United States Oftice of Research and EPA/620/R-93/001
Environmental Protection Development January 1993
Agency Washington DC 20460
Arid Colorado
Plateau Pilot
Study-1992
Implementation Plan
Environmental Monitoring and
Assessment Program
-------
EPA/620/R-93/001
January 1993
ENVIRONMENTAL MONITORING AND ASSESSMENT PROGRAM
EMAP-Arid Colorado Plateau Pilot Study - 1992
Implementation Plan
Susan E. Franson (Editor)
Desert Research Institute
Cooperative Agreement Number CR-816385-02
This study was conducted in cooperation with
U.S. Department of the Interior
Bureau of Land Management
Reno, Nevada 89520
U.S. Department of Agriculture
Soil Conservation Service
Washington, DC 20013
U.S. Forest Service
Fort Collins, Colorado 80526-2098
U.S. Department of Energy
Idaho Falls, Idaho 83402
Project Officer
William G. Kepner
Exposure Assessment Research Division
Environmental Monitoring Systems Laboratory
Las Vegas, Nevada 89193-3478
Environmental Monitoring Systems Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Las Vegas, Nevada 89193-3478
U.S. Environmental Prrcdion Agency ^
Region 5, j ibi'ffv /r'.' -."'•''•• l$X$> Printed on Recycled Paper
77 VVc-st Jacbon '.:<•.
Chicago, IL 60SiX-,: .,
-------
ENVIRONMENTAL MONITORING AND ASSESSMENT PROGRAM
Arid Ecosystems Resource Group
William G. Kepner, Technical Director
Carl A. Fox, Principal Scientist
EMAP-Arid Colorado Plateau Pilot Study- 1992
Implementation Plan
Susan E. Franson, Editor
CONTRIBUTORS
Robert P. Breckenridge, Idaho National Engineering Laboratory (INEL); Roger Clark, Grand
Canyon Trust; Carl A. Fox, Desert Research Institute (DRI); Douglas'G. Fox, U.S. Forest
Service (FS); Susan E. Franson, U.S. Environmental Protection Agency (EPA); Harold C.
Fritts, University of Arizona (U of AZ); Nancy L Hampton, INEL; William G. Kepner, EPA;
Robert O. Kuehl, U of AZ; Stephen G. Leonard, Bureau of Land Management; Richard D.
McArthur, DRI; Vern Meentemeyer, University of Georgia; Timothy B. Minor, DRI; David A.
Mouat, DRI; Martin R. Rose, DRI; George J. Staidl, Soil Conservation Service; Nita G.
Tallent-Halsell, Lockheed Engineering and Science Corporation (LESC); Robin J. Tausch, FS;
Carol B. Thompson, DRI; Richard D. van Remortel, LESC; James D. Wickham, Bionetics
Corporation; and Peter E. Wigand, DRI
NOTICE
The information in this document has been funded in part by the U.S. Environmental Protection
Agency through Contract #68-CO-0049 to Lockheed Engineering and Science Company, Cooperative
Agreement #CR-816385-02 to the Desert Research Institute of the University and Community College
System of Nevada, Interagency Agreement #DW 89934398 to the Department of Energy (Idaho National
Engineering Laboratory), Interagency Agreement #DW 14935509-01-0 to the Bureau of Land
Management, Interagency Agreement #DW 12935623-01-0 to the Soil Conservation Service and
Purchase Order#2V-0489-NAEX to the University of Arizona. It has been subject to the Agency's peer
and administrative review, and it has been approved for publication as an EPA document.
Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
Proper citation of this document is:
Franson, S.E., ed. 1992. Environmental Monitoring and Assessment Program: EMAP-Arid
Colorado Plateau Pilot Study - 1992: Implementation Plan. EPA/600/7-92/XXXX. U.S.
Environmental Protection Agency, Washington, DC.
-------
TABLE OF CONTENTS
NOTICE ii
ACRONYMS vi
LIST OF FIGURES viii
LIST OF TABLES ix
ACKNOWLEDGEMENTS x
1. 0 INTRODUCTION TO THE EMAP-ARID COLORADO PLATEAU
PILOT STUDY 1
1.1 INTRODUCTION 1
1.2 ARID ECOSYSTEMS 2
1.3 LAYOUT AND OVERVIEW OF THE IMPLEMENTATION PLAN 9
1.4 REFERENCES 11
2.0 CONCEPTUAL APPROACH 12
2.1 INTRODUCTION 12
2.2 STUDY OBJECTIVES AND QUESTIONS 13
2.3 REFERENCES 17
3.0 SITE SELECTION AND DESCRIPTION OF THE STUDY AREA 18
3.1 THE SITE SELECTION PROCESS 18
3.1.1 Introduction 18
3.1.2 K-T Analysis 19
3.2 DESCRIPTION OF THE COLORADO PLATEAU 22
3.3 LOCATION OF THE 1992 COLORADO PLATEAU PILOT STUDY 27
3.4 REFERENCES 28
4.0 DESIGN 30
4.1 EMAP DESIGN OVERVIEW 30
4.2 EMAP-ARID DESIGN OVERVIEW 31
4.2.1 EMAP-Arid Population and Subpopulations 31
4.2.2 EMAP-Arid Frame and Extent Estimation 33
4.3 EMAP-ARID PILOT STUDY SUBPOPULATIONS AND DESIGN 33
4.3.1 Pilot Study Subpopulations 33
4.3.2 Great Basin Desertscrub 33
4.3.3 Great Basin Conifer Woodland 35
4.3.4 Pilot Study Design 36
4.4 PLOT DESIGN FOR MEASURING INDICATORS 37
4.5 REFERENCES 41
5.0 INDICATORS 42
5.1 INTRODUCTION 42
5.1.1 Assessment Endpoints and Indicators 42
5.1.2 Selection of Potential EMAP-Arid Indicators 47
5.1.3 Selection of 1992 Pilot Indicators 49
5.2 SPECTRAL PROPERTIES INDICATORS 53
5.2.1 Introduction 53
in
-------
5.2.2 Relationships Between Remote Sensing Measurements and
Ecological Variables 55
5.2.3 Image Acquisition and Remote Sensing Measurements 58
5.2.4 Details for Specific Spectral Property Indicators 59
5.2.5 Ground-Based Measurements of Spectral Properties 59
5.3 VEGETATION COMPOSITION, STRUCTURE, AND ABUNDANCE
INDICATORS 62
5.3.1 Introduction 62
5.3.2 Details for Specific Indicators 63
5.3.3 Sampling Design 64
5.4 SOIL PROPERTIES INDICATORS 69
5.4.1 Introduction 69
5.4.2 Soil Properties Indicators 70
5.4.3 Erosion Index Indicator 76
5.4.4 Data Sources and Additional Data Requirements for Soils
Indicators 80
5.5 REFERENCES 82
6.0 LOGISTICS 85
6.1 DESIGN CONSIDERATIONS 85
6.2 SAMPLE SITE ACTIVITIES 87
6.3 REFERENCES 89
7.0 QUALITY ASSURANCE 90
7.1 INTRODUCTION 90
7.2 APPROACH TO QUALITY ASSURANCE 90
7.3 DATA QUALITY OBJECTIVES 91
7.4 QUALITY ASSURANCE OBJECTIVES 92
7.5 QUALITY ASSURANCE DURING THE 1992 PILOT 92
7.5.1 Training 92
7.5.2 Site Location QA 93
7.5.3 Vegetation Composition, Structure, and Abundance QA 93
7.5.4 Spectral Properties QA 94
7.5.5 Soil Properties QA 94
7.5.6 Field Audits 96
7.5.7 Pilot Study QA Reports 96
7.6 REFERENCES 96
8.0 INFORMATION MANAGEMENT AND CIS 97
8.1 ROLE OF INFORMATION MANAGEMENT 97
8.2 OPERATIONAL ASSUMPTIONS 98
8.3 OVERVIEW OF INFORMATION MANAGEMENT FUNCTIONS 98
8.4 PRE-FIELD FUNCTIONS 99
8.5 FIELD FUNCTIONS 100
8.6 CENTRAL OFFICE FUNCTIONS 101
8.7 EXTERNAL DATA SETS 102
8.8 ASSIMILATION, REVIEW, AND ASSESSMENT 102
9.0 ANALYSIS AND REPORTING 103
9.1 INTRODUCTION 103
IV
-------
9.2 QUESTIONS THAT THE PILOT IS DESIGNED TO ANSWER 103
9.3 REPORTING 112
9.4 REFERENCES 113
APPENDIX A. CANDIDATE INDICATORS 114
1.0 INTRODUCTION 114
2.0 LANDSCAPE INDICATORS 114
2.1 Introduction 114
2.1.1 Levels of Constraint 116
2.2 Landscape Indicators in EMAP-Arid 116
2.2.1 Introduction 116
2.2.2 Habitat/Cover Type Proportion 117
2.2.3 Spatial Distribution of Agricultural and Riparian Vegetation
Per Stream Reach 118
2.2.4 Fractal Dimension 119
2.2.5 Abundance/Density of Key Physical Features 120
2.2.6 Spatial Distribution of Grazing Intensity 121
2.2.7 Riparian Condition 122
3.0 RETROSPECTIVE HISTORY 124
3.1 Introduction 124
3.2 Retrospective Indicators 125
3.2.1 Tree-ring Series 125
3.2.2 Meteorological Data 132
3.2.3 Pollen Record 133
3.2.4 Packrat Middens 133
3.2.5 Fossil Charcoal Record 134
3.2.6 Stable Isotopes, Fossil Woodrat Midden Materials, and
Tree Ring 135
3.2.7 Repeat Photography as a Retrospective Indicator 136
3.3 Data Sources 137
3.3.1 Tree-ring Series Data Sources 138
3.3.2 Packrat Midden Data Sources 138
3.3.3 Pollen Data Sources 138
3.4 Data Analysis 138
3.4.1 Tree-ring Series Data Analysis 140
3.4.2 Pollen Data Analysis 141
3.4.3 Packrat Middens Data Analysis 143
3.4.4 Fossil Charcoal Records Data Analysis 144
3.4.5 Stable Isotopes, Fossil Woodrat Midden Materials, and
Tree Ring Data Analysis 144
3.5 Conclusions 145
4.0 REFERENCES 147
-------
ACRONYMS
AVHRR Advanced Very High Resolution Radiometer
BLM Bureau of Land Management
CEC Cation Exchange Capacity
CPPS Colorado Plateau Pilot Study
DQO Data Quality Objective
DRI Desert Research Institute (University and Community College
System of Nevada)
EMAP Environmental Monitoring and Assessment Program
EMAP-Arid Arid Ecosystem Component of EMAP
EOSAT Earth Observation Satellite Corporation
EPA Environmental Protection Agency
EROS Earth Resources Observation Satellite
ESP Exchangeable Sodium Percentage (solid indicate)
FIA Forest Inventory Analysis
FOTM Field Operations and Training Manual
FS Forest Service
FWS Fish and Wildlife Service
GAP FWS GAP Program (to identify gaps in protection of species and
habitats)
GIS Geographic Information System
GPS Global Positioning System
HLAS Habitat Linear Appraisal System
IM Information Management
INEL Idaho National Engineering Laboratory
K-T Kepner-Tregoe (Decision Analysis Technique)
LAI Leaf Area Index
LESC Lockheed Engineering and Science Company
LTER Long-term Ecological Research
MQO Measurement Quality Objective
MSS Multi-Spectral Scanner
NALC North American Landscape Characterization
NCSS National Cooperative Soil Survey
NDVI Normalized Difference Vegetation Index
NGDC National Geophysical Data Center (Boulder, CO)
NIR Near InfraRed
NOAA National Oceanic and Atmospheric Administration
NPP Net Primary Productivity
NPS National Park Service
VI
-------
NRC National Research Council
NRI National Resource Inventory
NSH National Soils Handbook
PC Personal Computer
PDR Personal Data Recorder
PDSI Palmer Drought Severity Index
PS-II Personal Spectrometer II
QA Quality Assurance
QAPjP Quality Assurance Project Plan
RUSLE Revised Universal Soil Loss Equation
SAR Sodium Absorption Ratio
SCS Soil Conservation Service
SSM Soil Survey Manual
TM Thematic Mapper
UCAR University Corporation for Atmospheric Research
USDA United States Department of Agriculture
USLE Universal Soil Loss Equation
WE Wind Erosion
WEPP Wind Erosion Prediction Project
VII
-------
LIST OF FIGURES
Figure 1-1. Aggregated Arid Ecoregions of the U.S 3
Figure 1-2. EMAP-Arid Biogeographic Provinces of North America 5
Figure 1-3. Arid Ecosystems Conceptual Model 6
Figure 2-1. The Elements of Designing and Implementing a Monitoring Program 13
Figure 2-2. Indicator Selection, Prioritization, and Evaluation Approach for EMAP. 14
Figure 3-1. Proposed Pilot Study Regions 20
Figure 3-2. Map of the Colorado Plateau 24
Figure 3-3. Potential Location of the Arid EMAP Sampling Points for 1992 Pilot 29
Figure 4-1. EMAP-Arid Sample Plot Design 38
Figure 4-2. EMAP-Arid Sample Plot Conceptual Hectare 40
Figure 5-1. EMAP-Arid Conceptual Model 43
Figure 5-2. Indicator Selection, Prioritization, and Evaluation Approach for EMAP. 48
Figure 5-3. Typical Spectral Response Curves 54
Figure 5-4. Relationship Between LAI and Two Vegetation Indices 57
Figure 5-5. Positions of Spectral Measurements 61
Figure 5-6. Nested Quadrat Design 65
Figure 5-7. Surface Type Classes 67
Figure 5-8. Conceptual Model 71
Figure 6-1. Flow Chart of Daily Activities 88
VIII
-------
LIST OF TABLES
Table 3-1. Criteria ("Wants") for EMAP-Arid Pilot Study Site Selection 21
Table 3-2. Study Area Selection for EMAP-Arid FY92 Pilot via K-T Decision
Analysis 22
Table 5-1. Association Between EMAP-Arid Assessment Endpoints and Societal
Values 44
Table 5-2. Association of Indicators and Their Types with Assessment Endpoints 45
Table 5-3. List of "Musts" and "Wants" for Selection of Candidate Indicators for
1992 EMAP-Arid Pilot 50
Table 5-4. Modified Daubenmire Cover Classes 63
Table 5-5. List of Soil Indicators and Associated Measurements for Soil Quality
Assessment Endpoints and Their Importance to Societal Values 72
Table 5-6. Estimates of Costs and Labor Requirements for Soil Properties Indicators
for 92 Pilot Study 77
Table 6-1. Responsibilities of Field Crew Members 86
Table A-1. Levels of Constraint for a Spatial Hierarchy of Ecosystems 115
IX
-------
ACKNOWLEDGEMENTS
The EMAP-Arid Indicator Workshop was held at Utah State Universtity
October 28-30, 1991 to evaluate indicators of condition of arid ecosystems. As a
result of that workshop a broad list of proposed candidate indicators was organized
into a more cohesive grouping of indicator categories. The discussions during that
workshop have been extremely valuable in the selection of indicator categories for
the pilot study and to the development of this Implementation Plan and are greatly
appreciated. Participants in that workshop included: Timothy Ball, Desert Research
Institute, University and Community College System of Nevada (DRI); Terrence P.
Boyle, National Park Service, Colorado State University; Robert P. Breckenridge,
Idaho National Engineering Laboratory (INEL); Carl A. Fox, DRI; Douglas G. Fox,
U.S. Forest Service Rocky Mountain Range Experiment Station; Susan E. Franson,
Environmental Monitoring Systems Laboratory-Las Vegas (EMSL-LV), U.S.
Environmental Protection Agency (EPA); Harold C. Fritts, Laboratory of Tree-Ring
Research, University of Arizona; Nancy L Hampton, INEL; Carolyn T. Hunsaker,
Oak Ridge National Laboratory; Wes Jarrell, Department of Environmental Science
and Engineering, Oregon Graduate Institute; Dale Johnson, DRI; K. Bruce Jones,
EMSL-LV, EPA; William G. Kepner, EMSL-LV, EPA; Robert O. Kuehl, College of
Agriculture, University of Arizona; Stephen G. Leonard, Department of Range,
Wildlife, and Forestry, Bureau of Land Management; James A. McMahon,
Department of Biology, Utah State University; Vern Meentemeyer, Department of
Geography, University of Georgia; David A. Mouat, DRI; Keith Mussallem, DRI;
Martin R. Rose, DRI; Carol Simmons, Colorado State University; Stan Smith,
Biology Department, University of Nevada - Las Vegas; George J. Staidl, Soil
Scientist, National Soil Range Team, Soil Conservation Service; Don Stevens,
Ecological Research Organization, Mantech Environmental Technology, Inc.; Robin
J. Tausch, USDA Forest Service Intermountain Research Station; Richard D. van
Remortel, Lockheed Engineering and Sciences Corporation; Fred Wagner,
Department of Range Science and Ecology Center, Utah State University; Neil E.
West, Department of Range Science and Ecology Center, Utah State University;
Walt G. Whitford, Department of Biology, New Mexico State University; James D.
Wickham, Bionetics Corporation; Peter E. Wigand, DRI.
The preparation of this plan has been a combined effort requiring the
contributions of a number of scientists from various universities, research institutes,
public interest groups, and federal agencies. This manuscript has benefited from
the comments of many outside reviewers: James A. McMahon, Department of
Biology, Utah State University; Duncan T. Patten, Center for Environmental Studies,
Arizona State University; Anthony J. Krzysik, Construction Engineering Research
Laboratory, U.S. Army Corps of Engineers; Anthony Olsen, Technical Coordinator
EMAP Design and Statistics, U.S. EPA Environmental Research Laboratory-
Corvallis; Richard E. Francis, U.S. Forest Service, Rocky Mountain Forest and
Range Experiment Station; Marvin LeNoue, Service Center Director, Bureau of
-------
Land Management; and K. Bruce Jones, EMAP Associate Director for Terrestrial
Ecosystems, U.S. EPA Environmental Monitoring System Laboratory - Las Vegas.
In addition, many of the contributors reviewed the entire manuscript: William G.
Kepner, Robert O. Kuehl, Steven G. Leonard, Richard D. McArthur, George J.
Staidl, Carol B. Thompson, and James D. Wickham. Appreciation also goes to
Julie K. Muhilly, Barbie Nauroth, and Debbie Wilson of the Desert Research Institute
(DRI) for document processing and to Glenda Mahin, also of DRI, for technical
editing.
XI
-------
1. 0 INTRODUCTION TO THE EMAP-ARID
COLORADO PLATEAU PILOT STUDY
(William G. Kepner)
1.1 INTRODUCTION
In response to the growing awareness of regional and global-scale environmental
degradation brought about by the combined actions of all peoples on Earth, nations
throughout the world are acknowledging the need to obtain critical scientific information
and are establishing environmental monitoring networks to assess the condition of their
important ecological resources.
The U.S. Environmental Protection Agency (EPA), in collaboration with other federal
agencies, research institutes, and university systems, has initiated the Environmental
Monitoring and Assessment Program (EMAP) to develop a long-term approach to assess
and periodically document the condition of ecological resources at regional and national
scales and to create innovative methods for anticipating emerging problems before they
reach crisis proportions. The goals of EMAP are to:
1. Monitor and report on the condition of the Nation's ecological resources.
2. Evaluate the effectiveness of the sum total of current environmental
policies and programs.
3. Identify emerging environmental problems before they become
widespread or irreversible.
To achieve these goals, EMAP will: (1) estimate the status, extent, changes, and
trends in ecological condition using an environmental indicator strategy; (2) seek
associations between human-induced stresses and ecological condition; and (3) provide
statistical summaries and interpretive reports on ecological status and trends to resource
managers and the public. The program is focused on linking with existing environmental
monitoring programs, where possible, and collecting new information as needed to achieve
1
-------
its objectives. EMAP is not intended as a substitute for ongoing programs, however, it may
enhance their value by placing local monitoring results in the perspective of a larger
geographical context.
To accomplish its goals and objectives, EMAP has established seven ecosystem
monitoring and research groups (i.e., estuarine and marine; Great Lakes; surface waters;
wetlands; forests; agroecosystems; and arid ecosystems) and seven cross-system
program groups (i.e., design and statistics; quality assurance; information management;
landscape characterization; indicators; logistics; and integration and assessment).
1.2 ARID ECOSYSTEMS
Arid ecosystems, as defined by EMAP, are 'Terrestrial systems characterized by a
climatic regime where potential evapotranspiration exceeds precipitation, annual
precipitation ranges from <5 to 60 cm, and daily and seasonal temperatures range from
-40 to 50° C. The vegetation in arid ecosystems is dominated by woody perennials,
graminoids, succulents, and drought-resistant trees. Physiognomy is generally low-form
and canopies typically open. Arid ecosystems include associated riparian communities,
however, intensively managed agriculture, such as irrigated farmland, is excluded even
though it may occur in the same climatic region" (Kepner and Fox, 1991). Arid ecosystems
in the United States occupy nearly all the land surface area (excluding high-elevation
forests) west of 95°W longitude (Figure 1-1). Historically, dramatic urbanization and
overexploitation of natural resources have resulted in rapid desertification, i.e., the decline
or loss of biotic productivity in arid and semi-arid lands due to certain natural phenomena
and man-induced stresses (Bender, 1982). Once arid ecosystems are degraded
significantly, they are generally unlikely to return to their prior state and hence are often
termed "fragile"because they exhibit little resistance or resilience in the face of
anthropogenic insult (UCAR, 1991). Desertification, livestock grazing, biodiversity, water
resource management, air quality, and global climatic change have been identified as
regionally important issues in arid ecosystems (Kepner and Fox, 1991).
-------
-------
The objectives of the EMAP Arid Ecosystems resource monitoring and research group
(EMAP-Arid) parallel those established for the overall EMAP program. When fully
implemented, EMAP-Arid will address the following objectives related specifically to arid
systems:
1. Measure the status and trends and estimate the extent of arid
ecosystems using synoptic, retrospective, and sample-based indicator
measurements.
2. Determine the spatial and temporal correlation between environmental
stressors and ecological condition.
3. Provide information to decision/policy makers and management, and
regulatory and research agencies and institutes that can be utilized for
comprehensive regional planning and management.
4. Develop a regional interagency communication and data transfer
network.
It is the intent and purpose of EMAP-Arid to measure and report on the extent,
condition, and trends in eight vegetation formation types (i.e., biomes) within the
conterminous U.S. portion of the seven biogeographical provinces of Nearctic and
Neotropical North America that reside in an arid or semi-arid climatic regime (Brown et al.,
1979; Figure 1-2). These include five upland formations (desertscrub, grassland,
scrubland, woodland, and tundra) and three lowland formations (riparian forest, riparian
scrub, and strandland). Under full implementation, EMAP-Arid will also include the tundra
of Alaska.
EMAP-Arid will utilize a set of environmental indicators that collectively can describe
the condition of an ecosystem (Hunsaker et al., 1990). The operating strategy is to identify
regional issues and critical questions; link them with ecological endpoints that have both
social and biological relevance; and identify indicators derived from conceptual models
that, when measured and integrated, can evaluate the status and trends in the condition of
arid ecosystems (Figure 1-3). Three societal values are currently identified as significant
to arid ecosystems and have served to focus the conceptual development of the monitoring
-------
D Great Basin
D Mohavian
E3 Sonoran
Californian
Chihuahuan
Mogollon
Plains
Figure 1-2. EMAP-Arid Biogeographic Provinces of North America.
-------
Figure 1-3. Arid Ecosystems Conceptual Model
-------
and research strategy for EMAP-Arid, especially relative to its selection and use of
indicators:
1. Sustainability
2. Biodiversity
3. Aesthetics
Sustainability is the ability of an ecosystem to sustain its potential or actual biological
productivity over the long term. Within arid and semi-arid lands, loss of Sustainability is
usually the result of desertification. Desertification can result from an extended period of
extreme drought, severe mismanagement of land, or evaporation of water leading to soil
salinization. Desertification is accompanied by a change in species composition or general
loss of biomass, loss of soil nutrients with remaining nutrients concentrated under shrubs,
and increased removal of soil materials by wind and water erosion - all of which contribute
to a loss of Sustainability of the ecosystem.
Biodiversity (species richness) recently has been recognized as an important global
resource. The preservation of species, communities, and ecosystems to provide natural
resources (e.g., food, medicine, shelter) for daily life and ecological services (e.g., climate
moderation, water and nutrient cycling, and breakdown of wastes) now and into the future
has become an issue of global proportions. In essence, to ensure survival of life on Earth
we must protect biodiversity.
Aesthetics as a societal value of EMAP-Arid can be broadly defined as the quality of
life. It is related to the above values but focuses on the human perception of the
ecosystem. Many people value arid systems not only for their ecological role but also for
their beauty: the desert in bloom, mountain sheep on rocky crags, the spectacular views of
the Painted Desert or Grand Canyon. Aesthetics is the societal value that ties all the others
together. Without Sustainability and biodiversity the beauty of the desert would be lost.
Of course the quality of life also depends on air quality, water quality and quantity, and
many other things. As the EMAP-Arid develops, other such societal values will be
-------
identified, along with assessment endpoints and indicators to monitor their status and
trends.
The framework for selecting and testing indicators follows a process outlined in
Hunsaker and Carpenter (1990). Each EMAP Resource Group is expected to select and
test a number of indicators in limited field tests or pilot exercises. These tests are intended
to evaluate the ability of selected research indicators to discriminate, separately and in
combination, environmental conditions and ultimately to determine which indicators are
retained and moved to a higher indicator category, rejected, or held for further evaluation. If
indicators are retained, they will be further tested on a regional scale via a demonstration
project. A final set of core indicators will be selected for long-term implementation based
on the results of regional demonstration projects and external peer review (Hunsaker and
Carpenter, 1990). New or improved versions of indicators can be added to the core set
following periodic revaluation and testing of indicator performance.
This document presents a proposed plan for pilot testing three indicator categories of
arid ecosystem condition (spectral properties; vegetation composition, structure, and
abundance; and soil properties). These indicators were selected through a number of
workshops and peer reviews, and are likely to meet all the criteria, such as being
applicable and interpretable on a regional scale, suggested by Hunsaker and Carpenter
(1990). Thus, they are of high priority in EMAP-Arid. Although these indicators appear to
demonstrate the highest potential or capability for diagnosing ecosystem change (i.e., the
ability to be merged with other data sets to make integrated assessments of ecosystem
condition at the regional level), they must be considered developmental in status and
subject to field testing prior to their long-term implementation.
The purpose of this study is to focus on answering important questions of indicator
performance such as determining components of variance. Other important information
such as requirements for methods development, logistics, data management, and quality
8
-------
assurance will also be gleaned from this type of study. This study is not intended to provide
a regional estimate of condition.
The preparation of this plan has been a team effort of a number of scientists from
various universities, research institutes, public interest groups, and federal agencies. The
success of the pilot study is equally dependent on the participation of this mixture of
affiliations, but particularly rests with the EPA, U.S. Bureau of Land Management (BLM),
National Park Service (NPS), Forest Service (FS), Soil Conservation Service (SCS), Fish
and Wildlife Service (FWS), and Navajo and Ute Nations.
In summary, the Implementation Plan for the Colorado Plateau Pilot Study is a
significant first step towards regional and national implementation of EMAP-Arid. This plan
provides the mechanism for coordination of indicator development and evaluation with
members from participating agencies and the external scientific community via the planning
and peer review process. Additionally, it provides the foundation and direction to team
members who are responsible for executing the plan. We fully anticipate that a number of
pilot and demonstration projects will be required in the next couple of years prior to
achieving full implementation. We also are confident that the selected indicator suite and
the location of the Colorado Plateau for this pilot affords us every opportunity to achieve a
success upon which to further develop this program.
1.3 LAYOUT AND OVERVIEW OF THE IMPLEMENTATION PLAN
This Implementation Plan gives an overview of the pilot study from a technical
perspective. A companion document, the Field Operations and Training Manual (FOTM)
(Franson and Pollard, 1992), presents the operational aspects of the field study. These
operational aspects include (1) detailed protocols for each step of the field work; (2) a
Safety Plan that documents the hazards that may be encountered in the study area, safety
procedures to be followed, and information about what to do in case of emergency; and (3)
a Quality Assurance Project Plan that details the steps to be taken to ensure that data
collected are of sufficient quality to address the objectives of the study. The indicator
-------
evaluation pilot study includes as an objective the development and evaluation of
measurement protocols and quality assurance (QA) procedures. Thus, the applications of
the operational details in the field data collection effort will serve as part of the review of
these operational plans.
Following this Introductory Section, the Implementation Plan addresses the
Conceptual Approach for the pilot study in Section 2. Questions that will be addressed in
order to meet each study objective are outlined as part of this Conceptual Approach.
Section 3 describes the site selection process that resulted in the choice of the Colorado
Plateau for this pilot study. A description of the entire Colorado Plateau is included in
Section 3, along with the exact geographic location of the pilot study. Section 4 gives an
overview of the EMAP design, the overall EMAP-Arid design, and the sampling and plot
design for the pilot study.
Section 5 discusses the indicators, including a rationale for their selection and use
and a general description of the measurements to be made. Specific protocols for
collecting the indicator data will appear in the FOTM. The intent of Section 5 is to give
sufficient information to review the selected indicators from a scientific standpoint, but not
to burden the reader with extensive details of the field protocols.
In a similar fashion, Section 6 on Logistics, Section 7 on Quality Assurance, and
Section 8 on Information Management and GIS present overviews of these topics and are
not intended to give the specific details required for the field activities that will appear in the
FOTM.
Section 9 concludes the Implementation Plan with a discussion of Analysis and
Reporting of the results of this pilot study. As with any exploratory research, the exact
analyses that will be employed are not always known before the data are in hand. Rather,
the approach to data analysis is to address the questions given in the Conceptual
Approach.
10
-------
1.4 REFERENCES
Bender, G.L., ed. 1982. Reference handbook on the deserts of North America. Greenwood
Press, Westport, Connecticut.
Brown, D.E., C.H. Lowe, and C.P. Pase. 1979. A digitized classification system for the
biotic communities of North America, with community (series) and association
examples for the Southwest. Journal of the Arizona-Nevada Academy of Science 14
(Supplement 1):1-16.
Franson, S.E. and J.E. Pollard, eds. 1992. Environmental Monitoring and Assessment
Program: Arid ecosystems 1992 indicator pilot study. Colorado Plateau: field
operations and training manual. U.S. Environmental Protection Agency, Washington,
DC. (In press).
Hunsaker, C.T., D.E. Carpenter, and J.J. Messer. 1990. Ecological indicators for regional
monitoring. Bulletin of the Ecological Society of America 71 (3):165-172.
Hunsaker, C.T. and D.E. Carpenter, eds. 1990. Ecological indicators for the Environmental
Monitoring and Assessment Program. EPA 600/3-90/060. U.S. Environmental
Protection Agency, Office of Research and Development, Research Triangle Park,
North Carolina, 416 pp.
Kepner, W.G., and C.A. Fox, eds. 1991. Environmental Monitoring and Assessment
Program. Arid ecosystems strategic monitoring plan. EPA 600/4-91/018. U.S.
Environmental Protection Agency, Washington, DC.
University Corporation for Atmospheric Research. 1991. Arid ecosystem interactions:
recommendations for drylands research in the Global Change Research Program.
Office for Interdisciplinary Earth Studies, Boulder, Colorado.
11
-------
2.0 CONCEPTUAL APPROACH
(Carl A. Fox, William G. Kepner, and Susan E. Franson)
2.1 INTRODUCTION
In 1991, EMAP-Arid developed a monitoring strategy for full scale implementation of
EMAP across arid ecosystems (Kepner and Fox, 1991). This strategic plan was based in
part on guidelines developed by the National Research Council (NRC, 1990) for designing
and implementing environmental monitoring programs (Figure 2-1). The NRC process
provides a formula that leads from defining goals to disseminating information to decision
makers. EMAP-Arid is closely following the NRC strategy and has completed Step 2 of this
process.
The strategic plan was evaluated through a peer review process. The expectations,
goals, and strategy of EMAP-Arid presented in the plan were approved by the review
panel, with the recommendation that the strategy be evaluated through field evaluations
and pilot studies, that is, to begin Step 3 of the NRC process.
Through the development and execution of the Colorado Plateau Pilot Study,
EMAP-Arid will begin Step 3, Conduct Exploratory Studies. These exploratory studies are
generally of two types: pilot studies and demonstration projects. Pilot studies are generally
intended to answer specific questions about indicator performance, including sensitivity,
components of variance, methods, and logistical requirements. Pilot studies are not
intended to provide regional estimates of ecological condition. Demonstration projects,
while addressing many of the same questions as pilot studies, are specifically designed to
demonstrate EMAP's ability to estimate the condition of regional populations.
The Colorado Plateau Pilot Study will be conducted to evaluate and field test a
number of issues related to design, ecological indicators, quality assurance, logistics,
information management, and analysis and reporting before full scale implementation. The
12
-------
(Source: NRC1990)
Stepl
Define
Expectations and Goals
Rethink
Monitoring
Approach
Step 2
Define
Study Strategy
Step 4
Develop
Sampling Design
Step3
Conduct Exploratory |
Studies if Needed
Refine
Objectives
Can
Changes Be
Detected
StepS
Implement Study
Make Decisions
Step 6 E
Produce Information!
Is Information
Adequate?
Step 7
Disseminate
Information
Figure 2-1. The Elements of Designing and Implementing a Monitoring Program.
Colorado Plateau Pilot Study will test nearly all aspects of the monitoring program with a
limited suite of indicators. Results will be used to plan future pilot studies and to develop
regional demonstration projects leading to full scale implementation.
2.2 STUDY OBJECTIVES AND QUESTIONS
The overall goal of the Colorado Plateau Pilot Study (CPPS) is to evaluate a selected
subset of ecological indicators to address issues of desertification and global climatic
13
-------
change as they relate to sustainability of arid ecosystems, a principal societal value. While
this goal will not be achieved in this study, it serves to focus EMAP-Arid activities and
guide the CPPS.
Specific objectives and associated questions that will be addressed in the pilot study
are as follows:
Objective 1: To gather and evaluate information to move selected ecological indicators
from the "research" category to the "development" stage in the indicator implementation
process (Figure 2-2).
CANDIDATE INDICATORS
(Expert Knowledge
Literature Review
Peer Review
RESEARCH INDICATORS
Analysis of Existing Data
EVALUATE EXPECTED PERFORMANCE Sld-Sc'a.e Re.d Tests
Peer Review
DEVELOPMENT OF INDICATORS
Regional Demonstration Projects
EVALUATE ACTUAL PERFORMANCE ' Peer Review
CORE INDICATORS
IMPLEMENT REGIONAL AND NATIONAL MONITORING
PERIODIC REVALUATION
Figure 2-2. Indicator Selection, Prioritization, and Evaluation Approach for EMAP.
