  operations.  Future books
           in this series will cover strip mining and reclamation,*
           coal preparation, management, and maintenance"]	
      As noted in Section 8.1 the agricultural areas described as impacted
 represent the irrigated southwestern desert, the short grass irrigated and
 dryland agriculture on the west  side of the American prairies and grasslands,
 and from the  midcontinent  east to  the northeastern middle-west of the United
 States  formerly covered by hardwoods and  river  bottom deciduous forests.
 The desert  area represents a transect  across the American  Southwest,  extend-
 ing east  from the  Imperial Valley  of California  to those irrigated parts of
 New Mexico  and  southwest Texas.
     The  impact of the  cleanup treatments on domestic  animals  is  not  covered
 here.   It is  presumed that rehabilitation of contaminated  areas for  produc-
 tion of agricultural crops is a  prerequisite to any animal  husbandry  enter-
 prises.  The  general crops for which cleanup impacts are described generally
 fall within three  groups of cultivars:  (1)   dryland and irrigated forage
 crops,  (2) commonly produced row crops and small grains, and  (3) tree  crops
 ranging from  the citrus orchards of the irrigated  desert to the stone  fruit
 and pome orchards in originally  deciduous forest areas.

     (0-1)  Natural Rehabilitation
     Unlike natural ecosystems, agricultural ecosystems have no capacity to
replace the destroyed commercial  crop by the same species  after cleanup.
Since agricultural  crops are grown  at an energy cost,  when  this energy is

*Emphasis added.
                                    8-15

-------
 no longer supplied to the system the crops will no longer grow in commer-
 cially productive quantities.  Waiting for natural rehabilitation to replace
 an agricultural crop is an unacceptable procedure.  The time course of re-
 turn from agricultural to natural ecosystems (Old Field Studies)  has been
 defined in Chapters 1, 2, and 3 under the heading Man Made Perturbations.
 Further details for recovery sequences are in each of those chapters in
 sections headed Recovery after Cleanup.  Those descriptions should be re-
 viewed before forecasting the vegetational changes that will occur on aban-
 doned farmland.  Ecologic journals are replete with flora succession studies
 conducted on Old Field farmlands;  they do not return to cultivars.

      (1-1)   Chem-ioal  Stabilization
      Containing the contaminant to diminish its redistribution  has  an imme-
 diate and obvious  benefit in both  hazard  reduction and in minimizing the  size
 of the area  of ecological impact by later cleanup  methods.   The function  of
 chemical  stabilization in an agricultural system is  to prevent  redistribution
 of the deposited contaminant.   More than  seventy stabilizers are  reviewed in
 Appendix  A;  those  suitable for  use  in  agricultural  fields have  been  desig-
 nated in  Table A-l.   With the exception of water-soluble  emulsions,  applica-
 tion  of a preferred chemical  stabilizer requires a number of preparatory
 steps before  the field can be treated.  The area to be treated  must  be  cleared
 of vegetation or clearcut, stumped,  and grubbed  for tree  crops, the  slash and
 vegetative refuse  removed, and  the  fields  smoothed and either graded or shal-
 low plowed.   Most  of  the  "preferred" stabilizers require  direct mixing with
 friable soils  before  they  become effective  in preventing  resuspension of  soil
 and contaminant  particles.  Because of  this requirement for  similar pretreat-
ment, the after  effects from most of the chemical stabilizers will be very
 similar on the farmlands treated.
     Chemical  stabilizers which deflocculate soil particles, such as sodium
chloride, sodium hydroxide, sodium carbonate, and Calgon, are unsuitable for
use in western agricultural fields.  Although this class of alkaline and high-
sodium dispersants would stabilize surface soil, they can only be  used in
areas which have high  levels  of precipitation or access to free water in basin
irrigation.   The requirement  for large amounts of water is to assist plants
in withstanding the effects of the saline  soil that would be produced by these
                                    8-16

-------
 stabilizers.  Leaching of the soil for a long period of time before green
 manure crops would compound the damage these stabilizers produce by removing
 free salts, resulting in the deflocculated surface soil becoming impervious
 to water infiltration.  Sodium adsorbed on clay disperses mineral particles
 and prevents water infiltration.   Lack of soil moisture prevents subsequent
 plant growth.   Growth of plants prior to extensive water leaching has been
 found the simplest way to remove the effects of sodium deflocculants and
 reconstruct soil tilth.   Surface incorporation of high quantities of organic
 matter into the soil is  mandatory at the time recovery is instituted.   Crop
 residues from plants produced for organic matter (green manure), or addition
 of mulch which is worked into the surface soil are standard procedures.
      Thirty-six "preferred" stabilizers for agricultural  use are listed in
 Appendix A;  29 of these  require reseeding or replanting of the crops to  which
 they are applied.   The remaining  seven would not destroy the crop to which
 the stabilizer was applied.   Appendix A has an additional list of 22 "accept-
 able" stabilizers that are  acceptable as alternates  for use when the pre-
 ferred stabilizer cannot be used.
      Only 9  of the 58  stabilizers  can be applied directly on growing crops
 without  totally endangering their  commercial  value.  Two of the  nine plant-
 safe  stabilizers  require special handling for cropland  application;  three
 of  the nine  that  are safe for plants  are  water  emulsions  of asphalt or
 resin, and four  are colloidal suspensions of  protein.

      (1-2)   Clearautting Vegetat-ion
      Clearcutting  orchards  has the same  effect as clearing  coniferous and
 deciduous forests  insofar as crop production  is  concerned.   It is very
 improbable that the operators of commercial orchards would  depend on stump
 sprouting, for those species in which  it  occurs, as a method of creating
 a replacement orchard.   It is more likely that the trees would be removed
 in toto  (treatment 1-3).   Clearcutting of irrigated and nonirrigated for-
 age is the normal harvesting procedures and would have no significant im-
pact  in cleanup considerations.
     Row crops would be eliminated by Clearcutting if undertaken before the
normal harvest time.  Clearcutting is the method of harvest for a number of

                                    8-17

-------
 row crops.  Examples are machine-harvested tomatoes, combine harvesting of
 all the small grains, the harvesting of the various species of maize and
 sorghums, and the mechanical picking of cotton crops.
      The physical structure of agricultural soils would be unaffected by
 this treatment.   Organic matter content would be undisturbed and the fer-
 tility levels unchanged.  There would be no change in  the  suitability of
 seed beds for subsequent crop production.

      (1-3)  Stumping and Grubbing
      With the exception  of orchards,  this  treatment is inappropriate for
 agricultural crops.   The effects of  stump  chippers and root  rakes  on east-
 ern orchards would be similar to the  effects  described for deciduous forests
 (Section 3.4).   This technique would,  of course,  eliminate any productive
 capacity of  an orchard by eliminating  the  producing plants.
      The physical  effects on the soils in  the eastern  United States  would
 be  due  primarily to  mixing the fertile surface  layers  with the relatively
 infertile subsoil  layers which the new tree roots  would normally penetrate.
 For desert orchard soils,  there  would  be a  small  increase  in fertility of
 the lower segments of the profile and  little  decrease  in fertility of the
 upper parts  of the profile.   Irrigated desert citrus would be eliminated
 as  a crop, but the effects upon  the soil structure  would be  relatively mild
 for most  undeveloped  irrigated desert  soils.  The principal  post-cleanup
 effects would be the  reduction in the  fertility of  the profile utilized  in
 the new mixed soil surface.  This would need  correction before reestablish-
 ing orchards in the eastern United States.  A secondary impact on eastern
 orchards would be  displacement of wildlife  from hedgerows and fencelines
 surrounding the orchards.  The measure of any economic impact would be the
potential  for a reduction  in production of  those crops dependent upon dis-
placed animals for pollination.

     (1-4)  Scraping and Grading
     The consequences of surface scraping and grading upon any agricultural
crop is seasonally dependent.  Performing this operation when fields have
been harvested and are normally clear of crops would result in removal of

                                     8-18

-------
 substantial portions of the limited organic matter in the desert.  It would
 have essentially no effect in the grassland areas but clearcutting would make
 it be an extreme perturbation in eastern orchard areas and on any of the grow-
 ing eastern row crops.  Surface soil would be laid bare to wind and water ero-
 sion.  If the soil were moist at the time of scraping and grading (presumably
 by heavy equipment such as road graders, land scrapers, and bulldozers), sig-
 nificant compaction would be likely to result (the effects of soil compaction
 are described in cleanup treatments 2-3, 2-4, and 2-5).
      The physical-chemical properties of most agricultural soils would not
 be greatly changed by scraping and grading (excluding compaction effects).
 The principal effects would be increased susceptibility to wind or water ero-
 sion at  the loosened surface.   However,  there would  be little opportunity for
 excess  sedimentation since the basic runoff water diversion structures would
 be unchanged.

      (1-5)   Shallow Plowing
      Shallow plowing would eliminate an  unharvested  cultivated crop and,  if
 the crop  were  near  maturity, would  create  a trashy surface.   In the noncrop-
 ping season its  impact  would be  minimal  or  insignificant  since plowing  is  a
 normal operation of agriculture.   Plowing with disc  equipment  would have  lit-
 tle impact  on  standing  orchards  in  either  irrigated  or  rainfed agriculture.
 The most  likely effect  would be  soil  compaction  in the  wheel  tracks, and  fur-
 row bottoms  resulting from  the use  of plowshare  equipment,  if  the  soil were
 significantly  moist  at  the  time  operations  were  carried out.
     The  effects overturning the plow layer has  on the  seed sources, vegeta-
 tion, and soil erosion  processes of  semi-arid rangeland sods are overcome in
 one or two  seasons of regrowth.  Shallow plowing of less than  10 cm has little
 impact on survival of grasses but afterward the roughened surface is unsuit-
 able for stabilizing by mechanical compaction because the surface is trashy.
 Clays and deflocculants would fall among the sod chunks and be lost to subse-
 quent stabilization techniques.
     Runoff is reduced by the increase of surface storage in the roughened
area but groundwater recharge in the western shortgrass prairie would  be un-
changed  because the subsoils are permanently dry. Surface erosion by wind

                                     8-19

-------
 and water would be unchanged as grass cover is the cure recommended for both
 erosion mechanisms.
      Damage to woody plants is of short  duration.   Seed stocks would remain
 and stump and root sprouts  would soon revegetate  the  plowed rangeland.   Ani-
 mal grazing on the new growth may be  a problem in establishing regrowth.

      (1-6)   Deep Plowing
      Plowing to a depth of  40 cm would entirely disrupt any of the  agricul-
 tural  crops during their  growing season.   If  the  land were  fallow,  it would
 eradicate standing orchards,  pastures, and  rangelands.   The fallow  season  im-
 pact  in areas with row crops  and small grains would be  relatively small  and
 approximately analogous to the consequences  of shallow plowing  (treatment 1-5).
 Its effect  in pastures  and  rangelands  would be a  reduction  in  productivity
 of  one  or more growing  seasons.   On shallow profiles  in the short grass prai-
 ries  it is  likely to  produce  a mixing  of the  A- and B-horizons, reducing plant
 growth  in subsequent  seasons  unless fertilization, and  possibly reseeding  in
 the drier areas  (less than  35  cm mean  annual  precipitation), is carried out.

      (1-7)  Soil  Cover Less than 25 am
     The  significance of the  impact on agricultural crops is a direct function
 of  the  quality of the soil used  for the cover.  A soil  high  in clay would  sig-
 nificantly  impair  aeration  in  orchard crops and production decline would begin
 in  a few  months.   Attempts to  add soil to areas with row crops would be im-
 practical in  terms of attempting  to save the  crop from  the equipment.  The
 minimum impact would occur in  those areas which are grassed—either rangelands
 or  irrigated  grasses.  Sweetclover forages would probably survive but the  sur-
 vival of  irrigated forages,  such as alfalfa and ladino  clover,  is uncertain.
 The basic soil fertility in the root zone would be unchanged by this depth of
 cover.  The principal effects on seed bed suitability for cropping would be
 fertility changes  from the soil added.  The soil used for covering would prob-
 ably not be equal  in tilth and fertility to the soil it covers  up.

     (1-8)  Soil Cover 25  to 100 cm
     It is improbable that commercial  agricultural crops would  survive the
addition of a soil cover of  this depth.
                                     8-20

-------
       (.2-1)  Remove Plow Layer

       This layer is about the minimum depth that can be removed as a controlled
 cut by ordinary heavy earthmoving equipment.  Soil removal would create a
 radical disturbance in agricultural land, almost regardless of the crops that
 were  involved.  It would totally destroy rangeland grasses, forage crops,
 irrigated forage crops, small grains, and standing tree crops normally found
 in orchards.  It would represent an especially severe disturbance east of the
 2-cm runoff isogram (Figure 8-1).  In the short grass prairie this treatment
 accomplishes complete removal of the A- and a substantial part of the B-horizon
 with the corresponding loss of the normal organic accumulation upon which the
 fertility of the land is based.   Removal of the organically rich A-horizon in
 this area would leave the soil with a severe nitrogen deficit.
      It is unlikely that any of the standing orchard species would survive
 removal of the surface 6 inches  of soil in the long term,  outside of irrigated
 areas.  Minimum effect would be  anticipated in those deep-lying desert soils
 that are normally  irrigated for  crop  production.   These  generally coarse-
 textured soils allow  a more rapid deep  penetration of plant nutrients  from
 surface and  shallow subsoil applications into  the  soil profile than  occurs
 with nonirrigated  agriculture.   However,  removal of this  layer in irrigated
 agriculture  would  destroy row crops produced there.

      (2-2)   Remove  Shallow  Boot  Zone
      On  the  shallower  soils,  removal of  up  to  40 cm of soil profile would
 eliminate  the  fertilized root  zone where  the major  dry-farmed  crops are grown
 (i.e.,  the infertile and highly  calcareous regions  of  the  short grass prairie
 and  in  some  of the  caliche  areas underlying irrigated  deserts).   The resulting
 truncated  soil profile would more nearly resemble an unpaved roadway after
 deep cuts to maintain a uniform  grade in hilly terrain.  Shallow  root zone
 removal in the tall grass prairies would have relatively little effect since
 the profile would still contain  significant depth and  amounts of organic mat-
 ter below the soil cut away.  In crop lands formerly occupied by deciduous
 forests, this treatment would destroy the agricultural potential of many of
 the farmed areas, eradicating the entire organic matter content of those soil
profiles as well as any residual artificial fertility elements applied to pro-
mote production of crops.  The phosphate-rich plow layer  would be removed and
the surface remaining  would  be more suitable for nonagricultural purposes.
                                     8-21

-------
      Prevention of erosion is the immediate problem in soil management when
 the root zone has been stripped to a depth of 40 cm.   West of the 2-cm runoff
 isogram surfaces not naturally compacted or cemented  are vulnerable to wind
 erosion.  East of the 2-cm isogram,  post-cleanup surface erosion by runoff
 water would be a severe problem throughout the stripped areas.   Uniformly
 removing 40 cm of soil surface would impair runoff diversion structures nor-
 mally incorporated into farmlands.   Removal of the shallow root zone west of
 the 2-cm runoff isogram would leave  a smooth erodible surface in the place
 of the roughened trashy surface normally utilized to  prevent wind erosion;
 that is, planting ridges,  grass waterways,  lister ridges,  shallow terraces,
 and the basic contour structures would be removed.  The basic contour struc-
 ture could  be maintained if care were taken during the soil  stripping to take
 the soil along field contours previously constructed  for dry farming and
 irrigation.
      The most severe impact in the excavated agricultural  areas would be soil
 erosion.  Other serious  impacts relate to  crop production,  i.e.,  loss of the
 soil  organic  matter,  and hence loss  of soil  fertility.

      (2-3)  Remove Scraping and Grading, Mechanically Stabilize
      This treatment  would have little  impact  on the internal  structures  nor-
mal to  control  of overland  water  flow  in agriculture.    The dikes, berms,
drainage ditches, grassed waterways, and overfall  structures normally present
would be devegetated but not  basically  eradicated.  Scraping and grading on
grasslands in pasture- and  rangelands would remove the  surface organic matter
layer,  subsequently reducing  nitrogen availability in the grasslands.  The
principal consequence would be reduction of sedimentation onto lands  adjoin-
ing the cleaned up area.  Scraping and grading, removing the loosened mate-
rial, and stabilizing the surface in areas as large as  10 km2 would signifi-
cantly reduce sediment loss that would otherwise occur from scraping and
grading large areas.
     The effects of this treatment and removal of deeper layers of soil to
40 cm were previously described in cleanup treatments  2-1 and 2-2.
     The basic physical effects of this cleanup treatment on desert, grass-
land, and crops grown on the formerly deciduous forest lands have been de-
scribed in Chapters 1.4, 2.4, and 3.4  (under cleanup  treatment 2-3) to which
                                    8-22

-------
 the reader is referred for added detail.  The effects and probabilities of
 wind erosion and runoff erosion by surface flow are discussed there and are
 not repeated here.  The descriptions in Chapter 2  (Prairies) are especially
 pertinent to cleanup in areas producing grasses under dryland conditions and
 for both native and reseeded vegetation on rangelands.
      The number of mechanical stabilization procedures that can be used in
 agricultural areas to be returned to croplands is limited.  The slopes in
 cropped areas are not usually steep enough to require that meshes,  nettings,
 or plastic films be laid in the areas involved.   Furthermore,  the areas to
 be protected in farmed fields may be so large that substantial problems would
 arise from attempts to use plastic films as covers.  In a relatively short
 time the plastic would rip and tear and lose its effectiveness in containing
 contamination.   The otherwise usable mechanical  stabilizations would subse-
 quently require removal  of an additional layer of soil because they are
 variations of the compaction  process.
      The success of compaction methodology is directly dependent  on the
 amount  of clay in the  soil  which  is  compacted.   It is  unsuccessful  in  the
 coarse-textured,  sandy soils  frequently found in the  irrigated desert  where
 reduction of  pore space will  be sufficient to destroy  agricultural  crops.
 Removal  of shallow  soil and mechanical  stabilization by  compaction  in  an
 orchard  would kill  the trees  in a  short  time  due to the  combined  effects of
 poor  aeration and severe moisture  deficits brought  about  by preventing pre-
 cipitation  infiltrating into  the soil profile  of the orchard.   If there is
 enough clay in the  surface soil for compaction to  be effective, compaction
 would be  relatively permanent  in agricultural terms.   Return of the area to
 agricultural  productivity would require  rippers, subsoil  chisels, repeated
 plowing,  and  other  action to break apart the compacted layer and free  the
 soil aggregates from the effects of prior  compression.  The offsite surface
 flow of runoff would have a major impact on adjacent areas.  For areas  as
              2
 large as  10 km , uncontrolled runoff could be a significant factor down-
 stream.    It is also probable that there would be a drastic reduction of the
 groundwater in an area where 10 km2 was mechanically compacted for stabili-
 zation purposes.  It is unlikely that deflocculants such as sodium chloride
would be tolerated for lands that  were to be returned to agricultural pro-
duction,  so mechanical  stabilization or short-term chemicals are needed.
                                    8-23

-------
     (2-4)  Remove Plow Layer^ Mechanically Stabilize
     The physical effects on the suitability of the cleaned-up area for agri-
cultural crops under this cleanup treatment are essentially the same as those
described under cleanup treatment 2-1.  The effects of mechanical stabiliza-
tion would be very similar to those described in treatment 2-3.  The princi-
pal differences relate to changes in organic matter and soil fertility induced
by compaction.  Where there is a relatively high water table and a high or-
ganic matter content below the compacted area (e-g-j from a large root mass),
the fields would be subject to anaerobic decomposition of the organic matter
and the nitrogen would be reduced to ammonia forms with the potential for its
slow volatilization out of the soil profile.  Effectively, the nitrogen con-
tent of the cropland would be reduced compared to that which existed prior
to the cleanup, as in treatment 2-3.  Crop production could not be resumed
until the compacted area had been broken apart by mechanical means.

     (2-5)  Remove Shallow Root Zone, Mechanically Stabilize
     The effects of this treatment, insofar as agricultural use of the land
is concerned, are very similar to those for plow layer and shallow root zdhie
removal.  Soil fertility would be reduced more than with removal of the plow
layer.   The deeper truncation of the soil profile would render revegetation
of the rangeland or native pastures unlikely without renovation in the semi-
arid short grass prairie region.  The desert soils on alluvial fans and
deeper profiles would be least affected by this treatment.  The deeper cut-
ting is likely to have significant effects on surface runoff control struc-
tures in the sub-humid and humid agricultural regions.  These control and
diversion structures would have to be reinstituted before surface stabiliza-
tion was undertaken if severe erosion effects are to be avoided.  The same
soil preparation methodology as described for treatments 2-3 and 2-4 would
have to be undertaken to neutralize soil compaction disadvantages prior to
reestablishing agricultural production.

     (.2-6)  Remove Scraping and Grading, Chemically Stabilize
     The physical effects upon agricultural production from this treatment
are about the same as the combined effects of cleanup treatments 1-4 and 2-3.
The effects of chemical stabilization would not be as long-lasting as the

                                    8-24

-------
 mechanical stabilizing treatments, and in a number of cases immediate reveg-
 etation with some variety of the grasses is required where runoff may cause
 surface erosion.   Infiltration into the soil profile is controlled by the
 soil profile physical and chemical characteristics rather than by the chemi-
 cal stabilizer applied;  hence, it is unaffected by this treatment.
      The most suitable chemical stabilizers to use in an agricultural area
 are those which break down without leaving toxic residues that would impair
 food values or plant growth at a later time.   Some of the chemical stabili-
 zers described in treatment 1-1 would be the stabilizers of choice in this
 treatment also.

      (2-7)   Remove Plow  Layer.,  Chemically Stabilize
      This is a composite technique described  separately under  treatments  1-4
 and 2-6.   The overall consequence is one of reduced soil fertility with seed
 bed conditions less  suitable for reintroduction of agricultural  crops than
 those conditions  resulting from treatment 2-6.

    .  (2-8)   Remove Shallow Root "Lone,  Chemically Stabilize
      Slicing away the surface  profile  to  a depth of approximately  40  cm would
 require the  greatest  application  of chemical  stabilizers.   Loss  of the or-
 ganic  debris  leaves  the  grassland soils  susceptible to  wind erosion,  and re-
 quires immediate  reseeding with grasses to  stabilize the soil  against runoff
 in  sub-humid  to humid regions.   Removal of  the  loosened  soil minimizes sedi-
 mentation problems downstream  in  the runoff area but increases flooding po-
 tential from  runoff out  of areas  as  large as  1  to  10 km2.

      (3-1)  Barriers  to Exclude People
     The most effective  fencing for the exclusion of people would be a chain
 link fence 6  to 10 feet high posing a physical restraint to trespassers.

     (3-2)  Barriers to Exclude Large Animals
     Design of fences to restrain cattle  is a well developed technology.   Re-
 straining of  large game animals,  such as elk and moose,  is not likely to be
a problem in most agriculturally developed areas.  Preservation of mechani-
cally and chemically stabilized surfaces would require that large animal

                                    8-25

-------
traffic be restricted.  The hooves of domestic cattle and sheep would rapidly
break down either of the stabilized surfaces, especially since the animals
are prone to travel single file which concentrates the cutting action of the
hooves in a small path.

     (3-3)  Barriers to Exclude Large and Small Animals
     Attempts to eliminate small burrowing animals, such as ground squirrels
and gophers, would probably not be successful in large areas.   Containment
of these animals within the area would be more successful than elimination.
Consequently, fencing should be directed towards retention of animals pre-
sumed to have become contaminated in the area rather than to keep small mam-
mals out of the area.  Horizontally oriented small-mesh fencing in ditches
around the cleaned-up area and backfilling to provide a mesh ledge will pre-
vent the escape of both large and small burrowing animals.  Because these
animals have a relatively limited home range there is not much danger of
them transporting large quantities of contaminants significant distances.
Fencing of successive containment areas may be the most practical method to
deal with burrowing animals.  It is assumed that larger animals can be suc-
cessfully excluded from contaminated areas until the appropriate cleanup
treatment has been selected for the area.

     (4-1)  Asphalt Hard-Surface Stabilization
     Consideration of the use of this treatment in an agricultural area
should be based on the understanding that it will be a very temporary cover.
Since the surface in most agricultural areas will inevitably be disrupted
by stray cattle at some time soon after the material is laid down, it cannot
be considered to be a permanent way of securing a contaminant.  In areas
where surface soil experiences freezing and thawing, such as the grassland
and deciduous forest areas, frost heaving is likely to be highly destructive
to an asphalted or road-oil treated area in a relatively few seasons.
     Application of hot road oil or asphalt to an existing crop in agricul-
tural production would be disastrous.  In most circumstances the standing
agricultural crop would be removed, with the possible exception of tree
crops.  An agricultural crop must be considered a total loss if the appli-
cation of asphalt or hot road oil is necessary to hold the contaminant in
                                    8-26

-------
 place until subsequent cleanup treatments are undertaken.  Few agricultural
 crops can resprout and puncture heavy asphalt surfacing.  The more damaging
 effects would include compression tracks introduced by the heavy equipment
 needed for transporting asphalt or hot road oil across the areas to be
 treated.   Reclamation of the sites will be an extremely expensive process
 since the oil-soaked or asphalt layers must be removed before agricultural
 production can be resumed.

      (4-2)   Concrete Hard-Surface Stabilization
      This is the most severe cleanup treatment proposed for use in agricul-
 tural areas.   The expense of removing the concrete  at  a later date almost
 entirely  eliminates  use of this cover from consideration.   It is inappropri-
                                            o
 ate  for use in areas as large as 1 to 10 km .   The  transportation costs  of
 recovering  and disposing of the large amount  of concrete that would be re-
                        2
 quired for  even 0.01 km  is almost certain to eliminate it  from serious  con-
 sideration  in areas  where agriculture is to be resumed.

      (5-0)  Application of Sewage Sludge
      The  properties  of  urban sludge  are  significantly  different  from farm
 manures and crop  residues.   The use  of sludge  requires highly specialized
 information in agricultural  management practices.   Outright  recommendations
 for the application  of  sludge  are unsafe.   Urban sludge  is  still  suspected
 of containing  viable viruses and  must not  be allowed to  come  into  direct
 contact with the  edible  portions  of  plants.
      Sludge contains high quantities of  industrial  wastes at  times.  Such
 wastes contain  heavy metals which  are poisonous to man:  lead, cadmium,
 chromium, copper, and nickel.   Other elements hazardous  to man  (unless proper
 soil management practices are used)  are  aluminum in various compounds, ar-
 senic, selenium, antimony, and mercury.  Wastes from industrial electroplat-
 ing processes are extremely high  in  these particular metals.

     Sludge supplies moderate amounts of nitrogen to cropland and increases
the organic matter content of soils.  Sludge application is not safe for
crops grown on acid soils of the former deciduous forest areas for either
human consumption*or as animal feeds.  However, the low solubility of the
                                    8-27

-------
heavy metal elements in neutral and alkaline soils minimizes these hazards.
Sludges contain from 1 to 6 percent nitrogen on a dry-weight basis.  Moder-
ate amounts of sludge would supply nitrogen to croplands and significantly
increase the organic matter content of the soil.  Five to 40 tons of dry
matter per hectare will supply nitrogen and organic matter in the alkaline
grassland and desert soils.  A decay series worked out for breakdown of the
organically contained nitrogen indicates that approximately 25 percent of
the nitrogen will be available to plants in the first four years the sludge
is in the field.  Approximately 30 to 60 percent of the nitrogen in the
sludge is in an inorganic form and would be readily available to plants in
the first year after the sludge application.  This soluble nitrogen poses
a potential hazard to groundwater supplies.  For this reason, large excesses
of sludge should not be applied in renewing the organic content of soils
following soil removal, scraping and grading of desert soils, or the addi-
tion of soil covers at a depth of 25 to 100 cm over croplands.

     (6-1)  High Pressure Washing (<3 mm)
     This cleanup treatment is outside the scope of work for this report.

     (6-2)  Flooding to 30 cm
     One-time flooding for redistribution of the contaminant into the soil
profile is outside the scope of work for this report.

     (7-0)  Soil Amendments Added
     Experiments in which a variety of materials have been added to soils
for the purpose of preventing plant uptake of hazardous material have been
unsuccessful.  Contaminants that are adsorbed onto the exchange surfaces
of soil minerals become available to plants either by contact exchange be-
tween the mineral surface and the root or by normal solution processes.
Once a contaminant is no longer retained on the surface soil but has, in
fact, penetrated the root zone of agricultural crops, the suggested reclama-
tion procedure is one of management that will produce luxuriant vegetative
growth that is discarded.  This is accomplished by normal management prac-
tice, soil liming where needed in the deciduous area, and large additions
of nitrogen, potassium and phosphate to promote luxuriant plant growth where

                                    8-28

-------
 water supplies are adequate.   In the prairie and desert areas the limiting
 factor in land reclamation is the water supply.   Attempts to immobilize
 heavy metals to avoid plant uptake in the deciduous area have been unsuc-
 cessful.   Breakdown of the sludge releases nitrates,  sulfates,  and other
 acid forms which will readily consume the added  lime and result in highly
 acid soils from which plants  readily extract the injurious metals contained
 in the sludge.

 8.5  RECOVERY FOLLOWING CLEANUP
 8.5.1  Irreversible Changes
      The  complement of cleanup techniques discussed in  this report has  no
 irreversible changes.   The limits to recovery of croplands are  economic;  the
 question  is  whether to rehabilitate  the spill area,  soils,  and  crops.   The
 limits to the use  of the most damaging  cleanup practices,  asphalt (treatment
 4-1)  and  concrete  (treatment  4-2), are  the cost  and effort required to  in-
 stall the asphalt/concrete cover  and remove it later.   The soils  under  these
 two hard  surfaces  are damaged more by compaction than by seepage  of chemi-
 cals.   Compaction  can be rectified mechanically,  as highway reconstruction
 and relocation  have shown.

