xvEPA
United States
Environmental Protection
Agency
Environmental Research
Laboratory
' Corvallis OR 97330
EPA-600/3-80-052
June 1980
Research and Development
The Bioenvironmental
Impact of a Coal-Fired
Power Plant
Fifth Interim Report,
Colstrip, Montana
-------
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 INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 221.61.
-------
-------
DISCLAIMER
This report has been reviewed by the Corvallis Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for publication.
Mention of trade names or commercial products does not constitute endorsement
or recommendation for use. . .
ii
-------
-------
PREFACE
The Environmental Protection Agency has recognized the need for a ra-
tional approach to the incorporation of ecological impact information into
power facility siting and management decisions in the Northern Great Plains.
Two capabilities need to be developed. First, in evaluating alternative
sites, planners need to be able to predict the kinds and magnitudes of impacts
to be expected so that adverse consequences to the environment can be mini-
mized. Secondly, for routine facility management after siting, environmental
monitoring methods are needed for detecting incipient ecological damage in
time for mitigation efforts to be effective.
Research funded by the Colstrip, Coal-Fired Power Plant Project is a
first attempt to generate the methods needed to predict the bioenvironmental
effects of S02 emissions from coal-fired power plants before damage is sus-
tained. The work can be subdivided into three chronological phases: 1) the
identification of information requirements and the expansion of data and
information bases, to fill these requirements, 2) the integration and synthesis
of newly generated data with existing information to define relationships
permitting maximum predictive capability, and 3) the provision of information
in a format useful to planners and decision makers involved in siting coal-
fired power plants.
Research effort for Phase 1 falls roughly into two broad categories: 1)
ecological effects monitoring in the vicinity of two 350 megawatt coal-fired
power plants at Colstrip, Montana, and 2) field and laboratory process
studies designed to elaborate the mechanisms through which coal-fired power
plant emissions cause their effects.
Pre-construction documentation of the environmental characteristics of
the grassland ecosystem in the vicinity of Colstrip, Montana began in the
summer of 1974. This documentation continued until Colstrip generating unit 1
began operation in September, 1975. Since then, key characteristics of the
ecosystem have been monitored regularly to detect possible pollution impacts.
The current results are in Sections 1-5.
In 1974, a Zonal Air Pollution System (ZAPS) was designed to stress 0.5
hectare areas of native grassland with measured concentrations of S02. In the
summer of 1975, field stressing experiments were begun to provide the data
necessary to develop dose-response models of SQ% stresses on a grassland
ecosystem. A second Zonal Air Pollution System (ZAPS II), was constructed in
1976. Both ZAPS I and ZAPS II were operated during the growing seasons of 1976,
1977, and 1978. The design of these experimental systems has been described in
previous interim reports. Their behavior is further described in the present
report. Effects on microorganisms, producers and consumers have been monitored
throughout the stressing experiments. The results of these experiments to date
are summarized in Sections 6 through 22.
The final two sections represent preliminary attempts to utilize informa-
tion presented in previous sections and previous interim reports to address the
project's objectives.
iv
-------
-------
-------
-------
Total sulfur In the forage increased in proportion to the median sulfur diox-
ide concentrations. Sulfur dioxide caused no significant decreases in forage
quality.
Large scale aerial photographs of the ZAPS plots were analyzed visually
and densitometrically. A general reduction in density with increased SOa
exposure was observed in red, green, and blue wave lengths. Photographic
density differences were clearly related to SOa fumigation. The red density
was especially depressed by S02 during the active growth season.
In summary, the 1978 ZAPS studies suggest that total living vegetation,
total graminoids, western wheatgrass, lichens and perennial species all show
decreasing cover with increasing sulfur dioxide exposure rates. SOa treatment
combined with nitrogen fertilization enhanced herbage yield. Chlorophyll a
and b concentrations in western wheatgrass appear to be sensitive indicators
of SOa exposure. The decomposition rates of western wheatgrass decreased and
grasshopper, beetle, and possibly tardigrade numbers were reduced on various
treatment plots. Biomass dynamics of dominant grasses, net primary produc-
tion, forage quality, and species composition have not been demonstrably
altered by SOa treatments.