14
-------
Questions:
1. Are the indicators for a) spectral properties, b) vegetation composition,
structure, and abundance, and c) soil properties, separately or in
combination, correlated with independent evaluations of site conditions?
BLM, NFS, SCS, and FS all have condition assessments for some arid
lands. While these assessments are based on different sets of criteria,
EMAP-Arid will compare pilot study results with other condition
assessments to begin to develop the assessment process.
2. What is the correlation between the remotely sensed spectral properties
data from AVHRR, MSS, or TM and spectral properties data acquired
through field sampling using a personal spectrometer?
3. Do remote measures of spectral properties correlate with other field
measured indicators of vegetation and soils or existing assessments
done by the NFS, FS, SCS, and BLM?
4. Which of the remote platforms (AVHRR, MSS, or TM, or a combination)
appears to be most effective in obtaining the required spatial and
temporal data necessary to link remotely sensed indicators with ground
measures arid existing data?
Objective 2: Evaluate the utility of using classified Thematic Mapper imagery and other
data acquired from the FWS GAP Program to select frame materials for the pilot study and
future studies and to provide data for extent estimation of arid ecosystems.
Questions:
5. Do the Biotic Communities Map (Reichenbacher and Brown, 1992) and
the GAP data correctly identify the plant communities found at each of
the pilot study sample grid points? If not, what is the level of
misclassification and can this level of misclassification be compensated?
6. Do the GAP data provide adequate information to describe the extent of
arid ecosystems in the pilot study area?
Objective 3: Evaluate sampling plot designs appropriate to the selected indicators.
Questions:
7. What are the sampling design between site, subplot, and sample
variance components of each of the selected indicators?
8. What are the costs associated with indicator measurement? Costs to be
evaluated include labor, equipment, laboratory analyses, image
15
-------
analyses, cost of imagery, data analyses, etc. Costs will be evaluated
relative to specific sizes of sampling units (subplots, samples, lab
replicates, etc.) as well as overhead costs for a site.
9. What are the optimum numbers of subplots and samples to have a good
estimate of each indicator at a site?
10. How many sites cross a vegetation/soil complex boundary? Does the
addition of quadrats provide a large enough sample to allow for
estimates of the vegetation indicators?
Objective 4: Evaluate the logistical, quality assurance, information management, data
analysis, and reporting requirements and constraints based on the pilot study data.
Questions:
11. What specific logistical constraints restrict the implementation of each
indicator? What logistic attributes favor or enhance indicator
measurement (e.g., use of a helicopter)?
12. Based on the results of the pilot study, can data quality objectives be
established for each indicator tested?
13. Does the information management system effectively and efficiently
provide for the movement of data from the field to the analysis stage?
14. Do the methods of collecting, transferring, and analyzing cata meet the
reporting requirements for an EMAP pilot study?
15. What are the special logistical requirements involved with fielding
multi-agency sampling crews?
The Colorado Plateau Pilot Study is not intended to be a full implementation of the
EMAP-Arid monitoring program but will provide information essential to the successful
development of regional demonstration projects. The pilot study will be an interagency
effort to evaluate selected indicators, sampling plot designs, logistics, quality assurance,
information management, and analysis and reporting. It represents EMAP-Arid's first field
study after the successful development of a strategic monitoring plan. The Pilot Study is
designed to fully consider those issues critical for the success and implementation of the
EMAP-Arid program.
16
-------
2.3 REFERENCES
Kepner, W.G. and C.A. Fox, eds. 1991. Environmental Monitoring and Assessment
Program. Arid ecosystems strategic monitoring plan. EPA 600/4-91/018. U.S.
Environmental Protection Agency, Washington, DC.
NRC. 1990. Managing Troubled Waters: The role of marine environmental monitoring.
National Academy Press, Washington, DC.
Reichenbacher, F.W. and D.E. Brown. 1992. Biotic communities of North America, Central
America, and the Islands of the Caribbean Sea. Map 1:8,000,000. U.S. Environmental
Protection Agency, Las Vegas, Nevada. (In press).
17
-------
3.0 SITE SELECTION AND DESCRIPTION
OF THE STUDY AREA
(Carl A. Fox, Robert P. Breckenridge, and Roger Clark)
3.1 THE SITE SELECTION PROCESS
3.1.1 Introduction
The most arid regions (i.e., deserts) on the earth are generally found within two
well-defined bands 20 to 30 degrees north and south of the equator and at the poles
(Bender, 1982). In North America, the major deserts (excluding the arct c) are the
Chihuahuan, Great Basin, Mojave, and Sonoran, all located in the western United States.
The Great Basin is considered a "cold" desert, while the others are often termed "hot"
deserts.
In selecting a regional focus for the 1992 pilot study, EMAP-Arid first looked at the
availability of relevant existing data and monitoring sites to assess sites located in "data
rich" areas. This was done to enhance the use of existing data for interpretation of
indicators and to foster interagency collaboration. Eight general areas were identified:
D Southwest New Mexico (Chihuahuan Desert)
n Great Basin National Park/Desert Experimental Range
n Northeast Utah and Southeast Idaho (Great Basin)
n Colorado Plateau (Great Basin)
D Southeast Arizona (Sonoran/Chihuahuan Deserts)
n Central California/Western Nevada (Western Great Basirt)
D Central Nevada (Great Basin/Mojave Deserts)
D Northwest Arizona (Mojave/Sonoran/Great Basin)
These were identified by evaluating the sources of data relevant to the indicators
identified in the EMAP-Arid Strategic Plan (Kepner and Fox, 1991) against a list of criteria
18
-------
that considered data quantity (number of years), data quality, site and data access, cost,
and multiple agency collaboration. From this list the top four areas were selected based on
their ability to meet the criteria. These four (Figure 3-1) became the final set from which a
pilot location was to be selected.
3.1.2 K-T Analysis
To insure that the selection of the study region for the pilot was done as objectively as
possible, the EMAP-Arid team employed a decision analysis procedure, the
Kepner-Tregoe (K-T) Analysis (Kepner and Tregoe, 1981). K-T Analysis is a decision
analysis technique that provides a quantitative methodology to insure that decisions are
made in a highly systematic and logical manner, but without inhibiting creativity and
innovation. The technique is focused around a decision statement that clearly defines the
desired outcome. A list of "must have" and "want" objectives (criteria) is developed. The
"want" objectives are weighted by giving the most important objective a score of 10 and all
others a score from 10 to 1 based on their importance relative to the most important
objective. The decision is then made by scoring each choice as to how well it meets the
wants and compiling a weighted score.
In the case of the EMAP-Arid pilot study site selection, the four candidates were
selected because they offered the opportunity of interagency collaboration and they fell
within the EMAP definition of what constitutes an arid ecosystem. Thus, there were no
musts in the K-T analysis. The "want" criteria and the respective weights developed by the
EMAP-Arid team that included representation from the BLM, FS, EPA, DRI, SCS, and
INEL are shown in Table 3-1. These criteria were then applied to the four candidate
regions. The regions were rated for how well they met each criterion, with 10 assigned for a
good match for the criterion and 1 assigned for a bad match for that criterion. The total
weighted scores were calculated for each region (Table 3-2). Based on this analysis, the
Colorado Plateau offered a study region that would best meet the selection criteria of the
pilot study.
19
-------
in
O
a
cc
o
s
Ul
<0
O
&
O
ffi
0.
Figure 3-1. Proposed Pilot Study Regions.
20
-------
TABLE 3-1. CRITERIA ("WANTS") FOR EMAP-ARID PILOT
STUDY SITE SELECTION
"WANT"
Weights
1. Relationship of Study Area to the Issue of Desertification 5.3
2. Relationship of Study Area to the Issue of Global
Climatic Change 4.9
3. Availability of Retrospective Data (Tree Ring Chronologies,
Historical Climate Network, Fossil Pollen Records, Packrat
Midden Radiocarbon Records) 4.9
4. Quality & Quantity of Historical (> 5 years) Data Related to
Sustainability 4.9
5. Opportunity for Collaboration with Other EMAP Groups 3.2
6. Availability of Classified Remote Sensing Imagery 2.8
7. Complexity of Study Area Logistics 1.9
Before a final decision was reached, however, the probability and seriousness of
adverse consequences (e.g., political, social, economic) in selecting the Colorado Plateau
were evaluated as the last step in the K-T Analysis. These adverse consequences were
rated high, medium, or low relative to the probability of occurrence and the seriousness of
their effect if they did occur. This part of the K-T Analysis focused strictly on the tentatively
selected site and did not allow for between site comparisons or ranking. It provided an
opportunity to review all the potential consequences and greatly improved the potential for
success. The end result of the final part of the K-T Analysis was that the Colorado Plateau
remained the region of choice for the pilot study.
21
-------
TABLE 3-2. STUDY AREA SELECTION FOR EMAP-ARID FY92 PILOT
VIA K-T DECISION ANALYSIS
SCORES FOR CANDIDATE SITES FOR CRITERIA*
Selection Criteria
"Wants"
1
2
3
4
5
6
7
. Desertification
. Global Climate Change
. Retrospective Data
. Historical Data (Sustainability)
. Collaboration within EMAP
. Classified Remote Imagery
. Logistics Complexity
Great
Basin
47.7
45.9
42.6
32.3
13.4
22.4
11.2
Colorado
Plateau
45.9
49.5
43.1
29.4
32.0
28.0
13.1
Sonoran
Desert
39.7
37.2
35.8
36.3
1.3
18.8
15.0
Chihuahuan
Desert
45.6
44.4
33.3
44.6
16.3
18.2
13.3
Total Weighted Score
215.5
241.0
184.1
215.7
* Scores reflect how the EMAP-Arid team rated the ability of the various sites to satisfy the
selection criteria from 10 (good match) to 1 (poor match). For example, the Great Basin received
a value of 9 for meeting desertification criterion (that was highest rated criterion with weight of
5.3); thus weighted score for Great Basin was 9 x 5.3 = 47.7.
3.2 DESCRIPTION OF THE COLORADO PLATEAU
The Colorado Plateau is an arid and semi-arid tableland in the American Southwest.
It is a place where climatic and geologic forces collide into monuments, mesas, canyons,
badlands, spires, arches, and landscapes unlike any other on our planet. It is a terrain of
sublime and stark beauty.
Life is harsh on the cold deserts of the Colorado Plateau. And yet it sustains alpine
tundra, hanging gardens, xeric woodlands, blackbrush and sage shrublands, and
cryptogamic communities of mosses, lichens, fungi, and cyanobacteria comprising most of
the living biomass on much of its otherwise sterile soil. It is also home to scores of unique
invertebrates, fish, reptiles, and amphibians.
22
-------
The Colorado Plateau was first delineated by Major John Wesley Powell shortly after
the American Civil War. Powell described the Plateau as a 170,000 square mile region
encompassing what is today western Colorado, northern Arizona, northwestern New
Mexico, the eastern two-thirds of Utah, and southwestern Wyoming. More recently,
geologists have reduced Powell's definition of the physiographic province to a 130,000
square mile area slightly larger than New England (Figure 3-2).
Major Powell was attracted to the Colorado Plateau for several reasons. It was the
last unmapped region of what was to become the lower 48 states; it was home to native
people who had inhabited the region for at least three centuries before European contact; it
revealed a unique panorama of the earth's geological history; and it held an enormous
potential of natural resources for furthering the progress of a young nation. And it was as
good a place as any to launch the career of a one-armed veteran, explorer, anthropologist,
geologist, topographer, botanist, paleontologist, hydrologist, and entrepreneur of his own
curiosity.
Today, the Colorado Plateau remains a remote region of undiscovered and forgotten
places. Slightly more than a million people live in dozens of small communities which are
concentrated along the Colorado River and its tributaries. Overall, its population density is
about seven persons per square mile, about one-tenth that of the rest of the nation.
The Plateau's traditional economic base has been ranching and mining and to a
lesser extent farming and logging. But by 1980, fewer than one in ten jobs was in
agriculture (including ranching), forestry, and mining. One-quarter of the region's residents
are employed in services related to tourism, recreation, and retirement. The second leading
employment sector is government, where one in five residents was working in 1987. About
15 percent of land on the Colorado Plateau is privately owned. Although the proportion of
private land is relatively small, the Plateau's sparse population makes the amount of
privately owned acres per resident considerably higher than in much more populous
23
-------
Colorado Plateau, FY92
Ints: w:
Figure 3-2. Map of the Colorado Plateau.
24
-------
regions. For example, California's amount of private land per person averages 2.1 acres,
while the Plateau's average is 16.9 acres.
Approximately 85 percent of the Colorado Plateau is under some form of government
jurisdiction. State lands comprise six percent of the land. The Bureau of Land Management
has jurisdiction over 29 percent of the Colorado Plateau; Indian tribal lands encompass 23
percent; and the U.S. Forest Service is responsible for 22 percent.
The National Park Service manages a mere four percent of the Colorado Plateau and
yet its 26 units attract over 30 million visitors each year. As is generally the case in these
rugged landscapes, park visitors are concentrated onto less than five percent of the land.
Consequently, most of the lands of the Colorado Plateau remain relatively unchanged by
direct human contact.
Nonetheless, direct and indirect impacts of human activities are widespread and
increasing. The National Park Service has concluded that in all Plateau parks, haze from
distant cities and nearby coal-fired generating stations reduces visibility during part of the
year. NPS has also measured biological impacts from air pollution in some of its remote
sampling sites. One indicator of the pervasive nature of domestic livestock on the Colorado
Plateau was provided when The Nature Conservancy could only find a few relic areas in all
of Utah that had not been grazed by cattle or sheep. Similarly, archaeological surveys of
remote sites in Glen Canyon National Recreation Area have documented extensive
livestock use and damage to prehistoric dwellings and artifacts.
Signs of human-induced change are most evident along the rivers of the Colorado
Plateau. Dams, reservoirs, and diversions have eliminated most of its native aquatic and
riparian habitat. Salinity in the Colorado River has increased as the result of upstream
mining and farming, and riparian habitat along the Colorado River in the Grand Canyon has
been completely altered by Glen Canyon Dam. Its reservoir, Lake Powell, has become a
trap for selenium, mercury, and other trace elements which have been found in bioassays
of introduced striped bass. Tamarisk, camelthorn, Russian olive and thistle, and many other
25
-------
introduced species now dominate the Plateau's riparian corridors. While the list of alarming
stories goes on, it is important to recognize that for all of the insults, the Plateau has
endured. And for all of its timeless and seemingly changeless beauty, evidence of change
is everywhere.
Despite its aridity, water is the primary force of change on the Colorado Plateau. One
reason is that most of the annual precipitation can occur during a single event. During the
late summer when moisture from the south accumulates into massive afternoon
thunderheads, periods of intense rainfall are scattered across the Plateau. In areas
characterized by thin soils and barren rocks, runoff from these high intensity storms is
quickly channeled into raging floods. Rockfalls, erosion, fresh sediments, and other signs
of major and frequent flooding are ubiquitous.
The geology of the Colorado Plateau reveals that erosion and deposition have been
shaping its landscape for hundreds of millions of years. The Plateau contains the most
voluminous, areally extensive, and continuous series of continental sediments in the world.
These sandstones, mudstones, and shales are the remnants of glaciers, streams, lakes,
marshes, mud flats, and dunes. Interspersed among the deeper continental sediments are
thick layers of marine limestones which date back over a billion years.
The Plateau's geology also reveals a dramatic history of climatic and ecological
change. Even in recent times where evidence has not had time to become fully fossilized,
researchers have documented that a cooler and much wetter climate supported quite a
different array of flora and fauna than is found there today. Prior to 11,000 years ago,
mammoths, musk oxen, ground sloths, and tapirs lived on the lush plant life of the
Colorado Plateau. Sixty-five million years before that, the region hosted the demise of the
great dinosaurs and the disappearance of now petrified forests.
The Plateau has a varying climate based upon elevation. In general, rainfall is low
(<50 cm/yr), with averages around 25 cm/yr. Most precipitation falls between October and
April as snow or rain. Due to high summer temperatures, evaporation usually exceeds
26
-------
precipitation on an annual average. Temperatures range from below 0°C in the winter to
over 50°C (122°F) in the summer. The warmest month is July with maximum temperatures
in excess of 50°C (USDA Soil Surveys, 1980a).
Soil is the most widely used natural resource on the Plateau. Soils on the Plateau
vary widely in their characteristics. Soils on strath terraces, alluvial fans, glacial outwash
fans, moraines, and talus slopes have a high content of rock fragments. The soils that
formed in aeolian (wind) deposits, alluvium (water) derived from sedimentary rock, and
shale landslide material have few rock fragments. The soils formed in recent aeolian
deposits commonly are sandy loam, loamy sand, or sand, while soils formed from shale
material are clay loam or clay. Deep soils are on mountainsides, alluvial fans, valley fills,
and gently sloping mesas, benches, and cuesta dip slopes. Shallow soils and exposed
sandstone are on escarpments, rims, desert benches, and sloping to moderately steep dip
slopes of anticlines and synclines (USDA Soil Survey, 1970, 1980a, 1980b).
The Colorado Plateau is a landscape of topographic and climatic extremes where
ecological and geological change are constant. While geologists have discovered many of
its secrets during more than a century of investigation, we know relatively little about the
ecology of the region. In contrast to the Plateau's hot desert neighbors to the south, not
much is known about what lives there or how it lives. And despite abundant evidence of
change, we have no long-term programs to monitor or to understand processes of
ecological change on the Colorado Plateau.
The dynamic tapestry of life on the Plateau remains an enigma. Partly due to its
isolation and protective topography, one of America's most desolate and remarkable places
may hold a rich repository of ecological information about living under extreme conditions.
This project launches a new expedition into what Major Powell called "the great unknown."
3.3 LOCATION OF THE 1992 COLORADO PLATEAU PILOT STUDY
The Colorado Plateau as described above will be the site of a demonstration study in
1993. However, the entire region of 130,000 square miles is much more extensive than
27
-------
needed to fulfill the requirements of the intended indicator evaluation pilot. Thus, a small
portion of the Plateau was chosen.
The study area for the 1992 indicator evaluation pilot activity lies between 37° and 39°
North latitude and between 109° and 111° West longitude. Figure 3-3 shows the location of
the EMAP-Arid grid points on this area, superimposed on a map of rivers and major roads.
The land ownership status of each sampling point is also indicated. (Data on the land
status/ownership of Utah was provided by the Utah State University Fish and Wildlife
Cooperative Research Unit of the US Fish and Wildlife Service GAP program.)
In this region are federal lands (i.e., BLM, NPS, and FS), part of the Navajo Nation,
state lands, and private lands. This diversity of ownership and jurisdiction will allow the pilot
to be interagency in its implementation. The area is bisected by the Colorado River and
includes many canyon lands that will allow for evaluation of logistical requirements in some
of the most difficult terrain that EMAP-Arid will face. The area includes both Great Basin
Desertscrub and Great Basin Conifer Woodlands, the two formation types chosen for
indicator evaluation.
3.4 REFERENCES
Bender, G.L., 1982. Reference handbook on the deserts of North America. Greenwood
Press, Westport, Connecticut.
Kepner, C.H. and B.B. Tregoe. 1981. The new rational manager. Princeton Research
Press, Princeton, New Jersey, 224 pp.
Kepner, W.G. and C.A. Fox, eds. 1991. Environmental Monitoring and Assessment
Program. Arid ecosystems strategic monitoring plan. EPA 600/4-91/018. U.S.
Environmental Protection Agency, Washington, DC.
USDA Soils Survey. 1970. Soil survey, Carbon-Emery Area, Utah. Soil Conservation
Service.
USDA Soils Survey. 1980a. Soil surveys of Canyonlands area, Utah, parts of Grand and
San Juan Counties. Soil Conservation Service.
USDA Soils Survey. 1980b. Soil survey of Navajo Indian Reservation, San Juan County,
Utah. Soil Conservation Service.
28
-------
in1
teg*
110'
109'
FIGURE 3-3
POTENTIAL LOCATIONS OF EMAP-ARTO SAMPLING POINTS
FOR 1992 PILOT STUDY
DRAFT
UAJOl 10AM
DATA souRcea usas
EPA • CORVAUJB
UTAH 3TATH UMVBBfTY
lOCAIUM
ILM - 10UAD O» UNB HAHACTMBIT
n . roun KBVKX
XAT . NAVAHD noXAK mnVATSW
nm - lunonAL ucuAmH AHA
ITATnl
IT -RAH
Un ~ VT1 W91AM
PRODUCED FOR EMAP - ARID BY
TME LABORATORY FOR SPATIAL
ANALYSIS. DESERT RESEARCH
INSTITUTE, RENO, NEVADA.
DATE: JUy. IMt
29
-------
4.0 DESIGN
(Susan E. Franson)
4.1 EMAP DESIGN OVERVIEW
The EMAP design has been described in detail elsewhere (Overton et al., 1990;
Kepner and Fox, 1991; White et al., 1992), but a brief summary follows.
The objectives of EMAP require a design with the following criteria:
• Consistent representation of environmental reality by use of probability
samples.
• Representation of all ecological resources and environmental entities.
• Sufficient flexibility to accommodate post-aggregation for many
alternative subpopulations.
• Provision for the capacity to respond quickly to a new question or issue.
• Spatial distribution of the sample of any resource according to population
distribution of the resource.
• Periodic revisiting of all sampling sites.
In order to meet these criteria, EMAP has adopted a sampling design based on a
random systematic triangular grid. The base density of this grid has about 12,200 points,
with neighboring points separated by 27.1 km, over the conterminous United States. Each
grid point is the center of a hexagon that is 40 sq km in area. The 12,200 hexagons
represent a one-sixteenth sample of the area of the conterminous United States.
The 40 sq km hexagons provide the first stage in a double sampling approach.
Information about the land cover within these hexagons is used to structure the field
sampling, or second stage, conducted by each Resource Group. Thus, every sample is
from an equal support base.
The geometry of the triangular grid allows changing of the density of sampling points
to meet particular needs. For example, for rare or locally abundant resources, the grid
30
-------
density may need to be enhanced to provide for an adequate sample size for estimation.
The triangular grid density can be enhanced by factors of 3, 4, 7, or multiples of these
numbers. Similarly, the density of the grid may need to be reduced for some purposes. In
particular, the dual objectives of estimation of both current status and long-term trends
lead to the sampling of one-fourth of the base grid points each year. This sampling can still
occur on a triangular grid, so that status estimates for the entire United States can be made
each year, and each site is revisited after four years, so that trends can be determined.
4.2 EMAP-ARID DESIGN OVERVIEW
4.2.1 EMAP-Arid Population and Subpopulations
The first step in developing a sampling design is to identify the population of interest.
The EMAP-Arid population was defined in Section 1.2 as follows:
Definition of Arid Ecosystems
Terrestrial systems characterized by:
1. Potential evapotranspiration exceeds precipitation.
2. Annual precipitation ranges from <5 to 60 cm.
3. Air temperatures range from -40 to 50° C.
4. Vegetation
D Dominated by woody perennials, graminoids, succulents, and
drought-resistant trees
D Low-form physiognomy, open canopies
a Includes riparian communities
a Excludes intensively managed agriculture
Within the population of arid systems, several subpopulations are of interest. These
subpopulations are defined to be the following formation types:
31
-------
EMAP-Arid Formation Types (adapted from Brown et a/., 1979)
Desertscrub | Riparian Forest
Grassland | Riparian Scrub
Scrubland | Strandland
Woodland |
Tundra |
While each of these formation types represents a subpopulation of interest to
EMAP-Arid, other subpopulations can be defined, such as pinyon-juniper woodlands of
the Mojave. These other subpopulations will be defined as the EMAP-Arid program
evolves.
The upland formation types on the left (except for tundra) are extensive resources.
Simulation studies for the Southwest have determined that adequate sample sizes for
estimation of these subpopulations will result from the base density of the grid applied to
arid ecosystems, without further delimitation of these at the frame development stage.
The lowland formation types, the riparian and strandland systems and tundra (at least
in the alpine tundra of the conterminous U.S.) and riparian communities represent
subpopulations that should be included as primary resources. Primary resources need to
be explicitly defined for frame development and may require an enhancement of the grid to
ensure that adequate sample sizes are obtained for estimation. It is possible that the
lowland formation types could be combined into one primary resource with the three
formation types representing subpopulations of that lowland primary resource.
This separation of subpopulations versus primary resources has been confirmed on a
preliminary basis by overlaying the EMAP base grid on the Brown and Lowe Biotic
Communities map of the Southwest. Whether the primary resource of lowland formations
and tundra will yield sufficient samples for those subpopulations is yet to be determined.
Riparian systems and strandland represent elongate resources, while alpine tundra is
a discrete resource. Frame materials, sampling design, and plot design for these special
resources represent continuing challenges yet to be solved by the EMAP-Arid team. Other
Resource Groups, primarily the Surface Water group, share some of these challenges with
32
-------
elongate and discrete resources. EMAP- Arid continues to interact with the other Resource
Groups and the Design and Statistics team to find solutions to these challenges.
4.2.2 EMAP-Arid Frame and Extent Estimation
For its extensive resources, EMAP-Arid will rely on the EMAP grid, locating sampling
points with respect to the grid center points by a pre-determined rule. For implementation,
it is doubtful that the exact center will be used. Using the exact center point would result in
generally known locations for all sample points. For reasons that are beyond the scope of
this Implementation Plan, it may be beneficial to randomly offset sampling points from the
grid center. This could be done by using the Forest Inventory Analysis (FIA) photo point
closest to the center of the hexagon, or using some other rule. These options are being
examined.
EMAP-Arid expects to use the vegetation mapping data available through
collaboration with the GAP program of the Fish and Wildlife Service for extent estimation.
The feasibility and methods for so doing will be a part of this pilot investigation.
4.3 EMAP-ARID PILOT STUDY SUBPOPULATION AND DESIGN
4.3.1 Pilot Study Subpopulations
The goal of the pilot is to obtain information about indicators. EMAP-Arid has chosen
to test its indicators in two of its subpopulations, desertscrub and woodland. Limiting the
pilot to these two formation types will allow evaluation of indicator performance in two
diverse systems while maintaining an adequate sample size in each. These subpopulations
are represented on the Colorado Plateau by Great Basin Desertscrub and Great Basin
Conifer Woodland.
4.3.2 Great Basin Desertscrub (adapted from Brown et al., 1979)
Great Basin Desertscrub is characterized by low, widely spaced hemispherical shrubs.
The major plant dominants are sagebrushes (Artemisia), saltbushes (Atriplex), and
winterfat (Ceratoides lanata). These are joined in varying degrees by Rabbitbrush
33
-------
(Chrysothamnus), Blackbrush (Coleogyne), Hopsage (Gray/a), and Horsebrush
(Tetradymia). The major series within the Great Basin Desertscrub biome are those
dominated by various species of Sagebrush (Artemisia), Shadscale (Atriplex confertifolia),
Blackbrush (Coleogyne ramosissima), Winterfat (Ceratoides lanata), Greasewood
(Sarcobatus vermeiculatus), or Rabbitbrush (Chrysothamnus).
These principal scrub species are much-branched, non-sprouting, aromatic
semishrubs with soft wood and evergreen leaves. These shrubs are mostly without spines.
There are few cacti — either in numbers of individuals or species. Those present tend to
be of short stature or prostrate and include a few chollas (Opuntia whipplei, O. pulchella),
prickly pears (Opuntia polyacantha, O. gracilis, O. erinacea), and hedgehog cacti
(Echinocereus triglochidiatus var. melanacanthus, E. fendlerivar. fendleri). Small cacti
(Pediocactus, Sclerocactus) and Echinocactus polycephalus var. xeranthemoides occur in
more southern locales.
Species diversity is characteristically low in all major communities of this biome, with a
dominant shrub occurring to the virtual exclusion of other woody species. Another feature
setting this desert apart from others of the region is the absence of characteristic desert
plants in minor waterways; nor is there a fringe of more closely spaced upland plants along
these habitats of slightly more favorable moisture conditions. There are, however, both
cosmopolitan and characteristic plants along flood plains of the larger waterways: included
here are Greasewood (Sarcobatus vermiculatus), Four-wing Saltbush (Atriplex
canescens), and New Mexican Forestiera (Forestiera neomexicana). The introduced
Russian Olive (Elaeagnus angustifolia), and in the warmer regions Saltcedar (Tamarix
chinensis), may be present along wetland stream channels.
Sagebrush Series - usually have Big Sagebrush (Artemisia tridentata var. tridentata),
Bigelow Sagebrush (A. bigelovii), or Black Sagebrush (A. arbuscula ssp. nova) as
dominants, although any of 18 other closely related taxa of Artemisia may be dominant.
34
-------
Sagebrush communities are regarded by many as steppe or shrub steppe because of the
usual importance of grasses.
Shadscale Series - has Shadscale (Atriplex confertifolia) as dominant. The general
appearance of this community is one of open starkness with the dominant woody plants
attaining heights of only 0.3 to 0.6 m. Although widely scattered, perennial grasses are
commonly found in the Shadscale community.
Blackbrush Series - has Blackbrush (Coleogyne ramosissima) as dominant. Perennial
grasses are commonly prevalent in unburned stands.
Other Series - several additional communities may be found within the Great Basin
Desertscrub biome. These may have as dominants: Sand Sagebrush (Artemisia filifolia),
Greasewood (Sarcobatus vermiculatus), Four-wing Saltbush (Atriplex canescens),
Fivehook Bassia (Bassia hyssopifolia), Inland Saltgrass (Distichlis spicata var. stricta),
Common Russian Thistle (Salsola kali), seepweeds (Suaeda spp.), or Winterfat
(Ceratoides lanata).
4.3.3 Great Basin Conifer Woodland (adapted from Brown et al., 1979)
This cold-adapted evergreen woodland is characterized by the unequal dominance of
two conifers — juniper (Juniperus) and pinyon (Pinus). These trees rarely, if ever, exceed
12 m in height and are typically openly spaced (woodland), except at higher elevations and
other less xeric sites where interlocking crowns may present a closed (forest) aspect. The
shorter, bushier junipers ("cedars") are generally more prevalent than pinyons, but either
may occur as an essentially pure stand. Structurally, these juniper-pinyon woodlands are
among the simplest communities in the Southwest.
Several species of juniper may assume or share dominance in the Southwest. These
include Rocky Mountain Juniper (Juniperus scopulorum), Utah Juniper (J. osteosperma),
and One-seed Juniper (J. monosperma). Rocky Mountain Pinyon (Pinus edulis) is the
common pinyon almost throughout, although west of longitude 113.5 it is largely replaced
35
-------
by the single needled form (P. monophylla) or the Four-leaved Pinyon (P. quadrifolia). Not
included as Great Basin Conifer Woodland species are Alligator-bark Juniper and Mexican
Pinyon.
The understory typically is composed of grasses (e.g., Bouteloua gracilis) and shrubs,
e.g., Threadleaf Groundsel (Senecio longilobus) and Snakeweed (Gutierrezia sarothrae).
Also well represented in many of these grass understories are Galleta Grass (Hilaria
jamesii), Indian Ricegrass (Oryzopsis hymenoides), and Western Wheatgrass (Agropyron
smithii). Other grasses include several muhleys (Muhlenbergiaspp.), dropseeds
(Sporobolusspp.), and Junegrass (Koeleria cristata).
4.3.4 Pilot Study Design
From the above descriptions, it may seem that the decision about which formation
type a site lies within is somewhat subjective. One of the questions that the pilot will
address is the frequency of sites crossing a boundary between vegetation types. Thus, the
pilot will, in part, test whether these descriptions are sufficient to unambiguously define the
subpopulations. In addition, EMAP-Arid is exploring the development of a dichotomous key
for use in on-ground identification of formation types.
The area of the pilot is restricted to approximately the area bounded by 37 to 39
degrees North latitude and 109 to 111 degrees West longitude (see Figure 3-3). To locate
the points for pilot testing of the indicators in this area, the base grid was enhanced by a
factor of four, then every fourth point was chosen. This has the same effect as locating the
sampling points one-half the distance to the nearest point to the northeast. This strategy of
point location allows the triangular grid to be maintained, but does not compromise the
sampling points that will be used for implementation.
Forty points lie within this area, which was selected to achieve approximately 20 sites
each of Great Basin Desertscrub and Great Basin Conifer Woodland. Each of the 40
chosen points will be visited. If the sample site falls within either Great Basin Desertscrub
or Great Basin Conifer Woodland, sampling will proceed. If the sample site falls into some
36
-------
other subpopulation, it will not be sampled. It is expected that sampling will occur on
approximately 30 sites, with about 15 in each subpopulation.
4.4 PLOT DESIGN FOR MEASURING INDICATORS
The sample plot consists of specific plot designs for each indicator overlaid on one
another resulting in a hexagon shaped plot (Figure 4-1). This plot design resembles that
being used by the Forest Health Monitoring Program (Kucera and Martin, 1991) with
additional features and modifications to accommodate arid ecosystems. Greater details of
the sampling procedures are found within the sections for the specific indicators and in the
Protocols of the Field Operations and Training Manual. However, general information on
the plot design is provided here to orient the reader for the upcoming sections and to
present the sampling of all indicators as a part of an overall design.
Figure 4-1 shows a central circular subplot, MD, centered on each designated
EMAP-Arid sampling point. Six satellite subplots are located with their centers 40 m from
the center point and oriented at 0, 60, 120, 180, 240, and 360 degrees relative to compass
North. Each of these circular subplots is 7 m in radius, with an area of 154 sq m. Radial
transects, AR, BR, and CR, extend from the center point to the centers of subplots A1, A2,
and A3, respectively. Exterior transects, AE, BE, and CE, extend between centers of
subplots A1 and A2, B1 and B2, and C1 and C2, respectively. Soil sampling locations are
at AP, BP, and CP, each 20 m from the associated subplot center point.
Shrubs and trees greater than 1.5 m in height within subplots MD, A1, B1, and C1 will
be identified and measured as a part of the vegetation composition, structure, and
abundance indicator.
The vegetation composition, structure, and abundance of shrubs less than 1.5 m in
height will be measured in 1 m x 2 m quadrats, aligned with their long axis parallel to each
of the six transects. The quadrats along each transect are separated by 1 m intervals, with
12 quadrats sampled along each of 6 transects, for a total of 72 quadrats sampled on the
37
-------
_R - radial transect
_E - external transect
_P - soil pit clockwise from radial transect
10m
Figure 4-1. EMAP-Arid Sample Plot Design.
38
-------
plot. Within each of the 1 m x 2 m quadrats is a 20 cm x 50 cm subquadrat that will be
evaluated to determine the vegetation composition, structure, and abundance of forb and
grass ground cover, and surface soil attributes.
Three soil sampling areas are located at AP, BP, and CP. At one of these areas, a soil
pit will be dug to evaluate characteristics of the soil profile and to collect soil samples to be
sent to the laboratory for analysis of physical and chemical properties. At the two remaining
areas, soil will be described to 50 cm, augered and described below this depth to 1.5 m or
bedrock, and samples collected from the top two horizons and sent to the laboratory for
analysis. The above approach will be used at half of the study plots. At the other half, all
three soil areas will have surface soils described to a depth of 50 cm, augered and
described below this depth to 1.5 m or bedrock and samples collected from the top two
horizons. The study plots for complete soil profile determination were randomly chosen
from the 40 available sites.
