United States
Environmental Protection
Agency
Environmental Research
Laboratory
Corvallis OR 97330
EPA-600, 3-80-002
January 1980
Research and Development
&EPA
Photochemical
Oxidant Air Pollution
Effects on a Mixed
Conifer Forest
Ecosystem
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7 Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ECOLOGICAL RESEARCH series. This series
describes research on the effects of pollution on humans, plant and animal spe-
cies, and materials. Problems are assessed for their long- and short-term influ-
ences. Investigations include formation, transport, and pathway studies to deter-
mine the fate of pollutants and their effects. This work provides the technical basis
for setting standards to minimize undesirable changes in living organisms in the
aquatic, terrestrial, and atmospheric environments.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/3-80-002
January 1980
PHOTOCHEMICAL OXIDANT AIR POLLUTION EFFECTS ON A
MIXED CONIFER FOREST ECOSYSTEM
Final Report
Editor
0. C. Taylor
Principal Authors
R. N. Kickert, J. R. McBride, P.R. Miller,
C. P. Ohmart, R. J. Arkley, D. L. Dahlsten,
F. W. Cobb, Jr., J. R. Parmeter, Jr.,
R. F. Luck, and 0. C. Taylor
University of California
Riverside, California 92521 and
Berkeley, California 94720
Contract No. 68-03-2442
Project Officer
R. G. Wilhour
Terrestrial Division
Corvallis Environmental Research Laboratory
Corvallis, Oregon 97330
This study conducted under the direction of
0. C. Taylor, Project Principal Investigator and Associate Director
Statewide Air Pollution Research Center
University of California
Riverside, California 92521
CORVALLIS ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CORVALLIS, OREGON 97330
-------
DISCLAIMER
This report has been reviewed by the Corvallis Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for pub-
lication. Approval does not signify that the contents necessarily reflect
the views and policies of the U.S. Environmental Protection Agency, nor
does mention of trade names or commercial products constitute endorsement
or recommendation for use.
ii
-------
FOREWORD
The San Bernardino National Forest (SBNF) has been under stress from
photochemically produced oxidant air pollutants for more than three decades.
With the rapid industrial and urban growth in the South Coast Air Basin
during the past 20 to 30 years the impact on forest species has intensified.
Loss of ponderosa and Jeffrey pine trees has increased dramatically as
pollutant levels have risen and frequency and length of pollutant attacks
have expanded. Pollutant effects on interrelated subsystems of the SBNF
ecosystem have been studied in 18 plots established in selected regions of
the forest. The plots were selected to represent sites of varying pollutant
dosages while retaining as much uniformity of plant species and environment
as possible. Studies by scientists from the Berkeley and-Riverside campuses
of the University of California collected data which will be used for a
group of linked models. The models will aid in describing pollutant impact
on subsystems and such models should be useful in anticipating or predicting
responses in other areas under similar conditions. Data gathered during the
period of this contract will add significantly to information collected
under previous contracts and during the subsequent two years of an EPA grant
for the purpose of refining the models of a western coniferous forest
ecosystem under stress from long—term exposure to photochemically produced
air pollutants.
iii
-------
ABSTRACT
EPA Contract 68-03-2442 provided support for three years of the studies
to determine the chronic effects of photochemical oxidant air pollutants
on a western mixed conifer forest ecosystem. A progress report for the
years 1974-'75 and 1975-'76 was published in the Ecological Research Series,
EPA-600/-3-77-104. The report being submitted deals specifically with the
year 1976-'77 and is the final report for EPA Contract 68-03-2442 which has
funded a three year portion of the study initiated in 1972 and is scheduled
to terminate May 31, 1980.
A computer data bank was partially developed in the early years of
the study at the Lawrence Livermore Laboratory and was subsequently revised
and moved to the computer at the University of California, San Francisco.
Verification and auditing of datasets is well underway and several sets are
now ready for cross-disciplinary analysis for modeling. Computer simulation
programs have been written for some of the subsections.
Subsystems which received greatest attention during this study period
were: major tree species response to oxidant dose; tree population dynamics;
tree growth; moisture dynamics; soil chemical and physical properties;
tree mortality relative to disease, insects and other factors; epidemiology
of forest tree pathogens with emphasis on Fomes annosus; cone and seed
production; tree seedling establishment; litter production and litter
decomposition relative to microfloral decomposer populations. Progress is
being made in preparation of models for the purpose of describing the
behavior of the interlinked subsystems. Since much progress has been made
in verifying accuracy of data and of identifying information in the data
bank the study of subsystems interaction should be accelerated.
This report in conjunction with the Ecological Research Series report
EPA-600/3-77-104 is submitted to fulfill the requirement for a final report
for EPA Contract 68-03-2442.
iv
-------
CONTENTS
Page
Disclaimer Notice • ii
Foreword iii
Abstract • iv
Contents v
Figures vi
Tables x
Abbreviations and Symbols ..... xiv
Acknowledgements .......... xv
1. Introduction 1
2« Ecosystem Simulation Modeling 3
3. Tree Population Dynamics Subsystem 28
4. Canopy Response Subsystem 38
5. Tree Growth Subsystem 65
6. Physical and Chemical Properties of Soil, Including
Moisture Dynamics 77
7. Stand Tree Mortality Subsystem: Bark Beetle Popu-
lation Dynamics 86
8. Forest Tree Pathogen Epidemiology Subsystem 115
9. Cause and Extent of Tree Mortality 132
10. Tree Seedling Establishment Subsystem 136
11. Cone & Seed Production For Dominant Conifer Tree
Reproduction 143
12. Litter Production Subsystem 150
13. Foliage Litter Decomposition Subsystem: Microbial
Activity and Nutrient Cycling 167
References • 179
Appendix 185
-------
FIGURES
Number Page
1 The system simulation modeling process with bold arrows
showing information flows which were behind schedule and
consequently were retarding modeling progress 4
2 Information structure for the computerized SBNF data base
accesssible from remote dial-up terminal via UCSF/CMS
time-sharing mode 8
3 Example of Part III of the SBNF data dictionary: Dataset
descriptors 11
4 Data processing environments used for simulating modeling
and data base manipulation on IBM 370/145 computer at the
U.C. San Francisco, via 30 character per-second telecommuni-
cations 13
5 Simulation sequence between subsystems showing associated
datasets for submodel quantification and validation ..... 21
6 Facies map of Dogwood Plot 31
7 Phenogram illustrating cluster formation 35
8 Territorial map of discriminant score 1 (horizontal) vs.
discriminant score 2 (vertical) . .' 36
9 Trend of seasonal oxidant dose at a representative San
Bernardino mountain station from 1968-1977 50
10 Seasonal pattern of potential transpiration, ratio of
actual over potential transpiration and cumulative oxi-
dant dose at Deer Lick, 1976 52
11 Comparison of injury to the 1975 needle whorl with poten-
tial transpiration and the ratio of actual over potential
transpiration at Camp Angelus in 1976 53
12 Relationship of daily potential transpiration and daily
oxidant dose at Camp Angelus during May through September,
1976 54
vi
-------
Figures Continued
Number Page
13 Changes of 1975 needle whorl injury in 1976 and 1977 at
Camp Angelus 56
14 Changes in the percent of the total needle length of
selected ponderosa pines with chlorotic mottle in relation
to two measures of cumulative oxidant dose 58
15 Average height growth of ponderosa and Jeffrey pine sap-
lings at plots experiencing different levels of chronic
oxidant injury 61
16 Height growth of ponderosa pine saplings in three injury
categories at Camp Paivika between 1967 and 1976 62
17 Growth trends for trees on the COO plot from 1946 to 1975 in
the San Bernardino National Forest 74
18 Growth trends for trees on the BP plot from 1946 to 1975
in the San Bernardino National Forest 74
19 Growth trends for trees on the GVC plot from 1946 to 1975
in the San Bernardino National Forest 75
20 Growth trends for trees on the DL plot from 1946 to 1975
in the San Bernardino National Forest 75
21 Growth trends for trees on the NEGV plot from 1946 to 1975
in the San Bernardino National Forest 76
22 Growth trends for trees in the HV plot from 1946 to 1975
in the San Bernardino National Forest 76
23 Moisture retention curves for Dogwood plot, Site 2 85
24 Graphic summary of the population sampling procedures used
for the western pine beetle showing datasets and the type
of information included for the San Bernardino study .... 88
25 Cost of sampling for fixed variances as a function of bark
sample size for total larvae of the Jeffrey pine beetle, 2
bark sample units per height, San Bernardino National Forest,
1974 109
26 Sample variance as a function of bark sample size for fixed
costs for total larvae of the Jeffrey pine beetle, 2 bark
sample units per height, San Bernardino National Forest, 1974 HO
vii
-------
Figures Continued
Number Page
27 Variance as a function of cost for different number of
1000 cm^ samples per height for gallery length of the
mountain pine beetle in pondrosa pine, San Bernardino
National Forest, 1974 Ill
28 Relationship between oxidant air pollution injury and
proximal colonization of inoculated pine roots by
Fomes annosus 118
29 Age class distribution of ponderosa and/or Jeffrey pine trees
on three plots ..... 145
30 Crown class distribution of ponderosa and/or Jeffrey pine
trees on three plots 146
31 Diameter class distribution of ponderosa and/or Jeffrey pine
trees on three plots 147
32 Height class distribution of ponderosa and/or Jeffrey pine
trees on three plots 148
33 Total pine needle litter collected related to oxidant injury
score 1975, 1976 152
34 Size of pine needles in litter related to oxidant injury
score—high and low rainfall plots 154
35 Nitrogen content of pine needle litter related to oxidant
injury score—high rainfall plots 156
36 Nitrogen content of pine needle litter related to oxidant
injury score—low rainfall plots 157
37 Potassium content of pine needle litter related to oxidant
injury score—high rainfall plots 158
38 Potassium content of pine needle litter related to oxidant
injury score—low rainfall plots 159
39 Phosphorous content of pine needle litter related to oxidant
injury score—high rainfall plots 160
40 Phosphorous content of pine needle litter related to oxidant
injury score—low rainfall plots 161
viii
-------
Figures Continued
Number Page
41 Calcium content of pine needle litter related to oxidant
injury score—high rainfall plots 162
42 Calcium content of pine needle litter related to oxidant
injury score—low rainfall plots 163
43 Soluble phosphorous content of surface soils related to
oxidant injury score of pine trees—high rainfall plots . . . 165
44 Soluble phosphorous content of surface soils related to
oxidant injury score of pine trees—low rainfall plots . . . 166
45 Source and destination of 960 decomposition study envelopes . 169
46 Source and destination of 160 decomposition study envelopes . 170
47 Evenness (E) in a three species community 173
48 Percent weight loss (30C) incurred by needle litter on four
plots, representing the range of air pollution impact on
ponderosa and Jeffrey pines 175
49 Percent weight loss (30C) incurred by needle litter follow-
ing transfer between and within species 176
IX
-------
TABLES
Number Page
1 Part 1 of the SBNF data dictionary: The dataset index
CMS Time-sharing commands used for on-line file management
of SBNF data base 9
2 CMS Time-sharing commands used for on-line file management
of SBNF data base 14
3 CMS Time-sharing edit-environment commands used for finding,
correcting, and displaying the contents of a dataset or
program file in the SBNF data base 15
4 Status of distributed versus centralized data processing
of SBNF data base for various subprojects as of August, 1977 • 17
5 Distribution of datasets by development stage for the SBNF
data base at U. C. Berkeley 20
6 Number of facies identified on the 18 permanent plots used
to monitor air pollution injury to forest trees 32
7 Species composition, tree height, and cover of facies on
Dogwood plot 33
8 Major inputs, internal operations and outputs of the trans-
piration simulator 41
9 Description of information collected to describe the within
season development of oxidant injury symptoms on ponderosa
pine (PP), Jeffrey pine (JP), white fir (WF), black oak
(BO) 43
10 Subjective categories for description of oxidant injury
symptoms on ponderosa (PP), Jeffrey pine (JP), white fir
(WF) , and black oak (BO) 45
11 Description of data types and frequency of data collection
for each type at major vegetation plots in 1976 and early
1977 47
12 Frequency of different transitional combinations of five
classes of spring and summer days 51
-------
Tables Continued
Number Page
13. Trends in ponderosa and Jeffrey pine chronic injury scores
and tree death at eighteen permanent plots 64
14. Annual precipitation (cm) at Squirrel Inn #2 weather station
and an analysis of variance among the ten year intervals
that correspond to the ten year growth period 67
15. Rainfall data for the six study plots (mean annual pre-
cipitation cm) 68
16. Mensurational descriptions of the six sample plots 70
17. Age groups, soil types, mean annual radial growth incre-
ment (mm) and standard deviation for each growth period, F
values from ANOVA, and explained variance (w^) for each strat—
fied group of sample trees on each plot 72
18. Soil water content at various matric suctions for three pro-
file depths 80
19- Soil water content at various matric suctions for profile
depths of 200 cm or greater 83
20. Available soil water and clay content as percent of the
whole soil for tree soil depth intervals 84
21. Height, diameter, and length of infestation for western
pine beetle sample trees 89
22. Western pine beetle infested ponderosa pine that were sampled
between 1973 and 1976 ranked by oxidant damage class 95
23. Western pine beetle mean egg dissection variable by year
and generation for whole ponderosa pines, San Bernardino
National Forest, 1973-1976 96
24. Variables calculated from western pine beetle egg disc
sample dissection by generation from 1973 to 1976. San
Bernardino National Forest . 97
25. Correlation of western pine beetle egg disc dissection
variables with year, generation and tree oxidant ratings.
San Bernardino National Forest, 1973-1976 97
26. Means by generations on the last sample date for western
pine beetle brood, parasites and predators from X-ray analysis
of sample bark discs. San Bernardino National Forest, 1973-
1976 98
xi
-------
Tables Continued
Number Page
27. Significance of multiple regression coefficients for the
last sample date of X-ray, rearing, and sticky cartons for
the western pine beetle and its natural enemies. San Ber-
nardino National Forest, 1973-1976 99
28. Mean western pine beetle and natural enemy emergence by
generation for the last sample date of laboratory reared
discs. San Bernardino National Forest, 1973 to the first
generation of 1975 100
29. Mean western pine beetle and natural enemy emergence by
generation for the last sample date of sticky cartons (field
research). San Bernardino National Forest, 1973 to the forest
generation of 1976 100
30. Results of using Tang's calculation on Jeffrey pine beetle
data: Tree effect 101
31. Results of using Tang's calculations on Jeffrey pine beetle
data: Height effect 101
32. Results of using various tests on attack density, sample
size of data, to determine differences between pairs of
trees, values in attacks/cm^ 102
33. Preliminary summary of final smog damage ratings for pines
killed insects on established vegetation plots, 1973-1975 . . 102
34. Infection of inoculated pine roots with Fomes annosus in
relation to the severity of air pollution injury 116
35. Colonization of inoculated pine roots by Fomes annosus in
relation to the severity of air pollution injury 117
36. Number of pine stumps inoculated with Fomes annosus by
site and species 120
37. Relationship between the surface colonization of inoculated
pine stumps by Fomes annosus and the severity of air pollution
injury 121
38. Downward colonization and colonization rate of Fomes annosus
in inoculated pine stump related to air pollution injury ... 122
39. Volume of inoculated stumps colonized by Fomes annosus in
relation to air pollution injury 123
xii
-------
Tables Continued
Number Page
40. Influence of ozone on the linear growth rate and conidial
production of Fomes annosus 125
41. Influence of ozone on conidial germination of Fomes annosus
(Isolate: JL1) 126
42. Influence of ozone on conidial germination of Fomes annosus
(Isolate: PP1) 127
43. Influence of ozone exposure on conidial germination and
germ tube extension of Fomes annosus 128
44. Effects of ozone on colonization of pine discs by Fomes
annosus 129
45. Influence of ozone on Fomes annosus conidial germination
through successive generations of exposure 131
46. Tree mortality by cause and forest type in the San Bernardino
mountains in 1976 134
47. Tree mortality by pest complexes and species in the San
Bernardino mountains in 1976 134
48. Percent of mortality centers in the mixed conifer and yellow
pine forests in relation to oxidant injury 135
49- Pathogenicity test summary 137
50. 1975-1976 SBNF seedling establishment field study, germinated
seeds, means per screened mini-plot in May, 1976 • 139
51. 1975-1976 SBNF seedling establishment field study, sur-
viving seedlings, means per screened mini-plot as of July
1976 139
52. Status in acquiring plot data on the 19 study plots located
in the San Bernardino mountains of southern California .... 144
53. Needle size and content of N, P, K and Ca in needle fall . . . 155
54. Estimated total radiation (R) and temperature (T) accumu-
lated during one clear day at measurement points beneath
the integrated field decomposition study trees 177
xiii
-------
LIST OF ABBREVIATIONS AND SYMBOLS
ABBREVIATIONS
SCAB — South Coast Air Basin
SBNF — San Bernardino National Forest
CP — Camp Paivika vegetaton plot
BP — Breezy Point vegetation plot
TUN 2 — Tunnel 2 vegetation plot
DWA — Dogwood A vegetation plot
DWB — Dogwood B vegetation plot
DL — Deer Lick
SF — Sky Forest vegetaton plot
UCC — University Conference Center vegetation plot
COO — Camp 0-Ongo vegetation plot
GVC — Green Valley Creek vegetation plot
NEGV — Northeast Green Valley vegetation plot
SV — Snow Valley vegetation plot
BL — Bluff Lake vegetation plot
SC — Sand Canyon vegetation plot
HV — Holcomb Valley vegetation plot
CA — Camp Angeles vegetation plot
SCR — Schneider Creek vegetation plot
BF — Barton Flats vegetation plot
CAO — Camp Osceola vegetation plot
HB — Heart Bar vegetation plot
PP — ponderosa pine
JP — Jeffrey pine
WF — white fir
dbh — diameter at breast height
ppm — parts per million
SYMBOLS
„£ — microgram
Oo — ozone
xiv
-------
ACKNOWLEDGMENT S
Administrative services for the project were coordinated by 0. Clifton
Taylor, Associate Director, Statewide Air Pollution Research Center, Univer-
sity of California, Riverside, CA 92521.
Portions of this report were contributed by the following authors
and assistants:
Pages
iii, iv
1
Author
0. Clifton Taylor
Contribution
Foreword, Abstract,
Introduction
Ronald N. Kickert
Assisted by
Robert Thomson
Ecosystem Simulation
Modeling
28
38
65
77
86
115
132
Joe McBride
Paul Miller
Cliff Ohmart
Reviewed by Rick Laven
and project to be con-
tinued by Rick Laven
Rodney J. Arkley,
P- L. Gersper and
R. Glauser
Donald L. Dahlsten
Fields W. Cobb and
Robert L. James
Joe R. McBride,
Donald L. Dahlsten,
and Fields W. Cobb
Tree Population
Dynamics Subsystem
Oxidant Dose — Canopy
Response Subsystem
Oxidant - Tree Growth
Physical and Chemical
Properties of Soils and
Moisture Dynamics
Stand Tree Mortality
Subsystem - Bark Beetle
Population Dynamics
Epidemiology of Forest
Tree Population
Cause and Extent of
Tree Mortality
136
Fields W. Cobb
Forest Tree Seedling
Establishment
xv
-------
Acknowledgments continued
Pages Author
Robert F. Luck
Rodney J. Arkley
John R. Parmeter
and J. N. Bruhn
Contribution
Cone and Seed Production
for Dominant Conifer Tree
Production
Litter Production
Foliage Litter Decom-
position
Particular thanks is extended to the District Rangers and Staff of
the Arrowhead, Big Bear, and San Gorgonio Districts of the San Bernardino
National Forest for their interest and cooperation in regard to field study
activities.
Special thanks is also extended to the many research assistants,
staff research associates and others who have made this study possible. A
few of these participants include:
Maureen A. Thomas
Kathrynn E. Banbury
Nancy L. Bruhn
David L. Rowney
Isabel F. Alvarez
G. Nick McKibben
William A. Copper
Diana Doyal
Edith Reisner
Robert Van Doren
A special expression of thanks is extended to Donna M. Shaw for typing
of the manuscript, to Paul Miller for proofing and invaluable assistance
in compiling the report and to Maureen Thomas for drafting and preparation
of many of the figures.
xvi
-------
INTRODUCTION
This is a progress report for the final year of a study partially
supported by EPA Contract 68-03-2442. This report and a previous report
designated as EPA-600/3-77-104 constitute the final report for the three
year contract which has supported a multidisciplinary research effort
initiated in 1972. The study is expected to terminate May 31, 1980.
This continuing-long term study is an effort to identify and quantify
chronic effects of oxidant air pollutants on individual units of a forest
ecosystem and to model the interactions between units which are initiated,
stimulated or driven by the pollutant impact. The units of the ecosystem
selected for study were those considered to be most susceptible to pollutant
impact those units which are known to play a very important role in struc-
turing the total ecosystem character. We anticipate that models developed
for this study will be useful: to design forest and recreational area
management programs; to predict long term changes in ecosystem structure
when air pollutants are present; and perhaps to aid in establishing reason-
able pollutant standards to protect against serious undesirable changes in
the ecosystem.
The forest ecosystem is subjected to numerous stresses which favor
the development of some organisms and suppress development of others.
Air pollutants represent an additional man made stress in the complex,
therefore, any study of long-term chronic air pollutant effect must be
accompanied by evaluations of impact from other stress factors. Since the
intensity of stress produced by any of these factors varies widely over time
and since plant response is affected by interaction of the stress factors
evaluation of the air pollutant involvement in ecosystem changes become a
very complex study.
The San Bernardino National Forest (SBNF) has been exposed to an
increasing annual dosage of photochemical oxidants during the preceding 3 or
4 decades as industrial and urban development in the South Coast Air Basin
(SCAB) expanded at a phenomenal rate. Abnormalities, later identified as
oxidant air pollutant injury, were causing concern among residents and U.S.
forestry officials in the early 1950's. The injury on ponderosa pine was
initially thought to be associated with hydrogen fluoride and perhaps other
air pollutants released by specific industries which were relatively new to
the SCAB. Research during the 1950's largely dispelled this theory and
implicated the oxidant air pollutants which are the responsibility of a
broad sector of man's activities.
The SBNF, located at the east end of the South Coast Air Basin, is
subjected to an ebb and flow of polluted air from the SCAB as the alternate
diurnal "pumping" of high desert and marine air through the SCAB occurs.
-------
Typically, during the recognized smog season, air flow during the morning
is from the east toward the coastal region; and as the day warms, the
flow is essentially reversed to deliver marine air laden with oxidant
pollutants to the SBNF.
Evaluation of direct effects of oxidant pollutants in the SBNF has
for the most part been confined to the dominant tree species although it
is known that a wide variety of species of green plants are adversely
affected. These primary producers are critical elements in an ecosystem to
provide food and shelter for all other organisms in the system. Chronic
injury to the green vegetation may over time significantly change the source
of energy, protection and general habitat of numerous consumer organisms.
An understanding of changes in plant communities suffering from air pollu-
tant injury is essential if one is to predict the fate of an ecosystem
impacted by a growing and changing industrial and urban complex*
This study, including the modeling effort, is based on the assumption
that the effect of air pollutants or any other stress element will be
ultimately transferred to numerous other units of the ecosystem. It would
then be expected that a gradual or insidious but significant change in the
ecosystem might be expected.
-------
ECOSYSTEM SIMULATION MODELING OF MIXED CONIFER FOREST UNDER PHOTOCHEMICAL
AIR POLLUTION
Introduction
Can experiments be conducted on forest land, using different kinds of
long-term air pollution trends to determine ecosystem response, without
actual manipulation of the forest, and without decades of waiting for the
results? This question leads to the following objectives of the SBNF
program: 1) to design forest ecology systems models for forecasting ecolog-
ical effects of photochemical air pollutants in southern California mixed
conifer forest ecosystems; 2) to evaluate the adaptability of systems models
to other pollutant types and other forest types; and 3) to evaluate the
forecasted consequences of photochemical air pollutants in forest ecosystems
in terms of human welfare effects. A discussion of these objectives,
together with a review of other scientists' published thoughts related to
them, can be found in a previous progress report (Kickert, 1977).
Methods
General systems philosophy and systems analysis techniques involving
digital computer simulation modeling methods were applied in working toward
the objectives of the project (Figure 1).
System Model Development--
The first step in the development process was to clearly define the
various problem solving goals for which the simulation models would be used.
This was done by extensive discussion with all project scientists and
consideration of the results reported by Kickert (1977). The discussions
were directed at drawing out of each investigator a qualitative description
of the relations between changeable properties of the forest as a system, as
he conceived them to exist in the subsystem(s) pertinent to his major
role(s) in the project. The relationships were then diagrammed and inte-
grated into a graphic flow model (Figure 28 in Kickert, 1977). The next
step is to convert the set of graphic models into sets of transfer functions
to describe annual rate of change expressed as finite difference equations.
Conditional logic for threshold and time-delay conditions typical of biolog-
ical phenomena will be included. The mathematics and associated qualitative
conditional logic is written in a high-level computer language. We will
then draw upon the project data base, assembled from the data collected by
the investigators, and use quantitative relations discovered by them, to
refine the mathematical form of transfer functions in the various submodels,
to set the values of species and site parameters, and to serve as external
data for driving certain functions. Methods of handling the data base are
discussed under achievements below because these have been reorganized and
-------
Model Planning
CO
1
o
u_
Q.
t-
o
1
(GOAL DEFINITIONS J
1
TABLES
U
A
*
Model
n
Model
Development
A
r / \
^ FLOW PARADIGM _/ ™$™\
TIME-SPACE \^ Mnnns /"
RESOLUTION \ nuutLs /
V
J
1
1
! Ql
i
i
1 cot
1
1 ,
* 1
Application |
~{ 1
1 i
— . ,
SYSTEM MODEL ' 1
t !_
DERIVATION OF
PROJECTIONS ON
SUBSYSTEMS 6
WHOLE SYSTEM |
V
GpUBLICATION-\
lEMONSTRATION J
1
MODIFY
/
r
\
\ ORIGINAL
DEVELOPMENT
I
i 1
FLOW CHART
COMPONENTS
& CONTROLS
J
i '
JANTIFICATION
6 •<
;ONDITIONAL
LOGIC
! i
1PUTER CODING
DEBUGGING
*
VALIDATION
1
SENSITIVITY
ANALYSIS
Figure 1. System simulation modeling process with bold arrows showing
information flows which were behind schedule and consequently
were retarding modeling progress.
-------
re-defined during the period which this report covers.
Model Reliability Evaluation and Use—
After conversion to mathematical computer programs and the essential
debugging, the set of simulation programs are run and their behavior is
compared to observations of real forests change over a number of years.
When this evaluation of reliability is acceptable, we then perform experi-
ments on the simulated forest to determine the probability of long-term
responses of various forest properties under alternative future trends in
air quality, forest meteorology, and harvesting practices externally imposed
on the simulated ecosystem.
Transfer of Ecological Discoveries to Social Scientists—
Results of these experiments are discussed by project investigators
and subsequently communicated to scientists outside of the project who
are in a position to interpret the economic, social, and political signifi-
cance as indicated in Table 10 in Kickert (1977). A cost-benefit evaluation
of alternative air pollution control strategies in the South Coast Air
Basin might possibly be made in agreement with Westman's (1977) warnings
by the combined use of this forest simulation package with parts of the
MATHAIR (MATHTECH, 1976) computer model and an appropriate meso-scale
meteorological transport model. The MATHAIR model assumption for the local
source-local impact is not valid for the Los Angeles Basin - SBNF geography.
Documentation is produced for each submodel consisting of a
word model description of relationships, graphic flow chart, mathematics,
computer program source code, results of reliability evaluation, and results
of experiments performed using the forest simulation package.
Dependence of Simulation Modeling on Data Base Management—
Conversion of conceptual models of various subsystems into mathe-
matical algorithms and subsequently into computer programs has been designed
to be dependent as much as possible on the many datasets being acquired
in the various subprojects of the SBNF program. The data base is designed
to serve as documentary evidence for the reality and validity of (1) quan-
titative relations used to construct the simulation models; (2) the behavior
of the set of models when run on the computer; and (3) the kinds of eco-
systemsuccession forecasting experiments conducted by using the set of
simulators. The bold arrows in Figure 1 outline the broad framework of
dependence between the data base and the information requirements for
simulator development. Until recently, the lack of a documented and opera-
tional procedure for managing the SBNF data base delayed progress in the
development of our forest modeling activity.
Discussion of Developments and Achievements During 1976-1977
Progress is described in terms of the sequence of steps being followed
in the simulation modeling process presented above and management of the
computerized SBNF data base which supports the modeling process. These
-------
accomplishments all relate to the first general objective of validating
quantitative relations used to construct the simulation models. It is
necessary to completely meet this first objective before the second and
third can be accomplished. For this reason, we do not yet present results
of actual computer experiments using the forest system simulation models for
various possible future oxidant air pollution trends.
Organization of Computerized Data Base Management Procedures for Air
Pollution Ecological Effects—
Some recent historical perspective is helpful as an aid in presenting
the achievements in managing the SBNF data base.
From January 1974, through December 1976, the Corvallis Environmental
Research Laboratory and the Lawrence Livermore Laboratory (LLL) had an
interagency agreement for the LLL to design and develop a data management
system using data collected in the SBNF program.
In late 1975 and early 1976, it became apparent that such a task
would not be completed by December 1976, and it was anticipated that the
agreement would be renewed for at least an additional year. In August and
September 1976, it was discovered that this extension would not take place.
In January 1977, the Ecosystem Simulation Modeling subproject acquired
the added responsibility of retrieving all of the data files which project
investigators had submitted to the LLL, and of organizing a computerized
data management system useable on the IBM 370/145 at U.C. San Francisco,
The ecological modeling computer work was also being done on this computer.
Under the prior arrangement, data management was done on the computer at
Livermore, but the change to UCSF led to a significant improvement in model
design planning and data management. Both activities could now be done on
the same computer system at UCSF. The transfer of responsibility in January
1977, was an abrupt one due to circumstances beyond our control. Clear,
comprehensive documentation of the status and contents of datasets associ-
ated with the various subprojects was unavailable. These conditions,
together with the fact that development of a management procedure for
the data base inherited without a budget between January and June 1977, was
behind schedule with respect to modeling needs, required us to relax empha-
sis on model development in order to organize the data base problem.
Dataset verification and auditing— Although written for a corporate
business environment, we applied the philosophy of Wilkinson (1977) in
beginning a data processing audit, both through-the-computer, and around-
the-computer, for the correction of data sets which we had acquired.
We inherited virtually no documentation on the nature of any verification
which might have previously been done on individual datasets or between
datasets having common cross-referencing data elements, such as tree
tag numbers, or species codes.
One objective of the audit was to discover whether datasets which the
field investigators assumed we had acquired were missing. Another objective
was to uncover discrepancies in data elements between different years for a
given dataset, and between datasets where the same data elements were used
-------
in each. The intention was to assist field investigators in revealing any
errors which reside in the datasets.
The SBNF data base structure: the data dictionary— The general
organizational structure of the data base is shoxm in Figure 2. A signifi-
cant achievement has been the establishment of an on-line data dictionary.
Schussel (1977) describes this as "a repository of information about the
definition, structure, and usage of data. It does not contain the actual
data itself" (sic).
We have structured this into a data set index (Table 1), dataset
definitions (Appendix 1), and dataset descriptors (example in Figure 3).
A computer terminal user can simply log-on, and proceed through these
levels of increasing detail of information in search of specific kinds of
datasets on ecological effects under air pollution in the forst. This can
be done by using the procedure described under the discussion of Central-
ized Data Base Approach which appears later in this report.
The dataset progress status chart— There is a sequence of distinct
stages through which datasets advance, from the time that the decision
is made by a field investigator to collect a certain kind of data, until
that time when a written report is produced which contains the description
and results of analysis of the dataset. At any given time during this
research program, various datasets progress at various rates through all of
these stages. Merely tracking down the descriptive information on a dataset
by using the data dictionary does not inform one as to whether or not the
dataset has reached a stage where the data are presently analyzeable on the
computer. In order to assess the status of any of the datasets in the SBNF
data base (Figure 2) at a given instant, an on-line Dataset Progress Status
file has been established and is updated on a weekly basis. A listing of
this file, as of August 31, 1977, is presented in Appendix 2. This feature
of the SBNF data base enables us to track the status of a dataset through
the various stages of preparation, from left to right in Appendix 2, so that
analysis can then be performed, using that dataset. It also allows us to
see where we are in terms of stages of a dataset's analysis for systems
modeling, for data-sharing among investigators in the project, and subse-
quently, for external requests for data. Datsets whose entries extend
to the right of the vertical bold line in Appendix 2 are ready for, or
are presently under analysis. Those that do not are still in a stage of
data preparation. The information categories "NEW DATA" and "VERIFICATION"
pertain to datasets which have not yet been verified by the original
investigator. "FORMAT APPROVED" only pertains to new types of data
collection efforts as they may arise. This stage is intended to call
to the attention of the investigator that some aspect of his data form
format will induce subsequent delays in data processing^ If its being
altered presents no problem for the logistics of field data collection,
he is advised as to what change to make for his benefit later in the data
processing stage.
The category "DESCRIPTOR ENTERED?" shows a record of whether that
document has been placed in the data dictionary. Appendix 2 shows that a
number of datasets have not been covered in this way since we took on this
-------
.DATASET INDEX
DATA DICTIONARY
DATASET DEFINITIONS
00
SBNF
DATA BASE
DATASET PROGRESS
STATUS
DATASETS
•DATASET DESCRIPTORS
DATASET DESCRIPTORS
DATA
Figure 2. Information structure for the computerized SBNF data base accessible from remote
dial-up terminal via UCSF/CMS time-sharing mode.
-------
TABLE 1. PART I OF THE SBNF DATA DICTIONARY: THE DATASET INDEX.
age class
air pollutant
air temperature
basal area, tree
cations, soil
cones
cover
crown data, tree
density, tree
diameter, breast height
disease, tree
elevation
fire effects
foliage
geographic coordinates
height, tree
height growth, tree
index
insect risk
insects
litter, needle
location, tree
moisture, soil
mortality, tree
needle leaf, condition
needle leaf, length
needle leaf, retention
net radiation
nutrients
organic matter content,
soil
oxidant, ambient con-
centration
ozone, ambient con-
centrtion
pH, soil
plot, vegetation-,
logistics
plot, super-
plot, sapling-
STAGE, FIRESTAG, PLOTREGN, CTREE, STNDSITE
OXIDINDX, OXIDANT, PLOTOXID
FSMTINDX, FSMET, HMET, PLOTMET
TREEVEG, SPRMORT1, SPRMORT2, SPRMORX,
STNDSITE
SXSCAT
CONE, GCONE
SHRUBVEG, PLOTREGN, PLITR, STNDSITE
CTREE, SPRMORT1, SPRMORT2, SPRMORX
TREEPEST
TREEVEG, FIRETREE, SPRMORT1, SPRMORT2
TREE, SPRMORT1, SPRMORT2, TREEPEST
DISU, FASP, SPRMORT1, SPRMORT2, STNDSITE,
TREEPEST
PLOTINDX, STNDSITE
FIRETREE, FIRESTAG, FIRESRUB
STOMRES, OZFLUX, TREE, CTREE, SAPTREE,
SAPSURF, FLDECOMP
PLOTINDX
CTREE, SPRMORT1, SPRMORT2, STNDSITE, TREEPEST
SAPGRO
PLOTINDX, FSMTINDX, OXIDINDX
TREE, SAPTREE
ISURV, BTREE, EGG, REAR, STIK, XRAY, SPRMORT1,
SPRMORT2, SPRMORX, STNDSITE, TREEPEST
PLTR, PNFALL, TREELIT, LITMAS, FLDECOMP,
LITRKEM, SLSS
TRID
MOIST, MATRIC, LSOIL
TREEMORT, TREE, SPRMORT1, SPRMORT2, SPRMORX,
STNDSITE, TREEPEST
TREE, SAPTREE
TREE, SAPTREE
TREE, SAPTREE
FSMET
SXSCAT, SFCSOLKM, LITRKEM, DRIP
STEXOM
OXIDINDX, OXIDANT, PLOTOXID
OXIDINDX, OXIDANT, PLOTOXID, OZFLUX
STEXOM
PLOTINDX
SPRMORT1, SPRMORT2, SPRMORX, STNDSITE,
TREEPEST
SAPTREE, SAPGRO, SAPSURF
-------
TABLE 1. (CONTINUED)
precipitation;
publications
radial growth, tree stem
regeneration, tree
relative humidity, air
seedling
seeds; see regeneration
shrub
site
slope
smog injury score
soil
species composition
species, tree
succession
texture, soil
tree tag number
wind direction
wind speed
HPREC, PLOTPREC, DRIP
EPA/SBNF contracts & grants; SBNFPUBS
TREEGRO, TREEGR02, BOGRO, SPRMORX,
TREEPEST
PLOTREGN, STAGE, CONE, GCONE, PLOTSEED,
SAPS, LSOIL, SSAS, SLSS, SPRXRGNS,
SPRXRGNC
FSMTINDX, FSMET, HMET, PLOTMET
PLOTREGN, SAPS, STAGE
SHRUB, SHRUBVEG, FIRESRUB, STNDSITE
PLOTINDX
TREESOIL, STNDSITE
TREE, SAPTREE, SPRMORT1, SPRMORT2, SPRMORX
SXSCAT, STEXOM, MOIST, MATRIC, TREESOIL,
SFCSOLKM, LSOIL, STNDSITE
TREEVEG
TRID
STAGE, PLOTREGN, FIRESTAG, STNDSITE
STEXOM, STNDSITE
PLOTINDX, TRID, TREE, CTREE, CONE, TREEGR02,
BOGRO, DISU, ISURV, TREEMORT, TREESOIL,
PNFALL, TREELIT, LITMAS, FLDECOMP, DRIP,
LITRKEM, SFCSOLKM, SAPTREE, SAPGRO, SAPSURF
FSMTINDX, FSMET, HMET
FSMTINDX, FSMET
10
-------
DATASET NAME: SAPGRO (PINE SAPLING ANNUAL HEIGHT GROWTH)
INVESTIGATOR; PAUL R. MILLER, U.C. RIVERSIDE, (7 14 ) - 787-366 1
STATUS OF EXTERNAL AVAILABILITY: CLOSED
DATASET DESCRIPTOR AUTHOR/DATE: R.N. KICKERT, FORESTRY, U.C. BERKELEY; 8/5/77
DATA SITES: BL (JEFFREY PINE), BP (PONDEROSA PINE), CA (PONDEROSA PINE),
CAO, CP, DWA, HB, HV, SF, TUN2;
DATA RECORD SEQUENCE: ONE RECORD PER TREE; RECORDS GROUPED BY PLOT;
DATA ELEMENT SEQUENCE: [X] POSITIONAL, [~~] FREE-FIELD, OR [""] KEY-IDENTIFIER?