Synthesis
Simulation models of SOa deposition and sulfur cycling in a grassland
were developed as conceptual tools to organize information. When models were
executed simultaneously with others describing primary production, decomposi-
tion and abiotic processes, they provided fair representations of the accumu-
lation of sulfur in live and dead leaves and soil. Under simulated exposure
to 10 pphm S02, many sulfur flows in the model grassland increased causing
substantial departure from initial equilibrium.
vnx
-------
CONTENTS
Page
Foreword ................................. ill
Preface ................................. ±v
Abstract ................................. v
List of Contributors .......................... . sii
Acknowledgements ............................. xiv
Section
ECOLOGICAL EFFECTS MONITORING IN THE COLSTRIP VICINITY
1. Air Quality Measurements in the Coal-Fired Power
Plant Environment of Colstrip Montana
J.D. Ludwick, D.B. Weber, K.B. Olsen and S.R. Garcia ... 1
2. Plume and Aerosol Properties Near Colstrip
C.C. Van Valin, R.F. Pueschel, D.L. Wellman,
arid G.M. Williams ..................... 20
3. Observations of Two Lichen Species in the Colstrip
Power Plant Vicinity
S. Eversman ......................... 49
4. Plant Community Monitoring in the Vicinity of Colstrip,
Montana
J.E. Taylor, W.C. Leininger and M.W. Hoard ..... ... 55
5. Accumulation and Transfer of Fluoride and Other Trace
Elements in Honey Bees Near the Colstrip Power Plants
J.J. Bromenshenk ............ .......... 72
FIELD AND LABORATORY EXPERIMENTS
6. Temporal Variation in SOz Concentrations on ZAPS
During the 1978 Field Season
E.M. Preston, T.L. Gullett, and D.B. Weber ......... 96
ix
-------
-------
21. The Effects of Low Concentrations of Sulfur Dioxide
on the Growth, Quality and Nitrogen Fixation of Alfalfa
D.T. Tingey, G.E. Neely, and M.L. Gumpertz 279
22. Remote Sensing of S02 Effects on ZAPS
J.E. Taylor, W.C. Leininger, and M.W. Hoard 292
PRELIMINARY SYNTHESIS
23. Simulation of a Grassland Sulfur Cycle
M.B. Coughenour, W.J. Parton, W.K. Lauenroth,
J.L. Dodd, and R.G. Woodmansee 304
24. Relationship of S02 Deposition to a Grassland Sulfur
Cycle
M.B. Coughenour 345
XI
-------
LIST OF CONTRIBUTORS
C.J. Bicak
Natural Resources Ecology Laboratory
Colorado State University
Fort Collins, Colorado 80521
J.J. Broraenshenk
Environmental Studies Laboratory
University of Montana
Missoula, Montana 59801
M.B. Coughenour
Natural Resources Ecology Laboratory
Colorado State University
Fort Collins, Colorado 80521
J.L. Dodd
Natural Resources Ecology Laboratory
Colorado State University
Fort Collins, Colorado 80521
S. Eversman
Department of Biology
Montana State University
Bozeman, Montana 59715
S.R. Garcia
Battelle Pacific NW Laboratories
Battelle Boulevard
Richland, Washington 99352
C.C. Gordon
Environmental Studies Laboratory
University of Montana
Missoula, Montana 59801
T.L. Gullett
U.S. EPA/CERL
200 SW 35th Street
Corvallis, Oregon 97330
M.L. Gumpertz
U.S. EPA/CERL
200 S.W. 35th Street
Corvallis, Oregon 97330
R.K. Heitschmidt
Natural Resources Ecology Laboratory
Colorado State University
Fort Collins, Colorado 80521
M.W. Hoard
Department of Animal and Range Sciences
Montana State University
Bozeman, Montana 59715
W.K. Lauenroth
Natural Resources Ecology Laboratory
Colorado State University
Fort Collins, Colorado 80521
J.W. Leetham
Natural Resource Ecology Laboratory
Colorado State University
Fort Collins, Colorado 80521
W.C. Leininger
Department of Animal and Range Sciences
Montana State University
Bozeman, Montana 59715
J.D. Ludwick
Battelle Pacific NW Laboratories
Battelle Boulevard
Richland, Washington 99352
T.J. McNary
Natural Resources Ecology Laboratory
Colorado State University
Fort Collins, Colorado 80521
D.G. Milchunas
Natural Resources Ecology Laboratory
Colorado State University
Fort Collins, Colorado 80521
G.E. Neely
U.S. EPA/CERL
200 S.W. 35th Street
Corvallis, Oregon 97330
xii
-------
D.W. O'Guinn
U.S. EPA/CERL
200 S.W. 35th Street
Corvallis, Oregon 97330
K.B. Olsen
Battelle Pacific NW Laboratories
Battelle Boulevard
Richland, Washington 99352
W.J. Parton
Natural Resource Ecology Laboratory
Colorado State University
Fort Collins, Colorado 80521
E.M. Preston
U.S. EPA/CERL
200 SW 35th Street
Corvallis, Oregon 97330
R.F. Pueschel
U.S. Department of Commerce
NOAA
Environmental Research Laboratories