Information on spectral properties will be measured on half of the 40 sites.
Measurements will be made within each of the seven circular subplots on a grid centered
on the subplot center. The grid illustrated for subplot C2 is a square of 4 x 4 points with
vertical and horizontal spacing between sampling points of 3 m. In addition, 3 spectral
measurements will be made evenly spaced in the 1 m x 2 m area in every other quadrat,
beginning with the first quadrat, for a total of 6 quadrats on a transect.
The entire sample plot represents a "conceptual hectare" (Figure 4-2). If one
imagines that each circular subplot represents an area surrounding it that has a radius of
20 m (one half the distance between the center points), then the entire plot represents
either a circle of radius 60 m and area of 11,310 sq m or a hexagon with 60 m from the
center to each vertex, with an area of 9350 sq m. This is an important conceptualization,
especially for the spectral properties indicator. TM pixels are 30 m on a side, so that a 3 x 3
cluster of pixels represents 8100 sq m and a 4 x 4 pixel cluster covers 14,400 sq m. Thus,
39
-------
Area of Circle: 11,310 sqm
Area of hexagon: 9,350 sq m
R - radial transect
E - external transect
P - soil pit clockwise from radial transect
N
h -I
10m
Figure 4-2. EMAP-Arid Sample Plot Conceptual Hectare.
40
-------
the conceptual hectare of the sample plot can be linked with remotely sensed spectral
data.
The plot structure represents a nested design with one plot at each site, several
subplots for each indicator, and potentially several samples within each subplot. This
structure allows for estimation of the sampling design components variance for each
indicator, to be discussed in greater detail in Section 9, Analysis and Reporting.
4.5 REFERENCES
Brown, D.E., C.H. Lowe, and C.P. Pase. 1979. A digital classification system for the biotic
communities of North America, with community (series) and association examples for
the Southwest. Journal of the Arizona-Nevada Academy of Science 14 (Supplement
Kepner, W.G. and C.A. Fox, eds. 1991. Environmental Monitoring and Assessment
Program. Arid ecosystem monitoring plan. EPA 600/4-91/018. U.S. Environmental
Protection Agency, Washington, DC.
Kucera, B.C. and B.E. Martin, eds. 1991. FY91 indicator evaluation field study for
Environmental Monitoring and Assessment Program-Forests (EMAP-F). U.S.
Environmental Protection Agency, Atmospheric Research and Exposure Assessment
Laboratory, Research Triangle Park, North Carolina.
Overton, W.S., D. White, and D.L. Stevens, Jr. 1990. Design report for EMAP
Environmental Monitoring and Assessment Program. U. S. Environmental Protection
Agency Environmental Research Laboratory, Corvallis, Oregon. EPA/600/3-91/053.
White, D., A.J. Kimerling, and W.S. Overton. 1992. Cartographic and geometric
components of a global sampling design for environmental monitoring. Cartography
and Geographic Information Systems 19(1):5-22.
41
-------
5.0 INDICATORS
5.1 INTRODUCTION
(Robert P. Breckenridge)
The following section discusses the relationship of indicators to the EMAP-Arid
assessment endpoints and societal values. A conceptual approach (Figure 5-1) was used
to identify the links between external components, stressors on ecological processes, and
retrospective and existing data relative to making statements about the sustainability,
biodiversity, and aesthetics of arid ecosystems. The EMAP-Arid Strategic Plan identified
seven groups of indicators (Kepner and Fox, 1991). These groups were further evaluated
by a group of arid ecosystem scientists at a workshop in Logan, Utah (October 28-30,
1991) to identify specific measurements and address peer review comments (related to
focusing the indicator effort) from the Strategic Plan. Nine groups of indicators were
evaluated at the Logan workshop. However, because of budgetary constraints, only a
subset of indicators could be selected for testing in the pilot. All the indicator groups are
discussed in this section so the reader can appreciate the scope of what is being
considered by EMAP-Arid. However, only those being tested in the pilot will be discussed
in detail.
5.1.1 Assessment Endpoints and Indicators
EMAP-Arid has identified three societal values that are of prime importance in
determining the condition of arid ecosystems. These societal values are discussed in
Section 1.0 (Figure 1-3) and include: 1) sustainability; 2) biodiversity; and 3) aesthetics.
Societal values are difficult to measure directly; thus, a set of assessment endpoints
associated with the various societal values has been selected.
The assessment endpoints are quantitative or quantifiable expressions of the
environmental value considered in an analysis (Suter, 1990). Eight assessment endpoints
have been identified for possible use in the EMAP-Arid monitoring program (Table 5-1).
42
-------
Chemistry
Socio-Economic
Biogeochemistry
Soil Chemistry, Leaf & Litter
Chemistry, Nutrient Transport fe
C, N, P Soil Interactions,
Methane Production,
Leaf & Utter N&R
Soil Salinity
(Biodiversity
Sustainability)
Snriatal IgiA
(Desertification,
Water Resources,
Global Change
Index of layer categories for Figure 5-1.
1 - External Driver Components
2 - Broad Resource Indicator Classes
3 - Major Indicator Components and Processes
4 - Measure Parameters to Assess Resource Status
5 - Resource Trend Analyses
6 - Ecological Risk Assessment
Figure 5-1. EMAP-Arid Conceptual Model. The model blends ecological modeling with the EMAP
assessment strategy. To move through the model, start with the external components (layer 1)
of atmospheric, stressors, and socio-economic factors, which drive the arid terrestrial system.
The indicator classes (layer 2) respond to these drivers, and interact with each other via major
indicator components (layer 3). Layer (4) holds examples of actual measurement parameters
that reflect indicator components. These measurements are used to assess resource status.
The current resource status is then put into a trend perspective (layer 5) by coupling to
retrospective indicators, long-term historical data and soil potential. Ultimately, resource status
and trend data are integrated into an ecological risk assessment (layer 6), to assess arid issues
and endpoints.
43
-------
These were identified through a series of workshops and peer reviews involving
EMAP-Arid team members, the academic community, and various Federal resource
agencies (Kepnerand Fox, 1991).
TABLE 5-1. ASSOCIATION BETWEEN EMAP-ARID ASSESSMENT ENDPOINTS
AND SOCIETAL VALUES
ASSESSMENT ENDPOINTS
Assessment conducted to measure
change in
1 . Net Primary Productivity
2. Species composition and
abundance
3. Soil quality
4. Landscape patterns/land use
5. Community structures
6. Surface water quality
7. Surface and subsurface water
quantity
8. Socio-economic factors
SOCIETAL VALUES
SUSTAINABILITY BIODIVERSITY AESTHETICS
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Indicators are characteristics of the environment that, when measured, quantify
magnitude of stress, habitat characteristics, degree of exposure to stressors, or the degree
of ecological response to an exposure (Hunsaker and Carpenter, 1990). Indicators serve
as the basis for quantification of the assessment endpoints (i.e., the actual measurements
to be made). For example, water holding capacity, bulk density, and surface soil
morphological types are indicators that serve to quantify the assessment endpoint of soil
quality. A decrease in water holding capacity, decrease in bulk density, and shift in soil
morphological types (e.g., from Types I and II, litter, vegetation, and cryptogamic crusts, to
Types III and IV, compacted desert pavement) (Eckert et al, 1986) could indicate a marked
44
-------
decrease in the soil quality assessment endpoint and suggest poorer societal values
related to sustainability (desertification) and biodiversity. The associations of various
candidate indicators to the assessment endpoints are identified in Table 5-2.
TABLE 5-2. ASSOCIATION OF INDICATORS AND THEIR TYPES
WITH ASSESSMENT ENDPOINTS
Assessment Endpoint
1. Change in net primary
productivity
2. Change in species
composition and abun-
dance
3. Change in Soil Quality
Indicator
Spectral reflectance - NDVI
Biomass
C:N:P ratio in plant tissue
Energy balance using Bowen
ratio
Dendrochronology
Breeding Bird Census
Abundance of field mice
Abundance and distribution of
ground beetles
Vegetation Composition and
Abundance
Bulk density
Soil salinity - saturation ex-
tract electrical conductivity
Extractable cations - Ca, Mg,
Na, and K
Extractable soil P
C:N:P ratio in plant tissue
Pedon description
Water retention
pH, carbonates, etc. (reaction)
Sodium absorption ratio (SAR)
Particle size analysis
Surface soil roughness
Surface soil cover
Cryptogams/lichens
Type of
Indicator
Response
Response
Exposure
Response
Response
Response
Response
Response
Response
Exposure
Exposure
Exposure
Exposure
Exposure
Response
Exposure
Exposure
Exposure
Exposure
Exposure
Exposure
Response
Measurement
Category
Synoptic/
sample
Sample
Sample
Synoptic/
sample
Sample
Sample
Sample
Sample
Sample
Sample
Sample
Sample
Sample
Sample
Sample
Sample
Sample
Sample
Sample
Sample
Sample
Sample
Status for
Pilot
Included
Possible
inclusion
Included
Included
Included
Included
Included
Included
Included
Included
Included
Included
Included
Included
Included
45
-------
Table 5-2. (continued).
4. Change in Landscape
patterns/Land use
5. Change in Community
Structure
6. Change in Surface Wa-
ter Quality
7. Surface and Subsurface
Water Quantity
8. Change in Socio-eco-
nomic factors
Soil erosion
Cation exchange capacity
Water soluble anions
CO3, HCO3, Cl, SO4, and NO3
Habitat/cover type proportions
Spatial distribution of agricul-
tural and riparian vegetation
per stream reach
Fractal dimension
Abundance/density of key
physical features
Spatial distribution of grazing
intensity
Riparian condition
Vegetation composition
Benthic macroinvertebrates
assemblage
PH
Alkalinity
Conductivity
N, P, and organic carbon
Toxins
Benthic macroinvertebrates
Annual Flow Duration Analysis
Mean Annual Discharge
Flood magnitude
Low Flow Magnitude
Ground water level
To be developed
Response
Exposure
Exposure
Habitat/
Stressor
Stressor/
Exposure
Stressor/
Exposure
Stressor/
Exposure
Habitat/
Response
Response
Response
Response
Exposure
Exposure
Exposure
Exposure
Exposure
Response
Response/
Stressor
Response/
Stressor
Response/
Stressor
Response/
Stressor
Response/
Stressor
Sample
Sample
Sample
Synoptic
Synoptic
Synoptic
Synoptic
Synoptic/
sample
Synoptic/
sample
Sample
Sample
Sample
Sample
Sample
Sample
Sample
Sample
Sample
Sample
Sample
Sample
Samples
Possible
inclusion
Included
Included
Possible
inclusion
Included
Evaluate other
agencies' data
Evaluate other
agencies' data
Evaluate other
agencies' data
Evaluate other
agencies' data
Evaluate other
agencies' data
46
-------
EMAP has identified four types of indicators for determining ecological condition:
response, exposure, habitat, and stressor. These categories have been provided as a
guideline for use in the selection, evaluation, and development of the proposed indicators
for EMAP-Arid.
• Response indicators are attributes that quantify the integrated response of
ecological resources to individual or multiple stressors.
• Exposure indicators are physical, chemical, and biological attributes that can be
used to suggest pollutant exposure and assist in the diagnosis of possible causes of
stress.
• Habitat indicators are attributes that describe the condition of the environment.
They are used to suggest whether alteration or disturbance of the physical habitat is the
possible cause of poor condition in response indicators.
• Stressor indicators are economic, social, or engineering attributes that are used
to identify the possible sources of environmental impairment or exposure to impact.
Table 5-2 lists candidate indicator measurements and their relative type proposed for
the EMAP-Arid Pilot. Those indicators selected for the 1992 Colorado Plateau pilot are
discussed in detail in Sections 5.2, 5.3, and 5.4. Two of the remaining candidate indicators,
landscape and retrospective, are discussed in a general nature in Appendix A to provide
the reader with a better understanding of where EMAP-Arid is headed under full
implementation. Selection of indicators for the pilot is discussed in the following section.
5.1.2 Selection of Potential EMAP-Arid Indicators
EMAP has adopted a process to move indicators from the candidate to the core level
(Figure 5-2). EMAP-Arid is at the initial stage of this process. At the recommendation of
peer review comments on the EMAP-Arid Strategic Plan (Kepner and Fox, 1991), a
workshop was held in Logan, Utah, to reduce the broad list of candidate indicators from the
Strategic Plan to a more selected subset (e.g., what was going to be measured). As a
47
-------
CANDIDATE INDICATORS
IDENTIFY AND PRIORITIZE
Expert Knowledge
Literature Review
Peer Review
t
RESEARCH INDICATORS
Analysis of Existing Data
EVALUATE EXPECTED PERFORMANCE
Field Tests
Peer Review
DEVELOPMENT OF INDICATORS
Regional Demonstration Projects
EVALUATE ACTUAL PERFORMANCE ' Peer Review
CORE INDICATORS
IMPLEMENT REGIONAL AND NATIONAL MONITORING
PERIODIC REVALUATION
Figure 5-2. Indicator Selection, Prioritization, and Evaluation Approach for EMAP.
result of the workshop, the following indicator categories and specific indicators were
recommended for consideration by EMAP-Arid.
1. Spectral Properties - including albedo and the Normalized Difference
Vegetation Index (NDVI) - measured using spectral reflectance and
verified with ground measurements.
2. Vegetation - composition and cover - including Leaf Area Index (LAI) as
a measure of productivity.
3. Biogeochemistry - including C:N:P ratio, soil, litter, and vegetation
samples.
48
-------
4. Energy Balance - includes air temperature, monthly precipitation,
Bowen's Ratio, and Palmer Drought Severity Index (PDSI) for assessing
energy balance.
5. Landscape Patterns.
6. Water Quality - conductivity, N,P,K, organic carbon, alkalinity, and
benthic macroinvertebrates.
Water Quantity - water balance on watershed and ground water basin
level.
7. Retrospective Index - developed using dendrochronology and PDSI
records.
8. Fauna - including the Breeding Bird Census, ground beetle abundance
and composition, and field mouse (Peromyscus maniculatus)
abundance.
9. Erosion index - developed using soil loss values from wind and water
loss equations.
10. Soil physical profile characterization - includes surface and soil profile
description and physical properties.
5.1.3 Selection of 1992 Pilot Indicators
Because the 1992 pilot has several primary objectives (Section 2.2), a balance was
required between the selection of indicators, available financial resources, and other
aspects of the project. The indicator categories identified from the Logan EMAP-Arid
Indicators Workshop were evaluated for inclusion in the 1992 pilot at a follow-up workshop
(held in Las Vegas, November 19,1991). The K-T analysis described in Section 3.1.2 was
again used as a decision tool to allow workshop participants to come to a logical,
documented decision as to the indicators which could be tested in the 1992 pilot.
A set of criteria was developed to guide the pilot indicator selection process. The
criteria were separated into primary criteria that the proposed indicators must satisfy
("musts") and those the workshop participants wanted ("wants") the indicator to meet. A list
of the "musts" and "wants" is presented in Table 5-3. The "must" criterion is a mandatory
requirement to advance through the decision process. If a proposed indicator did not
address the "must" criterion, then it would not be further considered. All proposed indicators
addressed the "must" criterion.
49
-------
The "wants" were then weighted according to their relative importance in meeting the
pilot objectives (see Section 2.2). The most important was given a weight of 13. All others
were then weighted in comparison to the first, from 13 (equally important) down to a
possible 1 (not very important). The weighting scale was used to make visible the
relationships among "wants" (i.e., what mattered most and what could be done without, if
necessary). The mean weight and standard deviations from 12 participants at the Las
Vegas workshop are presented in Table 5-3.
TABLE 5-3. LIST OF "MUSTS" AND "WANTS" FOR SEELECTION OF
CANDIDATE INDICATORS FOR 1992 EMAP-ARID
PILOT. WEIGHTS ON "WANTS" ARE MEANS AND
(STANDARD DEVIATION) FROM N=12.
MUSTS
1. Address the issue of sustainability (focus on desertification and/or
climate change) in arid systems
WANTS
1. Applies to a broad range of biogeographic provinces, BLP formation
types, and ecotones; 11.17 (2.04)
2. Connects or integrates with other indicators (as is); 9.25 (2.93)
3. Remote/automated monitoring (minimal field presence); 5.75 (2.56)
4. Cost effective; 6.17 (3.29)
5. Data availability (preferably electronic or summarized); (5.75 (2.99)
6. Connectivity - can relate or associate on site data collection with remote
sensing measurements; 5.08 (3.00)
7. Responsiveness to change; 9.75 (3.47)
8. Environmental impact of data collection efforts; 3.92 (3.06)
9. Methods with documented protocols including QA approach that could
be put directly into Implementation Plan; 6.00 (2.73)
10. Has documented sampling plot design; 4.83 (2.86)
11. Has existing information on variance; 4.5 (3.95)
12. Diagnostic of the general state of ecosystem health or specific distress
syndromes; 10.17 (3.51)
13. Serves data needs of other agencies for national/regional policy,
planning and management decisions; 7.67 (4.23)
50
-------
The candidate indicators were then evaluated relative to each other, against all "want"
criteria one at a time. A value of 10 was assigned to the indicator that came closest to
meeting the "want" criterion and all other indicators were scored relative to it. Once the
indicators were scored, a weighted score was generated for each candidate indicator by
multiplying the weight of each "want" criterion by the score for the indicator and summing
for all "wants" criteria. For example, spectral properties received the following scores:
"Want"
Weights
Score for
spectral
indicator
meeting
"want"
1
11.17
X
9.14
2
9.25
X
8.29
3
5.75
X
9.71
4
6.17
X
9.43
5
6.75
X
8.71
6
5.08
X
8.57
7
9.75
X
8.00
8
3.92
X
9.86
9
6.00
X
8.17
10
4.83
X
6.00
11
4.5
X
6.80
12
10.17
X
7.29
13
7.67
X
7.29
Weighted Score of 750
The weighted scores for the different candidate indicators were as follows:
Indicator
Spectral Properties
Vegetation Composition and Cover
Biogeochemistry
Energy Balance (Bowen's Ratio)
Landscape Patterns
Retrospective
Erosion Index
Weighted K-T Score
750
710
625
676
620
511
628
The final step in the K-T decision process was to explore consequences of not
selecting one of the candidate indicators. Several indicators were excluded from the
evaluation process: water quality/quantity, soil physical characterization, fauna, and
socio-economic. Water quality/quantity was excluded because the group determined that
data suitable for EMAP usage may be obtained by cooperating with other Federal agencies
(e.g., USGS monitoring programs), EMAP-Surface Waters, or from state water monitoring
51
-------
programs. Once data from these are obtained and evaluated, a decision will be made on
how complete existing water data are and how well they can be used to meet EMAP
objectives.
While socio-economic indicators may well be tied with ecological condition, they were
deferred for the pilot because this linkage is not well understood, may be very indirect, and
involves political considerations well beyond the expertise of the scientists involved in the
project. It is hoped that such indicators may be developed in the future with the
collaboration of economists.
Faunal indicators were deferred for the pilot because scientists at the Logan
workshop felt that animals in general may not be as directly diagnostic for change as the
other research indicators. This is due to the fact that only 10-25% of the energy input to
arid systems passes through animals. Key species may be very diagnostic of change,
however, such key species are not generally present in many different habitats across a
region. An approach that relies on guilds of animals that perform specific functions within
ecosystems needs to be identified. Animals were identified as being good indicators for
toxins and will be considered for this in the future. In addition, data from the Breeding Bird
Census will be evaluated for coverage in arid areas. If adequate, a similar census could be
incorporated by EMAP-Arid in the near future.
Soil physical profile characterization was determined to be a core indicator that is
needed to make associations with spectral properties and vegetation composition. Thus, it
was decided that soil profile data would be collected at each site along with descriptors of
surface soil characteristics. Soil profile physical and chemical aspects were identified as
baseline measurements that would only be resampled at follow-up site visits if the surface
characteristics indicated change.
At the conclusion of the K-T decision process, the workshop participants decided that
adequate funds would only be available to test 3-5 indicators. Thus, the decision was
made to test spectral properties; vegetation composition, structure, and abundance; and
52
-------
soil properties in the 1992 pilot. Upon additional consideration, the group decided to
include the erosion index as part of the soils properties indicator. This was done because
the erosion index could be calculated using data from the literature or collected via the
vegetation and soil properties indicators. The only additional data needed would be slope
and management practices for the plot.
5.2 SPECTRAL PROPERTIES INDICATORS
(David A. Mouat)
5.2.1 Introduction
The purpose of this indicator is to test and evaluate the use of spectral measurements
for the purpose of deriving information about arid ecosystem vegetation and soils. Spectral
measurements of vegetation and soils from satellite platforms will be compared to
concomitant spectral measurements obtained through the use of field instrumentation and
ground-based measurements of vegetation and soils. These ground-based
measurements will come from the vegetation and soils activities described elsewhere in
this Implementation Plan.
Electromagnetic radiation can provide information about the physical and chemical
properties of materials. While the spectral reflectance properties of objects tend to be
wavelength dependent, the determination of these relationships is critical for characterizing
or discriminating the objects. Vegetation, soils, and other materials have spectral
responses that are a function of a diverse array of properties of those materials. These
properties might include moisture content, shadowing, presence of other materials, etc.
Nevertheless, the overall spectral response of a material is largely a function of the material
itself.
Figure 5-3 illustrates the typical spectral response patterns of three different types of
materials: vegetation, soils, and clear water, from 0.4 to 2.4 um. The Landsat Multispectral
Scanner (MSS) and Thematic Mapper (TM) bandpasses have been superimposed on
53
-------
LEAF
PIGMENTS
CELL
STRUCTURE
WATER
CONTENT
DOMINANT FACTOR
CONTROLLING LEAF
REFLECTANCE
1.6 1.8 2.0 2.2 2.4 um
1600 laoo 2000 2200 2400 nm
|— Visible -f
Near-lnfrared
Wavelength
Figure 5-3. Typical Spectral Response Curves of vegetation, soils, and water
from 0.4 to 2.4 \im. Landsat MSS and TM bandpasses within the
spectral region have been superimposed on the graph for reference.
these spectral response patterns. In the visible portion of the spectrum, vegetation
response is largely a function of plant pigments such as chlorophyll, xanthophyll, carotene,
and beta-carotene; in the near infrared, vegetation response is largely a function of
internal leaf (mesophyll) structure; and in the shortwave infrared, vegetation spectral
response is largely a function of internal moisture content. The vegetation response curve
illustrated depicts a typically healthy green leaf. The strong chlorophyll absorption in red is
accompanied by a concomitantly high response in the near infrared. Soil spectral response
is a function of moisture and organic content throughout the reflectance spectrum and
chemical content at varying places, specifically in the shorter visible spectrum, the 400 to
54
-------
900 nm and in the 2000 to 2400 nm portions of the spectrum. Red soils, depicted in the
figure, have iron oxide absorptions in blue and green and thus appear red. The ability to
derive information about these surface materials based on their spectral reflectance
properties is a function of the spatial (the pixel size), spectral (both the placement within
the electromagnetic spectrum and the width of the bandpass), radiometric (the ability to
discern brightness), and the temporal (the timing and repeatability of measurements)
resolution properties of the remote sensing systems used.
Remote sensing involves measurements made in the electromagnetic spectrum using
instruments placed on satellite or aircraft-hosted platforms to characterize land, water,
and/or atmospheric phenomena. While the most widely known and used form of remote
sensing is the aerial photograph, other sensors using a variety of technologies can also be
useful. Multispectral scanners hosted on aircraft and satellites can derive information about
earth surface properties in parts of the spectrum beyond the capability of aerial
photography, preserved in digital format, and having much higher radiometric resolution
than photography. These systems, when placed aboard a satellite, can image wide areas.
One very common system, the Landsat satellite, images a swath 185 km wide. The digital
data recorded from these satellites are highly amenable to processing and interpretation in
a digital form, thus allowing consistent and repeatable measurements.
If ecosystem variables can be measured with any degree of accuracy from the
synoptic perspective of a satellite sensor, then a most effective and efficient method of
ecosystem structure and function may be obtained in a spatial context. A number of
researchers have investigated relationships among remote sensing - derived indices and
ecosystem variables.
5.2.2 Relationships Between Remote Sensing Measurements and
Ecological Variables
A number of researchers have shown very strong relationships between ecosystem
structural (such as biomass or LAI) and functional (such as Net Primary Productivity (NPP))
55
-------
features (e.g., Gholz, 1982; Waring et al., 1978). Gholz (1982), in reporting research on a
transect of coniferous forest ecosystems in west central Oregon, reported relationships
between LAI and Overstory NPP with an R2 of 0.96.
A rather extensive literature exists on the use of remote sensing technology for the
assessment of ecological processes including biosphere functioning (Hobbs and Mooney,
1990). A host of remote sensing measurements are used in this endeavor. Among these is
the use of vegetation indices based upon relationships involving a near infrared (NIR) and
a red (Red) channel. The relationships may take the form of a simple difference, NIR -
Red; simple ratio, NIR/Red; or a dimensionless index such as NIR-Red/NIR + Red (Cihlar
et al., 1991 and Running, 1990). The last relationship is known as the Normalized
Difference Vegetation Index (NDVI). This index has been used extensively to characterize
vegetation and has had a considerable history in remote sensing investigations of
ecosystem processes. Peterson et al. (1987) found strong relationships (R2 = 0.91)
between Landsat Thematic Mapper NIR/Red ratios and LAI in closed canopy, pure conifer
forests in west central Oregon. Nemani and Running (1989) used an Advanced Very High
Resolution Radiometer (AVHRR) -derived NDVI to estimate LAI (LAI == 3 to LAI = 10) in
conifer stands in Montana with an R2 = 0.88. Figure 5-4 illustrates the relationship between
LAI and remotely-sensed indices in forested ecosystems. Other researchers (reported in
Running, 1990) have shown that vegetation index and LAI relationships behave differently
in arid and semiarid environments (e.g., grasslands) than in more mesic environments
(e.g., forest ecosystems). An important element of this indicator pilot will be the
development of similar relationships for arid ecosystems.
Clearly a number of problems are inherent in using satellite derived vegetation indices
(e.g., NDVI) to estimate ecosystem variables. The optimal approach would be an intensive
research study similar to those cited above. Spectral reflectance measurements would be
obtained from trees and shrubs using a LI-COR LAI 2000. Light transmission derived using
Beer-Lambert's laws would be estimated, and LAI in turn estimated. Leaf sampling of
56
-------
JC 5
oc
\u
(a)
I
y-1.92x°-583
R2 = 0.91
S.E. = 0.77
I
I
_L
_L
6 8 10 12 14
LEAF AREA INDEX
16 18
0.65 T
0.50-
z
oc
OL
0.35-
NDVI-LN(LAI/1.625).0.34 o
R2«0.88
t^\ °-20
0>) 3
H-
6
LEAF AREA INDEX
12
Figure 5-4. Relationship Between LAI and Two Vegetation Indices a) illustrates LAI
vs. Landsat TM derived near IR/Red and b) illustrates LAI vs. AVHRR
NDVI (from Running, 1991).
57
-------
those same trees and shrubs would provide an independent estimate of LAI. This would
serve to calibrate the estimates of LAI made from spectral reflectance measurements.
Such an intensive research evaluation is beyond the scope of this pilot. Rather, we will
rely on findings of other researchers to calibrate our spectral reflectance measurements
with LAI until such time as funding allows these intensive calibration studies within the
EMAP-Arid program.
The pilot will focus on correlating the spectral reflectance measured on the ground
with that determined from various satellite platforms to estimate vegetation and soils
features.
5.2.3 Image Acquisition and Remote Sensing Measurements
The EOSAT Corporation will be the primary data source for the primary satellite
imagery. EOSAT provided data will be augmented from data acquired through the Utah
State University Fish and Wildlife Cooperative Unit of the FWS GAP Program and by the
EPA's North American Landscape Characterization (NALC) Program.
Satellite data for the approximate time of the field activities will be acquired.
The optimal time of data acquisition could be determined by an exhaustive study of the
phenological status of vegetation in the study area. Instead, an examination of local
weather station records, short-term AVHRR image assessment, and discussions with local
and regional university, NPS, FS, and BLM personnel will establish an approximate optimal
time for image acquisition. As the purpose of this pilot is to test selected indicators, this
strategy should be appropriate.
The contemporaneous AVHRR, MSS and TM imagery will be compared with similar
imagery, but from earlier years, in an attempt to determine image variance on an annual
basis. The use of imagery from similar dates but from different years could prove extremely
useful in applying these indicators on a regional basis. Imagery will be obtained from
EOSAT, the EROS Data Center, and other sources.
58
-------
5.2.4 Details for Specific Spectral Properties Indicators
The spectral properties indicator concept for the Arid Indicator Pilot proceeds from
some basic assumptions which have been made by other investigators as to the
relationships of vegetation structural and functional variables and spectral measurements.
The EMAP-Arid Pilot will include the testing and evaluation of spectral properties of
vegetation (primarily NDVI but other indices may be used or developed and tested) and of
soils (albedo) as determined by the use of satellite-hosted sensors. These sensors will
include AVHRR, TM, and MSS. Pixels extracted from the data sets will be chosen in such a
way that they coincide with ground observations and EMAP design grid center points. The
number of pixels needed to adequately characterize a given sample point will be tested
and evaluated. It has been suggested (Mike Scott, pers. comm., 1992; Mike Spanner, pers.
comm., 1992) that a 2 x 2 matrix may be adequate for the AVHRR pixels while a 3 x 3
matrix is probably necessary for the TM and MSS pixels.
In the 2 x 2 matrix, an assumption will be made that the center of that matrix will
coincide with the given EMAP grid point. In the 3 x 3 matrix, the assumption is that the
center pixel contains the EMAP grid point. That grid point, and its surrounding area, will be
sampled by the vegetation sampling team for vegetation composition, structure and
abundance and surface attributes (e.g., extent of bare soil). The variability of NDVI data as
gathered by the TM will provide an understanding of AVHRR NDVI variability.
5.2.5 Ground-Based Measurements of Spectral Properties
Ground-based spectral measurements will be made for two basic reasons: to
characterize the spectral measurements made by the satellite sensors to be examined
(AVHRR, Landsat TM, and MSS) and to determine relationships between ground-based
vegetation and soils measurements and their concomitant spectral responses. A
measurement made by a remote sensor integrates or "mixes" the heterogeneity of the
ground area being sensed. In the case of the Landsat TM, this area is 30 m x 30 m. The
ground area or "pixel" may be quite uniform or homogeneous or it may consist of a diverse
59
-------
array of cover types. In southeast Utah, for example, this could involve highly dissected
terrain (and widely varying soils and rock types), scattered shrubs, varying surface organic
matter content, shadows, and other factors. Ground-based spectral measurements of
these materials will determine the spectral composition of the integrated spectral
measurements made by the satellite. These measurements, made in the context of an
appropriate ground sampling strategy, will also help to determine the nature of spectral
variance within pixels.
Ground spectral data will be obtained for half of the study sites (randomly selected)
during the field sampling activity. A Personal Spectrometer II (PS-II) will be employed in
the field. This instrument, with a spectral range of 400 to 900 nm and a spectral resolution
of 2 nm, is a highly portable (3 kg) instrument capable of acquiring spectra in as little as
1/23 second. The PS-II will be used to acquire spectra within the circular subplots and
quadrats along transects of plants, litter, shadows, surface soils, and surface lithology for
the purpose of characterizing the sample site. This information will in turn be used to
correlate the satellite-derived information with the other ground measurements. The PS-II
measurements will also be used to determine the basic spectral properties of the materials
themselves. Spectral analysis software together with other statistical packages (Quattro
Pro) will be used to determine the spectral properties of the ground materials being
examined.
Sampling for spectral properties will proceed at each vegetation transect and at each
of the seven subplots described in Section 4.4 (Figure 4-1). A square grid of 4 x 4 points
with horizontal and vertical distance of 3 m between points will be centered on each circular
subplot center. On each vegetation transect, every other quadrat beginning with the first
quadrat will be sampled for a total of 6 quadrats. Each sample quadrat will be measured 3
times. Figure 5-5 illustrates the position of spectral measurements relative to a transect
and the circular subplot at each end. At each point the PS-II will be positioned
approximately 1 m above the surface and a set of 10 spectra acquired, averaged, and
60
-------
recorded. The PS-II has a field of view approximately 30 cm in diameter from a height of 1
m.
The 16 spectral measurements made on each subplot will be used to estimate the
value and spectral variance for each of seven TM pixels. Three subplots can be combined
to estimate the value and spectral variance in a 2 x 2 pixel cluster (replicated 4 times), or all
seven subplots can be combined to estimate the value and spectral variance in a 3 x 3
pixel cluster. The 220 measurements on the plot also can be used to estimate the value for
a 3 x 3 pixel cluster. Similarly, TM pixels can be randomly selected from MSS or AVHRR
pixels (or clusters of pixels) to estimate the values and spectral variance in the MSS or
AVHRR pixels or clusters. Thus, the spectral properties indicator portion of the pilot study
will evaluate the relationship between on-ground spectral and satellite spectral
measurements.
The spectral measurements of each quadrat can be related to the vegetation
composition and abundance and surface attributes of that quadrat. The 18 spectral
measurements for the entire transect can be combined and related to the vegetation
composition and abundance and surface attributes of that transect. The 220 spectral
measurements for the entire plot can be combined and related to the vegetation
composition and abundance and surface attributes of the entire plot. Similarly, the spectral
Circular
subplot
Circular
subplot
.«m\;":
jjy
i • ~ ~ ; i i • ~ "^ I I-": i i • ~ ~ ; rri i"ij«
\ TRANSECT
Spectral Measurements
* * * in 1 x 2 m quadrat
\ • •"•
V • • •
Figure 5-5. Positions of Spectral Measurements.
61
-------
properties measured from satellite imagery (TM, MSS, or AVHRR) for the entire plot can be
related to the vegetation composition and abundance of the plot. Thus, vegetation
properties and surface attributes can be related to spectral properties determined both from
on-ground measurements and satellite imagery.
In addition, a catalogue of spectral measurements of individual plants and surface
materials will be made. As time permits during the field activities, the botanist will work with
the spectrometer technician to record species identification, phenology, and condition (e.g.,
flowering, dead, withered, healthy) information for individual plants for which spectra are
acquired. Surface features will also be recorded and spectra acquired. Such a catalogue of
spectra for individual plants and surface features will be extremely useful in expanding the
use of remotely sensed spectral properties for vegetation mapping and determining
condition. This information, linked with information on the spectral variance averaged into
TM, MSS, or AVHRR pixels, will help to further develop spectral properties indicators for
future use in EMAP-Arid.