Li [ ] [J
RECORD FORMAT DESCRIPTION: (VARIABLE NAME7 COLUMN NUMBERS FOR VARIABLES' FIELDS,
PHYSICAL UNITS/ IF ANY/, ESTIMATED OBSERVATIONAL
ERROR TOLERANCE)
1. PLOT INDENTIFIER, COL 1, A(4);
2. TREE TAG NUMBER, COL 5, F(4);
3. ANNUAL INTERNODE LENGTH GROWTH, COL 9, 18(IX,F3,0), MILLIMETERS, +/- 30 MM;
DATA COLLECTION DATES: 1976;
NOTES, QUALIFICATIONS, LIMITATIONS (BY VARIABLE NAME):
2. INFORMATION ON SPECIES, EXACT LOCATIONS, 1975 TOTAL TREE HEIGHTS, STEM
DIAMETERS, AND ANNUAL CROWN CONDITION, FOR EACH TAGGED SAPLING CAN BE
FOUND IN THE DATASET NAMED 'SAPTREE';
3. FROM LEFT TO RIGHT, ALONG THE RECORD, A MINIMUM OF 10 INTERNODE LENGTH
VALUES CORRESPOND TO THE YEARS
1976, 19-75, 1974,... 1967. ADDITIONAL VALUES, UP TO A TOTAL OF 18 VALUES,
MAY BE FOUND IN A SINGLE LOGICAL RECORD, WHERE THE 18TH VALUE CORRESPONDS
TO THE YEAR 1959. THE VARIABLE RECORD LENGTHS PRIOR TO 1967 ARE A RESULT
OF THE INVESTIGATORS GOING BACK ONLY AS FAR AS THEY FELT THEY COULD
ACCURATELY DETERMINE ANNUAL INTERNODE INCREMENTS.
Figure 3. Example of Part III of the SBNF Data Dictionary: Dataset Descriptors.
-------
activity. If the category "KEYPUNCHED?" contains an "N" for a dataset, then
we have not been given a card deck for that dataset by the respective
investigator. The category "REDUCTION PLAN DEFINED?" refers to whether a
specific detailed plan for quantitative analysis of the dataset has been
defined for the purpose of identifying certain transfer functions in one of
the subsystem models, or for the purpose of evaluating the reliability of a
part of the modeling package.
The datasets— The third portion of the SBNF Data Base structure
(Figure 2) is the collection of datasets. Each one is a file stored on
magnetic disk, with a back-up copy on magnetic tape, and has a copy of
its descriptor located at the beginning of the file-
Centralized data base approach— The orientation used in maintaining
the SBNF data base has been a centralized approach in that all datasets are
kept collectively on only one computer system, the IBM 370/145 at the
University of California, San Francisco (Medical Center). Within the
Ecological Modeling and Data Management activity, the data manipulation
environments are diagrammed in Figure 4. Datasets are read in from card
decks to magnetic disk under the OS (Operating System) environment and
transferred to magnetic tape as backup. The datasets are transferred from
OS disk to mini-disk in the Conversational Monitor System (CMS) environment
in preparation for work. CMS is a general purpose time-sharing system
operating under VM/370. In addition to the datasets which are under
immediate use, other units of the SBNF data base such as the Data Dictionary
and Progress Status are maintained on CMS mini-disk for immediate tele-
phone access. These relationships are diagrammed in Figure 4. In keeping
with the dependence discussed earlier (Figure 1) of ecological systems
modeling on analysis of datasets in the SBNF Data Base, model development is
conducted in the same CMS environment, including storage of the computer
programs designed to simulate various subsystems. To dial up and interact
with the datasets in CMS, we use a DataMedia model 1520A video screen, key
board terminal, a Diablo 1620 HyTerm printer terminal, and an Execuport
320 portable thermal printer terminal.
Readily available CMS commands can be used to manipulate dataset files
and the contents within files. Rapid manipulation of data between datasets
can easily be done byusing the CMS commands shown in Table 2- Just about
any kind of searches desired can be made on data within a given dataset with
commands as shown in Table 3 which are immediately available at the terminal
in the TECO and CANDE interactive time-sharing environments on the DEC
PDP-10 and Burroughs computers respectively. This means that any eventual
use of the SBNF data base on other main frame computers elsewhere, perhaps
by other environmental scientists, should be just as useable as our capabil-
ity on the UCSF IBM computer.
Aside from using the CMS commands for retrieved and displayed various
kinds of data, summarization and analysis of data are done by entering
either the SPEAKEASY mode or SPSS mode in the on-line environment, or by
submitting a batch job, via remote job entry, in OS to use the BMDP statis-
tical programs (Figure 4)• SPEAKEASY is a simple interactive data manipula-
tion language containing an immense number of built—in functions for
12
-------
CMS INTERACTIVE
OS BATCH
St or age for
immed i at e
LINE
PRINTER
Evans HaJJ.
St or age for
frequent
use
Stor age for
infrequent use
& back
Corval1i s
Environment al
Research
Labor at ory
••^•^^MfeUC-A^^^VBV -^^*^H^^^^M«lmHI^M«HV^^^B
Ecological Modeling Subproject
U.C. Berkeley
Statewide Air Pollution
Research Center
U.C. Riverside
Figure 4. Data processing environments used for simulation
modeling and data base manipulation on IBM 370/145
computer at University of California, San Francisco,
via 30 character-per-second telecommunications.
13
-------
TABLE 2. CMS TIME-SHARING COMMANDS USED FOR ON-LINE FILE MANAGEMENT
OF SBNF DATA BASE.
Command
EDIT
READCARD
COPYFILE
RENAME
TYPE
PRINT
COMPARE
LISTFILE
ERASE
Function
construct a new file by inputting through the
terminal, or change, or examine data within, an
existing dataset file (see TABLE 3);
construct a new file by reading a card deck;
combine several files into one file; rearrange
the contents of records in a file; add one file
to the end of another;
change the name of a file;
type the contents of a file on the printer at
Evans Hall on the UCB campus;
type the contents of a file on the printer at
Evans Hall on the UCB campus;
compare all or part of the records in two files
and type the records that are not identical;
list information about the files which are stored
on disk;
delete the specified file from disk.
14
-------
TABLE 3. CMS TIME-SHARING EDIT-ENVIRONMENT COMMANDS USED FOR FINDING,
CORRECTING, AND DISPLAYING THE CONTENTS OF A DATASET OR PROGRAM
FILE IN THE SBNF DATA BASE.
Command
INPUT
Function
creates new lines typed into the file by the user
at the terminal
LOCATE
TOP
locates the next line in the file that contains a
specified character string (and types the line at
the terminal)+
moves the line pointer back a specified number of
lines in the file
UP
moves the line pointer back a specified numbr of
lines in the file
BOTTOM
moves the line pointer to the position following
the last line in the file
DOWN
moves the line pointer forward a specified number
of lines in the file
NEXT
CHANGE
REPLACE
DELETE
moves the line pointer forward one line (and types
the line at the terminal)+
changes a specified character string in the line
to a new character string; can be used to search
for a specified character string anywhere in the
file and then type out the line in which the
string is found
changs the current line content according to the
terminal user's request
beginning at current line, erases the specified
number of lines from the file
TYPE
beginning at current line, types out the contents
of the specified number of following lines in the
file
FILE
terminates the current editing session for the
file and stores the file on disk
+ assuming VERIFY command is ON
15
-------
performing mathematical and statistical analyses on data arrays and vectors,
for manipulating arrays in various ways, and for graphing data. SPSS is the
Statistical Package for the Social Sciences which is documented in Nie et
al. (1975). BMDP is the set of statistical analysis programs documented in
Dixon (1975).
While our base of operations is at U.C. Berkeley, we have easily
accessed, with no difficulties, the various units of the SBNF Data Base
just described, and some of the simulation programs, while at the Statewide
Air Pollution Research Center, U.C. Riverside, and the Corvallis Environmen-
tal Research Laboratory, Corvallis, Oregon. The potential exists for any
investigator in the SBNF program to directly interact with the UCSF computer
with those datasets that have been verified. The extent to which this
happens from now on depends upon the desires and motivations of the investi-
gators.
Distributed processing approach— While a centralized approach to
maintaining the SBNF data base has been employed, the approach to processing
of datasets has evolved in a distributed manner. Up to the present, several
investigators in the project have only operated on their own datasets
within their own data processing environments. This is evident in our
assessment, as shown in Table 4, of the amount of data processing which has
been done by investigators on computer systems other than the one used to
maintain the centralized SBNF data base since January 1977, and the distri-
bution of data-related requested from the investigators to this subproject.
Project investigators haven't reached the stage of conducting integrated
data analysis of their own datasets with those assembled by their colleagues.
This is probably because not enough years of data had accumulated prior to
this time and also because the entire data base was not in a readily access-
ible computer environment. In addition, it is natural that we will be doing
much of this for transfer function identification in system model develop-
ment. As the project approaches a stage of synthesis in the next few years,
a decision may be adviseable from the project investigators as to whether it
is in their best interest, from the viewpoint of trans-disciplinary data
analysis, for them to continue solely with a distributed data processing
approach.
Specific data processing tasks achieved— In 1*973, the Soils subproject
placed soil moisture sensors at various depths on 22 sites in the 18 vegeta-
tion plots. These have been interrogated at weekly, or biweekly, intervals
since that time by personnel out of U.C. Riverside. The Soils subproject
also took field soil samples to the laboratory to develope calibration data
so that the field data on electrical current passing through the moisture
sensors could be converted to log resistance values and then to percent soil
water values. Data processing to accomplish these steps was expected to be
finished under the previous EPA/LLL agreement during January 1974, through
December 1976, as discussed earlier in this report. The fact that this did
not materialize precluded the avilability of percent soil water data and
essentially halted progress in further development of our forest stand
moisture simulation model. In the first half of 1977, we tackled this
deliquent data processing task and by August 1977, the establishment of the
MOIST (reduced) dataset was 95 percent accomplished.
16
-------
TABLE 4. STATUS OF DISTRIBUTED VERSUS CENTRALIZED DATA PROCESSING OF
SBNF DATA BASE FOR VARIOUS SUBPROJECTS AS OF AUGUST 1977.
Submodel
WATER
CANOPY
TREEGROW
ROOTS
BEELTE
mortality
LITTER
LITDECAY
CONE
SEEDLING
STNDCOMP
Relevant
investigator
Arkley
Miller
Laven
Cobb
Dahlsten
superplot
Arkley /Miller
Bruhn
Luck
Cobb
McBride
McBride
Independent
data processing
little
much
little
some
much
little(?)
some
some
some
much
much
much
Dependence on
data management
subproject
much
little
much
some
some
much ( ? )
much
much
some
little
little
little
17
-------
Another task achieved was the conversion of raw field data on plot
tree ring widths (TREEGRO dataset) obtained as far back as 1920, to a form
which is now useable for analysis in developing the tree STEM growth sub-
system simulator.
Accomplishments in auditing datasets—In order to prepare the Data
Dictionary and Data Progress Status Chart, our audit of the datasets led to
results which in some cases revealed further data needs for modeling, and in
other cases led to improvements in the consistency of information within and
between datasets.
By auditing the data elements in SPRMORT1, SPRMORT2, SPRMORX, STNDSITE,
and TREEPEST, we discovered that data were not being obtained that would
enable calculation of the stand tree mortality as a percentage of the
total stand stocking density, by species. We would have the estimates of
numbers of recently killed trees, by apparent causal agents and by species,
but we would not be able to relate this to the population size of the stand
at the superplot spatial scale. This deficiency would hinder the evaluation
of reliability of the systems models.
The audit of data elements also revealed that we were not going to
obtain the kind of dead tree data necessary to tell the computer how to kill
a tree in a biologically reasonable way during simulation. Subsequently,
the data elements shown in the SPRMORX dataset (see Appendix 1) were defined
for field data collection. These include, for each mortality center dead/
damaged tree, recent radial growth increments, and height to lowest branch
bearing needles.
Additional supplementary datasets became identified as needing to
be established in the SBNF data base. These include SBNFPUBS, PLOTINDX,
FSMTINDX, OXIDINDX, and PLOTSEED. A description of each of these is found
in APPENDIX 1.
Several cases resulted in improvement in consistency of dataset con-
tents. Since several different datasets contained species identification and
tree tag numbers, we ran a program to compare for taxonomic agreement across
all the datasets, tree-by-tree, for all vegetation plots, and had a list
printed of any and all disagreements. The need to verify agreement between
datasets on this data element was fundamental in order to proceed with
any other tree-related data analysis. When the task was done, it became
evident that several datasets from various years and/or different invest-
igators contained discrepancies as to the taxonomy to be associated with
a given tree number. Two major reasons for these discrepancies seemed to be
the degree of hybridization which occurs in some areas between ponderosa
pine, Jeffrey pine, and Coulter pine, and the problem of mis-reading the tag
number of a tree when making and recording observations. Several project
investigators subsequently used these lists in the field to recheck specific
trees and plots, which led to improved consistency of data in TRID, CTREE,
TREE, DISU, ISURV, and TREEGRO datasets.
For both years, 1974, and 1976, of the Disease Survey (DISU) dataset,
a comparative listing was made by the computer, tree-by-tree, for any
18
-------
differences in the data entries for a given tree between the two surveys.
The results were returned to the plant pathology subproject for resolution
of inconsistencies.
Other internal examinations of TRID and TREE led to enhancements in
their contents.
Summary of data base status— We have used the Dataset Progress Check-
list (Appendix 2) to assess the present overall state of the project
data base. Table 5 contains a summary of the findings. The proportion
of the data base in various stages indicates that we are ready to concen-
trate more on cross—disciplinary data analysis for modeling in the next
year than has been possible since data collection began four years ago.
However, we urge the subproject leaders to make sure they prepare and submit
verified card decks as soon as possible following data collection, so that
joint analysis between their subprojects and the Ecosystem Modeling subpro-
ject can be done with a minimum of delay.
Developments in Ecological Subsystem Modeling—
A previous report (Kickert, 1977) highlighted the interrelational
structure between various submodels being developed. A population dynamics
accounting (STNDCMP) for trees in the forest stand is driven by submodels
dealing with tree regeneration (SEED, SEEDLING, LITTER, LITDECAY) and
stand mortality (ROOTS, BEETLE). Each of these model subsets is driven by a
stand moisture subystem (WATER), as well as by external inputs of air
quality monitoring data (OXIDANT dataset and PLOTOXID dataset).
In Figure 1, the first important link for use of the project data
base in the modeling activities is evaluating the quantitative nature of the
relations which have been hypothesized in the flow chart of the various
ecological subsystems (Kickert, 1977). A computer subroutine is being
written to simulate each subsystem is in the process. The sequence
in which each of the various subroutines will be activated on the computer,
for passing information from one simulated subsystem to another, is shown
from left to right in Figure 5. Prior to the computer terminal user telling
the simulators to begin running, the user is first given a series of options
for running the simulation. These options are for setting numerical values
for: starting year and ending year; site and tree species parameters
which will not change during the simulation; initial forest stand condi-
tions; the nature of long-term trends in meteorological conditions which the
user wants to drive the simulation; and the format of the output display to
be used. All of these options have default choices built-in, with the
additional option of being able to display the default values at the term-
inal, so the user may avoid making decisions to override these if so desired.
Figure 5 also shows which of the datasets, listed in Table 1 and Appendix 1,
are being, or will be, used to quantify relations in each of the subsystem
models, and which will be used to provide the external physical environ-
mental data to drive the set of simulators.
Computer simulation programs have been written for WATER, TREEGROW,
and partially for BEETLE. Details of the CANOPY submodel structure have
19
-------
TABLE 5. DISTRIBUTION OF DATASETS BY DEVELOPMENT STAGE FOR THE SBNF DATA
BASE AT UC BERKELEY.
Number of Percentage of
Development stage+ datasets data base
1A. Descriptor not yet
written for Data 49 79
Dictionary-H-
1B. Dataset not yet key-
punched (or converted)// 30 48
2. Dataset card deck entered
but not yet verified 16 26
3. Dataset verified and
corrected on disk## 10 16
4. Dataset presently in
analysis for modeling 6 10
within the Ecological
Modeling - Data Mgmt
subproject
+ does not pertain to updates for post-January 1977, for prior
existing datasets, but does include new datasets through August 1977;
-H- this stage is not mutually exclusive as the subsequent stages
are, so it should not be compared against them;
# raw data converted to a new form, or transferred from the U.S. Forest
Service to the SBNF data base;
## some datasets counted in this stage may be in the next stage at
UC Riverside.
20
-------
Simulation-User s Choice
Simulation Sequence
Response
N3
1 scare i
/Over A.
/ ride \
(defaults?
( Starting/Ending years)
— ta> r i 1 1 1 1
FI
1 L1
,cay|
.DECOMP, DRIP
LTKKEM, St'CSOLKM |
1 . _ Seedlinq 1
Litter 1 'ii •
1 0 I I
PLITR I
PNFALL
TREELtT
LITMAS
I
CTREE, CONE, GCONE
PLOTSEED
I-»J Water 1 _ Canopy 1 ». Treegrow 1 1
'••••U 3 1 4 j
/'parameters forA MQ
Vsite & speciesy Mi
STEXOM, SXSCAT
(initial Forest Standj
STAGE
(Driving data generators\
OXIDANT, PLOTOXID ^\
HPREC, PLOTPREC
HMET, FSMET, PLOTMET
(Output display format)
L —
1ST OZFLUX
TRIG STOMRES
INJBIWK
\
ha
^X^/ 0
TRID ' TREEGRO:
SAPS
LSOIL
SSAS
SLSS
PLOTREGN
»- r"
TREE BOGRO . , . |
SAPTREE SAPGRO mortality!
floot-lo I
BTREE, EGG,
REAR, STIK,
| XRAY
rvest ^ Roots I 1
3 1
•i tit fi
DISU
ISURV
TREEMORT
SPRMORT1
SPRMORT2
SPRMORX _ ..
1 FASP TREEPEST /!
,er \ ^~
Venter rupt^/
Stand
moosition
12
STAGE
TREEVEG
FIRESTAG
FIRETREE
'rimary \
)utput 1
)ispjay I
Figure 5. Simulation sequence between subsystems (numbered rectangles), showing
associated datasets (capitalized) for submodel quantification and
validation.
-------
been worked out, and the next step will be to convert it into program code.
ROOTS, LITDECAY, LITTER, SEED, and SEEDLING subsystems should have the first
versions of the simulation programs written in the next six months.
The stand moisture subsystem simulator— The WATER submodel is used
to take precipitation and air temperature data, and simulate biweekly
changes in soil water in the tree root soil depth. This is then used as
input to the CANOPY subsystem, along with ambient oxidant air pollution
data. The WATER submodel was a modification of that reported by Sollins et
al. (1976) from the Coniferous Forest Biome. We ran our version on the
computer and found it to behave in an intuitively reasonable fashion.
Further work in adapting this submodel to the southern California forest
sites had to be postponed because of unavailability of moisture data and the
entire data base problem encountered during the past year. Because of the
data base achievements described earlier in this report, we are now in the
position to resume development and application of this subsystem model to
our needs and goals. A new detailed flow chart was constructed for the logic
of this submodel. The one supplied by the authors of the original model
(Sollins et al., 1974) was too general to clearly portray the model's
relations in the midst of critical review sessions with the soils subproject
investigator.
The tree stem growth and canopy subsystems simulators— A second
example of our effort to build on the work of our colleagues elsewhere is
found in our subsystem simulators for tree canopy changes and tree stem
growth. We analyzed the structure of a forest succession simulator, called
SUCSIM, developed in the Coniferous Forest Biome. Reed (1976) has described
the theoretical ecology foundation for the model, and has documented the
mathematical details of the various functions (Reed and Clark, 1976). This
model was developed as an adaptation of the basic tree growth theory earlier
developed in the easterb United States for deciduous forest systems simula-
tion by Botkin et al. (1972). In contrast to the many mensurational forest
growth models based on site-index, Reed's simulator has trees grow in
response to physiological definition of tree species niches, with regard to
availability of needed environmental resources for growth, light, heat,
moisture and nutrients. Because of this particular theoretical base, this
model appears more suitable for restructuring for the study of growth
effects of environmental pollutants than do other tree growth models.
The structure of the Reed model can be viewed as consisting of two
modules relating to individual tree growth, a crown module, and a stem
module. These feed into three modules at the tree population or stand
level, a population regeneration module, a population mortality module, and
a population update module. The primary variables which describe the state
of the system in this model at any time are basal diameter of individual
trees, height of individual trees, total leaf biomass of individual trees,
age and species of individual trees, diameter at breast height of individual
trees, and number of live trees in the population. The only state-variable,
and respective mathematical functions, in this list for which we presently
have no data in the SBNF data base are those related to leaf biomass.
We have assembled the computer code for Reed's stem module as our
22
-------
TREEGROW subsystem, and are restructuring his crown module to respond to air
pollution injury so we can use the modified version as part of our CANOPY
subsystem. We are revising his simplified population regeneration and
mortality modules since the SBNF program was designed to investigate and
model these processes at a higher resolution (SEED, SEEDLING, LITTER,
LITDECAY, and ROOTS, BEETLE). What we have identified as his population
update module is similar to our STNDCMP subroutine. This is the master
calling procedure that keeps account of how many trees of various ages and
species exist at any given time during the simulation.
Quantification of site and species parameters and transfer functions—
As indicated under the previous discussion on the progress status of data-
sets, we have now organized the data base sufficiently to move more into
quantitatively analysis of transfer functions for the various submodels.
Our audit of the datasets and simulataneous conceptual development of
submodels has led to the discovery that certain kinds of analyses are
possible and data collection techniques are available that could improve the
quantification process for various subsystem models.
A design of foliar biomass analysis for trees under varying degrees of
oxidant air pollutant injury is suggested to improve useability of the tree
stem growth submodel. Functions in the original stem growth simulator now
used were constructed for essentially pollution-free forests. This analysis
is necessary to couple the oxidant-foliar injury submodel (CANOPY) with the
tree growth submodel (TREEGROW).
Development of the oxidant-uptake and crown foliar injury submodel
is designed on the hypothesis that changes in transpiration affect the
amount of oxidant taken into leaf tissue and the subsequent visible injury
found over long periods. We suggest that the appropriate investigator
strongly consider an experimental treatment on pine foliage to demonstrate
the degree of validity of this hypothesis. This suggestion involves repeat-
edly treating samples of new foliage, during Spring-time, at a known heavily
polluted site, to applications of an antitransparent, and then following
the seasonal changes in the various foliar injury symptoms typically assoc-
iated with ambient oxidants.
To apply the stand moisture simulator, WATER, to forest plots in the
SBNF, so that the hypothesis on transpiration control of oxidant uptake may
be simulted, we suggest that the soils subproject strongly consider perform-
ing a hammar seismograph analysis to determine the maximum soil depth on
those vegetation plots where this information is not already available as a
result of the past installation of soil moisture sensor profiles.
For a part of the organic LITTER dynamics submodel which responds
to the oxidant uptake crown injury subsystem, we suggest that the appropri-
ate investigator consider getting data on: 1) the distribution of various
areal densities of coarse woody litter on the ground on vegetation plots;
and 2) the time rate of change of the amount of woody matter which falls
from various tree species after individuals are killed by a pest complex.
Design of approach for evaluation of reliability of the simulation
23
-------
package—With regard to the simulation models developed in this project,
the project officer has advised that the stage at which the Corvallis
Environmental Research Laboratory desires to receive information is with
the documentation of the reliability of the models as compared to actual
real forest response. Reliability of the models can be examined using a
method of successive approximation. An easier, but less objective, method is
to obtain "the reactions of experienced field observers to the predictions"
of the simulation package (Botkin et al. 1972). A more objective, but more
difficult, method is to run the simulation with historical input data and to
compare the computer model output behavior with historical data on the same
output variables.
In some cases, there appear to be adequate types of data for evaluating
the reliability of individual submodels, such as WATER and CANOPY, over
the relatively short-term study period 1973 through present. However, we
are convinced that the reliability of the entire linked set of subsystem
simulators, producing simulated annual output on numbers of trees by species
by age, can only be evaluated over a long time span of data. If we knew what
the tree species composition on the vegetation plots were in 1920 and
1950, we could initialize the simulation package for 1920, and using histor-
ical weather data from HPREC and HMET datasets (Appendix 1), run the simula-
tors to 1950, covering the recent pre-air pollution era. We could then
compare simulated forest tree species composition at "1950" on the computer
with the actual 1950 known composition. By using quantitative techniques
developed by Miller for reconstructing the increasing oxidant air pollution
trend from 1950 to 1973, and using monitored data for 1973 through present,
we could run the simulators for the air pollution era and compare simulated
forest species composition at "1973" through "present" on the computer with
the known field composition data for the same years. For the latter, the
datasets STAGE and TREEVEG, collected in 1973, are in the SBNF data base.
Assuming these reliability evaluations showed that the simulators tracked
the field data closely enough for 1950 and the mid-1970's to preclude the
decision that the models' behavior were unreliable, then we could begin
using the simulation package to perform "what if..." experiments (Figure 1
herein, and 26 in Kickert (1977)) on the computer for the future period 1980
through 2025. We are using the philosophy that one cannot directly prove
that a simulation model is relizable; one can only fail to show that it is
unreliable, the initial a priori assumption, after repeated attempts to
discover the unreliable behavior.
All investigators in the SBNF project must be aware that the ultimate
usefulness and acceptability of all of the subsystem simulators and accept-
ability of all of the subsystem simulators to outside users, including
the Corvallis Environmental Research Laboratory, pivots on our ability
to evaluate the reliability of the simulators produced in this program.
Inability to evaluate reliability could lead to future potential users
regarding these ecological system simulators as simply academic exercises.
At present, the research design does not include obtaining the right
kind of numerical data on the SBNF which would allow for evaluating the
reliability of the simulation models over the time span 1920 through 1950
(the clean air era), and into 1950 through 1970/80 (the present air
24
-------
pollution era).
In order to evaluate reliability of the simulation package for prospec-
tive users, we urge the vegetation investigators to consider collect-
ing data, for the 18 permanent vegetation plots, on tree stump and snag
locations, species, diameter, year of death or cutting, and age when death
or cutting occurred. These data are needed for synthesis with the stand age
data (STAGE in Appendix 1) collected in 1973, in order that reliability
analysis can be performed as objectively as possible.
CONCLUSIONS
The project officer has advised that the stage at which the Corvallis
Environmental Research Laboratory desire to receive information on simula-
tion models developed in this project is with the documentation of reliabil-
ity of the models as compared to actual real forest responses. Reliability
of the models can be examined using a method of successive approximation.
The easiest, but least objective, method is to obtain "the reactions
of experienced field observers to the predictions" of the simulation pack-
age. A more objective, but more difficult, method is to run the simulation
with historical input data and to compare the computer model output behavior
with historical data on the same output variables. At present, the research
design does not include obtaining the kind of numerical data on the SBNF
which would allow for evaluating reliability of the simulation models
over the time span 1920 through 1950 (the clean air era), and the 1950
through 1970/80 (the present air pollution era). In some cases, there
appear to be adequate types of data for evaluating reliability of individual
submodels.
The proportion of the SBNF Data Base in various developmental stages
indicates that we are ready to concentrate more on cross-disciplinary
data analysis for model development in the next year than has been possible
since data collection began in the SBNF project 4 years ago.
RECOMMENDATIONS
Planning Future Project Activities In Manuscript Preparation
For purposes of planning data processing tasks and associated manu-
script preparation by the project investigators, we suggest that the princi-
pal investigator, in consultation with subproject leaders, define all of the
research activities felt to be necessary, and all of the manuscripts con-
templated, within this program during June 1978 through May 1980.
Reliability Evaluation of-Ecosystem Simulation Submodels
In order to evaluate the reliability of the simulation package for
prospective users, we urge the vegetation investigators make every effort
to collect data, from the 18 permanent vegetations plots to show: tree
stump and snag locations, species, diameter, year of death or cutting, and
age when death or cutting occurred. These data are needed for synthesis
with the stand age data collected in 1973, in order that reliability
25
-------
analysis of the simulation models can be performed as objectively as possi-
ble.
Development of Ecosystem Simulation Submodels
For usability of the tree stem growth submodel, we suggest that
a foliar biomass analysis be designed for trees under varying degrees of
of oxidant air pollutant injury. The functions in the original stem growth
simulator being used were constructed for essentially pollution-free forests
and analysis is necessary to couple the oxidant-foliar injury submodel
with the tree stem growth submodel.
Development of the oxidant-uptake and crown foliar injury submodel
is designed on the hypothesis that changes in transpiration affect the
amount of oxidant taken into leaf tissue and the subsequent visible injury
found over long periods. We suggest that the appropriate investigator
strongly consider an experimental treatment on pine foliage to demonstrate
the degree of validity of this hypothesis. This suggestion involves repeat-
edly treating samples of new foliage, during Spring-time, at a known heavily
polluted site, to application of an antitranspirant, and then following
the seasonal changes in the various foliar injury symptoms typically assoc-
iated with ambient oxidants.
In order to apply the stand moisture simulator to forest plots in
the SBNF, so that the hypothesis on transpiration control of oxidant uptake
may be simulated, we suggest that the soils subproject strongly consider
performing a hammarseismograph analysis to determine the maximum soil
depth on those vegetation plots where this information is not already
available as a result of the past installation of soil moisture sensor
profiles.
As a part of the organic litter dynamics submodel which responds
to the oxidant-uptake crown injury subsystem, we suggest that the appro-
priate investigator consider getting data on 1) distribution of various
areal densities of coarse woody litter on the ground on vegetation plots,
and (2) the time rate of change of the amount of woody matter which falls
from various tree species after individuals are killed by a pest-stress
complex.
For the datasets to be submitted to the Data Management subproject,
we urge subproject leaders to make sure that they prepare and submit veri-
fied card decks as soon as possible following data collection, so that
joint analysis between their subprojects and the Ecosystem Modeling sub-
project can be done with a minimum of delay.
Interested Cooperating Agencies
The funding agency should recognize and act upon the need of the
systems ecologist to have as clear a definition as possible of the ways in
which the funding agency could use the information being sought from the
research project. It is recommended that the project systems ecologist,
agency project officer, and other informed agency personnel obtain a
26
-------
conceptual information flow model describing how the environmental informa-
tion (data base, computer models, and ecological insights) resulting from
this research can be made available. Through a computerized information
transfer, delivery could be made to other environmental scientists, admin-
istrators, legislators, and interested general public, for the purpose of
evaluating secondary standards for photochemical air pollutants, determining
possible consequences of alternative ambient oxidant trends, and identify-
ing alternaive forest management practices.
Research proposals which indicate that a variety of different kinds
of related data collection is planned by more than one investigator, should
be required to show evidence in the proposal that a usable data base
management system and data dictionary processor are already available
and will be used at the time that data begin to accumulate. This should
expedite the rate at which collected data are analyzed.
27
-------
TREE POPULATION DYNAMICS SUBSYSTEM
Introduction
Environment and particularly stress strongly influence successional
change in plant communities and contribute significantly to structure
the composition of these communities. Photochemical oxidant air pollutants
are stress factors which invade the San Bernardino National Forest (SBNF)
and may play an important part in directing successional changes. A gradi-
ent in air pollutant exposure is recognized in the SBNF and differential
susceptibility of plant species to the pollutants has been demonstrated.
Therefore, it can logically be hypothesized that the oxidant air pollutants
may strongly influence successional change and through years or decades
of exposure may be an important factor in determining plant community
structure.
The vegetation subsystem project is focused on describing: 1) plant
communities within the mixed conifer forest type in the San Bernardino
Mountains, and 2) the impact of oxidant air pollutant on successional
changes in these communities. Initial characterizaton of major plant
communities has been reported by McBride (1977). This report summarizes two
studies conducted during 1976-77 of community description and successional
change. One study was designated to identify sub-units (facies) occurring
within major plant communities and the second was aimed at classifica-
tion of forest sites as a first step in the description of plant succession.
Research Objectives
1. To identify and map facies within the plant communities dominated
by yellow pines in the San Bernardino mountains.
2. To classify sites within the Jeffrey pine dominated forests of the
San Bernardino Mountains on the basis of environmental parameters.
Literature Review
Variation in forest composition in the San Bernardino Mountains has
been discussed by Horton (1960), Minnlch et^ a.1 (1969), Miller and McBride
(1973), McBride (1973), and McBride (1977). These authors identified a
variety of forest communities at the association level (Braun-Blanquet,
1932). The classification developed by McBride (1977) established five
associations within the general yellow pine types: ponderosa pine forest,
ponderosa pine-white fir forest, ponderosa-Jeffrey pine forest, Jeffrey pine
forest, Jeffrey pine-white fir forest, but these associations have not
previously been subdivided into facies.
28
-------
Forest succession in the San Bernardino Mountains was discussed
in general terms by Miller and McBride (1973), McBride (1973), and McBride
(1977). Wildfire was determined by these authors to have an important
control over forest regeneration and age structure. Minnich (1974) eval-
uated the role of major fires in initiating secondary succession over large
areas of the San Bernardino Mountains where the yellow pine type was adja-
cent to extensive chaparral areas. As suggested that both the rate and
pattern of recovery of forest tree species following five was variable.
Specific classification of sites on the basis of environmental parameters
has not been used previously to study forest succession.
Materials and Methods
Field Sampling Techniques— •
Eighteen permanent plots established in 1972 and 1973 (McBride, 1977)
were used in the identification and mapping of facies. The facies were
recognized on the basis of species composition, plant height, and cover
in the tree, shrub, and herb layers of the forest. Boundaries were es-
tablished when a change in species composition or cover (more than 25%)
occurred. The basic procedure was to walk the entire plot in order to
survey the variation in tree, shrub, and herb layers before mapping of the
facies. The recognition variable (species composition, plant height, and
cover) were recorded for each facies.
Data collected on 45 Jeffrey pine plots of the 83 temporary plots
established in 1974 to in investigate forest condition as a function of time
since the most recent fire (McBride, 1977) were used in the classification
of sites. No additional field data were collected from these plots in
1976-77.
Laboratory Analysis Procedures—
No laboratory analysis procedures were appplied to the facies identi-
fied and mapped on the 18 permanent plots.
The field data obtained from the Jeffrey pine plots along with data
obtained from published sources (i.e., U.S.G.S. topographic maps) were used
to classify sites with a numerical taxonomic clustering technique described
by Sneath and Sokal (1973). The following parameters were used for the
classification:
1. elevation
2. slope
3 radiation index
4. precipitation
5. soil depth
6. water surplus
7. water-holding index
8. percent clay
9. percent sand
10. soil fraction greater than 2 mm
29
-------
11. A horizon pH, color and chroma
12. C horizon pH, color and chroma
13. slope aspect
14. percent rock cover
15. percent bare ground
16. position of plot on slope
17. length of slope
18. microrelief (i.e., concave, flat, convex)
19- macrorelief (i.e., level, undulating, rolling, hilly, steep)
Based on the values of each of the above parameters similarity coefficients
were derived that compared each site with every other site. (A similarity
coefficient is a numerical representation of the overall similarity between
two sites). Sites that had the highest average similarity values were
grouped together thereby defining clusters containing members (sites)
that possessed a high degree of resemblance. These clusters are represented
in a tree-like diagram (i.e., phenogram) that is a two-dimensional represen-
tation of the interrelation of the study sites (Fig. 6).
Stepwise discriminant analysis was subsequently applied to determine
if any of the sites should be reassigned to other clusters and to evaluate
the relative importance of each parameter used to define these clusters.
Results and Discussion
A total of 189 facies were identifed on the 18 permanent plots. Seven-
teen of these occurred on more than one plot (Table 6). The facies map
(Fig. 6) and descriptive data (Table 7) for the Dogwood plot is presented
as an example of the maps and data prepared during the study- Maps and data
for all 18 permanent plots are available from the Forest Ecology Laboratory,
Department of Forestry and Conservation, University of California. The
facies maps will provide a basis for relating forest regeneration and plant
succession to local variations within the forest types.
The phenogram produced by cluster analysis revealed five fairly distinct
clusters (Fig. 7). As one moves along the horizontal axis the first
visually apparent cluster includes sites 78 through 50. The membership
for the second, third, and fourth clusters are respectively: sites 66
through 55; sites 62 through 59; and sites 71 through 46. The last cluster,
sites 60 through 64, can be viewed as a loose assemblage of sites that are
least like any of the other sites.
Results of discriminant analysis do not reveal any significant changes
in group membership. However, analysis of the territorial map produced by
discriminant analysis (Fig= 8), depicting a cluster summary of discriminant
scores 1 and 2 for each site, reveals with which parameters the clusters
are associated. The horizontal axis represents a moisture complex increas-
ing to the right. Precipitation and water surplus variables provided the
greatest influence in this complex. The veritical axis represents an
exposure complex decreasing upwaves. The greatest influence in this complex
was provided by aspect, radiation index, and macrorelief variables. The
map illustrates how the clusters are evaluated along these environmental
30
-------
DOG WOOD-165m
Figure 6. Facias map of Dogwood Plot (see Table 7 for description of
facies).
31
-------
TABLE 6. NUMBER OF FACIES IDENTIFIED ON THE 18 PERMANENT PLOTS USED TO
MONITOR AIR POLLUTION INJURY TO FOREST TREES.
Plot
Total
Number of Facies
Distinct Common to
to Plot Other Plots
- — -* ' —
Ponderosa pine
Breezy Point
U.C. Conference Grounds
N.W. Camp Paivika
10
4
8
9
4
5
1
0
3
Ponderosa pine-
White fir
Ponderosa-Jeffrey
pine
Jeffrey pine-
White fir
Jeffrey pine
Camp Angeles
Camp 0-ongo
Dogwood
Schneider Creek
Sky Forest
Tunnel Two
Barton Flat
Green Valley Ck
Camp Osceola
Bluff Lake
Heart Bar
N.E. Green Valley
Holcomb Valley
Deerlick
Sand Canyon
13
11
11
15
15
13
8
23
9
15
9
11
13
10
16
8
9
11
14
13
11
8
20
6
9
9
6
6
9
15
5
2
0
1
2
2
0
3
3
6
0
5
7
1
32
-------
TABLE 7. SPECIES COMPOSITION, TREE HEIGHT, AND COVER OF FACIES ON DOGWOOD PLOT,
FACIES ;
NUMBER ;
TREE LAYER
Upper Middle Lower
6-J!
rH H
« CD
•U >
0 O
H U
1 I 0
I
2
I
3
4
5
6
7
8
9
i
5-25
50-75
1-5
0
n
25-5
0
75-100
4->
•a
•rt
1
50+
50+
50+
50+
50+
>
?
en s^s : B^
cu ;
•H M r i-l H
on) ; cd a)
a) > : 4-j >
& o i o o
CO O ! H O
i
t
IC5-25
PP50-75
1C 1-5
PP25-50
1C 1-5
WF 5-25
1C 5-25
SP 5-25
i
0
25-50
,
50-75
0
0
0
50-75
75-100
50-75
i
4-1
•a
•H
3
15-25
25-50
10-25
10-25
25-50
Species
Cover %
1C 5-25
BO 5-25
IC50-75
PP 1-5
PP50-75
1C 5-25
WF 1-5
B075-100
PP 5-25
1C 1-5
PP50-75
PP 1-5
WF 1-5
1C 1-5
6^2
H M
tfl 0)
4-1 >
O O
H 0
0
25-50
0
1-5
0
0
5-25
75-100
i
CO &-S
4-1
0) ft O
Cd ; en u
10-25 WF 1-5
PP 5-25
1C 5-25
6-10 1C 1-5
BO 1-5
i
|
\
i
•j
6-10 PP 5-25
1C 5-25
6-10 PP75-100
1C 5-25
i
25-50 | 10-25 SP 5-25
1
GROUND
SHRUB LAYER LAYER
B-S
rH ^
rt a)
4-> >
0 0
H 0
0
0
1-5
0
0
0
0
0
0
4-»
•a
•H
3
>3
Species
Cover %
B01-5
CO &^
0)
•rl M
O
ft O
en o
BG 50-75
G 25-50
BF 1-5
L 75-100
BF 50-75
G 50-75
L 75-100
BF 1-5
BF75-100
F 1-5
BF 1-5
L 1-5
L 50-75
BG25-50
BF 5-25
G 1-5
BG75-100
L75-100
BF25-50
G 1-5
L75-100
BF 1-5
L75-100
BF 1-5
OJ
OJ
-------
TABLE 7. CONTINUED
FACIES
NUMBER
10
11
TREE LAYER
6^
T-l M
ctf CU
4J >
O O
H U
85-100
5-25
Upper Middle Lower
Height
50+
50+
Species
Cover %
PP75-100
PP 5-25
^s
i-i M
ti
O 0
H 0
0
50-75
4-1
&
M)
•H
CU
W
10-25
Species
Cover %
PP25-50
B025-50
SP 5-25
WF 5-25
&-S
.H h
ed OJ
•U >
0 0
H O
50-75
5-25
jj
,£5
bO
•H
a)
ffi
10-25
6-10
CO B-S
OJ
•H M
O 0)
0) >
a o
C/3 O
WF25-50
PP 5-25
BO 5-25
SP 5-25
1C 1-5
WF 5-25
PP 5-25
SP 5-25
SHRUB LAYER
6-5
i-l h
n) a)
J-> >
0 O
H 0
0
0
w
M
•H
a)
W
Species
Cover %
GROUND
LAYER
Species
Cover %
L75-100
BF50-75
G 1-5
L75-100
BF 5-25
*Height in feet
PP = ponderosa pine; BO = black oak; SP = sugar pine; WF = white fir; 1C = incense cedar;
BG = bare ground; G = grass; BF = bracken fern; L = litter.