 8.5.2  Rates  of Recovery
      In this  chapter  all the  cropping systems that  are  discussed  have been
 created by energy  input by man.   The  rates  of recovery  from cleanup are
 directly  proportional  to the  energy  input  expended  for  rehabilitation in
 agricultural  systems.  None of the agricultural  systems described will re-
 habilitate themselves  to commercially viable levels once cleanup has been
 performed.  Agricultural systems do not follow the concept  of species suc-
 cession through  intermediate  stages to climax as described  in Appendix B
 for natural ecosystems.  They follow the growth pattern of one individual
plant.
      Few of the environmental  insults that may have extremely large impacts
 in natural systems would have  significant impacts in nonorchard agricultural
 systems.  Agricultural systems normally have structures appropriate for con-
trol of erosion by wind or water runoff.  The suggested cleanup treatments
need not destroy structures.
                                    8-29

-------
      Rangeland and pasture grasses would need reseeding and modest nitrogen
 fertilization to recover their productivity.   Precipitation controls the
 rate of recovery to be expected.   Unusually intense rainfall would float
 the seed away and drought would kill the seedlings after germination.   In
 normal weather,  productivity on rainfed land  is regained in 2 to 3 growing
 seasons.
      Row crops,  small grains,  and field crops would need the same management
 they are given each season for normal productivity.   The same recovery scheme
 is appropriate to nontree irrigated crops.  Land leveling for irrigation is
 more severe than any of the recommended cleanup techniques;  only mechanical
 stabilization and asphalt/concrete resurfacing are more disturbing than re-
 sloping a field  for irrigation.
      Orchards would require replanting after  any but cleanup treatments 1-1,
 5-0,  and 7-0.  First fruit bearing would begin in 5  to  6 years and full pro-
 duction by 10 to 12 years,  including citrus.
      The general case for recovery of row crops,  small  grains,  irrigated,
 and  rainfed pastures follows the  sequence imposed by natural  disasters  such
 as drought,  fire,  or flooding.  Tree orchards  likewise  have  as  their recov-
 ery  model  the  growth and  management  steps that would follow  renovation  of
 an old,  declining  orchard.

 8.6   QUANTITATIVE  ASSESSMENT OF CLEANUP  IMPACTS
      The impact  assessment  assumes  in  cleanup  treatments  that  the  contaminant
 is on the  surface  soil  initially,  and  that clearcutting  is carried  out  before
 any of the  other temporary  containment and removal measures  (1-3 through 2-2,
 2-3 through  2-8, and  4-1  and 4-2)  are performed.  Tree crops are assumed to
 be entirely  removed  prior to soil manipulation, including root removal  (treat-
 ment  1-3).
      All erosion control devices for runoff are expected to remain  functional
 in the  eastern hardwood croplands.
      The impacts in this chapter require the intervention of man in almost
 all cases after  a cleanup treatment has been performed.   The multitude of
 corrective actions possible with four kinds of crops  (forages and pasture,
row crops, small  grains, and citrus, stonefruit and pome orchards), 24 cleanup
                                    8-30

-------
 techniques, and three ecosystems (desert, grassland, and deciduous forest)
 requires that the impact analysis be general and easily reproduced.   To that
 end the impact indexing system is shown here for reference.
                                                                   Index
      Response Required to Resume Production                      Number*
        No measurable impact,  no correction needed                  0
        Loss ends  with crop directly exposed to  contaminant          1
        Restored by normal agricultural  practice for  that crop       2
        One renovating technique required  for crop next  year**       3
        Two renovating techniques required for crop next year**      4
        Three or more renovating techniques required  for crop
        next year**                                                  5
        Renovation restores crop production in second growing
        season**                                                   6
        Greater  than  5 seasons  for commercial  crop                 7
        Greater  than  10 seasons  for  commercial crop                8
      The  soil profile and climatic  variables  for  pasture and grassed range
are discussed in  Chapter  2.6.   Removing 40 cm of  soil on the Rocky Mountain
side  of the shortgrass prairie  cuts  into  the  alkaline B-horizons.  In that
region  natural  precipitation is  too  low for maize and orchards.  On  the  Iowa
side  of the grasslands revegetation  will  recover  almost  spontaneously in
rainfed pastures.

      Cleanup techniques that are similar  in impact intensity to normal farm-
ing are mowing  (1-2)  forage and pastures;  shallow  plowing (1-5) and  scraping
and grading  (1-4, 2-1) are similar to land leveling and are readily  erased
by the usual seed bed  preparation in row crops and small grains.  Trees are
entirely traumatized by clearcutting (1-2); thus  stumping and grubbing and
subsequent operations have no additional effect on the tree crop.  Trees in
orchards in the broad  sense are totally destroyed by cleanup.   They survive
 *A11 treatments requiring unusual treatments for regaining productivity are
  equivalent to conditionally retrogressive (Index Number = N+2).
:*Renovating techniques are early reseeding,  early transplants,  annual fer-
  tilizer additions,  mulching,  unusual irrigation, reterracing,  recontouring,
  and other substantial,  unusual treatments such as early summer defoliation
  of orchards.
                                    8-31

-------
 contamination retention processes and soil amendment but not any of the
 standard decontamination procedures where soil is moved about.
      No significant physical impact on the land occurs when barriers (3-0)
 are used, but the economic impact caused to the owner by lack of resource
 availability may be severe.   This is true in all the systems reviewed and
 for normal fertilization (7-0).   Recommended chemical stabilizers (1-1)  have
 a similar impact across the  species and soils and cause the loss of the
 standing contaminated crop from  the sprays and machinery used in the fields.
 Clearcutting (1-1)  forage and pastures is ordinary mowing and likewise costs
 one cutting of one  crop.   On row crops the entire crop seasons'  product  is
 lost from Clearcutting (1-1)  unless it occurs at harvest time.   Sludge ap-
 plications (5-0)  are  risky and surface applications are limited  to  fields
 that grow animal  feeds or to citrus orchards if injected into the subsoil.
      Rendering the  surface impervious  to  precipitation entry has two effects,
 deficits in subsurface stored water and downstream channel  flooding,  whether
 it  is accomplished  by soil compaction  (treatments 2-3,  2-4,  and  2-5)  or  by
 paving (treatments  4-1  and 4-2).   Groundwater in the  treated zone is  cut off
 from local  recharge and channel  flow becomes proportional to the incidence,
 intensity,  and duration of precipitation  once the surface threshold  for  wet-
 ting is  reached.    Erosion control  structures would need to  be maintained on
 all  farms  to handle the increased  runoff  from soil  stabilizers.
      Loosening the  surface soil by  plowing  (1-5,  1-6)  and using  soil  covers
 (.1-7,  1-8)  can lead to  enormous runoff  erosion  losses  but are otherwise  non-
 traumatic to forage, pastures, and  small  grains.   Land  leveling  fills is
 similar  to  treatment  1-4, and  erosion control  terracing and  contouring fields
 are  analogous  to treatment 1-8.
     Hardsurfacing treatments  (4-1 and 4-2) are likely to be acceptable  on
 a few acres if they are removed later or the  land is condemned.  The concrete
 or asphalt residues are not especially harmful to plants where they have been
mixed with the soil.  Removal of slabs might be expensive but hard surfacing
has features of retention that other treatments lack.
     No treatments were judged entirely prohibited on a reclamation  basis,
except some chemical stabilizers.
                                    8-32

-------
      Successful soil compaction increases runoff from rainfall to high per-
 centages.   The proportion may increase to three-quarters of the local pre-
 cipitation, and downstream flooding is assured in areas as large as 1.0 km2.
 The degree of success in compaction sets the magnitude of impact afterward
 for degrees of hazard.   An effective resurfacing depends on low organic
 matter content in the soil and high clay.   These are soil characteristics
 of the deciduous forest  agricultural areas.   Cleanup treatments 2-3,  2-4,
 and 2-5 have been given  consistently high recovery time indices in all non-
 tree crops in deciduous  areas,  and  lesser impact recovery time indices in
 prairie.   Mechanical compaction is  impaired  in prairie fields  by the  high
 organic matter residue in the profile,  even  if 40 cm of topsoil is removed
 as in treatment 2-5.
      Lands in crops  will  have basic erosion  control  structures in place be-
 fore cleanup where wind  or water  erosion is  predictable;  however,  as  the
 area compacted increases  in size, effects  on underlying groundwater tables
 and adjoining water  sheds become  more  severe.   Community flood control
                                     2
 structures needed for 1.0- and  10-km  fields would exist in deciduous  agri-
 cultural zones.
      Desert  crop  system  impact  indices  are shown in  Table 8-6.   Removing
 soil  and chemically  or mechanically stabilizing  the  fields has  relatively
 low impact  because the coarsely textured  soils pack  together poorly and
 plant nutrients are  dispersed vertically throughout  the  profile.   Surface
 soil  removal  has less impact on fertility  than in other  land types, and  a
 lower impact  recovery time  index rating.   Soil renovation practices needed
 for recovery  are ordinary irrigation management  practices  in irrigated
 agriculture.   Soil chisels  and heavy fertilization are common.
     Grasslands in small  grains and  irrigated row crops have deep  lying
nitrogen reserves in the  soil profile and are less impacted than the shal-
 lower profile  on the forested land types.  The relative impact  index for
grasslands under cleanup  is given in Table 8-7.  Soil cover in grasslands
may add weed  seeds and be  infertile, which has more impact than in forested
lands which are of lower  fertility initially.
     Removing the upper profile has severe impact on later crops in the
deciduous forest land type by removing fertility in organic matter and nu-
trients.  The deeper cuts have greater impact as shown in Table 8-8.  Runoff

                                   8-33

-------
Table 8-6.
             Predicted recovery index of irrigated desert agriculture for cleanup treatments (see Appendix B).

Treatment
0-1 Natural Rehabilitation

1-1 Chemical Stabilization
1-2 Remove Vegetation by Clear Cutting
1-3 Remove Stumps and Grub Roots
1-4 Surface Scraping and Grading (<5 on)
1-5 Shallow Plowing (<10 on)
1-6 Deep Plowing (10 to 40 cm)
1-7 Soil Cover (<25 cm)
1-8 Soil Burial (25 to 100 cm)
2-1 Remove Plow Layer Soil (10 on)
2-2 Remove Shallow Root Zone Soil (<40 cm)
2-3 Scrape Surface, Remove (<5 era), Mechanically Stabilize
2-4 Remove Plow Layer (10 cm), Mechanically Stabilize
2-5 Remove Shallow Root Zone (<40 cm), Mechanically Stabilize
2-6 Scrape Surface, Remove (<5 cm), Chemically Stabilize
2-7 Remove Plow Layer (10 cm). Chemically Stabilize
2-8 Remove Shallow Root Zone (40 cm), Chemically Stabilize
3-1 Barriers to Exclude People
3-2 Barriers to Exclude Large Animals
3-3 Barriers to Exclude Large and Small Animals
4-1 Asphalt Hard-Surface Stabilization
4-2 Concrete Hard-Surface Stabilization
5-0 Sewage Sludge Application

6-1 Washing (<3 «n)
6-2 Flooding (3 to 30 on)
7-0 Soil Amendments Added
Forage/Pastures
Area Disturbed (km2)
0.01
NA

i
i
1
NA
4
3
3

3


3
3
3
3
3
3

0
0

5
2

*
*
0
0.1
NA


1
NA
4
3


2
3


3
3
3
3
3
3
3

0
0
0
UA
NA
NA
2

*
*
0
1.0
NA


1
1
NA
4


2
3

2
3
3
3
3
3
3
3

0
0
0

*


*
0
10.0
NA


1
1
NA
5
3
3
2
3

3
3

0
0
0

*


*
0
Row Crops
Area Disturbed (km2)
0.01
NA


1
1
NA
4
3
3
2
3

2
3
3
3
3
3
3
3

0
0
0

5
5
NA

*
0
0.1
NA


1
1
NA
4
3
3
2
3

2
3
3
3
3
3
3
3

0
0
0

NA
NA
NA

*
*
0
1.0
NA


1
1
NA
4
3
3
2
3

2
3
3
3
3
3
3
3

0
0
0

*
*
NA

*
*
0
10.0
NA


1
1
NA
5
3
3
2
3

3
3
4
4
4
4
4
4

0
0
0

*
*
NA

*
*
0
Citrus Orchards
Area Disturbed (km2)
0.01
NA


1
8
8
8
8
8
8
8

8
8
8
8
8
8
8
8

0
0
0

NA
NA
2

*
*
0
0.1
NA


1
8
8
8
8
8
8
8

8
8
8
8
8
8
8
8

0
0
0

NA
NA
2

*
*
0
1.0
NA


1
8
8
8
8
8
8
8

8
8
8
8
8
8
8
8

0
0
0

*
*
2

*
*
0
10.0
NA


1
8
8
8
8
8
8
8

8
8
8
8
8
8
8
8

0
0
0

*
*
2

*
*
0
Assumptions
Not appropriate for agriculture
Ranked fojr proteins and emulsions only

Ranked for proteins and emulsions only
Midsummer clearcut assumed





Sandy soils stabilize poorly, less impact
Sandy soils stabilize poorly, less impact
Sandy soils stabilize poorly, less impact




Assumes retrieval feasible before crops needed
Assumes retrieval feasible before crops needed
Inappropriate for unrefined human foods



NA - Not Appropriate
                                                                                                                                     8-34

-------
                                   Table 8-7.   Predicted recovery  index of agricultural  lands in grasslands  for cleanup treatments  (see Appendix B).
Treatment
0-1 Natural Rehabilitation
1-1 Chemical Stabilization
1-2 Remove Vegetation by Clear Cutting
1-3 Remove Stumps and Grub Roots
1-4 Surface Scraping and Grading (<5 cm)
1-5 Shallow Plowing (<10 cm)
1-6 Deep Plowing (10 to 40 cm)
1-7 Soil Cover (<25 era)
1-8 Soil Burial (25 to 100 cm)
2-1 Remove Plow Layer Soil (10 cm)
2-2 Remove Shallow Root Zone Soil (<40 cm)
2-3 Scrape Surface, Remove (<5 cm), Mechanically Stabilize
2-4 Remove Plow Layer (10 cm). Mechanically Stabilize
2-5 Remove Shallow Root Zone (<40 cm), Mechanically Stabilize
2-6 Scrape Surface, Remove (<5 cm). Chemically Stabilize
2-7 Remove Plow Layer (10 cm), Chemically Stabilize
2-8 Remove Shallow Root Zone (40 cm), Chemically Stabilize
3-1 Barriers to Exclude People
3-2 Barriers to Exclude Large Animals
3-3 Barriers to Exclude Large and Small Animals
4-1 Asphalt Hard-Surface Stabilization
4-2 Concrete Hard-Surface Stabilization
5-0 Sewage Sludge Application
6-1 Washing (<3 mm)
6-2 Flooding (3 to 30 on)
7-0 Soil Amendments Added

Forage/Pastures
Area Disturbed (km2)
0.01
NA
1
1
NA
3
2
4
4
4
3
4
2
3
4
2
3
4
0
0
0
5
5
2
*
*


0.1
NA
1
1
NA
3
2
3
4
4
3
4
2
3
4
2
3
4
0
0
0
NA
NA
2
*
*


1.0
NA
1
1
NA
3
2
3
4
4
3
4
2
3
4
2
3
4
0
0
0
*
*
2
*
*

0
10.0
NA
1
1
NA
3
2
3
4
4
3
4
2
3
4
2
3
4
0
0
0
*
*
2

*

0
Irrigated Row Crops
Area Disturbed (km2)
0.01
NA
1
1
2
3
2
3
4
4
3
4
2
3
4
2
3
4
0
0
0
5
5
NA

*

0
0.1
NA
1
1
2
3
2
3
4
4
3
4
2
3
4
2
3
4
0
0
0
NA
NA
NA

*

0
1.0
NA
1
1
2
3
2
3
4
4
3
4
2
3
4
2
3
4
0
0
0
*
*
NA

•

0
10.0
NA
1
1
2
3
2
3
4
4
3
4
2
3
4
2
3
4
0
0
0
*
*
NA

*

0
Dry Farmed Small Grains
Area Disturbed (km2)
0.01
NA
1
1
NA
3
2
3
4
4
3
4
2
3
4
2
3
4
0
0
0
5
5
NA
*
*

0
0.1
NA
1
1
NA
3
2
3
4
4
3
4
2
3
4
2
3
4
0
0
0
NA
NA
NA
£
*

0
1.0
NA
1
1
NA
3
2
3
4
4
3
4
2
3
4
2
3
4
0
0
0
*
*
NA

*

0
10.0
NA
1
1
NA
3
2
3
4
4
3
4
2
3
4
2
3
4
0
0
0
*
*
NA

*

0
Assumptions
Not appropriate to agriculture
Ranked for proteins and emulsions only
Midsummer clearcut assumed
Grass resprouts pptn dependent
Needs organic matter added
Needs stabilizing and reseeding
Stabilize, fertilize, and re seed
Needs early grass reseeding
Needs fertilizer and early grass reseeding
Not effective on trashy surface
Not effective on trashy surface
Needs fertilizer and plowing 
-------
                                            Table  8-8.
                                                           Predicted recovery index of agricultural  lands in  forests  for cleanup treatments  (see  Appendix  B).
                             Treatment
  0-1   Natural Rehabilitation
  1-1  Chemical Stabilization
  1-e  Remove Vegetation by Clear Cutting
  1-3  Remove Stumps and Grub  Roots
  1-4  Surface Scraping and Grading (<5 cm)
  1-5  Shallow Plowing (<10 cm)
  1-6  Deep Plowing (10 to  40  cm)
  1-7  Soil Cover (<25 era)
  1-8  Soil Burial  (25 to 100  cm)

  2-1  Remove Plow  Layer Soil  (10 cm)
  2-2  Remove Shallow  Root  Zone Soil  (<40 cm)
  2-3  Scrape Surface,  Remove  (<5 cm).  Mechanically Stabilize
  2-4  Remove Plow  Layer (10 en,). Mechanically Stabilize
 I'i  t^n6 I   l]°" R2ot Zon? (<4° <»)•  Mechanically Stabilize
 2-6  Scrape Surface,  Remove  <5 cm).  Chemically Stabilize
 2-7  Remove Plow  Layer (10 cm). Chemically Stabilize
 2-8  Remove Shallow Root Zone (40 on). Chemically Stabilize

 3-1  Barriers to  Exclude People
 3-2  Barriers to Exclude Large  Animals
 3-3  Barriers to Exclude Large  and Small Animals

 4-1   Asphalt Hard-Surface  Stabilization
 4-2  Concrete Hard-Surface Stabilization

 5-0   Sewage Sludge Application

 6-1  Washing (<3 ran)
 6-2  Flooding (3 to 30 cm)

 7-0  Soil Amendments Added

	'	•	.
•Outside scope of  work.
  NA - Not Appropriate
— 	 	 • 	 _^____ __
Forage/Pasture
Area Disturbed (km2)
0.01
NA
1
1
NA


2
3

3
4
4
4
6
4
4
4

0
0
0
5
5

2
*

0
••^m^BMM
0.1
NA
1
1
NA
4
3
3
2
3

3
4
4
4
6
4
4

0
0
0
NA
NA

2
*

0
— ««.^»™
1.0
NA
1
1
NA
4
3
3
2
3

3
4
4
4
6
4
4
4

0
0
0
*
*

2
*
*
0
i^—H^KM^K
10.0
NA
1
1
NA
5
3
4
2

3
5
4
4
6
4
4
4

0
0
0
*
*

2
*
*
0

Row/Field Crops
Area Disturbed (km2)
0.01
NA
1
1
NA
4
3
3
2

3
4
4
4
6
3
3

0
0
0
5
5

NA
*
*
0
~MM_B^_
0.1
NA
1
1
NA
4
3
3
2
3
3
4
4
4
6
3
3
3
0
0
0
NA
NA

NA
*
*
0
1.0
NA
1
1
NA
4
3
3
2
3
3
4
4
4
6
3
3
3
0
0
0
*
*

NA
*
*
0
^^^^^•Wi
10.0
NA
1
1
NA
5
3
4
2
3
3
5
4
4
6
3
3
3
0
0
0
*
*

NA
*
*
0
Area Disturbed
(km*)
0.01
NA
1
8
8
8
8
8
a
8
8
8
B
8
8
8
8
8
0
0
0
8

4
*
•
0
0.1
NA
1
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
0
0
0
8
8
4
*
*
0
1.0
NA
1
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
0
0
0
*
*
4
1
*
0
10.0
NA
1
8
8
8
8
8
8
8
8
8
8
8
8
8
a
8
0
0
0
*
*
4
ft

0


Assumptions
Not appropriate for agriculture
Ranked for protein or emulsions only
Midsummer clearcut assumed


Needs organic matter
Needs mulch
Needs fertilizer and mulch cover
Needs fertilizer and organic matter
Needs fertilizer, organic matter, and mulch
Requires erosion control maintenance
Requires erosion control maintenance
Requires erosion control maintenance
Needs fertilizer
Needs fertilizer and organic matter
Old field retrogression after one year
Old field retrogression after one year
Old field retrogression after one year
Assumes retrieval before needed
Assumes retrieval before needed
Not for unrefined food crops; for cotton and trees



   8-36

-------
 erosion  is more  of  a problem because  the deep cuts on pastures remove pro-
 gressively more  grass roots that have bound  soil together.  Row crops gen-
 erally are on  flatter slopes in well  managed farms where erosion  is  less
 severe than on scalped pastures.

 8.7  CONCLUSIONS
     The preferred  cleanup treatments are those which isolate the food chain
 from man.  For small areas this may include fencing  (treatment 3-1)  or apply-
 ing hard surfaces (treatments 4-1 and 4-2) to a small area with the  expecta-
 tion that they will be retrieved at a later date.  Shallow or deep plowing
 to 40 cm is acceptable with the exception of eastern orchards which  might
 suffer moderate root pruning from deep plowing.  Seven of the chemical sta-
 bilizers recommended for agricultural use would be appropriate in short-term
 response to a  contaminating event.
     Cleanup techniques which are inappropriate in an agricultural area ex-
pected to be returned to crop production are those which involve dispersing
soil by the addition of sodium compounds in large quantities,  particularly in
either desert or semi-arid grassland areas.   Removal of the shallow root zone
soil to 40 cm depth should not be considered for areas larger than 1 or 2 hec-
tares because the problem of disposing of the contaminated soil is a greater
problem than other ways of decontaminating.   Although fenced areas to exclude
cattle have been recommended in the natural ecosystems as a preferred treat-
ment, they are not recommended here because of their position near the top
of the food chain leading to man.   It is almost inevitable that domestic
cattle would breach fencing in grassland or deciduous forest areas if the
land is left fenced for more than a few months.
                                   8-37

-------
 8.8  AGRICULTURAL REFERENCES


  1.  Agricultural Research Service.  Predicting Rainfall-Erosion Losses from
      Cropland East of the Rocky Mountains:  Guide for Selection of Practices
      for Soil and Water Conservation.  Agriculture Handbook No. 282, U.S.
      Dept.  of Agriculture, 45 pp.

  2.  Agricultural Research Service.  A Universal Equation for Measuring Wind
      Erosion— An Aid to Conservation Farming in the Great Plains.   U.S. Dept.
      of Agriculture, Special Report ARS-22-69,  1961, 20 pp.

  3.  Agricultural Research Service.  Wind Erosion Forces in the United States
      and Their Use in Predicting Soil Loss.   Agriculture Handbook  No.  346,
      U.S.  Department of Agriculture, 42 pp.

  4.  Alberts,  E.E.,  G.E.  Schuman,  and R.E.  Burwell.   Seasonal Runoff Losses
      of Nitrogen and Phosphorus from Missouri Valley Loess Watersheds.
      J.  Environ.  Qual.,  Vol.  7,  No. 2,  1978,  pp.  203-208.

                  Don't  drift  .  .  .  avoid Wind Erosion.   Agribook Magazine,
     _
     Vol.  3,  No.  1,  1977,  p.  10.

  6.  Bar-Kochba,  Y. , and A.L.  Simon.   Factors  Affecting  Floods  from Watersheds
     in Humid Regions of Northeastern  Ohio.  Water  Resources  Bulletin   Vol   8
     No. 6, 1972, pp. 1235-1245.                                      '     '   '

  7.  Barrows,  H.L.,  and V.J.  Kilmer.   Plant Nutrient  Losses from  Soils  by Water
     Erosion.  Advances in Agronomy, 15: 303-316, 1963.

  8.  Bennett,  F.W.,  and R.L.  Donahue.  Processes, Procedures  and  Methods to
     Control  Pollution Resulting from  Silvicultural Activities.   Misc.  Publi-
     cation,  U.W. Environmental Protection Agency,  Office of  Air  and Water
     Programs, Washington, D.C. 1973.

  9.  Chepil,  W.S., and N.P. Woodruff.  The Physics  of Wind Erosion  and  Its
     Control.  Advances in Agronomy, 15: 211-302, 1963.

10.  Godfrey,  R.S., ed.  Building Construction Cost Data, 34th Annual Edition.
     Robert Snow Means Co., Inc., Duxbury, Mass., 1975, 330 pp.

11.  Hotes, F., L.K.H.  Ateshian, and B. Sheikh.  Comparative  Costs  of Erosion
     and Sediment Control, Construction Activities.   EPA-430/9-73-016, U.S.
     Environmental Protection Agency, Washington, D.C., 1973, 205 pp.*

12.  Jenny, H. Factors of Soil Formation.  McGraw-Hill, New York, 1941.

13.  Nuttonson, M.Y.  Wheat-Climate Relationships and the Use of Phenology in
     Ascertaining the Thermal and Photo-Thermal Requirements of Wheat.  Ameri-
     can Institute of Crop Ecology, Washington, D.C.,  1955,  388 pp.

14.  Ryden, J.C., J.K.  Syers, and R.F.  Harris.   Phosphorous  in Runoff and
     Streams.   Advances in Agronomy, 25:  1-45,  1973.

                                     8-38

-------
                       *' ^  ^^^ Vegetation'  Atlas of American Agricul-
                   1924


       r'lOQ-i^; 1962W'H* WiSChmeier'  Rainfa11 Erosion«  Advances in Agronomy,


17.  Wischmeier  W.H.  A Rainfall Erosion Index for a Universal Soil-Loss Equa-
     tion.   Soil Sci. Soc. Amer. Proc. 23: 246-249, 1959.

18.  Wischmeier, W.H.  Use and Misuse of the Universal Soil Loss Equation
     Soil Erosion,  Prediction and Control.  Soil Conservation Society of '
     America,  Ankeny, Iowa,  1977, pp.  371-378.
                                    8-39

-------

-------
 9.1   OVERVIEW
      Urban/suburban  areas  differ  from  the  agricultural  and  undeveloped  areas
 addressed  in other chapters  of  this  report in  a  number  of ways  that  affect  the
 reclamation decisions and  their impacts.

      The characteristics of  urban/suburban areas that may influence  the cleanup
 strategies and procedures  are diverse.  At one extreme, urban/suburban  areas
 include downtown areas of  major cities which are completely covered  by  build-
 ings  and pavements,  with residential and transient populations  numbering hun-
 dreds of thousands of people per  square mile at  any one time.   In contrast,
 urban/suburban areas also  include single-family  homes on spacious lots  with
 only a small fraction of the land area covered by buildings and paved walkways
 and roadways, and with populations of, at most,  a few thousand persons  per
 square mile.   Between these extremes are various other types of land use
 classes, such as commercial and industrial  sections.  In addition, within an
 area of certain overall characteristics, such as a residential area of  single-
 family homes, great diversity can exist.  Such places can have significant park
 and recreational areas,  scattered shopping centers and other commercial areas,
and undeveloped areas.*

     In addition to the diversity of population  density and land use and the
 degree of structural buildup, a particular area  will have other characteristics
 that influence the cleanup options.   Some of these factors are:   the surface
and subsurface hydrology of the area, the existing sewer and storm-drain sys-
tem, availability of equipment and manpower, the source of water supply (in-
cluding areas outside the affected area whose water supply might become
'Undeveloped areas within urban/suburban areas are not addressed in this chap
 ter.   For these areas,  the land type which best describes the undeveloped
 area  should be determined and the appropriate chapter in this report referred
 to while recognizing that the greater likelihood of human exposure in the
 vicinity of urban/suburban areas may require special considerations as to the
 cleanup techniques.
                                      9-1

-------
contaminated unless care  is taken), season of the year  (e.g., whether  the
ground is frozen or covered with snow), and short-term meteorology  (e.g.,
humidity, winds, likelihood of rain).

     From this discussion it must be recognized that for any particular inci-
dent that releases contaminating materials in an urban/suburban area,  the
courses of action and the effectiveness and consequences of such actions are
highly dependent on the specific situation.  However, it is useful  to  attempt
to describe a "typical" urban/suburban area.  One important characteristic is
the fraction of land area that is covered by (1) impervious surfaces such as
roofs, paved walkways and road areas, and (2) permeable surfaces consisting
of lawns, plants and trees, and bare earth.  Each of these types of areas
requires different cleanup techniques having different environmental impacts.

     The Census Bureau defines an "urbanized area" as a central city (or con-
tiguous cities) with a total population in excess of 50,000 plus the surround-
ing "closely held" populated areas, i.e., what is generally thought of as an
urban area and its suburbs (exclusive of large parks, golf courses, cemeteries,
etc.).  Based on 1970 Census data, there were 248 such urbanized areas in the
United States with a total land area of 35,081 square miles and a total popu-
lation of 118,446,566 people,  or 58 percent of the total U.S. population.
Thus, according to the U.S. Bureau of the Census , the average population
density in urbanized areas is 3,376 people per square mile.

     National statistics regarding the fraction of ground surface covered by
streets,  buildings and permeable surfaces in urban/suburban areas could not
be found, so residential areas in Santa Barbara, California, were randomly
selected and surveyed to develop these statistics.  The population density of
the "place" of Santa Barbara is 3,333 people per square mile6, which is in
excellent agreement with the average population density of urbanized areas,
so it is assumed that statistics for Santa Barbara are reasonably representa-
tive of typical urban/suburban areas of the Nation.