5.3 VEGETATION COMPOSITION, STRUCTURE, AND ABUNDANCE
INDICATORS
(Stephen G. Leonard and Robin J. Tausch)
5.3.1 Introduction
The composition, structure, and abundance of vegetation have been recognized as
useful indicators of environmentally induced changes in arid vegetation. The proposed
measures for the determination of these indicators are estimation of 1) the percent cover
and 2) height of the green vegetation on the site by species. Together, these can provide
an index of leaf area. These measures can provide sensitive indicators of change in
biological condition at the organism, population, community, and ecosystem levels. This
occurs through the relationships of cover and height to water availability and production
(Nemani and Running,1989; Tausch and Tueller, 1990; Tausch and Nowak, 1991).
Ground-based cover measurements can be related back to, and used for ground truthing
62
-------
of, remotely sensed spectral properties when specific timing requirements are met. The
proposed vegetation methods are rapid, have demonstrated levels of precision with proper
training, and are familiar to all land management agencies. Related site and soil
characteristics will be collected simultaneously with, and as a part of, the vegetation
sampling (Section 5.4).
Compatible data are available for many areas of arid/semi-arid vegetation. Similar
methods are in wide use by USDA Forest Service, Intermountain Research Station,
*
Inventory, Monitoring and Evaluation Program (O'Brien and Van Hooser, 1983; Born and
Van Hooser, 1988; Utah Forest Survey Field Procedures, unpublished; USDI Bureau of
Land Management, USDI, 1985).
5.3.2 Details for Specific Indicators
Vegetation Cover - Percent vegetation cover by species on a site provides information on
abundance, relative composition, and dominance in the community. We propose to sample
using the Daubenmire cover class method. The method will be modified as described by
Baily and Poulton (1968) by adding a seventh cover class (<1%) to better indicate trace
occurrences.
TABLE 5-4. MODIFIED DAUBENMIRE COVER CLASSES
CLASS
1
2
3
4
5
6
7
COVER RANGE
<1%
1%-5%
5% - 25%
25% - 50%
50% - 75%
75% - 95%
95% -100%
RANGE MID-
POINT
0.5%
3.0%
15.0%
37.5%
62.5%
85.0%
97.5%
Vegetation Height - Average height of each species on a site. Determined by species and
63
-------
subplot or quadrat, provides information on species dominance and vegetation structure in
the community.
Species Frequency - Quadrat sampling methodology provides for the determination of
frequency by plant species for compatibility with ongoing collection of monitoring data by
management agencies.
Ground Cover - Ground cover by total vascular plant cover, litter, rock, bare soil, and
cryptogams provides important information for soils and erosion analyses. Ground cover for
all but total vascular plant cover will be determined both for the total quadrat and for the
canopy interspace area.
Species Composition - Through both quadrat sampling and site survey, the species
composition of each site can be determined and monitored.
Essential Complementary Data - Includes the description of topography and landforms
surrounding the sample location, its slope and aspect, and information on land use in the
area.
Biomass - is an essential indicator for environmental conditions related to climate change,
desertification, and other processes. Biomass can be related to remotely sensed data for
monitoring purposes. However, sample based measurement will not be included in this pilot
because of cost and logistic constraints.
5.3.3 Sampling Design
Sampling design is similar to that for the EMAP-Forest and maintains functional
compatibility with it. It has been modified for the more spatially heterogeneous
arid/semi-arid communities to be sampled. Sample plot layout is a nested quadrat design
(Figure 5-6). Trees (>1.5 m in height) and widely spaced, large shrubs (>1.5 m in height)
will be measured in four fixed area subplots. Subplots are 7 m in diameter with one
centered on the plot center and with the centers of the remaining three 40 m from the plot
center on radial transects spaced 120° apart. All individual trees and large shrubs in each
64
-------
R - radial transect
.E - external transect
P - soil pit clockwise from radial transect
N
10m
Figure 5-6. Nested Quadrat Design.
65
-------
subplot will be measured for two crown diameters (Meeuwig, 1979), height, and either
diameter root crown or diameter breast height. The crown diameter measurements will be
used to determine percent cover by species in each subplot.
Shrubs, other than those sampled in the circular subplots, will be sampled by cover
class in the 1 m x 2 m quadrats systematically located along each of the three radial
transects and three exterior transects (Figure 5-6). Herbaceous forb and grass species will
be sampled by cover class in 20 cm x 50 cm quadrats nested within each 1 m x 2 m
quadrat. Data for surface features (total vascular plant cover, litter, rock, bare soil, and
cryptogams) are also collected in these 20 x 50 cm quadrats. This provides a total sample
of 72 for each quadrat size. All sample subplots and quadrats are located within a hexagon
of approximately 1 hectare. The entire hexagon will be used for the site description
write-up and searched to locate and identify all plant species present.
Standard practice of sampling programs of many land management agencies and
other ecological studies requires that plots be located entirely within the same vegetation
community/soil type for the resulting data to have adequate precision (EJonham, 1989). In
cases where the plot falls on different vegetation/soil complexes, the common practice is to
shift the plot location so that it is within only one vegetation/soil type. This shifting of plots,
while entirely appropriate for studies of a particular vegetation type, is inappropriate in the
setting of a probabilistic sampling design such as that employed for EMAP.
In order to maintain the systematic random grid and preserve the probabilistic nature
of the EMAP-Arid design, the plot will not be shifted so that it is within only one
vegetation/soil type. Instead, the plot will remain where specified by design and for plots
that contain more than one vegetation/soil type, additional data will be collected to ensure
adequate precision. For all vegetation/soil types that represent at least 16.7% of the total
plot (12 quadrats of 72), a minimum of 30 quadrats (each 1 x 2 m quadrat with its nested
20 x 50 cm quadrat is counted as one) will be sampled for surface features and vegetation
composition, structure, and abundance. These quadrats will be located along extensions of
66
-------
the existing transects. In addition, if the site is to have a complete soil profile (to 1.5 m or
bedrock), then a complete soil profile will be completed in the secondary vegetation/soil
type. If no soil pit falls within the secondary vegetation/soil type, an additional soil pit will be
added for either complete soil profile or surface soils, whichever is appropriate for the site.
Protocols that document the decision rules for when and how to extend transects to add
additional quadrats are included in the Field Operations and Training Manual.
Vegetation sampling will follow a procedure adapted from Rangeland Monitoring
Trend Studies (USDI, 1985).
The following data will be recorded within the 20 x 50 cm quadrat:
1. Cover class (see Table 5-4) for:
• Total vascular plant canopy cover
• Surface features for total plot and for canopy interspace areas:
n rock fragment cover by class: gravel (2-75 mm), cobbles (75-250
mm) and stones (>250 mm)
n litter cover
D surface type by class: I, II, III, IV, and unclassified (Figure 5-7)
M C B
M
M
M BMC
M
Figure 5-7. Surface Type Classes (from Eckert, et. al., 1986).
67
-------
n cryptogamic cover by class: moss, lichen, cyanobacteria
D bare soil
2. Plant species identification from National List of Scientific Plant Names (USDA, 1982
as amended) for each herbaceous grass/forb species.
3. Cover class of basal area for each herbaceous grass/forb species.
4. Average height to nearest 0.1 m for each herbaceous grass/forb species.
The following data will be recorded within the 1 x 2 m quadrat:
1. Plant species identification for each shrub (plants < 1.5 m in height) species.
2. Cover class of canopy for each shrub species.
3. Average height to nearest 0.1 m for each shrub species.
The following data will be recorded within the 7 m circular subplots (MD, A1, B1, and C1,
Figure 5-6) for each individual tree and shrub >1.5 m in height.
1. Species identification from National List of Scientific Plant Names (USDA, 1982 as
amended).
2. Longest crown diameter and the longest one perpendicular to it.
3. Height.
4. Basal trunk diameter for trees or diameter root crown for shrubs (>1.5 in height).
When trees or shrubs have multiple trunks measure all trunks.
For each tree species within each 7 m circular subplot, the number of seedlings and
saplings <1.5 m high will also be recorded.
Voucher specimens for any unknown species will be collected and preserved in a
plant press for later identification. Such species will be assigned a sample number to be
used on data forms until species identification can be confirmed.
To fully characterize the vegetation composition of the site, a general search of the
area will be made to identify any species present on the site but not encountered in the
vegetation transects or circular subplots. These species will be listed with no other data
recorded for them.
68
-------
5.4 SOIL PROPERTIES INDICATORS
(George J. Staldl, Robert P. Breckenridge, Richard D. van Remortel, and Nancy L.
Hampton)
5.4.1 Introduction
Soil quality is an important assessment endpoint that is closely associated with
EMAP-Arid societal values of sustainability and biodiversity. This endpoint is not measured
directly, but is directly associated with a number of indicators that are measured directly
(Table 5-2). The primary objectives of soil sampling in the 1992 field season are to:
• Demonstrate that the soil property indicators can be successfully
implemented in two different Brown, Lowe, and Pase formation types
using a cooperative effort among multiple agencies.
• Evaluate the variance components of surface soil parameters.
• Assess the benefit of describing a soil profile and collecting soil samples
relative to: 1) interpreting other EMAP-Arid indicators (e.g., vegetation);
2) comparing soil profile data to existing published soil survey data to
determine the most cost effective manner for obtaining soil descriptions;
and 3) determining associations with assessment endpoints and societal
values.
• Determine the cost-effectiveness and logistical feasibility of doing a
complete profile lab analysis versus selecting a subset of parameters.
Soil properties directly influence the amount, timing, and distribution (lateral and
vertical movement) of soil moisture available for plant growth. Soil infiltration properties and
surface characteristics also directly affect erosion processes, including overland flows
(runoff) and transport of suspended and dissolved solids. Disturbances and/or stresses to
surface and subsurface soil can influence flow velocity, routing, soil detachment, and
deposition. The result is accelerated soil erosion that further affects moisture infiltration
rates and patterns. Ultimately, physical changes to vegetation communities may result. An
altered soil moisture regime, in conjunction with changes in other soil properties through
erosion, can result in degradation to soil productivity, landscape features, and vegetation
composition and abundance. Consequently, biodiversity, sustainability, and the degree of
desertification of arid ecosystems are highly dependent on soil condition.
69
-------
A conceptual model was used to identify the linkages between soil properties and the
other arid indicators. Wood (1988) provides a conceptual model showing relationships
between land management practices and plant characteristics, soils, and infiltration rates.
This model was modified (Figure 5-8) to show the links among EMAP external driver
components (precipitation, energy, atmospheric chemistry, socio-economic factors, and
natural and anthropogenic stress), land management practices, and infiltration rates with
other key components (biogeochemistry, landscape, fauna, energy, water resources, and
vegetation).
5.4.2 Soil Properties Indicators
Three soil properties indicators will be measured in the 1992 pilot (Table 5-5): 1) Soil
profile (a) description and (b) analysis - the characterization of a vertical section of the soil
through all its horizons and extending into the parent material, 2) Soil surface (a)
description and (b) analysis - the top two soil layers, and 3) Surface soil attributes -
description of attributes of the topmost soil surface including vascular vegetation, rock
fragments, cryptogams, bare soil, litter, surface type, and surface roughness. Soil
properties encompassed by these three indicators control both soil moisture and
susceptibility to erosion processes.
Measurements associated with these three soil property indicators are discussed in
detail in the Field Operation and Training Manual. These measurements will be
incorporated with vegetation and spectral properties data to allow application of the Water
Erosion Production Project (WEPP) (Lane and Nearing, 1989; Flanagan, 1990, 1991),
Universal Soil Loss Equation (USLE) (Wischmeier and Smith, 1978), or revised USLE
(RUSLE) (Mills et al., 1985) and Wind Erosion (WE) (Fryrear, 1992) modules for
development of an Erosion Index as discussed in Section 5.5.
70
-------
SOIL (PHYSICAL) INTERACTIONS BETWEEN KEY COMPONENTS
EXTERNAL
DRIVER
COMPONENT
SOIL
DISTURBANC
SOIL SURFACE
TYPES AND
CONFIGU-
RATION
EVAPORATIVE
LOSS
SOIL
MOISTURE
LONG
TERM
INDICATORS
WATER
RESOURCES
BIOGEO-
CHEMISTRY
INPUT/
OUTPUT
PROCESSES
BROAD
RESOURCE
INDICATOR
CLASSES
Figure 5-8. Conceptual Model (adapted from Wood, 1988).
71
-------
TABLE 5-5. LIST OF SOIL INDICATORS AND ASSOCIATED MEASUREMENTS
FOR SOIL QUALITY ASSESSMENT ENDPOINTS AND
THEIR IMPORTANCE TO SOCIETAL VALUES
INDICATOR
1 a. Soil pro-
file descrip-
tion
1992 PILOT MEASUREMENTS
DESCRIPTION
Soil Taxonomy
Major land resource area
Slope (%)
Physiography
Water Table
Land use class
Surface rock fragments (%)
Rock outcrop (%)
Hydraulic conductivity
class
Drainage class
Elevation
Parent material
Hydrologic group
Erosion/runoff class
Particle size
Horizons
Color, moist/dry
Texture
Consistency
Structure
Mottles
Roots
Pores
Chemical reaction
properties
(e.g., pH, Ca CO3, etc.)
FIELD
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
LAB
SAMPLING SCHEME
FOR 1992 PILOT
Full soil profile will be de-
scribed in 20 of the sam-
pling hexagons. Half will
be in conifer woodland,
half in desertscrub.
Selection of the full pro-
file sites will follow the
statistical design consid-
erations discussed in
Section 4.0. At each site
3 holes, made using a 4"
bucket auger, will be
used to verify that the
soils are within the range
of characteristics of the
same series. The hole
most representative of
the soils unit will be dug
out for full profile descrip-
tion. The other 2 holes
will be described as in
2a. The remainder of the
20 hexagons that do not
get full profile descrip-
tions will be described
from 3 holes per site,
each made using a 4"
bucket auger.
IMPORTANCE TO
SOCIETAL VALUES
Status of the soil profile
description is key to sus-
tainability and diversity of
a site for human, animal,
and vegetative life. The
soil profile would be de-
scribed initially for a site,
and then resampled only
if short-term indicators
(e.g., surface layers,
vegetative composition)
show a dramatic change.
Changes in soil profile
would be associated with
major climatic, land use,
or land disturbance acti-
vities. Soil profile is con-
sidered a long-term
baseline indicator for
EMAP— Arid without
which other short-term
changes may not be in-
terpretable.
72
-------
TABLE 5-5. (continued).
INDICATOR
1b. Soil pro-
file analysis
1992 PILOT MEASUREMENTS
DESCRIPTION
Following 1 a, collect soil
samples from each layer of
a soil profile and submit for
analysis for:
Particle size distribution
Fabric Related Analyses
Bulk Density
Water Retention
Water Retention
Difference
Plasticity Index
Ion-Exchange Analyses
Cation Exchange
Capacity (CEC)
Sodium Absorption Ratio
(SAR)
Exchangeable Sodium
Percentage (ESP)
Chemical Analyses1
Organic Carbon
Total Carbon
Nitrogen
Extractable Bases and
Sum of Bases
Extractable Acidity
Calcium Carbonate
Gypsum
Soluble Cations and
Anions
Mineralogy
Crystalline Mineral
Components
Grain and Amorphous
Mineralogy
Miscellaneous Analyses
Soluble Salts
Total Soluble Salts
FIELD
LAB
X
X
X
X
X
x
X
X
X
X
X
X
X
X
X
X
X
X
X
X
SAMPLING SCHEME
FOR 1992 PILOT
Samples collected in 20
of the hexagons at the
same sites where full
profile description pits
are dug.
IMPORTANCE TO
SOCIETAL VALUES
These physical and
chemical analyses are
used to describe how
well a soil can be used
to sustain plant or animal
production or diversity.
Changes in values from
nominal to subnominal
indicate overuse or cli-
matic shift towards des-
ertification. Changes
from subnominal to nom-
inal indicate conserva-
tion and good land man-
agement. Some of these
may be better indicators
of change (e.g., organic
carbon, SAR, bulk densi-
ty, CEC), but the suite of
analyses is recom-
mended by SCS for in-
terpreting a soil relative
to its erosion rates, sus-
tainability, and biodivers-
ity uses.
1 Note: Not all chemical analyses will be run for all sites. Decisions on chemical analyses will be based on the
type of soil encountered. At some sites, conditions may not warrant use of analyses.
73
-------
TABLE 5-5. (continued).
INDICATOR
2a. Soil sur-
face hori-
zons de-
scription
2b. Soil sur-
face analy-
sis
1992 PILOT MEASUREMENTS
DESCRIPTION
All parameters as listed in
description in 1 a.
Collect soil samples from the
top two layers for analysis of
the following parameters
Particle size distribution
Organic carbon
CEC
Soluble salts
Bulk density
Water retention
Water retention difference
FIELD
X
LAB
X
X
X
X
X
X
X
SAMPLING SCHEME
COD -1 nftO DM /"\T"
rOH 1992 PILOT
At each site, soil-surface
profile will be described
at 3 different locations.
All holes will be dug to
50 cm and the top layers
described to 50 cm. An
auger will be used to de-
scribe the soil material
below 50 cm to a depth
of 1.5 m or bedrock.
Locations of the holes
are as described in the
design section (4.0).
Surface soil samples will
be collected from the top
two layers at three differ-
ent holes at each site.
IMPORTANCE TO
O/~\/"*ICTA 1 \/AI 1 ICO
SOCIETAL VALUES
Surface soil physical and
chemical properties
should be the first to
change relative to natu-
ral and anthropogenic
stresses. These will be
important indicators of
changes in land use, soil
moisture, erosion pro-
cesses, decay of organic
matter, and biodiversity
relative to sustainability
of an area. Changes in
status relative to deser-
tification should initially
be noted in surface soil
indicators. However,
long-term trends may
best be detectable by
changes in the soil pro-
file. Dramatic events
(e.g., floods, fires, short-
term drought) can be
noted via changes in
surface parameters.
Long-term changes
(e.g., shifts in water
table, climatic shifts) may
be better noted in deep-
er soil profiles.
Same as above
74
-------
TABLE 5-5. (continued).
Surface soil
attributes
Erosion In-
dex
Place a 20 x 50 cm plot
frame at 12 locations along
six 40 m transects. This is
the same frame used for
vegetative sampling. Inside
frame record measurements
on:
Surface morphology type
(collected with vegetation
parameters)
Surface cover type
(collected with vegetation
parameters)
Surface roughness
Calculate T using field data
and compare to SCS Soil
Survey value
X
X
X
X
X
These attributes will be
measured along the
sampling transect estab-
lished for sampling ve-
getation. They will be de-
scribed with the same 20
x 50 cm sampling frame
used to assess vegeta-
tion type and distribution.
Descriptions will be
made along six 40 m
transects at each site.
Coordination with the ve-
getation group is impera-
tive to avoid duplication
of data gathering.
Collect slope, vegetative
and surface cover data
from plots
Changes in surface at-
tributes have been
associated with seedling
survival of sagebrush
communities (Eckert et
al. 1986). Presence or
absence of cryptogams,
lichens, or rock frag-
ments are good indica-
tors of how much a site
has been disturbed by
natural, man induced (4-
wheelers), or grazing ac-
tivities. These can be
associated with sustain-
ability and biodiversity.
Areas with excessive
erosional rates will have
reduced sustainability
and biodiversity and be
more prone to drought
Due to a variety of considerations (i.e., budget, logistics), not all soil indicators will be
measured at each hexagon-site. The sampling plans for the various indicators in the pilot
are described in Table 5-5. Relative to soil indicators, all hexagon-sites will fall into one of
two groups (A or B). The following indicators will be measured for each group:
Group A - at each site conduct:
• Full soil profile description and analysis from one hole dug to 1.5 m or
bedrock.
• Auger hole description of soil profile and soil-surface description and
analyses of surface soil at 2 satellite locations.
• Surface soil attributes described along six 40 m transects (these data are
collected by the vegetation group).
Group B - at each site conduct:
• Auger hole description of soil profile at three locations.
• Soil-surface description and analyses at three locations.
• Surface soil attributes described along six 40 m transects (these data are
collected by the vegetation group).
These groups were established based on cost and the rationale that the collection of
complete profile descriptions at 10 sites per formation type (20 total Group A) when
75
-------
considered with the 20 sites with auger hole description will provide enough information on
variability to evaluate the status of the indicator categories. Soils in the remaining
hexagons (Group B) will be described in sufficient detail to make associations with
vegetation and spectral properties indicators.
The sampling plot design is presented in Figure 5-6. Plot location, layout, and field
procedures are discussed in detail in the Field Operations and Training Manual. Estimated
costs and labor requirements associated with the two different types of sites A and B are
presented in Table 5-6.
5.4.3 Erosion Index Indicator
(Robert P. Breckenrldge)
Soil erosion is almost universally recognized as a serious threat to the sustainability of
ecosystems and man's well-being. This is shown by the fact that most governments in the
world give active support to programs of soil conservation.
The objectives of testing the erosion index indicator are to:
1. Determine if field data collected by EMAP-Arid personnel can be used
(along with published data) to calculate wind and water erosion rates for
arid ecosystems.
2. Compare calculated erosion rates for a soil series to those published in
SCS Soil Surveys to determine condition of a site relative to erosion
rates.
3. Compare two water erosion models (USLE/RUSLE) for their precision
and cost effectiveness to provide EMAP level data to make large scale
statements on erosional condition of arid lands.
76
-------
TABLE 5-6. ESTIMATES OF COSTS AND LABOR REQUIREMENTS FOR SOIL
PROPERTIES INDICATORS FOR 1992 PILOT STUDY.
Indicators
1) SOIL PROFILE DE-
SCRIPTION AND ANALYSIS
1a. Pedon description in
field - all layers or horizons
(typifying)
1b. Laboratory evaluation
of all layers or horizons'*
1c. Site profile description
using bucket auger
2) SOIL-SURFACE HORI-
ZON DESCRIPTION AND
ANALYSIS
2a. Field description of top
50cm
2b. Laboratory Evaluation
of top 2 surface layers
3) SURFACE SOIL AT-
TRIBUTES
4) EROSION INDEX
USLE/RUSLE parameters
Slope characteristics
(WEPP)
Rangeland Management
Data
Subtotal per hexagon
Total per hexagon
Group A - Full Profiles Description and
Laboratory Analysis
Status3-
#holes/
site
X-1
X-1
X-2
X-3
X-2
X
Cost/Site
Labor
hours
6
1
2
2
1
3
Supplies'9
$30
$10
Analytical0
$928
$432
Group B - Soil Description
Status-
wholes/
site
X-3
X-3
X-3
X
Costs
Labor
hours
X-3
3
2
3
Supplies6
$15
Analytical0
$648
Some data exist in other agencies and additional data will be collected as part of the
vegetation sampling activities.
3
14
$40
Total for 20 hexagons
GRAND TOTAL FOR ALL SOIL INDICATORS
$1360
$1400
$28,000
$41,260
3
9
$15
$648
$663
$13,260
a X = will be measured
k Cost of sample containers, tape, and packing material.
c All laboratory analysis conducted by USDA, Soil Survey Investigation Lab in Lincoln, Nebraska, following all
SCS QA protocols.
d Cost based on analysis of an average of 5 layers/pedon.
77
-------
Soil erosion from wind and water can be measured directly or calculated using a
variety of models. To obtain accurate field measurements, extensive instrumentation and
sampling of a plot are often required (Larson et al., 1983; Breckenridge et al., 1991). Thus,
EMAP-Arid has decided to use modeling as an alternative approach to determine site
erosion. Three models will be evaluated as part of the Pilot Study: (1) the Universal for
Revised Universal Soil Loss Equation (USLE/RUSLE); (2) the Water Erosion Prediction
Project (WEPP); and (3) the Wind Erosion (WE) equation. These models will be evaluated
for their ability to construct an erosion index for a site. The erosion index (presented in tons
per acre per year) will be specific to a soil series and map unit. The calculated index will be
a combination of the water and wind contributions to the total site erosion rate. The index
will be compared to the erosion factors (T) found on the physical and chemical properties of
soils table (which is Table 12 in all soil surveys) in each of the published USDA soil surveys
for the various counties of the pilot area (see data sources for soil indicator Section 5.4.3).
The T factor is an estimate of the maximum average rate of soil erosion by wind or water
that can occur over time without affecting vegetative productivity over a sustained period.
This value is in tons per acre per year.
By comparing the calculated T for a site to the established nominal T from the soil
survey, EMAP-Arid hopes to make statements about the relative erosion condition of the
site on a regional basis. Thus, the expression for the soil index for a specific soil series
could be as follows:
Site Calculated Values SCS Recommended Value EMAP Association
(Tons/ac/yr)
Water erosion + Wind Erosion < T = nominal
> T = subnominal
The remaining discussion in this section describes the various equations and their inputs.
78
-------
Wind erosion is important and often the main form of erosion in large flat arid regions.
Many of the soils deposited in arid areas are aeolian materials (parent material
accumulated through wind actions). The WE equation is:
E = lxCxKxVxL (Israelsen et al., 1980) in which:
E = soil loss by wind in tons/acre/yr
I = soil wind erodibility factor
C = local wind erosion climatic factor
K = soil surface roughness factor
V = vegetative factor
L = length of the unshielded distance parallel to wind in the direction of the
wind fetch.
The inputs to these parameters will be obtained from field data or by using standard values
(Israelsen et al. 1980). Details of field collection techniques will be presented in the Field
Operations and Training Manual.
The USLE is a well established equation for water erosion originally developed in
1940 to predict long-term soil loss through sheet and rill erosion. USLE is widely used by
the SCS and conservation planners to determine appropriate soil management strategies.
The 1977 National Resource Inventory data base on erosion was derived form the USLE
(water erosion component) and WE (wind erosion component) models. The USLE is a
simpler equation than the WEPP. The USLE calculates soil loss as a product of six factors.
RxKxLxSxVMxP (Israelsen et al., 1980) where:
A = estimated soil loss in metric tons/hectare/yr
R = rainfall (a function of local rainstorm characteristics)
K = soil erodibility (a function of soil properties)
L = slope length
S = degree of slope
VM = Erosional Control Factor (vegetative and mechanical measure) (%
ground cover of grasses and stone, forb density)
P = erosion control (practices such as contouring, strip-cropping, or
terracing)
79
-------
The RUSLE is essentially identical to the USLE but incorporates additional parameters
addressing gully erosion.
WEPP is a new project designed to generate improved erosion prediction technology
on rangelands for use by multiple federal agencies. EMAP-Arid will collect the required
inputs for WEPP that include:
• Climate
• Soil properties
• Topography
• Land use
WEPP (Lane and Nearing, 1989; Flanagan, 1990,1991) is being designed ultimately
to replace USLE/RUSLE, but is in its trial stage.
EMAP-Arid will collect the required input for the WE, USLE/RUSLE, and WEPP for
those pilot sites that have published soil surveys (an exception could be bare rock). Based
on available funds, field data will be input to the 3 equations and erosion values calculated.
These values will then be compared to published T values from the soil surveys. Data
evaluation will be conducted to determine cost effectiveness of different models and
precision between USLE/RUSLE and WEPP.
5.4.4 Data Sources and Additional Data Requirements for Soils Indicators
Existing data - In order to assure the quality standards necessary for EMAP indicator
data collection, the procedures and methods identified in the National Soils Handbook
(NSH) as part of the National Cooperative Soil Survey (NCSS) will be used. Copies of the
NSH, Soil Survey Manual (SSM), Soil Taxonomy, and Soil Survey Investigations Report No.
1 will be obtained from the USDA-SCS and/or the U.S. Government Printing Office.
Existing data pertinent to the sample sites in the pilot area will be obtained from the
Utah State SCS. Additional data specific to the pilot study area will be obtained from SCS
field offices, BLM State or District Offices, and the FS Forest or Range District Offices. The
80
-------
SCS offices will be the main source of soil data because they are charged with soil
correlation, soil database maintenance, and manuscript publication responsibilities.
The soil data required for proper soil series identification at the sampling site are
available for: Grand County except for the area from Book Cliffs north; San Juan County
except for the southern section including the Navajo Nation; Eastern Garfield County,
including Escalante and Boulder and the Henry Mountains area which includes eastern
Garfield and Kane Counties; and southern Wayne County. The Emery County Castle Valley
is also available, but the balance of Emery County is in process. Soil Surveys for Grand
and San Juan Counties can be obtained from the Moab SCS office (801-587-2481); for
Emery County and Wayne County from Leland Sass - Price SCS Office (801-637-0041);
and for Garfield and Kane County from Gordon in the Cedar City SCS Office
(801-586-2429). This information includes the soil maps, map unit identification legend,
map unit descriptions, table of soil classification, typical soil pedon descriptions, soil
laboratory data, and soil interpretation tables or form SCS-SOI-5 for each soil component
in a map unit where the sample sites are located.
Access will be sought to the SCS-State soil survey database and existing GIS soils
data pertinent to the sampling sites. This is an important source of information that
identifies existing soil properties for the pilot area. The pilot will test the utility of existing
soils information.
Due to the kind, amount, and complexity of interpreting existing and collecting new
data, a journeyman level soil scientist familiar with soil survey field procedures and data
input and analysis will be on each field crew to collect data. Existing data will be used to
identify soil series at the sample sites. The SCS soil map for Utah will be used for the initial
determination as to whether the sample site is located on a homogenous soil series and
vegetation community. If not, soils will be sampled for all vegetation/soil types represented
on the site as discussed in Section 5.3.3. Determination of the field site will use the existing
soil data to pinpoint location, soil series, and vegetation community. Soil data collected on
81
-------
site will be compared with existing data to determine similarities or dissimilarities and fill
data gaps where EMAP is not collecting soil samples for further analysis. The pilot will test
the access, storage, and retrieval aspects of having EMAP-Arid soils data in the
SCS-National or State Soil Survey databases. This would provide a common database for
use by all agencies.
5.5 REFERENCES
Baily, A.W. and C.E. Poulton. 1968. Plant communities and environmental relationships in a
portion of the Tillamook burn, Northwestern Oregon. Ecology 49:1-13.
Bonham, C.D. 1989. Measurements for terrestrial vegetation. John Wiley and Sons, New
York, 338 pp.
Born, J.D. and D.D. Van Hooser. 1988. Intermountain Research Station remote sensing
use for resource inventory, planning, and monitoring. In: Remote sensing for resource
inventory, planning, and monitoring. Proceedings of the Second Forest Service
Remote Sensing Applications Conference, April 11-15,1988, Slidell, Louisiana.
American Society for Photogrammetry and Remote Sensing, Falls Church, Virginia.
pp 32-34.
Breckenridge, R.P., J.R. Williams, and J.F. Keck. 1991. Characterizing soils for hazardous
waste site assessment, U.S. EPA, Las Vegas, Nevada. EPA/540/4-91/003.
Cihlar, J., L. St.-Laurent, and J.A. Dyer. 1991. Relation between the normalized difference
vegetation index and ecological variables. Remote Sens. Envir. 35:279-298.
Eckert, R.E. Jr., F.F. Peterson, and ST. Belton. 1986. Relationship between
ecological-range condition and proportion of soil-surface types. J. Range. Manage.
39(5):409-414.
Flanagan, D.C., ed. 1990. WEPP Second Edition, Water Erosion Prediction Project -
hillslope profile model documentation corrections and additions. USDA-ARS National
Soil Erosion Laboratory. NSERL Report No. 4.
Flanagan, D.C., ed. 1991. WEPP Version 91.5, Water Erosion Prediction Project - hillslope
profile model documentation corrections and additions. USDA-ARS National Soil
Erosion Laboratory. NSERL Report No. 6.
Fryrenr et al. 1992. Revised wind erosion equation. (In press).
Gholz, H. 1982. Environmental limits on aboveground net primary production, leaf area,
and biomass in vegetation zones of the Pacific Northwest. Ecology 63(2):469-481
82
-------
Hobbs, R.J. and H.A. Mooney, eds. 1990. Remote sensing of biosphere functioning. New
York, Springer-Verlag, 312 p.
Hunsaker, C.T. and D.E. Carpenter, eds. 1990. Ecological indicators for the Environmental
Monitoring and Assessment Program. EPA 600/3-90/060. U.S. Environmental
Protection Agency, Office of Research and Development, Research Triangle Park,
North Carolina, 416pp.
Israelsen, C.E., C.G. Clyde, J.E. Fletcher, E.K. Israelsen, F.W. Haws, P.E. Packer, and E.E.
Farmer. 1980. Erosion control during highway construction, manual on principles and
practices. Transportation Research Board, National Research Board, National
Research Council, Washington, DC.
Kepner, W.G. and C.A. Fox, eds. 1991. Environmental Monitoring and Assessment
Program. Arid ecosystems strategic monitoring plan. U.S. Environmental Protection
Agency, Las Vegas, Nevada (in press).
Lane, L.J. and M.A. Nearing, eds. 1989. USDA - Water Erosion Prediction Project:
hillslope profile model documentation. USDA-ARS National Soil Erosion Laboratory.
NSERL Report No. 2.
Larson, W.E., F.J. Pierce, and R.H. Dowdy. 1983. The threat of soil erosion to long-term
crop production. Science 219:458-465.
Meeuwig, R.0.1979. Growth characteristics of pinyon-juniper stands in the western Great
Basin. USDA Forest Service, Intermountain Research Station Research Paper
INT-238, 22 pp.
Mills, W.B., J.D. Dean, S.A. Gherini, R.J.M. Hudson, W.E. Frick, G.L Rupp, G.L. Bowie.
1985. Water Quality Assessment: a screening procedure for toxic and conventional
pollutants in surface and groundwaters, Parts A and B. Athens, Georgia: U.S.
Environmental Protection Agency. Environmental Research Laboratory. Office of
Research and Development. EPA-600/6-85/002 a and b.
Nemani, R. and S.W. Running. 1989. Testing a theoretical climate-soil-leaf area hydrologic
equilibrium of forests using satellite data and ecosystem simulation. J. Appl. Meteorol.
44:245-260.
O'Brien, R. and D.D. Van Hooser. 1983. Understory vegetation inventory: an efficient
procedure. Research paper INT-323, Ogden, Utah, U.S. Department of Agriculture,
Forest Service, Intermountain Forest and Range Experiment Station, 6 pp.
Peterson, D.L, M.A. Spanner, S.W. Running, and K.B. Teuber. 1987. Relationship of
thematic mapper simulator data to leaf area index of temperate coniferous forests.
Remote Sens. Envir. 22:323-341.
83
-------
Running, S.W. 1990. Estimating terrestrial primary productivity by combining remote
sensing and ecosystem simulation, pp 65-86. in Hobbs, R.J. and H.A. Mooney, eds.
Remote sensing of biosphere functioning. New York, Springer-Verlag, 312 p.
Suter, G.W. 1990. Endpoints for regional ecological risk assessment. Env. Manage. 14(1):
9-23.
Tausch, R.J. and P.T. Tueller. 1990. Foliage biomass and cover relationships between tree-
and shrub-dominated communities in pinyon-juniper woodlands. Great Basin Nat.
50:121-134.