-------
.60 r-
.65
.70-
.75
.80
O
z
.85
^
at
I .90
1/1
.95
1.0
f?
OINJ
n
PLOT NUMBER
Figure 7. Phenogram illustrating cluster formation.
35
-------
-5
5.00-
3.75-
2.50-
1.25-
0-
•1.25-
2.50-
3.75-
•5.00-
•°° -2.50 0 2.50 5
11 ii/ii
t
i
3
i
i
<
/ 4
"**• -« ' N
"~x S "^^
^\ 5 ^^,
---'
\ *-~
0 \ *-'*
z v— - ^
1
1
1
1
1
'
1 1 1 I 1 1 1
-3.75 -1.25 1.25 3.75
.00
•5.00
-3.75
-2.50
-1.25
-0
—1.25
--2.50
--3.75
•-5.00
Figure 8. Territorial map of discriminant score 1 (hortizontal) vs.
discriminant score 2 (vertical). (Number indicates group
centroid of respective clusters.)
36
-------
complexes. Each of the four clusters are located in a different quadrant
indicating that these two environmental complexes successfully distinguish
the clusters. The fifth cluster, composed of anomalous sites according to
phenogram structure, is located in the center of the map thereby substanti-
ating its status.
Cluster #1 (located in the lower right quandrant of the territorial
map) is exposed to an evnironment that is associated with a high moisture
complex and a high exposure complex. Cluster #2 is located in the lower
right quadrant of the map. This indicates an environment that is associated
with a low moisture complex and a high exposure complex. Cluster #3,
located in the upper right quadrant, is distinguished by its association
with the low moisture and low exposure complexes. Cluster #4, situated
in the upper right quadrant, is characterized by a high moisture complex
and a low exposure complex»
Since objective techniques were used to group these sites, intra-
cluster variation was minimized and each cluster can therefore be considered
relatively homogenous. This allows for data collected from these sites
to be used to predict plant succession. However, since these clusters are
ordinated along the axes of the territorial map and hence exposed to
different environments, variations in plant successions between clusters is
likely. This assumption will be tested in the next phase of the study.
37
-------
OXIDANT DOSE - CANOPY RESPONSE SUBSYSTEM
Introduction
A number of different data gathering activities have been undertaken in
order to provide the background needed to reach the following objectives:
1) Determine the influences of the May through October oxidant
dose and climate on biweekly increases of injury symptoms on
the foliage of ponderosa and Jeffrey pines and black oak.
2) Investigate the consecutive year results of tree injury in terms of
both the oxidant injury index or score of overstory trees and the
height growth of sapling trees.
In objective 1, the data sets that provide hourly temperature and
relative humidity information near vegetation plots (FSMET) or on vegetation
plots (PLOTMET) were used to compute cumulative daily transpiration for the
whole season using a transpiration simulation model (Reed and Waring,
1974). Other data required by the model are designated as STOMRESIST.
Operationally they consist of biweekly predawn xylem water potential
measurements and associated minimum daily stomatal resistance measurements.
A second type of data needed for objective 1 are the hourly ozone or total
oxidant averages at stations nearest the vegetation plot (OXIDANT) and on
the vegetation plots (PLOTOXID). A third data set required by objective 1
is the biweekly change of injury to needles or leaves of selected trees at
each of five vegetation plots (INJBIWK). Injury data is gathered monthly at
the remaining vegetation plots.
It will be necessary to couple the output of objective 1 with other
data sets described for objective 2 (below) after sufficient testing has
established the relationships of transpiration (an integrater of climate),
oxidant dose and single season foliage injury.
In objective 2, several effects of multiple years of injury are being
investigated concurrently but independent of the output from objective 1.
The data sets employed here include the annual evaluation of injury to
all plot trees (TREE and TREEMORT) and to saplings nearby (SAPTREE). A
retrospective measure of annual height growth of saplings in relation
to oxidant injury (SAPGRO) is also included.
A supplemental data set (HMET) includes daily maximum and minimums
of temperature, relative humidity and a one time daily wind and precipita-
tion measurement from 9 Forest Service Stations in the San Bernardino
Mountains starting in 1976.
38
-------
Materials and Methods
Equipment and Calibration Methods—
Meteorological—The remote stations for measurement of temperature,
relative humidity, winds and net radiation, (FSMET) at Camp Paivika, Sky
Forest and Barton Flats and their locations with respect to vegetation plots
have been described earlier (Miller, et al. 1977). Hygrothennographs
in standard weather instrument enclosures were placed on or near vegetation
plots. In May 1976, enough instruments were available to place one from
June through October at each location: Camp Angelus (CA), Deer Lick (DL) ,
and Heart Bar (HB). These instruments were within 0.8 to 1.6 km of the
vegetation plot of the same name. In May of 1977, instrument shelters
with hygrothermographs were placed at the following vegetation plots for
the entire summer season: CA,DL, Tunnel 2 (TUN2) and Camp Osceola (CAO).
The three remaining hygrothermographs were rotated among the other vege-
tation plots in the spring of 1977 with measurements for periods ranging
from 2-3 weeks.
Air pollution—DASIBI Model 1003AH ozone photometers were maintained
at the three FSMET stations (CP, SF and BF) in both 1976 and 1977. These
instruments were calibrated at the California Air Resources Board Labora-
tory in El Monte. Mast Model 724-2 total oxidant analyzers were calibrated
by one DASIBI which was reserved for this purpose at the Statewide Air
Pollution Research Center (SAPRC). The altitude correction factor for
each mountain station was calculated by dividing the larger pressure height
(millibars) at the SAPRC by the smaller pressure height at each mountain
station. Placement of Mast analysers was coordinated with the schedule
described for hygrothermographs on the vegetation plots. Mast analysers
and their strip chart recorders had to be battery powered at the vegeta-
tion plots. Four, 6 volt, 217 amp-hr batteries were required to operate
the Mast analysers when used with a DC to AC inverter. Each set of batter-
ies lasted about 4 days. The strip chart recorders were powered by an
internal rechargeable battery pack. The delivery of these Datamart Model
D/M #755-M-M recorders was delayed because of a redesign required to
lower power drain. Consequently, the on-plot oxidant measurements did
not begin until May 1977. The original schedule specified an August
1976 beginning date.
Measurement of Biological Variables—
Data inputs to the transpiration simulation model—The Reed and Waring
(1974) transpiration simulator requires both meteorological data, namely
hourly temperature and relative humidity, and biweekly measurements of
predawn xylem water potential and the associated minimum daily stomatal
resistance (rs). Reed and Waring define xylem water potential in its
absolute terms (Ibs/in^) and refer to it as predawn plant moisture stress
(PPMS). As soil moisture becomes more depleted during the growing season
both PPMS and minimum daily resistance are expected to increase, thus,
the daily actual transpiration (Ta) decreases although atmospheric demand
or potential transpiration (Tp) may remain large.
39
-------
Stomatal resistance was related to the only available method of measur-
ing transpiration, namely, infiltraton pressure (IP) in a pressure porometer
(Fry and Walker, 1967). A branch of ponderosa pine was cut, mounted on a
clamp and exposed to a turbulent air flow at 2000 ft-c and 25 C in the
laboratory during a 220 minute period. Initially, and at 10 minute, and
later at 30 minute intervals, it was weighed to determine water loss due to
transpiration. At similar intervals a needle faside was removed to deter-
mine the corresponding IP. Resistance (rs) was calculated at each inter-
val:
rs = vapor density gradient = sec cm~l
transpiration
The resistance values ranged from 27 to 115 sec cm"1. The following
relationships from Reed and Waring (1974) were used to calculate rs as a
function of IP over the range available:
In rs = 1.433 + 0.053 IP
The key equation necessary to simulate transpiration was the regression
between PPMS and minimum daily IP, the equation also from Reed and Waring
(1974) was:
IP(lb) = (2.50 x PPMS [bars]) - 0.641, (r = .53)
Daily ppms values were interpolated between the biweekly data points.
Actual transpiration (Ta) was then calculated by solving for diffusion
conductance (DC) in the equation:
DC = 164.56
Potential transpiration (Tp) was similarly calculated by assuming a maximum
value or the largest observed for DC during the entire season at each plot
location. Table 8 summarizes the steps described above.
The regression equations above were used for a preliminary run of
the transpiration simulator for two plot locations, Camp Angelus and Deer
Lick during the May through September period, 1976. In addition, the ratio
Ta/Tp was calculated daily as a measure of the relative drought stress at
each location and at different times during the season. This ratio is
proposed as one criterion for testing the quantitative relationship between
plant moisture status throughout the season and ozone dose response.
Concurrently, an effort has been under way to improve the accuracy of
40
-------
TABLE 8. MAJOR INPUTS, INTERNAL OPERATIONS AND OUTPUTS OF THE TRANS-
PIRATION SIMULATOR
Variables
Source or Product Dataset
Inputs
Mean Hourly Temperature ( C)
Mean Hourly Relative Humidity (%)
FSMET and
PLOTMET
Predawn Plant Moisture Stress (PPMS)
(interpolated between biweekly
determinations)
Minimum Daily Stomatal Resistance (rs)
(as a function of PPMS)
STOMRESIST
Internally Generated Parameters
Vapor Pressure Deficit (VPD)
(night and day)
Diffusion Conductance (DC) = l/rc
Outputs
Daily and Cumulative:
Actual (Predicted) Transpiraton (Ta)
Potential Transpiration (Tp)
Ratio: Ta/Tp
TRANSPR
41
-------
the regression between minimum daily stomatal resistance rs and PPMS
without relying on the pressure porometer so that rs can be substituted
for IP directly. A diffusion porometer fabricated and calibrated by the
Lawrence Livermore Laboratory (Bingham and Coyne, 1977) was used to obtain
some mimimum rs data. Unfortunately the unusual amount of summer rain
during 1976 and 1977 prevented us from obtaining sufficient data points at
moderate and high levels of PPMS. STOMRESIST is designated as the data file
for PPMS and rs. Plot locations where PPMS and rs data have been
obtained include DWA, TUN2, DL, CA, CAO and HB. Trees were selected as close
as possible to the soil moisture sensor column in each plot or immediately
outside the plot. Two to three trees were selected at each plot; one
represented an overstory tree larger than 30 cm dbh and the second an
understory tree less than 30 cm dbh. Trees with moderate to severe oxidant
injury were not selected at this time because the limited amount of time
available during the predawn period (3 to 6 am) made it impossible to
observe all injury classes especially because travel time between the 3
plots was a big limiting factor. The selection of healthy or slightly
injured trees was done to provide the initial data because these trees are
expected to have the greatest survivorship and thus a requirement for
predictive information. The concurrent studies being done by Coyne and
Bingham (1978) include three injury categories, e.g. very slight, slight to
moderate and moderate to severe at one ponderosa pine stand near Crestline.
Their data will be helpful to us when it is reported.
Estimates of injury to foliage at intervals during the season—During
the 1976 and 1977 summer seasons five plots were visited every two weeks
to observe the amount of visible injury to selected ponderosa or Jeffrey
pines, white fir and black oak. A minimum of two trees of each species in
each of six oxidant injury score categories were selected throughout for
continuing observation. The ponderosa and Jeffrey pine categories included
0-8 (very severe), 9-14 (severe), 15-21 (moderate), 22-28 (slight), 29-35
(very slight) and 36 and higher (no visible injury). White fir and black
oak were also divided into six catagories. Usually about three categories
were present on each plot as determined from the oxidant injury scores from
1975. Preselected trees were first inspected so that disease or insect
problems which may interfere with the development of oxidant injury symptoms
could be avoided. Three branches on each selected tree species were tagged
except for white fir which had one tagged branch. On the ponderosa and
Jeffrey pines, eight needle fasicles were labeled in each successively
younger needle whorl starting with the 1975 needle whorl. Each of the eight
needle fasicles were enclosed in a loop made with one length of colored,
vinyl covered copper telephone wire. Table 9 summarizes the kinds of data
gathered either biweekly or monthly in 1976 and 1977 from mid-June through
the end of September.
On ponderosa and Jeffrey pines a "3 M" device was used to measure
total needle length affected by chlorotic mottle. This "metric mottle
measurer" was a transparent plastic tube about 30 cm long and 0.5 cm,
i.d. It was painted black on one third of the external surface for the
full length, and etched every 0.5 cm. This device which was developed
by Tom Quick was proven to be useful. When the needle faside was
42
-------
TABLE 9. DESCRIPTION OF INFORMATION COLLECTED TO DESCRIBE THE WITHIN
SEASON DEVELOPMENT OF OXIDANT INJURY SYMPTOMS ON PONDEROSA
PINE (PP), JEFFREY PINE (JP), WHITE FIR (WF), BLACK OAK (BO).
Species
Type of Data
Frequency
PP and PJ
WF
BO
For each labeled branch,
and whole needle whorls,
starting with 1975:
—Total number of June and
needles per whorl September
For the same labeled
branch and for 8 labeled
or "wired" needle fasicles
in each whorl since 1975:
—Number of the 8 needles Biweekly
remaining. or
—Total length of each Monthly
of the 8 needles (cm).
—Portion of the total
length with identifi-
able symptoms (cm).
—Intensity of injury
symptoms on each needle
(Table 10).
For the single labeled
branch:
—Number of annual needle Biweekly
whorls retained (first or
observation in 1975). Monthly
—Intensity of symptoms
on needles or each
whorl (Table 10).
For each labeled branch:
—Intensity of symptoms Biweekly
on representative or
leaves (Table 10). Monthly
43
-------
slipped into the tube for measurement (down to the top of the fasicle wrap)
the needle was isolated momentarily against a dark background which aided
the measurement of chlorotic mottle. In Table 10 the different subjective
categories are described for characterizing symptom intensity. These
categories are more refined than those used in the end-of-season oxidant
injury score. In that case 0 = severe chlorotic mottle and advanced necro-
sis, 2 = any discernable chlorotic mottle and 4 - an uninjured green needle.
In the refined version, the greater injury is assigned a larger number so
that injury resposes would have a positive slope. Scores of 7, 6, and 5 for
the pines, white fir and black oak respectively signaled needle or leaf
abscision.
Records of vertical growth on ponderosa and Jeffrey pine saplings
1967 to 1976—In September of 1976, the internodal growth of saplings
was measured with the aid of a fruit picker's ladder (SAPGRO). Internodes
preceding 1967 were also measured when they could be confidently ident-
fied. The oxidant injury scores of each of the 50 ponderosa or Jeffrey
pines in the plots was last determined in 1975 (SAPTREE). These plots
were located within or nearby the following major vegetation plots: SF,
CP, DWA, BL, BP, TUN2, CA, HV, CAO and HB.
Annual measurement of oxidant injury to trees larger than 10 cm dbh
at major vegetaton plots—The procedure has been described in earlier
reports. Comparisons between years is done by using the paired t test
to compare each tree with itself from one year to the next.
Summary of data gathering activities—Table 11 shows when and where
eight data sets are obtained. Daily and cumulative daily transpiration
(TRANSPR) is calculated from FSMET, PLOTMET and STOMRESIST but it is
included in the table thus bringing the total number of data sets to nine.
The oxidant injury scores for saplings (SAPTREE) is not included in the
table because this information has not been obtained on a regular annual
schedule as originally planned. SAPTREE and TREE must be obtained during
the September to November period each year; TREE has been given first
priority. Good weather and qualified manpower were available in 1973
and 1975 to allow SAPTREE to be obtained. It must be done in 1978 without
fail.
Results and Discussion
Meteorological Effects on Seasonal Oxidant Dose—
One method of summarizing the influence of seasonal climate on oxidant
dose was introduced in the last progress report (Miller, et al. 1977),
namely, the frequency of five classes of meteorological patterns in southern
California (McCutchan and Schroedor, 1973) and the oxidant dose related to
single and consecutive days of each type. The following are brief definition
of these types:
1) Hot dry continental air all day (Santa Ana)
44
-------
TABLE 10. SUBJECTIVE CATEGORIES FOR DESCRIPTION OF OXIDANT INJURY SYMP-
TOMS ON PONDEROSA (PP), JEFFREY PINE (JP), WHITE FIR (WF), AND
BLACK OAK (BO).
Species
Numeric Category
Description of Leaf Symptoms
PP and JP
-Completely grass green (PP)
Completely gray green (JP)
-Slight chlorosis or very
slight chlorotic mottle
-Distinct, bright yellow
chlorotic mottle
-More intense chlorotic mottle
and some uniform chlorosis
-Intense mottle with necrosis
appearing at needle tips, not
exceeding the distal 1/3 of
the needle
-Intense mottle with necrosis
occupying the distal 2/3 of the
needle
-Entire needle necrotic, appear-
ing dry and brown
-Needle abscission
WF
-Completely green or gray-
green
-Light green and/or chlorotic
mottle barely distinguishable
on the sides (thinnest part of
the elliptical cross section) or
at the needle tip
(continued on next page)
45
-------
TABLE 10. CONTINUED
Species
Numeric Category
Description of Leaf Symptoms
WF continued
5
6
-Mottle more definite (bright
yellow) sometimes uniformly
chlorotic
-Intense mottle and the tip 1/3
of the needle is necrotic
-Uniform yellow and at least 2/3
of the needle is necrotic
-Needle is entirely necrotic
-Needle abscission
BO
0
1
-Leaf completely green
-First evidence of interveinal
chlorosis, chlorotic mottle
or necrotic lesions mainly
on upper surfaces
-Moderate levels of interveinal
chlorosis, chlorotic mottle
and/or necrosis mainly on
upper surface
-More severe than 2 with
necrosis extending to the
lower leaf surface
-Whole leaf is necrotic, both
surfaces
46
-------
TABLE 11. DESCRIPTION OF DATA TYPES AND FREQUENCY OF DATA COLLECTION FOR
EACH TYPE AT MAJOR VEGETATION PLOTS IN 1976 AND EARLY 1977.
Plot Spec- FSMET/
Name ies OXIDANT
COO p,f,o
TCP r> f n _____
or p , t 9 u _— ___
CP p,o hourly
SF p,f»o hourly
rvTJA r\ f rt _____
UCC p , o
TTTKT7 n f n _____
TIT -i f n _____
rw r> T f n —————
BL j,f
VWfW -i f _____
Of ,, f n _____
TTU •? -F r> _____
5P.R n f n
Acronyms for data types
STOM- INJ-
PLOTMET PLOTOX RESIST BIWK TRANSPR SAPGRO
_____ _____ on/near
plot/
biwk
hourly hourly on/near
plot/
biwk
hourly hourly on/near
plot/
biwk
plot/mo
pxot/iiio _____ ann.
plot /mo ann.
plot- daily/ ann.
sap/ cum.
biwk
plot /mo
plot- daily/ ann.
sap / cum .
biwk
plot/ daily/ — — — —
biwk cum.
p j,o t./ mo ______ ___ .___
plot /mo -— — — — — —
pxo t. / mo —_-«.___. ann. •
rvl nt* lmr\ ..-..._.- -___.-__
TREE
ann.
ann.
ann.
ann.
ann.
ann.
ann.
ann.
ann.
ann.
ann.
ann.
ann.
O1"»f^
(continued on next page)
47
-------
TABLE 11. CONTINUED
Acronyms for data types
Plot Spec-
Name ies
FSMET/ STOM- INJ-
OXIDANT PLOTMET PLOTOX RESIST BIWK
TRANSPR SAPGRO TREE
CA
BF
CAO
HB
p,f,o hourly
p,j,o hourly
j,f,o
hourly
near plot- daily/ ann.
plot/ sap/ cum
biwk biwk
plot/mo
hourly on/near plot- daily/ ann.
plot/ sap cum
biwk biwk
near
plot
biwk
plot/mo daily/ ann.
ann.
ann.
ann.
ann.
Species present: p - ponderosa pine
j - Jeffrey pine
f - white fir
o - black oak
Comprised of biweekly predawn plant moisture stress (PPMS) and associated
stomatal resistance (rs) measurements made on 2 ponderosa or Jeffrey
pines
10 cm dbh; located in plots or immediately outside plots within 30 m of the
soil moisture sensor column.
Needle injury data taken from selected plot trees of all species >10 cm
dbh and from sapling ponderosa and Jeffrey pines < 10 cm dbh at 4 plots
at monthly (mo) or biweekly (biwk) intervals.
Sapling height growth (SAPGRO) and the oxidant injury score (TREE)
are determined annually (ann.).
48
-------
2) Relatively dry forenoon, modified marine air in the afternoon;
very hot (heat wave)
3) Moist, modified marine air, hot in the afternoon
4) Moist, modified marine air, warm in the afternoon
5) Cool moist, deep marine air throughout the day.
It was concluded that consecutive occurrences of Class 3 days were the
most numerous in 1974 and resulted in the highest daily dose of oxidant.
Consecutive Class 4 days followed very closely behind in frequency and the
size of the resultant oxidant dose. The primary reason for the higher dose
in 1974 (Fig* 9) was the larger number of Class 3 followed by 3 transitional
combinations (Table 12). The lower seasonal oxidant dose in 1976 (Fig.
9) was evidently associated with a higher frequency of transitional com-
binations which induce lower daily oxidant doses, namely, 5-5, 4-5, 5-4
and 1-1 (Table 12). In general, the 1976 season was cooler and marked by
greater than usual rainfall particularly on September 11, when a tropical
storm persisted over southern California. The cumulative oxidant dose
in June 1976 was particularly lower than the June of 1974 and 1975 (Fig = 9).
There were 13 Class 5 (cool, moist) days in June 1976 compared to 7 in 1975
and 4 in 1974. These results give a general view of the trends in seasonal
oxidant dose in relation to climate. The results in the following section
will attempt to show the coupling between seasonal climate, oxidant dose and
tree injury response.
Daily and Cumulative Transpiration—
Daily and cumulative transpiration at Camp Angeles and Deer Lick,
1976—The transpiration simulator provides both actual and potential
transpiration on a daily basis. Potential transpiration is a measure of
atmospheric demand. The ratio of actual to potential transpiration (Ta/Tp)
is a useful index of the relative drought stress at different locations as
well as throughout the season at the same location. Several observations
can be listed: First, the time series in Figures 10 and 11 suggest that
higher cumulative weekly oxidant doses coincided with periods of higher
potential transpiration throughout the season at both Deer Lick and Camp
Angeles. The scatter diagram (Figure 12) relating daily oxidant dose to
daily potential transpiration at Camp Angeles in 1976 further suggests a
relationship between transpiration demand (Tp) and concurrent oxidant dose.
Second, the ratio Ta/Tp at both DL and CA responded by becoming lower
following continuous periods of moderate to high potential transpiration and
higher following significant rain on Julian dates 211 and 254. Third, it is
possible that a high oxidant dose occurring during a period of high Ta/Tp
(higher actual transpiration) would be more injurious than in the opposite
circumstance, namely, a low Ta/Tp. The reason may be that ozone flux to
needle tissue may be larger during periods of higher actual transpiration
(high values of Ta/Tp). For example, in Figure 10 ozone flux could be
expected to be larger in early July than in late August. In Figure 11, the
first small peak of Ta/Tp coincides with a short period of high oxidant in
mid-June and the lowest value of Ta/Tp in late august also coincides with
49
-------
U1
o
ID
O
fO
5.0-
4.5-
4.0-
o»
£ 3.5
A|
k.
f 3.0-
KJ_
o>
a.
2.5-
UJ 2.0
CO
O
O
,_ 1.5
§'-°
O
< 0.5
O
H o-
PERCENT OF TOTAL POSSIBLE HOURS OF DATA-y
JUNE —SEPTEMBER, RIM FOREST/SKY FOREST /
89
84
SE
90
AU
90
JL
90
JU
90
82
94
86
85
THREE-YEAR MOVING AVERAGE
100
100
SE
too
AU
100
JL
too
JU
IOO 96
IOO
SE
100
AU
100
JL
100
JU
90
JU
1968 1969 1970 1971 1972 1973 1974 1975 1976 1977
Figure 9. Trend of seasonal oxidant dose at a representative San Bernardino mountain
station from 1968-1977.
-------
TABLE 12. FREQUENCY OF DIFFERENT TRANSITIONAL COMBINATIONS OF FIVE
CLASSES OF SPRING AND SUMMER DAYS-
Transitional
Combinations
4
3
5
2
3
4
4
5
1
2
3
2
4
fi
**/
5
2
4
5
3
4
1
3
2
4
2
n
32
53
22
8
11
9
8
9
6
6
4
1
2
1974
Percentage
18.7
31.0
12.9
4.7
6.4
5.3
4.7
5.3
3.5
3.5
2.3
0.6
1.2
n
54
17
22
16
8
9
8
7
9
6
6
4
2
1975
Percentage
32.0
10.1
13.1
9.5
4.8
5.4
4.8
4.2
5.4
3.6
3.6
2.4
1.2
n
37
23
34
8
8
16
7
13
12
4
3
3
0
1976
Percentage
22.0
13.7
20.2
4.8
4.8
9.5
4.2
7.7
7.1
2.4
1.8
1.8
0
+Day classifications were obtained from Morris H. McCutchan,
Project Leader, Fire Meteorology Project, Pacific Southwest
Forest and Range Experiment Station, Forest Service, U.S.D.A.,
Riverside, CA.
*The 12 remaining possible combinations occurred only 3 percent
of the time or less and are omitted.
**The dose (ppm-hrs) on the second day of the most common transi-
tional combinations is indicative of the pollution potential,
for example, the combined data from 1974 and 1975 (Miller, et
al., 1977) showed the following doses associated with these
combinations: 3-3 = 2.09, 3-4 = 1.90, 2-2 = 1.88, 4-4 = 1.76,
5-5 =0.78 and 1-1 =0.38 ppm-hr.
51
-------
V)
cc
<
CO
UJ
cc
I
2
QL
Q.
UJ
to
O
o
z
<
o
X
o
o
-------
Cn
CAMP ANGELUS — 1975 NEEDLES
(INJURY SCORES 0-18) Y-JSIN(I6.4-I-0.2IXJ]
<£
£
80-
Ul
Ul
z 40-
t- 20-
o
z
-I
_ 0-
Ul
^60-
>
- 40-
x
i
u 20-
o
z
u>
T—T—T^—n—r
ISO 180 260 220 240 260
$ *
I p | I p T-p
JUNE JULY AUG SEPT OCT
(INJURY SCORES 19-21) Y-|siW39.3+0.30x]2
JUNE ' 'JULY ' AUG SEPT ' OCT
JULIAN DATES
1976
~n—n—i i i" "ir—i—n
160 180 200 220 24" "~ -J
r"-2-^
0 260 280
-7?
-2ui
ui
02
Ul
Oz
Figure 11. Comparison of injury to the 1975 needle whorl with potential transpiration
and the ratio of actual over potential transpiration at Camp Angelus in
1976.
-------
2.4
2.2
2.0
1.8
1.6
1.4
•B
Q- 1.2
a.
u
-------
a high dose period. A scatter diagram (not shown) relating the weekly
oxidant dose (ppm-hr) to the average of the daily values of Ta/Tp for the
same week for 18 weeks at Camp Angelus in 1976 did not suggest a relation-
ship between two variables over a range of 0.14 to 0.35 for Ta/Tp. An
adjusted ozone dose for the whole season might be calculated by multiplying
daily Ta/Tp times the daily dose but it appears that a wider range for the
Ta/Tp variable will be needed to more carefully examine the modifying effect
that it may have on the development of needle symptoms. MuKammal (1965)
found that by multiplying the daylight oxidant dosage by the coefficient of
evaporation he could reduce the scatter of data points describing dose-
injury response of tobacco.
At this time, we have calculated seasonal transpiration at only two
of the six plots where data is available (CA, CAO, HB, TUN2, DWA and
DL). It will be helpful to find the seasonal ranges of Ta/Tp at other
plots which are expected to be both drier and more moist than CA and DL.
Since three of the plots are dominated by ponderosa pine (CA, TUN2 and
DWA) and the remaining three have Jeffrey pine it may be possible to
evaluate differences in transpirational behavior that may be related to
species. When the data from 1977 is considered in addition, we have pro-
cessed only one sixth of the available data. An improvement in the preci-
sion of the regression formula relating predawn xylem water potential and
minimum daily stomatal resistance will be incorporated as soon as possible.
The relationship between soil moisture content and transpiration
(Ta) can be determined, as proposed in the systems modeling plan by Kickert
in Miller, et al (1977), in cooperation with Rod Arkley using the relation-
ship in the model described by Thompson and Hinckly (1977).
Development of ozone injury to foliage of ponderosa and Jeffrey
pines, 1976 and 1977—The preliminary results showing the relationships
between Tp, Ta/Tp, and cumulative oxidant dose (weekly and seasonally) (and
development of oxidant injury symptoms to ponderosa pine needles) is shown
in both Figure 11 and 13 for Camp Angelus in 1976. The increase in the
percentage of needle length exhibiting injury of any intensity (Table
11) and the increase of injury severity on the same needles versus Julian
date show differing responses depending on the oxidant injury scores of
ponderosa pines at Camp Angeles. Needle injury was barely detectable on
Jeffrey pines at Deer Lick in 1976; no data are presented in graphic form.
A regression line was calculated for "injured length" at CA because it is
derived from a continuous metric scale, but not for the needle injury index
because it is the product of a subjective judgement for which scale units
cannot be assumed to be of uniform size at all points of the scale. It
is evident however, that injury intensity increased moderately for the more
sensitive group of trees (0-18) compared to the less sensitive group (19-21)
The 1975 needles began the 1976 season with some injury accumulated in
1975. The new needles produced in 1976 (not shown in Figure 11) did not show
injury until the September observation and then about 12 percent of the
needle length exhibited injury. The increase of injury to the 1975 needles
in 1976 (1 year old) and 1977 (2 years old) is shown in Figure 13. From
1976 to 1977, the percent of the needle length with injury (chlorotic
mottle) of 1975 needles increased with very slight recovery over the 1976-77
55
-------
Oi
100-
80-
H
§60-
i
J40-
Q
U
£
o20-
z
I-
(9
Sn
gao-
K
360-
Z
3
g40-
£
U.
°20-
o
z
3 0-
CAMP ANGELUS — 1975 NEEDLES
(INJURY SCORES 0-18) Y- |?IN( 16.4 4- 0.21 X TJ *
. . • • 100-
. :__^_— -
: !^--*^* — ^"""^
1 ' ' ' » t
• t T
4
• *
t +
;*»>**
_ 80-
.? £
-*- D:
-6-3 60-
% -
Z UJ
-4> o 40-
o: uj
,3 UJ
-3-j z
-2 g *«0-
"! S 0
111 "^™
/N -5 UJ XV
JUNE JULY AUG SEPT OCT
(INJURY SCORES 19-21 ) Y -[siN (-39.3+ 030 xT|2
80-
p
~
5 60-
i
. • ^s' w
' .s"^ D 40-
.s^ UJ
•^^•' I i z°-
r^'^ . '
• •
*,**** -i —
«• "J r:
•s§ |
-4>- " 0-
ir J
ll
"' Ul
Ul
_rt Z
(INJURY SCORES 0-18) Y« |JN(48.7 + -I3XT| 2
t ftit
v -; 1 » r
! :
* * * 1 *
*
4 4
* i t * +
4 4
4 4 * 4 T
t t
JUNE JULY ' AUG' 'S'EPT ' OCT
(INJURY SCORES 19-21) • • Y- QlN(-2l +^8xTj2
•
.
.
^^
^s^ .
* — . '.
' ^^^
r ; ' :
; 4 4 4 4
. « t t {
* $ i * t
7
±
6x
•S
-4>-
or
31
_i
iS
UJ
oz
X
tr
_i
~ UJ
UJ
*^^
— i — i n i ' i M^ n " < i i« i ' i 1 1 i i i
160 180 200 220 240 260 280 160 180 200 220 240 260 280
JULIAN DATES 1976 JULIAN DATES 1977
Figure 13. Changes of 1975 needle whorl injury in 1976 and 1977 at Camp Angelus.
-------
winter for trees in the oxidant sensitive (0-18) score range whereas the
less sensitive group of trees (19-21) may have recovered slightly over
winter. The regression lines for needle length with injury show increases
of injury in 1977; symptom intensity (needle injury index) increased similar-
ly. Needle abscision was not observed until more than 90 percent of the
needle length showed injury on trees in the sensitive group (0-18).
As illustrated in Figure 13, the fate of 1975 needles can be followed
year after year. The same is true for 1976 and 1977 needles. Other
information that can be derived from the INJBIWK data set includes the
rate of needle elongation of the new needles each year; total length
of needles; the longevity of needles; and the annual change in green or
mottled needle surface area relative to tree sensitivity.
The biweekly changes in injury to white fir are not reported because
the lower injury response compared to ponderosa and Jeffrey pine has
resulted in little change of needle condition or needle retention over
the single 1976 observation year. Black oak injury data and the monthly
observations of injury to all four species has been keypunched and given
preliminary processing. These results will be reported in conjunction
with the 1977 injury data.
Progress in development of a single season ozone dose, foliage injury
model—Transpiration, oxidant dose and foliage injury data sets for CA and
DL have been completed for 1976 and 1977. A preliminary analysis of the
relationships between injury to 1975 needles from the moderate injury (19-21
injury score) group of ponderosa pines at Camp Angeles in 1976 (Figure 13,
lower left) and the cumulative oxidant dose alone or the cumulative dose
multiplied by the ratio of actual to potential transpiration (Ta/Tp) is
shown in Figure 14. The biweekly mean of daily Ta/Tp was multiplied times
the dose increment for the same period. The effects of the multiplication
were to "weight" each biweekly dose increment and to change the total of the
cumulative seasonal dose from 160 to 38 ppm-hr. It is evident from the
small range of values for Ta/Tp at Camp Angeles in 1976 (Figure 11, lower
left) that the ratio was nearly a constant until the very end of the season
(Julian Date 250 to 265); it did not "weight" the dose in any significant
way. In this example, either dose injury relationship in Figure 14 would be
acceptable. The analyses that will be done with data from other plots may
test the utility of the transpiration ratio in describing dose if the range
of the ratio is larger. The cumulative dose documented on a daily or bi-
weekly basis seems to be an appropriate unit for expressing dose. The
geometric mean for biweekly increments of the whole season does not seem to
provide a scale with fine enough resolution but it should be tested.
Existing dose response models (Larsen and Heck, 1976) deal with single
concentrations for periods not exceeding 8 hr. In chronic exposures, the
dose pattern is usually characterized by a series of high concentration
episodes occurring at random intervals and linked by consecutive days with
lower or moderate concentrations of oxidant; this generates new questions.
What sequences or seasonal patterns of high concentration episodes are most
injurious? It is assumed that periods of high actual transpiration would
contribute to more rapid injury development because of a larger ozone flux
57
-------
to
40
OL
I30
(T
| 2O
UJ
o I0
o
X
£ o
O UJ
£ Z 40
UJ <
o5
ct gj 3O
z
o
o.
a: 20
10
0>
CAMP ANGELUS -1976
FIVE TREES, MODERATE INJURY (19-21)
267
141
JULIAN DATE
197
Y=CSIN( 1.5+ i.
6 10 14 18 22 26 3O 34
BIWEEKLY CUMULATIVE OXIDANT DOSE (PPM-HR) X
BIWEEKLY AVE. TQ/T,
38
141
JULIAN DATE
197
267
(2.8 + 0.22 X)]]2
20 40 60 80 100 120 140 I6O
CUMULATIVE OXIDANT DOSE, PPM-HR (BIWEEKLY)
Figure 14. Changes in the percent of the total needle length of
selected ponderosa pines with chlorotic mottle in rela-
tion to two measures of cumulative oxidant dose.
58
-------
to foliage, but is there a point on the lower end of the available soil
moisture curve where the lowered transpiration rate (that may limit ozone
flux to needles) increases to drought stress proportions? Drought stress
would mainly accelerate abscision of the needles already injured by ozone.
The worst case conditions that may cause the greatest injury in a
single season are hypothesized as follows: Abundant soil moisture at
the beginning of the season followed by hot weather in June and July which
would result in numerous closely spaced episodes of high ozone concentra-
tions, and a continuing high potential transpiration throughout August
and September. The Ta/Tp ratio would decrease gradually until late July
and oscillate only slightly at a low level for the remainder of the summer
season. The drought stress induced in August and September may decrease
carbon fixation in the remaining needles in addition to accelerating needle
abscision.
Some very important results have been obtained by Coyne and Bingham
(1978) that focus on changes in stomatal behavior of pole-size ponderosa
pines in the San Bernardino mountain near Camp Paivika and of container-
grown trees under greenhouse conditions. The most important variables
measured in the study included: predawn and daylight xylem water potential
and stomatal conductance in relation to amount of chronic injury to pon-
derosa pines under field conditions. Some of these results run counter
to the hypothesis that has been developed in our report, namely, that
injury during a given season is mediated mainly by stomatal resistance,
i.e., high predawn xylem water potential would result in a lower stomatal
resistance (increased conductance), hence increased pollutant flux to needle
tissue. Their data shows that as needles on individual trees showing
increasing amounts of chronic injury (lower oxidant injury scores) become
more injured as the summer season progresses, their stomatal resistance also
increases. This event is accompanied by drastic declines in apparent
photosynthesis. In summary: 1) Increased needle injury is associated with
increased stomatal conductance and decreased stomatal resistance and de-
creased water loss; 2) The relationship between xylem water potential
(container grown trees) and stomatal conductance is basically sigmoid for
both uninjured and ozone-injured trees but as xylem water potential in-
creases over a arange from 0 to 25 bars the steep part of the curve is
between -10 and -14 for uninjured trees and -14 to -16 for ozone injured
trees; 3) Increased stomatal resistance would decrease pollutant flux to
needles. 4) Because other internal variables may also influence stomatal
behavior, e.g. higher levels of abscissic acid, "the influence of water
stress can more readily be described as an operational limit beyond which
stomatal aperture can not increase."
The small amount of data that we have described in this report does
not dispute the results of studies done by Coyne and Bingham (1978) and we
will hasten to complete the analysis of the remaining data so that our
research can be directed towards the most crucial questions pertaining to
the description of the chronic ozone dose-injury relationship for ponderosa
and Jeffrey pine.