     Within single-family housing tracts of Santa Barbara,  the approximate
coverage of the ground surface area is 20 percent by buildings;  15 percent
by streets;  8 percent by driveways, sidewalks,  walkways, patios, etc.;  and
the remaining 57 percent by lawn and soil areas.  Thus within housing tracts,

                                      9-2

-------
  the  impervious  surfaces comprise approximately 43 percent of the  surface
  area.  Such single-family housing tracts that are fully developed have popu-
  lation densities of approximately 10,000 persons per square mile.

      The above  statistics are applicable for small areas within fully devel-
  oped single-family housing tracts.  Within a larger area, the permanent popu-
  lation density  is less because of commercial areas, open spaces, etc., and
  therefore .the proportion of the area which is covered by impervious surfaces
 may differ.  A recent study was concerned with the recharge of groundwaters
  in suburban areas of Santa Barbara.4  Within the area of interest of that
 study,  which can be described as primarily residential single-family suburban
 housing tract areas (upper-middle to middle income housing),  the population
 density is approximately 3,500 to 4,500 persons  per square mile (i.e., roughly
 in agreement with the typical population density of an urbanized area).   The
 ground  surface consists of approximately 37 percent impervious  surfaces  (build-
 ings, streets, etc.),  24 percent lawn and yard areas,  9 percent agricultural
 areas,  and 30  percent  open space and native vegetation areas.   Thus  it can be
 seen  that  a surprisingly large percentage of a suburban area  (39 percent  for
 the case of Santa Barbara)  consists  of areas having  "rural" or  "natural"  char-
 acteristics.   However,  the percent of surface which  is  impervious  is not  much
 less  than  the  value  of  43  percent within housing  tracts.

      It is  estimated that  approximately one-half  of  the rainfall from  small
 storms in Santa  Barbara  that  falls on  impervious  surfaces finds  its way to
 storm sewer systems with the  remaining  one-half soaking into the ground.4
 Therefore,  if  the cleanup option is  to  flush  the  contaminant from building
 roofs, streets,  and walkways with water, approximately one-half of the con-
 taminated water  can be expected to soak  into  the ground unless measures are
 taken to prevent this occurrence.

     Other  data for the upper Santa Ana Valley in California indicate that the
 fraction of land area which is covered by impervious surfaces varies linearly
 from approximately 20 percent for an area that is 40 percent urban developed
 to approximately 70 percent for an area that is 100 percent urban developed.4
 For the Santa Barbara area, which is  approximately 60 percent  urban developed,
 the ground surface should be approximately 38 percent covered  by impervious
surfaces.   This is in excellent agreement with the actual  value  of 37 percent.
                                     9-3

-------
      In  summary,  the characteristics of  small areas of  0.01 km2  in  size withir,
 an urbanized area may vary considerably.   If one randomly  selects a small area
 within an urbanized areas, the probability is approximately 40 percent that
 the area will have "rural" or "natural"  characteristics rather than urban
 characteristics.  As the area of contamination is increased to the  maximum of
      o
 10 knr used in this study, the characteristics of the area approach the aver-
 age characteristics of 60 percent urban  developed with 37 percent of the land
 covered  by impervious surfaces and a population density of approximately 3,400
 people per square mile.

 9.2   NATURAL PERTURBATIONS
      Natural perturbations that produce  some of the environmental effects that
 could occur from  cleanup treatments in urban/suburban areas include  most of
 the natural disasters, i.e., floods, fires, earthquakes, and severe  storms.
 In most  cases, these natural perturbations will have significant effects on
 man-made structures; whereas the reclamation treatments will have little ad-
 verse effects.  In contrast, some of the reclamation techniques will typically
 have more adverse effects on the "natural" ecology of urban/suburban areas than
 do natural disasters.  For example, removing 40 cm of soil will destroy essen-
 tially all vegetation, whereas no natural disaster is likely to produce this
 degree of total destruction.

 9.3  MAN-MADE PERTURBATIONS
     A number of activities of man can cause destruction to the man-made envi-
 ronment and the natural environment of urban/suburban areas.  Such activities
 include war,  urban development or redevelopment,  and spills of hazardous ma-
 terials.

     The normal construction activities associated with urban development or
 redevelopment usually employ most of the cleanup techniques that are discussed
 in Section 9.4.  Typically,  earthmoving and destruction of vegetation are re-
quired to prepare land for development.  By design,  the land is not allowed
 to return to its natural condition; instead, lawns are established and desir-
able trees and smaller plants are planted.   Within a few months to a couple
of years the desired vegetation,  except large trees,  is established.
                                     9-4

-------
      Spills of materials that are hazardous or ecologically damaging require
 cleanup techniques identical or similar to the treatments for cleanup of radio-
 active material.   Again, in the urban/suburban situation there is usually no
 significant lasting effect (except for social and economic impacts)  because
 the artificial man-made environment that is disturbed by the treatment can
 usually be reestablished relatively quickly.   An exception to this statement
 would occur if the contamination could not be cleaned up to the necessary de-
 gree,  forcing restriction or abandonment of an urban/suburban area or removal
 of structures.  However,  even in this  case the impact is primarily social and
 economic,  with little  additional effect on the natural environment.

 9.4  EFFECTS  OF CLEANUP PROCEDURES ON
      URBAN/SUBURBAN AREAS
      The urban/suburban environment is largely a man-made environment,  in
 contrast to the natural environments of undeveloped  areas.   Some cleanup op-
 tions  that  might cause significant environmental impact  in undeveloped  areas
 (e.g.,  vegetation  clearing,  removal of soil  layers,  and  applying soil  cover)
 are common  activities  in  urban/suburban areas.

      Because  of the greater  hazard to  humans  from contamination  of urban/
 suburban areas, a  greater cleanup  effort  is not  only justified but required.
 To  clean up the area to the maximum possible  extent,  a series of cleanup tech-
 niques  would  probably  be  used rather than  a single technique.  The effective-
 ness of the cleanup is  primarily limited only  by the monetary cost of carrying
 out  the tasks.  For example, contaminated  roofs  could be  stripped of shingles
 and  reshingled if  necessary.  Contaminated pavements could  even  be removed.
 In  an urban/suburban area, virtually the entire  surface area could be stripped
 and  removed without  significant impact  on  the  "natural" ecology—but with a
 high monetary and  social  cost.

     Treatments that can be used in urban/suburban areas to remove contami-
nants or to reduce  the hazard from  such contaminants can be grouped into
those that lend themselves to application for  lawn,  plant, and soil areas in
yards and parks (treatments 1-1 through 7-0 in the following list), and those
that lend themselves to application for hard, relatively impervious,  artifi-
cial surfaces, such as building roofs,  walkways, and paved areas (treatments
8-1 through 8-7 in the following list).

                                     9-5

-------
           Treatments  for  Lawn,  Plant  and  Soil  Areas*

           0-1   Natural  Rehabilitation-

           1-1   Chemical Stabilization
           1-2   Clear  Cutting Vegetation
           1-3   Stumping and Grubbing
           1-4   Scraping and Grading
           1-5   Shallow  Plowing
           1-6   Deep Plowing
           1-7   Soil Cover  Less  than 25 cm
           1-8   Soil Cover  25 to 100 cm

           2-1   Remove Plow Layer
           2-2   Remove Shallow Tloot Zone
           2-3   Remove Scraping  and Grading, Mechanically Stabilize
           2-4   Remove Plow Layer, Mechanically Stabilize
           2-5   Remove Shallow Root Zone,  Mechanically Stabilize
           2-6   Remove Scraping  and Grading, Chemically Stabilize
           2-7   Remove Plow Layer, Chemically Stabilize
           2-8   Remove Shallow Root Zone,  Chemically Stabilize
           3-1   Barriers to Exclude People
           3-2   Exclude Large Animals
           3-3   Exclude Large and Small Animals

           4-1   Asphalt Hard Surface Stabilization
           4-2   Concrete Hard Surface  Stabilization
           5-0   Application of Sewage  Sludge
           6-1   High Pressure Washing  (<3 mm)
           6-2   Flooding to 30 cm

           7-0   Soil Amendments  Added

          Treatments  for Hard,   Impervious, Artificial Surfaces

          8-1   Washing with High-Pressure Water (Firehosing)
          8-2   Vacuuming
          8-3   Sweeping
          8-4  Mechanized Street Flushing
          8-5   Surface Removal   Techniques
          8-6  Other  Removal Methods
          8-7  Containment

     In general, treatments 1-1 through 1-8 are not treatments for the removal

of a contaminant (although 1-2   is a removal treatment if contamination of
vegetation is significant).  Instead,  these treatments are those that might be
 This list of treatments applies to all land types,  and is included for a com-
 plete discussion.   Some of these treatments are not appropriate for urban/
 suburban areas.


                                     9-6

-------
 applied as a temporary measure to hold the contaminant in place to prevent its
 redistribution.  In the case of treatments 1-5 through 1-8, the immediate haz-
 ard is reduced because the contaminant is shielded by soil cover.

      Treatments 1-1 through 7-0 apply to land types in general.  Most of these
 treatments also apply for the lawn, plant, and soil areas (yards)  of urban/sub-
 urban areas.   However, in the urban/suburban situation some of the treatments
 may be more appropriate than others and the environmental impacts  may be sig-
 nificantly different than would be the case for an undeveloped or  agricultural
 area.

      Because  of the greater potential hazard  from a contaminating  incident in
 a  populated area,  several  of the  treatments might be used on  urban/suburban
 yards  so  as to  reduce  the  hazard  to the maximum practical extent.   For example,
 a  stabilizer  might  be  applied to  lawns to hold the contaminant in  place while
 the sod is removed  and any residual  contamination hazard  might be  further re-
 duced  by  turning over  the  soil  before resodding.

     Depending  on circumstances it  may be desirable or  it may  be undesirable
 to coat the ground  with a  sealant,  such as  liquid asphalt, so  that contami-
 nated  water from washing of  roofs and walkways  will run off into the  storm
 drain  system  rather than soak into  the soil.

     Flushing with  water or blowing with  compressed air may remove the  con-
 taminant from the foliage  of plants if the resuspension in air  or the washing
 into the soil of the contaminant is acceptable.   Otherwise, the contaminant
 can be fixed onto the  leaves of the plant by a stabilizer and then the plant
 can be pruned, defoliated, or removed.

     Some of the treatments which are used for cleanup of hard artificial
 surfaces,  such as sweeping and firehosing, can also be used with some limited
 effectivess on frozen ground.

     Each  of the potential treatments is discussed below in the order listed.
The discussion is based primarily on Owen et al.5 and to a lesser extent on
Bennett and Owen  and Cobb and Van Hemert.2
                                     9-7

-------
  9.4.1   Lawn,  Plant  and Soil Area  Treatments
       (0-1)  Natural Rehabilitation
       Because  of the proximity of  humans, it is assumed that doing nothing to
 clean up urban/suburban areas is  not a viable option; rather, any contamina-
 tion will have to be reduced to some acceptably low level.  Therefore, natural
 restoration does not apply for urban/suburban areas.

       (1-1)  Chemieal Stabilization
      The purpose of the chemical  stabilizers is to hold the contaminant in
 place, to prevent redistribution until the contaminant can be removed.  The
 properties of the general classes of chemical stabilizers are given in Appen-
 dix A, Soil Stabilizers.   Chemical stabilizers vary as to their effectiveness
 and longevity in holding the contaminant in place and in their effects on
 vegetation.  Some of the stabilizers cannot be used without necessitating
 subsequent removal of the vegetation.

      In lawn areas the sod will  undoubtedly be removed and the lawn area re-
 established with seed or  resodding,  in which case the  use of chemical  stabi-
 lizers has no  significant adverse  environmental  effect.

      (1-2)   Clear Cutting Vegetation
      For lawns and ground covers,  this  treatment  consists  of mowing  or perhaps
 renovating  down to the  soil.  There  is  no significant  adverse  effect from mow-
 ing  or renovating,  since  the root  zone  is not  damaged  and  the  grass  or ground
 cover  grows  back relatively quickly.

      In  general, clear cutting of  bushes and trees  is  to clear the ground sur-
 face for plowing or  soil removal.  Many  bushes, and some  trees, will reestablish
 from the root  systems of the clear-cut vegetation,  although not as quickly as
 grass.  The  time to reestablish varies from a year  for small shrubs  to many
 decades for  large  trees.

     (1-3)  Stumping and Grubbing
     This treatment contributes little additional impact to that which occurs
when the trees and bushes are cut down.  If trees are cut down, new trees would
probably be planted so removal  of stumps is  desirable.   Shrubs would have to be
replanted rather than allowed to  propagate from existing root systems.

                                 9-8

-------
      (1-4)   Scraping and Grading
      The removal  of up to 5  cm of soil  by scraping and grading is equivalent
 to removing the sod layer for lawns.  Removal  of this  layer will  not have any
 effect  on trees and larger shrubs unless  they  must be  removed to  provide  room
 for equipment.  Smaller plants may be damaged,  but they would likely be re-
 moved in any case to ease the task of scraping and grading.

      (1-5)   Shallow Plowing
      The disturbance of the  upper 10 cm of soil by shallow plowing will have
 more significant  effects than treatment 1-4.   If plowing is  done  close  to
 small plants they may be killed by root disturbance; but most trees will  not
 be killed or seriously damaged by plowing to this depth.

      (1-6)   Deep  Plowing
      Deep plowing (10 to 40  cm)  will have more  serious  effects on trees.   Most
 trees in yards  and  parks have shallow root systems because of improper  water-
 ing;  plowing to a depth of 40 cm will damage these root  systems.   However, be-
 cause of interference by larger  roots it  may not  be possible  to plow close to
 the  tree, so many trees that  might  otherwise be killed by  close plowing will
 survive.

      (1-7)   Soil  Cover less  than 25 am

      (1-8)   Soil  Covey 25  cm  to  100 am
      If  not  only  a  temporary  treatment, and if  not  removed within  a  relatively
 short time period,  soil  cover  in an urban/suburban  area will  change  grade  lev-
 els  and  affect  water  drainage  around buildings.  As a short-term measure,  add-
 ing  soil  cover  will not  affect trees and  larger bushes.  Smaller bushes may be
 covered  and  killed  and  lawns will die in  a short time period.

      Except  for palms,  trees cannot tolerate the permanent addition  of  soil
unless measures are taken to ensure adequate air circulation about the trunk
down  to  the  original  grade level.   If soil is placed around the trunk, rot
will  set  in  and the tree will eventually die.  The addition of 25 cm of soil
will have no effect on trees if  it  is kept from the trunk.  A 100-cm layer
requires a larger space around the trunk to ensure adequate air circulation,
and can be expected to affect the root  system of the trees by oxygen depletion.
                                     9-9

-------
       There may be practical  limitations on using soil cover.   To cover the
  ground to a depth of 100 cm  in one square kilometer of the typical  urban
  area,  whose surface  is  comprised of 63  percent  soil and vegetation,  requires
  824,000 cubic  yards  (630,000 cubic meters)  of fill  soil.

       Treatments  2-1  through  2-8  are designed  to  remove a  contaminant  from
  yards  so  that  it  can be  transported to  a  safe place.   These treatments  are
  straightforward-by  whatever means  are  practical, sod and layers  of  soil are
  removed until  the  residual contamination  is at an "acceptable"  level.   In  the
  urban/suburban situation, uncontaminated  topsoil will  usually have to be
  brought in  as fill to restore yards  to  the original grade  level.  Stabilizers
  might be used to prevent erosion, but stabilization techniques  which have
  adverse effects on the revegetation  of  the area, such  as  compacting, would
  not be  required.  Stabilization is required only for a relatively short time
  period while lawns become established.

      The depth to which soil  can be removed from around a tree without injur-
  ing or killing the tree depends on the proximity of the root system to the
 ground surface.  This, in turn, primarily depends on the type of tree, its
 location,  and the watering it has received.   Some of the most  desirable trees,
 such as maples  and some oaks  and conifers, have  relatively shallow root sys-
 tems.   Also, because  of frequent and shallow watering, trees in yards often
 develop root systems  relatively close to the ground  surface;  such trees are
 more susceptible  to damage  from soil removal  than are trees which rely on
 rainfall.

      For lawns, using sod cutters  and rolling  up  the sod for hauling  away is
 effective.   Past  experiments  indicate that this procedure  will  remove  98  per-
 cent  to  99  percent  of a contaminant.   Removing sod by  garden tractors  or  by
 hand  shoveling  appears to be  less  effective  in removing contaminant than  us-
 ing  a sod cutter, although the data  are  limited.

     Where access is  not restricted,  larger machinery  such  as bulldozers, back-
hoes, road graders, etc., may be used  to scrape off the top layers of soil.
Typically, for large  working areas, removal of 5 to 10  cm of soil will sig-
nificantly reduce contamination, although results are quite variable.  For
smaller working areas, such as typical urban/suburban yards, the effectiveness
of heavy machinery is  limited  and it must be supplemented by hand labor.
                                     9-10

-------
      The magnitude of the disposal problem associated with removing contami-
 nated soil,  and the task of hauling in uncontaminated soil for fill, must be
 appreciated.   Removing the plow layer of soil (10 cm) from the "typical"
 urban/suburban area requires the disposal of 82,400 cubic yards (63,000 cubic
 meters)  of contaminated earth per square kilometer, compacted to the original
 density.   Disturbed earth takes up more volume than undisturbed earth,  prob-
 ably by  a factor of 20 to 40 percent,  so removing the plow layer of soil from
 a 10-km2  typical urban/suburban area requires disposing of approximately one
 million  cubic yards of contaminated earth and hauling in the same volume of
 uncontaminated earth for fill.   Removing soil in the shallow root zone  quad-
 ruples these  requirements.

      (2-1)  Remove Plow Layer
      In the urban/suburban  situation there is little impact from removing the
 top  10 cm of  soil  from yards.   Lawns can be reseeded or resodded,  and plants
 and  small bushes can be replaced.   Only very shallow-rooted trees will  be
 seriously damaged  or killed  by  the  removal of soil  to this  depth.

      (2-2)  Remove Shallow Root Zone
      Whether  or  not  a tree is killed from removal of 40  cm  of  soil  depends
 on the particular  tree  and its  situation;  however,  in general,  the  removal
 of this amount of  soil  will  probably kill  most trees that do not  have signif-
 icant  taproots.  The  use of  heavy equipment  will cause more damage  than  care-
 ful use of hand  tools.

      (2-3)  Remove Seraping  and Grading., Mechanically Stabilize
     Removal of  the top 5 cm of  soil will  have no significant  effect on  trees
 and large bushes.  This is equivalent to removing the sod layer from lawns.
 The lawns can be reseeded or resodded, and smaller plants that have been
 removed can be replaced.  Compacting of the  soil for erosion control would
not be required; other means of  stabilization that are practical would have
no significant adverse effects.

      (2-4)  Remove Plow Layer, Mechanically Stabilize
     The  impact is the same as for treatment 2-1, provided compacting or hard-
surface stabilization techniques are not used.

                                    9-11

-------
      (2-5)  Remove Shallow Root Zone* Mechanically Stabilize
      The impact is the same as for treatment 2-2, provided compacting or
 hard-surface stabilization techniques are not used.

      (2-6)  Remove Scraping and Grading.,  Chemically Stabilize
      The impact is the same as for treatment 2-3.

      (2-7)  Remove Plow Layer, Chemically Stabilize
      The impact is the same as for treatment 2-1.

      (2-8)  Remove Shallow Root Zone,,  Chemioally Stabilize
      The impact is the same as for treatment 2-2.

      (3-1)  Barriers  to Exclude People
      This  technique is probably practical only  as a short-term measure while
 the  contaminant is removed.   Barricading  people from urban/suburban areas to
 which they normally have access for longer than a few days or weeks would
 cause considerable social  disruption and  adverse economic  impact.

      (3-2)   Exclude Large  Animals
      Any barrier to exclude  large  animals would also exclude  people,  which is
 unacceptable in an urban/suburban  area except for a brief  time period.   Because
 of the relative scarcity of  large  animals in urban/suburban areas,  there would
 be no significant  impact on  them.

      (3-3)   Exclude Large  and  Small Animals
     Except  for very small areas,  exclusion  of  small  animals  would  be difficult.
Household pets  would have  to be relocated with  their  owners or  impounded.  Wild
animals, primarily birds and rodents, might multiply.  They also might become
vectors for  spreading a contaminant to other areas.

      (4-1)  Asphalt Hard-Surfaoe Stabilization
     Probably in most cases asphalt hard-surface stabilization would be com-
bined with some degree of plant and soil removal, thus entailing additional
logistic and economic problems both for initial treatment and subsequent
                                     9-12

-------
 removal.   An asphalt surface would kill any unremoved vegetation.   Unre-
 moved trees would die from lack of water and oxygen to the roots.

      (4-2)  Concrete Hard-Surface Stabilization
      The  same comments apply as for treatment 4-1,  except that concrete would
 provide a more permanent surface or,  if removed,  be slightly more  difficult
 to break  up and transport.

      (5-0)  Application of Sewage Sludge
      Outside the scope of work.

      (6-1)  High-Pressure Washing
      Outside the scope of work.

      (6-2)  Flooding to 30  am
      Outside the scope of work.

      (7-0)  Soil Amendments Added
      Outside the scope of work.

 9.4.2  Impervious,Artificial  Surface Treatments
      The  treatments  that  may  apply to remove  a contaminant from the hard, im-
pervious  artificial  surfaces  that cover much  of the urban/suburban land areas
are:  washing with high-pressure water  (treatment 8-1), vacuuming  (treatment
8-2), sweeping  (treatment 8-3), mechanized street flushing (treatment 8-4),
surface removal  techniques  (treatment 8-5), and other removal techniques
 (treatment 8-6).

     As would be expected, smoother surfaces retain less residual contaminant
than rougher surfaces, larger particulates are more effectively removed than
smaller particulates, and the fraction of contaminant removed is greater for
initially  greater loading densities.   Data on the effectiveness of the cleanup
techniques are primarily based on results of simulations using sand.   The par-
ticulate sizes were often on the order of 100 microns or more and the surface
loading densities on the order of 100 grams/ft2.   For smaller particulates and
less dense surface loading,  effectiveness should be considerably less.

                                  9-13

-------
       (8-1)   Washing  with Elgin-Pressure  Water  (>3 mm)
       Large,  impervious  areas  such  as  roofs  and  streets  can be  cleaned  by high-
 pressure  streams  of  water.  The high-pressure water dislodges  and dilutes most
 of  the contaminating particles and carries  them away  in the runoff.  In this
 case, the contaminant is simply being moved to  a different location.   Deter-
 gents, manual or motorized scrubbing, and special hose  nozzles may be  employed.
 In  general,  firehosing has been envisioned as an emergency technique to be used
 in  catastrophic situations to clean up a relatively large populated area where
 concern with the runoff of contaminated water is secondary to reduction of the
 immediate hazard.

      Examination of results from numerous experiments summarized by Owen et al.5
 indicates that,  in general, firehosing will remove 90 percent or more of con-
 taminating particulates from areas paved with asphalt, concrete, and asphaltic
 concrete.   Often less than 1 percent of the residue remained after a single
 pass.  Repeated  passes with the water stream removed more contaminant,  but
 with decreasing  effectiveness.  Contaminants in slurrys appear to be removed
 as effectively as are particulates.  Although roofs are not  as effectively
 firehosed  as are street  areas, tar-and-gravel  and  composition shingle roofs
 can be 90  percent or  more decontaminated.   Limited  data on wood shingles  indi-
 cate 80  to 90 percent decontamination effectiveness.

      Firehosing  roofs and paved areas  requires  10 to  20 liters  of water per
 square meter of  area  per pass.  To  treat a  typical  suburban area in  this man-
 ner  would  require  1 to 2 million gallons  (3.7 to 7.4 million  liters)  of water
 per  square kilometer  per  pass, exclusive of  water needed to wash  down vertical
 surfaces.  Even  the smallest-sized  area of 0.01  km2 would  result  in  500 to
 1,000 50-gallon  drums of wastewater.   Thus,  except  for very small  areas, it may
 not  be practical to collect and store  contaminated wastewater from cleanup

     Hosing down impermeable surfaces simply moves the bulk of the contaminant.
 to a different location and/or reduces it to more acceptable concentrations.
 Runoff from structures moves the contaminant to the ground level onto other
 impermeable surfaces,  or onto soil and plants.   Flushing walkways and streets
moves the contaminant  onto soil surfaces or into storm drain systems.  If
                                     9-14

-------
  measures  are not  taken to  collect  the  runoff,  shrubs,  lawns,  and bare earth
  areas  will  receive  contaminated  water  as  will  storm drain discharge  points.

      Uncontaminated, permeable surfaces can  be  sealed  with substances such  as
  liquid asphalt  to prevent  contaminated runoff water from  soaking*into the
  ground.   However, it is unlikely that all splashes  can be contained;  in fact,
  it may be desirable in  some instances to  let contaminated runoff water soak
  into permeable  ground  surfaces as a means of contaminant  collection.  Radio-
  active  contaminants are often quite insoluble and can  be  effectively  filtered by
  soils,  and the  portions that are soluble are quickly adsorbed onto surface
  soil particles.  For example, as discussed in The Behavior of Radioactive
 Fallout in Soils and Plants,* approximately 99 percent of fallout is  insoluble
  (in distilled water) and a solution of soluble portions poured through the
 soil in one set of experiments was 80 to 85 percent adsorbed by the first few
 centimeters of soil.  Plutonium,  a radionuclide of great concern, is very in-
 soluble in water and might best be trapped by soil filtration and subsequent
 removal of a layer of soil.

      (8-2)  Vacuuming
      Removal of contaminating  particulates from hard surfaces by vacuuming  has
 not  been examined  in as much detail  as  other  techniques,  such as firehosing.
 The  reason for this  relative neglect may be because  of  the slow  working rate
 compared to  firehosing.  A  street can be firehosed at a rate of  300 to 500
 square  feet  per  minute  compared to a few tens of square feet per minute for
 vacuuming.   Nevertheless, examination of the  data from  Owen et al.5 indicates
 that vacuuming can be effective in removing particulate contamination  from
 hard surfaces.   In general, based on a limited number of experiments,  vacuum-
 ing can remove in excess of 90 percent, and sometimes in excess  of 99  percent,
 of the  particulate contamination  from composition  shingles  and concrete with
 a single "pass."  In the single experiment where repeated  vacuumings were
 employed to decontaminate concrete, the second vacuuming removed  90 percent
 of the residue from the first vacuuming and the third and  fourth vacuumings
 removed about one-half of the remaining residue.  Thus,  it appears that care-
 ful vacuuming can remove in excess of 99 percent of particulate contamination
from pavement and roofs.  In an experiment  for decontaminating frozen bare
ground,  vacuuming removed about 60 percent  of the contaminant after 4 passes
 (40 percent was removed on the  first pass).
                                     9-1-5

-------
      Details  on  the  design  of  the  vacuuming  experiments  are  not  available,
 but  it  is  presumed that  the particulates  were  "fresh," i.e.,  the particulates
 were not allowed sufficient time to "weather" and become  more firmly  attached
 to the  contaminated  surfaces.

      Vacuuming has a significant advantage over firehosing and other types of
 methods that  simply  move  the contaminant  to  another  location because vacuum-
 ing  picks  up  the contaminant and contains it for safe disposal.  Care must be
 taken that absolute  filters are used, in  addition to the coarse  filters  in
 the  typical vacuum cleaner, so that very  small particulates  will be  trapped
 rather  than resuspended  in  the air by the cleaner's  exhaust.  Otherwise, no
 adverse environmental effects appear to be associated with vacuuming contam-
 inants  from hard surfaces.

      (8-3)  Sweeping
      Mechanized  street sweepers can remove significant amounts of contaminants
 from  dry paved surfaces.  In experiments few passes of these  sweepers typically
 removed more than 99 percent of the contaminant.  Mechanical  street  sweepers
 have  a  realtively high rate of cleaning, e.g., a few thousand square feet per
 minute  per pass.   Again, in a large-scale emergency where the primary concern
 is to reduce the gross hazard as quickly as possible, street  sweeping may be a
 viable  technique.  However, in cases where it is less critical,  street sweepers
 probably would be a  less acceptable alternative due to resuspension  of paticu-
 late materials.  Dust can be kept down to some extent by sprinkling.   In any
 event, mechanized street sweepers can only be used on large paved areas where
 access  is available.   (Street sweepers can also be used, with  limited effective-
 ness, on open areas  of frozen ground.)   Hand brooms have some limited effective-
 ness for sweeping roofs.

     Environmental impact from sweeping is limited to particulate resuspension.

      (8-4)   Mechanized Street Flushing
     As is the case with mechanized street sweepers,  street flushers  have the
 advantage of a high rate of areal coverage,  e.g.,  a few thousand square feet
per minute per pass.   A single pass typically removes 90  to 99 percent of the
 contaminant and a second pass typically results in more than 99 percent removal
 of the original amount of contaminant.   Thus mechanized street flushers appear
                                     9-16

-------
 to be as effective as firehosing and faster.  However, their use is limited
 to streets or other large paved areas.

      Using mechanized street flushers poses the potential problem of flushing
 contaminants into the storm sewer system and thence into the sewage treatment
 plant or some other discharge point.  If this is not desired, the runoff must
 be captured.  Otherwise, there are no significant adverse ecological effects.

      (8-5)   Surface Removal Techniques
      Firehosing, flushing with water,  sweeping,  and vacuuming do not usually
 have any significant effect on the building or paving surfaces which are being
 decontaminated.   Another class of cleanup techniques for hard surfaces  consists
 of removal  of some of the surface to which the contamination adheres.   Surface
 removal  techniques would likely be applied only  for residual contamination
 which cannot be  removed  by the nondestructive techniques.   These techniques
 include  removal  of contaminated surfaces by chemicals,  scraping  or  wirebrush-
 ing,  sandblasting,  or  other abrasive techniques.   Sandblasting,  while very
 effective in removing  contaminants  from  hard surfaces,  poses the hazard of
 suspending  contaminants  in air unless preventive measures are taken.

      Using  surface  removal  techniques on artificial  surfaces poses  no signif-
 icant ecological  problems.   Paint  stripping requires  the use of  caustic chem-
 icals, but  the sludge  will  have  to be collected in any  case.   Sandblasting  is
 not a viable  technique unless measures are  taken to contain  the  particulate
 dust  which  would  otherwise  be resuspended in the air.