Boulder, Colorado 80302
L. Pye
Environmental Studies Laboratory
University of Montana
Missoula, Montana 59801
P.M. Rice
Environmental Studies Laboratory
University of Montana
Missoula, Montana 59801
J.E. Taylor
Department of Animal and Range Sciences
Montana State University
Bozeman, Montana 59715
D.T. Tingey
U.S. EPA/CERL
200 SW 35th Street
Corvallis, Oregon 97330
P.C. Tourangeau
Environmental Studies Laboratory
University of Montana
Missoula, Montana 59801
C.C. Van Valin
U.S. Department of Commerce
NOAA
Environmental Research Laboratories
Boulder, Colorado 80302
D.B. Weber
U.S. EPA/CERL
200 SW 35th Street
Corvallis, Oregon 97330
D.L. Wellman
U.S. Department of Commerce
NOAA
Environmental Research Laboratories
Boulder, Colorado 80302
G.M. Williams
U.S. Department of Commerce
NOAA
Environmental Research Laboratories
Boulder, Colorado 80302
L.T.K. Wong
Natural Resources Ecology Laboratory
Colorado State University
Fort Collins, Colorado 80521
R.G. Woodmansee
Natural Resource Ecology Laboratory
Colorado State University
Fort Collins, Colorado 80521
xiii
-------
ACKNOWLEDGEMENTS
Many individuals have contributed to the preparation of this document.
The editorial assistance of Susan With, Karen Randolph and others is much
appreciated.
Our work could not proceed without the help and support of the people of
southeastern Montana, especially the ranchers on whose land we are working
and the personnel and persons residing at and near Fort Howes Ranger Station,
Custer National Forest. The Kluver's and the McRae's have been particularly
supportive.
xiv
-------
.ECOLOGICAL EFFECTS MONITORING IN THE COLSTRIP VICINITY
SECTION 1
AIR QUALITY MEASUREMENTS IN THE COAL-FIRED POWER
PLANT ENVIRONMENT OF COLSTRIP MONTANA
J. D. Ludwick, D. B. Weber, K. B. Olsen, and S. R. Garcia
ABSTRACT
A network of up to six air monitoring and sampl-
ing stations were operated (4-23 km) downwind of the
700 MW coal-fired power plants at Colstrip, Montana.
Continuous monitoring identified background SOa levels
in the surrounding air at approximately 0.1 pphm (2.3
yg/m3). Average air concentrations measured in 1978
throughout the network increased to about 0.2 pphm
(4.6 yg/m3) and included the power plant contribution.
Meteorological and aerosol correlations with SOa
concentrations helped to identify and isolate the
major source. The high sensitivity for SOa measure-
ment provided resolution of all significant power
plant plume fumigations at ground level. The magni-
tude of S02 exposures were higher than predicted in
this rough terrain environment.
INTRODUCTION
Results are reported from an air sampling and monitoring network operat-
ing in the eastern portion of the state of Montana during 1978. This program
is an expansion and sophistication of earlier attempts to characterize the air
quality downwind of the two coal-fired power plants at Colstrip, Montana (700
MW). A year of continuous operations has yielded data which include signifi-
cant environmental effects from operation of these power plants 4-23 km down-
wind. The atmospheric models presently used to estimate pollutant exposures
in this rough terrain environment appear to have underestimated the frequency
and intensity of power plant emission exposures at the more distant locations.
Long <--erm studies have been reported concerning the downwind ground level
concentration of sulfur dioxide in the coal-fired power plant environment
(Moore, 1967; Martin and Barber, 1966; Martin and Barber, 1967; Martin and
Barber, 1973). In general, these studies have taken place in areas of consid-
erable pollution, where high background levels must be tolerated. Additional
problems relating to instrumental sensitivity capabilities, in real time have
limited the detection of low level plume fumigation. Lack of instrumental
-------
sensitivities may be responsible for high background estimates reported at
locations where incoming air appears to have no obvious pollution source
(Barnes, 1976). Long distance transport (100 km) of sulfur dioxide from major
industrial regions have been reported, however, and the influence of such
sources must be considered in some situations (Zeedijk and Velds, 1973). In
spite of elevated levels of sulfur dioxide in the background, meteorological
correlations have helped identify the larger sources in many instances (Raynor
et al., 1974, Zanetti et al,, 1977).