Tausch, R.J. and R.S. Nowak. 1991. Prediction of annual evapotranspiration from leaf
biomass on a burned sagebrush site in western Nevada. Abs. Ecol. Soc. Amer.,
August 1991.
USDI. 1985. Rangeland monitoring trend studies. Technical Reference 4400-4. USDI,
Bureau of Land Management, Service Center, Denver, Colorado, 130 pp.
Utah Forest Survey Field Procedures. 1991-1993. Unpublished document on file at U.S.
Department of Agriculture, Forest Service, Intermountain Research Station, Forestry
Sciences laboratory, RWU 4801, Ogden, Utah, 187 pp.
Waring, R.H., W.H. Emmingham, H.L. Gholz, and C.C. Grier. 1978. Variation in maximum
leaf area of coniferous forests in Oregon and its ecological significance. Forest
Science 24(1 ):131-139.
Wischmeier, W.H. and D.D. Smith. 1978. Predicting rainfall erosion losses - a guide to
conservation planning. U.S. Department of Agriculture, Agriculture Handbook No. 537.
Wood, M.K. 1988. Rangeland vegetation - hydrological interactions. In Tueller, P.T. ed.
Handbook of vegetation science, vol. 14, vegetation science applications for
rangeland analysis and management. Kluwer Academic Publishers, pp. 469-491.
84
-------
6.0 LOGISTICS
(Nita G. Tallent-Halsel!)
6.1 DESIGN CONSIDERATIONS
Field activities for EMAP-Arid will begin in June 1992 with a pilot study in the
southeastern Utah portion of the Colorado Plateau. The following assumptions are being
used for logistics planning and implementation of the Colorado Plateau Pilot Study.
1. The index period will be June, July, and August; a sampling window of
approximately ten weeks.
2. Approximately 40 sites will be sampled.
3. Selection of sample sites does not consider site access.
4. Four-wheel drive vehicles, backpacking, and helicopters will be used for
site access.
5. Field mobile laboratories are not required, sample preparation in the field
will be minimal.
6. Samples will be shipped to the laboratory weekly.
7. A field crew will consist of at least four people: a field supervisor, a soil
scientist, a botanist, and a field technician. A spectrometer technician will
accompany the crew to sites where spectral properties are to be
measured. In addition to the field crews, a field coordinator will be
located at the base site to coordinate activities and communications, ship
data and samples, and obtain site access permission. Responsibilities of
each team member are given in Table 6-1. There will be two crews
employed for the pilot.
8. EMAP-Arid will use and train qualified personnel selected from the
permanent staff of the NPS, EPA, SCS, BLM, and other Federal
agencies.
Based on these assumptions a field crew will evaluate proposed field methods at
each site. The terrain of the sample sites varies greatly, therefore, the best access methods
will also be evaluated. Sampling schedules will be tentative and subject to change as
needed. Sites difficult to access will require additional time and/or staff.
85
-------
TABLE 6-1. RESPONSIBILITIES OF FIELD CREW MEMBERS.
TEAM PERSONNEL
POSITION DESCRIPTION
Field Supervisor (2)
Scientist with responsibility for sampling effort of one crew
(performance & safety). Makes final decisions concerning
in-field protocol deviations or failure to sample site, conducts
and records daily debriefing, interacts with field coordinator,
transmits information for weekly reports, assists with
sampling. Preliminary verification of field forms and sample
condition prior to giving to field coordinator.
Soil Scientist (2)
Performs soil sampling and directs field technician in soil
measurement activities. Responsible for verification and
transmittal of soil data collected and care and maintenance of
field equipment for soil sampling.
Botanist (2)
Performs and directs vegetation sampling. Responsible for
verification and transmittal of vegetation data collected and
care and maintenance of field equipment for vegetation
sampling.
Field Technician (2)
Assists soil scientist (and botanist as needed). Assists field
coordinator with supply inventory and shipping. Responsible
for charging batteries and maintenance of GPS, PDR, and
radios, and care and maintenance of field equipment.
Spectrometer Technician (1)
(for sites where spectral
properties are to be measured)
Performs spectral properties measurements. Responsible for
for care and maintenance of Personal Spectrometer (PS-II)
and downloading data from PS-II. Assists botanist when not
taking spectral measurements.
Field Coordinator (1)
Ensures that permission to access sites has been obtained.
Assumes custody for samples. Processes and ships
samples. Photocopies and transmits completed data forms.
Coordinates daily and helicopter schedules. Maintains
inventory of supplies. Coordinates communications among
field crews and EMAP-Arid management. Arranges for repair
or replacement of equipment.
86
-------
Two field crews will sample all sites within the index period. Subsets of sites will be
sampled and/or resampled to evaluate the utility of various collection and measurement
techniques. An evaluation of activities will be conducted through daily debriefings to ensure
smooth operations throughout the sampling period. A final debriefing after the field study
will be part of the overall evaluation of the field sampling activities.
6.2 SAMPLE SITE ACTIVITIES
Crews will file an itinerary with the Field Coordinator before departing for the site.
Crews will contact the BLM District Office at Moab upon arrival at the site, at midday, and
before leaving the sampling site as a part of routine safety procedures. Daily activities for
the EMAP-Arid sampling crews are detailed in the Field Operations and Training Manual
(Franson and Pollard, 1992) are summarized in the following discussion and outlined in
Figure 6-1. Activities will start each morning by checking and calibrating the instruments
and ensuring (via equipment checklists) that all necessary equipment and supplies are
loaded into the vehicles.
At the site, the field crew will verify their location based on topographic maps,
landscape features, aerial photographs (if available), and GPS information (if satellite
coverage is available). If the site must be accessed by foot or helicopter, crews will
reinventory equipment prior to departure from a suitable vehicle parking or staging area.
Sampling protocols will remain consistent regardless of transport mechanism.
Upon arrival at the site, the formation type will be determined. If the site is Great Basin
Desertscrub or Great Basin Conifer Woodland, it will be sampled; if not, the site will not be
sampled. If the site is Great Basin Desertscrub or Great Basin Conifer Woodland but
cannot be sampled safely (e.g., due to cliff condition), an alternate site will be chosen from
the same formation type within 1 km of the designated site.
The plot will be established and permanently marked and the site characterization
form completed. Eight 35 mm color photographs will be taken to document the location of
87
-------
Check batteries for GPS, PS-II, and Motorola radio
Assemble gear and pack vehicle using daily check lists
File itinerary
Depart for site
Arrive at site
Verify site location
Establish plot
Begin site characterization
Photograph site
SOIL SAMPLING ACTIVITIES
Hole excavation, site photography,
soil profiling, soil sampling, complete
pedon coding form, and plot restoration
VEGETATION SAMPLING ACTIVITIES
Surface characterization, herbaceous
cover measurements, trees and shrub
measurements
SPECTRAL MEASUREMENTS
Begin at 10:30 sampling plots in the
same order as vegetation sampling
Check field forms for completeness
Field supervisor initials
Depart site and relocate to nightly lodging
±
Field forms and sample labels checked
for completeness and accuracy
Download electronic data
Equipment maintenance
Charge batteries
Figure 6-1. Flow Chart of Daily Activities.
88
-------
each site for ease in future visits. Sampling activities as described in Sections 5.2, 5.3, and
5.4 will proceed concurrently. Upon completion of sampling activities, data forms will be
reviewed for completeness, sample labels checked, the plot restored, and all equipment
collected prior to returning to the vehicles.
Upon relocation to nightly lodging, the Field Supervisor will debrief the sampling crews
and check the data forms, sample labels, and the condition of the samples. The sampling
crews will clean and prepare equipment and supplies for the next day. The Field Supervisor
will perform administrative tasks to prepare for ensuing sites.
Data forms will be held by the Field Supervisor until reunion with the Field
Coordinator. Data forms will be reviewed by the Field Coordinator, photocopied, and the
copies mailed to the Information Manager weekly for data entry. Original data forms will be
kept on file in the field office individually for each site. Soil and vegetation samples will be
shipped to the analytical laboratory by the Field Coordinator weekly.
6.3 REFERENCES
Franson, S.E. and J.E. Pollard, eds. 1992. Environmental Monitoring and Assessment
Program: Arid ecosystems 1992 indicator pilot study: Colorado Plateau: field
operations and training manual. U.S. Environmental Protection Agency, Washington,
DC. (In press).
89
-------
7.0 QUALITY ASSURANCE
(Richard D. McArthur)
7.1 INTRODUCTION
It is EPA's policy that environmental data collected under its auspices must be of
known and documented quality. Because of the magnitude and complexity of the EMAP
data collection effort, a quality assurance (QA) plan must be developed that ensures that
the type, amount, and quality of the data are adequate to meet EMAP's objectives. The
draft EMAP Quality Assurance Program Plan describes the overall QA strategy for the
program. EMAP-Arid will have its own specific QA plan, to be set forth in a Quality
Assurance Project Plan (QAPjP). The final document will be completed before the
full-scale monitoring and assessment effort begins. A draft QAPjP for the pilot study is
provided as an appendix to the Field Operations and Training Manual (Franson and
Pollard, 1992).
The pilot study that EMAP-Arid will conduct in 1992 is intended primarily as a field
test of a selected set of indicators. A rigorous QA program is not necessary or desirable at
this stage of the project. Thus, the level and scope of QA for the pilot study will be
considerably less comprehensive than will be necessary for the larger scale efforts to be
carried out in the future. Information obtained in this pilot study will be crucial in further
developing the QAPjP for future efforts.
7.2 APPROACH TO QUALITY ASSURANCE
The QA plan for EMAP-Arid will be based on a philosophy of guidance and
assistance rather than enforcement. The commitment of personnel at all levels to
maintaining and improving the quality of data is a key element of this approach. QA is not
the responsibility of any one person in the program. Rather, the responsibility is distributed
90
-------
among all personnel, each of whom has a specific role. Those roles must be clearly defined
and organized to ensure that an adequate level of quality is attained.
Each indicator lead will be responsible for the data quality for that indicator. This
responsibility will ultimately include defining Data Quality Objectives (DQOs), developing
standard operating procedures, training personnel, verifying data, and planning audits and
performance evaluations. The indicator leads will work closely with QA personnel in
planning and conducting these activities.
7.3 DATA QUALITY OBJECTIVES
EMAP is committed to the use of DQOs as a means of assuring data quality. DQOs
for EMAP-Arid will be defined in accordance with overall EMAP objectives and ecosystem
data requirements. At this early stage of the project, however, information is not adequate
to define DQOs on such a high level. The focus for the pilot study will be on Measurement
Quality Objectives (MQOs).
MQOs are defined for specific measurements. They may address a number of
attributes of data quality, including detectability, precision, accuracy, representativeness,
completeness, and comparability:
• Detectability - the lowest concentration of an analyte that a specific
analytical procedure can reliably detect.
• Precision - the level of agreement among multiple measurements of the
same characteristic.
• Accuracy - the difference between an observed value and the true
value.
• Representativeness - the degree to which the data collected accurately
represent the population of interest.
• Completeness - the quantity of data collected with respect to the amount
intended in the experimental design.
• Comparability - the similarity of data from different sources.
Some of these attributes can be assessed relatively easily, while others (particularly
representativeness and comparability) will likely be extremely difficult and costly to
determine.
91
-------
Each indicator lead will be responsible for defining the MQOs for the measurements
made during the pilot study. If existing data are not adequate for defining an MQO for a
particular measurement, one of the objectives for the pilot study should be to obtain such
data.
7.4 QUALITY ASSURANCE OBJECTIVES
The specific QA objectives for the pilot study reflect the preliminary nature of the
study:
Section 2.2 lists a number of questions that the EMAP-Arid 1992 pilot will address.
The primary purpose of the QA program during this pilot is to ensure that the data collected
during the study are of sufficient quality to be useful in answering these questions.
The general objectives of the QA program may be stated as follows:
• establish criteria to control and assess the quality of data collected in the
pilot study.
• ensure that sampling, analytical, and data management methods and
procedures are documented.
• use assessment samples and procedures to verify the quality of the
data.
• perform field audits to ensure that all activities are properly performed
and that discrepancies are identified and resolved.
• evaluate the QA data and document the results.
• establish procedures for documentation and data verification and
validation.
7.5 QUALITY ASSURANCE DURING THE 1992 PILOT
7.5.1 Training
A crucial element of quality control is sufficient training of the staff. An overall EMAP
orientation and task-specific training program will be conducted before the field study
begins to ensure that the field crews are technically competent and fully understand the
standard operating procedures. The training will be provided primarily by the indicator leads
92
-------
and support personnel. The indicator leads will be responsible for evaluating members of
the field crews after training and certifying that they are able to do the necessary tasks.
7.5.2 Site Location QA
When EMAP-Arid begins regional-scale monitoring, it will be important that the crews
be able to locate predetermined sites with a known degree of accuracy. The sites used in
the pilot study will not be resampled in the future, and locating them accurately is not so
important. The only QA effort directed at locating the sites during the pilot will be to ensure
that the protocol for using the GPS is written and followed. A site check exercise, where
each crew is sent to resample a site already done by the other crew, will test how well a
site can be relocated and thus give some indication as to how much effort should be given
to site verification in future studies.
7.5.3 Vegetation Composition, Structure, and Abundance QA
The quality of the vegetation data will be assessed primarily by resampling. The
resampling program will include two elements: plot checks and site checks. This activity
leads to estimates of crew comparability.
• Plot checks: Both field crews, and an expert if one is available, will visit the same plot
on the same day. All will measure the vegetation on the 6 transects and 4 subplots. Their
results will then be compared to determine how well the field crews agree with each other
and with the expert.
Suggested measurement quality objectives for this exercise will be presented in the
QAPjP appended to the Field Operations and Training Manual. Failure to meet some of
these objectives could mean that the crews need to be better trained or that the objective is
unrealistic. The appropriate action to be taken will be determined by the indicator lead.
Plot checks ideally would be made in each habitat type at the beginning of the study
and during both the third and the final week of the field season. This may not be possible,
but at least two plot checks will be conducted. Because the checks will cause
93
-------
aconsiderable amount of disturbance to the plots, they will be done at sites that are not
part of the main pilot study.
• Site checks: At least once during the field season for each formation type, each crew
will be sent to locate and measure a site that has already been measured by the other
crew. This exercise will not be a QA procedure in the strict sense because the crews will
probably not be sampling precisely the same subplots and transects, and it will not have
any measurement quality objectives. However, it will give a preliminary indication of the
amount of variability to be expected when sites are revisited, information that will be
essential later on when trends and changes are to be assessed.
7.5.4 Spectral Properties QA
The primary procedure for quality assurance of spectral measurements is ensuring
that the PS-II is calibrated properly and frequently. In addition, repeat measurements to
determine diurnal variability will be made on the two or more days devoted to plot checks
by the vegetation and soil crews. On these days, spectra will be recorded at the same spot
(for multiple spots) every 30 minutes from 10:30 am until 3:30 pm to see how much the
readings vary during the course of a day. In addition, the first quadrat of transect AR
(Figure 4-1) will be measured before beginning measurement of each transect and when
all measurements are completed (a total of 4 times during the sampling of the plot).
Variability among operators of the PS-II may also be evaluated during the pilot study.
Precisely how the spectra obtained during the repeat measurements will be
compared, and what criteria will be used to judge how "close" they are to one another, will
have to be specified by the indicator lead.
7.5.5 Soil Properties QA
The consistency of soil descriptions between field crews will be assessed during the
resampling exercises. At the test plot, each crew will dig a 1.5-m deep soil pit and prepare
a site and profile description. The crews will then change places and prepare a description
94
-------
for the other pit. The two descriptions of each pit will be compared by the indicator lead,
who will also establish the criteria for judging them.
Measurement quality objectives for the analytical results will be determined by the
indicator lead and the laboratory. The SCS Soils Laboratory has established QA
procedures for analysis of soil samples.
External QA on the soil analyses will be provided through reference and duplicate
samples. The reference samples will provide a check of the consistency of the laboratory
analyses over the course of the study. The duplicate samples will allow the variability of the
analytical results to be estimated.
The reference soil will be obtained from a soil pit dug during training exercises. A
volume of at least 241 will be removed from the pit, air dried, and mixed as thoroughly as
possible by hand. Eight aliquots of 3 I each will then be sealed in plastic bags similar to
those used to store the samples collected in the field. One bag of reference soil will be
included in each batch of field samples sent to the laboratory.
A duplicate sample will be collected from the deep soil pit dug at each site where a full
profile description is completed. The sample collected from one of the horizons in the pit
will be twice the usual volume. It will be mixed as thoroughly as possible by hand, then
divided between two sample bags. The choice of horizon for the duplicate sample is left up
to the field crew.
The labels attached to the bags of QA soil will be coded in such a manner that the
samples cannot be recognized as QA samples by the laboratory.
Results from the QA samples will be monitored by the QA Coordinator during the
study and reported at the end of the study. No corrective action will be taken unless
extremely unusual results appear.
95
-------
7.5.6 Field Audits
Each field crew will be audited at least once during the pilot study. The auditors will
interview members of the crew to make sure they have a clear understanding of the
standard operating procedures and are complying with them. They will also review data
forms and logs to verify that data are being recorded and documented properly. The audits
will be conducted by people who are not actively involved in the QA program. The results
will be reported to the Technical Director, the QA Coordinator, and the indicator leads.
Several corrective actions can be taken if the audits reveal problems, including
improving staff training and reviewing and revising the standard operating procedures.
Determining and taking the appropriate action will be the responsibility of the indicator lead
in consultation with the Technical Director and Statistician. A follow-up audit will be
performed to verify that the problem has been remedied.
7.5.7 Pilot Study QA Reports
The results of the plot check exercises will be evaluated by the QA Coordinator and
reported to the indicator leads and the Technical Director as soon as possible after each
exercise is completed. Prompt reporting of these results is essential so that any problems
found can be corrected before the field season ends.
After the pilot study is completed, the QA Coordinator will prepare a written report for
the Technical Director. The report will include documentation of changes made in the QA
Project Plan during the study, results of quality assessments and audits, discussion of
problems encountered and their resolutions, and discussion of whether measurement
quality objectives were met.
7.6 REFERENCES
Franson, S.E. and J.E. Pollard, eds. 1992. Environmental Monitoring and Assessment
Program: Arid ecosystems 1992 indicator pilot study: Colorado Plateau: field
operations and training manual. U.S. Environmental Protection Agency, Washington,
DC. (In press).
96
-------
8.0 INFORMATION MANAGEMENT AND GIS
(Carol B. Thompson and Timothy B. Minor)
8.1 ROLE OF INFORMATION MANAGEMENT
Information management plays a significant role in EMAP, EMAP-Arid, and the
Colorado Plateau Pilot Study. Its role is not defined by other aspects of the project. Instead
it shares complementary roles with the research objectives and quality assurance to define
how the project will be implemented. Information Management helps tie the research effort
together from data collection, through data analysis and trend assessment, to the use of
the information by decision makers and the public.
The efforts of Information Management in this pilot will not be as extensive as those
expected in later demonstration studies or in full implementation. Other EMAP resource
groups have had the opportunity to work through some of the same functions that
EMAP-Arid requires and EMAP-Arid will use those lessons learned where they are
applicable. There are specific lessons to be learned by EMAP-Arid because of its unique
monitoring requirements and a team with new partners that needs to learn how it can best
assign functions to make EMAP efforts effective and efficient for arid ecosystems.
One objective of the Colorado Plateau Pilot Study is to be able to determine how well
EMAP-Arid can actualize the functions necessary to accomplish its objectives on larger
scales. For this pilot, Information Management will test the following functions:
a use of GPS,
o utility of forms in the data collection effort,
D use of PDRs in selected data collection efforts,
D transfer of data between the field and central office,
n beginning requirements for the EMAP-Arid Information System,
D beginning requirements for interacting with other agencies
regarding existing data sets and providing assistance with new
97
-------
data sets (e.g., the SCS will be entering soils data during the
laboratory analysis),
D hardware and software requirements,
D information management needs of the design, quality assurance,
logistics, and statistics elements, and
n interactions of data into a GIS for spatial analysis and
cartographic presentation.
The focus will be to test these functions and to have the data available for evaluating
the results of the pilot study. The objective of EMAP-Arid Information Management will not
include testing the distribution of the pilot study data outside of EMAP-Arid.
8.2 OPERATIONAL ASSUMPTIONS
The following assumptions are the basis for the modes of operation to be described in
the procedures and the following sections:
1. The allowable error in locating a site is 100 m. This assessment does not
need to be made in real time.
2. Paper forms will be the primary mechanism for data collection. Paper
forms will serve as a backup mechanism when PDRs are used for data
collection.
3. Transfer of information from the field to the central office can be on a
weekly basis rather than daily basis.
4. Changes to field computer programs will be kept to a minimum after the
beginning of the field season.
5. The data collected will not be considered suitable for distribution as
EMAP assessment data. They will be used to evaluate indicators and as
a measure of performance for the chosen mechanics of implementation
in this pilot.
8.3 OVERVIEW OF INFORMATION MANAGEMENT FUNCTIONS
The functions carried out by information management in this pilot study fall into five
categories: Pre-field; Field; Central Office; External Data Sets; and Assimilation, Review,
and Assessment.
98
-------
The Pre-field functions relate to the tools required for data collection, both paper
forms and software, hardware set-up, and identification and coding schemes. The Field
functions include the use of equipment, methods for tracking the collected information,
backup of information taken, and transfer of information from the field to the central office.
Central Office functions include the tracking of information received, verification of data,
organization and archival of data sets, and transfer of information to the data base.
External Data Set functions include the coordination of the transfer, QA, and assimilation of
data sets that may be used in conjunction with the pilot data or are part of the pilot study
data and computerized by another agency. Assimilation, Review, and Assessment
functions include final documentation of the data processes and support for use of the data
by the researchers and project management.
8.4 PRE-FIELD FUNCTIONS
The functions in this category are developed before the field activities begin or relate
to activities that take place before work on a given site is begun. The following addresses
types of functions in this category for which procedures or documentation will be
developed.
The development of paper forms for data collection and instructions for completing
those forms is a combined effort of the Information Manager, indicator leads, and
Statistician. Forms must be designed to capture all information efficiently in the field and
set up to allow for efficient data entry. All forms are placed in an individual site packet with
the assigned site and soil sample numbers to be distributed to the Field Coordinator
approximately one week prior to scheduled sampling for that site. This is the first step in
data tracking.
Software programming for both PDRs and data entry involves the definition of
requirements, testing before use, development and testing of protocols for running and
using the programs, tracking versions of a program, and documentation (mostly internal,
minimal external) of programs.
99
-------
Hardware set-up activities primarily include developing protocols for the testing of the
GPS, spectrometer, PC, and PDR; the charging or replacement of batteries; reloading
software; and replacement of PDR memory cards.
Identification and coding schemes must be developed to allow for the tracking of film
rolls and site photography; PC disks; PDR memory cards; data forms; soil samples and
voucher specimens of unidentified plant species; sites, transects, pits, plots, and points;
and data file names.
8.5 FIELD FUNCTIONS
The functions in this category are carried out directly in the field or back at the base
site after field activities. These include use of equipment, methods for tracking collected
information, backup of information taken, and transfer of information from the field to the
Central Office. The following addresses the types of functions in this category for which
procedures or documentation will be developed.
Protocols must be developed for proper use of the equipment. These include: testing
or calibration of equipment before use (GPS, spectrometer, and PDR); collection of
measurements with the GPS, spectrometer, and PDR; downloading of measurements from
the GPS, spectrometer, and PDR; uploading programs and data from disk to PC and from
the PC to the PDR.
Tracking information collected in the field is accomplished in part by providing forms
and predetermined site and sample numbers. Protocols have been developed for the
continuation of soils data and vegetation data on additional pages when necessary and are
included in the procedures in the Field Operations and Training Manual.
In addition, instructions for recording vegetation composition, structure, and
abundance data with the PDR will be developed. These procedures will include basic data
collection; completing a partial save of data when there is an interruption of data collection;
handling a total loss of information; the continuation of a data set with PDR/PC; and editing
100
-------
data. Similar procedures for collecting spectral properties data with the PS-II have been
developed along with procedures for naming data files and recording data file names.
Tracking of samples is provided by having soil and plant specimen sample labels
included in each site packet with the basic site number and sample identification included.
A shipping form, included in each site packet, is used to track the samples sent in a given
shipment.
A backup of the information collected in the field is accomplished by providing
protocols for: downloading data from PDR, spectrometer, and GPS to PC; downloading
data from PC to disk; backup of disks; backup of forms; and printout verification of data
collected on PDR.
Protocols for transferring information from the field to the Information Manager include
mailing of disks and forms and mailing sample tracking information to the Information
Manager and to the analysis laboratory with samples.
8.6 CENTRAL OFFICE FUNCTIONS
Functions in this category are carried out by the information management staff after
information is received from the field. Materials, forms, and disks are logged as received.
Data received on forms are computerized and data received on disks are uploaded to the
main system with checks on the files and their sizes. Data in the main system will be
reviewed with computerized checks for completeness and consistency.
Information, forms, and original, edited, and combined files will be catalogued and
archived. GPS information will be combined with site information in the data files.
Relationships between files in a GIS context will be developed by incorporating and
integrating combined files into a GIS data base covering the entire pilot study area.
Transfer of information to the EMAP data base involves the development of record
structure, a data dictionary, a data set catalogue, and a data base scheme with defined
101
-------
relationships. Protocols will be developed for the addition and correction and retrieval of
data in the data base.
8.7 EXTERNAL DATA SETS
There are several functions that will be carried out in the use of external data sets,
including: assimilation of this information in the data dictionary, catalog, and directory
(DCD); tracking the versions of data sets received; assimilation with pilot study data;
making these data available for pilot study measurements; and QA on data sets
(annotations about data quality are stored in the DCD).
8.8 ASSIMILATION, REVIEW, AND ASSESSMENT
Following the pilot, the main task of information management is to provide support for
the reporting process and to document the steps taken and data sets and information
created in the process. The EMAP-Arid information management system will be improved
and expanded based on experience of the pilot study. Interactions with the overall EMAP
information management will be continued so that future data collected by EMAP-Arid may
be made easily accessible by other EMAP groups and interested researchers.
102
-------
9.0 ANALYSIS AND REPORTING
(Susan E. Franson and Robert O. Kuehl)
9.1 INTRODUCTION
The focus of the EMAP-Arid Colorado Plateau Pilot Study is to evaluate selected
indicators and the logistical, QA, and Information Management requirements of
implementing them. Ultimately, the indicator evaluation will determine which of the tested
indicators are retained and moved to a higher indicator category and further tested in a
demonstration project, and which will require more development and testing before
implementation. While this evaluation will rely heavily on statistical analyses of the variance
components of each measurement and indicator, other statistical analyses and subjective
considerations will be taken into account.
The evaluation will answer the questions posed in Section 2, Conceptual Approach.
This section will repeat each question and present the approach for data analysis to
answer each.
9.2 QUESTIONS THAT THE PILOT IS DESIGNED TO ANSWER.
Objective 1: To gather and evaluate information to move selected ecological indicators
from the "research" category to the "development" stage in the indicator implementation
process.
1) Are the indicators for a) spectral properties, b) vegetation composition, structure,
and abundance, and c) soil properties separately or in combination, correlated with
independent evaluations of site condition?
The site component of variance provides a measure of the variability for a measured
characteristic over the sample study area. If site variance is large, then there is variation in
103
-------
that characteristic over the sample area. Relating that variation to the variation in condition
among sites requires an independent measure of condition.
BLM, NFS, FS, SCS, and the Navajo Nation all have condition assessments for arid
lands in the study area. These assessments are based on different sets of criteria and do
not necessarily exist for all of the pilot study plots. Available data from other agencies will
be acquired and the pilot study results will be compared with these other condition
assessments to evaluate the selected indicators for their ability to relate to condition.
Because the data of other agencies are not available at this time, the exact form of these
comparisons cannot be detailed.
2) What is the correlation between the remotely sensed spectral properties data from
AVHRR, MSS, or TM, and field sampling of spectral properties data acquired through field
sampling using a personal spectrometer?
This question is perhaps the most exploratory of all those to be evaluated in the Arid
pilot. As described in the Spectral Properties Section, 36 ground-based measurements will
be collected from each of 7 square grids placed in the subplots. Each grid could represent
a single TM pixel, with 7 pixels of the 9 pixel cluster sampled; data from 3 subplots can be
combined to represent a 4 pixel cluster, with 6 possible resamplings (although the
resamplings would not be independent); or the entire collection of 7 subplots could
represent the 9 pixel cluster. Similarly, the plot data could be tied to clusters of different
numbers of TM pixels (4, 16, 25, etc.). Thus, ground-based measurements can be linked
with the TM measurements. The degree of correlation in each case will be determined.
In these analyses, it is recognized that differences between the spatial resolution of
ground-based measurements and satellite imagery lead to problems of georeferencing the
two measurement types. In general, the center point of the ground-based sampling grid
will be considered to coincide with the center of the TM pixel. The validity of this
assumption can be evaluated in the process of considering the size of the pixel cluster that
is most representative of particular collections of ground-based grid data.
104
-------
For MSS and AVHRR data, the entire plot must be considered as a sample of a single
pixel or the 4 (9,16, etc.) cluster of pixels. TM pixels can be used as samples of MSS data
and TM or MSS pixels can be used as samples of AVHRR data. In this manner, ground
observations can be linked directly with MSS or AVHRR data or indirectly through the TM
data.
3) Do remote measures of spectral properties correlate with other field measured
indicators of vegetation and soils or existing assessments done by the NPS, FS, SCS, and
BLM?
Correlating spectral properties with existing assessments done by the NPS, FS, SCS,
and BLM is part of answering question 1 above. In addition, spectral properties will also be
correlated with the vegetation and soils indicators measured in this field study.
The field estimates of proportion of cover of specific plants and surface types on a site
can be used as weights for the spectra of the component plants and surface types (from
the library of spectra acquired during the pilot study) in an equation to derive an estimate of
the spectrum that a remote sensor will record for that site. A correlation of these derived
estimates with remote spectra for the 40 sites will provide information about how accurately
remote data can measure on-ground vegetation (and surface) composition and
abundance.
Various indices (e.g., NDVI) will be calculated from the remote spectral data. These
indices also can be correlated with the ground-based measurements of soil and vegetation
indicators. The strength of these correlations will determine the feasibility of using remotely
sensed indices for estimating condition synoptically rather than relying solely on
ground-based sampling.
4) Which of the remote platforms (AVHRR, MSS, or TM, or a combination) appears to
be most effective in obtaining the required spatial and temporal data necessary to link
remotely sensed indicators with ground measures and existing data?
105
-------
The discussions in 2 and 3 above apply to data from all three remote platforms.
Comparisons of the results from the three platforms will allow a determination of the relative
precision and accuracy of the various platforms. However, a complete evaluation will
require both statistical analyses of variance, precision, and accuracy of remotely sensed
spectral properties, and a consideration of the relative costs of the imagery from various
satellites. Contemporaneous TM data are very expensive, with MSS and AVHRR data
being much less so. However, if it is determined that remote data do not need to be
current, but can be from previous years, then remote data acquisition cost could be greatly
reduced.
Objective 2: Evaluate the utility of using classified Thematic Mapper (TM) imagery and
other data acquired from the USFWS GAP Program to select frame materials for the pilot
study and future studies and to provide data for extent estimation of arid ecosystems.
5) Do the BLP Biotic Communities Map (Reichenbacher and Brown, 1992) and the
GAP data correctly identify the plant communities found at each of the pilot study sample
grid points? If not, what is the level of misclassification and can this level of
misclassification be compensated?
The Biotic Communities Map of North America developed by Brown and
Reichenbacher (Brown and Reichenbacher, 1992; Reichenbacher and Brown, 1992) has
been digitized. The EMAP-Arid sample grid points will be overlaid on this map, and the
BLP formation type for each point documented. This will be compared with the data
collected from the field visits for all 40 sites, and a percentage accuracy calculated.
Similarly, the vegetation map of the Utah State University Fish and Wildlife
Cooperative Research Unit component of the FWS GAP program will be used to determine
the formation type of each of the 40 pilot sample sites, and a percentage accuracy will be
calculated.
Based on only 40 sites, realistic accuracy assessments of the Biotic Communities
map or the GAP vegetation mapping are not possible. However, such information will be
106
-------
useful in evaluating whether in implementation, certain sites identified as not being arid by
either or both existing maps could be excluded from site visits. This could provide valuable
cost savings, providing that information was not lost.
6) Do the GAP data provide adequate information to describe the extent of arid
ecosystems in the pilot study area?
As above, 40 sites cannot be considered as an accuracy assessment of the GAP
vegetation maps. Indeed, the GAP program is performing its own accuracy assessment.
Rather, a qualitative judgement of whether the GAP program and the EMAP-Arid program
definitions of specific formations types concur will be made. If they are judged to be
sufficiently similar, the EMAP-Arid program will explore the possibility of relying on the GAP
program information to provide estimates of extent of arid ecosystems and subpopulations.
Objective 3: Evaluate sampling plot designs appropriate to the selected indicators.
7) What are the sampling design between site, subplot, and sample variance
components of each of the selected indicators?
The nested structure of the design allows this question to be addressed in a standard
fashion. Sites occur within one of two formation types and subplots are nested within a site,
with many samples being collected from each subplot and in some cases replicate samples
being taken.
Two assumptions will be made: (1) although the subplots are fixed relative to the plot
center point, the subplots will be considered random samples of the plot; and (2) similarly,
the samples will be considered as random samples of the subplot although their locations
are fixed when the subplot is established. The issue in this case is spatial correlation
between observations of samples and subplots and finite versus infinite population
assumptions. These issues will be addressed and spatial correlation determined as a part
of the data analysis, but for the following discussion, random samples (subplots) from an
infinite population will be assumed.
107
-------
Ultimately, observations on samples are averaged over all subplots for a mean value
to represent the site. The variance of a single observation, var(y), is:
var(y) = var(ft) + var(s) + var(sp) + var(b)
where var(ft) is the between formation type variance; var(s) is the site to site variance;
var(sp) is the subplot variance within a site; and var(b) is the sample variation within a
subplot. If replicate samples are included, an additional term for replicate variance is
needed. If only one sample is taken from a subplot (as in the case of some soil
measurements), then the sample variation term is removed.
The analysis of variance components will include the site, subplot, and sample
components. Where appropriate, the contributions of formation type, laboratory replicates,
teams, and time to the overall variance will also be considered.
The components of variance will provide information required to establish DQOs for
implementation. If the measured characteristic is combined with one or more other
measurements to arrive at an indicator value, then the site component of variance for the
indicator is what is of interest and the variance of the individual measurements is of lesser
importance.
8) What are the costs associated with indicator measurement?
Costs to be evaluated include labor, equipment, laboratory analyses, imagery
acquisition and processing, data analysis, etc. Costs will be evaluated relative to specific
sizes of sampling units, subplots, samples, lab replicates, etc., as well as overhead costs
for a site.