      (8-6)  Other Removal Methods
      Other methods for removing contaminants from hard  surfaces, primarily
 small areas,  include wiping with tac-cloths or similar material, applying
 strippable films, and perhaps even ultrasonic cleaning.  Strippable films and
wiping are probably only effective on very  smooth surfaces such as furniture
 tops or automobiles, but may have some application.  Ultrasonic cleaning poses
the possible hazard of resuspending small particulates.

      (8-7)   Containment
     Containment  techniques may be used  (1)  to temporarily immobilize contam-
inants onto  hard  surfaces while cleanup measures  are being formulated,  (2)  as
                                     9-17

-------
a long-term measure to seal off a contaminant from being a hazard, or  (3) as
a final step to seal off any residual radiation after cleanup measures have
been completed.  Containment techniques using an impervious sealant are 100
percent effective for reducing the hazard from alpha emitters since the ex-
ternal hazard from these radionuclides is negligible.  However, the sealant
must be maintained for as long as the underlying surface might be a hazard,
to avoid a hazard when the sealant weathers or otherwise deteriorates.  Paint,
tar, oils, and liquid asphalts are the commonly known containment materials,
based on 1950s and 1960s technology that was primarily concerned with nuclear
weapon fallout.  Many of the soil stabilizers shown in Table A-l (Appendix A)
could also be used to temporarily contain contaminants.

     If the containment is temporary, solvents or other techniques are needed
to remove the sealant to permit cleanup of the surface.  (The sealant itself
may be employed as a means of removing the contamination by "locking onto"
the contaminant so that it is removed with the sealant.)

     Applying techniques for containing contaminants in place on impervious
surfaces and then perhaps removing the containing material may be costly in
monetary terms and disruptive to human activities, but otherwise there is no
significant effect on the environment.

9.5  RECOVERY AFTER CLEANUP
     Most of the cleanup techniques for urban/suburban areas will not cause
significant long-term ecological damage.  Within a matter of months to a year
or two following the destruction of vegetation, lawns and small plants could
be reestablished.   The only long-term impact would be the lack of large trees
and shrubs, if they had to be removed.

     Many of the treatments involving earthmoving, destruction of vegetation,
or soil stabilization are commonly used in urban/suburban areas when preparing
land for development.

9.6  QUANTITATIVE ASSESSMENT OF CLEANUP IMPACTS
     Table 9-1 is an estimate of the relative impact on the ecology of an
urban/suburban area as a result of applying the cleanup treatments described

                                     9-18

-------
e"Vi~tal
                                                           of treatrcents on a scale of 0 (no l.pact) to 100
VO
I— 	 • 	 _ 	


Treatment
0-1 Natural Rehabilitation

1-1 Chemical Stabilization
1-2 Remove Vegetation by Clear Cutting
1-3 Remove Stumps and Grub Roots
1-4 Surface Scraping and Grading (<5 cm)
1-5 Shallow Plowing (<10 cm)
1-6 Deep Plowing (10 to 40 cm)
1-7 Soil Cover (<25 cm)
1-8 Soil Burial (25 to 100 cm)
2-1 Remove Plow Layer Soil (10 cm)
2-2 Remove Shallow Root Zone Soil (<40 cm)
2-3 Scrape Surface, Remove (<5 cm), Mechanically Stabilize
2-4 Remove Plow Layer (10 cm), Mechanically Stabilize
2-5 Remove Shallow Root Zone (^40 cm), Mechanically Stabilize
2-6 Scrape Surface, Remove (
-------
in Section 9.4.  The impacts are rated on a scale of 0  (no impact) to 100
(greatest impact) for the various treatments and the size of the area disturbed,

     In the urban/suburban case impact is assumed to be directly proportional
to the area affected, because there are relatively few wild animals, except
those considered pests, to be displaced and the vegetation will be replaced
rather than permitted to reestablish naturally.  Therefore, the impact on a
10-km2 area is assumed to be 1,000 times greater than the impact on the 0.01-
km-2 area.  This is in contrast to undeveloped land types where impact is not
assumed to be directly proportional to the size of the area.

     Treatments 4-1 and 4-2, stabilization with a hard surface, are judged to
be the most severe with a rating of 100.*  Not only will these treatments re-
sult in killing all vegetation and obviating most social and economic uses,
but the surface remains in place for a significant time period, in contrast
to the other treatments where the area can be restored to near its original
condition within a few years.

     Treatment 1-3, stumping and grubbing, and treatments 2-2, 2-5, and 2-8,
removing the shallow root zone, are rated next in severity with a rating of
75.  These treatments will result in the death of essentially all vegetation.
Treatment 1-2, clear cut vegetation, is rated less damaging at 70 because some
of the vegetation can propagate from existing root systems.

     Treatments 1-7 and 1-8, soil cover; treatment 1-6, deep plowing; and
treatments 2-1, 2-4, and 2-7, removing the plow layer, are judged roughly
equivalent with a rating of 50.  Most trees will not be seriously damaged or
killed by these treatments, but all other vegetation will be destroyed. Treat-
ments 1-4, 2-3, and 2-6, scraping and grading, and treatment 1-5, shallow plow-
ing, are rated approximately equal with a rating of 25.  Damage is to a lesser
degree than with removal of the plow layer.

     Treatments 3-1, 3-2,  and 3-3,  barriers to exclude people and animals,
are assigned impact ratings of 10.   Presumably the barriers are for a short
time period.   Application of sewage sludge may pose some problems in social
unacceptability.
*A11 ratings are for a 10 square km area.

                                    9-20

-------
      Treatment  1-1,  chemical  stabilization,  and treatments  8-1  through 8-7,
 treatment  for artificial  surfaces,  are  judged  to have  negligible  environmental
 impact  provided they are  done correctly so  that the  contaminant is  either con-
 tained  or  controlled in a desirable manner.

 9.7   CONCLUSIONS
      Although it may be very  expensive  and  socially  disruptive  to clean up
 contamination from a large urban/suburban area,  significant adverse  effects
 on the  natural  ecology are short term.   Many of the  treatments  are used during
 normal  construction  activities that occur in urban/suburban areas.   Within a
 matter  of months to  a year or two following  cleanup, lawns and  small plants
 will  be reestablished (except in the case of hard-surface stabilization).
 The only change  will  be the lack of large trees  and  shrubs if they have been
 removed.  The consequences of not effectively  cleaning up the contaminant
 (e.g.,  human health)  may  be more significant than concern for the environ-
 mental  impact of cleanup  treatments.  Thus,  environmental impacts must  be
 secondary to selecting the most effective treatment  or treatments.  The resid-
 ual contamination must be  reduced to some acceptably low level; if this neces-
 sitates the complete  destruction of all vegetation,  that impact must be ac-
 cepted.  In this context  there are no treatments that are unacceptable, ex-
 cept natural rehabilitation.   Circumstances will dictate the most desirable
 treatments for a particular situation.

 9.8  URBAN/SUBURBAN LAND AREAS REFERENCES
 1.   Bennett, C.B., and W.L.  Owen.   Planning Radiological Reclamation of Test
     Facilities at Kwajalein Contaminated by Plutonium, Volume II—Radiologi-
     cal Reclamation  Procedures.   U.S. Naval Radiological Defense Laboratory,
     San Francisco, CA 94135,  Report USNRDL-TR-67-68 (AD 820010),  29 May 1967
      (Official Use Only)
 2.   Cobb,  LTC F.C.,  and LTC  R.L.  Van Hemert.  USAF,  Source Book on Plutonium
     and its Decontamination.   Field Command, Defense Nuclear Agency, Technol-
     ogy and Analysis Directorate,  Kirtland Air Force Base NM 87115, 24 Sep-
     tember 1973.
3.   National Academy of Sciences,  National Research Council.   The Behavior of
     Radioactive Fallout in Soils  and Plants.  Washington,  D.C., Publication
     1092,  1963.
4.   Office of Environmental  Quality (in cooperation with Geotechnical  Consul-
     tants, Inc.).   Impact of Urbanization on Recharge  Potential of the Goleta
     Groundwater Basin.   Prepared  for the County of Santa Barbara,  Santa Barbara,
     California, March 1976 (revised July 1976).

                                     9-21

-------
5.   Owen, W.L., F.K. Kawahara, and L.L. Wiltshire.  Radiological Recalamation
     Performance Summary, Volume 1—Performance Test Data Compilation.  U.S.
     Naval Radiological Defense Laboratory, San Francisco, CA 94135, Report
     USNRDL-TR-967, OCD Work Unit No. 3114A, 13 October 1965.

6.   U.S. Bureau of the Census.  County and City Data Book-1972 (A Statistical
     Abstract Supplement).  U.S. Government Printing Office, Washington, D.C.,
     1973.
                                    9-22

-------
      PART III



      WILDLIFE








CHAPTER 10, WILDLIFE

-------

-------
                              CHAPTER 10
                               WILDLIFE
 10.1   OVERVIEW

     A large amount of  information must be  integrated to describe which
 impacts  to wildlife will occur  in particular  situations.  For example,
 the time span over which impacts occur, the type of disturbance, land
 area,  land types, vegetation component,  wildlife component, and wild-
 life ecology all must be considered.  In addition, feedback and inter-
 actions  among these factors must be considered.  So that this informa-
 tion can be more easily dealt with, several simplifications will be
 made.  These include a standardization of the above variables.  This is
 followed by a general discussion of characteristics which relate to im-
 pacts  on each wildlife group of birds and mammals.  This discussion is
 made first for immediate impacts, and then  repeated for long-term effects

     The term "ecological impact" implies two basic actions.  First,
 there  is  some disturbance to an area.   Second, there is a reaction
 (impact) which results from the disturbance.  Of interest are those im-
 pacts  caused by human disturbance which result in a change in population
 numbers  and/or species content of an area.   Since impacts can be subtle
 or gross, only those impacts which are quantitatively measurable in the
 field  are considered.

     The impact on an area usually has short-term and long-term com-
ponents.  The short-term component can be measured immediately following
the disturbance.   The long-term component must be examined for duration
as well as magnitude.   Some criterion  is needed to decide how much change
is required to "recover" the disturbed area (i.e.,  the quality of the
recovery), and how long the recovery will take.   However,  the product of
the recovery can vary.   If the land and vegetation return to a different
condition, a new wildlife community may become established which is very
                                10-1

-------
 different from the original  one.   Naturally,  the rate of recovery will
 vary  among land types  and  climates.

      This raises the question  of  the  location of the  end point  of the  im-
 pact.   Since  "impact"  is defined  as a change,  the end point  is  where
 changes become too small to  measure quantitatively.   However, natural  en-
 vironmental processes  are  dynamic; that  is, they fluctuate constantly.
 When  this fluctuation  is slow  the ecosystem is  considered stable,  or in
 equilibrium.   Likewise, wildlife  communities  are only relatively  stable.
 Thus  it is both unreasonable and  impossible to  restore a community to  its
 exact original  condition.  The practical solution is  to  restore the origi-
 nal natural conditions in  land and vegetation as nearly  as possible, and
 simulate  recolonization of wildlife.  A new, but somewhat different, wild-
 life  community  will become established in a dynamic equilibrium with the
 new and somewhat  different habitat.

     A  broad range of human  disturbances can result from the treatment of
 hazardous  material spills.   The least significant  disturbance is probably
 surface scraping  of a thin layer  of ground.  The impact  of this to wild-
 life might be unmeasurable.  Probably the greatest impact would result
 from defoliation  of plant  life,  followed by covering  the soil surface
 with cement or  some other  mechanical stabilizer.   If done on the maximum
 area size, this treatment  would cause immediate  local  extinction or emi-
 gration of the majority, if not all, of the wildlife in  the area.  It
 might also cause  abnormally high population densities  in areas contiguous
 to the  treated area due to emigration.

     In order to  evaluate  the impact of treatment disturbance in the dif-
 ferent  land types, a reference standard for "disturbance" is needed.
 This standard will be taken as the most drastic disturbance as described
 above.  In this case the immediate impact is that which follows the
 treatment.  The recovery phase is defined as following the erosion or
removal of the ground covering.  For simplicity this is assumed to be
removed at one point in time.  Less drastic treatments can be evaluated
across the land types by using the same rationale applied on a lesser
scale.
                                10-2

-------
     The impact to the local wildlife of a land treatment is heavily
dependent on the land area that is disturbed.   The areas of concern
range from 0.01 km  to 10 km^.  These areas can be envisioned as
squares:  the smaller is 100 meters on a side and the larger is approxi-
mately 3.16 km on a side.  Since a point-to-point discussion of this
entire size range is impractical, these end points plus a midpoint area
       2
of 1 km  (1 km or 1000 meters on a side) will  be discussed.  An addition-
al area 100 km^ is a*lso included because it approaches an upper "end
point", at which a measurable impact can be expected for almost all wild-
life species.  These should typify the desired range of areas.

     A wide variation in habitats exists for each land type under con-
sideration.  Again, some simplification is needed in order to organize
the wide range of impacts.  This can be done by using a procedure which
has proved successful in plant ecology; i.e.,  describing a typical
habitat type within each land type.  The following is an adaptation of
                  25
Whittaker's system   for this purpose:

  •  Deserts include a wide range of communities.  It is appropriate to
     note that two types of desert occur in the United States:  (1) warm-
     temperate deserts, represented by the widespread creosote bush com-
     munities and the floristically rich uplands of the Chihuahuan and
     Sonoran deserts, and (2) cool-temperate desert scrub, including the
     sagebrush semidesert of the Great Basin.   The floras of both these
     desert types share a scrubby habitat.  Also, in both a lack of
     moisture is critical to the maintenance of the various habitats.
     Thus deserts can be typified as those areas of extremely low
     annual rainfall.

  •  Prairie (or grassland) includes the great prairies, plains, and
     desert grasslands of the United States.  Despite reduction of
     vegetation structure to a single major horizontal stratum, plant
     species diversity in grasslands can be high compared with most
     forests.  These are areas of relatively low rainfall (but more
     than deserts) and experience a broad range in mild temperatures.
                                10-3

-------
•  Deciduous forests grow in moderately humid, continental, tem-
  perate zone climates, with summer rainfall and severe winters.
  Broad-leaved deciduous trees (oak, beech, maple, ash, basswood,
  etc.) are dominant in forests with trees 30 to 40 meters tall.
  Animal life of temperate deciduous forests is abundant but
  varies according to the season; diversity of plant and animal
  species is moderately low to moderately high.   Deciduous for-
  ests are extensively developed in the eastern United States.

  Coniferous (evergreen)  forests  occur in varied temperate zone
  circumstances.   Extensive needle-leaved evergreen forests occur
  in continental  climates  of the  western United  States,  and in
  places where  soil  characteristics,  or fire  frequency,  or both,
  favor pines over broad-leaved trees in the  milder climate of
  the eastern United States.  Coniferous forests are dominated
  by pine,  fir, spruce, etc.  A special  type  of  coniferous for-
  est is the subarctic-subalpine  needle-leaved forest  of the
  cold edge of the climatic  range of  forests.  These  areas exist
  in the northern  part of North America  and exist to  the south
  at  higher elevations in mountain areas.

 Mountain areas can be categorized as vertical  extensions  of
 other  land types, as well as those areas which are exclusively
 non-forested aeolian or alpine  regions.  Aeolian mountain areas
 are almost never the sole habitat of animals using them.   There-
 fore, wildlife responses detailed in the discussions of the
 other land types are applicable.  When mountain areas are ex-
 tensive, they may be very similar to tundra, and that land
 type should be referred to.  The reader is referred to these
 specific sections for treatment  of the wildlife of mountain
 habitats.

 Turidras are the  treeless  arctic  plains.  The vegetation forms
 are varied and there are  often complex patterns of dominance by
 dwarf-shrubs,  sedges and  gresses,  and mosses and lichens.  In
 many tundras the  deep layers of  the  soil are permanently  frozen,
 and only the surface soil  is thawed  and becomes biologically

                          10-4

-------
   active during the summer.   In many of these communities, re-
   peated freezing and thawing of the soil separates rocks from
   finer soil materials and arranges the rocks in parallel strips,
   or in networks with polygonal cells; striking internal pattern-
   ing of the plant communities in relation to soil differences
   results.  Vegetation cover decreases toward the sparse lichen
   communities, rock fields and ice of high elevations.

•  Coastal inter-tidal marshland can be salt-water or estuarine.
   These areas, especially salt-water marshes, are extremely rich
   energy sources.   Like prairies,  they are limited in vertical
   structure.  The plants of salt-water marshes are herb and forb-
   like.   Marshes are not usually extensive,  but are rather patchy in
   occurrence,  or form a narrow band along a lake or coast.

•  Agriculture  is unique,  being the substitution of a usually monotypic
   community for one or more natural plant communities.   Agriculture
   is practiced in nearly all climates,  although many hot and cold
   areas  are not yet occupied fully.   Agriculture can be characterized
   as managed fields with generally low vegetation (under 12 feet)
   dominated by 1 or 2 cultivated plant species.   The most common
   secondary species are those generally considered "weeds," i.e.,
   introduced herbs.   Agriculture is extremely widespread in the
   United States except where climate and/or terrain are exceed-
   ingly  harsh.

•  Suburban land types  are best described  as  those  areas which make
   up areas of  human habitation.   Like agriculture,  these areas are
   extremely widespread.   The numerous roads  and  human  activity
   associated with  this land type  severely restrict  the  type  of
   wildlife which can exist.   The human  dwellings are usually asso-
   ciated with  exotic introduced plants  and/or weedy  species.   As
   with agriculture,  this  land type  is an  imposition of  a new com-
   munity upon  a previous  and  older natural community.   This  favors
   a  low number  of  species,  but with high  densities.
                              10-5

-------
      Some species have extremely low population numbers and must be
 examined individually to assess potential cleanup impact.   Identification
 of these sensitive species can be obtained from Federal and State lists
 of threatened and endangered species.5'22  State lists are to be pre-
 ferred since they often include recent information on steps which have
 been taken to protect the species.   State agencies also usually have
 recent literature and staff knowledge which may bear on the problem.

      If a habitat which supports such a species is disturbed,  several
 things may be done to speed up the  recovery phase for the  animal.   This
 may be necessary for disturbances to rare species.   For example,  the
 Trumpeter Swan was close to extinction (about  65 birds)  in 1935.7   Areas
 where these birds breed were designated as off-limits to hunters.   Also,
 additional  food was  made  available  at  these  sites  in  winter.   Finally,
 the birds were introduced  to  other  potential breeding sites.   Due  to
 these steps,  the population  had  recovered  (to  about 915  in  the  U.S.)
 by 1976.    Other direct steps can be taken to  accelerate faunal recovery,
 that  is,  increasing  the rate  of  vegetative recovery.

      Faunal  recovery is strictly related to the  recovery of the vegeta-
 tion.  The  fauna cannot normally respond before  the vegetation has
 done  so.  There  are  three  components of the vegetation which are particu-
 larly important  to animals.   The  first is the plant species which colo-
 nize  a disturbed  area.  These are of especial importance to the herbivores,
 The other two  components are vegetative height and density, i.e., the
                                  q
 vegetational  structure  (see Emlen   for an exception).  All  three of the
 above  components  will change  over time in the successional  processes of
 the vegetation.   Faunal recovery must, to some degree, be an upward pro-
 cess  in terms  of  the food  chain.  Carnivores cannot colonize an area
 until their prey  is available.  Since abundances may  increase slowly by
 reproduction,  or  quickly by invasion, it is difficult to predict a re-
 covery for animals at the  top of the food chain.  However,   these effects
 can be mitigated  for those predators which can utilize a wide variety of
prey, so that food can be caught during several stages of the vegetation
succession.
                                10-6

-------
      Once vegetational succession has produced a new climax community,
 the more similar the new vegetation is to the original, the more similar
 the new fauna should be to the original fauna.  This does not require a
 vegetation be established that is identical to the original.  All that
 should be required is an approximation of the original vegetation which
 provi-es the same resource value in terms of food and structure.  The
 critical questions to wildlife restoration are the "quality" and duration
 of vegetation recovery.

      Plant  succession has  been  well  studied  by plant  ecologists25 but
 the  faunal  community  succession process which  occurs  in nature has not
 been studied  as much.  As  plant succession occurs, multiple  feedbacks
 are  possibly  among and within plants and animals  (not to mention soil
 content  changes, ground-level temperature regime cahnges, etc.).  The
 network  can be extremely complicated.  However, faunal changes can be
 closely  examined with respect to plant succession, and different animal
 species  will  show greater or lesser facility for recovery.

      In  most  cases some species which were not among  the original fauna
 of the area will move in during part of the  vegetative recovery.  These
 species  will  probably move out  of the disturbed area  as the  habitat
 changes  toward one more similar to the original one.  This causes a
 decrease in species diversity as the climax of the faunal succession is
 reached.  The pattern of plant  species diversity in succession emerges
 as a rise from the start of recovery to a pre-climax peak, and then a
 leveling off  at a slightly lower level as the dynamic equilibrium of
 the  climax ensues.
     The few  studies that have  been done on faunal succession draw
 evidence from the fossil record and emphasize evolutionary (very long
term) time.   Of concern here is ecological, not geological,  time, i.e.,
a relatively short period of time during which succession may occur,
but evolution cannot.  9  Thus this discussion emphasizes the responses
which are expected from individual species, rather than faunal succession.
                                10-7

-------
 10.2  EFFECTS  ON BIRDS
      Much information must  be  integrated  in  an examination  of impacts  on
 birds.   Unfortunately,  much of this  information deals  with  particular
 species,  making  generalizations difficult.   However, in  order to understand
 impacts  on birds some generalization is necessary,  as  a  discussion  of  each
 species  individually  is impractical.  Therefore,  impacts are  described
 for  groupings  of birds  which share some general life history  and/or feed-
 ing  characteristics,  or are closely  related.   These groupings are examined
 for  short-term effects  (immediate impacts) and long-term effects  (recovery
 phase),  then each land  type is  examined for  specific short- and long-term
 effects.
      Measurement of a change in the  numbers, or population, of a bird
 species  is  not an easy  or unambiguous process.  First  the term "popula-
 tion" must  be  defined in some way which is not necessarily  land-area de-
 pendent.  Once this is  done a  census method must  be picked  from those
 which have  been  shown to be accurate. '    But even after a standard for
 measurement is chosen,  and  the  data  taken, it  is  still necessary to put
 the measured impact into context.  What is really desired is  the effect
 on the local species.   This effect can only be ascertained by having a
 local wildlife expert analyze the data for the disturbed area and gener-
 alize the measured impact into  an impact value based on  local  abundances.

 10,2.1  Short-Term Effects
      In general,  a disturbance  to an area will  be much more serious to
 permanent residents than to migrators.  This is especially true where the
 disturbed area is along a migration route.  If adequate alternative routes
 of migration are  available  there may be no quantitative impact.  Where
 only one route is possible,  if  food is not taken  from the area by the
migrants the impact may be  very slight.   However, once alternative routes
 of travel are used, it may prove very difficult to reintroduce the dis-
 turbed area as a part of the migration route.  This will depend on how
many of the migrators are lost, whether the alternative route is really
 as adequate as the original, and whether the route is learned or innately
 known.  The slow rate of reintroduction of residents and migrants who
                                   10-8

-------
 feed on an area should be slightly greater for the more mobile migrant
 populations.
      The seasonality of a disturbance is of extreme importance to birds.
 In general,  disturbances in the spring will have a much greater initial
 impact than  those in other seasons for two principal reasons.   The first
 is that the  breeding season occurs in the spring for most  American bird
         24
 species.     The second reason is that the spring is also a very critical
 time of growth for plants.   Naturally,  a disturbance in another season
 can affect spring conditions;  however,  most bird species can react to  ad-
 verse habitat  conditions before breeding occurs.   For example,  they may
 nest elsewhere,  or at  least nest in lower densities.
      Among most  birds,  a disturbance to the breeding grounds is generally
 more serious than an equal  disturbance  to the  wintering range.   Many
 birds hold territories  during  the  breeding season,  but  form flocks during
 the winter.  Flocks  can lose members  to other  flocks  when  food  is  scarce,
 or can remain  cohesive  at  a small  cost  to each of the members.   This is
 not true of the  destruction of breeding territories.  The  area  lost here
 will  almost always result  in a lower  number of breeders.   Thus  an  immediate
 impact is  more likely to be of greater  magnitude  when breeding  habitats
 are destroyed.
      Some  birds  feed away from the nesting  area  or  hold  feeding terri-
 tories around  the nesting area.  The  nesting can  be done in tightly
 packed colonies  or, when food  is taken  at  some distance  away from  the
 area  of the nest, in looser  aggregations.   Those  species which  hold
 feeding-nesting  territories  are much more spaced  out.  Obviously,  dis-
 turbance of a  colonial breeding area will affect more birds than the
 same  area disturbance to a  feeding-nesting  territory.
     A strong  correlation between body size and feeding-nesting terri-
                                               20
 tory size has been shown for many bird species.    Thus a disturbance
of a given-sized area will usually affect more members of small species
than larger ones.  The loss of an area of 0.01 km  can cause the loss
of six song sparrows' feeding-nesting territories.  However, the same
area is only  0.25 percent of a red-tailed hawk's feeding-nesting terri-
             2
tory.  A 1  km  area is roughly a quarter of the size of a red-tailed hawk's
                                   10-9

-------
 territory,  and  is  a  reasonably  large area  in  terms of most  smaller birds.
 In  any  case,  there can be  little doubt that the  loss of an  area of 100 km2
 from  the breeding  range of almost any species will have a measurable impact

 10.2.2  Long-Term  Effects
      Birds  respond very quickly to habitat changes due to their high mo-
 bility.  When the  proper conditions prevail, bird dispersal is facilitated.
 However, different life histories will cause some differences.  Size dif-
 ferences will also be critical, as larger birds  can expand  their terri-
 tories more easily than smaller birds.  The following differences also
 hold  for some broad  groups of birds.
      In general, migrators being more mobile than permanent residents,
 will  colonize areas  more quickly.  However, where migration routes and
 summer/winter grounds do not border on a suitable area, colonization can
 be much slower.  Permanent residents have an advantage in being able to
 establish themselves at a  time when migrators are not present.  Some
 species, such as the Trumpeter Swan, can migrate or not, depending upon
 food  resources, and  will recover accordingly.
      The area used for breeding grounds by migrators may spread at a
 somewhat more rapid  rate than that of permanent residents.  However,  the
 choice of wintering  grounds is likely to be even more flexible.  The par-
 ticular migratory  species which uses an area will probably depend on the
 successional stage of the vegetation.  Some species of migrators may
 colonize new breeding habitats more quickly than others.  These effects
may not have corresponding effects on the wintering range except for a
 change in the number of young present.   However,  since the breeding ter-
ritories must be reestablished each year, complicated variable changes in
 the species pattern on the breeding grounds may occur.   The damping of
these changes may extend beyond the vegetative recovery phase.
      Feeding flocks are apt to be rapid colonizers since their foraging
method entails high mobility and allows them to exploit a thinly spread
         2
resource.    Birds with feeding territories usually colonize from a
"leading edge" of the population.   This movement  is necessarily slower.
The recruitment into flocks may be somewhat slower,  by reproduction,  so
                                  10-10

-------
that even though a flock may use an area, the vegetation is not exploited
to as great an extent as by the slower colonizing territorial species.
     Competition among species may be important to recolonization.  The
first bird species to colonize an area can move in unopposed.  However,
the species which replace it in faunal succession may have to compete
with the established species in the area.  The net effect is one of pro-
longing the faunal successional process.  Just how long the succession
takes will therefore be somewhat dependent on the intensity and duration
of the total amount of interspecies competition that occurs.
     The following "trophic/niche" groups of birds contain families whose
species share trophic level and lifestyle characteristics.  Naturally,
some families have members in more than one group.  These families are
divided into the respective groups in the following discussion.  Examples
of typical families are:
     1.  Large carnivorous birds:  American vultures, buteo hawks
         (Accipiter family), and large owls.
     2.  Medium-sized carnivorous birds:  medium-sized falcons, kites
         (Accipiter family), and medium-sized owls.
     3.  Small carnivorous birds:  kestrels (Falcon family), shrikes,
         and small owls.
     4.  Large primarily herbivorous birds:  grouse, pheasants, and
         turkeys.
     5.  Medium-sized primarily herbivorous birds:  mockingbirds,
         thrashers, thrushes, blackbirds, and quail  (Pheasant family).
     6.  Small primarily herbivorous birds:  hummingbirds, titmice,
         and sparrows (i.e., almost all of the Fringillid family).
     7.  Medium-sized insectivorous birds:  woodpeckers, tyrant fly-
         catchers, and swallows.
     8.  Small insectivorous birds:  wrens, vireos, and wood warblers.
     9.  All sized truly omnivorous birds:  ducks, jays, crows, and
         gulls.
    10.  All sized aquatic feeder birds: plovers, sandpipers, and
         phalaropes.
     Some land types take longer to restore than others.   This can be
considered the sensitivity of the land type to disturbance.   The reasons
for sensitivity can be the magnitude of the disturbance,  the amount of
vegetative structure,  and climatic conditions  present,  the length of the
                                  10-11

-------
 growing season, the soil types present, and other factors.  These many
 factors are integrated in the analysis of recovery of vegetation and land
 types over time.  Here, differences in the responses of the bird faunas
 to the immediate disturbance and the subsequent recovery of the different
 land types are described.
      Table 10-1 gives examples of bird species from each trophic/niche
 group in each land type which yield a measurable impact if the maximum
 disturbance has been made on the specified (or larger)  land areas.   The
 table uses area size differences among groups  of birds  to show how the
 groups are affected.   This  should not be equated with the qualitative
 "importance"  of a  species.   For example,  more  robins  would be  lost  than
 red-tailed hawks for a disturbance to a habitat for both.   But the  hawks
 may be more "important" than the robins in terms of numbers in existence
 or aesthetics.
      The table  assumes that  at  least  ten  breeding territories  must  be
 lost  to a  species  before  impact  can be  measured.   No  literature is
 available  upon  which to base an  estimate  of the number  of  breeding  pairs
 which must be removed  from a habitat  to yield  a measurable impact,
 The figure of ten  was  chosen because  it represents  a  number at which
 large and/or uncommon  species will  begin  to show a measurable  impact.
 The figure of ten  is thus a  minimum.  The  real  figure should be higher
 for many species.  However,  lower  figures  should  seldom occur.   A
 further assumption is made that measurement is  solely on a  local scale.
 Bird  breeding territory area is used because the breeding  grounds are
 of  the greatest immediate importance to the multiplication  (reproduction)
 of most populations.  Bird breeding territories may vary with  food
 resources  in some cases.  However, breeding area should generally be
 less variable than any other type of defended area used by individual
birds.