Another variable that may have an important influence on injury develop-
59
-------
ment is needle phenology. Earlier studies with container grown ponderosa
pines fumigated with ozone suggested that the injury from the same concen-
tration, 0.45 ppm, increased as the summer progressed. A large number of
trees were held in a filtered air greenhouse and at 4 week intervals (during
3 summers) a new group of 40 trees was fumigated. Smaller doses were
required to cause equivalent injury to current and one-year-old needles as
the summer season advanced (mid-June to mid-September). These trees were
not water stressed (Miller, 1973). If this observation holds true under
field conditions, it would suggest that higher dose episodes in late August
and September may result in increased injury under the regulation of unknown
and unspecified controls at the physiological level; these controls may be
independent of stomatal behavior.
Annual Shoot Growth of Ponderosa and Jeffrey Pine Saplings From 1967-1976—
Internode lengths of terminal shoots and oxidant injury score—The
link that may be the most practical in coupling oxidant injury to growth
of smaller trees is foliage surface area retained. For example, Kozlowski
and Winget (1964) showed by removal of various proportions of needles from
Pinus resinosa that the old (all except the current year whorl) needles
provided the food reserve that accounted for four-fifths or more of all shoot
growth. The combined reserves in the branches, main stem and roots accounted
for less than 15 percent of shoot growth.
The oxidant injury score used for both sapling and sawtimber sized
trees is comprised of the number of needle whorls retained on the main stem
and a mid-crown branch for saplings and for the upper and lower crown in
larger trees. It is presently a crude index of foliage retained but may be
expanded to foliage surface area estimates in combination with other parame-
ters. In Figure 15, the preliminary linear regression lines show the average
of the annual shoot growth (internode lengths) between 1967 and 1976. Most
of this period was characterized by high seasonal oxidant doses (Fig = 9).
The best correlation between injury score determined in 1975 and growth was
at those plot locations experiencing the highest doses namely CP, SF and BP.
The maximum value for each line may be an approximate measure of the site
quality where the 25 to 40 trees were growing. The ponderosa pine plots had
shown greater height growth than the Jeffrey pine plots (HV, CAO, HB and BL).
This difference is believed to be associated with lower rainfall and lower
mean temperatures at the Jeffrey pine plots compared to the ponderosa pine
plots. The regression lines in Figure 15 should be regarded as an approxima-
tion because the independent variable (oxidant injury score) is not based on
continuous variable data but is a composite of discontinuous variables,
ranked variables and attributes. One of the most important problems that we
face is to find acceptable ways to use selected variables in the index for
purposes of statistical analyses.
At the ponderosa pine plots experiencing the highest doses of oxidant
(Miller, et al. 1977) the sample populations could be divided into several
injury score groups which also exhibited growth differences commensurate
with their injury category (Fig. 16). The year to year variability in growth
of all injury score groups is definitely associated with rainfall amount but
the differences between groups is mostly attributable to seasonal oxidant
60
-------
CO
i
C0200
C/5
O
o
X
§100
o
UJ
X
50
<
<
• COEFFICIENT OF CORRELATION
.66
CA(PP)r-.l4
BP(PP) r-.6l
T-2 (PP) r«.29
SF (PP) r=.56
HV (PJ) r«.23
CAO (PJ) r-.3ft
HB (PJ) 1T-.2B
BL (PJ) r-.28
10 20 30
OXIDANT INJURY SCORE
40
Figure 15. Average height growth of ponderosa and Jeffrey pine sap-
lings at plots experiencing different levels of chronic
oxidant injury.
61
-------
330-
320-
310-
300-
290-
28O-
270-
26O-
250-
2* 240-
Z 230-
J 220-
o
z 2IO-
w
J 2OO-
z 170-
160
ISO
I40H
130
120
110 H
ANNUAL TERMINAL GROWTH
P. PONDEROSA - CAMP PAIVIKA
INJURY SCORE
(15-21)—8 TREES
(9-14) —IOTREES
(0—8) 8 TREES
100 76 75 74 73 72 71 70 69 68 67
YEAR
Figure 16. Height growth of ponderosa pine saplings in three injury
categories at Camp Paivika between 1967 and 1976.
62
-------
doses which increased towards the end of this period (see the three year
moving average in Figure 9 from the nearby Sky Forest station). In a pre-
vious study, saplings were placed in carbon filtered air and even the most
severely injured trees improved but there was a 2-3 year lag before the
improvement in terminal shoot growth was evident (Miller, et al. 1977). Lags
may make it difficult to recognize the effects of a single year with a low
oxidant dose and non-limiting soil moisture. Additional analysis of the
SAPGO data will be done in order to estimate the effects of the amount and
the timing of oxidant dose and precipitation relative to annual shoot growth.
Annual Changes in Oxidant Injury Scores of Ponderosa and Jeffrey Pines
and Tree Mortality at Major Vegetation Plots—
Oxidant injury scores—The changes in mean scores for each plot in
1976 included higher scores (improved tree condition) at most of the 18
plots. In Table 13, the changes at individual plots are shown between 1973
and 1976. In 1976, 10 plot scores were significantly higher (p = .05), 2 were
significantly lower and the remaining 6 plots did not change significantly.
The trend to improve tree condition matches the decreasing seasonal oxidant
doses in 1975 and 1976 (Fig. 9).
Tree mortality—Tree deaths related to chronic oxidant injury decreased
also from 26 in 1974 to 19 and 11 in 1975 and 1976 respectively. The
4-year accumulated mortality in Table 13 is expressed as percentage of
the original number of ponderosa and Jeffrey pines on the 30 m wide plots.
There are several exceptions where tagged trees fall outside the 30 m specifi-
cation because in some cases the stands contained fewer trees >30 cm dbh;
in such cases trees were selected at measured distances outside the plot (to
avoid making the plot excessively long) to achieve the purpose of having 50
ponderosa or Jeffrey pines ^30 cm dbh in each plot-
63
-------
TABLE 13. TRENDS IN PONDEROSA AND JEFFREY PINE CHRONIC INJURY SCORES
AND TREE DEATH AT EIGHTEEN PERMANENT PLOTS.
Plot
Name
COO
BP
CP
SF
DWA
UCC
TUN 2
DL
GVC
BL
NEGV
SC
HV
SCR
CA
BF(PP)
BF(JP)
CAO
HB
Score
1-8
9-14
15-21
1973
Injury Number
Score Dead
15.1
16.8
17.0
13.4
20.2
15.5
19.2
21.7
29.4
33.1
41.3
46.3
12.4
25.6
22.5
21.3
21.7
44.0
0
2
0
1
0
0
0
-
0
0
0
0
0
0
1
5
1
1
0
YEARS Accumulated
1974 1975 1976 Mortality
Injury Number Injury Number Injury Number 1973-1976
Score Dead Score Dead Score Dead Percent
12
16
16
13
16
15
17
18
20
31
32
47
47
11
17
19
20
24
39
.9
.5
.7
.8
.5
.9
.0
.6
.1
.8
.1
.3
.7
.7
.4
.3
.4
.6
.6
0
2
2
1
0
1
1
0
0
0
0
0
0
0
2
5
2
9
1
Interpretation:
= Very severe
-
Severe
= Moderate
12.5
14.5
14.7
15.3
16.0
16.3
17.0
19.2
23.1
30.5
40.6
44.5
44.6
12.5
17.9
20.5
21 = 5
25.7
33.9
22-28
29-35
36 &
4
1
0
2
2
0
1
1
0
1
1
0
0
1
1
2
' 1
1
0
above
13
15
17
15
18
16
18
20
21
34
36
46
50
16
17
23
23
33
38
=
=
.8 0
.5 0
.9++ 1
.0 0
.2++ 0
.0 0
.64"1" 0
.7 0
.0 0
.5* 2
.0 0
.0 0
.5^ 0
.7++ 1
.5 1
.0*"*" 6
• S4* 0
.2 0
.2^ 0
Slight
Very slight
No visible
6.
6.
3.
3.
2.
1.
2.
1.
0
2.
1.
0
0
4.
7.
10-
6.
8.
0.
injury
6
9
1
3
4
5
7
5
1
5
0
3
7
9
9
8
Significant Difference (p = 0.05) between 1975 and 1976
64
-------
EFFECT OF PHOTOCHEMICAL OXIDANTS ON TREE GROWTH IN THE SAN BERNARDINO
NATIONAL FOREST
Introduction
History of Oxidant Impact on Tree Growth—
Symptoms of air pollution injury were first observed in the San
Bernardino mountains on ponderosa pine (Pinus ponderosa Laws.) in 1953
(Asher 1956), although the direct link between cause and effect was not
discovered until early 1960. Symptoms included loss of all but the current
season's needles, reduction in number and size of needles, and yellow
mottling of the needles. Miller et al. (1963) confirmed that these injury
symptoms were produced by ozone. McBride et_ al. (1975) and Parmeter et al.
(1962) demonstrated significant reductions in both radial and height growth
increment of ponderosa pine as a result of oxidant exposure. Additional data
was necessary to more completely document the impact of oxidants on tree
growth and to properly calibrate a stand growth model for the SBNF.
Research Objective
1) Document the impact of photochemical oxidant injury on radial
growth increment of the major tree species occurring in the SBNF.
2) To examine the mortality rates and radial growth patterns of
forest trees in the SBNF with the aid of a stand development model.
Materials and Methods
Selection of Study Plots—
Six vegetation plots were selected for examination. A smog injury
rating system, developed by Miller (1973), was applied to the trees on
each plot. This system involves examination of a tree with binoculars
and scoring it as to needle retention, needle condition, needle length
and branch mortality. The six plots were then classified according to
the most common injury scores. Two plots contained a majority of trees
rated as severely injured, COO and BP. Two plots were rated as moderately
damaged, GVC and DL; and two plots had very slight or no visible smog
damage, HV and NEGV. The plots varied in length from 100 to 300 m.
Sampling Procedure—
During the summer of 1976 each conifer greater than 10 cm. dbh
65
-------
(diameter at breast height) on the six study plots was cored with an incre-
ment borer. The core was taken at a height of 1.5 m on a randomly chosen
side of the tree trunk. The cores were returned to the laboratory, dried at
70°C for 24 hours, mounted on boards and sanded, aged and the growth rings
measured to the nearest 0.01 mm back to and including the year 1920. Each
tree was classified as to the soil type on which it occurred.
Statistical Analysis—
Growth data from trees within and between stands was quite variable.
To reduce this variation as much as possible it was felt that the trees
should be stratified before analysis. Trees were divided into groups
delineated by species, plot, age and soil type. Within each group the
growth data for the past 30 years (excluding 1976) was considered for
analysis (i.e., 1946-1975). This period was divided into ten year growth
periods; 1946-1955, 1956-1965, 1966-1975. A period of this length (30
years) should be adequate to detect growth trends as influenced by photo-
chemical oxidants which have steadily been increasing in recent years
(Corn et al. 1975).
Examining the growth rings on a tree core is, in actuality, sampling
the growth of a given tree through time. This situation lends itself
to examination using a repeated measure analysis of variance (ANOVA) design
(Sokal and Rohlf, 1969).
No statistical tests were performed to compare growth between plots.
Since so many variables among them are different it was felt that they
were not quantitatively comparable. However, it was felt that, quali-
tatively, the growth trends for two similarly stratified tree groups on
different plots could be compared.
Results and Discussion
Impact of Oxidant on Radial Increment—
Radial growth of forest trees can be affected by many different phe-
nomena. In order to examine the effects of oxidants on tree growth it
is necessary to eliminate other variables that affect tree growth to
ensure that there are no confounding interactions among these variables.
One important variable influencing tree growth is precipitation. One
weather station in the vicinity of the SBNF was selected as being most
representative of the precipitation in the SBNF (Squirrel Inn #2). The
precipitation records for the years in question are presented in Table
14. A one way analysis of variance was performed on these data to detnnine
if any significant differences in precipitation existed among the three
growth periods and none were found. Therefore, significant differences in
radial growth among the three growth periods cannot be directly attributed
to rainfall.
It was necessary to insure that tree growth was not correlated with
precipitation during the years encompassed by this study. Mean annual
precipitation, recorded at 3 weather stations in close proximity of the
66
-------
TABLE 14. ANNUAL PRECIPITATION (cm) AT SQUIRREL INN #2 WEATHER STATION
AND AN ANALYSIS OF VARIANCE AMONG THE TEN YEAR INTERVALS THAT
CORRESPOND TO THE TEN YEAR GROWTH PERIODS.
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
n
X
Sx
Source
Between
Within
Total
35.89
83.29
109.32
55.85
99.26
147.32
36.68
120.19
78.97
63.35
10
83.01
34.61
d.f
Periods 2
Periods 24
26
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
SS
1599.915
43009.607
44609.522
125-78
113.46
62.26
73.74
45.16
87.05
86.59
76.07
169.88
90.27
10
93.03
33.73
MS
799.958
1792.067
1966 151.33
1967 49.05
1968 200.25
1969 91.59
1970 88.54
1971 35.13
1972 102.13
7
102,57
52.98
F
.446 ns
Reject H0 if F > F2;24/.05 = 3'40
67
-------
specific study plots, is presented in Table 15. Mean annual radial growth
from several trees on each plot was used in a regression analysis with mean
annual precipitation from the appropriate weather station. No significant
correlations were found between tree growth and precipitation.
Plot size, number of individual trees per plot, species composition,
density, and basal area for each plot are described in Table 16. Data for
the Deer Lick plot is incomplete.
COO and BP, the two plots that have a severe smog injury rating,
were the only two plots that showed significant changes in radial growth
trends (Table 17, Figs. 17 to 22). Growth in all of the tree groups
in COO shows a significant reduction (Fig; 17) but this trend was not as
distinct in the BP plot (Fig. 18).
In the first growth period (1945-55), the yearly radial growth incre-
ment for trees in COO was much greater compared to growth in the other
plots, and was still slightly greater in most groups even after drastic
reductions by 1975 (Table 17, Figs. 17-22). This could be explained,
in part, by the predominately younger age groups in the COO plot, since
younger trees tend to grow faster than older trees. The site class of COO
may also be higher. Trees in the BP plot were older than most of the
trees in COO and showed a more variable growth response. Older trees may
not react as strongly to oxidant exposure as young trees. The youngest
group of trees in BP were the fastest growing in 1945-56 and showed the
strongest decrease in growth through time. Faster growing trees may also
react more to oxidant exposure than slower growing trees. Since the young-
est trees were usually the faster growing individuals it is impossible to
determine which factor may be responsible for this strong reaction to
oxidant exposure.
TABLE 15. RAINFALL DATA FOR THE SIX STUDY PLOTS (MEAN ANNUAL PRECIPI-
TATION cm)
1946-55 1956-65 1966-75
Green Valley Weather Station
for COO, GVC, DL, NEGV
Panorama Point Weather
Station for BP
Big Bear Dam Weather
Station for HV
83.84
84.28
61.74
86 = 59
86.21
80.92
97,94
+No available data.
68
-------
White fir trees on COO are affected just as severely as the pine
(Fig. 17) while incense cedar on BP did not seem to be responding to
oxidant exposure (Fig. 18).
The two plots rated as moderately damaged (GVC and DL) showed no growth
response to oxidant exposure (Table 17, Figs. 19 and 20). The presence of
visible foliar damage on these plots may indicate a threshold phenomenon.
Visible injury, or oxidant exposure, may need to reach a certain level
before radial growth increment is affected. Another explanation may be that
oxidant impact on radial growth may not be expressed in the bole of the tree
at 1.5m but may cause measurable increment loss within the crown. Williams
(1967) demonstrated that spruce budworm defoliation reduced radial increment
of grand fir, Douglas-fir, and Englemann spruce to a greater extent within
the crowns than near the ground. The same simulation may exist here.
The two plots with slight or no visible smog injury (HV and NEGV)
showed some significant growth trends (Table 17, Figs. 21 and 22). The
Jeffrey pine over 200 years old in NEGV show a significant but small in-
crease in growth (Fig = 21). This defines no signicant trend in growth.
Jeffrey pine on the HV plot from 41-50 and 51-60 years old showed a signifi-
cant but small decline in growth. This reduction was probably due to the
natural, gradual decline in radial growth increment at breast height that
occurs as a tree ages (Duff and Nolan, 1953)•
The younger trees of each species are growing faster than the older
trees of the same species on any given plot. White fir is the fastest
growing tree on each plot when compared to other species of the same age.
It appears to be better adapted for faster growth than the other tree
species in the San Bernardino mountains but is just as susceptible to
oxidant injury as ponderosa pine.
Not all species occurred on each plot so it is not possible to make
statements on their susceptibility to specific levels of oxidant injury.
Stand Development Model—
The data collected from the six study plots has been sent to Dr.
A. R. Stage, U. S. Forest Service, Intermountain Forest and Range Experiment
Station, Moscow, Idaho to be run through his Stand Prognosis Model. No
results have been obtained from this effort to date. Work on this aspect of
the project will continue.
69
-------
TABLE 16. MENSURATIONAL DESCRIPTIONS OF THE SIX SAMPLE PLOTS.
PP SP
1C
WF
JP
BO
QW
CL
COO No. of Trees/
Plot
170m Spp.
Composi-
tion(%)
Severe Density
Damage (#/ha)
Basal Area
(m2/ha)
BP No. of Trees/
Plot
100m Spp.
Composi-
tion(%)
Severe Density
damage (#/ha)
Basal Area
(m2/ha)
GVC No. of Trees/
Plot
300m Spp.
Compos i-
tion(%)
Moderate Density
damage (#/ha)
Basal Area
(m2/ha)
NEGV No. of Trees/
Plot
180m Spp.
Composi-
tion(%)
No Density
damage (#/ha)
Basal Area
(m2/ha)
46
37
90.2
14.52.
71
43
236.7
28.10
1
1
2.0
.72
11
4
12.2
2.93
3
2
5.9
1.53
30
22
100.0
11.86
10
4
11.1
1.63
1
20 |
1
1
16 1
1
39.2 |
6.00 |
1
1
1
1
1
1
1
1
I
1
62 |39
1
1
26 |16
1
68.9 143.3
4.94 |12.20
1
1
8 |65
1
1
10 |89
1
14.8 |120.4
1.05 |35.00
1
1
39
31
76.5
4.80
31
23
103.3
4.08
82
35
91.1
4.70
1
1
3.3
0.04
1
1
1.1
0.02
1
1
1
70
-------
TABLE 16. CONTINUED.
HV No. of Trees/
Plot
290m Spp.
Compos i-
tion(%)
No Density
damage (#/ha)
Basal Area
(m2/ha)
DL No. of Trees/
Plot
120m Spp.
Compos i-
tion(%)
Moderate Density
damage (#/ha)
Basal Area
(m2/ha)
Table Legend
PP Ponderosa pin«
SP Sugar pine
1C Incense Cedar
WF White fir
PP
A
SP
22
15
1C
JI
B(
QV
CI
WF
24
11
27.6
3.4
32
21
—
3
)
1
J
JP
168
83
193.1
22.07
55
37
^••B
•M^B
Jeffre
Black
Quercv
Cercoc
BO
4
1
4.6
.21
40
27
^•^A
;y pine
Oak
is wisl:
:arpus ]
QW
izenii
LedifoJ
CL
1
5 \
1
1
2 I
1
5.7 |
0.13|
1
1
1
1
1
1
1
1
1
1
+
.ius
Note: From McBride, J. R. 1974. Annual report of the vegetation sub-
committee for the fiscal year ending. In: Taylor, 0. C. Oxidant air
pollution effects on a western coniferous forest ecosystem. Task D.
Statewide Air Pollution Research Center, Riverside, California.
71
-------
TABLE 17. AGE GROUPS, SOIL TYPES, MEAN ANNUAL RADIAL GROWTH INCREMENT
(mm) AND STANDARD DEVIATION FOR EACH GROWTH PERIOD, F VALUES FROM
ANOVA, AND EXPLAINED VARIANCE (co2) FOR EACH STRATIFIED GROUP OF
SAMPLE TREES ON EACH PLOT.
COO Age Soil Type
PP
PP
PP
PP
WF
BP
PP
PP
PP
1C
51-60
51-60
41-50
41-50
41-118
61-100
41-60
61-100
61-100
PxClCm
PxClDm
PxClEm
PxClDm
PxClEm
PHclCm
PHclCm
PHclDm
Phc IDm
n
9
9
9
9
7
15
8
38
15
1946-55
3.26 +
2.89 +
3.39 +
3.76 +
3.87 +
1.25 +
1.96 +
1.51 +
1.78 +
.92
.54
1.55
1.29
.75
.65
.58
.74
.85
1956-65
2.13 +
1.96 +
2.36 +
2.25 +
3.24 +
1.52 +
1.21 +
1.78 +
1.58 +
.65
.38
1.14
.97
.98
.65
.68
.67
.66
1966-75
1.33 +
1.64 +
1.48 +
.78 +
2.47 +
1.75 +
.56 +
1.17 +
1.87 +
.76
,90
1.00
.33
.98
.62
.21
.72
.81
F
23.24
23.77
16.22
55.87
6.92
n.s.
27.00
6.65
n.s.
CO2
.510
.401
.280
.620
.286
.542
.047
GVC
JP
WF
WF
SP
1C
DL
JP
SP
61-200
61-200
51-60
TsEm
ExsaDm
TsEm and
ExsaDm
41 or older
over 31
101-200
data
61-200
data
no
available
no
available
14
14
9
10
7
33
17
.87 +
1.20 +
1.80 +
1.28 +
1.85 +
-55 +
.88 +
.41
-53
.85
.58
.81
.33
.37
.76 +
1.07 +
1.90 +
1.29 +
1.23 +
.52 +
.89 +
.34
.53
.76
.73
.58
.34
.34
.69 +
1.33 +
1.92 +
1.55 +
1.38 +
.44 +
1.00 +
.50
.89
.81
.81
.82
.31
.39
n.s.
n.s.
n.s.
n.s.
5.81
10.94
3.39
.112
.022
-022
WF 61-200 no 19
data available
NEGV
JP over 200 TsEf 17
JP 101-200 TsEf 20
1.09 + .40 1.17 + .41 1.17 + .38 n.s.
.45+ .16
.77 + .49
.45+ .22
.79 + .42
.56 + .78
.77 + .34
4.37 .046
n.s.
72
-------
TABLE 17. (continued)
HV Age Soil Type n
JP
JP
JP
JP
WF
over 200
41-50
51-60
61-100
61-100
TAflDm
TAflCm
TAflCm
TAflBm
TAflBm
9
34
22
17
12
1946-55
.60 +
1.74 +
1.53 +
.75 +
1.00 +
.22
.68
.63
.39
.56
1956-65
.43 +
1.24 +
1.04 +
.64 +
.92 +
.14
.55
.53
.38
.46
1966-75
.43 +
1.21 +
1.11 +
.65 +
1.04 +
.18
.68
.64
.42
.64
F
7.48
43.43
26.77
n.s.
n.s.
C,2
.166
.126
.114
Table Legend:
Code
PxClCm
PxClDm
PxClEm
PHclCm
PHclDm
TsEm
ExsaDm
TsEf
TAflDm
FAfICm
TAflBm
Soil Type
Pachic Xerumbrept on 5-10% slopes
Pachic Xerumbrept on 19-23% slopes
Typic Xerorthent
Pachic Ultic Haploxerolls on 10-15% slopes
Pachic Ultic Haploxerolls on 15-30% slopes
Typic Xeropsamments
Entic Xerorthents
Typic Xeropsamments
Typic Argixerolls on 15-20% slopes
Typic Argixerolls on 9-15% slopes
Typic Argixerolls on 3-9% slopes
73
-------
o
_j
<
a
<
a:
z
z
UJ
3.0-
2.0-
1.0-
CAMP 0-ONGO
TxClEm
PxCIDm
PxCICm
WF
4ltoll8yrt
T
1946-55 1956-65
TEN YEAR GROWTH PERIODS
1966-75
Figure 17. Growth trends for trees on the COO plot from 1946
to 1975 in the San Bernardino National Forest.
^ 3.0 H
o
<
a
K 2.0-
_j
3
Z
<
1.0
UJ
BREEZY POINT
• PHclCtn
•PHclDm
pp
61 to 100 yn
1C 61 to 100
PP 61 to 100
PP 41 to 60
1946-55 1956-65 1966-75
TEN YEAR GROWTH PERIODS
Figure 18. Growth trends for trees on the BP plot from 1946
to 1975 in the San Bernardino National Forest.
74
-------
s
5
I 3.0'
(T
O
5 2.0-
_J
z>
•z.
<
1.0-
UJ
6REEN VALLEY CREEK
-ExsaDm
-TsEm
-Both Soils
-Entire Plot
JP 61 to 100
1946-55
1956-65
TEN YEAR GROWTH PERIODS
1966-75
Figure 19. Growth trends for trees on the GVC plot from 1946 to 1975
in the San Bernardino National Forest.
o 3.0-
cr
(S
2.0
<
UJ
1.0-
DEER LICK
NO SOIL CLASSIFICATION
WF
6l»o2OO»
biro zooyn
-•SP6H0200
•JP 101 to 200
1946-55
1956-65
TEN YEAR GROWTH PERIODS
1966-75
Figure 20. Growth trends for trees on the DL plot from 1946 to 1975
in the San Bernardino National Forest.
75
-------
I
t-
o
cr
3.0-
2.0-
<
UJ
1.0-
NORTH EAST GREEN VALLEY
-TsEf
JP
•lOltoZOOyrs
1946-55
1956-65
TEN YEAR GROWTH PERIODS
1966-75
Figure 21. Growth trends for trees on the NEGV plot from 1946 to 1975
in the San Bernardino National Forest.
o 3.0
K
O
o
<
or
z
UI
2.0-
1.0 H
HOLCOMB VALLEY
-TAflDm
-TAflCm
-TAflBm
•JP6MOIOO
'JPov»r200
1946-55
T
1956-65
TEN YEAR GROWTH PERIODS
1966-75
Figure 22. Growth trends for trees on the HV plot from 1946 to 1975
in the San Bernardino National Forest.
76
-------
PHYSICAL AND CHEMICAL PROPERTIES OF SOILS, INCLUDING MOISTURE DYNAMICS
Introduction
During 1976-'77 the questions addressed were (1) What are the soil
moisture and temperature regimes and how do they relate to the suscepti-
bility of the vegetation to damage during periods of high oxidant air
pollutant concentrations and do they affect the impact of oxidants on other
organisms such as pathogenic fungi, arthropods and litter decomposing
organisms? (2) How do physical properties of the soils affect these mois-
ture and temperature relationships? (3) Do the plant nutrients in and
chemistry of the soil affect the relationships described above? In this
context, three kinds of soil characterization measurements carried out were:
soil moisture retention properties; soluble soil phosphorous; and identifi-
cation of the soil and slope with each tree on the major vegetation plots.
The first two characterization measurements were descriptive of the system.
Soil moisture and temperature data collected continuously since summer
1973 can be used to document the complete water balance including periods
of soil moisutre deficit. This data will then be related to the physiolog-
ical conditions of the vegetation measured as xylem water potential and
transpiration rate.
Material and Methods
Soil Moisture and Temperature Measurement—
Soil moisture-temperature sensors (fiberglass moisture blocks) install-
ed at depths of 15, 31, 61, 92, 152, 214 and 274 cm were monitored every one
to three weeks depending upon the rate of change. Field readings from the
moisture sensor were converted to percent water content (Pw) based upon
calibration curves for soil at each sensor site. Soil samples were col-
lected at the sensor sites and water content was determined in the labora-
tory to construct the calibration curves. The values obtained were used
mainly to describe the shape of the curves of soil moisture change with
time. The values were then adjusted based upon soil moisture content
determined by mechanically sampling of the soil at the same depths as the
moisture blocks with an auger boring within 1.5 meters of the site of
the sensors. Moisture retention data also determined on the same sensor-
site samples were used as a further check upon the accuracy of the final
corrected values. This procedure was adopted because of the very large
number of moisture sensors emplaced in the field. Soil temperature was read
directly from the sensors by means of calibration data stored in the compu-
ter program. Both soil moisture and temperature can be monitored for
accuracy by comparing the results from a given sensor with those above and
77
-------
below it in the soil column. Any erratic behavior can be readily detected
and corrected. A few sensors failed and were replaced due mainly to the
wire leads being bitten off by pocket gophers. The sensors were remarkably
stable and consistent in their behavior.
Nature and Pattern of Soils—
Detailed slope and soil classification and mapping completed previously
on 18 major vegetaton plots were used to characterize the slope and soil
under each pine tree on the plots in order to develop relationships of
soil and stand dynamics.
Soil properties measured—Particle size distribution, bulk density,
exchangeable and soluble cations, soil pH, organic matter and nitrogen were
previously measured in the soils on major vegetation plots (Miller et al.,
1977) During the 1977-'78 period soluble phosphate was measured on 138
surface soil samples taken under 40 trees selected for their relationships
to other subprojects dealing with arthropods, litter decompositon, and
pathogenic fungi. The latter completes the analyses of the surface soil
samples, reported previously, which included organic carbon, nitrogen,
exchangeable sodium, potassium, calcium and magnesium.
Specific analytical procedures—Moisture retention was determined
on soils from all major vegetation plots and on two aspect plots at the
same sites used for soil moisture monitoring by fiberglass moisture block
sensors. Soil moisture tension was measured at 0.1, 0.33, 1, 5 and 15
atmospheres by the methods described by Richards in the book edited by
Black (1965). However, duplicate 1.0 cm rings were used rather than 5
cm diameter rings described in the method for holding the soil on the
tension and pressure plates. Soil moisture retention was determined on each
site at depths of 15, 61, 91 cm and over 200 cm for deep soils. Soluble
phosphorous was determined by a colorimetric method (Jackson, 1960) for the
surface soils collected to a depth of 7.5 cm of soil.
Results and Discussion
Soil Moisture Retention Characteristics—
The computer programs for processing the large amount of data (over
8,000 field readings on the 153 sensors) has been completed and complete
curves were plotted from summer of 1973 to date (summer 1977) . These
curves were not adjusted for the gravimetric control moisture sampling and
so are not included in this report - Adjusted curves are being prepared
and will be made available for all participants in the project soon.
However, inspection of the nonadjusted curves show that the moisture-
temperature sensors have been remarkably stable throughout the 5 year
period, in that the maxima of winter moisture and summer minima are nearly
the same each year. Also the shape of the moisture depletion curves vary
from year to year, and with increased depth of soil or variation in soil
texture. Further, errors in field readings are readily detected as indi-
cated by individual variations from the smooth curves not accounted for by
sudden rainfall or abrupt changes in air temperature which may affect the
78
-------
upper sensors.
It appears that the soil moisture-temperature monitoring during
the 5 year period has been eminently successful and can be used effectively
in evaluating soil relationships to the impact of oxidant air pollution
on the ecosystem. Rates of soil moisture depletion will be related to
transpiration by the vegetation and to moisture stress on the plants.
The results of the analyses are shown in Table 18 and 19 and a typical
set of moisture retention curves are shown in Figure 23. Water held
in the soil at low tension (0.1 atm.) is an approximation of field capacity
of the soil to hold water when freely drained. The minimum water content
which is available for plant use is considered to be about 15 atm; this
minimum is called the permanent wilting point. Difference is an approxima-
tion of the amount of water that can be stored in the soil which is avail-
able for plant growth. Values of the differences in percent water for the
major vegetation plots are shown in Table 20 and are called "available soil
water". However, these values are only approximations and do not always
agree well with measurement of samples taken directly from the field, and
should be used with caution. They will be used to help in calibrating the
soil moisture monitoring program, and for direct measurement of soil mois-
ture sampled when wet in the spring and when dry in the late summer or fall.
"Available soil water" indicated in Table 20 ranges from 7.17 and 7-35
percent by weight (Pw) in the two plots with the sandiest textures to nearly
24 percent in the surface of a number of plots where the soil is rich in
organic matter. Available soil water tends to increase with increased clay
content in the upper layers of these sandy soils, but there appears to be
little relationship between the two variables below 61 cm.
79
-------
TABLE 18. SOIL WATER CONTENT AT VARIOUS MATRIC SUCTIONS FOR THREE PROFILE
DEPTHS.
Site
BF
BL
CAO
DW2
GVC
HV
NE12
NEGV
SCI
SC2
SV2
UCC
BP
COO
DW1
DW3
HB
CP
S22M
SCR
SF
TUN2
0-30
31.73
18.31
20.79
22.48
13.87
16.72
25.63
13.22
11.73
14.22
16,56
20.44
26.87
22.30
26.29
24.04
15.79
27.73
26.15
10.54
29.40
19.11
pw
1/10 BAR
30-61
26.88
15.78
20.45
22.22
14.88
12.89
18.59
12.85
12,20
13.45
n.d.
21.04
25.00
18.40
18.80
22.20
11.90
24.50
20.30
10.90
25.90
17-20
61-91 cm
23.02
12.23
19.44
17.86
12.59
13.05
17.29
11.10
13.85
11.17
n.d.
20.07
21.66
17.00
19.02
17.66
12.36
n.d.
18.63
10.50
21.21
n.d.
80
0-30
20.97
11.55
14.98
17.48
10.64
11.50
18.03
9.21
8.74
9.47
13.05
15.90
18.18
17.03
17.40
16.77
11.19
19.83
20.97
6.38
21.36
13.91
pw
1/3 BAR
30-61
15.24
8.37
12.91
13.82
10.25
8.99
12.82
7.71
7.87
6.77
6.59
14.65
16.05
12.85
13.45
13.65
7.47
15.74
14.31
5.93
17.17
11.00
61-91 cm
13.84
7.42
13.77
12.83
8.82
8.94
11.98
6.65
8.28
5.83
n.d.
15.26
17.88
13.55
15.62
13.32
9.42
n.d.
14.56
7.07
17.71
n.d.
-------
TABLE 18. CONTINUED
Site
BF
BL
CAO
DW2
GVC
HV
NE13
NEGV
SCI
SC2
SV2
UCC
BP
COO
DW1
DW3
HB
CP
S22M
SCR
SF
TUN2
0-30
14.61
8.29
11.25
12,55
7.89
8.59
12.60
6.52
6.69
6.99
9.32
11.09
12.55
12.53
11.91
10.76
6.26
13.39
13.38
5.15
14.74
8.80
pw
1 BAR
30-61
10.15
5.31
9.04
10.21
7.52
7.25
8.83
5.36
6.09
4.66
4.58
11.60
12.10
10.34
9.24
10.43
5.72
12.45
11.65
4.56
13.34
7.71
61-91 cm
9.28
5.25
9.72
9.22
6.49
7.07
8.45
5.25
6.25
4.29
n.d.
12.27
13.29
9.86
10.30
8.54
6.39
n.d.
9.92
4.14
10.96
n.d.
0-30
8.81
5.39
6.33
7.46
3.97
6.23
7.59
4.30
5.00
5.66
5.21
5.60
7.96
8.50
7.08
7.06
4.09
8.86
7.48
4.11
9.00
4.74
pw
5 BAR
30-61
6.30
3.54
5.78
6.40
4.15
5.46
5.27
3.38
4.77
3.14
2.68
8.40
7.04
7.08
5.82
6.34
3.47
8.33
6.58
2.92
8.00
4.24
61-91 cm
5.79
3.54
6.38
6.29
4.00
5.36
5.44
3.15
4.74
2.93
n.d.
9.52
7.57
7.25
7.04
5.58
4.42
n.d.
6.29
2.66
6.94
n.d.
81
-------
TABLE 18. CONTINUED
Site
BF
BL
CAO
DW2
GVC
HV
NE13
NEGV
SCI
SC2
SV2
UCC
BP
COO
DW1
DW3
HB
CP
S22M
SCR
SF
TUN2
0-30
7.86
5.08
5.27
6.58
3.31
5.66
6.58
3.94
4.56
5.13
4.36
4.78
6.95
6.97
5.90
6.20
3.45
7.42
6.00
3.19
7.46
3.81
pw
15 BAR
30-61
5.84
3.32
5.25
6.16
3.71
5.15
4.87
3.25
4.64
3.14
2.37
7,91
6.47
6.44
5.23
5.80
2.95
7.84
5.99
2.44
7.18
3.93
61-91 cm
5.43
3.25
5.95
6.15
3.62
5.18
5.15
3.06
4.51
2.71
n.d.
9.09
8.16
8.64
6.92
5.06
4.07
n.d.
5.54
2.33
6.20
n.d.
82
-------
TABLE 19. SOIL WATER CONTENT AT VARIOUS MATRIC SUCTIONS FOR PROFILE DEPTHS
OF 200 CM OR GREATER.
Site
BP
COO
SCR
SF
BL
DW2
GVC
NE13
UCC
1/10 BAR
200 cm
12.33
10.73
10.09
22.94
10.50
17.69
11.45
15.56
17.06
1/3 BAR
200 cm
7.62
4.14
6.22
18.62
6.45
12.92
7,55
9.37
11.40
1 BAR
200 cm
6.42
2.97
4.36
12.43
3.86
8.85
4.89
6.50
9.23
5 BAR
200 cm
4.70
2.19
2.75
7.21
1.80
4.98
3.15
3.73
6.30
15 BAR
200 cm
4.23
1.87
2.49
6.65
1.63
4.71
2.84
3.50
5.89
83
-------
TABLE 20. AVAILABLE SOIL WATER AND CLAY CONTENT AS PERCENT OF THE WHOLE
SOIL FOR THREE SOIL DEPTHS.
Site
BF
BL
CAO
DW2
GVC
HV
NE13
NEGV
SCI
SC2
SV2
UCC
BP
COO
DW1
DW3
HB
CP
S22M
SCR
SF
TUN2
0-30
Pw
23.87
13.23
15.52
15.90
10.56
11.06
18.05
9.28
7.17
9.09
12.20
15.66
19.92
15.33
20.39
17.84
12.34
20.31
20.15
7.35
21.94
15.30
cm
% Clay
9.0
4.9
9.9
5.4
4.9
9.1
7.4
4.4
5.4
5.6
6.8
4.9
7.9
6.3
6.3
8.9
5.0
7.9
7.9
3.0
10.3
5.2
30-61
Pw
21.04
12.46
15.20
16.06
11.17
7.74
13.72
9.60
7.56
10.31
6. ,63
13.13
18.53
11.96
13.57
16.40
8.95
16.66
14.31
8.46
18.72
13.27
cm
% Clay
7.8
3.6
10.6
4.1
2.3
9.6
6.9
4.3
3.9
4.4
4.0
12.7
6.4
5.0
6.7
8.9
4.7
7.9
7.1
3.2
9.5
5.6
61-91 cm
Pw % Clay
17.59
8.98
13.49
11.71
8.97
7.87
12.14
8.04
9.34
8.46
n.d.
10.98
13.50
8.36
12.10
12.60
8.29
n.d.
13.09
8.17
15.01
n.d.
7.6
4.6
13.0
6.3
2.7
7.9
6.3
2.1
2.8
2.9
n.d.
14.3
6.2
8.6
11.2
8.4
9.4
n.d.
7.2
3.3
7.0
n.d.
84
-------
20
15.0 [-
10.0
5.0
10
DC
<
£ 2.0
O 1-5
en
1.0
0.8
0.6
6 0.4
CO
0.2
0,15
O.I
15 CM
I
5 10 15 20 25
SOIL WATER CONTENT (Pw) % BY WEIGHT
Figure 23. Moisture retention curves for Dogwood Plot Site 2.
85
-------
STAND TREE MORTALITY SUBSYSTEM—BARK BEETLE POPULATION DYNAMICS
Introduction
Bark beetles in the genus Dendroctonus attack and kill conifers.
Except in epidemics (the last large outbreak in California was in 1962-
63) these beetles attack and kill trees that have been weakened or stressed
physiologically. Some factors that predispose trees to attack are drought,
flooding, lightning strikes, mechanical injury due to logging, building or
road construction, and root disease. Photochemical oxidants likewise
predispose trees to attack by bark beetles and this has been reviewed by
Cobb ^t_ al. (1968).