The subplot and sample components of variance reflect the precision of measurement
at a site and will be used to arrive at a cost and time effective and statistically efficient
estimate of a measured characteristic at that site.
If ybar is the mean of the observations over all samples on all subplots of a site, then
the variance of the estimate is:
108
-------
var (ybar) = var(b)/r + var(sp)/nr
where n is the number of samples per subplot, r is the number of subplots per site, var(sp)
is the variance of samples within a subplot and var(b) is the variance of subplots within a
site. This equation can be extended for cases where there are replicate samples.
The total cost for measuring a characteristic at a given site is:
C = rC1 +nrC2
where C1 is the cost of establishing a subplot and C2 is the cost of collecting the sample
and measuring or analyzing the sample once the subplot is established. Once the plot
center is located, the C1 cost includes all the time it takes to lay out the subplots and all
other activities up to, but not including, establishing and measuring the samples.
The optimum number of samples per subplot is determined by minimizing the variance
for a fixed cost, or minimizing the cost for a fixed variance. In either case, the optimum
number of samples is
n = sq.rt.(C1*var(sp)/C2*var(b)).
If the allowable cost per site is fixed at CO, then the number of subplots is taken as:
r = CO/(C1 + nC2).
If the allowable variance of the site mean is fixed at VO, then the number of subplots is
taken as:
r = [var(b) + n * var(sp)]/n*VO .
Costs of laboratory replicate analyses can be considered in a similar fashion.
9) What are the optimum numbers of subplots and samples to determine each
indicator?
The pilot design is thought to provide for oversampling of the vegetation and spectral
properties indicators at most sites. Thus, the variance components of the indicators will be
109
-------
closely examined to determine if the number of samples per subplot or the number of
subplots can be reduced for implementation. The radial transects and the exterior transects
subplots for vegetation sampling will be compared for their ability to characterize the site.
BLM sampling procedure for vegetation sampling includes an on-site calculation of
the coefficient of variation after sampling a given number of quadrats. Based on the result
of that calculation, additional quadrats may be sampled to reduce the site variation to a
predetermined acceptable level. This approach has the advantage of limiting sampling to
what is required to achieve a sufficiently low variation. However, the approach means that
indicator values from different sites may not be based on the same support area and thus
adjustments for differing inclusion probabilities are required in combining these estimates
into regional estimates of condition. Through data manipulations after the pilot study, this
approach will be evaluated for its effect on EMAP-Arid data collection.
10) How many sites cross a vegetation/soil complex boundary? Does the addition of
quadrats provide a large enough sample to allow for estimates of the vegetation indicators?
As mentioned in Section 5.3.3 on the sampling design for vegetation indicators,
standard practice of sampling programs of many land management agencies and other
ecological studies calls for shifting sampling plots so that plots are entirely within one
vegetation/soil type. This ensures that the resulting data have adequate precision
(Bonham, 1989) but is not in keeping with the probabilistic nature of the EMAP design. The
pilot study will evaluate a compromise design for plots that cross a vegetation/soil
boundary, leaving the plot in the place predetermined from the EMAP grid and extending
the transects to sample additional quadrats so that at least 30 quadrats are sampled for
each vegetation/soil type that comprises at least 17% of the site.
A count of sites that cross a vegetation/soil boundary will answer the first part of the
question. Data will be evaluated in several ways. The additional quadrats will be excluded
to determine the estimates and their variances that would have resulted from a strict
application of the EMAP design. The data from quadrats that represent the individual
110
-------
vegetation/soil types will be evaluated separately to determine if values for each type are
adequately estimated. Inclusion probabilities needed to combine supplemented plots with
standard plots for "regional" estimates will be determined.
Objective 4: Evaluate the logistical, quality assurance, information management, data
analysis, and reporting requirements and constraints on the pilot study data.
11) What specific logistical constraints restrict the implementation of each indicator?
What logistical attributes favor or enhance indicator measurement (e.g., use of a
helicopter)?
This will be a qualitative evaluation based on notes from field log books, crew
debriefings, etc. The evaluation will include: site access, both permission and accessibility
constraints; performance of protocols for measurements; time requirements for plot
sampling, traveling to sites, calibration of instruments, site restoration, data and sample
processing, etc.
12) Based on the results of the pilot study, can data quality objectives be established
for each indicator tested?
Formulating data quality objectives depends upon estimates of the variance
components of each of the indicators. Whenever possible, variance estimates will be used
to establish data quality objectives for the indicators.
13) Does the information management system effectively and efficiently provide for
the movement of data from the field to the analysis stage?
Again, this will be a qualitative judgement based on the results of the pilot study. For
the pilot, it is expected that data for at least one indicator will be recorded electronically
using Personal Data Recorders (PDR). The operation of these PDRs and steps required to
upload data to Personal Computers for storage on disks will be evaluated. Time and cost of
entering data collected onto field data collection forms will also be evaluated.
111
-------
During the process of data analysis, the Information Management system will be
further evaluated for its ability to provide data in a reliable and efficient format. During this
process, the benefits of GIS for data evaluation and display will also be evaluated.
14) Do the methods of collecting and analyzing data meet the reporting requirements
for an EMAP pilot study?
The results of the pilot study will appear in a report (described below) compiled after
the data collection and sample and data analyses are completed. This report will be
subjected to peer review to determine if the pilot study met its objectives. This process will
determine how well the methods for collecting and analyzing data meet the reporting
requirements for EMAP. A goal of EMAP is to have statistical summaries available nine
months after the field activities for that season are complete. The process of data analysis
and reporting for this pilot study should lead to the development of computer programs that
will aid EMAP-Arid to meet this nine-month reporting goal in the future.
15) What are the special logistical requirements involved with fielding multi-agency
sampling crews?
This will be a qualitative assessment derived from experiences during the pilot study.
9.3 REPORTING
Following the pilot study, sample analysis, and data analysis, a report will be prepared
presenting the results of the pilot study. The questions presented above will be addressed
and any additional information that is relevant to indicator evaluation will be included. This
report will be peer reviewed and published as an interagency effort within the EPA
publication series.
Perhaps the most important result of the pilot study will be the information that will go
into the development of the demonstration study. It is expected that the implementation
plan for the demonstration study will build on this pilot study plan and incorporate many of
112
-------
the results of this pilot. Thus, the implementation plan for the 1993 demonstration could be
viewed, in part, as a report of this pilot study.
The Field Operations and Training Manual (with its component Safety Plan and
Quality Assurance Plan) developed in draft form for this pilot study will be revised on the
basis of the findings of this study and will form the basis for a similar manual for the 1993
demonstration study.
No attempt will be made to make an assessment of the condition of the study area,
the Colorado Plateau, or arid ecosystems. However, the data collected during the pilot will
be used to begin to design the formats for Yearly Statistical Summaries and Interpretive
Reports that are to be the products of EMAP-Arid implementation. This exercise will allow
analysis programs to be written; formats, graphics, and tables for data presentation to be
developed; and existing data needed for interpretation to be collected. These activities will
speed the reporting process in the future so that EMAP-Arid can meet the EMAP goal of
producing reports of results within nine months from the end of field data collection.
9.4 REFERENCES
Bonham, C.D. 1989. Measurements for terrestrial vegetation. John Wiley and Sons, New
York, 338 pp.
Brown, D.E. and F.W. Reichenbacher. 1992. A classification system and map of the biotic
communities of North America. U.S. Environmental Protection Agency. Las Vegas,
Nevada. (In press).
Reichenbacher, F.W. and D.E. Brown. 1992. Biotic communities of North America, Central
America, and the Islands of the Caribbean Sea. Map 1:8,000,000. U.S. Environmental
Protection Agency. Las Vegas, Nevada. (In press).
113
-------
APPENDIX A
CANDIDATE INDICATORS
1.0 INTRODUCTION
The Implementation Plan describes those indicators that will be a part of the 1992
EMAP-Arid Colorado Plateau Pilot Study. In the process of discussions and workshops,
several other indicators were proposed that have great promise for monitoring arid
ecosystem condition. However, budgetary constraints limited the number of indicators that
could be field tested to the three described in Section 5.
This appendix presents information on two indicators (landscape and retrospective)
that the EMAP-Arid group intends to develop for implementation in the future. In both cases,
these indicators could be derived in part from existing data or data being collected in the
field as part of other indicator measurements. Therefore, they could be developed for the
Pilot study after the field season, should funding allow.
2.0 LANDSCAPE INDICATORS
(James D. Wickham and Vern Meentemeyer)
2.1 Introduction
Landscape ecology has been described as the investigation of the ecological
consequences of the spatial mix (Milne, 1988) and temporal dynamics of land cover patterns
(Risser et al., 1984). This concept has been incorporated into the EMAP indicator selection
process by categorizing landscape indicators as exposure/habitat rather than response
indicators (Hunsaker and Carpenter, 1990). Thus, landscape indicators will serve as
measurements that characterize the environmental "setting" under which response
indicators, that more typically measure biotic and abiotic processes, are operating (Table
A-1). By using landscape indicators in this fashion, they can be combined with response
indicators to address EMAP's second objective to monitor indicators of pollutant exposure
and habitat condition and seek associations between human-induced stresses and
ecological condition (Hunsaker and Carpenter, 1990).
114
-------
TABLE A-1. LEVELS OF CONSTRAINT FOR A SPATIAL HIERARCHY OF ECOSYSTEMS
CONSTRAINTS
Climatic heat, moisture, seasonally
Climatic heat, moisture, seasonality,
topography, drainage, geology
Climatic heat, moisture, seasonality,
topography, drainage, geology, local
topography (slope, orientation,
exposure, land use, land drainage,
and soils)
Climatic heat, moisture, seasonality,
topography, drainage, geology, local
topography (slope, orientation,
exposure, land use, land drainage,
soils, nutrients/food, habitats, water,
and competition)
ECOLOGICAL
CLASSIFICATION
Biome
Region
Landscape
Community
EXAMPLES
Arid, Schlerophyll
Sonoran Desert
EMAP hexagon,
watershed
cottonwood-willow
An important challenge in the selection of landscape indicators is the relative newness
of some of the proposed indicators. Many are in the research category. EMAP-Arid
examines these indicators and argue for the selection of a few promising indicators, and
proposes measures of the parts of landscapes in the more arid portions of the U.S. which
are vulnerable to change and degradation.
Another challenging problem is the inclusion of the spatial dimension in the models and
descriptions of how ecosystems function across landscapes. Function across a spatial scale
is a problem somewhat different from the point measurements proposed for the other
portions of EMAP. Furthermore, one must deal with processes that "appear" to be slow.
Small (point) systems appear to change rapidly, but big systems such as landscapes appear
to change slowly (Meentemeyer and Box, 1987).
With the change in scale from individual organisms and species comes the task of
identifying higher level, aggregated units suitable for landscape scale analysis. A critical
problem is the degree to which changes in these units represent true changes in the
115
-------
landscape. This change may be "natural" or it may be anthropogenic. Simple change may
not be degradation. The first step is to select a landscape metric from the several that have
been developed (e.g., O'Neill et al., 1988; Forman and Godron, 1986; Burgess and Sharpe,
1981; Patton, 1975). These metrics link landscape patterns with ecological processes
(Turner, 1989).
2.1.1 Levels of Constraint
To better visualize the unique problem of identifying landscape metrics, a hierarchical
framework is beneficial. For each level, levels of constraint can be identified (Table A-1).
These constraints (or constraint variables) dictate the location, extent, and spatial
characteristics of the system at that level. The spatial characteristics of a biome are
controlled mostly by the broad scale constraints of climate (i.e., temperature, moisture, and
seasonality). For a biotic region such as the Great Basin, spatial patterns may be set by
climate, as well as additional factors such as geology, topography, and drainage. For a
landscape within this region, additional controls on spatial patterns could be attributed to
local topography and drainage, soils, and land use. It is at this scale that the spatial patterns
regulated by people become most evident. Within a landscape, communities and
populations can be identified as well as the detailed ecological processes in operation.
It seems logical then to identify metrics at the landscape level and to search for
information about landscape change by examining selected species at the level immediately
below — communities or land use types. For communities this should perhaps be the
common but vulnerable systems, such as riparian systems. A common land use type in the
EMAP-Arid region is grazing.
2.2 Landscape Indicators in EMAP-Arid
2.2.1 Introduction
Six landscape indicators are proposed to be evaluated by the EMAP-Arid group. Listed
in order of their priority for evaluation, they are: (1) habitat/cover type proportions, (2) spatial
distribution of agriculture and riparian vegetation per stream reach, (3) fractal dimension, (4)
116
-------
abundance/density of key physical features, (5) spatial distribution of grazing intensity, and
(6) riparian condition. These proposed indicators are rank ordered according to the degree
of development and the cost/benefit of implementation. Riparian condition, spatial
distribution of grazing intensity, and spatial distribution of agriculture and riparian vegetation
per stream reach are newly proposed; the remainder have been previously proposed for
EMAP and are discussed in detail in Hunsaker and Carpenter (1990).
2.2.2 Habitat/Cover Type Proportion
Habitat or cover types can be mapped and the proportion of the area in each type
calculated. From such measures landscape diversity and the change among types can be
assessed. It is a prerequisite for calculation of the connectivity of habitat types across
landscapes (Hunsaker and Carpenter, 1990). This may be the most important and basic of
all landscape metrics. It provides perhaps the most basic indicator of landscape change.
Land use and its change can be viewed as a zero-sum game because a change in one type
must occur at the benefit or cost of other types. Such changes can be used to determine
trends and the broad-scale geography of these trends across the EMAP-Arid region. It is
likely that this metric is exceptionally sensitive to human activities. It is robust because it can
be universally applied, it is simple to understand, and the types in a region not only can
change, they can be newly added or disappear entirely. Furthermore, land use and cover
types have a large impact on the distribution of pollution and sediment source and sink
patterns.
The habitat proportions will be generated using standard spectral pattern recognition
techniques applied to Landsat Thematic Mapper (TM) data. This technique encompasses
generation of spectral statistics using either a supervised or unsupervised approach to
represent the land cover categories being mapped. These statistics are then used as input
into a maximum likelihood classification algorithm. (A complete description of these image
classification techniques is provided by Lillesand and Kiefer, 1979).
117
-------
EMAP-Landscape Characterization (US EPA, 1990) has proposed a combined
supervised-unsupervised approach, in which a clustering algorithm (unsupervised) is used
to generate spectral statistics which are then matched to the desired land cover categories.
These spectral statistics are then augmented by analyst-derived (supervised) spectral
statistics. Both sets of statistics are then used as input into the maximum likelihood
algorithm.
2.2.3 Spatial Distribution of Agricultural and Riparian Vegetation Per Stream Reach
White et al. (1981) estimated increases in sediment yield, nitrogen (N), and
phosphorus (P) as a result of increases in the amount of land devoted to crop production in
Georgia between 1973 and 1976. Sediment yield was estimated to have increased 95
percent on lands newly devoted to crop production. The position of land cover types (e.g.,
forest, residential, agriculture) can have a significant effect on the net release of these
materials into the watershed. Peterjohn and Correl (1984) have shown that the presence of
riparian vegetation between streams and cropland significantly reduced N, P, and sediment
input to the stream compared to values expected with no riparian vegetation (based on the
export rates of these materials from the cropland). The studies of McColl (1978), Schlosser
and Karr (1981a,b) and Lowrance et al. (1984) have produced similar results.
The relationship between terrestrial/aquatic nutrient flux and land use pattern provided
by these studies suggests that a landscape measure that incorporates the amount of
agriculture and riparian vegetation should serve as an index of exposure against which
water quality sample data can be associated. Using Geographic Information Systems (GIS),
a data base could be constructed containing land cover, streams, watershed boundaries,
and soils and topographic information (at a minimum). Several indices could be generated
using these data, such as total stream length minus the stream length bordered by riparian
vegetation, or the ratio of stream length bordered by agriculture (cropland and rangeland) to
the stream length bordered by riparian vegetation. For example, to estimate the ratio of
riparian-bordered stream length to agriculture-bordered stream length for an entire
118
-------
watershed, a GIS proximity analysis could be applied to the stream data, and then this data
could be overlaid with the land cover data. The result of the GIS overlay analysis would then
be "clipped" to the extent of the watershed to determine the ratio values.
2.2.4 Fractal Dimension
Fractal dimension measurements have been used in conjunction with geostatistics to
interpret the complexity of environmental gradients (Phillips, 1985; Palmer, 1988), to
estimate home range or territory size of raptors (Pennycuick and Kline, 1986; Loehle, 1990),
and perhaps most commonly to describe landscape geometry (Turner, 1990; O'Neill et al.,
1988; Krummel et al., 1987; Gardner et al., 1987). The interest in describing the geometry of
land cover patterns arises from the observation that human and natural processes appear to
operate at different scales (Krummel et al., 1987). In general, complex patch shapes have
been associated with natural processes (or the absence of human influence), while more
simple, regular shapes have been associated with anthropogenic processes.
Krummel et al. (1987) have provided the first empirical evidence of reduction in size
and simplification of patch shape as a result of anthropogenic impact by computing the
change in fractal dimension of forest patches as a function of patch size. A sharp increase in
the fractal dimension was found at a patch size range of 60-73 hectares. Since patches
smaller than the 60-73 hectare threshold range were significantly less complex in shape
than patches above this threshold range, it was inferred that these smaller patches were
largely the result of anthropogenic processes. As patch size decreases, nutrient cycling
patterns may change as a result of changing edge to interior ratios; species typically found
at the forest's edge may have a greater likelihood of being found at more interior positions
within the patch; species diversity may decrease (Forman and Godron, 1986); and stand
dynamics may change. Rex and Malanson (1990) note that king-nut hickory (Carya
laciniosa) and swamp white oak (Quercus bicolot) develop only on interior sites and are
relatively rare in Iowa.
119
-------
It is proposed that for each natural cover type category the fractal dimension of the
"individuals" or patches of that cover type be computed to generate an index of
anthropogenic exposure. Per cover type fractal dimension will be calculated by regressing
the log transformation of patch area (the "X" or independent variable). Starting with a 10
percent subset of the smallest patches the fractal dimension (equal to 2x the slope of the
equation) will be iteratively computed by replacing the smallest patch with the next largest
patch. The change in fractal dimension will then be plotted against the change in patch size
to determine at what patch size anthropogenic impact is most apparent. The methodology
follows the approach taken by Krummel et al. (1987).
An additional benefit of this technique for calculating fractal dimension is that it requires
that patches are sorted by size. Information on the frequency distribution of patch size will
likely provide a picture of landscape heterogeneity. Patch size as an indicator was
previously recommended and discussed in Hunsaker and Carpenter (1990).
2.2.5 Abundance/Density of Key Physical Features
The abundance/density of key physical features as a landscape indicator is justified
because certain physical features such as rock outcrops, cliffs, springs, and talus slopes
and structural elements such as downed logs can control animal diversity. For example,
Short (1986) has noted that interpreted aerial photographs are useful for determining the
presence of high cliffs and cliff faces, which are habitat for peregrine and prairie falcons
(Falco peregrinus and F. mexicanus, respectively). However, the cost of obtaining the
necessarily large scale aerial photography (to interpret such features) and EMAP's
systematic sample design are not optimal for program-wide implementation of this
measurement. This measurement would best be implemented on a species specific basis,
selecting a keystone, endangered, or threatened species within the study region (for which
there is time series population data of good quality), with the acquisition of the aerial
photography directed by the synoptic coverage of satellite data and the species known
120
-------
habitat orientations. In the end, these prerequisites must be aligned with the EMAP
sampling design.
2.2.6 Spatial Distribution of Grazing Intensity
Eighty-six percent of western and Great Plains ecosystems are grazed by domestic
livestock (Short, 1986), but little is known about spatial distribution of range use (Senft et al.,
1985). Senft et al. (1987) have postulated that domestic sheep and cattle and many other
herbivores exhibit landscape-scale matching — a diet selection behavior where feeding
areas are chosen based on available plant communities and other pasture characteristics.
Senft et al. (1985,1983) applied this postulate in constructing regression equations to
predict the spatial distribution of range used by cattle. These are some of the first studies to
quantitatively predict cattle spatial distribution. Smith (1988) has successfully applied this
technique on control and experimental pastures in Australia. The results of these studies
suggest that grazing impacts are not uniform, but instead are controlled by such factors as
distance from water, topography, and availability of preferred browse.
These regression techniques are constrained, however, by having only local
application and the cost of collecting of input and verification data (Smith, 1988;
Rittenhouse, pers. comm.). By building a GIS model, based on the knowledge gained from
these and other studies that rank orders areas according to their use intensity, the obstacles
of cost and local applicability should be overcome. Such a model would not be unlike the
GIS model constructed by Chuvieco and Congalton (1989) for forest fire hazard mapping. In
this study, vegetation, slope, aspect, proximity to roads, and elevation were rank ordered (in
this listed order), weighted as 100, 30,10, 5, and 2, respectively, and used to generate a
hazard index according to the equation:
H = 1 + 100v + 30s + 10a + 5r + 2e.
The coefficients for each category were assigned a value of 0,1, or 2 based on the
attributes of that category for each cell.
121
-------
This ranking concept could be applied to develop a grazing intensity index. Based on
previous studies, the important categories for development of this index would include
distance from water, vegetation type or Resource Value Rating (RVR; BLM, 1984), distance
from salt, and slope (Roath and Krueger, 1982; Senft et al., 1983 & 1985; Smith, 1988;
Owens et al., 1991). The models could then be verified by field sampling fecal pat densities
(Lange and Willocks, 1978; Senft et al., 1983).
The indicator measurement is in the developmental stage, and therefore requires initial
testing on just a few hexagons. Construction of the model would require large scale aerial
photography concomitant with detailed vegetation mapping, topographic data at a scale
matching (or nearly so) that of the aerial photography, and data on location of water and salt
sources. After acquiring these data, actual construction of the GIS model remains. The most
efficient approach to testing the model would be to vary the weighting factors and then test
the resultant accuracies using the above mentioned field reconnaissance of fecal pat
densities.
2.2.7 Riparian Condition
Riparian vegetation in the western United States has been reduced to as little as 2 percent
of its original extent, largely as a result of dam construction (Johnson and Simpson, 1985).
Citing previous studies, Rex and Malanson (1990) report that only 23 percent of the
pre-European settlement forests in Iowa remain. Removal of riparian vegetation constitutes
elimination of perhaps the sharpest, naturally occurring environmental gradient in the
western United States (Gregory et al., 1991). Kauffman and Kreuger (1984) note that
western riparian vegetation supports higher plant and animal productivity, and higher
diversity, while occupying the smallest areal proportion of the western landscape. Given the
biological importance of riparian ecosystems and such extensive removal, the simple
measurement of remaining riparian extent should be supported by determination of whether
or not these remaining systems are in a self-maintaining state and measurement of
variation in habitat quality.
122
-------
The methodology for assessing the condition of extant riparian systems centers on
construction of age distribution diagrams for important riparian tree species.
Self-maintenance can be determined by constructing age distribution diagrams for dominant
species, where descending J-curves represent a self-maintaining state (Whittaker, 1975).
These population curves would then be related to the following landscape measures: patch
size, patch distance to stream, and amount of anthropogenic edge.
Rex and Malanson (1990) noted that along the Iowa and Cedar Rivers human impact
was more strongly correlated with riparian forest patch shape than measures of fluvial
geomorphology. The authors also noted that human impact resulted in reduced patch area,
and that some rare riparian tree species (Quercus b/co/orand Carya laciniosa) only
occurred on interior sites.
Brady et at. (1985) have hypothesized that mature cottonwood-willow communities
form through coalescence of seedling-stage sand bars which were previously colonized by
seepwillow and other shrubs. With a history of light and moderate flood volumes, these sand
bars continue to aggrade and ultimately grow together at some distance from the stream.
Glinksi's (1977) observations that all seedling cottonwoods occurred immediately along the
watercourse support this hypothesis. Brady et al. (1985) also noted that mature
cottonwood-willow stands showed no regeneration. Gebhart et al. (1990) have suggested
that mature cottonwood-willow stands will be succeeded by ash or other self-perpetuating
species. The measurements proposed to assess riparian stand development (discussed
below) follow these authors.
Given these constraints to the formation of cottonwood-willow riparian communities,
determination of a self-maintaining state requires time series mapping to document the rate
of formation of serai and mature stages and relating this to flood history information.
Through time series analysis, if it can be shown that seedling-stage sandbars coalesce to
form mature cottonwood-willow formations, then interpretation of aerial photographs for the
123
-------
presence of these sandbars should prove to be a reliable indicator of a self-maintaining
state.
The Habitat Linear Appraisal System (HLAS) should be considered as part of the
riparian condition indicator because it serves as an exposure/habitat indicator for field
measurement in riparian areas to determine variation in habitat quality, HLAS is a technique
for calculating vertical and horizontal structural diversity of vegetation using point intercept
techniques; it provides a comparatively simple way to calculate the vertical and horizontal
dimensions of habitat.
HLAS measures density, frequency, and dispersion of vertical diversity (understory,
midstory, overstory). It can be applied on a grid or transect design. For EMAP Arid, transects
could be established across the width of the floodplain and measurement points be
established that are appropriate to the community being sampled.
3.0 RETROSPECTIVE HISTORY
(Harold C. Fritts, Martin R. Rose, and Peter E. Wigand)
3.1 Introduction
A suite of retrospective historical and paleoenvironmental indicator measurements
offers an objective, repeatable, and quantitative method for determining trends in ecosystem
health. They also provide a framework for the understanding of natural variation and climate
conditions. Instrumental hydrometeorological data, tree-ring series, pollen data,
macrobotanical remains from packrat middens, and charcoal record can be used directly or
indirectly to produce retrospective reconstructions of climatic, vegetative, and disturbance
histories. Direct interpretation by statistical analysis of the observations available for each
retrospective measurement, or indirectly by calculating its covariance with other response,
exposure, or stressor indicators, can provide information about the mean of the process, its
variability, and long-term trends, i.e., normal variation. In addition, because retrospective
124
-------
measurements offer a lengthy temporal perspective, they are helpful in formulating
conceptual models of environmental change.
Retrospective indicators put current phenomena into the context of a longer
time-frame, thereby providing the opportunity to evaluate slowly operating processes, past
cyclic changes or unusual events, disturbance regimes, and historically constrained
phenomena (Schoonmaker and Foster, 1991). The primary advantage of retrospective
indicators is that they allow probabilistic evaluations of observations obtained in the first year
or two of monitoring. They are also applicable to and compatible with the same type of
observations made in other regions (cross-cutting); they make use of existing databases;
they are scalable up and down in compatibility with the tessellated sampling design; and
models can be created to establish their covariance with other non-retrospective response,
exposure and stressor indicators.
The indicators proposed for evaluation include 1) tree-ring series, 2) instrumental
hydrometeorological data, 3) pollen data, 4) macrobotanical remains from packrat middens,
5) charcoal record, 6) stable isotopes, and 7) repeat photography.
3.2 Retrospective Indicators
3.2.1 Tree-ring Series
Introduction - Tree rings from core samples can be visually compared to past rings on the
same specimens and to dendrochronologies for the area to identify growth increases,
decreases, or other morphological changes. A variety of ring features can be used to detect
a wide range of environmental changes. Where changes in ring characteristics are noted in
many individuals and sites, elaborate tests can be made using techniques from
dendrochronology, chemistry, isotopic analysis, statistics, and tree-ring modeling to help
identify the nature of the change and probable causes.
Time series of growth ring widths sampled from woody plants growing on climatically
stressed sites provide a proxy of past climatic variability, including seasonal and annual
temperature, precipitation, drought, and stream discharge. The long reconstructions (from
125
-------
500 years to several thousand years) provide a sound basis for obtaining more reliable
estimates of central tendency, variability, and time series characteristics than the normally
short period of instrumental data. These reconstructions of past climate provided by tree and
shrub rings can also be calibrated with other indicators of paleoenvironmental change
sampled at reasonably high frequencies.
It is proposed that tree rings be included as a core indicator because they integrate
factors limiting growth throughout the past life of the plant. EMAP-Arid program will be
considering energy and material balance measurements such as the Bowen's Ratio, the
atmospheric carbon dioxide record, the water balance, satellite information such as the
greenness index, productivity of the ecosystem, and leaf production, v/hich ultimately limit
physiological conditions in the plant controlling ring growth. The ring features that integrate
these causal factors can, in turn, be compared to the same features in past years to detect
changes and trends. Other than tree-ring data, no other core indicator will combine the
integration of current environmental conditions with a record of past conditions so that a
change in the current year can be detected without waiting for future measurements. The
isotopic composition of the rings will add information on water use efficiency and water
stress. The tree-ring information can be calibrated with a variety of related variables and
modeled to help evaluate whether the changing conditions are important and statistically
significant. The tree rings can reveal decade and century long variations while other
retrospective indicators can be used to relate the tree-ring variations to century and
millennia long changes in climate, fire occurrence, and community structure.
Background - Most woody plants from temperate regions produce distinct annual growth
layers in their stems and roots in response to the accumulation of photosynthates,
absorption of mineral nutrients, and water status of the plant (Fritts, 1976; Schweingruber,
1988; Baillie, 1982; Cook and Kairiukstis, 1990). The annual growth rings are an important
component of biomass, reflecting productivity of the local environment (Graumlich et al.,
1989). When environmental conditions are nominal, few environmental factors are limiting,
126
-------
processes are optimal, many cells are produced, and large rings are formed often with a
smooth transition from earlywood to latewood. When one or more limiting factors are
subnominal for a long period of time, the rates of processes affecting cell production,
enlargement, and wall thickening are more limited. Fewer cells are produced, and narrow
rings are formed. Often distinct cellular features are produced that indicate when and what
kind of factors have limited growth (Fritts, 1976; Schweingruber, 1988; Fritts, 1990; Fritts et
al., 1991).
It is relatively easy to observe ring width, which reflects the number of cells that are
produced in each growth layer (Fritts et al., 1991). Potential causes of ring width change
include: (1) temperature, moisture, sunshine, and carbon dioxide variations important to the
photosynthetic and growth processes, (2) toxic substances that inhibit growth (Cook, 1987;
Fox, 1980; Fox et al., 1986; Kienast, 1985), (3) disease (Schweingruber, 1988), (4) insect
infestation (Blais, 1958; Brubaker and Green, 1979; Swetnam and Lynch, 1989), (5)
defoliation by ice storms or hail (Travis et al., 1989), (6) fire (Swetnam, 1990; Swetnam and
Dieterich, 1985; Swetnam and Betancourt, 1990.), (7) earthquakes (Jacoby et al., 1988;
Atwater et al., 1991), (8) landslides and other geomorphic changes (Scuderi, 1984), (9)
floods (Gottesfeld and Gottesfeld, 1990; Henoch and Parker, 1972; Hupp, 1988; Payette,
1980; Yanosky, 1982 & 1983), (10) frost (LaMarch and Hirschboeck, 1984; Stahle, 1990),
and (11) a variety of other disturbances both natural and anthropogenic (Brubaker, 1987;
Conkey, 1984; Larson, 1990; Leavitt and Long, 1987; Payette et al., 1990). These include
the effects of forest management practices (Crone, 1987), fertilization (Thompson, 1981),
wind (Robertson, 1986 & 1991; Shiyatov, 1990; Wade and Hewson, 1979), as well as
pollution.
A sampling of the many tree ring studies of pollution includes: (Alekseyev, 1990;
Alekseyeve et al., 1988; Arndt and Wehrle, 1982; Baes and McLaughlin, 1984; Baes et al.,
1984; Bauch et al., 1985; Cain, 1978; Cook et al., 1987; Freyer, 1979; Gemmil et al., 1982,
Greve et al., 1987; Havas and Huttunen, 1972; Innes and Cook, 1989; Kienast, 1985; Lepp,
127
-------
1975; McClenahen and Dochinger, 1985; Mclaughlin etal., 1982; Peterson, 1985; Peterson
and Wakefield, 1987; Peterson et al., 1987; Phillips et al., 1977; Puckett, 1982;
Schweingruber, 1987; Serre-Bachet, 1987; Symeonides, 1979; Tan, 1980; Tessier et al.,
1990; Tian and Lepp, 1977; Treshow et al., 1987; Valkovic et al., 1979; Vins and
Pollanschutz, 1977; Waring, 1987; Yokobori and Ohta, 1983).
Often features other than ring width are clearly visible, such as frost rings, false rings,
eccentric reaction wood, sudden growth changes, and scars due to many causes, such as
fire, flooding, impact, or stripping of the bark. A simple hand lens can be used to identify
these features from cores and disks, and with training, important features can be identified
and recorded while in the field (Schweingruber, 1988; Fritts and Swetnam, 1989; Cook and
Kairiukstis, 1990).
For example, Schweingruber et al. (1986) used a rapid sampling and analysis
approach to study forest decline in Switzerland. They surveyed forest conditions by
observing the surface of samples while in the field, including 6,000 cores and 3,000 stem
disks, and they determined the percentage of trees exhibiting reduction or recovery together
with the date of onset and duration of each growth change. Also, Baillie and Munro (1988)
deduced large-scale climatic changes attributed to large volcanic eruptions from abrupt
changes in dated ring widths from wood buried in bogs in Europe.
Data Analysis - A great deal of information from Retrospective Indicators (RIs) such as tree
or shrub rings can be deduced from features along a core or cross section by simply using a
hand lens. Most conifers (Pinus, Junperus) and deciduous angiosperms (Quercus, Populus,
Juglands, Fraxinus) can provide reliable information, and shrubs like big sagebrush
(Artemisia tridentata Nutt.) and antelope bitterbrush (Purshia tridentata (Pursh) DC.)
produce well defined annual layers (Roughton, 1962 & 1972). Others, such as true mountain
mahogany (Cercocarpus montanus Raf.) offer potential but are difficult to work with
(Roughton, 1972). Shrubs of tropical origin may not produce annual rings, but techniques
128
-------
have been developed (Flinn et al. personal communication) for bringing out growth rings in
mesquite (Prosopis glandulosa var. glandulosa Torr.).
Any abrupt changes in growth would be clearly visible as a marked reduction or
increase in ring width of the outermost rings, and the percentage change in ring size can be
approximated by eye, tabulated, and averaged to reveal the extent and direction of the
changes (Schweingruber et al., 1986). The investigators making these surveys must be
sufficiently acquainted with dendrochronological procedures and materials to crossdate
them visually so that missing and double rings are identified and anatomical anomalies
within the annual growth layers that indicate stress can be recognized.
The cores and cross sections gathered from the tessellated sampling scheme that
show interesting changes can be processed further in the laboratory to gather more
information on the possible causes. Various analyses might be done depending upon the
species and type of change that was observed, drawing from techniques of
dendrochronology (Cook and Kairiukstis, 1990), densitometry (Schweingruber, 1990), image
analysis (Jagels and Telewski, 1990; Vaganov, 1990), chemistry (Baes and McLaughlin,
1984; Baes et al., 1984; Guyette et al., 1989 & 1991), isotopic analysis (Leavitt and Long,
1987; Leavitt. in press; Martin and Sutherland, 1990; Pilcher, 1990; Waring, 1987), statistics
and modeling. The objective of these analyses would be to identify causal agents, quantify
the relationships, examine longer-term variations, and test the changes for significance.