     10.2.2.1  Deserts—
     The impact of a disturbance in deserts will be greatest for small
herbivores, medium  sized insectivores, and all  carnivorous birds.  An
added impact on carnivores is the loss of a great diversity of reptilian
                                   10-12

-------
Table  10-1.
Examples of bird species  impacted  by maximum  disturbance
to  specified types  and  sizes of land areas.
MAJOR GROUPS
1 . LAND TYPE
DESERT
Large Carnivores
Medium Carnivores
Small Carnivores
Small Herbivores
Medium Insectivores
PRAIRIE
Large Carnivores
Medium Carnivores
Small Carnivores
Large Herbivores
Omni vores
DECIDUOUS FOREST
Large Herbivores
Medium Herbivores
Smal 1 Herbi vores
Medium Insectivores
Small Insectivores
CONIFEROUS FOREST
Small Herbivores
Medium Insectivores
Small Insectivores
TUNDRA
Large Herbivores
Aquatic Feeders
COASTAL INTER-TIDAL
Medium Herbivores
Small Herbivores
Medium Insectivores
Small Insectivores
Omni vores
Aquatic Feeders
2. LAND USES
AGRICULTURE
Large Carnivores
Medium Carnivores
Small Carnivores
Medium Herbivores
Small Herbivores
SUBURBAN
Medium Herbivores
Small Herbivores
Medium Insectivores
Omni vores
SIZE OF LAND AREA
0.01 km2




Franklin's gull

Least flycatcher
Whi te-eyed vireo





Song sparrow
Bank swal low


Robin
House sparrow
Cardinal
Song sparrow
Cliff swallow
1 km2


Loggerhead shrike
Black-throated sparrow
Ash-throated flycatcher

Greater prairie chicken
Common crow

Ruffled grouse
Baltimore oriole
Black-capped chickadee
Acorn woodpecker
Tennessee warbler

Pine siskin
Western flycatcher
01 i ve warbler

Wi 1 low ptarmigan
Whimbrel

Yellow-headed blackbird
Marsh wrens
Tree swallow
Yel low-throat
Pintail
Millet


Loggerhead shrike
Mockingbi rd
Chipping sparrow
Mockingbird
House finch
Black phoebe
Blue jay
10 km2


Elf owl

Sparrow hawk










Barn owl
Burrowing owl
Sparrow hawk

100 km2


Red-tailed hawk
Prairie falcon

Golden eagle
Prairie falcon
Common raven












   Territory sizes taken from Melty24 and Schoener20 or estimated from sizes of similar species.
                                in

   Species nomenclature follows Peterson,  where scientific names are also given.


   Member of the sparrow family.
                                       10-13

-------
 prey.   Large- and medium-sized herbivores occasionally occur in desert,
 usually on its fringes,  as is true of most small  insectivores and omni-
 vores  which live in or near deserts.   The desert  itself is  not the sole
 habitat of these species,  and impacts are thus  lessened.
     Resident desert bird faunas  should  contain a high proportion of
 specialist species which have low population  densities.   These birds
 will be extremely slow to  respond to  most recovery situations.   The
 generalist species which do occur in  desert can usually be  expected  to
 expand their  numbers beyond the normal level  upon colonizing  a disturbed
 area.   In  this particular  habitat  the presence  of generalists  in abundance
 will probably further retard  recolonization of  specialists  to  a measurable
 degree.  Management  may  be extremely  useful in  this unusual case.

     10.2.2.2 Prairies-
     Prairies are remarkable  for  having  a distinct  lack of  vegetative
 height.  Large herbivores  and omnivores  are the groups most impacted by
 a disturbance here.   These are closely followed by  carnivores  of all
 sizes.  Carnivores  feed  not only on bird  prey on  prairies,  but  also  on
 the high populations  of  mammals which are  common  herbivores.  Medium-
 sized  and  small herbivores, while not as  diverse  on prairies as  in other
 land types, may still be very abundant.  Once again, as smaller herbi-
 vores  are  examined,  the  impact of a disturbance is usually magnified.
Medium-sized  insectivores  often feed over prairie, but seldom nest there.
Thus the impact to this group should be reduced.  Even less impact should
be seen in aquatic feeders, who use prairie areas  as stopping-feeding
sites during migration.
     Prairies contain an imbalance of generalist  species,  mainly due to
 the high carnivore and omnivore diversity.  Thus the recoveries of prairie
bird faunas are largely secondary effects, since both the above groups
depend to some extent on animal prey.   Carnivore recovery may lag far
behind the vegetative recovery.  The omnivores,  on the other hand, are
generalists and will recover more quickly.  Much more omnivore feeding
than breeding occurs in prairie areas, also facilitating the return of
this group.  The remaining breeding species are  herbivores which are
                                  10-14

-------
 generally mobile species for their size.  They will recover relatively
 quickly.

      10.2.2.3  Deciduous Forests-
      Deciduous forests have good diversity and abundance of herbivores
 and insectivores of all sizes.   These groups will bear the major impact
 of a disturbance in most cases.   Here also large and medium sized
 carnivores will occur mainly where there are breaks in the forest,  and
 be impacted in proportion to the disturbed area of the forest breaks.
 Small carnivores and omnivores,  while lacking in diversity in these
 forests, may be locally abundant.   These last two groups should be  wide-
 spread where they occur, thus lessening potential impacts upon them.
      Deciduous forests are  generally richer in bird fauna than coniferous
 forests and relatively higher proportions  of specialist  species occur.
 Thus,  the faunal recovery will take longer in deciduous  forests.  However,
 these forests also  have the advantage of being extensive,  so  colonists
 are available.   These bird  faunas  may be extremely sensitive  to dis-
 turbance,  since the specialists  may have a more  "patchy"  distribution
 as  well as  lower population densities.   Because  of their  size  and lesser
 mobility,  the smaller herbivores and  insectivores  may  be  very  slow  in
 recolonization.

      10.2.2.4  Coniferous Forests—
     The  disturbance  of coniferous  forests  should  impact  small  herbivores
 and  insectivores  of all  sizes to the  greatest  degree.  A great  abundance
 and  diversity of  these  groups occur here.   In  breaks in the forest,  large
 and mediiM-sized  carnivores occur and will be  impacted where the breaks
 are disturbed.  Coniferous  forest supports relatively  less small carni-
 vores,  large  and medium-sized herbivores, and omnivore diversity.  Where
 these groups  occur, however, they may do so in abundance, and thus be
 impacted.  Again, these last groups (except the small carnivores) are
more likely to occur in breaks in the forest.
     Coniferous forests tend to contain a relatively limited fauna
 compared with the other land types.  However, the number of specialists
                                   10-15

-------
  and generalists is usually evenly balanced.   Naturally,  the generalists
  will respond to vegetative recovery more  quickly.   However, since  coni-
  ferous  forests  are generally extensive, surrounding forest  areas are
  almost  always present  to provide  colonists.   As vegetative  recovery nears
  completion,  more and more  specialists will move in  to  supplement the
  labile  generalists.

       10.2.2,5   Aeolian Mountain Peaks—
       A  disturbance  to a mountain-alpine habitat will affect only those
 few birds which  have adapted to this barren land type.  An  example of
 one of  the few  species of birds which are specialized for this type of
 habitat are the Ptarmigans  (in the Grouse family).  During winter,
 alpine birds move to lower elevations,  and live in other land types.
 Herbivores of all sizes will receive the greatest impact from an environ-
 mental disturbance, because of the disruption in their food supply.
 Large and medium sized carnivores may use  alpine areas for hunting  and
 breeding, but carnivores will be impacted  to a lesser degree due to
 their ability to move  into  other types  of  habitat  below treeline.   The
 other trophic/niche bird groups should  rarely occur  in alpine areas and
 thus receive  no  impact  from a disturbance  to it.
      Due to specialization  of herbivorous  alpine birds  and the patchy
 and  disconnected nature  of  the alpine habitat,  recovery from a disturbance
 will  be  an arduous process.   Mountains are to  a large  extent "islands"
 separated by  a "sea" of  forest;25  therefore, recolonization  of alpine
 areas  may only be possible  from within the original  population.  Many
 alpine birds  winter  in nearby forests, thus, the more isolated a mountain
 is,  the  longer it will take  for colonists  from  other mountains to arrive.
 Since  large and medium carnivores  of  alpine areas are generalists they
 will recover more  quickly.  However,  they  will  only be able  to recolonize
 alpine areas as  their prey also recolonizes.

     10.2.2.6  Jundra Areas--
     The tundra is commonly only the summer range of species which breed
there.  However,  a group of large herbivores  (ptarmigans) are resident
on tundra, and thus are subject to the greatest impact of any disturbance.
                                   10-16

-------
 Impact on the principal summer breeding migrants,  the aquatic feeders,  is
 also very serious.   Many of these birds nest in aggregation,  so impacts
 are either drastic  or considerably dampened, depending on the location
 of the disturbance.   A few omnivores (gull  species)  share this charac-
 teristic and nest on or near tundra.   Two large carnivores (the snowy
 owl and gyrfalcon)  feed on tundra and  breed nearby.   These last two  birds
 have large ranges which help to mitigate any impact  upon them.
      Tundra bird  faunas are largely migratory,  so  that bird recovery can
 be relatively rapid.   However,  since the bulk of breeding for the  aquatic
 feeders occurs here,  the recovery of breeding populations must  generally
 follow the recovery  of food resources.   The yearly freezing of the ground
 does not encourage invertebrates,  so that more  vegetation must  be  eaten.
 The only resident birds,  the large herbivorous  ptarmigans,  are  also
 closely tied to vegetative recovery.   Where tundra occurs it  is usually
 extensive,  so that the bird fauna can  recolonize an  area easily.

      10.2.2.7  Coastal  Inter-Tidal  Marshlands—
                                                2S
      Marshlands are usually rich  energy  sources    and  therefore all  non-
 carnivorous  groups except  large herbivores  are usually represented.
 The small  insectivores  and aquatic  feeders  are especially noticeable,
 the former even more  so  in fresh-water marshlands.   Since those species
 which live in marshes  are  often restricted  to it,  they will be  impacted
 severely by  a disturbance.   Large-  and medium-sized  carnivores  which
 frequent marshes  usually include other areas  in  their  hunting range,
 or  can  easily do  so.  They are therefore  less impacted.  Many aquatic
 feeders  and  omnivores use marshland as stopping  and/or  feeding  places
 during migration, and can  be subjected to a  lesser impact by changing
 their migration routes.
     Marshland  bird species are generally very sensitive  in terms of
 faunal recovery.  A very high relative proportion of specialists occurs
here.  Their recovery is complicated by the patches in which this habitat
occurs.  If the disturbance occurs over an entire patch of marshland,
recovery will be much slower than that expected for a disturbed area
within a patch.  Thus, disturbed area size effects  are especially
                                   10-17

-------
  important.   Size  effects will  be greatly magnified  for  smaller  birds,  but
  may have relatively  little  significance for  the more mobile  large-  and
  medium-sized carnivores.

  10.2.3  Land Uses
      10.2.3.1  Agricultural Areas--
      Agricultural areas contain either breeding or feeding areas for
  some members of all the trophic/niche groups except small insectivores.
 However, the groups which are most impacted by disturbance of agriculture
 are the medium and small herbivores and the carnivores of all sizes which
 prey upon them.  Large herbivores,  medium insectivores,  and omnivores are
 more likely to feed on agricultural areas than to breed  on them, and will
 usually be less heavily impacted.   Aquatic feeders are the least likely
 group  to be impacted  since  they occur only under special conditions
 (excess water or ponds available).   However,  the impact  on this  group
 increases  if they can breed where an area is  disturbed,  again a  special
 case.

     Agricultural  land mainly supports generalist  type bird  species  (i.e.,
 those with  a broad niche).   This land type  has  not existed long  enough  for
 specialists (birds with a narrow niche) to  evolve  necessary  adaptations
 for habitat specialization.   The few specialists found here  were pre-
 adapted  to  some special feature of  the agricultural  habitat,  and moved
 in from  other habitats.  The result  is an extremely  resilient  bird
 fauna.   This fauna will follow  vegetative recovery very  closely.  Since
 the species which  occur in agricultural areas have dense populations,
 dispersal from  surrounding populations  should send almost  constant waves
 of colonizers into the  disturbed area.
     A second alternative for agricultural land is to allow it to revert
 to a state more similar  to its preagricultural condition.  In this case
 the agricultural condition can be considered an additional disturbance
 from which the avifauna must recover.  In order to evaluate this particular
 type of recovery, the sections dealing with the preagricultural land
 type should be consulted.  Differences from these  land type recoveries
should  not vary greatly due  to the  agricultural use of the land.   The
                                  10-13

-------
avifauna will change from an agricultural one to one commensurate with
the particular land type's plant succession.

     10.2.3.2  Suburban Areas--
     Suburban birds are generally birds which were preadapted to live
with humans.  This implies a versatile lifestyle and a broad niche.
The groups which would be impacted (about equally) by a disturbance are
medium and small herbivores, medium-sized insectivores, and omnivores.
Groups which occur on the fringes of suburban areas are large carnivores,
large herbivores, small insectivores, and aquatic feeders.  These will
be lightly impacted by a suburban disturbance, if at all.
     Suburban bird faunas will also respond quickly in recovery of a
disturbed area.  Evidence for this is abundant in almost any spreading
city.  A heavy proportion of generalists exists in most suburban habi-
     9                                                      12
tats.   In addition, rather dense populations usually occur.    These
areas are usually extensive or at least connected by corridors such as
freeways, construction sites, and/or rural human populations.  Thus
colonization can be extreme during the recovery of the land type.  In
addition, the recovery of many species need not be tied to vegetation
recovery since human dwellings provide structural diversity, and may
provide food.
     As in the agricultural land use, the alternative of allowing the
suburban area to succeed to something more like the presuburban land
type exists.  However, this process is complicated by the presence of
semipermanent structures (e.g., buildings) which increase bird habitat
heterogeneity.  Without these structures a recovery similar to that
described in the appropriate land type is possible.  However, if suburban
land is simply deserted, species which depend on suburban structures for
nesting and/or roosting (or cover from predators) will be favored.
These will generally not be the species commonly found in the presuburban
land type, and recovery will be retarded.
                                  10-19

-------
 10.3  EFFECTS ON MAMMALS
      The impacts on mammals, like birds, can be described by examining
 mammals in groups which share similar lifestyles and feeding habits, as
 well as in groups which are closely related.  The general characteristics
 shared by these various groups will first be examined for short-term
 (impact) effects, and then for long-term (recovery phase) effects.
 Then each land type is examined for the specific characteristics which
 relate to the mammal fauna for short- and long-term effects.
      There are two types of classification of mammal living areas:  the
 defended territory,  and the total area traveled in by an animal during
 its  daily activities (excluding migration routes),  or "home range."   The
 latter area is of concern in a discussion of impacts since  it  is probably
 closer to the area size necessary for survival.   However, it may only be
 an approximation  of  the area which is essential.   In some cases mammals
 form aggregates,  or  colonies.   These  colonies  usually have  a home range,
 but  impacts  are difficult  to assess.   The loss of an area of habitat
 can  cause  either  the  loss  of a  colony,  or only the  loss  of  some of the
 colonial  individuals.   There are  little  data with which  to  even begin
 to examine these measurement problems.   However,  several  methods  of
 varying accuracy do exist  which can be used  for the  estimation  of mammal
 home range.

 10.3.1  Short-Term Effects
     The impacts to mammals  which share  similar size and  trophic  level,
or trophic/niche group, are generally consistent.  Therefore the following
groups are defined, with examples of the mammalian families which are
representative of the groups.
     1.  Large carnivorous mammals:  wolves and mountain  lions.
     2.  Medium-sized carnivorous mammals:  foxes and small  cats.
     3.  Small carnivorous mammals:  skunks and weasels.
     4.  Large primarily herbivorous mammals:  wild boar; deer;
         pronghorn; and bison, goats,  and sheep.
     5.  Small primarily herbivorous mammals:  10 rodent  families,
         pikas, and hares and rabbits.
     6.  Large omnivorous mammals:  bears.
                                  10-20

-------
      7.   Small  omnivorous mammals:   opossums,  raccoons,  ringtails,
          and  armadillos.
      8,   Insectivorous mammals:   shrews  and moles.
      9.   Flying mammals:  two bat  families.

      10.3.1.1   Large  Carnivores--
      The  primary  immediate  impact  of a land disturbance  to the  large
 carnivores  should be  negligible for  the  range  of area  sizes under
                                  2
 consideration.  An area of  100 km  is equivalent to only about  one to
                              23
 six mountain  lion home ranges.     Due to the wide ranging of these cats,
 even  a much larger disturbance would be  very hard to measure in terms of
 a  change  in population numbers.  However, a secondary  impact to this
 animal is the loss of its prey.  Again this effect is  probably  unmeasur-
 able, since the deer  they feed on  are also large animals which  are
 widely distributed.
      A slightly different situation  exists for the other large  carnivore
 group, the  wolves.  Measurement of populations is extremely difficult,
 even  for  the  exceedingly common coyote.  But in addition to measurement
 problems, the grey (timber) wolf also forms packs.  A  disturbance to the
      2
 100 km  area  size could eliminate  (from  the area) a minimum size pack of
 two,  but  this area is only  about one-tenth the home range of a  pack of
 . ,   23
 eight,    so that primary impacts are impossible to assess.  Some attempt
 can be made to  measure the  change  in pack size as an indication of impact,
 but this  is an  extremely variable characteristic.  Again, secondary im-
 pacts are possible, but they are unlikely to be measurable.  In all large
 mammalian species emigration is the most likely cause  of a drop in popu-
 lation numbers  after  a disturbance.

      10.3.1.2   Medium-Sized Carnivores--
      Impacts on medium-sized carnivores are also difficult to measure.
 Once again, the maximum disturbance to the maximum area  size specified
      2
 (10 km )  for consideration should not produce a measurable impact in
most cases.  The major prey of these carnivores,  small herbivorous birds
 and mammals, will be impacted at some area size,  so that it may be
possible to get some measure of secondary impacts.   The medium-sized
                                   10-21

-------
  carnivores are not as mobile as the  large carnivores, so that it may
  be possible to estimate an impact in some cases for a disturbed area
               2
 size of 100 km.
      10.3.1.3  Small Carnivores--
      Many of the small carnivorous mammals are nocturnal, and thus
 ignored by the majority of humans.  However, it is in this group that
 impacts in the larger area sizes can be measured.   Weasels will have
 measurable impacts where pre-cleanup prey densities are either high,
 or are moderate but extensive in area.   The other small carnivores
 have larger home ranges in general and  will be much harder to examine.
 Secondary (prey)  effects need not be examined for the groups which have
 measurable impacts, since these effects are included in the change in
 population size.

      10.3.1.4  Large Herbivores—
      With the large herbivores size of  the cleanup area again becomes
 important.   This  difficulty is compounded by the  ability of most  large
 herbivores to change home ranges  seasonally (by migration).   Local
 changes  in population size can probably be measured for disturbances
              2
 of the  100 km  area size.   However,  these changes  will  probably be
 measurable only immediately,  for  the high mobility of the  larger  herbi-
 vores will  soon reduce  the possibility  of finding  many  of  the  animals
 which reside  in the area,  since they may leave  it  or  perish.   Animals
 which successfully  leave  the  area are not  defined  as  being  impacted.

     10.3.1.'5   Small  Herbivores--
     Small herbivores are  generally restricted  to  living their entire
 lives in relatively  small  areas.  The once popular  idea that some small
mammal herbivores migrate  has been thoroughly disproved.13'14  These
animals are the most  important United States mammals in terms of species
diversity and abundance.  Their populations are easily sampled and moni-
tored in most cases.  In addition, they form an important link between  '
primary producers (plants) and the carnivores which depend on them for
food.11
                                  10-22

-------
      Table 1Q-2 gives examples of small herbivorous mammal species that
 occur in each land type which yield a measurable impact if the maximum
 disturbance has been made on the specified (or larger) land areas and
 types.  The table indicates the kind of rodents and lagomorphs (the
 rabbit order) which occur in the various land types and the general
 disturbed area sizes which will produce impacts.  The density of these
 animals is relatively easy to measure.
      Table 10-2 assumes that at least 12 to 15 home ranges must be lost
 in an area before a local impact can be measured.   As in the bird section,
 this home range (area)  amount is a minimum for measurable impact, except
 for large and/or uncommon mammals.   A special case of vole and lemming
 home ranges being very  small is somewhat counteracted by their excep-
 tional movement abilities,  so that  a larger area must be disturbed for
 these species than just 15  home ranges.   Home range size usually  varies
 with changes  in the density of food  resources.   Special  cases  in  specific
 locations are to be expected to vary from the predictions  of the  table.
      It may be that small herbivore  species can  be  used  as indicators
 to  measure  the effects  of various disturbances.  Small herbivore  species
 can change  at the  onset  of  a disturbance.16  As  succession ensues,  changes
 in  species  composition  and  density can  also occur.16  When the  original
 species distribution and density reoccur,  an  advanced stage  of  recovery
 has probably  been  entered.   A  distinction must be made,  however,  between
 those  small rodent  species  considered "pests," or those  deemed  undesirable
 by  man, and species which are more beneficial to man's interests.  Pest
 species are generally exterminated periodically where they exist near
 humans and  impacts  on them,  although measurable, may not be of  real
 concern.

     10.3.1.6  Large Omnivores--
     Large omnivores (bears) are another group for which it is exceedingly
difficult to measure impacts.  Only three bear species occur within the
designated land types in North America.  These will not be measurably
impacted for disturbances to the land area sizes under consideration.
For example, even without the problems of impact measurement, the  100 km2
                                   10-23

-------
Table 10-2.  Examples of small  herbivorous mammals impacted by maximum
             disturbance to specified types and sizes of land areas.a
  MAJOR GROUPS
                                   SIZE  OF  LAND AREAb
   0.1  km'
  LAND TYPE

     Desert
     Prairie
     Deciduous
     Forest
     Coniferous
     Forest
     Tundra
  LAND USE

     Agriculture
     Suburban
Cactus mouse
Pocket gophers
Pocket mouse
Harvest mouse
Grass voles
Jumping mouse

Deer mouse
Deer mouse
Phenacomys
Redback voles
Lemmings
Tundra redback vole
House mouse
Pocket gophers(
Cotton ratsc

Black ratc
Norway ratc
House mouse0
Antelope  squirrels
Kangaroo  rats
Wood rats
Jackrabbits

Ground squirrels
Prairie dogs
Jackrabbits
Tree squirrels
Flying squirrels
Chipmunks
Cottontail rabbits

Marmots
Tree squirrels
Flying squirrels
Chipmunks

Showshoe hare
Ground squirrels
Jackrabbits0
    Home range sizes  taken  from McNab,    Vaughn,    and  Burt  and
    Grossenheider,4 or estimated from  home  range  sizes  of  similar
    species.   Nomenclature  of the genera  and  species  follows Burt
    and  Grossenheider,^ where scientific  names  are  also given.
  b                          2
    For  land  areas  of 0.01  km  or less  no small mammal  herbivore
    should be impacted, above 1  km2  all should  be impacted.
  £
    Denotes pest species or genera.
                                10-24

-------
 area size is only about half the size of the Grizzly bear home range.23
 However, the actual effects on these animals can be serious since they
 are restricted to isolated patches of wilderness areas,4 and emigration
 may not always be possible.

      10.3.1.7  Small  Omnivores—
      The impacts from disturbance of the larger area sizes may be
 measurable for the small omnivores.   These animals are generally nocturnal
 and occur in areas which are human-disturbed (mainly agricultural).   Im-
 pacts should be easier to measure for opossums,  and more difficult  to
 measure for the wider ranging raccoons.17  These animals are widely dis-
 tributed where they occur,  so that immediate impacts are difficult  to
 measure due to a reassortment of individuals during movement away from
 the disturbed area.

      10,3.1.8  Insectivores--
      Insectivores  are small  mammals  with extremely limited mobility.
 Like small  rodents, they are closely tied to their area  of residence
 and can be  expected to  be measurably impacted in  disturbances  to  all  but
 the smallest  land  area  sizes.  In'addition,  these animals  depend  on insect
 and other small  animal  prey,  so  that  a disturbance  to the  prey can cause
 as  great an impact as a direct disturbance to the  insectivores themselves.

      10.3.1.9   Flying Mammals—
      The impact  from  a  disturbance on the  flying  mammals  (bats) is
 roughly comparable to that on insectivorous  birds.  These  nocturnal
 animals have excellent mobility, but are most sensitive to disturbances
 to  breeding areas.  In  addition, many bat species are colonial, so the
 breeding areas are patchy in occurrence.   A disturbance of the smallest
 designated area size could have a measurable  impact on a breeding colony.
On the other hand, much larger feeding areas would have to be disturbed
to yield a measurable  impact.  Bat species occur in several types of land
areas.
                                   10-25

-------
10.3.2  Long-Term Effects
     The recovery phase of a mammal community is closely tied to the re-
covery of the vegetation.  The food value of the vegetation is an impor-
tant determinant in mammal recovery.  In addition, the amount of cover
provided by the vegetation is important to small mammals which are subject
to avian predators.  The vertical height of the vegetation is actually of
less general importance since relatively fewer mammals are canopy dwellers
than ground dwellers.  The main benefit provided by the vegetation height
is the amount of feed which exists at each level for herbivores and om-
nivores.  Another extremely important benefit for some prey species is
cover.   Naturally, the habitat beneath the vegetation should vary with
vegetative height and density.  Therefore an important indirect benefit
is provided.

     10.3.2.1  Large Carnivores-
     Large carnivores will respond to the recovery of their prey following
a disturbance.  The functional response of carnivores, an increase in the
exploitation of prey, can be practically instantaneous.  However, the
numerical response, an increase in the number of predators in an area,
is necessarily slower.   This response can occur both through immigration
or large carnivores to the recovering area   and by reproduction, but
the latter process is much slower and may be masked by immigration.
Since large carnivores  are not measurably impacted by a disturbance,
their recovery in an area will be a qualitative rather than a quantita-
tive change.   Qualitative changes can be measured,  but only on a very
restricted local scale,  so that the measurements are of questionable
significance in assessing impacts to the individuals of highly mobile
species.   The pertinent  point  is that the quality of the local area
changes when the animals leave.

     10.3.2.2  Medium-Sized Carnivores--
     The recovery of medium-sized carnivores should share the same
characteristics as the  recovery of large carnivores, but on a more local
scale.   Since medium-sized carnivores have less mobility, they will be
slower in responding to  prey in the interior of a large disturbed area.
                                     10-26

-------
 However, the numerical response will still be made mainly by  immigration
 Again, population changes of these animals will be limited by changes in
 the populations of their prey.  The medium-sized carnivores are also
 unlikely to be impacted measurably, so that recovery will only be a
 qualitative process.

      10.3.2.3  Small Carnivores--
      Disturbances to the larger land area sizes will have measurable
 impacts on the small carnivores,  and therefore "recovery" can be demon-
 strated quantitatively.   Since these carnivores are less mobile than the
 larger classes,  much of the numerical response must come from reproduc-
 tion.   Small carnivores  should follow fluctuations  in prey density much
 more closely than larger carnivores due to their more limited mobility.

      10.3.2.4  Large Herbivores--
      For  large  herbivores,  size effects are again of  critical  importance
 in evaluating recovery from  a  disturbance and  their recovery  will  be
 qualitative rather  than  a measurable  change.   Since the  large  herbivores
 are highly  mobile  (some are  migratory), the rate of movement back  into
 an area will  depend on the reestablishment of  edible vegetation.   The
 number of large herbivores that a disturbed area can support will  change
 gradually,  and some disturbed areas may actually support higher densities
 than originally existed.   This will occur where the plant food is
 actually denser and/or of better quality during recovery than in the
 climax community.

     10.3.2.5  Small Herbivores—
     The recovery phase of small herbivore populations can be extremely
 complex. •    Interspecies competition effects will  be important in many
 cases.     The recovery of small herbivore numbers should depend on both
 immigration and reproduction.  Small mammals are noted for their ability
 to respond reproductively,  but  dispersal abilities can also play a
key role in their population dynamics.11  The recovery of small herbi-
vores should be  very closely tied  to changes in the  vegetation, and
several distinct  successional communities  may occur.
                                     10-27

-------
      The immediate disturbance to an area may well create ideal conditions
 for invasion by small "pest" rodents.   These animals may aid plant suc-
 cession by cycling nutrients in the community and thus provide a beneficial
 function.     However, competition may occur among "natural" species attempt-
 ing to immigrate and could have a very detrimental effect.   Careful manage-
 ment of small mammals can provide means for facilitating community recovery
 Predators  can be introduced to dampen  outbreaks  in small herbivore popu-
 lations, if control  measures are needed.   Small  mammal populations can
 provide an index for monitoring community recovery of both  the flora and
 fauna.

      10.3.2.6  Large Omnivores—
      Because  of their great mobility and  large home  range sizes,  the re-
 covery of  large omnivores  (bears)  after a disturbance  to an area  is,  like
 that  of other large  animals,  a qualitative  process.  The recovery of bears
 is most likely to  consist  of them  simply  using the area  to  forage  and/or
 den in  again.   The problem of range reduction is  a serious  one  for North
 American bear species.  Therefore  real  recovery may  be very  difficult if
 the maximum area size  disturbance  causes  the loss of one or  both  of  a
 breeding pair which  live in  an  isolated patch of  habitat.  This problem
 will  rarely occur, however,  and can be  overcome by introduction of one or
 more  outside  individuals as  appropriate.