Three species of bark beetle, Dendroctonus brevicomis Le Conte, D_.
ponderosae Hopkins, and _D. .leffreyi Hopkins, attack and kill pines in the
San Bernardino National Forest (SBNF). White fir engraver, Scolytus ven-
tralis Le Conte, Ips emarginatus and Melanophila California Van Dyke are
also important species in the pest complex in SBNF. Pines were chosen for
this study because ponderosa and Jeffrey pines are more susceptible to
oxidant air pollutants than other major tree species in the forest. A major
part of the study was focused on the Dendroctnous genera of insects because
they appeared to be more important in the SBNF than other genera. The
California flathead borer, M. California Van Dyke, on Jeffrey pine appears
to be of little importance but it may be recognized as a significant part of
the pest complex on ponderosa pine when the data are analyzed. Ips emar-
ginatus is found in the pest complex on Jeffrey pine 22% of the time (Fig. 1
in McBride, Dahlsten and Cobb, in Miller 1977) and could be considered
important. The three Dendroctonus species alluded to above were the only
bark beetles studied and by far the greatest effort was on the western pine
beetle, D_. brevicomis. An earlier study (Stark et_ al., 1968) on the SBNF
showed that as the severity of oxidant damage to ponderosa pine increased,
the incidence of western pine beetle and mountain pine beetle, D_. ponderosae
infestations increased. Further substantiation of the interaction between
oxidant damage to pines and the incidence of bark beetles was gathered by
Wood (1971). A historical analysis on the Lake Arrowhead District of the
SBNF showed substantial increases in bark beetle caused tree mortality since
1951. These data were obtained by examining the beetle control records.
Research Objectives
The general objectives of the study were to determine susceptibility of
predisposed trees to bark beetles and the nature of the interrelationship
between oxidant damaged trees and beetle populations. Specifically the
objectives were as follows:
1) To determine the degree of susceptibility of oxidant-injured
86
-------
ponderosa pine to the western pine beetle and the mountain pine beetle
and of Jeffrey pine to the Jeffrey pine beetle and the California flat-
headed borer.
2) To investigate the influence of oxidant-injured pine trees on
the success and productivity of broods of the four beetle species listed
in objective #1.
3) To study the direct and indirect influence of photochemical
oxidants on the biology of bark beetles, with particular emphasis on the
insect associates, parasitoids, and predators.
4) To develop life tables for bark beetles by oxidant injury categor-
ies and, based on these tables, to develop predictive models of beetle
activity with reference to stand type and pine oxidant-injury level.
5) To determine the biological impact and relative importance of
each of the beetle species in forest communities and what influence they
have on stand change and forest succession.
Methods and Materials
Western Pine Beetle—
Field sampling procedures and laboratory analyses have been described
in detail for the western pine beetle (Dahlsten, 1974 and Dahlsten, 1977)
and are summarized in, Figure 24. Basically each beetle generation was
sampled twice or three times if it was an overwintering generation. Four
different procedures were used so that each developmental stage of the
western pine beetle and the insect associates, parasitoids, and predators
could be accounted for. The type of information taken is shown in Figure 24
as well as in Tables 21 through 33.
Mountain Pine Beetle and Jeffrey Pine Beetle—
Procedures for the development of an optimum sampling design for
the mountain pine beetle on ponderosa pine and the Jeffrey pine beetle
on Jeffrey pine have been described previously (Dahlsten, 1974 and Dahlsten,
1977). A nested sampling design was used for paired samples taken at
the lower, mid and upper portions of six beetle infested Jeffrey pines
and five infested ponderosa pines. The optimum design is one which for
a fixed variance yields the lowest cost or for a fixed cost yields the
lowest variance. Four sample sizes were used and several population
attributes were measured, three are indicated on the figures and tables.
Variances were estimated for each sample size for each variable
within heights, within trees, and between trees using the following formu-
las:
87
-------
00
00
FIELD SAMPLE
2 egg discs/odd ht [—
2 lorvol discs/ht [•
LAB ANALYSIS
TREE
TREE
TBUG
SAMPLE
TREE
2 larval discs/ht.
2 emergence cartons | — '
TBUG
STREE
Egg Disc Dissection
^attacks
#eggs
gallery length
early larvae
X-ray
Rearing
X-ray
Rearing
parent adults
early larvae
predators
associates
brood adults
parasites
predators
associates
teneral adults
pupae
late larvae
parasites
brood adults
parasites
predators
associates
Sticky Cartons
brood adults
parasites
predators
associates
Figure 24. Graphic summary of the population sampling procedures used for the western
pine beetle showing datasets and the type of information included for the
San Bernardino study.
-------
TABLE 21. HEIGHT, DIAMETER, AND LENGTH OF INFESTATION FOR WESTERN PINE BEETLE SAMPLE TREES SAN
BERNARDINO NATIONAL FOREST, 1973-1976.
oo
VO
Tree
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Yr.-Gen.
73-1
73-1
73-1
73-1
73-1
73-1
73-1
73-1
73-1
73-1
73-1
73-1
73-2
73-2
Tree
Ht.
(m)
18.3
22.9
19.2
28.0
29.0
25.9
43.9
29.9
13.7
19.5
20.7
17.1
18.0
16.5
DBH
(cm)
34.4
44.9
41.7
50.9
—
34.4
87.9
52.5
48.4
36.0
31.5
28.0
29.6
23.2
Top infest.
Ht. Diam.
(m) (cm)
16.5
13.5
13.5
15.5
19.5
13.5
34.5
22.0
12.0
12.0
15.0
12.0
10.8
11.5
15.9
29.6
23.2
32.8
31.8
23.9
23.9
15.9
1.0
22.6
18.1
16.9
17.2
15.0
Bottom infest.
Ht. Diam.
(m) (cm)
1.5
1.5
.5
.5
3.0
.5
.6
.1
.4
.5
.2
.6
.4
.3
34.4
44.9
45.8
57.3
-^
33,7
90.7
62.4
53.8
39,5
34.4
29.9
32.2
27.7
Lowest
sample
On)
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1,
5
5
5
5
5
5
5
5
5
5
5
5
5
5
Highest
sample
(m)
15.
13.
13.
15.
19.
12.
34.
21.
10.
12.
15.
12.
10.
10.
0
5
5
0
5
0
5
0
5
0
0
0
5
5
(continued)
-------
TABLE 21. (continued)
Tree
No.
15
16
17
18
19
20
21
22
23
24
525
526
527
528
Yr.-Gen.
73-2
73-2
73-2
73-2
73-2
73-2
73-2
73-2
73-2
73-2
74-1
74-1
74-1
74-1
Tree
Ht.
(m)
24.0
17.7
27.5
18.7
34.0
23.0.
22.0
25.0
2Q.Q
25.5
22.8
16.0
—
30.0
DBH
(cm)
53.8
26.4
56.0
28.0
76.4
30.6
33.4
46.2
26.4
36.3
69. .7
39 ..2
73.2
___
Top
Ht.
Cm)
16.8
12.0
22.7
10.2
28. Q
13.1
15.0
18.2
10.0
14.2
15.0
10.9
27.0
24.0
infest.
Diam,
(cm)
23.9
12.4
25.8
21.0
27,1
19,7
22.3
19.7
19.. 1
21.7
51.6
24.2
27.1
45.8
Bottom
Ht.
(m)
.5
.1
.2
.1
.2
.4
,3
3.0
.4
.5
.5
0.0
0.0
—
infest .
Diam,
(cm)
60 .,8
38.2
60.5
34,7
76.4
32,2
43.3
58.9
32.5
39.5
75.4
43.6
87.5
— .,
Lowes t
sample
Cm)
1,5
1.5
1.5
1,5
1.5
1,5
1.5
1.5
1.5
1,5
1.5
1,5
1.5
1,5
Highest
sample
(m)
15.0
12.0
22.5
9.0
19.5
12.0
9.0
16,5
9,0
7.5
10.5
10.5
27.0
24,0
(continued)
-------
TABLE 21. (.continued)
Tree
No.
529
530
531
532
533
534
vo
In***
535
536
537
538
539
540
541
542
Yr.-Gen.
74-1
74-1
74-1
74-1
74-1
74-1
74-1
74-1
74-1
74-2
74-2
74-2
74-2
74-2
Tree
Ht.
(m)
16.2
25,0
22.0
30.2
14.4
19.2
28.4
25.0
26.0
13.7
12.8
13.7
28.9
16.3
DBH
(cm)
30,6
60.5
94.5
67.8
71.6
51.6
96.1
48.4
50. Q
30.2
25.2
53.5
69.1
79.3
Top
Ht.
Cm)
13.0
25.0
21.0
23.6
14.4
16.8
16.5
19.8
22.0
10.0
6.0
11.0
22.5
13.8
infest.
Diam.
(cm)
15.9
—
—
29.3
55.7
21.0
71.0
19.1
16.6
21.7
19.4
39.2
36.9
43.9
Bottom
Ht.
(m)
.4
0.0
—
.1
.7
0.0
.3
.9
.8
0.0
.2
0.0
.4
0.0
infest.
Diatn.
(cm)
35.0
65.3
93.0
82.4
74.8
64.0
110.8
50.0
49.3
33.7
29.9
57.9
78.3
88.5
Lowes t
sample
(m)
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
Highest
sample
(m)
12.0
18.0
21.0
22.5
13.5
16.5
16.5
19.5
21.0
9.0
6.0
10.5
19.5
13.5
(continued)
-------
TABLE 21.(continued)
l-o
Tree
No.
543
544
546
547
548
549
550
551
552
553
554
555
556
557
Yr . -Gen .
74-2
74-2
74-2
74-2
74-2
74-2
74-2
74-2
75-1
75-1
75-1
75-1
75-1
75-1
Tree
Ht.
(m)
26.8
22.9
—
20.0
25.0
22.6
19.0
27.0
18.0
21.3
15.6
20.1
21.9
12.5
DBH
(cm)
90.7
47.1
40.4
38.2
43.6
36.3
54.8 ,
63.0
42.0
45.8
39.5
65.3
52.8
32.8
Top infest.
Ht. Diam.
(m) (cm)
18.0
18.5
16.0
19.0
18.3
17.0
12.3
15.0
10.8
15.2
11.2
14.5
12.0
9.0
63.7
23.9
18.1
18.5
25.2
17.2
30.2
41.7
25.8
21.0
21.7
34.7
24.2
17.8
Bottom infest.
Ht. Diam.
(m) (cm)
.3
0.0
1.5
0.0
.5
0.0
2.0
.3
.1
1.5
.3
3.0
.1
•5
95.8
57.6
38.5
46.2
50.0
43.6
53.8
68.1
44.6
45.8
47.8
61.1
65.9
38.2
Lowest
sample
(m)
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
3.0
1.5
1.5
Highest
sample
(m)
18.0
18.0
15.0
18.0
18.0
16.5
12.0
13.5
10.5
15.0
10.5
13.5
12.0
9.0
(continued)
-------
TABLE 21. (continued)
Tree
No.
558
559
560
561
562
563
564
565
566
567
568
569
570
571
Yr.-Gen.
75-2
75-2
75-2
75-2
75-2
75-2
76-1
76-1
76-1
76-1
76-1
76-1
76-2
76-2
Tree
Ht.
On)
10.1
20.4
25.6
29.3
33.8
—
31.7
28.0
32.6
16.2
29.9
38.4
25.6
20.7
DBH
(cm)
48.4
36.9
67.5
62.1
75.4
—
74,5
37.6
111.7
38.8
52.5
95.2
71.0
44.9
Top infest.
Ht. Diam.
(m) (cm)
8.8
15.6
15.0
20.5
18.5
—
27.5
19.0
16.5
11.6
16.9
16.2
16.5
17.2
24.5
17.2
33.1
25.8
53.2
—
36.0
25.2
79.9
18.1
30.6
66.8
48.1
22.6
Bottom infest.
Ht. Diam.
(m) (cm)
0.0
Q.O
0.0
0.0
Q.Q
—
1.0
.4
.3
.1
0.0
0,0
.2
.1
54.1
44.2
77.7
63.7
84.7
—
—
42.0
121.0
40.4
65.9
111.4
78.3
54.8
Lowes t
sample
(m)
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
7.5
1.5
1.5
Highest
sample
(m)
7.5
15.0
15.0
19.5
18.0
19.5
27.0
18.0
16.5
10.5
16.5
15.0
16.5
16.5
(continued)
-------
TABLE 21. (continued)
Tree
No.
572
573
574
575
Yr . -Gen .
76-2
76-2
76-2
76-2
Tree
Ht.
(m)
18
32
18
27
.9
.3
.6
.7
DBH
(cm)
41.4
83.1
92.0
77.4
Top infest.
Ht. Diam.
(m) (cm)
11.2
23.1
13.5
24.8
22.9
49.0
71.9
25.8
Bottom infest. Lowest
Ht. Diam. sample
(m) (cm) (m)
1.2
3.0
.5
1.3
43
80
96
78
.6 1.5
.5 6.0
.8 1.5
.6 1.5
Highest
sample
(m)
10
22
13
24
.5
.5
.5
.0
\o
-------
TABLE 22. WESTERN PINE BEETLE INFESTED PONDEROSA PINES THAT WERE SAMPLED
BETWEEN 1973 AND 1976 RANKED BY OXIDANT DAMAGE.
D . brevicomis
generation
1973-1
1973-2
1974-1
1974-2
1975-1
1975-2
1976-1
1976-2
Years
combined
Generations
combined
Damage class
1-5 6-10
3
1 3
3 4
2
2
2
3
1 7
3 12
4 19
11-15
4
5
3
2
7
7
14
16-20
2
3
1
2
1
1
4
6
10
21-25
2
3
2
1
8
0
8
26-30
5
2
1
1
6
3
9
31-35 36+
2 1
2
1 1
1
1
2 2
2 3
4 5
95
-------
TABLE 23. WESTERN PINE BEETLE MEAN EGG DISSECTION VARIABLES BY YEAR AND
GENERATION FOR WHOLE PONDEROSA PINES, SAN BERNARDINO NATIONAL
FOREST, 1973-1976.
••"
Attacks
per dm^
Gallery
length
cm/dm^
Total eggs
per dm^
1st ins tar
larvae
per dm^
Eggs per
cm. of
gallery
Average
tree smog
rating
Year
Gen.
1
2
1
2
1
2
1
2
1
2
1
2
1973
2.
0.
79.
41.
71.
41.
59.
30.
0.
0.
28.
17.
88
93
3
7
1
2
4
4
90
97
5
3
1974
2,27
0.99
71.8
36.5
54.7
47.4
45.7
34.2
0.76
1.31
14.6
11.2
1975
2.89
1.48
58.7
46.5
70.9
63.8
53.5
51.8
1.24
1.35
13.8
19.3
1976
1.32
1.04
60.2
39.3
86.8
64.8
68.9
49.5
1.41
1.56
19.8
17.7
4 yr Sierra Nevada Miller
mean Blodgett Forest & Keen
1967-1970 mean 1960
2.
1.
67.
41.
70.
54.
56.
41.
1.
1.
19.
16.
34
20
5
0
9
3
9
5
08
30
2
4
1
1
52
45
74
78
47
56
1
1
.61
.56
.8
.8
.5
.8
.3
.3
.42
.69
1.25
69.8
1.65
96
-------
TABLE 24. VARIABLES CALCULATED FROM WESTERN PINE BEETLE EGG DISC SAMPLE
DISSECTION BY GENERATION FROM 1973 TO 1976. SAN BERNARDINO
NATIONAL FOREST.
•
Eggs per attack
Gallery length
per attach
1st instar lar-
vae per attack
Average tree
smog rating
Gen.
1
2
1
2
1
2
1
2
1973
25.0
45.4
28.2
48.3
20.9
32,7
28.5
17.3
1974
28.3
66.0
36.1
51.6
23.4
47.6
14.6
11.2
1975
47.8
56.7
38.2
41.0
33.5
44.8
13.8
19.3
1976
70.8
75.8
49.2
47.0
57.0
57.8
19.1
17.7
4 Year
Mean
43.0
61.0
37.9
47.0
33.7
45.7
19.2
16.4
TABLE 25. CORRELATION OF WESTERN PINE EGG DISC DISSECTION VARIABLES
WITH YEAR, GENERATION AND TREE OXIDANT RATINGS SAN BERNARDINO
NATIONAL FOREST, 1973-1976.
Attacks/dm2
Gallery length cm/dm^
o
Eggs, dm''
1st instar larvae/dm2
Eggs/cm, gallery L.
Gallery length/attack
Eggs/attack
1st larvae/attack
Generation
(-) -05
(-) .01
(-) NS
(-) -01
.01
.05
.05
.05
Year
NS
NS
.05
NS
.05
NS
.05
.05
Tree Smog Rating
.05
.05
.01
.05
NS
(-) .05
NS
NS
Notes: NS = Not significant at 5% level or better.
(-) = Negative correlation.
97
-------
TABLE 26. MEANS BY GENERATIONS ON THE LAST SAMPLE DATE FOR WESTERN PINE
BEETLE BROOD, PARASITES AND PREDATORS FROM X-RAY ANALYSIS OF
SAMPLE BARK DISCS. SAN BERNARDINO NATIONAL FOREST, 1973-1976.
Sierra Nevada
San Bermardino National Forest Blodgett Forest
Total D.b.
brood/dm^
Total para-
sites
Total preda-
tor s/dm^
Mean tree smog
rating
Year 1973
Gen.
1 15.1
2 2.18
1 0.82
2 0.30
1 3.36
2 0.74
1 28.5
2 15.4
1974 1975
16.8 26.5
4.52 3.57
0.59 0.51
0.59 0.34
3.07 1.13
1.25 0.57
14.0 13.8
11.2 19.3
1976
12.4
7.42
0.18
0.16
1.17
0.37
19.8
17.7
Mean 1966-1970 Mean
17.7
4.42
0.52
0.35
2.18
0.73
19.0
16.9
13.6
6.95
0.452
0.613
1.59
0.434
—
98
-------
TABLE 27. SIGNIFICANCE OF MULTIPLE REGRESSION COEFFICIENTS FOR THE LAST
SAMPLE DATE OF X-RAY, REARING, AND STICKY CARTONS FOR THE WESTERN
PINE BEETLE AND ITS NATURAL ENEMIES, SAN BERNARDINO NATIONAL
FOREST, 1973-19761.
Gen.
Smog
Year rating D.b. Pred. Para.
Live D. brevicomis
Predators
Parasites
iFor D.b.,
X-ray
Rear
Stik
X-ray
Rear
Stik
X-ray
Rear
Stik
the variables
.001
.001
.01
.001
.001
NS
.05
NS
NS
of
.05
NS
NS
.001
.05
.01
NS
NS
NS
generation,
.05 NS NS
.05 NS NS
.05 .001 .05
NS NS .001
NS .01 NS
NS .001 NS
NS .05 .001
NS .01 NS
NS .01 NS
year, and smog rating were analyzi
together, without the effect of predators and parasites. The predators and
parasites variables were then added to the equation to test their signifi-
cance.
99
-------
TABLE 28. MEAN WESTERN PINE BEETLE AND NATURAL ENEMY EMERGENCE BY GENERA-
TION FOR THE LAST SAMPLE DATE OF LABORATORY REARED DISCS, SAN
BERNARDINO NATONAL FOREST, 1973 TO THE FIRST GENERATION OF
1975.
Western pine
beetle per dm^
Predators
per dm
Parasites
per dm
Mean tree
smog rating
TABLE 29. MEAN
TION
SAN
1976
Western pine
beetle per dm^
Predators
per dm
Parasites
per dm
Mean tree
smog rating
Year
Gen.
1
2
1
2
1
2
1
2
- _ , ,. — — ________ — -•
1973 1974 1975
S
8.00 4.82 6.47
1.19 1.00
1.85 1.67 0.523
0.315 0.237
0.697 0.301 0.400
0.318 0.235
28.5 14.9 13.8
15.2 11.2
.B.N.F.
Mean
6.43
1.10
1.35
0.276
0.466
0.277
18.7
13.2
Blodgett
Forest
1966-1970
7.12
2.99
2.88
0.177
0.298
0.630
—
—
WESTERN PINE BEETLE AND NATURAL ENEMY EMERGENCE BY GENERA-
FOR THE LAST SAMPLE DATE OF STICKY CARTONS (FIELD RESEARCH),
BERNARDINO NATIONAL FOREST, 1973 TO THE FOREST GENERATION OF
*
Year
Gen.
1
2
1
2
1
2
1
2
1973 1974 1975 1976
3.86 3.30 1.77 1.60
0.810 0.297 0.763
2.11 1.50 0.746 0.759
0.91 0.459 0.583
2,20 1.33 1.75 0.146
0.359 0.343 0.358
30.4 14.0 13.8 19.8
15.4 11.2 19.3
S. B.N.F.
Mean
2.63
0.623
1.28
0.65
1.36
0.353
19.5
15.3
Blodgett
1966-1970
4.95
1.45
0.947
0.733
1.56
0.639
—
^ ,_
100
-------
TABLE 30 RESULTS OF USING TANG'S CALCULATION OF JEFFREY PINE BEETLE
DATA: TREE EFFECT*
Quantity
Measured
Gallery
Length
Density
Attack
Density
TABLE 31.
Quantity
Measured
Gallery
Length
Density
Attack
Density
Sample Calcu- 50
Size lated 6 trees
100cm2 1.02 1.7
100cm2 1.14 1.7
250cm2 1.52 1.7
500cm2 1.91 1.7
1000cm2 1.82 1.7
100cm2 .87 1.7
100cm2 1.28 1.7
250cm2 1.78 1.7
500cm2 1.78 1.7
1000cm2 2.47 1.7
50
18 trees
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
RESULTS OF USING TANG'S CALCULATIONS ON
DATA: HEIGHT EFFECT.
Sample
Size Calculated
100cm2 .484
100cm2 .609
250cm2 .611
500cm2 .442
1000cm2 .257
100cm2 .272
100cm2 .272
250cm2 .365
500cm2 .424
1000cm2 .474
* 50
3 heights
1.4
1.4
1.4
1.4
1.4
1.4
1.4
1.4
1.4
1.4
§ 50 Avg= Conf.
3 trees effect found
2.0 .177
2.0 .166
2.0 .181
2.0 .170
2.0 .128
2.0 .0018
2.0 .0026
2.0 .0018
2.0 .0016
2.0 .0016
JEFFREY PINE BEETLE
50 Conf.
6 heights found
1.15 .290
1.15 .42
1.15 .42
1.15 .25
1.15 .093
1.15 .10
1.15 .87
1.15 .18
1.15 .24
1.15 .28
67
41
95
99
99
50
87
98
99
99
101
-------
TABLE 32. RESULTS OF USING VARIOUS TESTS ON ATTACK DENSITY, SAMPLE SIZE OF
DATA, TO DETERMINE DIFFERENCES BETWEEN PAIRS OF TREES. VALUES IN
ATTACKS/CM2.
Trees
compared
Mean differ
with
confidence
For
mean
dif.
Actual
mean
dif.
Rel.
that
fer
likelihood
mean dif-
by 0.25
Means
differ by
at least
compared by
1.2
1-3
1.4
1.5
1.6
.17
.98
.83
.99
.95
.0033
.0033
.0035
.0029
.00339
.0003
.0040
.0023
.0043
.00333
D
0.69
13.0
2.3
89.5
0.81
2D
.47
.71
,55
.94
.21
.0068
.0014
D is the measured difference in means.
TABLE 33. PRELIMINARY SUMMARY OF FINAL SMOG DAMAGE RATINGS FOR PINES
KILLED BY INSECTS ON ESTABLISHED VEGETATION PLOTS, 1973-1975.
Tree
species"*"
PP
PP
PP
.
PP
JP
JP
Insect
species^
D.b.
D.p.
Mixed (D.b.
broods + D.p . )
Ips & M.c.
(combined)
D.j.
Mixed (D.j.
broods + Ips)
Number
of
trees
17
7
8
5
7
4
*
Oxidant injury score
Mean
9.9
10.4
11.8
15.2
11.6
13.0
SD SE
6.3 1,5
6.3 2.4
8.6 3.0
10.0 4.5
6.7 2.5
8.2 4.1
Range
1-21
6-25
1-30
2-32
3-19
4-23
*A11 scores given by P. Miller except those from 1973.
+pp = ponderosa pine; JP = Jeffrey pine.
~T).b. = Dendroctonus brevicomis; D.p. = D. ponderosa; D.j
jeffreyi; Ips = Ips sp; M.c. = Melanophila californica.
= D.
102
-------
Within levels
632
- 7i>
.1-1 i=l
6-3-2-1
Within trees:
6 3 - -2
0-2 . * 2<7jk - 7k>
WT k=l .1 = 1
6-3-1
Between trees:
6 - 3
0-2 = j (7k - y>
BT k=1
in which
value °f variable for itn bark sample, jtlri level, kth tree
= mean of 2 bark samples at jtn level, k*-" tree
mean of 6 bark samples for k tree
y = mean overall 6 trees
We assume, for a new population to be sampled, if the variances were
similar to those of this population, that the variance of an estimate
would be
2 2 2 2
BT __WT WL
n 3'n 3'm-n
103
-------
in which
n = number of trees
m = number of samples per each of the 3 levels
Costs of sampling were estimated by the following formula
C = $120 + $123 x n + Cinm
in which C^ is the cost of cutting and measuring a bark sample of size
i, $123 is the additional cost of measuring each tree and $120 is the one
time set-up cost of measuring each tree and $120 is the one time set-up
cost of sampling ; These costs were derived from tallies of the time spent
doing the sampling tasks, CIQQ = $2^17, ^25Q = $2>77> C500 = $4.07
and CIQOO = $6.10.
Using the above formulas sampling was simulated for various combina-
tions of numbers of trees, samples per tree and sample sizes. For each
combination the corresponding cost was also calculated.
Application of the Tang procedure — During the initial phase of the
analysis of the mountain pine beetle and Jeffrey pine data several types
of analyses were used. One procedure was found to be particularly valuable
and some calculations using data on the Jeffrey pine beetle were used
as an illustration of the application of the Tang procedure.
The Tang procedure supplies the probability that an analysis of variance
test will find significance for a certain test given an estimate of the size
of the effect of the treatment, an estimate of the variance and the confi-
dence desired. By use of this procedure the probability that an analysis
of variance calculation would show significance given that the effect is
of a certain size can be determined. Thus, use of this test allows one
to conclude that if an analysis of variance computation does not show sig-
nificance, the actual effect was probably less than a given size.
The Tang procedure is as follows: the quantity (j) is calculated,
where (j)2 is given by:
,2 m2d2
$ i
kd2
d is the size of the ith effect
a2 is the variance
k is the number of treatments
m is the number of replications of each treatment.
Separate computations are made for each effect in an experiment in which
more than one effect is tested. m,k,d,
-------
effects.
For a given combination of degrees of freedom, confidence level, and
value of Q-2, the probabilities of finding significance can be found in
tables in statistical textbooks. The probabilities are tabulated for
= 2,2.5,2,2,5,3,4,5,6,7,8.
Vegetation Plot Surveys—
Each numbered tree on each of the 19 vegetation plots was examined
twice a year, in July and November. The cause of mortality, if due to
an insect, was recorded if a tree had died since the previous survey date.
Results and Discussion
Western Pine Beetle—
The size, extent of infestaton, and the location of the top and bottom
sample is summarized for each Ponderosa pine sampled from 1973 to 1976 in
Table 21. Most of the sample trees fall in the middle size categories of
those attacked since the large trees are too difficult to sample. Also,
small trees were not sampled as the western pine broods were often mixed with
Ips spp. All of the sample trees were smog rated using Miller's (1973)
scoring system. The distribution of sample trees by smog score is shown in
Table 22.
Evidence for the direct or indirect influence of photochemical oxidants
on the dynamics of western pine beetle populations is circumstantial.
However, it is interesting to compare the San Bernardino populations
with other western pine beetle populations. Where possible the various
population parameters are compared with data from Blodgett Experimental
Forest (Dahlsten et_ _§!_•, 1974) where trees are not damaged by photochemical
oxidants but are weakened by a root disease, Verticicladiella wagneri
(Dahlsten and Rowney, 1974)• Some comparisons can be made with the data
summarized by Miller and Keen (1960) but their information does not separate
out generation effects, which are considerable. The Miller and Keen
data was collected before smog damage was prevalent in southern California
and the data was collected at sites throughout the state.
Egg disk dissection variables—The first samples taken at the beginning
of each generation were the egg disks. The mean values by generation
for each of the variables sampled is given in Table 23 along with mean
values from Blodgett Forest (Dahlsten et^ al., 1974) and Miller and Keen
(1960). Evaluation on a tree by tree basis or by height was not done for
this analysis. The main objective was to find possible photochemical oxidant
effects on the different life stages of the beetle.
Generally the first generation of beetles attacking in a given year
emerge from the overwintering brood trees. These beetles set up what is
referred to in the text as the first generation. This generation is more
discrete in terms of development, it is the most successful and the most
abundant* It is conceivable, too, that the condition of the host trees
105
-------
in mid-June to July when the first generation begins may have considerable
influence on the success of the generation. The second generation begins
from mid-August to September; at this time of year it is dry and warm,
and the incidence of photochemical oxidants is at its peak. Stress on
the trees would be greatest at this time of year and the beetles may there-
fore be affected. Note that the mean tree smog rating was usually (3 of 4
years) lower (a lower score means more oxidant damage) in the second genera-
tion (Table 23).
Attack densities per square decimeter (dm2) were much higher in SBNF
in the first generation than those previously recorded at Blodgett or by
Miller and Keen, Table 23. The second generation was much lower and this
could well be the influence of photochemical oxidants on the trees, i.e., it
takes fewer beetles to kill the trees because trees are weakened at this
time of year. The low attack densities in 1976, generation one at SBNF
could have been due to drought conditions or to a declining beetle popula-
tion. Note the higher densities in 1975, generation two, Table 22. This,
too, could be related to moisture conditions. There was a close relation-
ship between the cm of gallery length per dm2 and attacks and the same
relationships hold, Table 23.
The relationship between attacks and total eggs per dm^ was con-
sistent except at Blodgett. However, there were small outbreaks at Blodgett
in 1967 and 1969 which could explain the jump in the number of eggs in
generation two. It appears that whenever an outbreak occurs it is preceded
by very high densities of generation two in the previous year.
Density of first instar larvae was as to be expected based on the
preceding discussion. Egg mortality can be calculated by the difference
between total eggs and first instar larvae. The percent egg mortality at
SBNF was 19.7 in generation one and 23.6 in generation two. At Blodgett it
was 36.5% in generation one and 28.5% for generation two. There is no
explanation for these differences except perhaps the influence of natural
enemies. This illustrates the importance of developing life tables which
are not yet available. Of particular interest is the large difference in
egg mortality between the first generation at SBNF and that at Blodgett.
The eggs per cm of gallery length is usually higher in generation
two. This is undoubtedly due to competition, and is an inverse relation-
ship between attacks and eggs per cm of gallery length. The calculated
variables in Table 24 further demonstrate intraspecific competition.
Since all trees were ranked by oxidant rating it was possible through
covariance analysis to analyze for smog as well as generation and year
effects, Table 25. From the data in Table 24 it appears that a direct or
perhaps indirect effect could be attributed to photochemical oxidants in
generation two. Looking at the tree by tree smog rating shows significant
differences between high scoring (not damaged) and low scoring (oxidant
damaged) trees for most variables. There is a negative correlation for
gallery length/attack and this is to be expected and is due to competition
as described above, i.e., the lower the smog score, the greater the gallery
length as the attack density is lower, Table 24. There are some significant
106
-------
relationships by year and almost all correlations are significant by genera-
tion as would be expected by looking at Tables 23 and 24.
X-ray analysis—The second sampling procedure is designed to analyze
the density of western pine beetle brood, its parasites and predators,
Figure 24. As with the egg variables, generation one is always higher
than generation two, Table 26. Generation one at San Bernardino was higher
than Blodgett but generation two was lower. The very low densities for
the second generation could well be an effect of smog directly or indirectly
as described above.
The parasite and predator densities are very low but tend to be higher
in SBNF than Blodgett except for the second generation parasites, Table
26. It is interesting and perhaps important that the highest brood density
in the second generation at SBNF occurred when both the predators and
parasites were the lowest.
Generation, year, and oxidant ratings are significantly different
for western pine beetle brood, Table 27. Generation was significant for
the predators and parasites but there is no influence of smog at least
indirectly as determined by tree oxidant ratings.
Rearing analysis—Each of the x-ray discs was placed into a carton
for rearing and the individuals emerging were identified and recorded.
Only the data from the last sample date was considered, as it corres-
ponded more closely to actual emergence in the field and could be compared
with the sticky carton samples. The rear data has only been analyzed
through the first generation of 1975, Table 28. The same trends are shown
with this sampling procedure, i.e., generation one emergence was higher for
western pine beetle brood, parasites and predators than generation two.
Significant effects of tree smog rating are shown only for western pine
beetle brood, Table 27. There were significant effects of generation
and year on predators but not on parasites.
A possible link between western pine beetle success and smog is the
low mean emergence of beetles in the first generation, Table 28. The
densities for western pine beetle were higher than Blodgett up to emer-
gence, and then fell below Blodgett. The SBNF second generation emer-
gence was extremely low. Predator emergence in generation one was lower than
Blodgett, while generation two was higher; the relationship between para-
sites was just the opposite. Both parasitoids and predators were more
numerous in the first generation, this was true for x-ray and for sticky
cartons.
Sticky carton analysis—Analysis of the sticky carton data is
complete through the first generation of 1976. Results from this sample
procedure show the same trends as the rearing data, Table 29- The differ-
ential effects of environment can be seen by comparing the rearing
samples with the sticky carton samples. The mean densities for western pine
beetle were always higher in the rearing samples than the sticky carton
samples. However, parasites and predators were more dense in the sticky
cartons except for the first generation predators. The relationships
107
-------
were exactly the same at Blodgett, Tables 28 and 29. This suggests that one
method should be used for the bark beetle and another for the natural
enemies.
Significant differences are shown for generation and smog rating,
but not for years for the western pine beetle. With the natural enemies
only year was significant for predators.
While there is only circumstantial evidence for the direct effect
of photochemical oxidants on the western pine beetle, there is good evidence
of an indirect effect on the brood from each of the four sampling proce-
dures. Smog damaged trees had a depressing effect on western pine beetle
populations. The implications of this in the forest community in San
Bernardino are that the probability of an outbreak is lessened since it
takes fewer beetles to kill diseased trees and fewer beetles are produced in
such a system. Also, since it takes more beetles to kill healthier trees,
the healthier trees should be relatively safe. The incidence of drought,
mechanical injury, root disease, and fire could drastically alter these
generalizations for a particular site. It remains to be seen what this
means in terms of management in a recreation forest such as the San Bernar-
dino. The only conclusion that can be made now is that attempts to control
the western pine beetle are not necessary as this is treating a symptom
rather than the cause of the problem.
Mountain Pine Beetle and Jeffrey Pine Beetle—
The associates, parasitoids and predators reared from trees infested
by mountain pine beetle or Jeffrey pine beetle have been listed previously
(Dahlsten, 1977). In addition Tables giving means by height in the trees
and sample size for number of attacks and gallery length for each beetle
species have been recorded earlier (Dahlsten, 1977) •
A series of analyses was made on several variables (attacks, gallery
length, total larvae, emergence, and total pupae) for each beetle species.
For all runs, the lowest cost for a fixed variance or the lowest variance
for a fixed cost occurred with the 500 or 1000 cm^ disk sizes. Figure
25 shows a typical relationship for cost of sampling for fixed variances
as a function of bark sample size and Figure 26 is similar except that
variance is shown as a function of bark sample size for fixed costs. The
feasibility of using the larger bark samples are dubious since handling
becomes difficult beyond a certain size and consequently more measurement
error may be introduced. The implications are, however, that the larger
bark sample sizes are most cost effective.
Variance as a function of cost for a fixed bark sample size for various
numbers of samples per height are shown in Figure 27. In nearly every
case the sampling of more than 2 units per level did not increase cost
effectiveness. Even in cases where 3 or more units per height were better
the results were nearly as good for 2 as for more sample units per level.
Application of Tang's procedure—Only two of the variables measured
for the Jeffrey pine beetle were used to test the application of Tang's
108
-------
7
O
Q
-------
UJ
o
o:
-------
8
7-
LJ
O
z
< 4
tr
8 10 12
COST (in Thousands)
14
16
18
Figure 27. Variance as a function of cost for different numbers of 1000 cm2 samples per
height for gallery length of the mountain pine beetle in ponderosa pine, San
Bernardino National Forest, 1974.
-------
procedure. The first set of calculations was performed to check the sig-
nificance of the tree effect on attack density, i.e., to find out if there
was significant variation in attack density between trees. For the given
conditions (6 trees, 3 heights, 2 measurements at each height which were
considered replicates), $50* the value of cj) needed to have a 50% proba-
bility of finding significance is 1.7. If 18 trees were used 4>50 would
be approximately 1.0, and for 3 trees about 2.0. The cj>56 are given
approximately because the values must be interpolated from graphs. For
sample size 1(100 cm2 disk) ^was .871. Hence using sample size 1 there
would be little chance of finding significance even if 18 trees were used.
On the other hand for sample size 5(1000 cm2 rectangle) cj) was 2.47. Thus
using sample size 5 there was an excellent chance of finding a significant
effect with 6 or even with 3 trees. The reason for the difference is that
sample size 1 is much smaller, so that the fluctuations are a bigger per-
centage of the actual measured values. Hence the ratio of the between group
to the within group variance is smaller, and more measurements are needed to
establish significance.
Generalizations can be made from this example to construct a procedure
for using Tang calculations in planning sampling schemes. Suppose there
is a desire to determine if there is an effect due to treatment of at
least a certain size (on the average). Then 5Q should be calculated
for several plausible experimental designs, using values of 02 estimated
from previous experiments. Actually it might be preferable to calculate
85 or 95' When the measurements are made or relevant calculations
performed the significance test will show (with appropriate confidence)
whether or not the effect of the treatment is greater than a predetermined
size.
Table 30 shows the results of calculations for attack density and
gallery length density for the tree effect. Table 31 presents results
on the same data for the height effect, i.e., the significance of being
above, below, or at the center of the infestation.
It can be easily seen from the tables that the chance of finding an
effect from a given number of trees generally increases with the sample
size. Moreover, in cases where the Tang calculation predicted a high
probability of finding significance, the confidence found was generally
high. The number in the "confidence found" column is 1 minus the pro-
bability of obtaining at random the F value actually found.
Some additional tests were performed on this data. For selected
pairs of trees the t test was used to determine the significance of the
differences in attack density between trees. The t test is a special
case of the analysis of variance test in which there are only two treat-
ments, e.g., one degree of freedom in the numerator. The calculations
below are presented merely to illustrate the techniques on several pairs
of data sets. Multiple comparisons should not be made with a test on
several samples. This is because if there are many samples it is likely
that some pairs will have what appear to be significant differences, even if
there is no effect of treatment. In such cases an analysis of variance or
multiple comparison test should be used.
112
-------
In addition to the basic t test the following calculations were per-
formed to add information to that supplied by the t test:
1. Difference in population means that would be just large enough
to give a significant difference (95% confidence level) was computed and
compared to the actual difference in means.
2. If the means differed significantly, then a minimum difference
value, M was calculated. It can be said with confidence (95% confidence
level) that the means differ by at least M.
3« A calculation was made to find the ratio of the likelihood that,
given the observed results, the observed means differ by the measured mean,
to the likelihood that the means were the same. In addition, the ratio
of the likelihood that the observed means differ by twice the measured
mean to the likelihood that they differ by 9 was computed. In the formula
for t(equation 1) for the difference of the two populations the appropriate
difference was substituted and the differential probablity computed from
equation 2, the probability density function for the t variable.