In addition, some fundamental baseline investigations will be necessary to test
hypothesized changes. These may require further collection of replicated tree ring or
paleoenvironmental samples from relatively undisturbed habitats. Analysis may include (1)
various biophysical phenomena and observed cellular or chemical features mentioned in the
previous paragraph but observed in the natural system, (2) statistical transfer function
studies to relate and calibrate regional-scale Rl data to large-scale biological,
hydrometeorological, or other environmental indicators important to the EMAP program
(Briffa et al., 1988; Briffa et al., in press; Fritts, 1976; Fritts et al., 1990; Fritts, 1991), and (3)
129
-------
simulation modeling of the tree ring growth and environmental variables to build a predictive
capacity based upon known biophysical relationships (Dixon et al., 1990; Fritts, 1990;
Vaganov, 1990; Fritts et al., 1991; Gay, 1989).
Chemical composition of rings including stable isotopes of carbon, hydrogen, and
oxygen, detailed cell structure, density, and other features attributable to specific causes
requires laboratory preparation, chemical or dendrochronological treatment, and rigorous
analysis to identify and evaluate the actual conditions that produced them. A variety of
dendrochronological techniques have been used to identify potential causes of growth
change (Fritts, 1976; Baillie, 1982; Hughes et al., 1982; Schweingruber, 1988; Cook and
Kairiukstis, 1990).
It is mandatory that any qualitative features used for EMAP assessment be studied
carefully under conditions of known environmental changes to certify that the features and
assumed relationships are based upon quantifiable real-world phenomena. In addition,
some sites and species will be encountered in the first round of grid-point sampling for
which there is inadequate dendrochronological and ecophysiological information. New
chronologies may be required to evaluate these new species and environmental conditions.
Available chronologies may require updating to make comparisons to current conditions.
Some of this can be accomplished using materials collected in the tessellated sampling
scheme. Chronology development is tedious and time consuming work requiring skilled
dendrochronological workers.
Retrospective indicator applications can also be scaled up and down, in line with the
telescoping framework of the sampling design involving three tiers. At Tier 1 characterization
they can be coupled and calibrated with remote sensing information to evaluate the
statistical significance of observations with respect to long-term baseline behavior. At Tier 2
studies and site visits, physical-chemical and biological measurements of some resources
can be methodologically evaluated similar to Tier 1 characterization. Tier 3 special and
130
-------
intensive studies can be evaluated in terms of long-term trends using retrospective
indicators.
Modeling Ring Characteristics and Environmental Relationships - Two dendrochronological
growth models are being developed that could greatly facilitate the evaluation of newly
detected environmental conditions, ecological changes involving the productivity of sites, or
the presence (or absence) of stressors (Fritts et al., 1965; Fritts, 1990; Fritts et al., 1991).
The first is an empirical model that uses correlation, regression, or response function
technology, as described above, to calibrate tree-ring chronologies for any species with
associated monthly temperature, precipitation, Palmer Drought Severity Index, or any other
time series linked to growth-limiting conditions. Both linear and non-linear relationships can
be examined, and the calibration applied to questions of climatic change (Fritts and Dean,
1990) or pollution. The original measurements, estimates, or residuals from the first analysis
can be extracted and subjected to further analysis or compared to results from other growth
models. A Kalman Filter can be used to evaluate changing growth response over time.
This model is based upon empirical relationships, so it can be applied to species and
sites where there is little ecophysiological information and can be used in its present form to
address important EMAP questions involving environmental changes recorded by tree ring
chronologies. The model is easily modified and new modules can be developed to tailor it to
help answer other questions of importance to the EMAP program.
The second is a mechanistic model for conifer species. It uses biophysical
relationships between cambial activity and daily temperature data and soil water balance to
estimate cell growth and structure across a simulated radius of an annual ring. This model
requires a basic understanding of the biophysical relationships including field measurements
on specific sites. The coefficients of this model have been identified and validated for Pinus
ponderosa from Arizona and P. sylvestris and Larix siberica from Siberia. The model can be
calibrated with new sites and species as cell measurements for a species are obtained
along with daily climatic data and information on the beginning and ending of the growing
131
-------
season. Ringn/vidth estimates from this model can be compared to ring-width indices from
independent tree-ring chronologies using the empirical model, which serves as a unique
type of validation. The mechanistic model is more complex than the empirical model but it
has a greater potential for simulation.
Modifications of the model are planned to simulate features of hardwood, as well as
softwood species and cellular structure for an entire cross-section of a stem. Eventually the
model should be applied to the three-dimensional form of a tree including stem, roots, and
leaves to simulate both ring structure and the increment in biomass. The present version of
the model includes a simple photosynthetic module but should be expanded to utilize the
output from canopy models (Running and Coughlan, 1988; Weinstein and Beloin, 1990) to
estimate ring features expected from current or altered environments. With adequate
funding, an EMAP version could be developed and calibrated with known conditions, so that
it would be capable of simulating and investigating possible effects of suspected stressors
on woody plant growth in a resource sampling unit.
3.2.2 Meteorological Data
While tree-ring data provide site specific information, grids of tree-ring chronologies
can be analyzed and used to reconstruct large-scale climatic variations. For example, Briffa
et al. (1988, in press), Fritts et al. (1990), Fritts (1991), Stockton and Ivleko (1975 & 1983)
use principal component analysis of tree-ring parameters from tree-ring grids to extract the
large-scale variations through both space and time and to calibrate the variations with grids
of climatic data. Independent data are used to validate the calibration. Then, tree-ring data
from 1600-1900 are applied to the validated models to reconstruct seasonal variations in
climate for time periods and regions that lack instrumental measurements. This has
generated a 400 year long climatic record that can be used to evaluate the short-term
variations to be observed in the EMAP grid-point sampling.
The Palmer Drought Severity Index (PDSI) is a useful climatic integration that can link
modern climatic data to biological responses. It is a useful measure of recent (approximately
132
-------
100 years) climatic variations and is derived from a combination of monthly precipitation,
temperature, and soil moisture retention information. It offers an integrated measure of
moisture availability (i.e., effective precipitation). It frequently exhibits a higher covariation
than temperature or precipitation alone with a tree-ring series, because a tree-ring
responds to the integrated effects of temperature and precipitation through its interface with
the soil and atmosphere.
3.2.3 Pollen Record
Paleobotanical proxy data consisting of pollen data, as well as sub-fossil and fossil
plant remains recovered from the same depositional contexts, have traditionally been used
to reconstruct changes and long-term trends in past community composition and
predominance, often in response to climate change (Bryand and Holloway, 1985). As such,
fossil pollen records have served as retrospective indicators of long-term plant community
response to climate change.
Fossil pollen studies traditionally have been used to examine changes in the
dominance of certain plant communities on the landscape. The pollen record has been
limited by: 1) level of taxonomic identification and 2) the resolution that can be obtained
(Birks and Birks, 1980). At best, the pollen rain found in lake sediments can monitor only the
appearance or disappearance of specific plant species from a region. Therefore, most pollen
analyses consist of broad description of trends in community changes through time.
Because most pollen is identified only to genus (Moore and Webb, 1978), it is difficult
to identify the climatic limits and thus to interpret the fossil record in terms of specific climatic
conditions. At best it can be used only to estimate the direction of past climatic trends and to
compare the general climatic conditions to those in modern times.
3.2.4 Packrat Middens
Paleobotanical evidence from pollen records and woodrat middens can be used to
relate any change detected by tree ring analyses to century and millennia long variations in
133
-------
climate, the occurrence of fire, and changes in community structure (some yearly and
decadal information is available).
The contents of fossil dens (middens) of woodrats (Neotoma spp.) in the arid American
West are preserved due to the aridity of the region or protection in caves and overhangs.
The woodrats are known to forage plant remains (twigs, flowers, seeds, etc.) from within a
50 m radius, thus providing information on the species growing in the neighboring area. The
plant remains are identified, often to the species (information that pollen analysis cannot
provide), are radiocarbon dated, and detailed lists are constructed to provide unique
information on vegetational history for a very restricted locality at a specific point in time
(Spaulding, 1980 & 1985; Mehringer and Ferguson, 1969; Wells and Jorgensen, 1964;
Wells and Berger, 1967; Van Devender and Spaulding, 1979). Past climate and soil
conditions may be deduced from this information (Spalding, 1985). Because the distribution
of many plants is determined by specific moisture and temperature conditions, their
occurrences in the past at elevations above or below today's distribution provide clues to
changing temperature and precipitation patterns.
The geographical dispersion of fossil woodrat middens, although chronologically
discontinuous, can be assembled to reconstruct a history of both local and regional
vegetation response to climatic change (Spaulding, 1981; Spaulding, 1985). Fossil woodrat
midden records can also be used to investigate past vegetation distributions and their
change after the end of the Pleistocene and after fire (Wells, 1983; Thompson, 1990;
Mehringer and Wigand, 1990). Rates of plant migration can be reconstructed based upon
the changing occurrence of migrating species through time.
3.2.5 Fossil Charcoal Record
Fire, as well as disease and insect infestation, are disturbance phenomena reflecting
plant community health. Charcoal obtained from fossil pollen records can be used to
examine the relationships between climate, fuel build-up, and fire (Wigand, 1987).
Therefore, increases in charcoal abundance in the pollen records recovered from lakes and
134
-------
bogs indicate periods when plant communities are weakened and have become susceptible
to destruction by fire or when the climatic regime is more favorable to the generation of fires.
Increased amounts of charcoal with respect to pollen within a sample may record changes
in 1) the frequency of fire during a period, 2) a change in the extent of the fire, 3) proximity of
the fire, or 4) a change in the fuel type. The contemporaneous pollen record provides clues
as to which kind of change is responsible for the build-up of charcoal in the palynological
record.
Charcoal records from the American West reveal a distinct pattern. Fire becomes more
important during the wet periods of the early and late Holocene in the lower elevation
woodland, lower sagebrush steppe, and desert scrub. Little evidence of fire is present for
the middle Holocene (Mehringer and Wigand, 1987; Mehringer and Wigand, 1990). These
fires do not occur when it is wet, but are confined to droughts that interrupt periods of
generally wetter climate. Therefore, a pattern of fuel build-up during wet periods, followed
by drought and fire, is followed by renewed fuel build-up during succeeding wet periods.
This cycle seems to continue until a major shift to drier conditions occurs. At that time, fires
apparently thin the vegetation, reduce the fuel load and leave little opportunity for significant
fire activity during the succeeding period of dry climate.
Although fire is not an immediate indicator of community health, it is a proxy record of
past community health that often precedes changes in plant community composition. Fire
may be viewed as a key in opening niches for migrating species to occupy as a new climatic
regime is established.
3.2.6 Stable Isotopes, Fossil Woodrat Midden Materials, and Tree Ring
The core of stable isotope research concentrates on the ecophysiological and
environmental patterns of selected plant species. As with all RIs, calibration of the linkages
between stable isotopic storage in plant tissues and climate is required before applying the
analyses to fossil woodrat middens or to tree-ring chronologies (Long et al., 1990). This is
135
-------
accomplished by correlating isotopic measurements with environmental changes along
environmentally sensitive transects or from one time period to the next.
Water is the key factor for study, because it is the primary limiting factor for growth and
productivity in semi-arid environments (Smith and Nowak, 1990). Drought severity, as
characterized by the Palmer Drought Severity Index, reflects potential evapotranspiration.
Analyses of the hydrogen isotopic ratio (2H:11H (5D)), the oxygen isotopic ratio (18O:16O
(518O)), and the carbon isotopic ratio (13C:12C (813C)) should result in identificaiton of
changes in effective precipitation and possible changes in water use efficiency which affect
the growth and success of plants (Ehleringer et al., 1990; Farquhar and Richards, 1984; Toft
et al., 1989; Long et al., 1990; Siegel, 1983; Morecroft and Woodward, 1990).
3.2.7 Repeat Photography as a Retrospective Indicator
The advent of photography during the last century has provided the opportunity to
enhance the record of paleoenvironmental change provided by other retrospective indicators
by visually documenting environmental change over a variety of time spans. Comparison of
photographs of the same locality taken as much as a hundred years or more apart can be
used to track changes in: 1) plant community distribution (i.e., expansion and retreats of
lower and upper tree lines), 2) erosional and depositional processes, and 3) human land
use.
Environmental changes occurring over the last 150 years or so already have been
followed with photo series that already exist (Hastings and Turner, 1965; Rogers, 1982;
Rogers etal., 1984).
Time-lapse photography can track environmental change with intervals on the order of
minutes, hours, days, months, years, decades, scores of years, or more. At the opposite
end of the time scale, high-speed photography can be used to monitor much more rapid
changes on the order of seconds or fractions of a second. Although the range of time that
the environment can be monitored with photographs is limited to the last 160 years or so,
this record can be extended with landscape paintings that depict environmental changes
136
-------
stretching back into the 1500s (Ladurie, 1971) (e.g., the winterscapes of the Dutch painter,
Pieter Bruegel the Elder; and the depictions of greatly expanded glaciers during the "Little
Ice Age").
Paintings or lithographs of lakes, mountain scapes, etc. of the early period of
exploration of the American West often record conditions different than those that
predominate today. This source of environmental proxy data was generated not only by
artists on exploration and survey parties, but also by artists of the Romantic Style who were
fascinated by the "pristine" or "primeval" nature of the American West. Whereas much of the
exploration and survey photography and illustrations can be obtained from government
sources, paintings, lithographs, and other illustrations are often housed in private collections
or in local historical museums, which often curate historical photographs containing
environmental data spanning the period of settlement of the American West.
3.3 Data Sources
The use of retrospective indicators is not limited exclusively to arid lands, nor even to
the United States. This cross-cutting attribute is a desirable feature of EMAP indicators.
Comparable, although sometimes less well developed, information exists on a continental
and worldwide basis. An extensive tree ring network is available for arid regions from
western Canada to northern Mexico, and there is reasonable coverage for the eastern
United States. Between one and two thousand chronologies already exist, although not all
extend to the present (as each year passes all existing chronologies are one more year out
of date). For evaluating trends exceeding the age of trees and shrubs sampled as a part of
the EMAP program, it will be necessary to update some existing chronologies and
paleoenvironmental data. In addition, it may be necessary to develop new time series for
evaluating changes appearing in the EMAP samples in cases where no paleoenvironmental
time series exists today.
Long-term macroclimatic patterns are similar over wide regions, so the rings from
climate stressed old trees growing on relatively undisturbed habitats in neighboring areas
137
-------
can be sampled and analyzed for their paleoclimatic information. These sites could include
long-term ecological research sites (LTERs) where other baseline information has already
been assembled that could be calibrated with the tree-ring information and related to past
and current conditions.
3.3.1 Tree-ring Series Data Sources
An extensive database of tree-ring information is available from the NOAA National
Geophysical Data Center (NGDC) in Boulder, Colorado. Seven hundred twelve chronologies
are available from North America, with several hundred of those representing sites from the
western United States. The chronologies specific to our study area need to be identified.
Other sources have not been specified.
3.3.2 Packrat Midden Data Sources
While no formal data base exists for packrat midden data, numerous studies are
ongoing and have been conducted and published (e.g., Betancourt et al., 1990). These are
largely focused in arid ecosystems in the U.S. and provide a basis for utilizing midden data
in the EMAP-Arid pilot study.
3.3.3 Pollen Data Sources
An American Pollen Database that will incorporate all North American pollen data
(stratigraphic and surface-sample data) is under development. The database coordinator is
Eric Grimm at the Illinois State Museum. Whether this database will be useful for the pilot is
unknown.
3.4 Data Analysis
Analysis using retrospective indicators can be accomplished using both statistical
techniques and simulation modeling. For example, a comparison can be made of the current
year's shrub- or tree-ring index for a species and site with the time series of indices for the
previous several hundred years. The current year's value can be evaluated with respect to
long term central tendency, variability, and persistence structure; and a probabilistic value
138
-------
can be associated with its interpretation. In the absence of such retrospective information,
the current year's observations can be compared only to spatial variation. If retrospective
information is unavailable, it is impossible to state with any degree of confidence what the
current year's values represent in terms of temporal behavior.
In addition, a statistical transfer function can be developed that will relate statistically
independent variables of retrospective information (tree-ring indices) to dependent variables
of seasonal or annual climate (Fritts, 1976; Fritts et al., 1990 & 1991; Fritts, 1991). This is
accomplished using a portion of the recent period of common overlap between the two types
of variables to calibrate the relationship and then using another portion to validate the
relationship (Fritts et al., 1990). The transfer function can be applied to the tree-ring series
predating the climatic record to reconstruct past climatic conditions for several centuries. A
long-term record of a stressor indicator (climate) is produced against which future values
can be evaluated. This technique can be used to reconstruct a variety of new retrospective
indicators (NRIs) gathered by remote sensing or monitored at collecting stations as long as
the variable is linked in some way to environmental factors directly affecting the
retrospective indicators (RIs) (Fritts, 1976). For example, Fritts (1991) reconstructs spatial
variations of sea-level pressure, as well as temperature and precipitation from a spatial grid
of tree ring chronologies. Atmospheric pressure is not known to influence growth, but it is
related to the movement of storms that deliver the precipitation, affect the sunlight, and
control the winds influencing the temperature that, in turn, affect ring-width variations.
Including retrospective indicators (RIs) for as many resources as possible could
supplant, or at the very least, complement the wait-and-see approach for a variety of
important indicators. Such RIs observations already exist for decades, centuries, or
thousands of years depending upon the variable and the Rl. Timely interpretation of
observations is pragmatic scientifically, because results can be introduced into peer review
publications and subjected to wide distribution and criticism within a short time period. Such
immediate statistical interpretation and evaluation of indicators could be an attractive
139
-------
marketing feature to funding sources (e.g., Congress) unwilling to wait years for results. It
will take four years before sample results are available for all locations under current design
guidelines. From a political survivability perspective it may be advantageous to be able to
examine some measure of the relative changes or lack of changes (however limited) within
the first few years of the EMAP effort.
3.4.1 Tree-ring Series Data Analysis
A variety of ring characteristics and climatic variables can be analyzed. Indices of
maximum latewood density were used in the first two studies of Briffa et al. (1988, in press)
to reconstruct warm-season temperatures over Europe and western North America. Fritts et
al. (1990), Fritts (1991) used indices of ring widths to reconstruct seasonal and annual
temperature and precipitation for North America and sea-level pressure for the North Pacific
and North American sectors. Similar indices were used by Stockton and Meko (1975 &
1983), but they reconstructed Palmer Drought Severity Indices over the western United
States. In addition, there are numerous dendroclimatic reconstructions for local regions,
historical references to climatic conditions or events, and basic information from all RIs that
can be assembled with the large-scale tree-ring reconstructions to study the range of
variation for each region, time period, and variable. This can be used to generate a type of
probability distribution to place confidence limits on the paleoclimatic estimates and to test
observed changes.
Instead of developing transfer functions, the tree-ring index can be considered the
dependent variable and climatic data the independent variables to obtain response
functions. The index is calibrated with principal components of monthly climatic data and the
resulting regression coefficients are multiplied by the eigenvectors of the climatic data to
obtain response function coefficients identified with each monthly climatic variable (Fritts,
1976). However, in the original version, the errors of the response function coefficients were
found to be biased because stepwise regression procedures were applied to collinear
variables. This problem has now been resolved by using a bootstrap method (Efron, 1979,
140
-------
1983) to obtain unbiased estimates of the response function coefficients. Response
functions obtained by using this new technique or by using more conventional multiple
regression techniques are used to assess the quality, structure, and amount of climatic
information in tree-ring chronologies or to remove the climatic signal in the record so that
the effects of non-climatic factors can be evaluated (Fritts et al., 1991). This approach can
be used to calibrate a tree-ring chronology with climatic variation for a period known to be
free of pollution and then to estimate the effect of climate on growth through the interval of
possible pollution. Any differences between the growth estimated solely from climate and
the actual chronology during the pollution period are identified as a possible effect of
pollution (Fritts et al., 1991). The errors from the bootstrap analysis can be used to test the
significance of the differences. The same technique can be used to detect the possible
effect of rising atmospheric CO2 on ring characteristics (LaMarche et al., 1986), as well as
other growth promoting or inhibiting influences present in the tree-growth record.
Replicate cores, two radii per tree, are usually sampled from 20 to 50 old trees,
growing under similar environmental conditions. These samples are mounted, the surface
prepared, all rings crossdated, ring characteristics measured, and subjected to quality
control to assure accuracy of dating and measurement. These data are standardized to
remove trends due to increasing tree-age and stand dynamics, indices are computed, and
the information combined to obtain a mean chronology. Autoregressive Moving Average
(ARMA) modeling may be required to simplify the time-series characteristics. A variety of
ring characteristics can be measured, some using simple optical procedures and others
using more complex technology involving x-rays, image analysis, isotopic analysis, or
analytical chemistry.
3.4.2 Pollen Data Analysis
The resolution of the pollen record is influenced by the deposition rates of the
sediments, the accuracy of dating, and the sample collection process. Except for annual
varved sediments, pollen cores can only be dated using radiocarbon of organic materials
141
-------
deposited in the sediments. Most pollen samples are taken at approximate 50 to 100 year
intervals along the length of the sampled core because of the limitations of time and money.
In addition, the pollen rain in a sediment may include several years due to the height of the
column that is sampled and the annual turnover and redeposition of sediments at the bottom
of the lake.
Deep steep-sided lakes with reducing conditions (meromictic lakes) physically prevent
the annual turnover of sediments that mixes lake-bottom sediments and pollen. However,
this still does not prevent materials from being redeposited from the slopes of the lake basin.
Decreasing the distance between samples and reducing sample height will increase the
resolution of the record, but not its precision. Use of x-rays to locate annual or semi-annual
breaks in the record is possible but costly. However, even a slight decrease in the sample
interval has beneficial results in changing the pollen record from one reflecting long-term
plant community expansions and contractions to one reflecting near-annual changes in
pollen production. This can be accomplished by finding a locality that has a very rapid
deposition rate, e.g., 2 m per 1,000 years or so. In such a situation, greater detail can be
achieved without decreasing sample spacing and height. When the number of years
between samples is reduced, much shorter term variations in pollen production, unrelated to
expansion and contraction of specific plant communities, may be revealed. These are the
result of interannual changes in pollen production of already established plants and reflect
changing climatic conditions (Grosse-Brauckmann, 1978). This change in pollen production
from one year to the next can be applied both to modern and fossil pollen records to
examine possible environmental changes and community health.
Multivariate transfer functions can also be established between a spatial array of
locations where surface pollen has been collected and arrays of either quantitative
vegetation information or climatic data. These types of analyses have been developed most
extensively and convincingly in the eastern, northeastern, and central United States
(Bernabo and Webb, 1977; Webb and Clark, 1977; Webb and Bryson, 1972). The transfer
142
-------
functions established with the spatial data arrays can then be used to interpret the time
series of palynological data available from lake sediments, bogs, and other depositional
environments with good preservation. However, different rates of plant migration after
deglaciation also influence community structure and can seriously distort the climatic signal
in some pollen samples.
3.4.3 Packrat Middens Data Analysis
In order to assess accurately the information contained in fossil woodrat middens, the
collecting bias of the woodrat must be taken into account (Dial and Czaplewski, 1990).
When the contents of modern middens are compared with surrounding plant communities
there is not necessarily a one to one correlation. This relationship, as well as how plant
species abundance may vary throughout the nest, must be studied carefully (Finley, 1990).
Unfortunately, community composition may lag behind climatic change due to the time
required for plant migration. Thus, traditional fossil woodrat midden data appears to be
insensitive to short-term climatic changes. The time required for species favored by a new
climatic regime to enter into an area is affected by (1) distance from the area of dispersion,
(2) dispersal mechanisms, and (3) the opening of niches for the new species to fill. In
addition, because plant materials from fossil woodrat middens cannot be adequately
quantified, and because their original incorporation within the nest is biased by the collecting
behavior of the woodrat, their use as sensitive indicators of plant community change has
been ineffectual.
However, isotopic analysis provides new possibilities of examining immediate and
sensitive plant response to rapid climate change with fossil woodrat midden materials (Long
et al., 1990). Information gained on the physiological responses of modern species to
environmental variation can be used to deduce past plant responses that may indicate
climate changes.
143
-------
3.4.4 Fossil Charcoal Records Data Analysis
Increased abundance of charcoal followed by clear decreases in arboreal and shrub
pollen and increases in grasses and weed pollen indicate local fire. Additionally, local fires
often introduce larger charcoal into the record, because with distance the larger fragments
have settled out. Significant increases or decreases in the mean size of charcoal through
time in either local or regional fires also indicate a change in the kind of fuel being
consumed. Changes in fire frequency only can be revealed through x-ray analysis of the
pollen core, which may reveal changes in the number of charcoal laminae per time period
being noted.
3.4.5 Stable Isotopes, Fossil Woodrat Midden Materials, and Tree Ring Data Analysis
The ratio 5D:818O in physical systems is inversely related to drought severity
(Dansgaard, 1964; Hoefs, 1987). This also should hold true for plant systems (Yakir et al.,
1990). Research presently being conducted by Dr. R.S. Novak of the University of Nevada,
Reno is addressing this questions in order to verify that 13C of plants is related to water use
efficiency, leaf gas exchange measurements of photosynthesis, transpiration, and leaf
conductance are being used to determine real time measurements of water use efficiency,
as well as the slope of the relationship between photosynthesis and conductance. The
relationship between 513C and growth also is being determined by correlation of the carbon
isotope ratio with tree-ring measurements. Using this information, it will be determined if
these isotopic ratios can be used to examine whether particular plant species are restricted
to specific environments or not. For these relationships to be usefully revealed, species such
as Utah juniper and pinyon pine, with wide spatial and deep time distribution, are being
selected for study.
Unfortunately, changing abundance of individual plant species in a community may not
evidence only the physiological response of an apparently homogeneous species, but it may
reflect genotypic variation. Genotypic changes of this kind across space and through time
can be monitored by studying DNA using Restriction Fragment Length Polymorphism
144
-------
(RFLP) techniques on both modern materials and materials from fossil woodrat middens
(Sambrook et al., 1989). This can provide information on conifer phylogeny, population
structure, and genotypic distribution of natural populations (Kiem et al., 1989; Nybom and
Schall, 1990; Strauss and Doerksen, 1990; Strauss et al., 1990) which can produce
substantial differences in response over the EMAP sampling area.
Currently this diversity is being studied in three ways by Dr. R. Tausch (Project Leader
for the Intermountain Research Station of the U.S. Forest Service). First, the geographical
variation of a particular species across an area is being determined. Second, the role of
topography in determining the geographical distribution of genetic patterns will be
determined. Third, paleoecological materials will be examined to add information on
variations through time. Such an approach provides the opportunity to integrate genetics,
physiological ecology, palaeoecology, population biology, and dendrochronology in the
EMAP program.
3.5 Conclusions
The cells in annual rings of woody plants serve as biomarkers of past limiting
conditions. We must become aware of the many kinds of environmental information
contained in the ring structures of woody plants and utilize this information in planning for
and detecting future changes.
Thus, the annual rings of woody plants should be one of the EMAP core indicators. All
materials should be examined at the time of collection and obvious growth changes or
evidence of particular stress be recorded and tabulated. All samples should be saved for
potential laboratory analysis of any changes that are identified.
Relevant literature should be reviewed and plant species identified that are likely to
provide useful information. Additional species encountered in EMAP sampling should be
investigated as to whether their rings also contain useful information.
145
-------
Dendrochronological simulation models should be developed and applied to EMAP
questions. This would help to identify ecosystem responses to past environmental changes
that could be used to predict responses to future changes.
Grids of long tree-ring chronologies can be calibrated with appropriate diagnostic
indicators to reconstruct base line conditions. This would extend the data sets over a longer
period and provide for more rigorous interpretation and testing of future changes.
The seasonal and annual variability of climatic data can be reconstructed from
dendrochronological data. Also, available climatic information about the past from all
retrospective indicators and historical accounts can be assembled, compared, and
summarized for EMAP evaluation of current climatic trends and conditions.
Pollen data reveal information on community structure which can sometimes be
interpreted in terms of climatic history, but rates of plant migration following deglaciation also
affect community structure. A finer sampling of sediments can reveal shorter-term changes
related to yearly pollen production and changes in the local environments. Charcoal
fragments in the pollen profile provide clues to periods of wetness and dryness that affect
fuel loading and health of the community.
The potential of stable isotopic and DNA analysis of paleobotanical micro— and
macro—fossil data and tree rings has afforded the opportunity to examine plant physiology at
the organismal level through time. This has enhanced the use of these data for the detection
of organismal response to environmental change and the use of this information for
projecting future botanical responses to global changes. The method is not only relatively
easy to perform, but it is also cost effective.
Repeat photography provides visual documentation of the environmental changes that
are evidenced by other retrospective indicators. Not only can it record that change occurred,
but it can be used to document the degree of change, so that the actual biotic response can
be equated with the environmental parameters that caused them.
146
-------
4.0 REFERENCES
Alekseyev, A.S. 1990. Analysis of annual radial tree ring growth in tree stands in
atmospheric pollution. Lesovedenie (0)2:82-86.
Alekseyev, A.S., N.I. Lairand, and Y.I. Leplinsky. 1988. Quantitative analysis of the impact of
atmospheric pollution on the tree stands using the indices of the radial growth
increment. Botanicheskii Zhurnal 73(6):911-917.
Arndt, U. and M.M. Wehrle. 1982. Ergebnisse dendrochronologische Untersuchungen an
Eichen zur Indikatoren von Immissionsbelastungen. [Results on dendrochronological
investigations on oak as indicators of pollution stress.] Staub-Reinhalt Luft
42(2):64-68.
Atwater, B.F., M. Stuiver, and O.K. Yamaguchi. 1991. Radiocarbon test of earthquake
magnitude at the Cascadia subduction zone. Nature 353:156-158.
Baes, C.F. Ill and S.B. McLaughlin. 1984. Trace elements in tree rings: evidence of recent
and historical air pollution. Science 224:494-497.
Baes, C.F. Ill, S.B. McLaughlin, and T.A. Hagan. 1984. Multielemental analysis of tree rings:
temporal accumulation patterns and relationships with air pollution. In: Proceedings
Symposium on Air Pollution and the Productivity of the Forest. Isaac Walton League,
Arlington, VA, pp. 273-286.
Baillie, M.G.L. 1982. Tree-ring dating and archaeology. Groom Helm, London, UK.
Baillie, M.G.L. and M.A.R. Munro. 1988. Irish tree rings, Santorini and volcanic dust veils.
Nature 332:344-046.
Bauch, J., P. Rademacher, W. Berneike, J. Knoth, and W. Michaelis. 1985. Breite und
Elementgehalt der Jahrringe in Fichten aus Waldschadensgebieten. [Width and
elemental contents of tree rings from fir in pollution-stressed forest regions.]
VDI-Berichte (Verien Deutscher Ingenieure) 560:943-959.
Bernabo, J. C. and T. Webb III. 1977. Changing patterns in the Holocene pollen record of
northeastern North America: a mapped summary. Quaternary Research 8:69-96.
Betancourt, J.L., T.R. Van Devender, and P.S. Martin. 1990. Packrat midden: the last 40,000
years of biotic changes. University of Arizona Press, Tucson, AZ, 467 pp.
Birks, H.J.B. and H.H. Birks. 1980. Quaternary paleoecology. Edward Arnold, London, 289
pp.
Blais, J.R. 1958. Effects of defoliation by spruce budworm (Choristoneura fumiferana Clem.)
in radial growth at breast height of balsam fir (Abies balsamea (L.) Mill.) and white
spruce (Picea glauca (Moench) Voss.). Forestry Chronicle 34(1):39-47.
147
-------
BLM. 1984. Rangeland monitoring: actual use studies. United States Department of Interior,
Bureau of Land Management, TR44000-2. Denver, CO. 8 pp.
Brady, W., O.K. Patton, and J. Paxson. 1985. The development of southwestern riparian
gallery forests. Pages 39-43 in: Johnson, R.R., C.D. Ziebell, D.R. Patton, R.F. Ffolliott,
and R.H. Hamre (Technical Coords.). Riparian ecosystems and their management:
reconciling conflicting uses. Proceedings, First North American Riparian Conference,
USDA, Forest Service General Technical Report RM-120. 523 pp.
Briffa, K.R., P.O. Jones, and F.W. Schweingruber. 1988. Summer temperature patterns over
Europe: a reconstruction from 1750 A.D. based on maximum latewood density indices
of conifers. Quaternary Research 30:36-52.
Briffa, K.R., P.O. Jones, and F.H. Schweingruber. in press. Tree-ring density reconstruction
of summer temperature patterns across western North America since A.D. 1600.
Journal of Climate.
Brubaker, L.B. 1987. Forest disturbance and tree-ring analysis. In: G.C. Jacoby, Jr. and
J.W. Hornbec, eds. Proceedings of the International Symposium on Ecological Aspects
of Tree-Ring Analysis. U.S. Department of Energy, Publication CONF-8608144, pp.
101-118.
Brubaker, L.B. and S.K. Green. 1979. Differential effects of Douglas-fir tussock moth and
western spruce budworm defoliation on radial growth of grand fir and Douglas fir.
Canadian Journal of Forestry Research 9:95-105.
Bryand, V. M. Jr. and R. G. Holloway, eds. 1985. Pollen records of late quaternary North
American sediments. American Association of Stratigraphic Palynologists, 423 pp.
Burgess, R.L. and D.M. Sharpe, eds. 1981. Forest island dynamics in man-dominated
landscapes. New York, Springer-Verlag.
Cain, W.F. 1978. Carbon-14, tree rings, and urban air pollution. Environment International
1:167-171.
Chuvieco, E. and R.G. Congalton. 1989. Application of remote sensing and geographic
information systems to forest fire hazard mapping. Remote Sensing of Environment
29:147-159.
Conkey, L.E. 1984. X-ray densitometry: wood density as a measure of forest productivity
and disturbance. In: D. D. Davis, ed. Air pollution and the productivity of the forest.
Izaak Walton League of America, Inc., Arlington, VA, pp. 287-296.
Cook, E.R. 1987. The use and limitations of dendrochronology in studying effects of air
pollution on forests. In: T.C. Hutchinson and K.M. Meema, eds. The effects of acid
deposition on forests, wetlands, and agricultural ecosystems. NATO ASI Series G16,
pp. 277-290.
148
-------
Cook, E.R., A.M. Johnson, and TJ. Biasing. 1987. Forest decline: modeling the effect of
climate in tree rings. Tree Physiology 3:27-40.
Cook, E.R. and L Kairiukstis, eds. 1990. Methods of dendrochronology: applications in the
environmental sciences. Kluwer Academic Publishers, Dordrecht, 394 pp.