      10.3.2.7   Small Omnivores--
     The small  omnivorous mammals are for the most part  opportunistic
 species.  Both  the opossum and raccoon are familiar  species which have
managed to survive in areas of human disturbance due to their versatile
 feeding abilities.  The armadillo has actually extended its range of
occurrence in the Southwestern United States during the last hundred
      23
years.    Quantitative recovery of these species should occur closely
in pace with changes in both vegetation and prey species on the largest-
size area.   These species are reasonably mobile and can probably spread
as well by immigration as by reproduction.
                                     10-28

-------
      10.3.2.8  Insectivores—
      Insectivore recovery will  depend on the  recovery of insectivorous
 prey,  which in turn  will  probably depend largely on  vegetative  recovery.
 Thus the return of insectivores to a  disturbed area  is a second-order
 effect and will require a longer time lag than that  which will  occur
 for the similarly impacted small rodents.   The recovery of moles will be
 stimulated where macroinvertebrates are  added to the soil .o  assist
 vegetative recovery.   Few insectivores breed  more than twice  a  year, and
 many breed but once.   This will tend  to  retard recovery,  as reproduction
 is  an  essential part of colonization  due to the animal's  limited mobility.

      10.3.2.9  Flying  Mammals—
     The recovery of bats  after a  disturbance  to an  area  will vary with
 the nature of the disturbance.   Like  many birds,  bats  require the re-
 establishment of vertical  structure in the vegetation.  However, several
 bat species roost in dead  trees,  so that  available roosts may actually
 increase from a disturbance.  Bats  are commonly known  for their mobility.
 Several  species  migrate seasonally, or nightly,  to food sources.  For
 these  reasons,  bat recovery can  occur rapidly  after  recovery of insect
 prey in  feeding  habitats,  and densities may actually increase given the
 proper circumstances.  Changes  in bat populations will usually be due
 to  emigration  and immigration rather  than reproduction.

 10.3.3   Land  Types
     Table  10-3 gives examples of mammal species other than small herbi-
vores  from  each trophic/niche group in each land type which are measurably
 impacted if the maximum disturbance has been made to the specified land
areas,   Large mammal groups are extremely mobile, and are therefore im-
pacted only by area sizes greater than 100 km2.  Many of the  species occur
in more  than one  land type.  This phenomenon is more evident  in this table
than it  is for birds (Table 10-1) and  small herbivorous mammals (Table  10-2)
The greater body size and mobility of  many of the mammal species may be the
reason for this.  Again, the number of home ranges lost is used in estimat-
ing the types of species lost when a given size area is disturbed.
                                     10-29

-------
Table  10-3.   Examples  of mammals  other  than small  herbivores  impacted  by
                 maximum  disturbances  to  specified types  and  sizes of  land
                 areas.3
MAJOR GROUPS
1. LAND TYPE
DESERT
Large Carnivores
Medium Carnivores
Small Carnivores
Large Herbivores
Bats
PRAIRIE
Large Carnivores
Medium Carnivores
Small Carnivores
Large Herbivores
Small Omnivores
Insectivores
DECIDUOUS FOREST
Large Carnivores
Medium Carnivores
Small Carnivores
Large Herbivores
Large Omnivores
Small Omnivores
Insectivores
Bats
CONIFEROUS FOREST
Large Carnivores
Medium Carnivores
Small Carnivores
Large Herbivores
Large Omnivores
Small Omnivores
Insectivores
Bats
TUNDRA
Large Carnivores
Medium Carnivores
Small Carnivores
Large Herbivores
Large Omnivores
2. LAND USES
AGRICULTURE
Medium Carnivores
Small Carnivores
Large Herbivores
Small Omnivores
Insectivores
Bats
SUBURBAN
Small Omnivores
Bats
SIZE OF LAND AREA
0.1 km2





L
Yuma myotis






Vagrant shrew







Pacific mole
Little brown myotis







Shorttail shrew
Mexican freetail bat












Eastern mole
Evening bat'3

h
Pallid bat0
1 km2




















Armadi 1 lo





Least weasel








Least weasel













10 km2




Hoqnose skunk





Spotted skunk

Opossum




Long-tail weasel


Opossum





Short-tail weasel


Opossum





Short-tail weasel





Spotted skunk

Opossum



Opossum

• 100 km2


Mountain lion
Kit fox

Mule deer


Coyote
Red fox

Pronqhorn



Mountain lion
Lynx
Marten
Whitetail deer
Black bear
Raccoon



Grey wol f
Lynx
Marten
Mule deer
Grizzly bear
Raccoon



Grey wolf
Arctic fox

Barren ground caribou
Grizzly bear


Red fox

Mule deer
Raccoon



Raccoon

        Home range sizes taken from McNab,   Vaughn,   and Burt and Grossenheider,4 or estimated from home range sizes
        of similar species.  Nomenclature of the genera and species follows Burt and Grossenheider.5 where scientific
        names are also given.

        Bats, although migratory, appear in the smallest area category because of their colonial or semicolonial
        nesting characteristics.
                                              10-30

-------
      Impacts  can be defined more  precisely as  smaller mammals  are  ex-
 amined,  due to  their reduced mobility which smaller  size  usually causes.
 Therefore,  in the discussion of the mammal fauna,  small mammals are
 emphasized.   In addition,  some  land types,  notably agricultural and
 suburban,  support more  "pest" specimens  than others.   A special con-
 sideration regarding these land types  is the desirability of restoring
 these mammals.

      10.3.3.1  Deserts—
     The single most important  and prevalent group of mammals  in the
 desert habitat  is the small  herbivores.   Although  large,  medium, and
 small carnivores,  large herbivores, and  bats may occur in deserts, their
 diversity does  not  approach  that  of the  small herbivores.  When a desert
 area is  disturbed,  some measure of the impact on small mammals is often
 obtainable.   However, the  impact must be  inferred  from a  comparison of
 the disturbed area  with some  nearby area.  The return  of  herbivores will
 keep pace with  the  vegetation return, a  distinctly slow process.  Deserts
 are continuous,  so  that colonization by  both dispersal and reproduction
 can occur from  the  edges of the disturbed area.  Intermediate faunal com-
 munities are  not  expected  for succession  in most desert disturbances.

     10.3.3.2   Prairies—
     The prairie  habitat supports a diversity of large herbovores.   Car-
 nivores of all  sizes as well as small omnivores may occur here, but by far
 the most abundant and diverse group in prairies is the small herbivores.
 The importance  of small herbivores in this habitat is greater than in any
 other habitat except tundra, where their importance is roughly equal.12
A disturbance to prairie will certainly have a measurable impact on some
 small herbivores for all but the smallest area size.   Therefore some
measurement of  the recovery may be possible.  The recovery of the  small
herbivores  may  involve changes in the densities and distributions  of the
species present, as the vegetation changes.  The community recovery in
this habitat can be relatively rapid.   The continuous nature of prairies
should aid  in this process.
                                    10-31

-------
      10.3.3.3  Deciduous Forests--
      Deciduous forests follow a pattern of response similar to coni-
 ferous forests, but with some important differences.  These forests can
 be expected to be somewhat richer in their mammal fauna.  The seasonal
 change in canopy cover allows a richer understory of forage for both
 large and small herbivores.  Bats and insectivores occur here, but are
 seldom a major component of the fauna.   Medium size carnivores and
 small omnivores should occur more frequently in deciduous forests.  The
 response of the mammal fauna to a disturbance may again involve the
 movement of large mammals  out of the area,  and an increase in  small
 herbivores.  In comparison to forests,  the importance of small mammals
 will  be greater than in nonintensive agriculture during the recovery;
 but in  the mature forest,  small  mammals will  have a similar importance
 to that in nonintensive agriculture.

      10.3.3.4   Coniferous  Forests—
      In coniferous  forests  during  the recovery cycle following a  dis-
 turbance the number  of small  mammals will probably increase rather than
 decrease.   Large  mammals  (carnivores, herbivores,  and  omnivores)  will
 usually leave the area, at  least until  restoration is  well  under  way,
 and then return.  Thus in coniferous forest,  impacts are  very  difficult
 to measure  but  recovery is  easier  to assess.   This habitat  is  ideal for
 the use  of  small  mammals as indicators  of recovery.
      Bats and insectivores  are relatively unimportant  components  of the
 mammal  fauna, though they do  occur in coniferous  forest.   If insectivores
 are present they may feed on  insects which attack  the dominant trees,
 and thus  increase their importance.  The medium carnivores  and small
 omnivores are more likely to  occur at the forest border,  and thus will
 recolonize from other  habitats.  The continuity of coniferous forests
 should help most species recolonize  from undisturbed areas.

     10.3.3.5  Tundra Areas--
     In tundra,  small herbivores are key species to the ecosystem
economy.    Small carnivores depend upon these herbivores for prey.
                                    10-32

-------
Large herbivores, large omnivores, and large and medium carnivores may
also occur seasonally in tundra.  A disturbance to tundra will certainly
impact the small mammals measurably in most years.  The well known cyclic
fluctuations in population size of the dominant herbivores may cause an
inability to measure immediate impact in some years.  However, the
recovery phase measurements of following years will show the impact
belatedly.  Tundra is extremely sensitive to disturbance and a long
period of time is required for the vegetation to recover.  Whether
the recovery of the tundra habitat and fauna is possible in ecological
time is questionable, and predictions about recovery cannot be made.

     10.3.3.6  Coastal Inter-Tidal Marshlands--
     The marshlands are not expected to support extensive mammal faunas.

10.3.4  Land Uses
     10.3.4.1  Agricultural Areas--
     Herbivorous species are generally detrimental to man's interests
where they occur in agricultural areas.   The most common pests are, of
course, the small herbivores.   However,  as the intensity of agriculture
increases, the role of small herbivores decreases.    The other trophic/
niche groups are less frequent in occurrence in agriculture, but small
omnivores and medium-sized and small carnivores are considered pests
where animal husbandry is involved.  Large omnivores and carnivores
should occur only rarely.   Farm buildings are usually excellent places
for bat "nursery" breeding places, so that their occurrence in agri-
culture areas may be relatively high.   Insectivores are a relatively
unimportant, though occurring, component of the mammal fauna here.
     The recovery of mammals,  particularly small herbivores, on dis-
turbed agricultural land can be extremely rapid.  Most, if not all, of
the responding animals will be pest species.   Most recovery problems
will involve having too many animals,  at least in density,  rather than
too few.   Some beneficial  effects may result  from nutrient  cycling
through small herbivores.   However, these benefits are minimal because
of the control that man can exert on the nutrient balance.
                                    10-33

-------
      Here again, another alternative exists of allowing succession to
 something similar to the preagricultural land type.   This is best thought
 of by considering agriculture as an additional disturbance to the treat-
 ment under consideration.   The mammal recovery will  be similar to the
 recovery described for the specific preagriculture land type, except
 where precluded by fencing.   Fences which restrict large mammal access
 will not greatly retard recovery,  but may have subtle effects.   For
 example, if predators  are  excluded,  prey species may be affected.   A
 fence which excludes all ground-moving mammals will  eliminate the re-
 covery for all  mammals except bats,  and in some cases squirrels.

      10.3.4.2   Suburban Areas—
      The suburban habitat, much  like agriculture,  is typified by pest
 species.   No large mammals of any  type and no carnivores of any size
 occur regularly in this habitat.   Small omnivores, small herbivores,
 and bats make up the rather  depauperate mammal  fauna of the suburban
 habitat.   The small  omnivores have principal  habitat elsewhere,  and
 the easily measurable  impact  to  small  herbivores is  probably not  of
 concern.   Feral  domestic species,  especially  cats, may  be  impacted.
 Again this  is probably not of concern.   What  may be  very important,
 however,  is  the  increase in  rodent pests,  and the  diseases  they  carry,
 during the  recovery.   Introduction of  predators  such as  feral cats may
 be  the best  technique  for managing these  outbreaks.   The importance of
 mammals  in  suburban  faunas is  likely to be  quite low, perhaps on a
 par with intensive agriculture.
      If  the  suburban land use  is abandoned, succession to a  land type
 similar  to the presuburban one may occur.  However,  the  rate and quality
 of  this  recovery will  depend  on whether man-made structures remain or
 not.  As  in  agriculture, this  land use  itself may be  considered a dis-
 turbance.  Fencing out  large mammals will simply cause the  loss of these
 animals as participants  in the recovery.  However, fencing out all
 ground-moving mammals may also fence surviving pest species in.   These
pests may require control measures to avoid health hazards to humans
which might periodically occur in the fenced area.
                                     10-34

-------
10.4  CONCLUSIONS
     The impact of a disturbance to a species was carefully defined at
the beginning of the wildlife section, but many people may simply want
to know if the disturbance will have a significant impact on the wild-
life species.  This demands not only  a definition of impact, but also an
evaluation of impact significance.   If significance is defined as whether
or not a disturbance puts any species in danger of becoming extinct, for
the vast majority of wildlife species and the designated land areas the
answer is "no".
     The real concern must be for those species so small in numbers that
extinction is a danger to them.  These species fill the rare, threatened,
                             5 22
and endangered species lists. '    Many of these species are nearly
extinct due to man's encroachment upon their habitats.  Other species
may be on the way to extinction with or without man's help.  These two
classes of endangered species may be difficult to tell apart.  Whether
or not species which are becoming extinct should be preserved is a social
decision.  This decision, conscious or unconscious, will govern their
fate.
     Many sensitive species may not be recoverable after a disturbance
to their range.  Other sensitive species may require expenditure of a
great effort in order to assure their revival.  This must be decided
upon in each case where an endangered species is affected, presumably
on the basis of the social and political values which predominate at
the time of the disturbance.   Any effort directed to this end may also
have a fringe benefit of aiding in the recovery of the total floral and/or
faunal community.
     Table 10-4 rates the relative impacts on wildlife, immediately after
cleanup, of the maximum disturbance possible in the individual treatment
categories in each land type for the stated area sizes.  Impacts on
mammal and bird wildlife are scaled from 0 (no measurable impact) to 5
(elimination of most original species) for each combination of the three
axes of  land type, treatment, and area size with the exception of fencing,
which divides impacts on mammals and birds.           •
                                 10-35

-------
Table  10-4.   Relative  impacts  of  land  cleanup  treatments  on  wildlife
                following maximum possible land disturbance.
MAJOR GROUPS
LAND TYPE
Desert




Prairie


Deciduous
Forest



Coniferous
Forest



Mountains
(Alpine)



Tundra



Coastal
Inter-Tidal

LAND USES
Agriculture



Suburban



TREATMENT

Stabilization, asphalt
Vegetation removal
Soil removal, stabilization
Fencing (mammals only)
Fencing (birds only)
Vegetation removal
Soil removal, stabilization
Fencing (mammals only)
Fencing (birds only)
Stabilization, asphalt
Vegetation removal
Soil removal, stabilization
Fencing (mammals only)
Fencing (birds only)
Stabilization, asphalt
Vegetation removal
Soil removal, stabilization
Fencing (mammals only)
Fencing (birds only)
Vegetation removal
Soil removal, stabilization
Fencing (mammals only)
Fencing (birds)
Flooding
Vegetation removal
Soil removal, stabilization
Fencing (mammals only)
Fencing (birds' only)
Vegetation removal
Soil removal, stabilization
Fencing (birds only)

Stabilization, asphalt
Vegetation removal
Soil removal, stabilization
Sludge application
Fencing (mammals only)
Fencing (birds only), soil additive
Stabilization, asphalt
Vegetation removal
Soil removal, stabilization
Fencing (mammals only)
Fencing (birds only)
TREATMENT3
NUMBER

4-0
1
2
3-2
3-2
1
2
3-2
3-3
4-0
1
2
3-2
3-3
4-0
1
2
3-2
3-3
4-0
1
2
3-2
3-3
1
2
3-2
3-3
1
2
3-3

4-0
1
2
5-0
3-2
3-3
4-0
1
2
3-2
3-3
RELATIVE IMPACTS FOR DIF- 5
FERENT AREA SIZES (in km2)
0.01 0.10 1.0 10 100

3 45 c c
J •» 0 b b
1 23 45
2 45 t; «;
*- •* 3 D D
019 in
it 1 U
00001
1 ? 1 A A
1 £- J 44
2 14 K c
c J «t 3D
01221
00011
24555
1 19 t -j
1 1 t J J
? 1 A C c
t j 4 b 5
010 in
It 10
00011
24555
1 19 9 •}
1 It c 6
? 1 A c c
£ J 1 b 5
0 1 9 in
u ' t 1 U
00011
34555
3 45 ti c
J *» 3 3 b
1 ? "\ 71
' CO £_ \
On n -i -1
U U I)
00 0 00
•J AC c c
J * 3 55
T 4i; EC
° * b b 5
1 ? 1 91
1 t j <: i
00011
34555
1 4 c. c c
0 H b b 5
01111

2 45 c c
' ** o b b
9 91 /I c
t c J 45
1 19 •) •}
1 i c J 3
001 22
01211
00011
1 9 1 >l c
1 t J 45
01 O -3 *
I c 34
Oil o o
II 22
001 00
000 00
       Treatment number 1 implies treatments 1-2 through 1-8,  in general.
       Treatment number 2 implies treatments 2-1 through 2-8,  in general.

       Relative impact, based on reduction in biomass

               0 - No measurable impact on species, < 5 percent reduction
               1 - 5 to 35  percent reduction
               2 - 35 to 60 percent reduction
               3 - 60 to 80 percent reduction
               4 - 80 to 90 percent reduction
               5 - Original, species eliminated, >  90 percent reduction
                                          10-36

-------
     The  assignment of the values of the table is a subjective process
which necessarily assumes that all wildlife responds homogeneously to the
disturbance.  Although such an assumption is unrealistic, the table still
provides  insights toward the truth, i.e., reasonable estimates of how a
wildlife  community responds to a disturbance.  The table can also be viewed
as a ranking of treatments within each land type and area size.
     Table 10-5 estimates the relative time for recovery of wildlife from
the maximum disturbance possible and the individual treatment categories
of each land type for the stated area sizes.  The relative time to recovery
is scaled from 0 (about one year) to 5 (2 or 3 centuries).  The one-year
period is required as a baseline to measure whether a community has changed
to a greater degree than would be expected under "natural" conditions.
The 5 rating puts recovery sufficiently far into the future that it is not
reasonable to consider that recovery can take place, since evolution may
change the community before it reaches equilibrium.
     One major assumption has been made in the derivation of relative im-
pacts:  that is, the colonization rate must remain constant for all species
in the successional process.  While this is not true in most individual
cases, it may be true on the average.   That is, the species in the coloni-
zation process may have varying movement rates.  But if the rates do not
vary widely, then violating the assumption may not be very serious.  The
mathematical reasoning implies a linear relationship between area size and
recovery time.   The constant colonization rate assumption is actually
another form of this assumption.
     Wildlife is dependent on vegetative recovery;  in fact, it is not
reasonable to try to uncouple these two types of recovery.   Therefore,
Table 10-5 includes biases which are a direct result of considering the
total environment in which the wildlife lives.   These biases are necessary
in order to improve the accuracy of impact assessment,  and the computa-
tion of recovery times.
                                     10-37

-------
Table 10-5.
       Relative  time  for  recovery of wildlife  following
       maximum disturbance  to  land and cleanup treatment.
MAJOR GROUPS
LAND TYPE
Desert
Prairie
Deciduous
Forest
Coniferous
Forest
Mountains
(Alpine)
Tundra
Coastal
Inter-Tidal
LAND USES
Agriculture
Suburban
TREATMENT
Stabilization, asphalt
Vegetation removal
Soil removal, stabilization
Fencing (mammals only)
Fencing (birds only)
Vegetation removal
Soil removal, stabilization
Fencing (mammals only)
Fencing (birds only)
Stabilization, asphalt
Vegetation removal
Soil removal, stabilization
Fencing (mammals only)
Fencing (birds only)
Stabilization, asphalt
Vegetation removal
Soil removal, stabilization
Fencing (mammals only)
Fencing (birds only)
Vegetation removal
Soil removal, stabilization
Fencing (mammals only)
Fencing (birds only)
Flooding
Vegetation removal
Soil removal, stabilization
Fencing (mammals only)
Fencing (birds only)
Vegetation removal
Soil removal, stabilization
Fencing (birds only)
Stabilization, asphalt
Vegetation removal
Soil removal, stabilization
Sludge application
Fencing (mammals only)
Fencing (birds only), soil additive
Stabilization, asphalt
Vegetation removal
Soil removal, stabilization
Fencing (mammals only)
Fencing (birds only)
TREATMENT3
NUMBER
4-0
1
2
3-2
3-3
1
2
3-2
3-3
4-0
1
2
3-2
3-3
4-0
1
2
3-2
3-3
4-0
1
2
3-2
3-3
1
2
3-2
3-3
1
2
3-3
4-0
1
2
5-0
3-2
3-3
4-0
1
2
3-2
3-3
RELATIVE TIME FOR RECOVERY FORb
DIFFERENT AREA SIZES (in km2)
0-01 0.10 1.0 10 100
5 55 55
22344
34555
21100
ooooo
11222
12344
21100
ooooo
45555
22333
23455
21100
ooooo
45555
22333
23455
21100
ooooo
34555
33455
21100
ooooo
ooooo
34555
33455
21100
ooooo
12344
11222
ooooo
45555
1 1 1 1 1
11111
11111
11000
ooooo
34555
12233
11122
11000
ooooo
    Treatment number 1 implies treatments 1-2 through 1-
    Treatment number 2 implies treatments 2-1 through 2-

         1 year
                                       !, in general
                                       !, in general
    '     i jr cu (
    1 - 1  to 3 years
    2 - 3  to 5 years
    3 - 5  to 50 years
    A _ Kn <-^ i nn ,,„,..
   4 -
   5 -
  *
 to 100 years
100 years
                                    10-38

-------
  10.5   WILDLIFE  REFERENCES


  1.  Adams,  D.  L. ,  and  G. W.  Barrett,  1976.   "Stress  Effects  on  Bird  Species
          Diversity  Within Mature  Forest Ecosystems."  American Midland
          Naturalist,  96(1), p.  179.                   ~"~

  2.  Belrose,  F.  C.,  1976.  Ducks, Geese,  and Swans of North  America.
          Harrisburg,  Penn.: Stackpole  Books.                       ~

  3.  Bunnell,  S.  D.,  and D. R.  Johnson, 1974.  "Physical Factors Affecting
          Pika  Density and Dispersal."  Journal of Mammology,  55(4),
          pp. 866-69.

  4.  Burt, W.  H., and R. P. Grossenheider, 1964.  A Field Guide to Mammals.
          Boston,  Mass.: Houghton  Mifflin Co., 284 pp.

  5.  California Department of Fish and Game,  1976.  At the Crossroads.

  6.  Cook, Sherburne  F., Jr., 1959.  "The Effects of  Fire on  a Population
          of Small Rodents."  Ecology 40(1), pp. 102-108.

  7.  Delacour, J.,  1954.  The Waterfowl of the World.  London: Country
          Life  Ltd.        ~~~   '

  8.  Emlen, J. M.,  1973.  Ecology:  An Evolutionary Approach.  Reading,
         Mass.: Addison-Wesley  Pub. Co.

  9.  Emlen, J. T.,  1974.  "An Urban Bird Community in Tucson, Arizona:
         Derivation,  Structure,  Regulation."  Condor  72(2), pp.  184-97.

10.  Frauzreb, K. E., 1976.   "Comparison of Variable Strip Transect and
         Spot Map Methods for Censusing Avian Populations in a Mixed
         Coniferous Forest."  Condor 78(2), pp.  260-62.

11.  Golley,  F. B.,  K. Petrusewicz, and L.  Ruszkowki,  editors, 1975.
         Small Mammals:   Their Productivity and Population Dynamics.
         Cambridge,  Mass.:  Cambridge University Press.

12.  Joule, J., and G. N. Cameron, 1975.   "Species Removal Studies I.
         Dispersal Strategies of Sympatric Sigmodon hispidus and
         Reithrodontomus fulvescens Populations."   Journal of Mammology
         56(2), pp.  378-96.

13.  Krebs, C.  J., 1964.  "The Lemming  Cycle at  Baker  Lake,  Northwest
         Territories, during 1959-62."   Arctic Inst.  N.  Amer.  Tech.  Paper
         No.  15,  104 pp.                   ~	—

14.  Krebs, C.  J., 1966.  "Demographic  Changes in  Fluctuating  Populations
         of Microtus californicus."  Ecological Monographs,  36(3),  pp.
         ^ O i/™" / O •
                                   10-39

-------
 15.   Krebs,  C.  J.,  1972.   Ecology.  New York:  Harper  £  Row,  Inc.

 16.   Lawrence,  George C.,  1966.   "Ecology  of Vertebrate  Animals  in Relation
          to  Chaparral Fire in the Sierra Nevada Foothills."   Ecology
          47(2), pp.  278-91.                                   	&JL'

 17.   McNab,  B.  K.,  1963.   "Bioenergetics and the Determination of Home
          Range  Sizes."  American  Naturalist, 97(3), pp.  133-40.

 18.   Peterson,  R. T., 1964.  A Field Guide to Western  Birds.  Boston,
          Mass.: Houghton Mifflin  Co.

 19.   Pianka, E. R.,  1974.  Evolutionary Ecology.  New  York: Harper-Row,
          Inc.

 20.   Schoener, T. W., 1968.  "Sizes of Feeding Territories Among Birds."
          Ecology, 49, pp.  123-41.

 21.   Twigg, G. I., 1975.  "Marking Mammals."  Mammal Review,  5(3), pp.
          191-16.                              	

 22.   U.S. Department of the Interior,  Fish and Wildlife Service, 1976.
         Endangered and Threatened Wildlife and Plants.  Federal Register,
         Wed. 27 Oct. 1976, pt.  IV.

 23.  Vaughn, T. A.,  1972.  Mammology.   Philadelphia, Penn.: W,B. Saunders
         Co.

 24.  Welty, J. C., 1962.  The Life of Birds.   Philadelphia, Penn.:  W.B.
         Saunders Co.

25.  Whittaker, R. H.,  1970.  Communities  and  Ecosystems.  New York:
         MacMillan.
                                  10-40

-------
              PART IV
            APPENDICES
APPENDIX A,  STABILIZERS
APPENDIX B,  IMPACT ASSESSMENT
APPENDIX C,  LAND TYPES
APPENDIX D,  CLEANUP TREATMENTS
APPENDIX E,  GLOSSARY

-------

-------
                                  APPENDIX A
                                 STABILIZERS
 A.I  INTRODUCTION
      This appendix contains a brief description of the major categories  of
 soil  stabilizers,  the subgroupings  of stabilizers,  and the  values  assigned to
 the stabilizer  subgroupings,  land types,  and  land-use  classes given  in Table
 A-l.  The use of Table A-l  is also  explained.

 A.2  TYPES OF STABILIZERS
      For  purposes  of  this report, "soil  stabilizers" are defined and  categorized
 to determine their practical  application  to soil  stabilization with the  land
 types defined herein.   Stabilizers  are generally  classified  into three major
 categories according  to  Ingles and  Metcalf:1  chemical, mechanical, and  physi-
 cal.  A fourth  category, chemical with mechanical characteristics, is also
 considered here.
      Individual stabilizers are shown at  the top  of Table A-l.  The major
 category  of each is shown by  a symbol underneath  in one of the first two
 horizontal rows; these rows additionally  denote whether the  stabilizer is
 organic or inorganic.  The  stabilizers listed in Table A-l are listed in
 Table A-2 with  information on manufacturers provided for those that have
 patented name.

 A. 2.1   Chemical
     These stabilizers are liquid or solid additives to soil that effectively
alter the physical properties of the soil being treated.
A.2.2  Mechanical
     These stabilizers mechanically provide sufficient stability to bare soils
to retard and/or prevent soil erosion caused primarily by wind and water,
                                      A-l

-------
TYPC3  OF  S7AIILIZCH3

  O  Chamital

  •  Uoehtnical

      Pttftical

  ©  Chomitol with motnonieol
        cnorocttritlitt
                                                 I   toft
                                                I
                                                3
                                                4

•HI
»
d'ling procoutiont roaurtd
trtt anrironmontol tttoctt
1/r notardou*
vc i 111 riMi iraua ur &l AHILIZ
/ manual - normal
I manual - difficult
3 oauipmont - normal
4 oowpmtnt - tptcialutd
cK uHOUrlNGo
/ pormonont
2 inttrmtdioto
3 tnorl -lorm, 1 yoor or *M
yooatotion roco*,.
1 immadioto, natural
2 roauirtt itodina, normal ntpentt
3 campion no* totdotd rootttrod
4 no rocoirory

Land ttoti and clou
a profonod
ft occoprooit at a olrornoti*o
e not oeeoatotto, loot roaort moomrt
t not applicant
                                    K-Z

-------
 without modifying the physical  properties of that soil.   The process con-
 sists primarily of compacting soils to a density that rejects water.  It also
 includes hard surface materials,  manufactured soil  retention materials,  and
 chemical aids to compaction.
      Materials used for mechanical  stabilization of soils include,  but  are not
 limited to,  hard surface paving such as concrete and bituminous  surfaces.   Other
 manufactured materials are  meshes,  polyvinyl  films,  and  erosion  control  nettings.
 Materials such,as burlap bags with  concrete,  sandbags, rip-rap,  and cribbing
 are  also means of mechanical  stabilization.
      Other forms of this type of  stabilizer  include  the  use  of grouts,  sand,
 and  other filler compounds  that can be used  to  physically fill voids in  the
 soil  by absorbing water and swelling to seal  off the surface area from  water
 infiltration.
      Equipment used for mechanical  soil  stabilization is  divided into two  types:
 (1)  compaction equipment  such as  pneumatic-tired rollers,  tandem rollers,  sheeps-
 foot  rollers,  vibratory rollers,  vibratory packers;  and  (2)  application  equip-
 ment  such as  concrete  pavers, asphalt  pavers, rock spreaders,  distributors, and
 sprayers.
      A  third  subcategory  of equipment  also should be  considered for  mechanical
 stabilization  of  contaminated areas.   This includes  equipment  such  as bull-
 dozers,  turn-a-pulls,  and scrapers  which are used to  construct diversion channels,
 ditches,  berms, and retention basins around the  affected  area.  (These structures
 would be  temporary  to  collect,  divert, or retard  surface  drainage from adjacent
 areas or  to prevent water from  leaving affected  areas and causing downstream
 contamination.)