(1) t =*
X-Y
s 2
n
X is the mean for the first population
Y is the mean for the second population
s is the estimate of the variance
n is the number of degrees of freedom in each population
(2) f(u) = const. (1 + u2) = -|f + 1
This test can provide a lot more information additional to that in
a significance test because it can provide confidence levels on the differ-
ences in the means. It can give the relative likelihood that the popula-
tion means actually differ by any combination of values. If it is found
that the means are about as likely to differ by 0 as by the measured means,
then the two populations are likely to be the same. If, on the other
hand, it is twenty times as likely that the populations differ by the mea-
sured mean as by 0, it can be concluded that they are significantly different
and also have an idea of the size of the difference.
Table 32 shows some of the results of these calculations. These are
for comparisons of various trees for attack density using sample size
4(500 cm^ rectangle).
113
-------
Vegetation Plot Surveys—
The results of the vegetation plot surveys through 1975 (Table 33)
show that the western pine beetle and the Jeffrey pine beetle were the two
most important species in terms of pine mortality. A more complete analysis
of the cause of tree mortality on the San Bernardino National Forest is
currently in progress (McBride et^ al., 1977). There is no indication as yet
that the smaller vegetation plot survey is representative of the mortality
occurring on the San Bernardino National Forest.
114
-------
EFFECTS OF PHOTOCHEMICAL AIR POLLUTION ON THE EPIDEMIOLOGY OF FOREST TREE
PATHOGENS
Introduction
Since the last report, several studies that were in progress have been
completed and others initiated during the period 1976-1977 have also been
completed. The results of these studies are summarized as follows.
Susceptibility of Roots of Mature Trees—
All live tree root inoculations have been completed. Data (Table
34) for ponderosa pine indicate differences in susceptibility to infection
among trees from various air pollution damage categories. Roots from
trees showing very severe to severe damage were much more susceptible
to proximal infection by JF. annosus than roots from trees with slight
air pollution injury. The term proximal infection is used to indicate that
the portion of root between the point of inoculation and the main stem of
the tree was invaded by the organism. Distal infection or colonization
refers to fungal invasion toward the root tip from point of inoculation.
When distal and proximal infection are considered together, no clear
differnces among air pollution damage classes emerge. Results for Jeffrey
pine do not provide enough basis for air pollution damage comparisons because
of the absence of severely injured trees in the sample.
Colonization results for inoculated roots are presented in Table 35.
For ponderosa pine, data indicate definite differences in proximal coloniza-
tion rate in roots from trees showing various levels of air pollution
damage. The trend is like that described for infection; i.e., roots from
trees severely damaged were more susceptible to colonization by the pathogen.
Distal colonization rates show no distinct air pollution related trends.
Regression lines for comparisons between oxidant injury scores and
rates of proximal colonization are shown in Figure 28. For both ponderosa
pine inoculation trials, relationships between these two variables were
statistically significant (P = 0.01).
Greater infection and more rapid colonization of roots by _F. annosus
in severely injured trees could have significant effects on disease epi-
demiology in affected stands. In such stands, greater tree mortality in
shorter time periods could be expected.
Stump Inoculation Studies—
Studies described in the previous progress report were completed.
115
-------
TABLE 34. INFECTION OF INOCULATED PINE ROOTS WITH FOMES ANNOSUS IN RELATION
TO THE SEVERITY OF AIR POLLUTION INJURY.
Air Pollution No. Roots Root Infection
Site Species3 Damage Inoc. A" % Bc %
Snow Valley JP
Holcomb Valley JP
Breezy Point PP
Camp Paivika PP
Moderate
Slight-No Injury
Slight-No Injury
Very Severe-Severe
Moderate
Slight-Very Slight
Very Severe-Severe
Moderate
Slight
12
12
19
16
8
16
22
17
10
5
7
10
7
5
7
15
6
5
41.7
58.3
52.6
43.7
62.5
43.7
68.2
35.3
50.0
0
4
3
5
0
1
15
5
1
0
33.3
15.8
31.3
0
6.3
68.2
29.4
10.0
aJP = Jeffrey pine; PP = ponderosa pine
°No. of roots showing proximal and/or distal infection from the point of
inoculation
cNo. of roots showing proximal infection from the point of inoculation
116
-------
TABLE 35. COLONIZATION OF INOCULATED PINE ROOTS BY FOMES ANNOSUS IN RELA-
TION
Site
Snow Valley
Holcomb Valley
Breezy Point
Camp Paivika
TO THE SEVERITY OF AIR POLLUTION INJURY.
Air Pollution
Speciesa Damage
Moderate
JP
Slight-No Injury
JP Slight-No Injury
Very Severe-Severe
PP Moderate
Slight-Vey Slight
Very Severe-Severe
PP Moderate
Slight
Distal
Colonization"
14.7
17.4
27.4
11.3
20.5
10.4
9.4
6.1
6.4
Proximal
Colonization"
0
3.1
0.3
5.6
0
0.05
2.0
0.7
0.07
aJP = Jeffrey pine; PP = ponderosa pine
mm/month
117
-------
oo
o
E
OTMB ^' f
o^
P o
< Z
N Z
o£
-J >-
< CD
X H
O O
tr o
a. tr
I2J
10-
8-
6-
4-
2-
\
\
LEGEND
_ - - — Ponderosa Pine (Breezy Point)
Statistically Significant (P=0101)
Regression Equation:
y = 10.7 - 0.5 x
•in i»—. Ponderosa Pine (Camp Pavika)
Statistically Significant (P=0.01)
Regression Equation:
y = 3.8 - 0.2 x
" Jeffrey Pine (Snow Valley & Holcomb
Valley
Not Statistically Significant
Regression Equation:
y = 0.1 + 0.03
12 16 ' 20 ' 24 ' 28 ' 32
OXIDANT INJURY SCORE
(DECREASED INJURY
»
Figure 28. Relationships between oxidant air pollution injury and proximal colonization of
inoculated pine roots by Fomes annosus.
-------
Also, two additional twenty-stump trials were done (one for Jeffrey pine
and one for ponderosa pine). Table 36 summarizes characteristics of stumps
which were inoculated.
All inoculated stumps became infected with J?. annosus. Stumps from
trees severely injured by air pollution were about twice as susceptible
to _F. annosus, based upon percent of surface area colonized (Table 37),
as stumps from trees showing slight or no injury. Surface colonization
percentages were similar for ponderosa pine at Barton Flats and Jeffrey
pine at Amphitheatre. Much greater surface colonization occurred in ponder-
osa pine stumps at Camp Paivika. This may be explained in part by the
different _F. annosus isolate used there.
Regression analyses comparing oxidant injury score with surface coloni-
zation by JF. annosus were completed. For ponderosa and Jeffrey pine,
as the oxidant injury score decreased (corresponding to greater air pollu-
tion injury) stump surface colonization increased. This correlation was
significant for ponderosa pine at both sites tested (Barton Flats P=0.025;
Camp Paivika P=0.01). The correlation for Jeffrey pine at Amphitheatre
was significant only at the 0.25 level.
Table 38 presents JF. annosus colonization rate and extension in stumps
at least six months after inoculation. The pathogen colonized stumps of
ponderosa pine trees severely injured by air pollution at a 30% greater rate
than stumps from trees showing slight or no injury. A 50% rate differential
was found in colonization of Jeffrey pine stumps. Stump colonization rates
at Barton Flats were less than at the other two sites. This is probably
caused, in part, by season of inoculation and isolate differences from other
trials.
Regression analyses, comparing downward colonization and coloniza-
tion rates with oxidant injury score were done. As oxidant injury score
decreased, indicating greater injury colonization rate and downward fungal
extension increased. For both dependent variables, relationships were
statistically significant at the .05 level for ponderosa pine at CP and
Jeffrey pine at the Amphitheatre. They were significant for ponderosa pine
at BF only at the 0.25 level.
The stump volume colonized by _F. annosus six (BF and Amphitheatre)
and ten (CP) months after inoculation are summarized in Table 39. At
BF, stumps from trees severely injured by air pollution had twice as much
volume colonized by _F. annosus as stumps from trees slightly injured.
Amphitheatre and CP had nearly a 3:1 volume colonization ratio of stumps
from severely injured trees to stumps from slightly injured trees.
Regression analysis showed that the volume of stump colonized in-
creased with decreasing oxidant injury scores. This correlation was
statistically significant at the .01 level for ponderosa pine at CP,
but not at BF (P = .25). The correlation was significant at the 0.10
level for Jeffrey pine (Amphitheatre).
119
-------
TABLE 36. NUMBER OF PINE STUMPS INOCULATED WITH FOMES ANNOSUS BY SITE
AND SPECIES.
Barton Flats Amphitheatre
Pollution Damage* (Ponderosa Pine) (Jeffrey Pine)
Very Severe
Severe
Moderate
Slight
Very Slight
No Injury
Total
Ave. DBH (dm)
Ave. Hgt.(m)
*Equivalent Numerical
Very Severe:
Severe:
Moderate :
Slight :
Very Slight:
No Inj ury :
5 1
5 4
0 5
3 2
3 7
4 1
20 20
1,54 2.27
8.64 11.81
Ratings:
1-8
9-14
15-21
22-28
29-35
36+
Camp Paivika
(Ponderosa Pine)
5
5
0
9
1
0
20
2.53
10.60
120
-------
TABLE 37. RELATIONSHIP BETWEEN THE SURFACE COLONIZATION OF INOCULATED
PINE STUMPS BY FOMES ANNOSUS AND THE SEVERITY OF AIR POLLUTION
INJURY.
Site1
BF
BF
AMP
AMP
CP
CP
Species^
PP
PP
JP
JP
PP
PP
F. .annosus
Isolate
SV1
SV1
SV1
SV1
JL1
JL1
Air Pollution
Damage
Very Severe-
Severe
Slight-No
Injury
Very Severe-
Moderate
Slight-No
Inj ury
Very Severe-
Severe
Slight-Very
Slight
% Surface
Colonization^
36.9
17.6
33.1
15.8
86.0
45.9
Ratio
2.10
2.09
1.87
1BF = Barton Flats; AMP = Amphitheatre; CP = Camp Paivika
2pp = ponderosa pine; JP = Jeffrey pine
^One month after inoculation
121
-------
TABLE 38. DOWNWARD COLONIZATION AND COLONIZATION OF FOMES ANNOSUS IN
INOCULATED PINE STUMP RELATIVE TO AIR POLLUTION INJURY.
Site1
BF
BF
AMP
AMP
CP
CP
Species2
PP
PP
JP
JP
PP
PP
Air Pollution Downward
Damage Colonization3
Very Severe-
Severe , 22 Q
Slight-No 94 5
Injury
Very Severe- ,
... . jUb.l
Moderate
Slight-No
In j ury
Very Severe- ,07 _
Severe
Slight-Very
Slight Jli-4
Colonization
Rate* Ratio
20.5 U3Q
15.8
51.0
1.49
34 2
40 7
1.31
31 1
Barton Flats; AMP = Amphitheatre; CP = Camp Paivika
PP = ponderosa pine; JP = Jeffrey pine
months after inoculation for Barton Flats and Amphitheatre;
ten months after inoculation for Camp Paivika (Given in mm)
^Given in mm/month
122
-------
TABLE 39. VOLUME OF INOCULATED STUMPS COLONIZED BY FOMES ANNOSUS IN
RELATION TO AIR POLLUTION INJURY.
Air Pollution % Volume
Site^ Species^ Damage Colonized^ Ratio
BF PP Very Severe-
„ 1 j « o
Severe
1.95
BF PP Slight-No
Injury 8'°
AMP JP Very Severe-
i, i ._ f-O • 1
Moderate
3.19
AMP JP Slight-No _ 0
_ . o «o
Injury
CP PP Very Severe- , , n
_ bl «U
Severe
3.10
CP PP Slight-Very in _
Slight 19'7
Barton Flats; AMP = Amphitheatre; CP = Camp Paivika
2pp m ponderosa pine; JP = Jeffrey pine
^For Barton Flats the figures indicate % volume colonized in the top
130 mm of each stump at the end of six months. For Amphitheatre and
Camp Paivika, the figures indicate colonization for the entire height
of each stump at the end of six and ten months respectively.
123
-------
Effects on Growth, Spore Production, Spore Germination, and Adaptability
of Fomes Annosus—
Photochemical air pollutants probably influence J> annosus directly
during portions of the disease cycle, such as during spore production,
dispersal and germination. Penetration of stump surfaces by the fungus
may also be affected.
Direct effects of ozone on characteristic behavior of J> annosus
in culture and during wood disk colonization were investigated. Growth
chambers were used to control other environmental factors. Five character-
istics of _F. annosus potentially affected by ozone were studied: linear
growth, conidial production, conidial germination, colonization of freshly-
cut pine discs and adaptability over successive generations.
Four San Bernardino Mountain isolates were used to study ozone effects
on J[. annosus linear growth rate and conidial production. Ten-day-old
cultures were fumigated at four ozone concentrations (0.05 - 0.45 ppm = 98
- 88z ug/m3) for three days. Growth and conidial production were determined
after fumigation. As ozone dosage increased, growth rate and conidial
production decreased (Table 40). Significant differences in growth rate
between the controls and ozone-exposed culture occurred for ozone concentra-
tions of 0.10 ppm (196 ug/m3) or greater. Differences in conidial produc-
tion occurred at all ozone concentrations.
Experiments were conducted with 2 isolates to ascertain direct ozone
effects on conidial germination. Spores on water agar petri plates were
fumigated with ozone at three concentrations for various time periods.
As ozone dosage increased, percent germination generally decreased, (Tables
41 and 42) with the greatest change at higher ozone levels, such as 0.45 ppm
for 4 and 8 hours (4301.7 and 9544.2 ug/m3/hr).
In another experiment to determine ozone effects on conidial ger-
mination and germ tube extension, conidia were fumigated for 12 hours
at concentrations of from 0.09 - 0.72 ppm (176.4 - 1411.2 ug/m3). As
ozone dosage increased, germination decreased (at 0.72 ppm, no conidia
germinated), average germ tube length decreased, and percentage of germ
tubes with branches was less (Table 43)•
Two JF. annosus isolates were used to evaluate effects of ozone on
the colonization of wood discs. Discs inoculated with conidia were fumi-
gated at four concentrations of ozone for seven days. Extent of coloniza-
tion by J?. annosus over the surface of discs were determined after ozone
exposure. Colonization of discs exposed to ozone was significantly less
than the controls at ozone concentrations of 0.11 ppm (16868.2 ug/m3/hr)
and greater (Table 44)•
Two F. annosus isolates were used to investigate adaptability to
ozone through a number of successive generations. J?. annosus conidia
were ozone-fumigated for three successive generations to ascertain whether
germination percentages changed after each fumigation. No significant
differences in conidial germination occurred during the sequences of ozone
124
-------
TABLE 40. INFLUENCE OF OZONE ON THE LINEAR GROWTH RATE AND CONIDIAL
PRODUCTION OF FOMES ANNOSUS.
Ozone
Cone.
Isolate (ppa)
0.045
JL1 °a°
0.22
0.45
0.05
JP1 °-10
0.22
0.45
0.045
ci oao
0.22
0.45
0.05
PP1 °a°
0.22
0.45
Ozone
Dosage*
(ug/m3-hr)
2401.0
5634.0
11477.0
23814.0
2646.0
6350.4
11477.0
25401.6
2401.0
6350.4
13230.0
23814.0
2646.0
5634.0
13230.0
25401.6
Growth
Rate
(% of Controls)
77. 62
10.3
12.0
7,5
75.52
19.6
11.3
4.6
86. 82
42.6
18.9
14.8
82, 12
30.8
19.0
5.5
Conidial
Production
(% of Controls)
16.3
1.2
0.0
0.0
14.6
8.3
5.9
0.0
10.3
8.6
3.6
0.0
34.3
28.0
7.5
0.0
^•Cultures were fumigated 9 hours daily for 3 days (total 27 hours).
2Not statistically significant (P=0.05) using one-way analysis of
variance comparing control and ozone fumigated cultures. All other
such comparisons were statistically different.
125
-------
TABLE 41. INFLUENCE OF OZONE ON CONIDIAL GERMINATION OF FOMES ANNOSUS (ISOLATE: JL1).
Ozone
Cone .
(ppm)
0.10
0.10
0.22
0.10
0.45
0.22
0.10
0.22
0.45
0.22
0.45
0.45
Exposure Ozone
Time Dosage
(hr) (ug/m3-hr)
1
2
1
4
1
2
8
4
2
8
4
8
183.9
396.9
444.7
749.2
888.0
931.6
1573.9
1770.6
2058.0
4167.5
4301.7
9544.2
Control
Ave. %
Germin . a Sxb
90.3AB
89. 6A
9 2. 2 ABC
94.6BC
89. 3A
91.7ABC
91.1ABC
93.4BC
91.5ABC
90.2AB
90.1AB
88. 8A
2.34
3.92
2.48
1.90
2.79
1.95
1.64
2.63
2.37
2.62
2.13
1.87
Ozone Fum.
Ave. %
Germin.3 Sxb
93. 4F
85.4BC
87. 1C
86. 4C
82.1ABC
85. 7C
83.8ABC
83.5ABC
79 . 7AB
78. 9A
57. 8E
29. ID
1.95
5.60
3.18
3.13
4.15
2.87
3.85
3.17
1.89
5.22
3.12
5.26
% of
Control
103.4
95.3
94.5
91.3
91.9
93.5
91.9
89.4
87.1
87.5
64.1
32.8
F
Value0
10. 2d
3.8e
16. Od
50. ld
20. 7d
29. 9d
29. 5d
57. 7d
151. 7d
40. ld
730. 7d
1143. 5d
aMeans followed by the same capital letter are not significantly different (P=0.05) using the
Studentized Range Test for Multiple Comparisons.
^Standard Deviation
cBased on one-way analysis of variance comparing control and ozone fumigation mean germination
values:
Statistically Significant (P=0.01)
eNot Statistically Significnat
-------
TABLE 42. INFLUENCE OF OZONE ON CONIDIAL GERMINATION OF FOMES ANNOSUS (ISOLATE: PP1).
Ozone
Cone.
(ppm)
0.10
0.10
0.22
0.10
0.45
0.22
0.10
0.22
0.45
0.22
0.45
0.45
Exposure Ozone
Time Dosage
(hr) (ug/m^-hr)
1
2
1
4
1
2
8
4
2
8
4
8
183.9
396.9
444.7
749.2
888.0
931.6
1573.9
1770.6
2058.0
4167.5
4301.7
9544.2
Control
Ave. %
Germin.3 Sxb
91. OB
91. 8B
73. 1A
90. 6B
88. IB
72. 4A
89. 4B
72. 1A
89. 6B
71. 4A
90. 2B
88. 2B
2.34
2.78
3.14
2.07
3.41
4.06
1.65
5.34
2.01
2.22
1.93
2.90
Ozone Fum.
Ave. %
Gerroin.3 Sxb
94. 1G
80 . 2EF
75.7DE
85. 2F
66.9.C
68.3CD
81.9.EF
55. 6B
56. 6B
40. 9A
54. 3B
35. 2A
1.95
4.47
3.31
6.68
4.01
5.91
4.72
5.93
2.17
13.00
2.79
2.53
% of
Control
103.4
87.4
103.4
94.0
75.9
94.3
91.6
77.1
63.2
57.3
60.2
39.9
F
Value0
6.6e
48. 6d
3.2§
6.0f
161. 9d
3.3g
22. 7d
42. 7d
1266. 7d
53. 5d
1118. 5d
1956. 7d
aMeans followed by the same capital letter are not significantly different (P=0.05) using the
Studentized Range Test for Multiple Comparisons.
^Standard Deviation
cBased on one-way analysis of variance comapring control and ozone fumigation mean germination
values:
Statistically Significant (P=0.01) Statistically Significant (?=0.05)
^Statistically Significant (P=0.025)
Statistically Significant
-------
TABLE 43. INFLUENCE OF OZONE EXPOSURE ON CONIDIAL GERMINATION AND GERM TUBE EXTENSION OF
FOMES ANNOSUS.
NJ
00
Isolate
JL1
HB11
Ozone
Cone.
(ppm)
0
0.09
0.18
0.25
0.72
0
0.09
0.18
0.25
0.72
Ozone
Dosage
(yg/mP-hr)
0
2205.0
4145.4
5884.9
16993.2
0
2205.0
4145.4
5884.9
16993.2
Percent
Germination3
41.8
23.5
4.0
2.2
0
39.1
21.0
4.8
3.4
0
Ave. Germ
Tube Length*3
(mm)
0.077E
0.039CD
0.038CD
0.022B
OA
0.079E
0.04 ID
0.36CD
0.029BC
OA
S_c
X
0.051
0.020
0.025
0.018
0
0.047
0.020
0.025
0.026
0
Percent
BGT
20.0
7.0
11.2
4.0
0
13.4
2.6
5.6
4.0
0
aAfter 12 hours incubation
t>Mean followed by the same capital letter not significantly different (P=0.05) using the
Studentized Range Test for Multiple Comparisons.
GStandard Deviation
^Percent of germ tubes with branches.
-------
TABLE 44. EFFECTS OF OZONE ON THE COLONIZATION OF PINE DISCS BY FOMES ANNOSUS.
NO
vO
Isolate
JL1
HB11
Ozone Ozone
Concentration Dosage
(ppm) (ug/m3-hr)
0.06
0.10
0.11
0.27
0.06
0.10
0.11
0.27
8736.2
12595.0
16868.2
31928.4
8736.2
12595.0
16868.2
31928.4
Disc Colonization
Control3 03 Exposed3
71. 1A
75. SAB
80.4BC
84. OC
79.0ABC
77.5ABC
76.2ABC
70. 8A
75. 6C
59 . 3AB
59.6AB
55. OA
73. 8C
67.9BC
62.6AB
57. OA
% of Control Fb
106.3
78.5
74.1
65.5
93.4
87.6
82.1
80.5
2.1
10. 7C
29. 2C
134. 6C
0.5
2,3
35, 6C
39.8°
aGiven in percent—Means followed by the same capital letter are not significantly
different (P=0.05) using Duncan's Multiple Range Comparison Test.
F values based on one-way analysis of variance comparing control and ozone exposed percent
colonization values.
Statistically significant (P=0.01) ,
-------
fumigation (Table 45). Spores selected for their germination ability in an
ozone environment did not give rise to colonies which subsequently had
spores with improved germinative capabilities.
In summary, ozone influenced certain cultural characteristics of
_F. annosus under growth chamber conditions. Growth rate, sporulation,
spore germination and colonization of wood discs were all limited by
ozone, with the most dramatic effects occurring at high dosages. No
evidence of adaptation by the fungus to an environment with ozone was
found. Sensitivity of the fungus to ozone may have little effect on
epidemiology under field conditions since the pathogen, because of its
occurrence within host tissues and because it commonly sporulates at
night and during moist periods, is not often exposed to the gas during
its life cycle.
Epidemiological Model
Development of a model to simulate _F. annosus behavior in a pine
ecosystem impacted by photochemical air pollution was a goal of this
research. From such a model, prediction of potential disease buildup
and future losses might be made.
The model predicts tree mortality expected in stands subjected to
high dosages of photochemical air pollution relative to stands not im-
pacted. Experimental data and regression equations were used at various
disease cycle stages to quantify air pollution effects. The model pre-
dicts that _F. annosus root disease would be expected to increase at a
rate 7 times greater than in a stand with no injury. For example, the
model indicates that within 50 years of initial stump infection, 31 times
more trees would die from JF. annosus infection in stands severely injured
by air pollution (average rating = 14) than in stands with no injury.
This prediction needs verification by further experimentation and/or testing
in the field.
Disease Survey
Results of the surveys of 1974 and 1976 on all vegetation plots
were keypunched and input to the data management system. A print-out
was made of the discrepancies, if any, in disease observation of each
tree between the two years. Each discrepancy was screened to determine
if:
1) An observational error was made (a highly improbably change
indicated this).
2) An incorrect code was recorded for a good observation.
3) An actual change in disease status was observed.
In the case of (1) and (2) a certain number of trees required re-
checking in the field. This was done during the early summer of 1977.
The list of corrections have not yet been input to the computer data file.
130
-------
TABLE 45. INFLUENCE OF OZONE ON FOMES ANNOSUS CONIDIAL GERMINATION THROUGH SUCCESSIVE GENERATIONS
OF
Isolate
Designation
JL1
JL1-1A
JL1-1B
JL1-2A
JL1-2B
HB11
HB11-1A
HB11-1B
HB11-2A
HB11-2B
EXPOSURE .
Ozone
Concentration
(ppm)
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
Cumulative
Dosage
(ug/m3-hr)
2353.
4704.
4704.
7056.
7056.
2353.
4704.
4704.
7056.
7056.
Percentage
Controls
40. OA
40. 6A
40. 2A
39. OA
42. OA
39. 4A
39. 2A
40. 4A
36. 2A
35. 6A
Germination*
03 Exposed
27. OA
22. 4A
26. 6A
27. 4A
27. 8A
23. 4A
24. 2A
24. 8A
22. 4A
25. OA
Test
Gl
G2
G2
G3
G3
Gl
G2
G2
G3
G3
*Within each category, means followed by the same Capital letter are not significantly
different (P=0.05) using Duncan's Multiple Range Comparison Test.
-------
CAUSE AND EXTENT OF TREE MORTALITY
Introduction
Tree mortality is an important regulator of stand composition and
density. It also plays a major role in forest succession. The study
reported here was initiated in 1976 to determine the cause and extent
of tree mortality in the mixed conifer forest of the San Bernardino Moun-
tains. The focus of the study was to identify the major causes of mortality
and relate their occurrence to the incidence of oxidant pollutant injury to
the forest.
Research Objectives
1. To identify the causes of mortality of the conifer species in the
mixed conifer forest of the San Bernardino Mountains.
2. To determine the extent of mortality in these conifer species.
3. To relate mortality to the incidence of air pollution injury to
the forest.
Literature Review
The general impact of forest insects, pathogens, and oxidant air
pollutants on the forests of the San Bernardino Mountains has been re-
viewed by Wood (1973) and Miller and McBride (1973). Observations of
tree mortality and insect occurrence on the 18 permanent plots established
to observe air pollution injury on forest trees in the EPA study have
been reported by Miller (1977) and Dahlsten et al (1974). The major insect
pests, western pine beetle (Dendroctonus brevicomis Le Conte), mountain
pine beetle (D_. ponderosae Hopkins) and Jeffrey pine beetle (I), jeffreyi
Hopkins), and some aspects of their population dynamics in the San Bernar-
dino Mountains have been discussed in relation to oxidant air pollutants
by Dahlsten et al (1977). James et al (1977) has reported several experi-
ments which explored the relationship between the pathogenicity Fomes
annosus (Fr.) Cke. and oxidant air pollutants.
All of these studies have focused on selected plots within the San
Bernardino Mountains which were established in an attempt to determine
the extent and cause of tree mortality as well as air pollution injury
(McBride, 1974; Dahlsten et al, 1974; Cobb et al, 1974). The plots unfor-
tunately were not large enough to sufficiently measure the extent and cause
of tree mortality.
132
-------
Materials and Methods
The method used to determine the cause and extent of tree mortality
was developed by Byler, Hart and Wood (1976). In brief, the method involved
the following steps:
1. Random selection of plots, of approximately 285 acres each, to
provide a 15% sample of the mixed conifer and yellow pine forest
types in the San Bernardino Mountains as defined by the U.S.F.S.
2. Aerial photography of each plot using Ektachrome MS film.
3. Photointerpretation of each plot to identify and locate all dead
trees.
4. Ground check of a sample of dead trees reported by photointer-
preter. This survey was designed to check the accuracy of the
photointerpretation and to determine the cause of tree mortality.
5. Measurement of stand conditions on a sample of the ground check
plots to determine stand and site conditions and pests.
6. Calculation of extent of mortality by each pest complex.
Laboratory Analysis Procedure
Verification of the field identification of pathogens was made on
cultured samples of root material. This root material was plated on
agar and incubated for 14 days under sterile conditions in the laboratory.
Identification of fungal species was based on the appearance of fruiting
bodies and/or hyphal configurations examined under the microscope.
Results and Discussion
A total of 220 dead trees were inspected during the ground check
portion of the study. It is estimated, by projecting this sample to the
entire forest, that 11,243 trees died during 1976. An overall summary
of the mortality percentages is shown in Table 46.
Specific pest complexes were identified for the major forest species.
Only Jeffrey pine and white fir are reported here because an insufficient
number of mortality centers for the other species (Table 47). Examination
of the data indicate that an overwhelming number of trees have succumbed to
the combined attack of disease and forest insects rather than a single pest
species. The combined effects of the Jeffrey pine beetle and dwarf mistle-
toe were the most common cause of mortality in Jeffrey pine while the fir
engraver beetle and Fomes root and butt rot were most common in white
fir.
The correlation between tree mortality and air pollution injury was
much lower than anticipated (Table 48). The data collected reflects the
distribution of the samples rather than what is believed to be the actual
133
-------
TABLE 46. TREE MORTALITY BY CAUSES AND SPECIES IN THE SAN BERNARDINO
MOUNTAINS IN 1976.
Mortality Cause
Percent of Total Mortality
Jeffrey Pine
Combined
Jeffrey Pine and
White Fir
White Fir
Pathogens and
Insects
Insects
Pathogens
Other
61
34
1
4
70
21
1
10
78
9
3
8
TABLE 47. TREE MORTALITY BY PEST COMPLEX AND SPECIES IN THE SAN BERNARDINO
MOUNTAINS IN 1976.
Pest
Complex
Percent of Total Mortality
Jeffrey Pine
Combined
Jeffrey Pine and
White Fir
White Fir
Root Disease
and Insects 26
Root Disease,
Mistletoe and
Insects 18*
Mistletoe and
Insects 22*
Insects alone 19
Insects and
Mechanical Injury 15
Pathogens alone 1
No insect, path-
ogen, or injury
apparent 4
39
16
15
14
7
1
10
53
19**
6
9
0
3
*Dwarf Mistletoe
**Tree Mistletoe
134
-------
relationship between air pollution injury and tree mortality. It does
not fit the trends observed by Miller (1977) on the 18 permanent plots
established to observe air pollution injury. The samples selected for study
fell outside of the zone of severe air pollution. Most of the samples were,
in fact, in areas of slight oxidant concentration. Subsequent surveys of
tree mortality will be stratified in order to obtain samples from zones
of high oxidant concentration.
TABLE 48. PERCENT OF MORTALITY CENTERS IN THE MIXED CONIFER AND YELLOW
PINE FORESTS IN RELATION TO OXIDANT INJURY.
Oxidant Injury Percent of Sampled Mortality Centers
Mixed Conifer Yellow Pine
Rating Class Forest Forest
1- 8
9-14
15-21
22-28
29-35
sen-
very severe
severe
moderate
slight
very slight
no visible symptoms
0
1.1
3.4
8.0
20.4
67.1
0
0
7.9
7.9
38.6
45.6
135
-------
EFFECTS OF PHOTOCHEMICAL AIR POLLUTION ON FOREST TREE SEEDLING
ESTABLISHMENT
Introduction
The purpose of this subproject is to investigate: (1) the rate of
tree seedling establishment in forest stands exposed to different levels of
photochemical oxidants; (2) the influences, both direct and indirect, of
oxidants on the establishment of ponderosa and Jeffrey pine seedlings; (3)
the individual and concurrent joint effects of biotic and physical factors
on seedling establishment, and (4) the influence of an oxidant gradient on
these biotic and physical factors.
The investigations initiated in the 1976-1977 period included studies
(1) to determine the rate of seedling establishment in forest stands along
an oxidant gradient; (2) to evaluate the individual and joint effects of
vertebrates, arthropods, and pathogens on seeds and seedling establishment;
(3) to evaluate the interactions between litter depth and type and the
effects of the biotic agents; (4) to determine the relationships between
litter decomposing organisms and those organisms causing loss of seeds and
seedlings; (5) to further analyze data from studies initiated during the
1975-1976 contract period.
The following is a summary of studies and results obtained during
the 1976-1977 period.
Pathogenicity Test of Litter Organisms
This test was designed to determine which fungi isolated from pine
litter were pathogenic to pine seeds and seedlings. Seven fungi were
tested in the initial study: Mucor, Curvularia. Aureobasidium, Alternaria,
Ulocladium, Penicillium, and Fusarium roseum.
Summary of Test Procedure—
Inoculum of the seven fungi was grown on a litter substrate in dishes
for 12 days; 10 seeds then planted in each plate. The plates were watered
and incubated in a cool environment (4.4'C) for 20 days; then they were
removed during the day to room temperature (IS^C) and returned to the
cool environment at night for an additional 57 days. They were watered
periodically, and dying seedlings were removed and isolated from the culture
dishes. At the conclusion of the study remaining seeds, seedlings, and
samples from the surface and subsurface of the litter substrate were iso-
lated from each of the substrate dishes.
136
-------
Test Results—
A summary of the results for each fungus and the control are given
in Table 49. Seed germination was observed to be lower in all inoculated
plates than in the controls,.but it was only substantially lower for
the Penicillium plates. Although the fungus was not isolated from a high
percent of the nongerminated seeds.
The Alternaria plates were observed to have a substantially higher
number of diseased seedlings than the control; however, Alternaria was
only isolated from about half of the diseased seedlings. The 31.7% diseased
controls contained some organisms used in the other plates, indicating some
contamination may have occured between plates.
Final conclusions from this experiment will be drawn after a statis-
tical analysis is performed on the data. Initial results indicate that
Alternaria was pathogenic on seedlings. The other organisms showed an
effect on the seedlings, but because of the contamination of the con-
trols, we have not yet been able to draw conclusions about these organ-
isms.
TABLE 49. PATHOGENICITY TEST SUMMARY
# Plates
(10 seeds
Test Organism per plate)
Control
Mucor
Curvularia
Aureobasidium
12
6
6
6
Alternaria
Ulpcladium 6
Penicillium 6
Fusarium roseum 5
Mean % Di-
Mean % Mean % seased Seed-
Mean % Seeds Seedlings lings with
Germination Diseased Diseased Test Organism
95.8
80.0
88.3
83.3
91.7
85.0
60.0
82.0
0.8
3.3
1.7
3.3
8.3
6.7
5.0
0
31.7
25.0
26.7
38.3
45.0*
33.3
21.7
32.0
NA
93.3
55.8
53.3
49.6
92.2
93.8
74.7
137
-------
Field Study, 1975-1976.
The field study of 1976-1976 involved 4 vegetation plots located
across the air pollution gradient, CAO, BF, HB, and HV. Seeds were planted
in 0.4 meter square areas (mini-plots) with 49 seeds per square in these
plots. The mini-plots were either open or screened, and some were litter
covered while others were bare mineral soil. The basic design is given on
page 188 of the 1975-1976 EPA Progress Report.
Results—
Predation by birds and small mammals was extremely heavy on seeds
in unscreened miniplots. Only about 3% of the seeds on unscreened plots
without litter survived to germination, and only 1% of the seeds survived
to germinate on the unscreened plots with litter. It appears that animals
were primarily the cause of the mortality in the plots without litter,
while in the plots with litter a combination of animal predation and seed
pathogens depressed survival even further. There was no significant differ-
ence among the four vegetation plots because of high variability among
the individual mini-plots. Results for the screened mini-plots are given
in Table 50 and 51. Germination as checked in May 1976 averaged about 62%
of the seeds in the screened plots without litter had survived to germinate,
while only about 20% of the seeds in the screened, plots with litter had
survived (Table 50). Two-way analysis of variance among the 4 vegetation
plots and between litter and no litter showed the differences to be non-
significant among the vegetation plots but the differences were highly
significant (.001 level) between litter and no-litter plots. It is believed
that seed pathogens found mostly in the litter were primarily responsible
for this difference in germination rate.
In July of 1976 the numbers of living seedlings remaining alive had
been greatly reduced from the number germinated. Table 51 gives the means
per mini-plot of surviving seedlings in July. Compared to the number
germinated, about 16% had survived in the no-litter mini-plots while 5% had
survived in the litter mini-plots. Seedling damping-off pathogens vrere
believed responsible for part of the mortality, especially in the late
spring, but by mid-summer drought was the major factor. The driest plot,
HV, had no surviving seedlings by July in the no-litter mini-plots, and
only 1 in the litter mini-plots. The damping-off fungi appeared to have
a greater effect on the seedlings in the litter mini-plots. An analysis
of variance showed the difference between litter and no-litter to be
highly significant (.001 level) while the differences among vegetation
plots was marginally significant (.10 level). The vegetation plot differ-
ences were due primarily to the high mortality at HV because of drought. No
smog effects were apparent; if smog effects were present, they were obscured
by the rainfall differences between plots and the small sample size.
The results of this study pointed to the need of an expanded study
with higher sample numbers and the addition of a control variable for
soil moisture content.
138
-------
TABLE 50. 1975-1976 SBNF SEEDLING ESTABLISHMENT FIELD STUDY, GERMINATED
SEEDS, MEANS PER SCREENED MINI-PLOT IN MAY, 1976. FORTY-NINE
SEEDS WERE PLANTED IN EACH MINI-PLOT.
Vegetation Plot
CAO
BF
HB
HV
Mean of Plots
No Litter
Litter
30.0 34.0 32.0 26.0
13.0 12.0 7.1 6.4
30.0
9.7
TABLE 51. 1975-1976 SBNF SEEDLING ESTABLISHMENT FIELD STUDY, SURVIVING
SEEDLINGS, MEANS PER SCREENED MINI-PLOT AS OF JULY, 1976.
FORTY-NINE SEEDS WERE PLANTED.
Vegetation Plot
CAO
BF
HB
HV
Mean of Plots
No Litter
Litter
5.9 8.1 5.6 0
0.1 1.3 0.3 0.1
4.9
0.4
139
-------
Field Studies, 1976-1977.
The seedling establishment field studies of 1976-1977 consisted of
one primary study (Seedling Air Pollution Study - SAPS) and two supple-
mentary studies (Supplementary Seedling-Animal Study -SAAS. and Supple-
mentary Litter-Seedling Study - SLSS).
The objective of the SAPS study was to determine air pollution and
moisture effects on the survival of young seedlings in the San Bernardino
mountains. The first supplementary study (SSAS) was designed to determine
relative rates of seed predaton by small mammals vs. birds on the forest
floor. The second study (SLSS) was designed to test the potential for seed
survival and germination in different types of litter under natural condi-
tions .
Primary design (SAPS)—
Healthy Jeffrey pine seeds were planted on four vegetation plots,
two (BF and CAO) with relatively high air pollution and two (HV and HB)
with low air pollution. Each vegetation plot was subsampled with four
subplots, and each subplot had 16 mini-plots consisting of screened 15 x 15
inch frames planted with 49 seeds per frame. Twelve of the 16 mini-plots
had the litter layer completely removed prior to planting, while 4 were
placed on undisturbed, natural pine litter. One of the 4 subplots on each
vegetation plot was chosen as a "water" subplot. These subplots were
periodically (once a month or as necessary) watered with an amount of water
necessary to simulate a heavy rainfall month. This amount was defined as
the mean rainfall for each specific month over the past 25 years at Big Bear
Lake Fire Station, plus one standard deviation. On each field trip to the
study plots (done every 3 to 4 weeks during the spring, summer, and fall of
1977) soil samples were taken from 4 mini-plots (two with litter and two
without) on each water subplot, and 2 mini-plots (one litter and one no-
litter) for each of the other subplots. Moisture content of the soil
from these samples was determined in the lab. On each field trip data was
taken on surviving seedlings and dead seedlings, and all dead seedlings were
removed. Possible mortality causes and plot disturbances were noted.