The Great Plains contains a vast reservoir of fossilized energy. The
most direct purpose for its use is in supplying the power needs of the region
it underlies. It was anticipated that the expanded use of coal for supplying
electrical power will present environmental problems which must be addressed.
In the Colstrip area a consortium of public utilities have site applications
which propose to triple the generating capacity of'this area with two addi-
tional units. Each of these proposed power plants will have the combined
capacity of the two original units. Experimental data such as that reported
here, as well as much more, will be necessary to provide accurate siting
criteria information on the expected impact of large pollutant emitters. The
bulk of the air monitoring and sampling data reported here includes the entire
year of 1978.
A review of the collected data illustrates the sensitivity of the air
monitoring stations to assess the very low sulfur dioxide air concentrations
found in this clean environment. As a result pertinent data are available to
estimate the extent of the power plant impact upon the Class 1 region located
geographically similar to our more distant monitoring stations.
MATERIALS AND METHODS
The relative positions of the air monitoring stations and the power plant
are shown in Figure 1.1. Not more than four sites were operated simultane-
ously; however, this figure illustrates all locations that were monitored
during the course of 1978. The initial site locations selected were based
upon a wind-rose analysis of this area. Although all stations were not ident-
ical in capability, a typical station description follows.
I
Air monitoring was conducted on oxides of sulfur utilizing a Meloy Model
285 sulfur analyzer, which was selected for its low noise characteristics.
This was particularly significant since.the bulk of the ambient data was
expected to be below 1.0 pphm (23 yg/m3). Ozone levels were measured with a
Bendix Model 8002 ambient air analyzer. Particulates were monitored with a
Model Rich 100 Environment/One Condensation Nuclei Detector. Ambient air was
supplied to the monitoring instrumentation from a common Teflon manifold.
High velocity flow was maintained through this manifold. Information from the
air monitoring equipment, as well as data from meteorological instrumentation
located at the sites, were collected by a microprocessor/datalogger and inter-
faced to a paper punch tape for storage and computer processing. The meteoro-
logical information included windspeed and direction, temperature, humidity
and solar radiation. All the equipment was housed in small, highly insulated
trailers. Electrical power for most stations was derived from 5.5 kW pro-
-------
BURLINGTON
NORTHERN RR
WIND
DIRECTION
TO POWER
PLANT
i i i i i i i i i
012345678
km
A
GREENLEAF
I
106°40'
106°35'
106°30'
106°25'
Figure 1.1. Monitoring station location.
pane-fueled generators. The power was divided into two sections: rough power
for the heating, air conditioning, and pumps as well as stable power for the
instrumentation. An uninterruptible power system supplied highly regulated
115 volt 60 Hz electricity necessary for the stable operation of the air
monitoring instrumentation. This equipment also provided power continuity for
station operation during periods of short term generator outage. One-thousand
gallon storage containers provided a 45-day fuel supply for the generators.
This was necessitated by the inaccessibility of some of the stations during
the winter months in Montana.
A two-stage cyclone separator was operated in conjunction with the moni-
toring instrumentation. This provided particulate information which could be
correlated with the gas data. The air monitoring data were collected in incre-
ments of four minutes and these values were averaged for different pertinent
intervals. The data were edited for spurious information and excessive instru-
mental fluctuations.
-------
A daily quality assurance program for verifying the integrity of the
collected air monitoring data was in effect throughout the operation of this
program. In addition, Rockwell International, under contract from the En-
vironmental Protection Agency (EPA), performed quarterly audits on a routing
basis. The extent of each audit included all operational air monitoring
instrumentation and other calibration equipment at the sites visited. A
record of the audit reports is available as well as a more complete record of
other parameters associated with this report (Ludwick et al., 1979).