 A.2.3   Physical
     These stabilizers  stabilize soils by modifying their physical  properties
with heat, electricity, and cold.

 A.2.4   Chemical with Mechanical  Characteristics
     This term applies to stabilizers that are chemically produced  or chemical
compounds that bind surface soils for short periods of time.   Primarily, they
                                     A-3

-------
 are  compounds used  by  the  agricultural  industry  to  promote  germination  and
 then break  down  once germination  has begun.  They are not true  "mechanical--
 stabilizers because of their  lack of relative permanency, and because there
 are  brief periods between  application and  stabilization during  which chemical
 reaction with soil occurs.

 A.3   STABILIZER  GROUPINGS
     The four categories of stabilizers can be further categorized into group-
 ings.  These groupings define the physical characteristics and  treatment types
 to promote  "prescription-type" treatment of affected areas within the land
 types defined in the report.  These groupings are shown at the  left of Table
 A-l, separated by broken lines.  To determine the rating of a grouping cate-
 gory for a particular stabilizer,  use the number immediately above the broken
 line for the grouping category.  The key to these number ratings is at the
bottom of Table A-l.
     The following definitions are provided for the groupings of soil sta-
bilizers:

     Organic and Inorganic  —  the  category given  to  stabilizers  that  generally
     define  the physical  composition of  the specific material.
     Hazard  Levels — primarily the  degree of safety with  which  a stabilizer
     can be  handled  by  personnel during  the application  periods.  It  also
     includes  the "environmental safety" of the material after application.
     Application  —  the method or  procedure of applying  the  prescribed
     stabilizer  to a defined area  as it  relates to available normal applica-
     tion  equipment  and methods.
     Durability  — the  ability of  a  chemical  or mechanical stabilizer to
     maintain  a  specified treatment  for  a  prescribed time period.  This
     includes  the ability of a specified stabilizer  to withstand the effects
     of  the  seasonal environmental changes.
     Vegetation Recovery Period — the ability of a  chemical  or  mechanical
     stabilizer to disintegrate in the normal revegetation process where
     localized natural  seed  sources  are  present (does not include conventional
     reseeding methods).
                                    A-4

-------
 A.4  RATINGS OF STABILIZER GROUPINGS

      The ratings of the stabilizer groupings  noted on Table A-l  are  defined
 as follows:

      1 >   Safe  - a  m^erial  or  technique  that is  generally  harmless  to  the
      applicator,  environment, and  equipment.
      2,   Handling Precautions Required  - requires  compliance with manufac-
      turers  instructions for physical applications  during handling,  mixing,
      applying,  and  temperature  extremes.
      3,   Adverse  Environmental  Effects  - materials where there  is no oppor-
      tunity  for  the area to recover from the  use of a stabilizer without
     major rehabilitation operations.   It does not  include consideration as
     a toxic substance to man.
     iz	Highly Hazardous — materials or techniques where the "cure" is
     more dangerous than the "disease."  However,  these can be situations
     where highly specialized personnel can be called upon to perform this
     type of stabilization.

A.4.1  Application

     L—Manual (normal)  — hand-operated  equipment that can be  used  effec-
     tively and is frequently available  to the home gardner.
     2,   Manual (difficult)  - stabilizers that must be  applied  to  a  small
     area, are  caustic  to operators and  equipment,  have  difficulty  staying
     in  solution,  require difficult or  specialized  mixing  techniques, or
     require  special carrier  substances  for proper  application.
     5,   Equipment (normal) - can  be applied  to large affected areas using
     standard agricultural  or construction equipment;  restricted  to areas
     where there  is  space for the equipment to be used effectively.
     4,   Equipment (specialized)  - requires specially manufactured equipment
     for  proper application or standard  equipment greatly modified to accom-
     modate the stabilizer to be used.
                                    A-5

-------
 A.4.2  Durability

      1,   Permanent - will remain in place without radical physical change as
      a result of normal climatic conditions.   This includes the use of chemical
      stabilizers that permanently change surface soil  characteristics.   Any
      alteration  of an area where this stabilizer is used would require  major
      construction.

      2,   Intermediate -  broad  terminology meaning that  natural  recovery will
      eventually  take  place.  This  also  applies  to  stabilizers  that  persist
      for  at  least  5 years  when  considered  for  surface  applications.  These
      same  stabilizers  can  persist  for periods of more  than  5 years  when  in-
      corporated  into  a  compacted soil.
      5,  Short Term (1  year or  less! -  stabilizers that effectively control
      soil movement resulting from normal rainfall.  It also includes stabilizers
     generally accepted in the agricultural industry as erosion retardents
     during germination periods of most crops.

A.4.3  Vegetation Recovery
     L>	Immediate (natural! — stabilizers which have little resistance to
     climatic conditions,  and which permit natural invader plants to germinate
     and grow on  disturbed areas.
     2,  Requires Seeding  (normal response! - effectively hold surface  soils
     intact for a period of at  least  6 months,  providing  the applied membrane
     is not disturbed.   Revegetation  requires  normal seedbed preparation and
     planting of  selected  varieties.   Residues  of these stabilizers,  after
     they  are incorporated,  are  biodegradable  and  have  no  effect  upon the new
     seedlings.
     3,  Complete New  Seedbed Required  -  stabilizers that were used must be
     completely removed  or  buried at  sufficient  depth with new  "cold soil" so
     that an  affected  area  can provide suitable  growing conditions for vegeta-
     tive reestablishment.
     4,  No Recovery -  change the physical makeup of the surface soil suf-
     ficiently so that the plant zone cannot subsequently support any type of
    vegetation, and the stabilizer is not effectively altered by normal  freeze
    and thaw conditions.
                                     A-6

-------
 A.5  LAND TYPES AND LAND USE CLASSES
      The seven land types and two land use classes discussed in this report
 are shown at the right of Table A-l, separated by solid lines.   To determine
 the rating of each stabilizer on each land type or land-use class, use the
 letter above the solid line for the particular land type or class.  The key
 to these letters (ratings)  is located at the bottom of Table A-l.   Each of
 these ratings is defined as follows.
      a,  Preferred —  materials that are short-term effective,  generally
      environmentally safe naturally and sociologically,  and subject to the
      general moisture  and climatic conditions of a land  type.
      b,  Acceptable as an Alternate — basically the same characteristics as
      those classified  "preferred," however,  because of the effectiveness of
      the material  or application procedures  they should  not be  considered
      exclusively for stabilization of affected areas.
      c,  Not Acceptable (last resort measure)  — materials that  should not be
      used because  of the resultant damage, primarily to  the natural  environ-
      ment.   When there are  situations where  the  natural  environment  cannot be
      protected  from irreversible effects of  "accidental  spillage," these
      materials  may be  used.  This  assignment  is  also made to materials when
      applied to vari'ous land types because of  the variability within the land
      types themselves.   For example,  "prairies"  can  be divided into many
      types of "prairie" and a particular stabilizer  can  be effective for
      one  of  these  "prairie" types  without serious  environmental  damage.   In
      other "prairies,"  however,  this  same stabilizer would produce  irreversible
      effects and must not be used.   Based upon this  type  of comparison with
      each of the land types, the "not  acceptable"  assignment is  made where
      there is an irretrievable commitment of the natural  resources present.
      d,   Not Applicable  — materials  that cannot be  applied  because of soil
      conditions, temperature extremes, or moisture conditions.

A.6  APPENDIX A REFERENCE
1.   Ingles,  O.G.,  and  J.B.  Metcalf.  Soil Stabilization Principles and Practice
     J.  Wiley §  Sons, New York-Toronto,  1973.   374 pp.
                                      A-7

-------
                        Table A-2.   Manufacturer and  application
c»
               STABILIZER
            4-ter Butyl  Pyrocatechol

            Aerospray  52 Binder
            (resin/water emulsion)

            Agri-mulch
            (63% asphalt)

            Anionic  Asphalt  Emulsion


            Aquatain
            (sodium  polypectate/glycerin
            ammonia)

            Asphalt Emulsions
           Asphalt (1 liter)-Plastic
           Asphalt (1 literj-Chipcoat


           Bentonite

           Calcium Sulfate/Calcium
           Chloride

           Carboxymethyl  Cellulose

           Coherex
           (resin/water  emulsion)

           CMC-7-L
           (sodium carboxymethyl cellulose,
           low viscosity)
                                                   MANUFACTURER
                        information for  soil  stabilizers.
                                 	•	

                                  APPLICATION  RATE
 American Cynamid  Co.


 Douglas Oil  Co.


 Phillips Petroleum  Co.


 Larutan Corp.
American Colloid, Inc

Dow Chemical Co.


Hercules, Inc.

Golden Bear Oil  Co.


Hercules, Inc.
 1  liter  per  sq. meter
 surface

 Surface  -  2.25 liters
 per  sq.  meter

 Mix  with soil
                              2.0 liters per sq.  meter,
                              surface applied

                              14 kg per sq.  meter applied
                              phasing & compaction
                              required
2% by weight incorporated
into upper 0.15 meter

Incorporate into soil  surface

Incorporate into soil  surface


Incorporate into soil  surface
                                                                                                               REFERENCE
                                                             Ingles & Metcalf
                                                                                                              (continued)

-------
                                                     Table  A-2.   (continued)
               STABILIZER
                                        MANUFACTURER
                                                                                 APPLICATION RATE
                                                                                                     REFERENCE
I
UD
CMC-7-M
(sodium carboxymethyl cellulose,
medium viscosity)

CMC-7-H
(sodium carboxymethyl cellulose,
high viscosity)

CurasolR AE
(polyvinyl acetate copolymer
emulsion)

CurasolR AH
(high polymer synthetic
resin dispersion)

DCA-70
(water dispersible polyvinyl
acetate emulsion)

Dialkyl Quaternary Ammonium
Chloride

E802 Corn Extract,
Maxofern Brand (corn
steep water;  N, P, Ca,  Mg)

Electro-osmosis and
Electrohardening

Elvanol 50-42
(polyvinyl  alcohol)

"Experimental"
(styrene-butadiene emulsion
in mineral  oil)
                                                Hercules, Inc.
                                                Hercules, Inc.
                                                American Hoechst Co.
                                                American Hoechst Co.
Union Carbide Corp.
                                                Corn Products Sales,  Inc
                                                E.  I.  Dupont  de  Nemours
                                                & Co.,  Inc.

                                                Ashland  Chemical  Co.
                              Incorporate into soil  surface
                              Incorporate into soil  surface
                              Incorporate into  soil  surface
                              Incorporate  into  soil  surface
Incorporate into soil  surface
                              3%  by  weight  incorporated
                              upper  0.15 meter

                              Incorporate into  soil  surface
                                                                                                             Ingles & Metcalf
                             Surface  1.5 liter/sq. meter
                                                                                                                (continued)

-------
                                          Table A-2.   (continued)
    STABILIZER


 Fatty Amine Acetate
 Foams
 (e.g., polystyrene)

 Gantrez An-119
 (polymeric anhydride)

 Gantrez ES-3351
 (monoester resin)

 Geon  652
 (vinylchloride-vinylidene
 chloride latex)

 Goodrite 2570  XI
 (styrene-butadiene latex)

 Grouting

 Gypsum  Hemihydrate

 Hydrofluoric Acid


 Krilium

 Lignins


Meshes
    MANUFACTURER
                                  APPLICATION RATE
                                                                 REFERENCE
                               100% surface  applied,
                               approx.  1  liter  per  sq.
                               meter

                               100% surface  applied


 General  Analine  &  Film  Corp.   Incorporate into  soil surface


 General  Analine  &  Film  Corp.   Incorporate into  soil surface


 B.  F.  Goodrich Chemical Co.



 B.  F.  Goodrich Chemical Co.
National Gypsum Co.
Monsanto Chemical  Co.
                              3% solution incorporated
                              into upper 0.15 meter
                              6% by weight  incorporated
                              into  upper  0.15 meter

                              100%  cover  of disturbed
                              areas
                                                             Ingles & Metcalf
                                                                                                      (continued)

-------
                                        Table A-2.    (continued)
   STABILIZER
   MANUFACTURER
APPLICATION RATE
REFERENCE
Metallic Soaps

NC 1556L
(modified pojyacrylamide)

Oil/Latex Polymer Emulsion

Ortho Soil Mulch
(asphalt emulsion)

Paraffin Wax
PetrosetR SB
(rubber emulsion)

Phosphoric Acid and
Phosphates

Polyacrylamides (PAA)

Polyco 2460
(styrene-butadiene
copolymer)

Polyco 2605
(vinylchloride-vinylidene
chloride copolymer)

Polyethylene, Vinyl,  Butyl
Rubber, and Other Impervious
Membranes

Polyvinyl Alcohol
                              100% disturbed  areas,  spray
Dow Chemical Co.


B. F. Goodrich Chemical  Co.

Chevron Chemical Co.
Phillips Petroleum Co.
Dow Chemical Co.

Bordon Chemical  Co.



Bordon Chemical  Co.
                              1/8 inch film over entire
                              area
                              7% aqueous  solution
                              Total  cover with  1%
                              overlap  of joints, pinned
Swift & Co.
                                                                                                      (continued)

-------
                                            Table A-2.   (continued)
    STABILIZER


 Quaternary Ammonium Salts
 (e.g., Cetyl trimethyl ammonium
 bromide)

 Resin Adhesive Z-3876
 (mainly polyvinyl  alcohol)

 Road  Oil
 Rubberized  Asphalt
 SBR Latex  S-2105
 (styrene-butadiene  latex)

 SEPARAN NP-10
 (polyacrylamlde, high
 molecular weight)

 Soda Ash
Sodium Chloride
Sodium Hydroxide
(caustic soda;  a compac-
tion aid)

Sodium Methyl  Silanolate
    MANUFACTURER
      -  ••  —  . —

 Dow Chemical  Co.




 Swift & Co.
    APPLICATION  RATE
                                   REFERENCE
Shell Chemical  Co,


Dow Chemical  Co.
2.25 liters per sq. meter,
surface applied

1.0 liter per sq.  meter,
surface applied

Surface applied
                              7%  by weight,  incorporated
                              into upper 0.15 meter

                              4%  by weight,  incorporated
                              into upper 0.15 meter

                              3% minimum by weight to
                              surface; 2% by weight when
                              incorporated 6 inches deep

                             3% by weight incorporated
                             into upper 0.15 meter;
                             spray applied
                                                                                                       (continued)

-------
                                           Table A-2.   (continued)
    STABILIZER
   MANUFACTURER
                                                                      APPLICATION RATE
                                                                REFERENCE
I

u>
 Sodium  Polyphosphate


 Sodium  Rosinate


 SOILGARD
 (elastomeric polymer)

 Soil Seal
 (formulation of copolymers)

 Sulphonates and Lignosulphonates
 (compaction aids)

 TACSS or Misui Stopper
 (class of polyurethane monomers)

 Technical Protein Colloid 1-V

 Technical Protein Colloid 5-V

 Technical Protein Colloid 2236

 Technical Protein Colloid 2260

 Thermal  Stabilization

 Volclay

WICALOID Latex 7035(AD)
 (carboxylated styrene-butadiene
 latex)
Alco Chemical Corp.
Soil Seal Corp.
Swift & Co.

Swift & Co.

Swift & Co.

Swift & Co.



American Colloid Inc.

Wica Chemicals
                                                                              15%  by weight  incorporated
                                                                              into upper 0.15 meter

                                                                              5% by weight incorporated
                                                                              into upper 0.15 meter

                                                                              Mix  with soil
                                                                             Mix with soil
                              6% by weight into upper
                              0.15 meter

                              Surface applied, 100%
                              coverage
                                                                                                            Ingles & Metcalf

-------

-------
                                   APPENDIX B
                               IMPACT ASSESSMENT
 B.I  INTRODUCTION
      This appendix presents the steps to be followed to obtain information from
 Sections 1 through 3 of each chapter to derive area recovery index numbers.
 These numbers are then used to identify preferred and unacceptable cleanup treat
 ments.   The ranking of the impacts of these cleanup treatments is also discussed

 B.2  DERIVATION OF RECOVERY INDEX FROM SUCCESSIONAL SERE
      The following steps describe the methods  to  be used to  obtain the informa-
 tion  shown on Tables B-l and B-2  under column  3,  titled "Duration of Succession
 Stage:"
      1.   List the stages in the sere  for the land type,  starting  from bare
          ground and continuing  through climax  (herbs,  shrubs,  softwoods, and
          hardwoods)  in  a loosely  tabular format.   Indicate the major stages
          of co-dominance and assume secondary  succession.
      2.   List  the physiognomy stages  appropriate  to  the  land type (canopy
          closure,  subclimax,  climax)  opposite  their  successional  equivalents,
          where  both occur.
      3.   List  in  a  third  column the range in years that  successional  stages and
          physiognomy stages  last  in the  sere.  The years do not need  to be
          continuous between  two ranges,  e.g.,  0-3 followed by  9-13 years.

      Inflection points on the curves of  Figure B-l prepared for the  species
diversity, cover, and primary productivity changes over the years from to to
100 or 200 years for 0-1 Natural Rehabilitation (Sections)  are examples.  The
Recovery Scenario of Section 5 applies a similar assessment routine to derive
descriptions of recovery at specific points in time:  1 year, 5 years, 50 years,
and 100 years after cleanup.  For example;
     1.   Select a median year for each time range  of column 3 (Duration of
         Succession Stage).   Where no median is available in  the literature
                                      B-l

-------
  Table  B-l.   Hypothetical  forest sere.
Succession
Mixed herbs
(grasses & forbs)
Grasses
Shrubs
Softwoods
Co-dominants
(Softwoods &
Hardwoods)
Hardwoods
Physiognomy
Stage
Pioneer
invasion
Cover
closure
Canopy
closure
Crown
cover
Subclimax
Climax
Duration of
Succession
Stage
0-10
10-20
20-30
30-100
100-150
150-300
Median
Year of
Succession
3
15
25
70
120
200
Median
Years to
Climax
200
185
175
130
80
0
Relative
Recovery
Index
5
4
3
• 2
1
0
Table B-2.  Hypothetical grassland sere.
Succession
Herbaceous
Annual Grasses
Co- dominance
Bunch grasses
Tall grass
Prairie
Physiognomy
Stage
Pioneer herbs
Covered
surface
Transition
Subclimax
Climax
Duration of
Succession
Stage
0-2
9-13
18-20
25-75
100- ?
Median
Year of
Succession
1
12
20
60
200
Median
Years to
Climax
200
190
180
40
0
Relative
Recovery
Index
4
3
2
1
0
                   1-2

-------
          0-1 NATURAL REHABILITATION


             HERB I SHRUB I TREE STAGES
                                MANAGED REHABILITATION
o

i-
LU
I


O
           LL
                                         V)

                                         V*
                                         O
                                         O
                                         a:
                                         Ul
                    10
                             100 200
                                                            10
                                                                      100 200
    Figure  B-l
Example  of times for recovery from  cleanup in coniferous

forest by  natural and managed rehabilitation.

-------
          assume the years are on a logarithmic scale and the median is the log
          of the mean (skewed towards the high side).  List the medians in column
          4 (Median Year of Succession).
      2.   Subtract each of the medians in column 4 from the climax median and
          enter the differences in column 5 (Median Years to Climax), rounding
          years >50 to the next decade either forward or backward.  Label column 5
          as "Years to Climax," observing that the entry for climax is zero.
      3.   Label column 6 as "Relative Recovery Index" and enter zero opposite
          climax in the physiognomy column.   Select significant stages in the
          sere (for which quantitative data in years are available)  as guideposts
          for the Index.   Starting from climax,  which is zero in the Index,
          assign numbers to the guideposts so the  sequence increases upwards  to
          the start of the sere.   Hardening surface becomes N+l and  N+2 becomes
          "Intervention  required to enter sere."  For practical use,  N+2 should
          be less than 10.

 B.3   RANKING OF DISTURBANCE  FROM CLEANUP TREATMENTS
      The  following paragraphs  describe the  ranking of cleanup  impacts  upon 0.01
 km2.
      To evaluate the  time  required for recovery from the  cleanup, define  the
 physical  disturbance  caused upon 0.01 km2 of  the  land type  by  performing  each
 one of the  treatments given in  the  working  matrix,  between  0-1  Natural  Rehabili-
 tation and  7-0  Adding Soil Amendments.   Enumerate  the potential  disbenefits
 that  will impact natural revegetation and specify  which kinds  of vegetation
 will  be helped  and  which will be harmed.   Consider erosiveness,  soil moisture
 relationships,  fertility and soil  tilth  reconstitution requirements for revege-
 tation to reenter  into succession  towards climax.   This is  the material for
 Section 4,  Effects  of Cleanup  (damage assessment).
      Discuss the retrogressive effects the physical  disruption causes upon
 climax conditions in the land type  for each of  the  cleanup procedures.  Esti-
mate which part of  the sere revegetation will start with, which parts of suc-
 cession can be skipped over, and which cannot be omitted in the physiognomy
progression.
                                    B-4

-------
      Rank  the  impacts  caused by each cleanup on specific parameters important
 in  the  ecosystem  and compile a  summary of the rankings  as in  Table  B-3.   Define
 the assumptions made during  the ranking and  record  them (see  Table  B-4,  column
 7)  for  the final  tabulation  in  Section 6,  Quantitative  Assessment.   Assemble
 the known  data, on  years  to  return  to climax after  specific impacts,  from Sec-
 tions 1  through 3 and  enter  the years in Table B-4,  Years Estimated to  Reach
                                                    2
 Climax.  Estimate the  missing years  for the  0.01  km  column of Table B-4  by
 using the  ranked  impacts  in  Table B-3.
     Next  consider  the effect that  increasing the area  cleaned up will  have
 upon the years estimated  to  pass before the  sites reenter climax.   For  each
 cleanup  treatment extrapolate,  the  recovery  years following retrogression from
                                    2
 climax to  areas larger than  0.01 km   and evaluate the impacts  at 0.10,  1.0,
 10.0 and 100 km.  Discuss those cleanups as  a unit  that  reattain climax by
 passing  through similar numbers of  stages  of succession.   This discussion
 forms the  interpretive backbone of Section 6,  Quantitative Assessment.  Fill
 in  the remaining  four  columns of Table  B-4,  Years Estimated to Reach  Climax.
     Pick  a sequence of points  in time  to  discuss that extends from the first
 year after cleanup until  climax has been reattained.  Compare  that  scenario to
 columns  1  and 2 of Tables B-l and B-2,  at  frequent  intervals in succession,
 and  group  those treatments together that pass  through the  same physiognomy
 stages at  the same time.  These assessments,  when written up,  constitute  the
 text of Section 5, "Recovery  After Cleanup."   The text should  also  include
 comparisons of the times  like groups  require  to attain climax  and should  iden-
 tify irreversible impacts.

 B.4  RECOVERY INDEX NUMBER ASSIGNMENT
     Use the median year of succession  for the cleanup treatment (Table B-4)
                                  2
 as the entry into the  treatment-km  matrix, basing the median  on the  stage of
 succession or stage of physiognomy generated under the heading  "Derivation of
 Recovery Index from Successional Sere."  Median years are required for 0.01,
                     2
 0.1, 1.0, and 10.0 km  areas  for each treatment in the matrix  (Section 6).
Those treatments that were deleted should be indicated by an asterisk.  Those
which require man's intervention before they enter into succession are to be
given #  symbols.   An NA should be used  in the matrix for treatments that should
not be used in the land type, otherwise defined as irreversible impacts. Trans-
fer the  "Assumptions" from column 7, Table B-4 to column 7, Table B-6.

                                      B-5

-------
                                    Table B-3.   Evaluation  of the effects of  cleanup procedures.
03
I
CTi
             Environmental
             Parameter
           Plant Succession
           Wildlife (small animals)
           Wildlife (large animals)
           Water Runoff
           Soil  Erosion*
           Microclimate
                                               Ranked Treatment Effect (least effect
4-1 >» 2-6 >  2-3 > 2-2  >   2-1  >  1-4 > 1-8 > 1-7 > 1-3 >  1-6 > 1-2 =

4-1 >» 2-6 >  2-3 > 1-8  >   2-1  >  1-7 > 1-4 = 2-2 > 1-5 >  1-3 > 1-6 >

4-1  >  3-0 >  2-6 = 2-3  »  1-8  >  2-1 > 2-2 > 1-7 > 1-4 >  1-3 = 1-5 >
                                                                                                1-5

                                                                                                1-2

                                                                                                1-6
                         4-1 >» 2-6 >  2-3 >  2-2  >   2-1  =  1-4 = 1-7 _> 1-8 > 1-5 >  1-3 > 1-6 >   1-2

                         2-2  >  1-4 =  2-1 >  1-7 _>   1-8  >  1-5 > 1-6 > 1-3 > 1-2 » 5-0 > 3-0 »  2-8

                         4-1  >  2-6 =  2-3 =  1-2  =   1-3  =  1-4 = 1-5 = 1-6 = 1-7 =  1-8 = 2-1 =   2-2  >»  3-0   =
                                                                   5-0  >   3-0

                                                                   5-0  >   3-0

                                                                   1-2  >   5-0

                                                                   3-0  >   5-0

                                                                   2-3 »>  4-1

                                                                           5-0
           *The effects are assumed  to  be  limited to erosion on site, off-site sedimentation  is  not  considered.
Key to effects  ranking:
  1-2
  1-3
  1-4
  1-5
  1-6
  1-7
  1-8
Clear Cutting
Stumping and Grubbing
Scraping and Grading
Shallow Plowing
Deep Plowing
Soil Cover (<25 cm)
Soil Cover (>25 cm)
                                                                              2-1
                                                                              2-2
                                                                              2-3,4,5

                                                                              2-6,7,8
                                                                              3-1,2,3
                                                                              4-1,2

                                                                              5-0
                                                                             Soil Removal (<15 cm)
                                                                             Soil Removal (>15 cm)
                                                                             Soil Removal, Mechanically Stabilize
                                                                             by Compaction
                                                                             Soil Removal, Chemically Stabilize
                                                                             Animal Barriers (including people)
                                                                             Hard Surface Stabilization (asphalt,
                                                                             concrete)
                                                                             Sewage Sludge Application

-------
      Table B-4.   Years  estimated  for hypothetical  ecosystem  to reach climax after the  indicated cleanup.
Action Evaluated
0-1 Natural Rehabilitation
1-1 Chemical Stabilization
1-2 Remove Vegetation by Clear Cutting
1-3 Remove Stumps and Grub Roots
1-4 Surface Scraping and Grading (<5 cm)
1-5 Shallow Plowing (<10 cm)
1-6 Deep Plowing (10 to 40 cm)
1-7 Soil Cover (<2S cm)
1-8 Soil Burial (25 to 100 cm)
2-1 Remove Plow Layer Soil (10 cm)
2-2 Remove Shallow Root Zone Soil (<40 cm)
2-3 Scrape Surface, Remove (<5 cm), Mechanically Stabilize
2-4 Remove Plow Layer (10 cm), Mechanically Stabilize
2-5 Remove Shallow Root Zone (<40 cm), Mechanically Stabilize
2-6 Scrape Surface, Remove (<5 cm), Chemically Stabilize
2-7 Remove Plow Layer (10 cm), Chemically Stabilize
2-8 Remove Shallow Root Zone (40 cm), Chemically Stabilize
3-1 Barriers to Exclude People
3-2 Barriers to Exclude Large Animals
3-3 Barriers to Exclude Large and Small Animals
4-1 Asphalt Hard-Surface Stabilization
4-2 Concrete Hard-Surface Stabilization
5-0 Sewage Sludge Application
6-1 Washing (<3 mm)
6-2 Flooding (3 to 30 cm)
7-0 Soil Amendments Added
Area Disturbed
(km2)
0.01
150
90
60
70
80
60
70
150
150
160
180
200
200
200
200
200
200
0
0
0
>200
>200
0
0
0
0
0.1
150
90
60
70
80
60
70
150
150
160
190
200
200
200
200
200
200
0
0
0
>200
>200
0
0
0
0
1.0
160
100
60
80
90
60
70
170
170
180
200
200
200
200
200
200
200
0
0
0
>200
>200
0
0
0
0
10.0
170
110
70
90
100
70
80
180
180
>200
200
200
200
200
200
200
200
0
0
0
>200
>200
0
0
0
0
100.0
200
>200
90
100
150
90
100
>200
^200
>200
>200
200
200
200
200
200
200
0
0
0
>200 {
>200 (
0
0
0
0
Assumptions
(Precleanup state is climax)
No seed reserve, average seed bed
Seed reserve, average seed bed
Seed reserve, good seed bed
Seed reserve, good seed bed
Some seed, poor seed bed, erosion
Seed reserve, good seed bed
Some seed, good seed bed
No seed reserve, average seed bed, erosion
No seed reserve, average seed bed, erosion
No seed reserve, poor seed bed, erosion
No seed reserve, very poor compacted seed bed
No seed reserve, very poor compacted seed bed
No seed reserve, very poor compacted seed bed
No seed reserve, very poor compacted seed bed
No seed reserve, very poor compacted seed bed
No seed reserve, very poor compacted seed bed
Applied to climate
No significant effects on climax*
No significant effects on climax
At least 50 years required for
concrete breakup
No significant effects on climax*
No significant effects on climax*
No significant effects on climax*
No significant effects on climax*
*The sampling (variability between sites) error greatly exceeds the instrumental (treatment) error.

-------
       Go to Section 7 and select the least damaging treatments and discuss their
  advantages over other treatments as part of the Conclusions Section (Section 7)
  Perform the same analysis for treatments that are not usable on the land type,
  rated # or NA in Tables B-5 and B-6.