Seedling data was not taken on those mini-plots used for soil sampling.
At the end of the field season, analyses will be made on seedling
mortality by plot, considering the variables of soil moisture and rain-
fall, for each 3 to 4 week period. Rainfall and oxidant exposure data
from other projects will also be used to try to relate periods of greatest
mortality with corresponding environmental conditions.
Supplementary design (SSAS)—
Four subplots were chosen on pine litter covered areas of CAO, and
6 mini-plots (15 x 15 inches) were planted with 49 sound Jeffrey pine
seeds each in the fall of 1977. On each subplot, three of the mini-plots
had the litter layer completely removed, while on the other three the
litter was undisturbed. On each litter type, one plot was completely
screened to exclude all vertebrates, one was closed except for a narrow gap
140
-------
at ground level, which would allow small mammals to enter, and one plot
was completely open.
The following table is a summary of the percent of seeds that sur-
vived to germinate:
No Litter Litter
Completely screened 20.9% 3.6%
Screened top, open base 0.5% 0.5%
Open top 1,5% 0%
The seeds used in this test showed an 85% germination rate under
sterile lab conditions. The results indicate that in the case of the
completely protected mini-plots, 65% of the additional mortality was due
to pathogens and/or arthropods. In the case of the protected litter
mini-plots, an additional 16% mortality occurred. The results from the open
base mini-plots and open top mini-plots indicate nearly complete mortality
(the total number of seeds surviving to germination on these mini-plots was
only 5 of 784, or 0.6%). Since the open base mini-plots had about the same
survival as the open top mini-plots, it can be concluded that small mammals
were probably most important in seed predation. Broken seed coats and scat
(mouse droppings) were observed around most of these mini-plots.
Supplementary design (SLSS)—
On one vegetation plot (CAO), seven different major types of litter
cover were selected:
1. Jeffrey light litter (with PP and Oak)
2. Jeffrey heavy litter (with PP and Oak)
3. Oak litter (with PP and JPO
4. Mixed Jeffrey - Ponderosa - Oak litter
5. White Fir (with JP, PP and Oak)
6. Bare ground with some JP, PP needles
7. Bare ground with some grass
Two areas were found for each of the above litter types. On each
areas, 6 mini-plots, (.4 x .4m) were established at random and marked
by stakes at two corners. On each of three of the mini-plots 49 seeds
were placed, while the other three received 16 seeds each. Litter depth
and % cover were recorded for each mini-plot.
The seeds were planted on November 3, 1976. All the plots were
rechecked on April 20, 1977. This check showed no visible seedlings on
any of the mini-plots. Further checks will be made in the fall of 1977
to see if any seedlings have become established.
Miscellaneous field activities—
141
-------
During the fall of 1976 cones were collected from several healthy
Jeffrey pines in the SBNF. No cone trees with significant smog damaged
were found at this time. Seeds from the healthy tree cones were used
in the studies described above.
Cones were again collected in August, 1977. This time both smog
damaged and healthy trees were located (both Jeffrey and Ponderosa pines).
The cones for each tree were measured, and the numbers of cones and yield
of seeds were recorded for each tree. Tests of the cone data, seeds,
and seedlings grown from them in the greenhouse will be made to determine
if there is a relationship between cone, seed, and seedling quality land
the degree of smog damage of the parent tree.
A pilot study was made in the fall of 1976 to determine if seed drop
under natural conditions could be determined on the vegetation plots.
Very few cone bearing trees were found on any of the plots. A few trees
were found on CP, and a 30 x 30 meter subplot was established under these
trees. Fifteen seed traps were placed randomly in this subplot, but no pine
seeds were recovered. It is believed that seed drop was too small to be
detected by the density of traps used.
142
-------
CONE AND SEED PRODUCTION FOR DOMINANT CONIFER TREE REPRODUCTION
Introduction
The two primary objectives of this study are:
1) To test the hypothesis that cone crop abundance and frequency
in ponderosa and Jeffrey pines are affected by ozone injury;
and
2) To describe the probability that a tree with specified charac-
teristics will produce a cone crop in a given year.
The rational for the cone study and the study design, methods, and tree
characteristics used to classify the trees have been previously reported
(Luck, 1977).
Recent Research Progress
Status in acquiring plot data other than the cone counts is presented
in Table 52. A total of 9 pots have been completed, the data punched
on 1MB cards and the values on those cards verified against the origi-
nal data.
These data sets have been forwarded to the data management group
for entrance into the data bank. The other ten plots are in various
stages of completion.
The annual cone crop data, 1973-1976, obtained by visually counting
cones within the tree crowns, have been punched on IBM cards, verified
and forwarded to the data management group for entry into the data manage-
ment system. These counts have been reported by Luck (1977). The 1977 cone
crop is currently being counted. Ground counts have been made of the whole
cones, and of those eaten by squirrels during the 1976 cone crop. Insect
damaged cones and cones which aborted before maturing were placed in rearing
containers. Cones from the 1975 crop that were placed in rearing are being
processed. The insect species assciated with a particular type of damage is
being identified so that damaged cones can be identified in the future, even
though the insects causing; the damage are absent. Curation of insects
reared from damaged cones obtained from the 1974 and 1975 cone crops has
been partially completed. We are waiting for the return of some specimens
sent to several specialists for identification.
As an example of the way in which the data is being analyzed, Figures
29, 30, 31 and 32 present annual cone crop data (combined visual counts for
the 1974-76 cone crops) for ponderosa and/or Jeffrey pine trees on three
143
-------
TABLE 52. STATUS IN ACQUIRING PLOT DATA ON THE 19 STUDY PLOTS LOCATED
IN THE SAN BERNARDINO MOUNTAINS OF SOUTHERN CALIFORNIA.
Data Type
Plot
ID
UCC
HB
COO
SF
CP
DW
SV.I
TUN2
GVC
BP
NEGV
HV
DL
CAO
CA
BF
SC
SCR
BL
Tree
age
X
0
X
X
—
X
X
X
X
X
X
X
X
X
X
0
0
X
—
Crown
height
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
0
0
0
0
Tree
height
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
0
0
X
0
Keen
class
X
0
X
X
0
X
X
X
X
X
X
0
0
0
—
0
0
0
0
Crown
class
X
X
X
X
—
X
X
X
X
X
X
X
0
X
X
—
0
0
X
Data verified and entered
into data management system
X
X
X
X
X
X
X
X
X
Key: X = complete
— = partially done
0 = not started
144
-------
II
II
k
41
NEGV
No. of Trees 65
X Age 162
12
GVC
No. of Trees 63
X Age 148
I
22
15
13
13
SF
I 2
44
35
86
70
V/.
66
No. of Trees 113
X Age 81
18
<30 50 70 90 110 130 150 170 190 >200 % CONES
AGE (years)
Figure 29. Age class distribution of ponderosa and/or Jeffrey pine
trees on three plots. Solid bars = % of trees in each age
class; Open bars = % of cone bearing trees in each age
class; Horizontal hashed bars = % cones produced per age
class; Diagonally hashed bars = % of the combined three
plot cone production.
145
-------
40-
NEGV
No of Trees 65
X Age 162
28
sT
28
11
96
GVC
No. of Trees 63
X Age 148
66
2728
d
SF
No. of Trees
X Age 81
15
113
DO CD II 10 IS SS
CROWN CLASS
% CONES
Figure 30. Grown class distribution of ponderosa and/or Jeffrey pine trees
on three plots. Solid bars = % of trees in the six specified
crown classes; Open bars = % of trees bearing cones in a crown
class; Horizontally hashed bars = % of cones with respect to
crown class of cone bearing trees; Diagonally hashed bars = % of
the combined three plot cone production. DD-Dominant; CD-Co-
dominant; Il-Intermediate; 10-Intermediate open; IS-Intermediate
suppressed; and SS-suppressed.
146
-------
NEGV
No. of Trees 65
X Age 162
14
16
GVC
No. of Trees 63
X Age 148
15
66
1
43
SF
No. of Trees 113
~ Age 81
13
I
IB
15 25 35 45 55 65 75 85 >90
DBH (cm)
%CONES
Figure 31. Diameter class distribution of ponderosa and/or Jeffrey pine on
three plots. Solid bars = % of trees in specified diameter
classes; Open bars = % of trees in a diameter class that produced
cones; Horizontal hashed bars = % of cones with respect to diam-
eter of cone bearing trees; Diagonally hashed bars = % of the
combined three plot cone production.
147
-------
NEGV
GVC
No. of Trees 63
X Age 148
16
I
32
SF No. of Trees 113
X Age 81
13
20
MJ 10
No. of Trees 65
X Age 162
27
JD
64
45
24
13
16
66
i
18
15 25 35 45
TREE HEIGHT (m)
% CONES
Figure 32. Height class distribution of ponderosa and/or Jeffrey pine on
three plots. Solid bars = % of trees in the five height
classes; Open bars = % of trees in each height class that pro-
duced cones; Horizontally hashed bars = % of cones produced
with respect to height of cone bearing trees; Diagonally hashed
bars = % of the combined three plot cone production.
148
-------
plots, Northeast Green Valley (NEGV), Green Valley Creek (GVC) and Sky
Forest (SF). The trees from these plots are classified into groups based
on 20 year age intervals (Fig. 29), crown classes (Fig; 30) (see Luck,
1977, for a description of the crown classes), 10 cm intervals in diameter
(at breast height) (Fig. 31), and 10 m intervals in tree height (Fig,
32).
The first point to note is that NEGV and GVG have older trees pres-
ent on the plots (Fig* 29). Trees 200 years or older are absent from SF.
The lack of older trees on SF is due to the logging history on that plot.
The 200+ year class of trees, however, produces more than its share of
cones. For example, trees in the 200* year class at NEGV represent 35 per
cent of the trees bearing cones and bear 44 per cent of the cone crop (total
number of cones) while they make up only 29 per cent of the stand. This
pattern is even more pronounced at GVC. Trees in the 200* year class repre-
sent 45 per cent of the stand but account for 70 per cent of the trees
producing cones and bear 85 per cent of the cone crop. The older trees, age
classes 130, 150, 170, show the same trend at SF. Clearly the older trees
are the greatest source of cones and, thus, of seed as well.
Crown class is perhaps the most important tree characteristic linked
to cone production (Fig. 30). Clearly, dominant trees are the greatest
source of seeds because they contribute more than their share of cones
and seeds based on their representation in the stand.
The larger diameter (Fig. 31) and taller trees (Fig. 32) are also
disproportionate contributors to the seed crop. However, both these
variables are correlated with tree age and crown class (dominant trees
are usually the taller and larger diameter ones); hence, these variables
may explain little or no variation in crown crop abundance when they
are corrected for covariance with other variables.
Ozone damage classes were not represented by a sufficient number
of trees on these plots to permit assessment of the effect of this variable
on cone production. It should be used to reclassify; tree only within a
crown class or age class. For example, given that a tree belongs to the
dominant crown class, how does ozone damage effect the cone crop produced by
that class? Thus, before the effects of ozone damage can be assessed, the
pattern of cone produciton as influenced by a number of other variables,
such as age and crown class, needs to be determined first. Furthermore,
since cone crops vary substantially in abundance, a number of years of data
are essential before the pattern of cone production can be revealed.
149
-------
LITTER PRODUCTION SUBSYSTEM
Introduction
Oxidant air pollution stimulated development of the abscission zone and
thus induced heavy defoliation, especially on yellow pines. The rate of
needle fall in the early stages of impact increased markedly and later
decreased to zero as the tree was increasingly defoliated and finally
killed. The question here is how does the change in rate of needle and
branch mortality and the consequent change in addition of needle and branch
litter affect the nature and amount of organic litter covering the soil?
Further, are there consequent effects upon: the soil moisture and tempera-
ture; seed germination; seedling survival; and the forest floor as a habitat
for micro- and macroorganisms? Since defoliation also results from other
plant damage (pathogens and insect attack) there is also a question as to
the effect of air pollutants in relation to other causes of defoliation.
As foliage of the canopy is damaged and the density is altered, it is
postulated that both chemical composition and amount of crown drip falling
on the soil during periods of precipitation will be altered. Here the
question is, do changes in crown drip and throughfall precipitation affect
the soil and the forest floor; and if so, are there consequent effects?
Research Objectives
The objectives of this project relative to oxidant injury gradient
were as follows:
1. Continue and expand the measurement of litter production from
individual trees on all major plots.
2. Continue to measure nutrient content of needle fall and accumulate
litter under individual trees variously affected by oxidant air pollutants.
3. Verify with expanded data the effect of oxidant pollutants on
needle size in the litter.
4. Measure accumulated litter under all trees for which data were
being collected.
Materials and Methods
Measurement of litter production—
Litter was collected as it fell on 46 cm square screens under 50
yellow pines (P^. ponderosa, £- Jeffrey) on 16 plots. Collections
150
-------
from the screens were made in the late spring and late fall in both 1975
and 1976, expanding the number over that of 1974 which was 39 trees.
Two to four screens were used under each tree depending upon crown size.
The litter was oven-dried, separated into needles and other material (twigs,
etc.) and weighed. A large number of intact needle-fascicles were sepa-
rated, counted and weighed to obtain the mass per fascicle as a measure of
needle size. The total needle component of the litter was analysed for (N,
P, K, Ca and Mg).
Needles were digested by the Johnson and Ulrich (1959) procedure
in preparation for nutrient analyses. Nitrogen was measured by a modified
Kjeldahl method from Black (1965) and phosphorous was determined by a
colorimetric method of Richards in a book edited by Jackson (1960). Atomic
absorption was used to measure the concentration of cations.
Results and Discussion
Pine Litter Production—
Litter collected under pine trees during 1975 (47 trees) and 1976 (41
trees) on 18 major vegetation plots was analysed for total litter deposit,
total needle component and size of needle fascicles (mass per fascicle) in
relation to oxidant injury rating. Needles were also analysed for content
of N, P, K, Ca and Mg.
Amount of litter—Dry weight of needles collected during late autumn of
1975 and 1976 as related to oxidant injury score is shown in Figure 33. The
increase in needle fall from relatively healthy trees (score >31) of 87.5
gm/m^ to 256.5 gm/m^ from severely affected trees (score 8 to 15) was
statistically significant (P=.01) as is the decline to 125.7 gm/m^ from
severely affected trees (score 8 to 15) is statistically significant (p =
.01) as is the decline to 125.7 gm/m^ from very severely affected trees
(score 0 to 7). The total amount of litter of all kinds collected on the
screens followed almost exactly the same pattern, with litter other than
needles (twigs, branches, etc.) contributing about 33 percent of the
total.
An important impact of oxidant air pollutant was to increase the litter
fall during the period of tree injury from the onset of clearly discernable
injury (score 30) until the tree died. If the duration of this period is
assumed to be 8 years, then the litter fall would be almost exactly double
that expected from unaffected pine trees. After the first year following
death of the tree, needle fall is very small, but a vast increase in dead
branches (woody litter) may be added. The net result is predictable,
namely that the amount of undecomposed litter on the forest floor is
increased and thus the fire hazard increased during the injury period and
for a number of years following the death of the pine trees.
In addition to the fire hazard created by increase litter fall, the
extra thickness of fresh loose litter created on the forest floor can
be expected to have a marked detrimental influence on seed germination and
seedling survival during and following the period of oxidant injury and tree
151
-------
IB
0>
O
CJ
d
UJ
O
UJ
111
z
UJ
z
Q.
U_
O
CO
CO
O
70
60
50
40
30
20
10
On-17
on=2?
Oh
i
e
7 15 23
OXIDANT INJURY SCORE
31
Figure 33. Total pine needle litter collected related to oxidant injury
score 1975 and 1976.
152
-------
mortality: this aspect of litter fall is being investigated by others
on the project.
Size of needles in the litter—The size of needles as measured in
grams per fascicle (cluster of 3 needles) by counting and weighing a
large number of inteact fascicles separated from the total litter collected
by the screens was determined for each tree. The data showed clearly that
as oxidant injury increased, size of the pine needles decreased. The rela-
tionship was almost identical for 1975 and 1976 (difference between the
regression lines was not significant). Least squares linear regression for
the combined years is Nm = .01665 + .0073 1R, where Nm is mass per
fascicle in g/m^, and IR is the oxidant impact rating scaled with increas-
ing scores indicating decreasing injury. The correlation r = .392 is very
highly significant (P = .001) although there is considerable scatter about
the regression line (Fig= 34). Considering the usual variability in biolog-
ical systems, it appears safe to conclude that the oxidant injury rating
system does reflect the general health of the pine trees. Stratification of
the data by grouping into either high and low rainfall plots or into pon-
derosa pine and Jeffrey pine does not give increasing reliability, although
the correlation for the high rainfall plot was higher (r = 0.536) and the
regression line somewhat steeper (Nm = 0.145 + .00158 IR). However, the
difference from the general equation does not appear to be very significant.
Plant Nutrient Content of Needle Litter—
The content of nitrogen (N), potassium (K), phosphorous (P), calcium
(Ca), and magnesium (Mg) in the needles collected on the screens in 1975 and
1976 were determined and the relationship with oxidant injury ratings and
needle size was analysed by linear regression.
The content of N, P, and K in the needles which fell tended to increase
with increasing oxidant injury and declining needle size. Ca, on the
other hand, decreased and Mg appeared to be variable, but with no con-
sistent trend. It appears that as needle mass decreased the carbohydrate
content and N, P and K content per fascicle remained about constant so that
the concentration increased, a simple dilution relationship. This can be
seen from the ratios given in Table 53 where for a 49-5 percent increase in
needle size in high rainfall plots, concentration of N decreased by 50 per
cent. The relationship was not quite as exact for phosphorous (P), and was
still less exact for potassium (K) especially in the high rainfall plots.
The increase in Ca concentration with increasing needle size appeared to be
explainable in that Ca is involved in carbohydrate in cell wall material so
the larger needles take in more calcium as the cells enlarge. Relationships
between N content, versus oxidant injury score are shown in Figures 35 and
36; for K content in Figures 37 and 38; for P in Figure 39 and 40; and for
Ca in Figures 41 and 42.
Phosphorous in the Surface Soil—
A discussion of phosphorous in the surface soil is included here
as it is postulated that plant nutrients in the surface soil may be affected
by the nutrient content of the litter. The soluble soil phosphorous in the
153
-------
(JI
0.3
0.25
0.0
f
T
T
0 — 1978 H< 0.146 «• 0.00150
r « 0.64
0--I976
O"1978 t«0.168 1- 0.00076 F
T» 0.392
D —1976
1
10 15 20 25 30
OXIDANT INJURY SCORE (S)
35
40
45
Figure 34- Size of pine needles in needle litter collected related to oxidant
injury score.
-------
TABLE 53. NEEDLE SIZE AND CONTENT ON N, P, K AND Ca IN NEEDLES COLLECTED
ON SCREENS.
High Ratio Low Ratio
Rainfall Smaller/ Rainfall Smaller/
Plots Larger Plots Larger
Oxidant Injury
Score* 0 50 0 50
Needle Size
gm/fascicle* 0.124 0.25 0.495 0.168 0.20 0.84
Percent N in
needles* 0.688 0.34 0.50 0.542 0.438 0.81
Percent K in
needles* 0.48 0.11 0.23 0.36 0.23 0.64
Percent P in
needles* 0.096 0.042 0.44 0.076 0.055 0.72
Percent Ca in
needles* 0.294 0.453 0.65 0.243 0.582 0.42
*NOTE: 0 oxidant injury score is very severe injury and 50 is a health
tree with no evident injury. The numbers in the body of the
table are average values derived from regression analysis.
155
-------
Ul
0.8
CO
3 0.7
H
z
UJ
ij>.6
|j_r Q
O o
0.5
0.4
O
o
§
oc.
0.3
8
O
a
O
D
j I
a
o
T
T
O- 1975 N= 0.676 - 0.0067 S
Q-1976 r =0.51
J I
10 15 20 25
OXIDANT INJURY SCORE (S)
30 35 40 45
50
Figure 35. Nitrogen content of pine needle litter related to oxidant injury
score—high rain fall plots.
-------
Ui
U|
I 0.6
Q.
0.5~
O
O
ui 0.4
o
o
(C
0.3
a
"1 I I
O-I975 N =0.543- 0.0021 S
D-1976 --0.43
D
OO
10 15 20 25
OXIOANT INJURY SCORE (S)
30 35 40 45
50
Figure 36. Nitrogen content of pine needle litter related to oxidant injury
score—low rainfall plots.
-------
Ul
00
0.6
V)
UJ
a 0.5
UJ
£L
0.4
0.3
55 0.2
I
o
Q.
O.I
D
"1 I I
O-I975 K= 0.48 -0.00739 S
D-1976 T=-0.59
10
15 20 25 30
OXIDANT INJURY SCORE (S)
35
40
Figure 37. Potassium content of pine needle litter related to oxidant injury
score—high rainfall plots.
-------
0.6
Ul
VO
I I I
O-I975 K* O.36-0.0026 S
D -1976 r = 0.65
10
15 20 25
OXIDANT INJURY SCORE
30
Figure 38. Potassium content of pine needle litter related to oxidant injury
score—low rainfall plots.
-------
0.02
10 \5 20 25 30
OXIDANT INJURY SCORE (S)
35
40
45
50
Figure 39. Phosphorous content of pine needle litter related to oxidant iniurv
score—high rainfall plots.
-------
o\
o
UJ
UJ
0.10
Q.
£ 0.08
Z *
g 0.06
CO
I
Q.
0.04
0.02
I I I I
O-1975 P - 0.0761 - 0.000418 S
D-1976 r=-0.52
O
D
O
D
no
I I
o
I
o
I I
10 15 20 25 30
OXIDANT INJURY SCORE (S)
D
35 40 45 50
Figure 40. Phosphorous content of pine needle litter related to oxidant injury
score—low rainfall plots.
-------
£0.7
o 0.6
tn
0.5
a.
O
0.4
O.3
o
O-I975 CA =0.294 + 0.00319 S
D-1976 r * 0.40
O
10 15 20 25 30
OXIDANT INJURY SCORE (S)
40
45
Figure 41. Calcium content of pine needle litter related to oxidant injury
score—high rainfall plots.
-------
0.7
I
« 0.6
! 1 i
-
1 l
0-1975
0-1976
n
1 l i 1
CA= 0.245 + 0.00654 S
r • 0.81
„ •
1
^
UJ
10
15 20 25 30
OXIDANT INJURY SCORE (S)
Figure 42. Calcium content of pine needle litter related to oxidant injury
score—low rainfall plots.
-------
upper 7.5 cia of soil sampled under pine trees of varying oxidant injury
rating is shown in Figures 43 and 44. It can be seen that as the oxidant
injury score approached zero (tree death) the soluble phosphorous in the
soil increased. By comparing Figures 39 and 40 with Figures 43 and 44 it is
evident that the slope of the regression lines for needle phosphorous
content with oxidant injury score are similar to the regression lines for
soluble phosphorous in soils with oxidant injury score. There is probably a
causal relationship here, as postulated. Similar relationships were found
for Ca, K and Mg however neither the Ca nor Mg content of needle litter
increased with increasing injury by oxidant air pollutants.
164
-------
Ui
3 °'5
o
CO
u_
0 0.4
H
UJ
8 _ 0.3
(0
o
3
CD
O
0)
0.2
O.I
0.0
O
I
I
P» 0.288-0.0042 S
r = 0.34
I
10 15 20 25 30
OXIDANT INJURY SCORE (S)
35
40
45
50
Figure 43.
Soluble phosphorous content of surface soils related to oxidant in-jury
score of pine trees—high rainfall plots.
-------
o\
0.5
o
CO
fe 0.4
o
0.3
ai
x~ 0.2
ft
O.I
O
CO
0.0
O
_L
°o
1
P= 0.254-0.00316 S
r = 0.59
O
1
O
10 15 20 25 30
OXIDANT INJURY SCORE (S)
35
40
45
50
Figure 44. Soluble phosphorous content of surface soils related to oxidant injury
score of pine trees—low rainfall plots.
-------
FOLIAGE LITTER DECOMPOSITION SUBSYSTEM: MICROBIAL ACTIVITY AND NUTRIENT
CYCLING
Introduction
There is a continual turnover of tree biomass in the forest as old
foliage, branches, and trees die and fall and other &row to replace
them. Litter decomposition and nutrient cycling are the means by which
the living forest recovers much of the nutrition incorporated in this
organic matter. Recovery of these vital nutrients is the function
of the vast populations of litter and soil microflora and microfauna.
This portion of the overall study is directed toward determining the
influence of oxidant air pollution on microfloral populations and leaf
litter decomposition (primarily of ponderosa and Jeffrey pines). Al-
though previous work in this field is relatively sparse, reviews are
available (Dickinson and Pugh, 1974).
Research Objectives
Three major research approaches are being implemented in this sub-
system study. They include the following:
1) To quantify needle litter decomposition in the field. This
will help determine the degree to which oxidant air pollution
affects (a) the quality of needles as substrates for decomposers
and as sources of nutrients for cycling, and (b) the capacity of
naturally-occurring populations of litter micro-organisms to
decompose pine needle litter.
2) To characterize and quantify microfloral inhabitant populations
of pine needles from the period needle elongation through decom-
position on the forest floor. This will suggest effects of oxidant
air pollution on decomposer communities and provide a basis
for laboratory studies of fumigation effects on decomposition.
3) To conduct laboratory fumigation experiments on both fungal
growth and needle decomposition. These are expectd to clarify
results obtained from field studies by eliminating such var-
iables as moisture and temperature from consideration.
Materials and Methods
Needle Litter Decomposition in Natural Stands—
For this study, relatively isolated co-dominant and dominant trees
were selected. Two each of the least and most oxidant-injured Jeffrey
167
-------
pine trees were selected on each of two sites. These sites were Holcomb
Valley ("no" oxidant injury) and Camp Osceola ("moderate" oxidant^injury).
Similar selections of ponderosa pines were made at Barton Flats ("moder-
ate" oxidant injury) and Camp Congo ("moderate-heavy" oxidant injury).
These sites represent the range of oxidant injury to each species on
the study plos in the SBNF. During the autumns of 1974-1976, freshly
fallen litter was sampled randomly beneath each selected tree and subsam-
ples of approximately 15 gm were made at random. One random 20-gm subsam-
ple from each tree per year was analyzed by Drs. Gersper and Arkley for
nutrient content. An additional random sample was dried to 30 C, allowing
calculation of fresh/dry weight ratio for the entire sample. The remaining
subsamples were placed in labeled nylon mesh (3 mm) envelopes (approximately
15 cm x 30 cm) and disbursed as related in Figure 45. Each arrow represents
similar treatments consisting of 5, 10, and 30 envelopes left in the field
for one winter, one year, and two years, respectively- The 1974 samples
comprised the two-year experiment; the 1975 samples comprised the one-year
and first one-winter experiment. The first one-winter experiment was re-
peated, and the second involved the exchange of ponderosa and Jeffrey pine
litter between their respective sites (Fig. 46). The rationale for this is
explained in the following discussion.
After each treatment was completed, envelopes were retrieved. The
needles were: (1) brushed lightly to remove excessive inorganic and
fungal materials; (2) dried to a constant weight at 30 C and (3) weighed.
A composite sample of the retrieved litter for each treatment in
the two, one and first overwinter experiments was analyzed for percent
content of N, P, K, Ca and Mg by Drs. Gersper and Arkley. Percent change in
weight and nutrient content were then calculated.
Among the environmental variables which may influence litter de-
composition are solar radiation, litter temperature, litter moisture
and litter depth. Although litter moisture has not been measured, it
is felt that careful interpretation of soil moisture depletion curves
and precipitation, soil moisture and litter temperature data will permit
the determination of the relative litter moisture relations between the four
plots under consideration. Litter depth measurements have been taken for
comparison between trees and plots. Solar radiation and litter temperature
were measured once per hour, at four locations beneath each study tree from
sunrise to sunset on a cloudless day in late August and early September.
Solar radiation was measured with a Weather Measure Model R401 Mechanical
Pyranograph. Temperature in the upper one cm of the organic horizon was
measured with a calibrated mercury centigrade thermometer. Each measurement
point corrsponded to the midpoint of one quarter of the arc along which
decomposition envelopes were placed beneath each tree.
Envelopes of litter from the two and one year experiments were
rated visually for activity in 14 categories of signs of decomposition.
The total length of needles on each two and one year envelope was esti-
mated by (1) counting the number of fascicles in each envelope, (2)
determining the mean length per fascicle for one envelope per treatment
and (3) multiplying this mean value by the number of fascicles in each
168
-------
Pinus Jeffrey!
Holcomb Valley
Camp Osceola
Pinus ponderosa
Camp Oongo
Barton Flats
1974 oxidant score (Go868 was killed by bark beetles and replaced in
this study by Si449.)
Figure 45. Source and destination (tree tag-1976 oxidant score) of 960
decomposition study envelopes. Each arrow represents 30
envelopes and points from their source to their destination.
169
-------
Pinus Jeffrey!
Pinus ponderosa
Holcomb Valley
Barton Flats
Figure 46. Source and destination (tree tag-1976 oxidant score) of 160
decomposition study envelopes. Each arrow represents 5 en-
velopes and points from their source to their destination.
170
-------
envelope per treatment. Because each envelope per treatment constituted
a random subsample from the same source, and each envelope contained
approximately the same initial weight of needles (15 g), it was felt
that the estimated total length of needles per envelope reflected their
relative surface area.
All of the above data types will be examined for meaningful rela-
tionships to decomposition.
Decomposer microorganisms populations on pine needles in natural stands
experiencing different oxidant doses—Oxidant air pollution may alter
the rate of pine needle decomposition by affecting the composition of
microbial populations on senescing and fallen needles. To study the
succession of pine needle microorganisms, two lines of investigation were
followed in the field. Microbial succession in living needles was deter-
mined by isolation of fungi from surface-sterilized needles of various
ages, while succession in litter was determined by isolating fungi from
surface-sterilized needles that were on the forest floor for varying
periods of time.
Before litter fall in 1974-1976, one square meter of nylon mesh
was placed beneath each of the trees involved in the integrated field
needle compositon study at approximately two-thirds of the crown radius out
from the stem. In 1974, four trees were tagged at each of two locations in
the University of California Blodgett Experimental Forest, El Dorado County,
California. They were tagged GR 1-8, and each received a nylon mesh square
prior to litter-fall in 1974 and 1975. In 1975, four Jeffrey and four
ponderosa pines were selected for study and tagged GR 9-16 on the Stanislaus
National Forest, near Pinecrest, California. Each received mesh squares
prior to litter-fall in 1975. It was felt that pine stands outside the SBNF
must be considered for comparison in terms of air pollution impact. Having
separated annual increments of litter-fall in this manner, we collected
periodic samples of litter from these nets, surface-sterilized them, and
incubated them on water agar in petri dishes. The populations of micro-
organisms were then recorded.
To determine the succession of microrganisms on living pine needles,
the lowest healthy twigs on the north, south, east and west sides of
the stem were clipped not only from trees involved in the integrated
field needle decomposition study, but also from trees at Blodgett and
Pinecrest. The annual needle increments on each twig were separated in the
field. A subsample of each increment was surface-sterilized and incubated
on water agar in petri dishes. The populations of microorganisms were
then recorded.
For the purposes of this study, a community will be considered to
consist of the microbial population inhabiting (1) the needles of a
given age on a given twig, or (2) the needles representing a single
annual increment of litter-fall beneath a specific tree. Each community
will be characterized by an ordered pair of values (S, e) representing
taxonomic richness and evenness (Williams, 1977) and a "species" list
171
-------
ranked by abundance. Taxonomic richness is represented by the number
of taxonomic categories isolated from each community. Taxonomic evenness
is defined as the distance from the point of perfect evenness to the
point represented by a given community in n-space. Figure 47 illustrates
the concept of evenness in a three species community. Mathematically,
S ±2 - 1
1-1
S-l , where
S = the number of taxonomic categories ("species")
in the community, and
s
TT = the species frequency vector such that .2 = 1-
This degree of data summarization will permit comparison of similar
communities in different situations.
Laboratory tests of the effect of oxidant dose on decomposers and de-
composition—This phase of the project is designed to determine who
air pollution affects (1) growth and reproduction of microbial agents
of litter decomposition, and (2) rates of needle decomposition by major
microorganisms.
Two clear and twelve opaque plexiglass fumigation chambers were
constructed and installed inside two walk-in Percival growth chambers
on the Oxford Tract, UCB. These walk-in chambers have been renovated to
permit control of light, temperature, and relative humidity. The plexi-
glass chambers permit control of ozone concentration.
Species of fungi isolated from litter samples will be fumigated
at a number of ozone concentrations. The fungus will be inoculated
onto sterilized pine needle sections placed on cellophane-covered cellu-
lose agar (Eggins and Pugh, 1962) in petri dishes. The effects of ozone
on such factors as (1) colony growth rate, (2) spore production, (3)
spore germinability, and (4) cellulose decomposition will be quantified.
The ability of a microflora to affect decomposition might be altered
by oxidant air pollution. To test this possibility, preweighed, steri-
lized pine needles from the Sanislaus N.F. were inoculated by placement on
a mixture of the organic horizons from BF, CAO, COO, HV, and the Stanis-
laus N.F. Moistened by constant subirrigation, one-half of the experiment
was exposed to ozone-enriched air. Weight loss was determined after
14 weeks.
Over a prolonged period of exposure, microbial populations might be
altered by oxidant air pollution in ways which affect their ability to
172
-------
(0,1,0)
(0,0,1)
(1,0,0)
Figure 47. Evenness (£} in a three species community.
173
-------
decompose litter. In this experiment, preweighed sterilized pine needles
from the Stanislaus N.F. were inoculated by placement on mixtures of
the organic horizons from BF, CAO, COO, HV or the Stanislaus N.F.
Moistened by subirrigation, the entire experiment was conducted in fil-
tered air. Weight loss was determined after 22 weeks.
Experiments have been designed which will test (1) the ability^of
a diverse microflora to decompose needles produced by healthy and di-
seased trees and (2) the effects of moisture and temperature on litter
decomposition.
Results and Discussion
Needle Litter Decomposition in Natural Stands—
Quantification of integrated needle litter decomposition in the
field has been concluded. Data sets are complete for weight losses
incurred by 320, 320, 160, 320 mesh litter envelopes over two years,
one year, one winter and the next winter, respectively. The 32 treat-
ments involved in each of these experiments are diagrammed in Figure
45. The 32 interspecific treatments involved in the second overwinter
experiment are diagrammed in Figure 46.
Ponderosa pine litter lost more weight («; = 0.001) during the
two year, one year, and first overwinter experiments than did Jeffrey
pine litter (Figure 48). Within each species, greater weight loss occur-
red on the site receiving the greatest oxidant air pollution dosage.
To determine whether the difference in decomposition rate between
ponderosa and Jeffrey pines was due to some species-specific factor or
an environmental factor, litter envelopes were placed on plots of the
opposite species to decompose. Data from this overwinter experiment
(Fig. 49) showed that the species and source of litter was not signifi-
cantly related to decomposition while the site of decomposition was. It
is felt that the overwinter weight losses experienced in these experiments
were relatively light, probably reflecting the below-normal precipitation.
The sparse literature relevant to this point suggests that weight and
nutrient losses by pine needle litter during its first winter on the
ground are relatively great (Stark, 1972, 1973; Millar, 1974).
Change in nutrient status (per cent N, P, K, Ca, Mg on a dry weight
basis) has been calculated for a pooled sample from each treatment in
the two year, one year, and first overwinter experiment. Though both species
ultimately gained calcium and nitrogen, ponderosa pine litter consistently
acquired both nutrients at a greater rate than did Jeffrey pine litter.
This may correspond to the greater fungal activity and subsequent decomposi-
tion observed in ponderosa pine litter.
Measurement points at HV received the greatest total daily radiation
and temperature input (Table 54), while those at COO received the least.
CAO and BF lie intermediate, CAO received more radiation but approximately
the same temperature input as BF. This gradient matched the oxidant air
174
-------
JEFFREY PINE
PONDEROSA PINE
CAMP OSCEOLA HOLCOMB VALLEY
-MODERATE -HEALTHY
CAMP CONGO
-SEVERE
BARTON FLATS
-MODERATE
2 YEARS
10.84 > 9.41
1 YEAR
11.61 >* 9.43
0.5 YEAR
2.80 > 2.10
2 YEARS
19.99 >**
1 YEAR
14.61 >*
0.5 YEAR
8.86 >**
14.64
11.62
5.79
Figure 48. Percent weight loss (30 C) incurred by needle litter on four
plots, representing the range of air pollution impact on
ponderosa and Jeffrey pine. *-significant at 10% level;
**-significant at 5% level; and ***-significant at 1% level.
175
-------
JEFFREY PINE PONDEROSA PINE
0.5 YEAR
HV COO (8.85) BF COO (9-86)
BF (3.69) BF (4.27)
CAO (7.08) CAO (6.40)
HV (4.07) HV (4.27)
CAO COO (8.89) COO COO (9.60)
BF (3.57) BF (5.57)
CAO (3.58) CAO
HV (2.45) HV (5.27)
Figure 49. Percent weight loss (30 C) incurred by needle litter following
transfer between and within species. Transfers made according
to plant shown in Figure 48.
176
-------
TABLE 54. ESTIMATED TOTAL RADIATION (R) AND TEMPERATURE (T) ACCUMULATED
DURING ONE CLEAR DAY AT MEASUREMENT POINTS BENEATH THE INTE-
GRATED FIELD DECOMPOSITION STUDY TREES.
Species Plot
Jeffrey HV
Jeffrey CAO
Ponderosa BF
Ponderosa COO
X
Tree
1537
1561
1574
1598
X
1865
1934
1964
807
X
2600
2625
2755
X
852
875
894
449
X
R
107.53
292-14
253.91
400.57
263.54
167.59
213.93
211.12
159.14
187.95
182.03
258.96
126.06
189.02
202.04
158.33
79.47
133.44
143.32
166.17
T
15724.4
22919.4
21113.6
24984.1
21185.4
16755.5
18626.5
15466.0
15750.0
16649.5
15988.9
20362.2
14181.12
16844.08
17864.9
17477.1
13917.4
16978.6
16559.5
16701.8
177
-------
pollution, decomposition and species gradients determined for these four
plots. This suggests that radiation (and temperature) might be inversely
correlated with weight loss. Correlation analysis, however, showed that the
only significant within-plot relationship (<* = 0.05) between litter weight
loss and either radiation or temperature occured at HV. Further, at HV
litter weight loss was directly proportional to both incident radiation and
temperature. The elevation of HV is the highest of the four plots studied
and depends heavily on winter precipitation for moisture. Perhaps points
receiving greater radiation at HV were exposed for a greater length of time
to temperature and moisture conditions conducive to decomposition.
Decomposer Microorganism Populations on Pine Needles in Natural Stands
Experiencing Different Oxidant Doses—
One complete experiment during late summer of 1975 provided information
on microbial succession in pine foliage in the SBNF, Blodgett Forest, and
the Stanislaus N.F. Data from a similar experiment in the SBNF during
spring, 1976, will be analyzed shortly. One additional experiment of this
type is in progress. In this experiment, it is hoped that nutrient status
will be determined by Drs. Arkley and Gersper for each annual increment of
foliage. Results of the first experiment are tentative, requiring further
analysis and confirmation. Collection of data on microbial succession in
litter continues. The value of the data increases from each successive
sampling; there are now four annual increments of litter-fall separated
(except where nets have been vandalized) in the SBNF. All data collected
from these studies will be interpreted in the light of the existing environ-
mental data.