RESULTS AND DISCUSSION
The monthly averages .of the data from sites that had sufficient on-line
time are shown in Tables 1.1, 1.2, 1.3, and 1.4. The concentrations of sulfur
dioxide are quite low, with respect to the urban environment. Except for
individual short-term events discussed later, the impact of oxides on nitrogen
from the power plant upon the ozone averages would be masked by the ubiquitous
ambient levels of this species. In general, the levels of ozone are consis-
tent with those observed in other relatively clean environments (Ludwick
et al., 1976).
The concentration fluctuations of gaseous sulfur at the stations is
sufficient in some instances to be of interest. For example, in March the
averaged sulfur dioxide concentrations at Hay Coulee, 12 km from the power
plant was 0.56 pphm. This represents a contribution of almost 0.05 pphm on an
annually averaged basis from this single month. The Northern Cheyenne Indian
Reservation extends to within 22 km of the power plant, in the same general
direction as Hay Coulee. This land bears a Class 1 air quality designation.
With reference to Table 1.5, the Air Quality Standards, we see that an eleva-
tion of no more than 0.07 pphm is acceptable before significant air quality
deterioration is suggested. Since values of this magnitude were not expected
from EPA models utilized in this area, the question of the sulfur dioxide
source and downwind dilution between 12 and 22 km becomes rather important.
No reliable ambient sulfur dioxide background measurements are available
prior to power plant operation. These data suggest that averaged present day
background levels decline to about 0.07 pphm and are typically about 0.1 pphm
(2.31 yg/m3). Values of this magnitude correspond to those observed in the
cleanest areas of industrialized nations (Junge, 1963). Assuming this to be a
reasonable conservative minimum, the magnitude of the effect from a pollution
source can then be considered. Lower background estimates would, of course,
infer more potent anthropogenic sulfur dioxide sources. A summation and aver-
aging of this data on an annual basis indicate that both KE and KW stations
observed an increase of 0.08 pphm (1.9 yg/m3) even though KE is several kilo-
meters further downwind of the Colstrip area. A similar increase was observed
at BNW-1, very close to Cplstrip. A slightly larger increase of 0.12 pphm
(2.66 yg/m3) was observed at Hay Coulee, downwind of the BNW-1 site. Although
variations in operational periods exist, it appears that an overall elevation
of 0.1 pphm of sulfur dioxide throughout the area were measured during 1978.
-------
TABLE 1.1. HAY COULEE MONTHLY AVERAGES
Date
2/78
3/78
4/78
5/78
6/78
7/78
8/78
9/78
10/78
11/78
12/78
Wind
Speed
kmph
7.0
7.0
12.7
10.8
9.5
9.0
9.2
9.3
7.7
5.7
9.5
Relative
Humidity
%
70.9
61.3
54.6
68.2
56.6
52.4
43.0
56.0
50.0
61.5
55.4
Temperature
°C
-11.2
-2.4
8.2
12.6
19.4
24.1
22.3
15.8
7.3
-8.7
-13.0
Ozone
pphm
4.3
4.2
3,9
4.1
4.0
4.3
3,3
2.8
2.7
3.3
3.4
SO 2
pphm
0.37
0.56
0.18
0.13
0.23
0.16
0.10
0.15
0.17
0.15
0.17
TABLE
1.2. KLUVER
EAST MONTHLY AVERAGES
Date
3/78
4/78
5/78
6/78
7/78
8/78
9/78
10/78
Temperature
°C
8.9
9.1
12.3
16.5
27.4
22.3
18.9
12.3
Ozone
pphm
4.0
3.9
4.5
3.9
4.6
3.6
3.0
3.2
SO 2
pphm
0.18
0.18
0.22
0.34
0.14
0.09
0.07
Evidence as to the source of the pollution comes from several areas
throughout this report. Figure 1.2 illustrates the generally correlating
character of the monthly sulfur dioxide fluctuations of the individual sta-
tions. This is indicative of a common pollution source influencing the area.
-------
TABLE 1.3. KLUVER WEST MONTHLY AVERAGES
Date
3/78
4/78
5/78
6/78
7/78
Temperature
°C
7.0
8.5
12.8
18.8
21.3
Ozone
pphm
4.0
3.9
4.4
4.1
4.3
SO 2
pphm
0.19
0.16
0.15
0.26
0.14
TABLE 1.4. BNW-1 MONTHLY AVERAGES
Date
7/78
8/78
9/78
10/78
11/78
12/78
Wind
(Degrees)
—
I
229
250
231
257
Temperature
°C
24.3
24.0
17.2
10.8
-3.5
-5.4
Ozone
pphm
4.5
4.1
3.4
3.4
3.6
3.6
S02
pphm
0.20
0.14
0.14
0.13
0.20
0.26
In addition to the longer, time averaged values mentioned, many interest-
ing features of the monitoring information are apparent in the real time data.