       Once  the recovery  times and the area effects are evaluated  complete the
  Recovery Index and  Guide  Numbers,  columns 6  and 7 in  Table  B-5.   Cleanup that
  leaves  the site at  climax is jrated as  zero impact even though  it  may be  some
  time  before it regains  productivity, for example  mowing bunchgrass.  The Index
  Numbers  "count" the stages  in a  sere.  The largest one represents the greatest
  site  retrogression and  the  longest  recovery  time  to regain  climax; its Guide
  Number is  N.   Cleanup recovery that will progress through succession but de-
  lays  in starting recovery is more  severe than those which begin succession
  immediately.   Nevertheless,   they are of the  same order  and  kind of recovery
 as those with  a smaller Recovery Index, and  the Guide Number is increased one
 to account  for the greater time for recovery, from N to N+l.  An example of
 an N+l Guide Number is that  of breakup  of asphalt hard surface that delays
 succession  20 years in the hypothetical coniferous forest  of Table B-5.
      In some ecosystems, cleanup can provide  changes  which allow secondary im-
 pacts  severe enough to prevent a succession occurring  that leads  to  the  pre-
 cleanup climax.  One example is  removal of the plow layer  from  conifers  on a
 shallow soil on a steep  hillside,  taking  the  accumulated organic  matter  and
 surface horizon mineral  soil. Subsequent erosion  over time  can eliminate the
 contaminated,  cleaned-up site by  replacing it with a gully that continues to
 the barren  bedrock.  The original  forested site  no longer  exists.  In this
 case,  the timely intervention by  site renovators could return the  site of
 cleanup  into conditions  that  bring  back the prior  sere.  These  cleanup impacts
 are defined as  the "Conditional Stage"  and are another  one of the pre-succession
 stages.  They are given  a  Recovery.Index  that is non-numerical  (the symbol used
 in Table B-5 is  #); the  analogous Guide Number is N+2,  which should be less
 than 10 for convenience.

     The final  Guide Number is that assigned  to cleanup  impacts which produce
 "Irreversible"  impacts  and in Table B-5  the acronym is NA.   This concludes
the mechanism for generating   Index and Guide Numbers to represent the impacts
caused by contamination cleanup.
                                      B-8

-------
                      Table B-5.  Hypothetical coniferous sere as  represented  by  the  development of a
                                  Spruce-hemlock forest on mineral  soil  exposed by  glacial  retreat.*
Succession Stage
deciduous shrubs
mixed shrubs
Brush thicket
deciduous trees
intolerant conifers
mixed conifers
muskeg
delayed
conditional
irreversible
retrogression
Physiognomy
Stage
pioneer invasion
cover closure
crown cover
crown cover
sub climax
climax
retrogression
(asphalt breakup)
site retrogression
excessive
site retrogression
excessive
Duration of
Succession Stage
(years)
0-7
7-1Q
10-60
60-90
90-170
130-200
>200
10-25
until renovated**
permanent
Median Year
of Succession
2.5
8.4
35.0
74.0
130.0
170.0
200
220
—
—
Years
to Climax
200
190
165
125
70
30
0
>200
—
—
* After Lawrence
Recovery
Index
6
5
4
3
2
1
0
7
#
NA

Gui de
Numbers
N





0
N+l
N+2
N+3

**Requires one or more site renovation treatments: terrace, drain, reseed, transplants, fertilizer,
irrigation, mulching (human intervention).
CD
I

-------
                       Table  B-6.   Recovery  indexes  for  cleanup  of  hypothetical coniferous  forest.
Action Evaluated
0-1 Natural Rehabilitation
1-1 Chemical Stabilization
1-2 Remove Vegetation by Clear Cutting
1-3 Remove Stumps and Grub Roots
1-4 Surface Scraping and Grading (<5 cm)
1-5 Shallow Plowing (<10 cm)
1-6 Deep Plowing (10 to 40 cm)
1-7 Soil Cover (<25 cm)
1-8 Soil Burial (25 to 100 cm)
2-1 Remove Plow Layer Soil (10 cm)
2-2 Remove Shallow Root Zone Soil (<40 cm)
2-3 Scrape Surface, Remove (-5 cm), Mechanically Stabilize
2-4 Remove Plow Layer (10 cm), Mechanically Stabilize
2-5 Remove Shallow Root Zone (-.40 cm), Mechanically Stabilize
2-6 Scrape Surface, Remove (<5 cm), Chemically Stabilize
2-7 Remove Plow Layer (10 cm), Chemically Stabilize
2-8 Remove Shallow Root Zone (40 cm), Chemically Stabilize
3-1 Barriers to Exclude People
3-2 Barriers to Exclude Large Animals
3-3 Barriers to Exclude Large and Small Animals
4-1 Asphalt Hard-Surface Stabilization
4-2 Concrete Hard-Surface Stabilization
5-0 Sewage Sludge Application
6-1 Washing (<.3 mm)
6-2 Flooding (3 to 30 cm)
7-0 Soil Amendments Added
Area Disturbed
(km2)
0.01
4
2
2
2
2
2
2
4
4
4
5
7
#
#
4
4
4
o
o
0

i
0
*
*
*
0.1
4
2
2
2
2
2
2
4
4
4
5
7
#
4
4
4
o
o
0
n

0
*
*
*
1.0
5
3
2
2
2
2
2
4
4
4
5
7
4
4
4
n
u
n
u
0

#
0
*
*
*
10.0
5
3
2
2
3
2
2
C
j
5
c
o
7
/
7
n
c
j
c
J
5

n
u
0


0
ir

*
100.0
6
c
o
2
2
4



6
7
/
7
/
7
7
/
7
/
7


0

J
0

*
*
*0utside scope of work.
Assumptions
(Precleanup state is climax)

























ro
i

-------
B.5  APPENDIX B REFERENCE

1.    Lawrence, D.  B.   The Glaciers and Vegetation in Southeastern Alaska,
     American Scientist, Summer 1958.  pp. 89-122.
                                   B-ll

-------

-------
                                APPENDIX C
                                LAND TYPES
Aeolian Mountain Peaks — High mountain terrain above trees and flowering
        plants, above alpine zones.

Coastal Inter-Tidal Marsh — Low, flat marshlands traversed by interlaced
        channels amid tidal sloughs and subject to tidal inundation;
        normally, the only vegetation present is salt-tolerant bushes and
        grasses.

Coniferous Forest — A plant association predominantly of trees of the
        order Coniferae, having evergreen leaves shaped like needles or
        scales; such trees produce softwood lumber.  Understory may be
        highly variable, from humid ferns to semiarid grasses and shrubs.

Deciduous Forest — A plant association predominantly of trees and other
        woody vegetation which seasonally loses all its leaves.

Desert — An area of land that has an arid, hot to cool climate with
        vegetation that is sparse and usually shrubby.

Prairie — A tract of level to hilly land that has a dominance of grasses
        and forbs, has a scarcity of shrubs, and is treeless.  The natural
        plant community consists of various mixtures of tall, mid- and
        short-growing native species, also known as true prairie, mixed
        prairie, and shortgrass prairie, respectively.

Salt Marsh — Low area adjacent to the sea that is covered with salt-tolerant
        vegetation and regularly flooded by the high tide; similar inland
        areas near saline springs or lakes, though not regularly flooded.

Tundra — The treeless land in arctic and alpine regions;  varying from bare
        area to various types of vegetation consisting of grasses, sedges,
        forbs, dwarf shrubs, mosses, and lichens.
                                   C-l

-------

-------
                             APPENDIX D
                         CLEANUP TREATMENTS
Treatment   Treatment
 Number       Name

  1-1      Chemical
           Stabilization
  1-2
  1-3
  1-4
  1-5
  1-6
1-7
1-8
         Clearcutting
         Stumping and
         Grubbing

         Surface
         Scraping and
         Grading

         Plowing
           Soil  Cover
Erosion prevention through the use of soil addi-
tives to promote increases in soil cohesion or
shear strength, bulk density, water retention,
and/or surface cohesion  (see Appendix A).

A method of cutting that removes the entire
vegetational stand; usually refers to lumbering
operations.  Synonyms here include mowing.

The operation of removing stumps and roots.
Removal of the upper soil layers to 5 cm by means
of bulldozers, road graders, or other heavy or
medium equipment with controlled cutting.

Employment of a device to cut, lift, break up, and
turn over the soil, usually leaving parallel
furrows on the surface.  Depth ranges are:
Shallow (<10 cm) and Deep (10 — 40 cm).

Soil collected from uncontaminated borrow pits,
surface areas, road excavations;  generally expected
to be a poorer medium for plant growth than the
soil covered up.
  2-1       Removal of
           Plow  Layer
           Soil
  2-2      Removal of
          Shallow Root
          Zone Soil

  2-3      Surface Soil
  2-4      Removal with
  2-5      Mechanical
          Stabilization
                           "Plow layer soil" defined here as from surface to
                           depth of about 10 centimeters.  Removal can be by
                           hand shoveling,  bulldozers,  scrapers,  front-end
                           loaders.

                           "Shallow root  zone" defined  as that  part of the
                           soil,  up to a  depth of 40 cm,  that is  penetrated
                           or can be penetrated by plant  roots.

                           Physical removal of contaminated surface and soil
                           from the affected area with  depth-controlled equip-
                           ment,  dozers,  front-end loaders,  or  scrapers.   The
                           exposed surface  remaining is mechanically compacted
                           to a maximum density by heavy  equipment or is  resur-
                           faced with impervious materials such as soil-cement
                           or NaCl.  The  resultant effect is to remove the con-
                           taminant and to  seal the exposed surface against
                           water infiltration.
                                2-3  The  cut removes less than  5  cm of material
                                2-4  Up to  10 cm of material may  be removed.
                                   D-l

-------
 2-6
 2-7
 2-8
 Surface Soil
 Removal with
 Chemical
 Stabilization
3-1
3-2
3-3
Barrier
4-0

4-1
4-2
Mechanical
Stabilization
(asphalt)
(concrete)
 2-5  Removes the shallow root-zone and may ex-
      tend to 40 cm.  The deepest soil removal
      may cause temporary impoundment of surface
      waters.

 The physical  operations for these treatment pro-
 cedures are similar to those discussed for 2-3
 to 2-5.  The  basic differences between the two
 treatment types are:

    a.   the method of  subsurface stabilization
    b.   the size of the treated area.

 Chemical  subsurface stabilization is the applica-
 tion of one of the chemical  forms of stabilizers
 as defined in "Appendix A,  Stabilizers" to create
 change  in the physical soil  structure.

 Most chemical stabilizers  are restricted to areas
 in size where the surface  soils were  removed by
 hand methods  or with  small  tractor end-loader
 equipment (from 0.01  km2 to  0.1 km2).   The usual
 application of the selected  chemical  stabilizer
 is by hand methods.   This  foregoing description
 must not  be construed to preclude the use  of
 chemical  stabilization on  larger areas  where
 heavy equipment  can be used  for the chemical
 application.   However,  the  treatment  is  generally
 restricted to smaller areas  or  to urban  situations.

 A  vertical  impediment  such as a wall or  fence which
 serves  to  prevent  horizontal  animal movements.
 A  reinforced  heavy high barrier for large  animals
 (bison, moose, elk) and a buried edge fine mesh
 barrier  (including sheet metal  flashing) to  impound
 small animals  (pocket  mouse, moles, grass  vole,  but
 excluding  bats).  Horizontal barriers  (screens  or
 sheets) may also  be employed to  prevent  vertical
 animal movements.

 Erosion prevention through compaction of the  surface
 soil through  tamping,  rolling,  vibration,  electro-
 static methods, plastic sheeting,  and other non-
 chemical means.  Hard-surface stabilization  (paving)
 is  included here.
5-0
Sewage
Sludge
Settled sewage solids combined with varying amounts
of water and dissolved materials that are removed
from sewage by screening, sedimentation, chemical
precipitation, or bacterial digestion.

Application may serve to increase soil permeability,
increase soil fertility, and increase soil water
retention, thus reducing soil erosion and contaminant
mobility.
                                     D-2

-------
6-1      Washing           A system to spray water under high pressure over
                           streets, buildings and impermeable surfaces to sus-
                           pend and remove contamination in a volume of water
                           equivalent to 3 mm or less over the washed surface.

6-2      Flooding          Inundation of a surface by water in an effort to
                           wash a contaminant downslope or into deeper soil
                           regions, using 3 to 30 cm of water.

7-0      Soil              Any material, such as lime, gypsum, sawdust, or
         Amendment         synthetic conditioner, that is worked into the
                           soil to make it more amenable to plant growth.

                           Plant growth is stimulated to act as a surface
                           cover and thus as an erosion preventative; some
                           amendments may also serve to prevent contaminant
                           uptake by plants or to promote contaminant uptake
                           by plants, in which case the plants could be
                           removed for storage as contaminated wastes, having
                           incorporated the contaminant in their tissues.
                                     D-3

-------

-------
                             APPENDIX E

                              GLOSSARY



 Aeolian Soil  Material  — Soil  material  accumulated through wind action.

 Adsorption — The molecular attraction  of a substance to the surface
        of a solid or liquid.


 Afforestation — The artificial establishment of forest crops by planting
        or sowing on land that  has not previously,  or recently,  grown trees

 Agricultural  Land — Land in  farms regularly used  for agricultural pro-
        duction.

 A Horizon —  See Soil  Horizon.


 Algae  —  Simple  plans,  many microscopic,  containing chlorophyll;  forming
        the base  of the  food chain in  aquatic environments.

 Alluvial  — Pertaining  to material that  is  transported and  deposited by
        running water.


 Alpine  —  That portion  of mountains above tree  growth or organisms living
        there.


 Animal  Unit — A measurement of livestock numbers  based  on  the  equivalent
        of  a mature  cow  (approximately 1,000  pounds  live  weight).

 Annual  Plant  —  A plant  that completes its  life  cycle and dies  in  1  year
        or  less.


 Ap — The  surface layer  of  a soil  disturbed  by cultivation  or pasturing.

 Arborescent — Resembling a tree  in properties,  growth,  structure, or
        appearance.


 Arid — Regions  or climates that  lack sufficient moisture for crop pro-
        duction without irrigation.

 Aspect  (forestry) — The direction that  a slope  faces.

Autogenic — Orginating within.


Autotroph — An organism that  manufactures its own food, such as a plant.

Benthos — The plant and animal life whose habitat  is the bottom of a
       sea, lake, or river.


Berm — A shelf o^r flat area that breaks the continuity of a slope.

B Horizon — See  Soil Horizon.
                                E-l

-------
 Biocoenosis - The plants and animals comprising a community.

 Biomass - The total amount of living material in a particular habitat
        or area.


 Biome — A unit of plant and animal communities having similarities
        in form and environmental conditions.

 Biota —  The flora and fauna of a region.


 Blowout - (1)  An excavation in areas of loose soil produced by wind;
        (2) a rupture of soil surface attributable to hydraulic pressure.

 Bole — The trunk of a tree.

 Boreal — Of,  relating to,  or located in northern regions.

 Brunizem (Prairie)  Soils —  The group of soils developed under tall
        grasses in a temperature,  relatively humid climate.

 Buffer Strips  —  Strips of  erosion-resisting vegetation between culti-
        vated strips or fields.

 Canopy —  The  cover of leaves  and branches formed by the tops or crowns
        of plants.


 Catch Crop —  1)  A  crop produced  incidental to the main crop,  usually
        occupying  the  land for  a  short  period;  2)  A crop grown to  replace
        a main  crop  which has failed.

 Check Dam  — Small  dam  constructed  in  a  small  watercourse to  decrease
        the streamflow velocity, minimize channel  scour,  and promote
        deposition of  sediment.

 Chiseling  — Breaking or loosening  the soil  with  a chisel cultivator  or
        chisel plow.

 C Horizon  — See Soil Horizon.


 Ciliated — Provided  with a  minute  hairlike  process of many cells that
        is  capable of  lashing movement.

 Clearcutting — A cutting method that removes the entire timber stand.

 Climate — The total of all  atmospheric or meteorological influences,
       principally temperature, moisture, wind, pressure, and evaporation

 Climax — The terminal  stabilized system of  an ecological succession.

Clone — A group of organisms derived by asexual reproduttion from a
       single parent.

Coarse Texture —  The soil  texture exhibited by sands.
                                E-2

-------
Co-dominant Trees — Trees with crowns forming the general level of the
       forest canopy and receiving full light from above but comparatively
       little from the  sides.

Colluvial — Material that has moved downhill and has accumulated on
       lower slopes and/or at the bottom of the hill; material of
       avalanches.

Community — An aggregation of organisms within a specified area.

Compaction — The process by which soil grains are compressed, thereby
       increasing the weight of solid material per cubic foot.

Conifer — A softwood tree belonging to the order Coniferae with cones
       and evergreen leaves of needle shape or "scalelike."

Contour Ditch — Irrigation ditch laid out approximately on the contour.

Contour Farming — Plowing, planting, cultivating, and harvesting on the
       contour.

Coppice - A growth of small trees originating mainly from sprouts or root
       suckers rather than from seed.

Copse — Coppice.

Cover, Percent — The area covered by plants and mulch expressed as a
       percent of total area.

Cropland — Land used primarily for the production of cultivated crops.

Crown (forestry) — The upper part of a tree, including the branches
       and foliage.

Crown Class — All trees in a stand with tops or crown occupying a
       similar position in the canopy.  Crown classes usually distinguished
       are:

       Dominant — Trees with crowns extending above the general level
            of the forest canopy.

       Co-dominant — Trees with crowns forming the general upper level
            of the forest canopy.

Crown Cover — The canopy formed by the crowns of all the trees in a forest,

Cutting Cycle — The planned interval between major cutting operations
       in a managed woodland tract.

Deciduous Plant — A plant that sheds all its leaves every year at a
       certain season.
                                E-3

-------
                                             particies  °

 Delta - An alluvial deposit formed where a stream or river drops its
        sediment load on entering a body of more quiet water.

 Dendritic — Tree-like.


 Detritus — Matter worn from rocks by mechanical means; generally
        alluvial deposits.

 Diatom — Algae with silicified skeletons.

 Dominant (ecology) - Species which by their activity, behavior, or
        number have considerable influence or control upon the conditions
        of existence of associated species.

 Duff —  The organic layer  on top of mineral  soil.

 Ecology  —  The  study of interrelationships of  organisms to one another
       and  to  their environment.


 Ecosystem — A  community,  including  all  the  component  organisms,  together
       with  the environment,  forming an  interacting  system.

 Ecotone  — A transition strip of  vegetation  between  two  communities,  having
       characteristics  of  both  kinds of  neighboring  vegetation as well  as
       characteristics  of  its own.

 Environmental Impact  Statement  — A document detailing the  environmental
       impact of a proposed  action that may  significantly  affect the  quality
       of the environment.

 Erosion — The  wearing  away of  land surface by running water,  wind, ice,
       or other geological agents.

       rill erosion —  a process  in which numerous small channels only
            several inches deep are formed.

       sheet erosion  — the removal of a fairly uniform layer  of soil
            from the  land surface by runoff water.

       splash erosion — the spattering of small soil particles caused
            by  the impact of raindrops on wet soils.

Evergreen — Perennial plants that are never entirely without green
       foliage.


Fallow — Allowing cropland to lie idle,  either tilled or untilled,  during
       the whole or greater portion of the growing  season.

Farm Forestry — The practice of forestry on  farm or  ranch lands generally
       integrated with other farm or ranch operations.
                                E-4

-------
 Fertilizer  — Any organic or inorganic material that is added to a soil
        to supply elements essential to plant growth.

 Fibrous Root System  — Having a  large number of small, finely divided,
        widely spreading roots but no large individual roots.

 Field  Planting  (forestry) — The establishment of woody plants on land
        essentially free of trees.

 Forb — An  herbaceous plant which is not a grass, sedge, or rush.

 Fragipan — A natural subsurface horizon, low in organic matter, with
        high bulk density; seemingly cemented when dry but showing a
        brittleness when moist.

 Frost  Heave — The raising of a  surface due to the accumulation of ice
        in the underlying soil.

 Gabion — A wire mesh cage filled with rock and used as a protecting
        apron against erosion.

 Gley (soil) — pale  gray to bluish white horizon high in ferrous
        compounds and other oxygen deficient decomposition products.

 Grass  — A member of the botanical family Gramineae,characterized by
        bladelike leaves arranged on the stem in two ranks.

 Grove  — A small group of trees, usually without understory, planted or
       natural.

 Gully  Control Plantings — The planting of forage in gullies to establish
        a vegetative  cover adequate to control runoff and erosion.

 Habitat — The environment in which the life needs of a plant or animal
       organism are supplied.

 Hardwoods — Trees that have broad leaves, in contrast to the conifers;
       also wood produced by trees of this group regardless of texture.

 Herb — Any flowering plant except those developing persistent woody
       bases and stems above ground.

 Heterotrophic — Pertaining to organisms that are dependent on organic
       material for food.

Homeostasis — Ecologic inertia of a population,  under stress by changing
       conditions,  which restores a prior equilibrium through self-adjust
       ment.

Hydric —  Requiring an abundance of moisture.

Hydrophyte —  A plant that grows in water or in wet  or saturated soils.
                                E-5

-------
  Illuvial  —  A soil  layer or horizon in which material  carried from an
         overlying layer has been precipitated from solution or deposited
         from suspension.   The layer of accumulation.

  Insolation —  Incoming  solar radiation.

  Interstices  — The  pore space or voids  in  soil  and rock.

  Invader Plant Species -  Plant species  that  were absent in  undisturbed
         portions  of the original  plant  community and will invade under
         disturbance or  continued  overuse.

  Invasion  -  The migration and establishment  of  organisms from  one  area
         to another  area.

  Irreversible — Incapable of  correction or of being reversed.

  Kame — A conical hill  or short irregular ridge or gravel or sand
         deposited in contact with glacier ice.

 Mesic (soil)  — formed under a mean annual temperature range of 15° to
         22°C (59° to 72°F) at 50 cm depth.

 Mesophyte  —  A plant that grows under intermediate moisture conditions.

 Microflora — Bacteria,  including actinomycetes, viruses,  and fungi.

 Mitigation (wildlife)  - The reduction or elimination of damages to
         fish  and  wildlife resources.

 Montane -  (1)  pertaining to  mountain conditions; (2)  the lower vegetation
       belt on mountains.

 Moraine  - An accumulation of rock debris by  the direct  action of glacial
         ice.
Mulch  - A natural or  artificial  layer of plant  residue  or other materials
        such as  sand or paper, on the soil  surface.

Muskeg — Bog; especially a sphagnum bog of northern  North America.

Mycorrhizal - The fungal complex with the  roots of a seed plant.

Native Pasture - Land on which the climax  plant community is forest  but
        which is used and managed primarily  for the production of native
        species  for forage.

Native Species - A species that  is a part of an area's original fauna
        or flora.

Niche - A habitat that supplies the factors necessary for the existence
        of an organism or species.
                                 E-6

-------
Old-Field — An abandoned agricultural field.

Outwash — The material, cheifly sand or gravel, washed from a glacier
   by the action of meltwater.

Overburden — The earth, rock, and other materials that lie above a
   mineral deposit.


Overstory — The portion of the trees in a  forest  stand forming the
       upper crown cover.

Phenology — The study  of the time of appearance of characteristic
       periodic events  in the life cycles of organisms in nature and how
       these events are influenced by environmental factors.

Phreatophyte — A plant deriving its water  from subsurface sources;
       commonly used to describe nonbeneficial, water-loving vegetation.

Phytoplankton — Unattached microscopic plants of  plankton, subject to
       movement by wave or current action.

Physiography — A description of nature or  natural phenomena in general.

Piping — Removal of soil material through  subsurface flow channels developed
       by seepage water.

Pitting — (1)  making shallow pits to retain water from rainfall or snowmelt
       on rangeland or pasture;  (2) small cavities in a surface created by
       corrosion,  cavitation,  or subatmospheric pressures.

Plant Succession — The process whereby an  area becomes successively
       occupied by different plant communities of higher ecological order.

Primary Productivity — The rate (grams per day) at which organic matter
       is stored by photosynthesis and chemosynthesis of producer organisms
       (autotrophs).

Protozoa — Single-celled animals.

Pyric — Resulting from, induced by, or associated with burning.

Range — (1)  Rangeland,  and forestlands with understory suitable for grazing;
        (2)  the geographic area  occupied by an organism.

Reclamation — The process of reconverting disturbed lands to their former
       uses or other productive  uses.

Rehabilitation —  Returning of land to productivity in conformity with
       prior land use,  including a stable ecological state consistent with
       surrounding aesthetic values.
                                    E-7

-------
  Relict  -  A remnant  or fragment  of a flora that  remains  from a former
          period when  it was  more  widely distributed.
 Retrogression - Reverse succession under environmental stress.

 Savanna - A grassland with scattered trees; often a transitional type
         between true grassland and forest.

 Scarify — To abrade, scratch, or modify the surface.

 Scenario - The order of events in time describing interacting forces
         environmental influences and organisms.

 Scree - (1) Pebble, stone;  (2) a heap of stones or rocky debris.

 Seedbed - The soil prepared to promote the  germination of seed and
         the growth  of seedlings.

 Sequum -  The generally gradual changes in soil properties in horizontal
         or vertical directions from the soil body in  question.

 Seratinous —  Late, especially in  developing or flowering.

 Sigmoidal  —  A slanted  S-shaped curve.

 Site  (ecology) - (1) an area considered for  its ecological factors with
        reference to capacity  to produce vegetation; the combination of
        bl°^lc? climatic, and  soil conditions of an area;  (2) an area
        sufficiently uniform in soil, climate, and natural biotic con-
        ditions to produce a particular climax vegetation.

 Snow  Fence  - A fence used in  winter to intercept drifting snow, thus
        protecting  roads and other areas from snowdrifts.  Also used to
        impound snow where melting in place  adds to soil moisture.

 Soil Amendment - Any material, such as lime, gypsum,  sawdust, or synthetic
Soil Horizon - A layer of soil or soil material approximately parallel
        to the land surface and differing from adjacent genetically related
        layers in physical, chemical,  and biological properties or charac
        teristics, such as color,  structure,  texture, Consistence^ kinds

        e'tc nTeLrS,naf°rganirS PieSent' ****** °f acidit^ Or Alkalinity,
        as foHowsf      S     Pr°Perties of  three  major soil  horizons  are
       A Horizon.  Topsoil,  the  zone of maximum concentration of  soil
           organisms and biotic  activity.  This is the horizon  in which
           organic debris becomes converted into humus and mixed  with
           mineral matter.   It is the zone of eluviation in which pre-
                                 E-8

-------
              cipitation, mixed with acids from decomposing organic matter
              dissolves and carries further down into the soil the more
              soluble materials

          B Horizon.  The lowest true soil, or subsoil.  This zone lying
              below the horizon of maximum leaching is an area of illuvi-
              ation, or deposition for minerals leached out of the topsoil,
              and in particular it is a layer in which clays tend to accum-
              ulate.  In arid regions it will be a layer of accumulation
              for calcium carbonate, magnesium carbonate, gypsum, or other
              soluble salts.

          C Horizon.  A layer of unconsolidated material and rock fragments
              lying above the unmodified parent rock and below the true
              soil.  It usually contains little material modified by living
              organisms, although it may be broken, split or cracked by the
              roots.

Solum — The upper and most weathered part of the soil  profile;  the A and B
        horizon.


Spoil — Soil  or rock material excavated from a canal,  ditch,  basin,  or
       similar construction.

Stabilized Grade —  The slope of a channel  at which neither erosion nor
       deposition occur.

Stratification — The process or arrangement or composition in strata
       or layers.

Strip Mining — A process in which rock and  top soil strata overlying min-
       eral deposits are scraped away by mechanical shovels.   Also known
       as surface mining.

Substrate — (1)  in biology,  the base of substance upon which an organism
       is growing;  (2)  in chemistry,  a substance undergoing oxidation;
       (3) in  hydrology,  the bottom material of a waterway.

 Succession  — The progressive  development  of  vegetation  toward  its  highest
        ecological expression,  the  climax;  replacement  of  one  plant  com-
        munity by another.

Surface Compaction — Increasing the dry density of surface soil by
       applying a dynamic load.

Surface Water — All water whose surface is exposed to the atmosphere.

 Tacking  — The  process of binding mulch fibers together by the addition
        of a sprayed chemical compound.

 Taxonomy —  (1)  the  science of  classification;  (2)  classification of
        animals  and  plants, such as species, genus, family, and order.
                                     E-9

-------
 Tilth  - The physical  condition  of soil  as related to its ease of tillage
        fitness as  a  seedbed,  and impedance to seedling emergency and root'
        penetration.

 Topsoil - (1) earthy material used as top-dressing  with favorable charac-
        teristics for production  of desired kinds of vegetation-  (2)  the
        surface plow layer of a soil; (3) the  dark-colored upper  soil  that
        ranges in depth with different kinds of soil;  (4) the A horizon
        varying widely among different kinds of soil.

 Trophic Level - The level in a  nutritive  series of an ecosystem in
        which a group or organisms in a certain stage  in the food chain
        secures food in the same general manner.  The  first or  lowest
        trophic level consists of producers (green plants); the second
        level of herbivores; the third level of secondary carnivores- and
        the last level of reducers.

 Truncated Soil  Profile -Soil profile that has been cut down by  accelerated
        erosion  or by mechanical means.   The profile may have lost part or
        all  of the A horizon and sometimes the B horizon, leaving only the
        C  horizon.

 Undergrowth  (forestry)  -  Seedlings,  shoots,  and small saplings under an
        existing  stand of trees.

 Understory  - That  portion of the trees  in a forest below the  upper
        crown  cover,  also called underwood.

 Urban Land — Areas  so  altered  or obstructed  by urban  works  or structures
        that  identification of soils  is not  feasible.  A miscellaneous land
        type; towns  of over 2,500  are  defined  as  urban  by the U.S. Bureau
        of Census.

 Vegetation - Plants  in  general or the sum  total of  plant  life  in an  area.

 Vertical Stratification  — The process of arrangement  or composition
        in strata or layers vertically, as in understory  shrubs  and over-
        story tree canopy.

Wildlife — Undomesticated vertebrate animals, except  fish, considered
       collectively.
                                       * U.S. GOVERNMENTHMTNK OFFICE: 1971- 720-335/6170
                                   E-10

-------