Through the study of annual increments of living foliage and strati-
fied litter, successions of microfloral populations are being determined
for individual trees representing plots and regions impacted by varying
amounts of smog. Comparisons among these population successions will
help to explain patterns in (1) litter decomposition and (2) the inci-
dence of fungus-caused damping-off of pine seeds and seedlings (under
study by the seedling establishment investigators). Species of fungi
for which population data have been collected, to date, include phyco-
mycetes, ascoymcetes, basidiomycetes, and fungi imperfect!. A method
for the culture of fungi on microscope slides has been employed for the
grouping and identification of important isolates (Riddell, 1950).
Laboratory Tests of the Effect of Oxidant Dose on Decomposer and Decom-
position—
Two growth chamber decomposition experiments have been completed.
The data from these experiments will be analyzed shortly and both will
be repeated. The first experiment tested the ability of BF, COO, CAO,
HV, and Stanislaus N.F. litter microorganisms to decompose a standard
sterile needle litter in filtered air. The second experiment tested
the effect of ozone on the decomposition of standardized litter by the
organisms in a standardized mixture of the organic horizons from BF, COO,
CAO, HV and the Stanislaus N.F.
178
-------
REFERENCES
Asher, J. E. 1956. Observations and theory on "X" disease or needle
dieback. File Rep. Arrowhead Dist., San Bernardino Natl. For.,
Calif.
Bingham, G. E. and P. I. Coyne. 1977. A Portable, Temperature Control-
led Steady State Porometer for Field Measurements of Transpiration
and Photosynthesis. Photosynthetica 11(1):148-160.
Black, C. A. (ed.). 1965. Methods of Soil Analysis Agronomy No. 9,
Part II. Madison, Wis., Amer. Soc. Agronomy, Inc. pp. 771-1572.
Botkin, D. B., J. F. Janak and J. R. Wallis, 1972- Some Ecological Conse-
quences of a Computer Model of Forest Growth. Journ Ecol. 60:849-872*
Braun-Blanque, J. 1932* Plant sociology. Hafner. N.Y. pp. 22-25.
Byler, J. W., D. R. Hart and R. E. Wood. 1976. Project plan: Biologi-
cal evaluation of insect and disease caused mortality in the San
Bernardino National Forest. U.S.F.S. Region 5. San Francisco,
CA Mimeo 68 p.
Cobb, F. W., Jr., D. L. Wood, R. W. Stark, and J. R. Parmeter, Jr. 1968.
Photochemical oxidant injury and bark beetle (Coleoptera: Scolytidae)
infestation of ponderosa pine. IV. Theory on the relationship between
oxidant injury and bark beetle infestation. Hilgardia 39(6):141-152.
Cobb, J. W., Jr., et al. 1974. Effects of photochemical air pollutants
on the epidemiology of root pathogens in the mixed conifer forests.
In: Oxidant air pollutant effects on a western coniferous forest
ecosystem. Annual Progress Report 1973-74. 0. C. Taylor (ed.)
Statewide Air Pollution Research Center. Univ. of Calif. Riverside,
CA Section III-E.
Corn, M., R. W. Dunlap and L. A. Goldmuntz. 1975. Photochemical oxidants:
sources, sinks and strategies. J. Air Poll. Contr. Assoc. 25(1):
16-18.
Coyne, P. I. and G. E. Bingham. 1978. In Annual Progress Report on the
Studies of Air Pollution Effects on Vegetation, Lawrence Livermore
Laboratory, pp. 12-20.
Dahlsten, D. L. 1974. The role of bark beetles, Dendroctonus spp., in
the succession of the mixed conifer ecosystem exposed to photochemical
air pollutants, Section III-C. In: Annual Progress Report 1973-74,
179
-------
References Continued
for the U. S. Environmental Protection Agency, Corvallis, Or. 10
pp. and Appendix.
Dahlsten, D. L., C. J. DeMars, D. L. Rowney, and P- A. Felch. 1974.
Sampling techniques and preliminary analyses for the western pine
beetle in the central Sierras of California. (Unpublished manu-
script, in preparation).
Dahlsten, D. L., and D. L. Rowney. 1974. The influence of Verticicladiella
wagenerii on western pine beetle, Dendroctonus brevicomis, populations.
(Unpublished progress report, Western Regional Project W-110). 5
pp.
Dahlsten, D. L. 1977. Western pine bark beetle population dynamics—
stand tree mortality subsystem, pp. 159-176. In: Photochemical
oxidant air pollutant effects on a mixed conifer forest ecosystem—
a progress report. P. R. Miller and M. J. Elderman, editors. Con-
tract Nos. 68-03-0273 and 68-03-2442. U. S. Environmental Protection
Agency, Corvallis Environmental Research Laboratory, Oregon 338-
pp.
Dickinson, C. H., and G. J. F. Pugh. 1974. Biology of Plant Litter
Decomposition, Vols. I and II. Academic Press, New York. 775 pp.
Dixon, W. J. 1975. BMDP Biomedical Computer Programs. University of
California Press, Berkeley, California 792 p.
Duff, G. H. and N. J. Nolan. 1953. Growth and morphogenesis of the
Canadian forest species. I. The controls of cambial and apical
activity in Pinus resinosa. Can. J. Bot. 31:471-513.
Eggins, H. 0. W. and G. J. F. Pugh. 1962. Isolation of Cellulose-decompos-
ing Fungi from the Soil. Nature 193:94.
Fry, K. E. and R. B. Walker. 1967. A Pressure Infiltration Method for
Estimating Stomatal Opening in Conifers. Ecology 48:155-157.
Horton, J. S. I960. Vegetation types of the San Bernardino mountains.
U.S. For. Ser., U.S.D.A. Tech. Paper No. 44. 29 pp.
Jackson, N. L. I960. Soil Chemical Analysis. Prentice-Hall Inc., Engle-
wood Cliffs, N.J. 498. pp.
James, R. L. et al. 1977. Root pathogen dynamics and stand mortality
subsystem. In: P. R. Miller and M. J. Elderman (eds.). Photochemical
oxidant air pollutant effects on a mixed confier forest ecosystem—
a progress report. Contract Nos. 68-03-0273 and 68-03-2442 (EPA),
Univ- of Calif., Riverside, CA. pp. 177-185.
Johnson, C. M. and A. Ulrich. 1959. II. Analytical Methods for Use
180
-------
References Continued
in Plant Analysis. Calif* Agr. Exp. Sta. Bull. 766. 77 pp.
Kickert, R. N. 1977. Definition of the Conifer Forest Ecosystem as a Group
of Coupled Ecological Models. In: Photochemical Oxidant Air Pollutant
Effects on a Mixed Conifer Forest Ecosystem. Miller, Paul R. (ed.)«
U.S. Environmental Protection Agency, Corvallis, Oregon. 71-105.
Kickert, R. N., P. R. Miller, 0. C. Taylor, J. R. McBride, J. Barbieri,
R. Arkley, F. W. Cobb, Jr., D. L. Dahlsten, W. W. Wilcox, J. M. Wenz,
J. R. Parmeter, Jr., R. F. Luck and M. White. 1976. Photochemical
air pollutant effects on mixed conifer ecosystems: a progress report.
U.S. Environmental Protection Agency Office of Res. and Dev. Corvallis
Environ. Res. Lab., Corvallis, Oregon. CERL-026. 275 pp.
Kozlowski, T. T* and C. H. Winget. 1964. The Role of Reserves in Leaves,
Branches, Stem and Roots on Shoot Growth of Red Pine. Amer. J. Botany
51:522-529.
Larsen, R. A. and W. W. Heck. 1976. An Air Quality Data Analysis System
for Interrelating Effects Standards and Needed Source Reductions.
Part 3. Vegetation Injury. J. Air Pollut* Contrl Assoc. 26:325-333.
Luck, R. F. Cone and seed production for dominant conifer tree repro-
duction. In: P. R- Miller and M. J. Elderman (eds.). Photochemical
Oxidant Air Pollutant Effects on a Mixed Conifer Forest Ecosystem.
a progress report* U. S. Environmental Protection Agency, Office of
Research and Development. Corvallis Environmental Research Laboratory
CERL EPA-600/3-77-104: 338 pages.
MATHTECH. 1976. Computer Simulation Model for Analyzing Mobile Source
Air Pollution Control Strategies. U. S. Environmental Protection
Agency, 600/5-76-010, Washington, D.C. 180 p.
McBride, J. 1973. Vegetation of principle study sites in the San Bernar-
dino mountains. In: Oxidant air pollutant effects on a western
coniferous forest ecosystem. Task C Report. Statewide Air Pollution
Research Center, Univ. of Calif-, Riverside, CA.
McBride, J. R. 1974. Annual report of the vegetation sub-committee.
In: Oxidant air pollutant effects on a western coniferous forest
ecosystem. Annual Progress Report 1973-74. 0. C. Taylor (ed.).
Statewide Air Pollution Research Center, Univ. of Calif., Riverside,
CA Section III-B.
McBride, J. R., V. Semion, and P. R. Miller. 1975. Impact of air pollution
on the growth of ponderosa pine. Cal. Agr. 29(12):8-10.
McBride, J. R. 1977. Tree population dynamic subsystem. In: P- R.
Miller and M. J. Elderman (ed.)« Photochemical oxidant air pollutant
effects on a mixed conifer forest ecosystem—a progress report.
181
-------
References Continued
Contract Nos. 68-03-0273 and 68-03-2442 U.S. Environmental Protection
Agency, Univ. of Calif., Riverside, CA pp. 106-121.
McCutchan, M. H. and M. J. Schroeder. 1973. Classification of Meteoro-
logical Patterns in Southern California by Descriminant Analysis.
J. Appl. Meteorol. 12:571-577.
Millar, C. S. 1974. Decomposition of Coniferous Leaf Litter. In: Biology
of Plant Litter Decomposition. Dickinson, C. H. and C. J. F. Pugh
(eds.) Academic Press, New York. pp. 105-128.
Miller, J. M., and F. P. Keen. 1960. Biology and control of the western
pine beetle. A summary of the first fifty years of research. U.
S. Department of Agriculture Forest Service, Washington, D.C. Misc.
Publ. No. 800. 381 pp.
Miller, P. R., J. R. Parmeter, Jr., 0. C. Taylor and E. A. Cardiff. 1963.
Ozone injury to the foliage of ponderosa pine. Phytopath. 53:1072-
1076.
Miller, P- R. 1973. Oxidant damage to conifers on selected sites, 1972.
In: Taylor, 0. C. Oxidant air pollution effects on a western coni-
ferous forest ecosystem. Task C Report. Statewide Air Pollution
Research Center. Univ. of Calif*, Riverside, CA.
Miller, P- R. 1973. Susceptibility to ozone of selected western conifer.
2nd Inter. Cong. Plant Pathology, Minneapolis, Abstract 0579.
Miller, P- R. and J. R. McBride. 1973. Vegetation committee report
(Sec. A). In: Oxidant air pollutant effects on a western coniferous
forest ecosystem. Task B Report. Statewide Air Pollution Research
Center. Univ. of Calif., Riverside, CA.
Miller, P. R. and J. R. McBride. 1975. Effects of air pollution on
forests, pp. 175-235. In: Mudd, J. B. and T. T. Kozlowski (eds.),
Responses of Plants to Air Polluton. Academic Press, N.Y.
Miller, P. R. 1977. Oxidant flux canopy response subsystem. In: P. R.
Miller and M. J. Elderman (eds.). Photochemical oxidant air pollutant
effects on a mixed conifer forest ecosystem—a progress report.
Contract Nos. 68-03-0273 and 68-03-2442. U. S. Environmental Protec-
tion Agency. Univ. of Calif., Riverside, CA pp. 122-138.
Miller, P. R., R. N. Kickert, 0. C. Taylor, R. J. Arkley, F. W. Cobb,
Jr., D. L. Dahlsten, P. J. Gersper, R. F. Luck, J. R. Parmeter, Jr., J.
M. Wenz, M. White and W. W. Wilcox. 1977. Photochemical oxidant
air pollutant effects on a mixed conifer forest ecosystem—a progress
report, 1975-1976. U.S. Environmental Protection Agency Contracts Nos.
68-03-0273 and 68-03-2442. Corvalis Ecological Research Laboratory,
Corvallis, Oregon EPA-600/3-77-104: 338 pp.
182
-------
References Continued
Minnich, R. D. et al. 1969- Mapping montane vegetation in southern
California. Status Report 3, Contract No. 14-08-0001-10674. U.
S. Dept. Interior, Tech. Report 3, Dept. Geog. Univ. Calif., Riverside,
CA 40 pp.
Minnich, R. A. 1974. The impact of fire suppression on southern California
conifer forests: a case study of the Big Bear fire, November 13-16,
1970. In: Symposium on living with the chaparral proceedings. Sierra
Club. San Francisco, CA pp. 113-126.
MuKammal, E. I. 1965. Ozone as a Cause of Tobacco Injury. Agr. Meteorol.
2:145-165.
Nie, N. H., C. H. Hull, J. G. Jenkins, K. Steinbrenner and D. H. Bent.
1975. SPSS Statistical Package for the Social Sciences. McGraw-
Hill Book Co., New York. 675 p.
Parmeter, J. R., Jr., R. V. Bega and T. Neff. 1962. A chlorotic decline
and death of ponderosa pine in southern California. Plant Dis. Rep.
46:269-273.
Reed, K. L. and R. S. Waring. 1974. Coupling of Environment to Plant
Response: A Simulation Model of Transpiration. Ecology 55:62-72.
Reed, K. L. 1976. Prediction of Tree Growth Response Surfaces in an
N-Dimensional Habitat Coordinate System. (in prep.).
Reed, K. L. and S. G. Clark. 1976. Succession SIMulator: A Coniferous
Forest Simulator. Model Documentation. Coniferous Forest Biome
Bulletin No. 11, University of Washington, Seattle, Wash. 73 p.
Riddell, R. W. 1950. Permanent Stained Mycological Preparations Obtained
by Slide Culture. Mycologia 42:265-270.
Schussel, G. 1977. The role of the data dictionary. Datamation. 23(6):
129, 133, 136, 142.
Sneath, P. M. A. and R. R. Sokal. 1973. Numerical taxonomy. W. H. Freeman
and Company. San Francisco. 573 p.
Sokal, R. R. and F. J. Rohlf. 1969. Biometry. W. F. Freeman and Co.
San Francisco. 776 pp.
Sollins, P., R. H. Waring, and D. W. Cole. 1974. A Systematic Framework
for Modeling and Studying the Physiology of a Coniferous Forest
Biome. Bulletin No. 5. Waring, R.H. and R. L.Edmonds (eds.). Univer-
sity of Washington, Seattle, Wash. p. 7-20.
Stark, R. W., P. R. Miller, F. W. Cobb, Jr., D. L. Wood, and J. R. Parmeter.
1968. Photochemical oxidant injury and bark beetle (Coleoptera;
183
-------
References Continued
Scolytidae) infestation of ponderosa pine. I. Incidence of bark beetle
infestation in injured trees. Hilgardia 39(6):121-126.
Stark, N. 1972. Nutrient cycling pathways and litter fungi. Bioscience
22:335-360.
Stark, N. 1973. Nutrient cycling in a Jeffrey pine ecosystem. Montana
Forest and Conservation Experiment Station. 389 pp.
Thompson, D. R. and T. M. Hinckley. 1977. A simulation of water relations
of white oak based on soil moisture and atmospheric evaporative
demands. Canad. J. Forest. Res. 7:400-409.
Westman, W. E. 1977. How much are nature's services worth? Science.
197:960-964.
Wilkinson, B. 1977. An application audit. Datamation. 23(8):51-54..
Williams, C. B., Jr. 1967. Spruce budworm damage symptoms related to
radial growth of grand fir, Douglas-fir and Engelman Spruce. For.
Sci. 13(3):274-285.
Williams, F. M. 1977. Model-free evenness: An alternative to diversity
measures. Paper S23-6. 4 pp. presented 8/8/77 to Second International
Ecological Congress, Satellite Program in Statistical Ecology, Satel-
lite A: NATO-Advanced Study Institute and ISEP Research Workshop.
Wood, D. L. 1971. The impact of photochemical air pollution on the
mixed conifer forest ecosystem arthropods. In: Oxidant air pollutant
effects on a western coniferous forest ecosystem. Task B Report:
Historical Background and Proposed Systems Study of the San Bernardino
Mountain Area- U.S. Environmental Protection Agency, Corvallis,
Oregon, Section C. 26 pp.
Wood, D. L. 1973. The impact of photochemical air pollution on the mixed
conifer forest ecosystem arthropods. In: Oxidant air pollution
effects on a western coniferous forest ecosystem. Task B Report.
Historical Background and Proposed Systems Study of the San Bernardino
Mountain Area. U.S. Environmental Protection Agency, Corvallis,
Oregon. Section C. 26 pp.
184
-------
APPENDIX 1. PART II OF THE SBNF DATA DICTIONARY: DEFINITION OF DATA-
SET NAMES.
Dataset
name
Definition of data content
SBNFPUBS
PLOTINDX
SXSCAT
STEXOM
TREE
STAGE
SHRUB
TREEVEG
SHRUBVEG
PLOTREGN
FIRETREE
FIRESTAG
Bibliography of publications, manuscripts in preparation,
and intended manuscripts from scientists conducting re-
search under EPA contracts 68-02-0303, 68-03-0273, 68-03-
2442, R805410-01 from 1972 through present.
For each vegetation plot, gives year established, eleva-
tion, geographic coordinates, length, width, azimuth,
general hillslope aspect, tree tag colors, lowest and
highest tree tag numbers, forest type, and number of
tagged trees;
Soil exchangeable and soluble cations, various soil depths
on each vegetation plot;
Soil texture and organic matter; percent of soil in
various texture classes, pH, and organic carbon and nitro-
gen content at various depths on each vegetation plot;
Tree identification; locations and species of tagged
trees on each vegetation plot;
Stand age data; number of trees by ten year age class,
by species, on vegetation plots;
Locations of shrub cover, by species, along a transect
across each vegetation plot;
Tree vegetation data; density, basal area, species com-
position of trees, greater than, or equal to 10 cm diameter
at breast height, on each vegetation plot;
Shrub vegetation data; frequency, density, percent cover,
by species, on each vegetation plot;
Plot regeneration data; age of tree seedlings, saplings,
poles, and age classes of trees in larger size-classes,
by cover types, on each vegetation plot;
Tree data on 85 special study plots which had burned
at various dates;
Fire stand age data; similar to STAGE (above) but on
85 special study plots which had burned at various dates;
185
-------
APPENDIX 1. PART II OF THE SBNF DATA DICTIONARY: DEFINITION OF DATA-
SET NAMES. (CONTINUED)
Dataset
name
Definition of data content
FIRESHRUB
CTREE
FSMTINDX
FSMET
HMET
HPREC
OXIDINDX
OXIDANT
PLOTMET
PLOTPREC
PLOTOXID
OZFLUX
Fire shrub data; similar to SHRUBVEG (above), but on
85 special study plots which had burned at various dates;
Crown tree data, such as Keen crown class, crown position
in canopy, gross geometrical crown volume, tree height,
and crown ratio, for each tagged tree on each vegetation
plot;
Forest Service meteorological data index; days, by month
and year, in FSMET dataset, which have valid usable data;
Forest Service meteorological data which were telemetered
from 3 sites near the vegetation plots; net radiation,
air temperature at 4 ft and 6 ft, relative humidity, wind
direction, wind speed, hourly resolution;
Historical meteorological data, incuding relative humidity,
wind direction, wind speed, air temperature maximum and
minimum on a daily basis, for sites at various distances
from vegetation plots;
Historical precipitation data on monthly and annual basis
for 23 sites operated by the San Bernardino County Flood
Control District, and ranging as far back as 92 years;
Oxidant index; days, by month and year, which have valid
data in OXIDANT dataset;
Hourly data on oxidant and ozone concentration at sites
in the vicinity of vegetation plots;
Plot meteorological data on air temperature and relative
humidity from hygrothermographs located on the vegetation
plots;
Plot precipitation data from snow gauges and summer rain
gauges located on each vegetation plot;
Plot oxidant ambient concentration data obtained during
various time periods on vegetation plots;
Ozone flux data for sample conifer trees near a vegeta-
tion plot;
186
-------
APPENDIX 1. PART II OF THE SBNF DATA DICTIONARY: DEFINITION OF DATA-
SET NAMES. (CONTINUED)
Dataset
name
Definition of data content
STOMRES
INJBIWK
TREE
TREEMORT
SAPTREE
SAPGRO
SAPSURF
PNFALL
TREELIT
PLITR
LITMAS
TREEGRO
TREESOIL
Stomatal resistance data for sample conifer trees near
a vegetation plot;
Air pollution- conifer tree foliar injury on a biweekly
basis for sample trees near a vegetation plot; needle
retention, needle length, injury length;
Annual foliar injury observations on each tagged tree on
each vegetation plot; needle retention score, needle
condition score, branch mortality score, needle length
score, total smog injury score, disease observation, insect
risk category, tree diameter at breast height;
Tree mortality data for each vegetation plot;
Sapling tree data, analogous to the tree dataset (above),
but for sapling plots located very near the vegetation
plot;
Sapling height growth data on an annual basis on the
sapling plots;
Sapling foliage surface data;
Pine needle fall from tree crowns on an annual basis,
beneath selected pine trees on all but 2 of the vegetation
plots;
Tree litter thicknesses on the ground in 4 directions
at various distances from the tree trunk, under selected
trees on selected plots;
Plot litter thickness at 2 m intervals down the center-
line of each vegetation plot in 1973;
Litter mass (dry weight) at various distances from trunks
of selected trees on selected vegetations plots;
Tree growth data, cumulative radial, for conifer trees
at least 10 cm diameter, breast height, on 6 vegetation
plots representing extremes of air pollution exposure;
Soil type, hillslope gradient, and hillslope aspect for
each tree in the TREEGRO dataset;
187
-------
APPENDIX 1. PART II OF THE SBNF DATA DICTIONARY: DEFINITION OF DATA-
SET NAMES. (CONTINUED)
Dataset
name
Definition of data content
TREEGR02
BOGRO
MOIST
MATRIC
ISURV
BTREE
EGG
REAR
STIK
XRAY
DISU
FASP
Amalgamation of TREESOIL dataset with TREEGRO, and con-
verted to give annual radial stem growth, at breast
height, as far back as year 1920;
Black oak growth data, similar to TREEGRO dataset des-
cribed above;
Soil water percentage, by weight, at various depths at
weekly, or biweekly, intervals for 22 sites, including
the vegetation plots;
Soil water MATRIC potential at various depths for data
from the MOIST dataset, converted by means of soil water
retention curves determined in the lab, using respective
field soil samples;
Insect survey data taken in early summer and late fall,
from tagged trees on vegetation plots;
Beetle tree data; infestation heights and stem circum-
ferences at sampling heights for western pine bark beetle
on killed trees;
Western pine bark beetle attack densities, gallery lengths,
egg counts, from bark disc samples off of killed trees;
Emergent bark beetle densities, parasites, and predator
data from bark field samples reared under laboratory
conditions;
Emergent bark beetle densities, parasites, and predator
data from bark into sticky cartons on tree trunks under
field conditions;
Potential bark beetle brood, as determined from XRAYs
of bark sample discs;
Disease survey data; plant diseases found on tagged
trees on vegetation plots;
Fomes annosus spread plots data;
188
-------
APPENDIX 1. PART II OF THE SBNF DATA DICTIONARY:
SET NAMES. (CONTINUED)
DEFINITION OF DATA-
Dataset
name
Definition of data content
SPRMORT1
SPRMORX
STNDSITE
TREEPEST
SPRXTREE
SPRXRGNS
SPRXRGNC
SPRXGRND
SPRXSHRB
SPRMORT2
CONE
PLOTSEED
Super plot mortality dataset #1; data on dead and severely
damaged trees, obtained in cooperation with the Pest
Damage Inventory, U.S. Forest Service, Region 5, San
Francisco;
Super plot mortality extra data not part of the USFS/PDI
(above); smog score components, basal area detail, addi-
tional insect related symptoms, recent annual stem radial
growth rate, height to lowest live branch bearing needles,
all on dead/damaged tree mortality centers;
Stand site data for trees around each dead/damaged tree
mortality center in a superplot;
Tree morphological and growth data, and insect and disease
data, for each tree in the stand surrounding a mortality
center;
Super plot extra tree data for trees within a 30 m x 30
m plot around each mortality center;
Super plot extra regeneration tree data in square plots
(10 m x 30 m) upslope from each mortality center;
Super plot extra regeneration tree data in circular plots
(6.5 m diameter) around mortality centers and paired
living trees;
Super plot extra ground cover data in 3 plots (10 m x
30 m) upslope from each mortality center;
Super plot extra shrub intercept data along 2 lines
(30 m) upslope from each mortality center;
Super plot mortality dataset #2; data on dead and damaged
trees obtained from superplots located near air quality
and meteorological monitoring sites; done separately
from SPRMORT1 (above);
Annual cone counts from trees on vegetation plots;
Average seed production per cone for cone-bearing trees
on each vegetation plot;
189
-------
APPENDIX 1.
PART II OF THE SBNF DATA DICTIONARY:
SET NAMES. (CONTINUED)
DEFINITION OF DATA-
Dataset
name
Definition of data content
SAPS
LSOIL
SSAS
SLSS
FLDECOMP
LITRKEM
DRIP
SFCSOLKM
Seedling air pollution study; numbers of emerged pine
seedlings and seedling mortality from various ground
cover and wildlife exclusion treatments in mini-plot
on selected vegetation plots;
Laboratory determined soil moisture percent by weight
for soil samples taken from seedling mini-plots;
Seedling supplementary animal study; data on proportional
loss of seeds and seedlings associated with different
kinds of wildlife exclusion treatments in mini-plots
on select vegetation plots;
Supplementary litter seedling study; data on seedling
emergence and mortality associated with different litter
depth treatments;
Field decomposition data for pine needles in net bag
samples under select trees on select vegetation plots,
over 1, 2, and 3 year intervals;
Litter chemistry; elemental content of pine needle fall
at various distances from tree stems, beneath select
sample trees on select vegetation plots;
Intercepted precipitation crown drip elemental content,
at various distances from tree stems, beneath select
sample trees on select vegetation plots;
Surface soil chemistry; elemental content of surface
soil at various distances from tree stems, beneath select
sample trees on select vegetation plots.
190
-------
APPENDIX 2. ON-LINE DATASET PROGRESS CHECKLIST FOR SBNF DATA BASE OF AUGUST 31, 1977.
IDENTIFICATION
DATASET
NAME
CRMppITTJO
ouiNr JT uoo
PT OTTWTIY
L Lt\J i. J.1NLJ4V
SXSCAT
STEXOM
TRID
STAGE
SHRUB
TREEVEG
SHRUBVEG
PLOTREGN
FIRETREE
FIRESTAG
FIRESHRUB
*CTREE-FD
CTREE- RD
PRINCIPAL
INVESTI-
GATOR
PROTVfT
IT jxv/ij EJ\J j.
PPD TETT
JTJXVJU Ej\_*.L
ARKLEY
ARKLEY
MCBRIDE
MCBRIDE
MCBRIDE
MCBRIDE
MCBRIDE
MCBRIDE
MCBRIDE
MCBRIDE
MCBRIDE
LUCK
LUCK
DATA
(MONTH)
YEAR
HKFPTr
WiN V_*Ij
ONCE
ONCE
ONCE/ 73
ONCE
ONCE
ONCE
ONCE
1977
ONCE
ONCE
ONCE
ONCE/ 75
ONCE/ 75
NEW
DATA
D A
A P
T F P
A 0 R
R 0
F M V
0 A E
R T D
M ?
Y
D
E
S E
C N
R T
I E
P R
T E
0 D
R ?
9
N
N
N
N
N
N
N
N
N
Y
Y
K
E
Y
P
U
N
C
H
E
D
?
Y
1
«
•
•
N
N
RPT
RPT
12/77
N
N
N
M10
M10
VERIFICATION
D
A
U T
N A
V
E E
R N
I T
F E.
I R
E E
D D
?
Y
Y
Y
10/77
UCR
V
E
R D
I E
F B
I U
C G
G
P E
R D
0 ?
G
R
DATA
LISTING
TO
P.I,
(.DATE)
Y
Y
5/77
AT UCR
C
0
R E
R N
E T
C E
T R
I E
0 D
N ?
S
N
N
M10
N10
ANALYSIS
R
E D
D E
U P F
C L I
TAN
I N E
0 D
N
?
N
N
N
R
E
D
U
C
T
R.
P A
R N
0 ?
G
R
INFORMATION
TRANSFER
ANALY-
SIS
OUTPUT
TO
P.I.
?
D
0
C
U
M M U
0 E P
D N D
ETA
L A T
T E
I D
0 ?
N
I
N
T
E
W R R
R P E
IRC
TEE
T T I
E A V
N T E
I D
0
N
-------
APPENDIX 2. ON-LINE DATASET PROGRESS CHECKLIST FOR SBNF DATA BASE AS OF AUGUST 31, 1977 (CONTINUED)
FSMTINDX
FSMET
HMET
HPREC
a^VT T^T'NT'n V
UAJ-UIINJUA
OXIDANT
PLOTMET
PLOTPREC
PLOTOXID
OZFLUX
STOMRES
INJBIW1C
TREE
TREEMORT
SAPTREE
SAPGRO
SAPSURF
PNFALL
TREELIT
PLITR
LITMAS
*TREEGRO
MILLER
MILLER
MILLER
ARKLEY
A/fTT T T?T3
WlLiijJiK
MILLER
MILLER
MILLER
MILLER
MILLER
MILLER
MILLER
MILLER
MILLER
MILLER
MILLER
MILLER
ARKLEY
ARKLEY
ARKLEY
ARKLEY
OHMART
75-76
73-76
TO 9/75
10/75-
9/77
1967-76
MAY 77
WINTER? 5
WINTER? 6
MAY 77
JUN 77
?
JUN 76
77
FALL 7 3
74
75
76
77
?
ONCE
1967-76
77
736J4
75
76
77
73/75
ONCE 73
73/74
rris\ o / "7 £
TOo/ 76
m/-\Q / -7 "7
TOo/ / /
— _ — — —
Y
N
N
N
N
N
N
N
N
•
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
.
N
•
Y
Y
Y
N
N
N
•
N
Y
t
.
N
»
N
,
.
N
N
N
Y
N
Y
Y
Y
Y
N
8/77
N
•
9/77
»
N
N
N
.
•
*
*
Y
(
•
UCR
•
UCR
.
*
UCR
UCR
UCR
UCR
UCR
UCR
UCR
UCR
UCR
UCR
UCR
Y
N
Y
Y
Y
•
N
»
AT UCR
AT UCR
,
AT UCR
AT UCR
AT UCR
AT UCR
AT UCR
AT UCR
AT UCR
AT UCR
AT UCR
AT UCR
AT UCR
AT UCR
AT UCR
Y
N
Y
Y
Y
•
Y
Y
Y
UCR
Y
Y
Y
N
N
N
N
N
N
Y
UCR
N
N
N
N
"
N
N
N
N
N
N
N
N
N
N
N
•
-------
APPENDIX 2. ON-LINE DATASET PROGRESS CHECK LIST FOR SBNF DATA BASE AS OF AUGUST 31, 1977 (CONTINUED)
ATRFF^OTT
.L iX.CjJ_i C? \J JL JLt
TREEGR02
*BOGRO-FD
*CMOIST
*POINTS
*GRAV
&MHT QT— TPD
iviu j_ o i r u
MOIST-RD
*H20CURVS
MATRIC
ISURV
BTREE
EGG
REAR
STIK
OHMART
wniiruxj.
OHMART
LAVEN
ARKLEY
ARKLEY
ARKLEY
APVT T7V
/\r\Jxljij j.
ARKLEY
ARKLEY
ARKLEY
DAHLSTEN
DAHLSTEN
DAHLSTEN
DAHLSTEN
DAHLSTEN
fVMPF
VJ1N v»J_j
T08/76
T08/77
12/77
ONCE
ONCE
7
7 *} 7 A
/ -J— / D
77
73-76
77
ONCE
73-76
73
74
75
76
77
73-1/2
74-1/2
75-1/2
76-1/2
73-1/2
74-1/2
75-1/2
76-1/2
73-1/2
74-1/2
75-1/2
76-1/2
73-1/2
74-1/2
75-1/2
76-1/2
.
Y
.
„
N
•
„
Y
.
N
N
N
N
N
N
N
N
.
.
N
•
N
.
N
.
N
B
.
^
,
N
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
N
%
,
•
N
.
N
,
B
,
(
.
N
N
B
.
*
Y
N
.
,
%
*
10/77
a
B
B
.
.
^
.
.
.
.
*
.
i
.
B
.
B
#
4
•
Y
Y
Y
N
N
Y
Y
Y
Y
.
.
*
.
.
.
.
.
.
•
Y
Y
Y
N
N
Y
N
N
N
N
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
N
?
7
?
7
?
7
7
?
7
7
7
7
7
7
7
7
SOME
6/77
-------
APPENDIX 2. ON-LINE DATASET PROGRESS CHECKLIST FOR SBNF DATA BASE AS OF AUGUST 31, 1977 (CONTINUED)
vO
XRAY
DISU
FASP
SPRMORT 1
SPRMORX
STNDSITE
TREEPEST
SPXTREE
SPRXRGNS
SPRXRGNC
SPRXGRND
SPRXSHRB
SPRMORT2
CONE
GCONE
PLOTSEED
SAPS
LSOIL
SSAS
SLSS
FLDECOMP
LITRKEM
DRIP
SFCSOLKM
DAHLSTEN
COBB
COBB
MCB/D/C
MCB/D/C
MCB/D/C
MCB/D/C
MCBRIDE
MCBRIDE
MCBRIDE
MCBRIDE
MCBRIDE
MCB/D/C
LUCK
.
LUCK
LUCK
COBB
COBB
COBB
COBB
PARMETER
ARKLEY
ARKLEY
ARKLEY
73-1/2
74-1/2
75-1/2
76-1/2
JUN 74
JUN 76
?
76
77
77
77
77
77
77
77
77
77
76
111
73-76
77
SPRING78
76
77
77
77
77
?
73-75
76??
74-75
75
10/77
Y
N
Y
Y
N
N
N
N
Y
N
N
N
N
N
N
N
N
N
Y
N
,
N
W
N
N
N
N
N
N
•
N
Y
Y
Y
Y
.
•
N
*
N
N
N
N
N
N
N
N
N
*
N
•
N
N
RPT
N
N
N
N
N
Y
N
.
Y
•
•
N
,
N
N
N
N
N
N
N
N
N
.
N
UCR
UCR
UCR
N
N
N
N
N
Y
N
.
Y
Y
Y
N
f
9
•
4/77
kill
Y
Y
AT UCR
AT UCR
AT UCR
Y
Y
Y
f
,
•
N
N
7
7
Y
UCR
4/78
N
N
N
Y
Y
Y
Y
N
N
N
N
N
Y
Y
Y
7
?
?
?
N
N
N
-------
APPENDIX 2. ON-LINE DATASET PROGRESS CHECKLIST FOR SBNF DATA BASE AS OF AUGUST 31, 1977 (CONTINUED).
LEGEND FOR HEADINGS:
"VERIF PROGR" IS VERIFICATION PROGRAM (COMPUTER),
"P.I." IS PRINCIPAL INVESTIGATOR,
"REDUCT PROGR" IS (DATA) REDUCTION PROGRAM (COMPUTER),
LEGEND FOR COLUMN ENTRIES:
"*" PRECEDING DATASET NAME INDICATES ORIGINAL DATA USED TO DERIVE SUBSEQUENT DATASET, SO ORIGINAL
MAY NOT BE DIRECTLY MEANINGFUL TO OTHER INVESTIGATORS.
"N" IS NO; THIS NEEDS TO BE DONE, BUT HAS NOT BEEN COMPLETELY FINISHED YET (AS FAR AS IS KNOWN),
"Y" IS YES, THIS STEP HAS BEEN ACCOMPLISHED,
"RPT" MEANS THAT THIS IS TO BE GOTTEN OUT OF AN EXISTING REPORT,
"?" MEANS THE STATUS OF THE INFORMATION ITEM HAS NOT BEEN DETERMINED YET,
"ONCE" MEANS THE DATA ARE ONLY COLLECTED ONCE, NOT REPEATEDLY,
"UCR" MEANS THAT THIS STEP IS BEING HANDLED AT THE UC RIVERSIDE CAMPUS,
"8/77" IS THE APPROXIMATE DATE WHEN THE P.I. PLANS TO COMPLETE THE STEP,
"M10" MEANS THAT DATA FOR SOME PLOTS HAVE BEEN RECEIVED, BUT WE ARE STILL "MISSING 10" PLOTS'
DATA WHICH P.I. IS STILL PROCESSING,
"." MEANS PROGRESS HAS GONE BEYOND THIS STEP; LOOK AT COLUMNS TO THE RIGHT >
" " MEANS THIS STEP IS NOT APPLICABLE IN THE MODELLING/DATA MGMT SUBPROJECT; BEING DONE
ELSEWHERE.
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
. REPORT NO.
EPA-600/3-80-002
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Photochemical Oxidant Air Pollution Effects on a Mixed
Conifer Forest Ecosystem
5. REPORT DATE
January 1980 issuing date
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
O.C. Taylor, Editor; Principal authors: R.N. Kickert,
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
University of California, Riverside 92521
and
University of California, Berkeley, 94720
10. PROGRAM ELEMENT NO.
1AA602
11. CONTRACT/GRANT NO.
Contract # 68-03-2442
12. SP.ONSORING AGENCY NAME AND ADDRESS ,,. _„
Environmental Research Laboratory-Corvallis, OR
Office of Research and Development
U.S. Environmental Protection Agency
Corvallis, Oregon. 97330
13. TYPE OF REPORT AND PERIOD COVERED
extramural. final
14. SPONSORING AGENCY CODE
EPA-600/02
15. SUPPLEMENTARY NOTES
Project Officer: R.G. Wilhour, Environmental Research Laboratory, Corvallis, OR 97330
FTS 420-4634 (503-757-4634
16. ABSTRACT
EPA contract 68-03-2442 provided support for three years of the studies to de-
termine the chronic effects of photochemical oxidant air pollutants on a western mixed
conifer forest ecosystem. Progress reports were published for years 1974-75 and 1975-7!:
This report deals with the year 1976-77 and is the final publication on EPA contract
68-03-2442. A computer data bank was partially developed in the early years of the
study at the Lawrence Livermore Laboratory and was subsequently revised and moved to th<
computer at the University of California, San Fransisco. Verification and auditing of
datasets is underway and several sets are ready for cross-disciplinary analysis for
modeling. Computer simulation programs have been written for some of the subsections.
Subsystems which received greatest attention during this study were: major tree species
response to oxidant dose, tree population dynamics, tree growth, moisture dynamics, soi
chemical and physical properties, tree mortality relative to disease, insects and other
factors, epidemiology of forest tree pathogens with emphasis on Fomes annosus, cone and
seed production, tree seedling establishment, litter production and litter decompositior
relative to microfloral decomposer populations. Progress is being made in preparation
of models for the purpose of describing the behavior of interlinked subsystems. Since
much progress has been made in verifying accuracy of data and of identifying informatior
in the data bank the study of subsystems interaction should be accelerated.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
COS AT I Field/Gioup
plant ecology
ecological succession
plant growth
forest land
plant reproduction
forest trees
pine trees
photochemical oxidants
conifer ecosystems
interdisciplinary invest!
gations
ecological responses
2/F
6/F
6/C
8. DISTRIBUTION STATEMENT
Release to public
19. SECURITY CLASS (ThisReport)
unclassified
21. NO. OF PAGES
20. SECURITY CLASS (Thispage)
unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77)
196
------- |