Crone, A. 1987. Tree-ring studies and the reconstruction of woodland management
practices in antiquity. In: G.C. Jacoby, Jr. and J.W. Hornbeck, eds. Proceedings of the
International Symposium on Ecological Aspects of Tree-Ring Analysis. U.S.
Department of Energy, Publication CONF-8608144, pp. 327-336.
Dansgaard, W. 1964. Stable isotopes in precipitation. Tellus 16:436-468.
Dial, K.P. and N.J. Czaplewski. 1990. Do woodrat middens accurately represent the animals'
environments and diets? The Woodhouse Mesa study. In: J.L. Betancourt, T.R. Van
Devender, and P.S. Martin, eds. Fossil packrat middens: the last 40,000 years of biotic
change. University of Arizona Press, Tucson, pp. 43-58.
Dixon, R.K., R.S. Meldahl, G.A. Ruark, and W.G. Warren, eds. 1990. Forest growth: process
modeling of responses to environmental stress. Timber Press, Oregon.
Efron, B. 1979. Bootstrap methods: another look at the jackknife. The Annals of Statistics
7:1-26.
Efron, B. 1983. Estimating the error rate of a prediction rule: improvement on
cross-validation. Journal of the American Statistical Association 78:316-331.
Ehleringer, J.R., J.W. White, D.A. Johnson, and M. Brick. 1990. Carbon isotope
discrimination, photosynthetic gas exchange, and transpiration efficiency in beans and
range grasses. Acta Oecologia 11:611-625.
Farquhar, G.D. and R.A. Richards. 1984. Isotopic composition of plant carbon correlates
with water use efficiency of wheat genotypes. Australian Journal of Plant Physiology
11:539-552.
Finley, R.B. 1990. Woodrat ecology and behavior and the interpretation of paleomiddens. In:
J.L. Betancourt, T.R. Van Devender, and P.S. Martin, eds. Fossil packrat middens: the
last 40,000 years of biotic change. University of Arizona Press, Tucson, pp. 28-42.
Forman, R.T.T. and M. Godron. 1986. Landscape ecology. New York, John Wiley & Sons,
Inc.
Fox, C.A. 1980. The effect of air pollution on western larch as detected by tree-ring
analysis. PhD dissertation. Arizona State University, Tempe, 98 pp.
Fox, C.A., W.B. Kincaid, T.N. Nash III, D.L. Young, and H.C. Fritts. 1986. Tree-ring variation
in western larch (Larix occidentalis) exposed to sulfur dioxide emissions. Canadian
Journal of Forest Research 16:283-92.
149
-------
Freyer, H.D. 1979. On the carbon-13 record in tree rings. Part II. Registration of a
microenvironmental carbon dioxide and anomalous pollution effect. Tellus
31(4):308-312.
Fritts, H.C. 1976. Tree rings and climate. Academic Press, London, 567 pp. Reprinted 1987
as: "Tree rings and climate". Methods of dendrochronology Vols. II and III, L.
Kairiukstis, Z. Bednarz, and E. Feliksik, eds. International Institute for Applied Systems
Analysis and the Polish Academy of Sciences, Warsaw.
Fritts, H.C. 1990. Modeling tree-ring and environmental relationships for
dendrochronological analysis. In: R.K. Dixon, R.S. Meldahl, G.A. Ruark, and W.G.
Warren, eds. Forest growth: process modeling of responses to environmental stress.
Timber Press, Oregon, pp. 368-382.
Fritts, H.C. 1991. Reconstructing large-scale climatic patterns from tree-ring data: a
diagnostic analysis. University of Arizona Press.
Fritts, H.C., D.G. Smith, and M.A. Stokes. 1965. The biological model for paleoclimatic
interpretation of Mesa Verde tree-ring series. American Antiquity 31(2):101-21.
Fritts, H.C. and T.W. Swetnam. 1989. Dendroecology: a tool for evaluating variations in past
and present forest environments. Advances in Ecological Research 19:111-88.
Fritts, H.C., J. Guiot, G.A. Gordon, and F. Schweingruber. 1990. Methods for calibration,
verification and reconstruction. In: E. Cook and L. Kairiukstis, eds. Methods of
dendrochronology: applications in environmental sciences. Kluwer Academic
Publishers, Dordrecht, pp. 163-217.
Fritts, H.C. and J.S. Dean. 1990. Dendrochronological modeling: the effects of climatic
change on ring width variation. Presented in: "Climate and culture: the response of
Chaco Canyon people to climatic change." 1990 AAAS Annual Meeting, New Orleans,
February 20. Priscilla Reining, Organizer.
Fritts, H.C., E.A. Vaganov, I.V. Sviderskaya, and A.V. Shashkin. 1991. Climate variation and
tree-ring structure in conifers: empirical and mechanistic models of tree-ring width,
number of cells, cell size, cell-wall thickness and wood density. Climate Research
1:97-116.
Gardner, R.H., B.T. Milne, M.G. Turner, and R.V. O'Neill. 1987. Neutral models for the
analysis of broad-scale landscape pattern. Landscape Ecology 1(1): 19-28.
Gay, C.A. 1989. Modeling tree level processes. In: R.D. Noble, J.L. Martin, and K.F. Jensen,
eds. Proceedings of the Second US-USSR Symposium on: "Air Pollution Effects on
Vegetation including Forest Ecosystems." pp. 143-157.
Gebhart, K., S. Leonard, G. Staidl, and D. Prichard. 1990. Riparian area managment:
riparian and wetland classification review. U.S. Dept. of Interior, Bureau of Land
Management, TR1737-5, Denver, CO.
150
-------
Gemmill, B., J.R. McBride, and R.D. Laven. 1982. Development of tree-ring chronologies in
an ozone air pollution stressed forest in southern California, USA. Tree-Ring Bulletin
42:23-32.
Glinski, R.L. 1977. Regeneration and distribution of sycamore and cottonwood trees along
Sonoita Creek, Santa Cruz County, Arizona. Pages 116-123. In: R.R. Johnson and
D.A. Jones (Technical Coords.) Importance, preservation and management of riparian
habitat: a symposium. USDA, Forest Service General Technical Report RM-43, 217
pp.
Gottesfeld, A.S. and L.M.J. Gottesfeld. 1990. Floodplain dynamics of a wandering river,
dendrochronology of the Morice River, British Columbia, Canada. Geomorphology
3:159-179.
Graumlich, L.J., L.B. Brubaker, and C.C. Grier. 1989. Long-term trends in forest net primary
productivity: Cascade Mountains, Washington. Ecology 70:405-410.
Gregory, S.V., F.J. Swanson, W.A. McKee, and K.W. Cummins. 1991. An ecosystem
perspective of riparian zones. Bioscience 41(8):540-551.
Greve, U., D. Eckstein, R.W. Aniol, and F. Scholz. 1987. Dendroklimatologische
Untersuchungen an Fichten unterschiedlicher Immissionsbelastung in Nordostbayern.
[Dendroclimatological investigations on Norway spruce under different levels of air
pollution in northeastern Bavaria.] Allgemeine Forst-und Jagdzeitung 157(9):174-179.
Grosse-Brauckmann. 1978. Absolute jahrliche Polleniederschlagsmengen an
verschiedenen Beobachtungsorten in der Bundesrepublik Deutschland. Flora
167:209-247.
Guyette, R.P., P.E. Cutter, and G.S. Henderson. 1989. Long-term relationships between
molybdenum and sulfur concentrations in red cedar tree rings. Journal of
Environmental Quality 18:385-389.
Guyette, R.P., P.E. Cutter, and G.S. Henderson. 1991. Long-term correlations between
mining activity and levels of lead and cadmium in tree-rings of eastern red-cedar.
Journal of Environmental Quality 20:146-150.
Hastings, J.R. and R.M. Turner. 1965. The changing mile: an ecological study of vegetation
change with time in the lower mile of an arid and semiarid region. University of Arizona
Press, Tucson, 317 pp.
Havas, P. and S. Huttunen. 1972. The effect of air pollution on the radial growth of Scots
pine (Pinus sylvestris L). Biological Conservation 4(5): 361-368.
Henoch, W.E.S. and M.L Parker. 1972. Dendrochronological studies relating to climate,
river discharge, and flooding in several regions of western Canada. In: W.P. Adams
and F.M. Helleiner, eds. International Geography 1972, Vol. 1. University of Toronto
Press, pp. 231-233.
151
-------
Hoefs, J. 1987. Stable isotope geochemistry. Springer-Verlag, Heidelberg.
Hughes, M.K., P.M. Kelly, J.R. Pilcher, and V.C. LaMarche, Jr. 1982. Climate from tree rings.
Cambridge University Press, Cambridge, 223 pp.
Hunsaker, C.T. and D.E. Carpenter, eds. 1990. Ecological indicators for the Environmental
Monitoring and Assessment Program. EPA 600/3-90/060. U.S. Environmental
Protection Agency, Office of Research and Development, Research Triangle Park, NC.
Hupp, C.R. 1988. Plant ecological aspects of flood geomorphology and paleoflood history.
In: V.R. Baker, R.C. Kochel, and P.C. Patton, eds. Flood geomorphology. John Wiley
and Sons, New York, pp. 335-356.
Innes, J.L. and E.R. Cook. 1989. Tree-ring analysis as an aid to evaluating the effects of
pollution on tree growth. Canadian Journal of Forest Research 19:1174-1189.
Jacoby, G.C. Jr., P.R. Sheppard, and K.E. Sieh. 1988. Irregular recurrence of large
earthquakes along the San Andreas Fault: evidence from trees. Science 241:196-199.
Jagels, R. and F.W. Telewski. 1990. Computer-aided image analysis of tree rings. In: E.
Cook and L. Kairiukstis, eds. Methods of dendrochronology: applications in
environmental sciences. Kluwer Academic Publishers, Dordrecht, pp. 76-93.
Johnson, R.R. and J.M. Simpson. 1985. Desertification of wet riparian ecosystems in arid
regions of the North American Southwest. Pages 1383-1393. In: Proceedings of Arid
Lands: Today and Tomorrow. Westview Press, Boulder, CO.
Kauffman, J.B. and W.C. Krueger. 1984. Livestock impacts on riparian ecosystems and
streamside management implications ... a review. Journal of Range Management
37(5) :430-438.
Keim, P., K.N. Paige, T.G. Whitham, and K.G. Lark. 1989. Genetic analysis of an
interspecific hybrid swarm of Populus: occurrence of unidirectional introgression.
Genetics 123:557-565.
Kienast, F. 1985. Tree-ring analysis, forest damage and air pollution in the Swiss Rhone
Valley. Land Use Policy 2, pp. 74-77.
Krummel, J.R., R.H. Gardner, G. Sugihara, R.V. O'Neill, and P.R. Coleman. 1987.
Landscape patterns in a disturbed environment. Oikos 48:321-324.
Ladurie. 1971. Times of feast, times of famine: a history of climate since the year 1000.
Doublday, Garden City, New York.
LaMarche, V.C. Jr. and K.K. Hirschboeck. 1984. Frost rings in trees as records of major
volcanic eruptions. Nature 307:121-128.
152
-------
LaMarche, V.C. Jr., D.A. Graybill, H.C. Fritts, and M.R. Rose. 1986. Carbon dioxide
enhancement of tree growth at high elevations. In: Technical Comments. Science
231:859-60.
Lange, R.T. and M.C. Willocks. 1978. The relation between sheep-time spent and egesta
accumulated with an arid zone paddock. Aust. J. Exp. Agr. Anim. Husb. 18:764-767.
Larson, D.W. 1990. Effects of disturbance on old-growth Thuya occidentalis at cliff egdes.
Canadian Journal of Botany 68:1147-1155.
Leavitt, S.W. in press. Isotopes and trace elements in tree rings. Proceedings of "Tree Rings
and Environment" held at Ystad, Sweden, Sept. 3-9,1990.
Leavitt, S.W. and A. Long. 1987. Influence of site disturbance on 13C isotopic time series
from tree rings. In: G.C. Jacoby, Jr. and J.W. Hornbeck, eds. Proceedings of the
International Symposium on Ecological Aspects of Tree-Ring Analysis. U.S.
Department of Energy, Publication CONF-8608144, pp. 119-129.
Lepp, N.W. 1975. The potential of tree-ring analysis for monitoring heavy metal pollution
patterns. Environmental Pollution 9:49-61.
Lillesand, T.M. and R.W. Kiefer. 1979. Remote sensing and image interpretation. John Wiley
and Sons. New York. 612 pp.
Loehle, C. 1990. Home range: a fractal approach. Landscape Ecology 5(1):39-52.
Long, A., L.A. Warneke, J.L Betancourt, and R.S. Thompson. 1990. Deuterium variations in
plant cellulose from fossil packrat middens. In: J.L. Betancourt, T.R. Van Devender and
P.S. Martin, eds. Fossil Packrat Middens: The last 40,000 years of biotic change.
University of Arizona Press, Tucson, pp. 380-396.
Lowrance, R.R., R.L Todd, and LE. Asmussen. 1984. Nutrient cycling in an agricultural
watershed: I. Phreatic movement. Journal of Environmental Quality 13:22-27.
Martin, B. and E.K. Sutherland. 1990. Air pollution in the past recorded in width and stable
carbon isotope composition of annual growth rings of Doublas-fir. Plant, Cell and
Environment 13:839-844.
McClenahen, J.R. and LS. Dochinger. 1985. Tree ring response of white oak to climate and
air pollution near the Ohio River Valley. Journal of Environmental Quality
14(2):274-280.
McColl, R.H.S. 1978. Chemical runoff from pastures: The influence of fertilized pasture
zones. New Zealand Journal of Marine and Freshwater Research 12:371-380.
McLaughlin, S.B., R.K. McConathy, D. Duvick, and L.K. Mann. 1982. Effects of chronic air
pollution stress on photosynthesis, carbon allocation, and growth of white pine trees.
Forest Science 28(1):60-70.
153
-------
Meentemeyer, V. and E.G. Box. 1987. Scale effects in landscape studies. In: Turner, M.G.,
ed. Landscape heterogeneity and disturbance, pp. 15-33, Springer-Verlag, New York.
Mehringer, P.J. Jr. and C.W. Ferguson. 1969. Pluvial occurrence of Bristlecone Pine (Pinus
aristata) in a Mohave Desert mountain range. Journal of Arizona Academy of Science
5:284-292.
Mehringer, P.J. Jr. and P.E. Wigand. 1987. Western Juniper in the Holocene. Everett, R.L
(Compiler). Proceedings: Pinyon-Juniper Conference, Reno, NV. January 13-16,
1986. Intermountain Research Station General Technical Report INT-215, pp.
109-119.
Mehringer, P.J. Jr. and P.E. Wigand. 1990. Comparison of Late Holocene environments from
woodrat middens and pollen: Diamond Craters, Oregon. In: J.L. Betancourt, T.R. Van
Devender, and PS. Martin, eds. Fossil packrat middens: the last 40,000 years of biotic
change. University of Arizona Press, Tucson, pp. 294-325.
Milne, B.T. 1988. Measuring the fractal geometry of landscapes. Applied Mathematics and
Computation 27:67-79.
Moore, P.O. and J.A. Webb. 1978. An illustrated guide to pollen analysis. John Wiley and
Sons, New York, 133 pp.
Morecroft, M.D. and F.I. Woodward. 1990. Experimental investigations on the environmental
determination of 513C at different altitudes. Journal of Experimental Botany
41 (231 ):1303-1308.
Nybom, H. and B.A. Schall. 1990. DNA "fingerprints" reveal genotype distributions in natural
populations of blackberries and raspberries (Rubus, Rosaceae). American Journal of
Botany 77:883-888.
O'Neill, R.V., J.R. Krummel, R.H. Gardner, G. Sugihara, B. Jackson, D.L DeAngelis, B.T.
Milne, M.G. Turner, B. Zygmunt, S. Christensen, V.H. Dale, and R.L. Graham. 1988.
Indices of landscape pattern. Landscape Ecology 1:153-162.
Owens, M.K., K.L. Launchbaugh, and J.W. Holloway. 1991. Pasture characteristics affecting
spatial distribution of utilization by cattle in mixed brush communities. Journal of Range
Management 44(2):118-123.
Patton, D.R. 1975. A diversity index for quantifying habitat edge. Wild. Soc. Bull.
394:171-173.
Palmer, M.W. 1988. Fractal geometry: a tool for describing spatial patterns of plant
communities. Vegetatio 75:91-102.
Payette, S. 1980. Les grandes crues glacielles de la rivire aux Feuilles (Nouveau-Quebec):
une analyse dendrochronologique. [Great ice floods of the Leaf River, New Quebec: a
dendrochronological analysis.] Naturaliste Canadien 107(4):215-225.
154
-------
Payette, S., L. Filion, and A. Delwaide. 1990. Disturbance regime of a cold temperate forest
as deduced from tree-ring patterns: the Tantar Ecological Reserve, Quebec. Canadian
Journal of Forest Research 20:1228-1241.
Pennycuick, C.J. and N.C. Kline. 1986. Units of measurement for fractal extent, applied to
the coastal distribution of bald eagle nests in the Aleutian Islands. Oecologia
68:254-258.
Peterjohn, W.T. and D.L. Correll. 1984. Nutrient dynamics in an agricultural watershed:
observations on the role of a riparian forest. Ecology 65(5):1466-1475.
Peterson, D.L. 1985. Evaluating the effects of air pollution and fire on tree growth by tree
ring analysis. In: L.R. Donoghue and R.E. Martin, eds. Weather-the drive train
connecting the solar engine to forest ecosystems. Proceedings of the Eighth
Conference on Fire and Forest Meteorology, Detroit, Michigan, April 29-May 2,1985.
Society of American Foresters and American Meteorological Society, pp. 124-131.
Peterson, D.L. and V.A. Wakefield. 1987. Detecting the effects of air pollution on tree
growth: ozone and Jeffrey pine in the Sierra Nevada. In: A. Lucier, ed. Tree rings and
forest mensuration: how can they document trends in forest health and productivity?
National Council of the Paper Industry for Air and Stream Improvement, Technical
Bulletin 523, pp. 60-63.
Peterson, D.L, V.A. Wakefield, and M.J. Arbaugh. 1987. Detecting the effects of ozone
pollution on growth of Jeffrey pine in the Sierra Nevada, California. In: G.C. Jacoby, Jr.
and J.W. Hornbeck, eds. Proceedings of the International Symposium on Ecological
Aspects of Tree-Ring Analysis U.S. Department of Energy, Publication
CONF-8608144, pp. 402^09.
Phillips, J.D. 1985. Measuring complexity of environmental gradients. Vegetatio 64:95-102.
Phillips, S.O., J.M. Skelley, and H.E. Burkhart. 1977. Growth fluctuation of loblolly pine due
to periodic air pollution levels: interaction of rainfall and age. Phytopathology
67:716-720.
Pilcher, J.R. 1990. Radioactive isotopes in wood. In: E. Cook and L. Kairiukstis, eds.
Methods of dendrochronology: applications in environmental sciences. Kluwer
Academic Publishers, Dordrecht, pp 93-96.
Puckett, L.J. 1982. Acid rain, air pollution, and tree growth in southeastern New York.
Journal of Environmental Quality 11(3):376-381.
Rex, K.D. and G.P. Malanson. 1990. The fractal shape of riparian forest patches. Landscape
Ecology 4(4):249-258.
Risser, P.G., J.R. Karr, and R.T.T. Forman. 1984. Landscape ecology: directions and
approaches. Illinois Natural History Special Publication Number 2.
155
-------
Roath, L.R. and W.C. Krueger. 1982. Cattle grazing and behavior on a forested range.
Journal of Range Management 35(3):332-338.
Robertson, A. 1986. Estimating mean windflow in hilly terrain from tamarack (Larix laricina
(Du Roi) K. Koch) deformation. International Journal of Biometeorology 30(4):333-349.
Robertson, A. 1991. Centroid of wood density, bole eccentricity, and tree-ring width in
relation to vector winds in wave forests. Canadian Journal of Forest Research
21:73-82.
Rogers, G.F. 1982. Then & now: a photographic history of vegetation change in the central
Great Basin Desert. University of Utah Press, Salt Lake City, Utah.
Rogers, G.F., H.E. Malde, and R.M. Turner. 1984. Bibliography of repeat photography for
evaluating landscape change. University of Utah Press, Salt Lake City, Utah.
Roughton, R.D. 1962. A review of literature on dendrochronology and age determination of
woody plants. Technical Bull 15:1-99. Colorado Department of Game and Fish.
Roughton, R.D. 1972. Shrub age structure on a mule deer winter range in Colorado.
Ecology 53:615-625.
Running, S.W. and J.C. Coughlan. 1988. A general model of forest ecosystem processes for
regional applications: I. Hydrologic balance, canopy gas exchange and primary
production processes. Ecological Modelling 42:125-154.
Sambrook, J., E.F. Fritsch, and T. Maniatis. 1989. Molecular cloning. A Manual. 2nd ed. Cold
Spring Harbor Laboratory Press. Cold Springs Harbor, N.Y.
Schlosser, I.J. and J.R. Karr. 1981 a. Water quality in agricultural watersheds: impact of
riparian vegetation during base flow. Water Resources Bulletin 17:233-240.
Schlosser, I.J. and J.R. Karr. 1981b. Riparian vegetation and channel morphology impact on
spatial patterns of water quality in agricultural watersheds. Environmental Management
5:233-243.
Schoonmaker, P.K. and D.R. Foster. 1991. Some implications of paleoecology for
contemporary ecology. Botanical Review 57:204-245.
Schweingruber, F.H. 1987. Potentials and limitations of dendrochronology in pollution
research. In: G.C. Jacoby, Jr. and J.W. Hornbeck, eds. Proceedings of the International
Symposium on Ecological Aspects of Tree-Ring Analysis. U.S. Department of Energy,
Publication CONF-8608144, pp. 344-<352.
Schweingruber, F.H. 1988. Tree rings, basics and applications of dendrochronology. D.
Reidel Publishing Co., Dordrecht, 292 pp.
156
-------
Schweingruber, F.H. 1990. Radiodensitometry. In: E. Cook and L. Kairiukstis, eds. Methods
of dendrochronology: applications in environmenal sciences. Kluwer Academic
Publishers, Dordrecht, pp. 55-63.
Schweingruber, F.H., H. Albrecht, M. Beck, J. Hessel, K. Joos, D. Keller, R. Kontic, K.
Lange, M. Nidedrer, C. Nippel, S. Spang, A. Spinnler, B. Steiner, and A.
Winkler-Seifert. 1986. Abrupt growth changes in tree ring sequences as ecological
indicators. Dendrochronologia 4:125-183.
Scuderi, L.A. 1984. A dendroclimatic and geomorphic investigation of late-Holocene
glaciation, southern Sierra Nevada, California. Ph.D. dissertation, University of
California, Los Angeles, 247 pp.
Senft, R.L, M.B. Coughenour, D.W. Bailey, LR. Rittenhouse, O.E. Sala, and D.M. Swift.
1987. Large herbivore foraging and ecological hierarchies. Bioscience 37(11):789-799.
Senft, R.L, L.R. Rittenhouse, and R.G. Woodmansee. 1983. The use of regression models
to predict spatial patterns of cattle behavior. Journal of Range Management
36(5):553-557.
Senft, R.L, L.R. Rittenhouse, and R.G. Woodmansee. 1985. Factors influencing patterns of
cattle grazing behavior on shortgrass steppe. Journal of Range Management
38(1):82-87.
Serre-Bachet, F. 1987. Tree-growth, ARMA modeling, and pollution. In: G.C. Jacoby, Jr.
and J.W. Hornbeck, eds. Proceedings of the International Symposium on Ecological
Aspects of Tree-Ring Analysis. U.S. Department of Energy, Publication
CONF-8608144, pp. 591-598.
Shiyatov, S.G. 1990. The use of dendrochronological methods in determining tree windthrow
time. Lesovedenie (0)2:72-81.
Short, H.L. 1986. Rangelands. In: A.Y. Cooperrider, R.J. Boyd, and H.R. Stuart, eds.
Inventorying and monitoring of wildlife habitat. U.S. Dept. of Interior, Bureau of Land
Management, Service Center, Denver, CO, 858 pp.
Siegel, R.D. 1983. Paleoclimatic significance of D/H and 13C/12C ratios in Pleistocene and
Holocene wood. Unpublished M.S. Thesis, Department of Geosciences, University of
Arizona.
Smith, M.S. 1988. Modeling: three approaches to predicting how herbivore impact is
distributed in rangelands. New Mex. Agr. Exp. Sta. Reg. Res. Rep. 628.
Smith, S.D. and R.S. Nowak. 1990. Ecophysiology of plants in the intermountain lowlands.
In: C.B. Osmond, LF. Pitelka, and G.M. Hidy, eds. Plant biology of the basin and
range. Ecological Studies, Vol. 80. Springer-Verlag, Heidelberg, pp. 179-241.
157
-------
Spaulding, W.G. 1980. The presettlement vegetation of the California Desert. Riverside,
Desert Planning Staff, California Bureau of Land Management unpublished report, 97
pp.
Spaulding, W.G. 1981. The late Quaternary vegetation of a southern Nevada mountain
range. Unpublished Ph.D. dissertation. University of Arizona, Tucson, 271 pp.
Spaulding, W.G. 1985. Vegetation and climates of the last 45,000 years in the vicinity of the
Nevada Test Site, south-central Nevada. U.S. Geological Survey Professional Paper
1329,83pp.
Stahle, D.W. 1990. The tree-ring record of false spring in the southcentral USA. Ph. D.
Dissertation. Arizona State University, 272 pp.
Stockton, C.W. and D.M. Meko. 1975. A long-term history of drought occurrence in western
United States as inferred from tree rings. Weatherwise 28:244-249.
Stockton, C.W. and D.M. Meko. 1983. A long-term history of drought occurrence in the
Great Plains as reconstructed from long-term tree-ring records. Journal of Climate and
Applied Meteorology 22:17-29.
Strauss, S.H. and A.M. Doerksen. 1990. Restriction fragment analysis of pine phylogeny.
Evolution 44(4):1081-1096.
Strauss, S.H., A.M. Doerksen, and J.R. Bryne. 1990. Evolutionary relationships of
douglas-fir and its relatives (genus Pseudotsuga) from DNA restriction fragment
analysis. Canadian Journal of Botany 68:1502-1510.
Swetnam, T.W. 1990. Fire history and climate in the southwestern United States. In: J.S.
Krammes, ed. Effects of fire management of southwestern natural resources. USDA
Forest Service General Technical Report RM-191, pp. 6-17.
Swetnam, T.W. and J.L Betancourt. 1990. Fire-Southern Oscillation relations in the
southwestern United States. Science 249:1017-1020.
Swetnam, T.W. and J. Dieterich. 1985. Fire history of ponderosa pine forests in the Gila
Wilderness, New Mexico. In Proceedings - Symposium and Workshop on Wilderness
Fire USDA Forest Service General Technical Report INT-182, pp. 390-397.
Swetnam, T.W. and A.M. Lynch. 1989. A tree-ring reconstruction of western spruce
budworm history in the southern Rocky Mountains. Forest Science 35(4):962-986.
Symeonides, C. 1979. Tree-ring analysis for tracing the history of pollution: Application to a
study in northern Sweden. Journal of Environmental Quality 8(4):482-486.
Tan, T.K. 1980. The physiological basis for producing lead pollution histories from tree rings.
Ph.D. dissertation, Council for National Academic Awards.
158
-------
Tessier, L, F. Serre-Bachet, and J. Guiot. 1990. Pollution fluore et croissance radiale des
conifres en Maurienne (Savoie, France). [Fluoride pollution and radial growth of
conifers in Maurienne (Savoy, France).] Annales Sciences Forestires 47:309-323.
Thompson, M.A. 1981. Tree rings and air pollution: A case study of Pinus monophylla
growing in east central Nevada, USA. Environmental Pollution, Series A: Ecological
and Biological 26(4):251-266.
Thompson, R.S. 1990. Late Quaternary vegetation and climate in the Great Basin. In: J.L.
Betancourt, T.R. Van Devender, and P.S. Martin, eds. Fossil packrat middens: the last
40,000 years of biotic change. University of Arizona Press, Tucson, pp. 200-239.
Tian, T.K. and N.W. Lepp. 1977. Further studies on the development of tree-ring analysis as
a method for constructing heavy metal pollution histories. Trace Substances in
Environmental Health 11:440-447.
Toft, N.L., J.E. Anderson, and R.S. Nowak. 1989. Water use efficiency and carbon isotope
composition of plants in a cold desert environment. Oecologia 80:11-18.
Travis, D.J., H.D. Grissino-Mayer, and P.W. Suckling. 1989. The impact of ice storms on
tree-ring widths of loblolly pine in northern Georgia. Proceedings, Sixth Conference on
Applied Climatology, American Meteorological Society, pp. J38-J40.
Treshow, M., E.K. Sutherland, and J.P. Bennett. 1987. Tolerance and susceptibility in air
pollution, a new direction in sampling strategy: A case study of Pinus strobus L. in
Acadia National Park, Maine. In: G.C. Jacoby, Jr. and J.W. Hornbeck, eds.
Proceedings of the International Symposium on Ecological Aspects of Tree-Ring
Analysis. U.S. Department of Energy, Publication CONF-8608144, pp. 392-400.
Turner, M.G. 1989. Landscape Ecology: the effect of pattern on process. Annual Review of
Ecological Systematics 20:171-197.
Turner, M.G. 1990. Landscape changes in nine rural counties in Georgia. Photogrammetric
Engineering & Remote Sensing 56(3):379-386.
U.S. Environmental Protection Agency. 1990. EMAP Landscape Characterization research
and implementation plan. Peer Reviewed Draft, EPA-600/X-90/1099A, Environmental
Monitoring Systems Laboratory, Las Vegas, NV.
Vaganov, E.A. 1990. The tracheidogram method in tree-ring analysis and its application. In:
E.R. Cook and L. Kairiukstis, eds. Methods of dendrochronology: applications in
environmental sciences. Kluwer Academic Publishers, Dordrecht, pp. 63-76.
Valkovic, V., D. Rendic, and E.K. Biegert. 1979. Trace element concentrations in tree rings
as indicators of environmental pollution. Environment International 2:27-32.
Van Devender, T.R. and W.G. Spaulding. 1979. Development of vegetation and climate in
the southwestern United States. Science 204:701-710.
159
-------
Vins, B. and J. Pollanschutz. 1977. Erkennung und Beurteilung Immisionsgeschadigter
Walder an Hand von Jahrringanalysen. [Detection and evaluation of air pollution in
forests from tree-ring analysis.] Allgemeine Forstzeitung 88(6): 146-148.
Wade, J.E. and E.W. Hewson. 1979. Trees as a local climatic wind indicator. Journal of
Applied Meteorology 18:1182-1187.
Waring, R.H. 1987. Distinguishing pollution from climatic effects by the analysis of stable
isotope ratios in the cellulose of annual growth rings. In: S.H. Bicknell, ed.
Proceedings, California Forest Response Program Planning Conference California Air
Resources Board, Western Conifers Research Cooperative, pp. 90-98.
Webb, T. and R. A. Bryson. 1972. Late and post-glacial climatic change in the Northern
Mid-West USA: quantitative estimates derived from fossil pollen spectra by
multivariate statistical analysis. Quaternary Research 2:70-115.
Webb, T. Ill and D.R. Clark. 1977. Calibrating micropaleontological data in climatic terms: a
critical review. Annals of the New York Academy of Science 288:93-118.
Weinstein, D.A. and R. Beloin. 1990. Evaluating effects of pollutants on integrated tree
processes: a model of carbon, water, and nutrient balances. In: R.K. Dixon, R.S.
Meldahl, G.A. Ruark, and W.G. Warren, eds. Forest growth: process modeling of
responses to environmental stress. Timber Press, Oregon, pp. 313-323.
Wells, P.V. 1983. Paleobiogeography of montane islands in the Great Basin since the
glaciopluvial. Ecological Monographs 53(4):341-382.
Wells, P.V. and C.D. Jorgensen. 1964. Pleistocene woodrat middens and climatic change in
the Mohave Desert: a record of juniper woodlands. Science 145:1171-1174.
Wells, P.V. and R. Berger. 1967. Late Pleistocene history of coniferous woodland in the
Mohave Desert. Science 155:1640-1647.
White, F.C., J.R. Hairston, W.N. Musser, H.F. Perkins, and J.F. Reed. 1981. Relationship
between increased crop acreage and nonpoint-source pollution: a Georgia case study.
Journal of Soil and Water Conservation 36:172-177.
Whittaker, R.H. 1975. Communities and ecosystems. New York, MacMillan Publishing Co.,
Inc.
Wigand, P.E. 1987. Diamond Pond, Harney County, Oregon: vegetation history and water
table in the eastern Oregon desert. Great Basin Naturalist 47(3):427-458.
Yakir, D., M.J. DeNiro, and J.R. Gat. 1990. Natural deuterium and oxygen-18 enrichment in
leaf water of cotton plants grown under wet and dry conditions: evidence for water
compartmentation and its dynamics. Plant Cell and Environment 13:49-56.
160
-------
Yanosky, T.M. 1982. Hydrologic inferences from ring widths of flood-damaged trees,
Potomac River, Maryland. Environmental Geology 4:43-52.
Yanosky, T.M. 1983. Evidence of floods on the Potomac River from anatomical abnormalities
in the wood of flood-plain trees. U.S. Geological Survey Professional Paper 1296,42
pp.
Yokobori, M. and S. Ohta. 1983. Combined air pollution and pine ring structure observed
xylochronologically. European Journal of Forest Pathology 13:30-45.
• US GOVERNMENT PRINTING OFT1CE 1 993 -75 o -002' 6 o 176
161
-------
m
-D
£
O5
ro
o
33
ib
co
o
o
ft ~° O
§11
Sr CO
>< —
-~ 03
O c
"* CO
-D 5'
~2- CD
< CO
CO CO
CD
CO
CD
o o m
O
en
CO
C
D
CO
•*>• o ' "°
DO >
CD CQ
CO CD
CD 13
0) O
O
—\
c:
"O
H
CD
^r
IT
U
a
o
o
3
CD
Q.
CD
Eu
o
3-
o
8
~D
*<
IT
to
O
O
5
W
13
Q.
3
13
O
5
CD
address i
^
31
tD
•<
O
C
Q.
O
o
5
in
=r
O
CD
O
CD
§
3-
CD
cvi
CD
CD
•§
y?
O
X
m
O
7^
X
^Q
m
D
II
-V ">
1*
0. 0
§8
U "O
5^
0)
13
cx
ro
^
n to the ai
LJ.
a
CD
W
U
31
CD
C
~o
"a
Q
"*
-a
CD
fa
ffl
CD
3
ytr
CD
w.
13
O
CD
C/)
C/)
OJ
ry change
CO
o
13
3~
o~
CD
i
w
o~
CD
T)
O
-D W
m rj
30 o£
^ C) C
= _ m r—
2:
o
O
co
en
------- |