Due to the remote rural location of the stations, most of the air being moni-
tored contains little more than hemispheric concentrations of gaseous and
particulate materials. When meteorological conditions prevail that are condu-
cive to downward transport of the power plant plume to the stations, plume
strikes are unmistakably prominent features upon this data base. Figure 1.3
is a classical plume strike situation in which elevated levels of sulfur
dioxide are correlated with particulates from the power plant stack when the
wind vector orients from the plant direction. Excesses of nitric oxide in the
plume from high temperature coal burning processes are often apparent in the
data as a result of their interaction with atmospheric ozone, as follows:
03 + NO -*- N02 + 02 (3)
-------
TABLE 1.5. SULFUR DIOXIDE AIR QUALITY STANDARDS
Averages
Annual
24 Hour
3 Hour
Federal*
EPA (Classified Areas)
Prevention of Significant
Deterioration
Class 3 Industrial
Class 2 Most of USA
Class 1 Pristine
(Natl. Parks,
Wilderness
Indian Lands)
State of Montana
3.0 pphm '
5.7 yg/m3
0.7 pphm
(2 yg/m3)
20 yg/m3
14.0 pphm
38 yg/m3
1.9 pphm
(5 yg/m3)
lOOt yg/m3
50.0 pphm
(1300 yg/m3)
* Not to be exceeded (NTREX) more than once per year.
t NTBEX - 1 day/3 month period.
tt 1 hr average NTBEX - 1 hr/4 consecutive months
50.0 pphm
267 yc/m3
9.5 pphm
(25 yg/m3/
250tt yg/m;
0.5
0.4
E
f-0.3
0.2
0.1
I
I
I
I
I
I
I
KW *- *
KE • •
HC • •
I
JAN FEB MAR APR MAY
JUN JUL AUG
1978
SEP OCT NOV DEC
Figure 1.2. Monthly S02 correlations between air monitoring stations,
7
-------
pphm
pphm
10K
0
1.0
0
3.0
0
360
degrees
0
2.0
LANG LEY
mm
0
+ 15
-5
IOC
90
kmph
CONDENSATION NUCLEI
WIND DIRECTION
j i
bOLAR RADIATION
TEMPERATURE
RELATIVE HUMIDITY
WIND SPEED
12 0 12
APRIL 18, 1978 THROUGH APRIL ?0, 1978
HAY COULEE
0 hrs
Figure 1.3. Air monitoring information: typical plume strike.
The anticorrelations between sulfur dioxide and ozone are further evidence of
the presence of the power plant plume. Not all sulfur plume strikes bear such
evidence, however. Sulfur dioxide excursions are occasionally accompanied by
unchanged or increasing ozone concentrations. Masking of the ozone anticorre-
lation can take place in several ways. At first, while consuming ozone, NO is
converted to NOa and adds to that already emitted. A second process, creating
ozone, is known to become important during daytime hours as the concentration
of NOa increases (Davis et at., 1974):
NO 2 -
hv
NO + 0*
(2)
0' + 02
03
(3)
-------
10K
0
1.0
pphm
0
3.0
pphm
0
360
degrees
(';
2.0
. LANGLEY
mm
0
30
'10
100
90
kmph
12
CONDENSATION NUCLEI
(MISSING DATA)
SO,
SOLAR RADIATION
TEMPERATURE
RELATIVE HUMIDITY
WIND SPEED
0 12 0
AUGUST 11, 1978 THROUGH AUGUST 13, 1978
HAY COULEE
12hrs
Figure 1.4 Air monitoring inforjnation: multiple plume strikes.
Thus, the age of the pollutant plume may influence the net effect of NO
and N02 on the correlating ozone concentrations.
Elevated ozone concentrations have also been observed above stable under-
lying air masses (Miller, 1974; Ludwick et al., 1976). Convective turbulence,
which appears important in transferring the pollutant plume to the monitoring
station may also, on occasion, be responsible for elevated ozone levels and
interference with anticorrelations that were initially created by the pollu-
tion source.
A 2-day period is illustrated in Figure 1.4 in which three plume strikes
were evident each time the wind cycled through the plant direction. It is ap-
parent from the frequency of appearance of these conditions that plume strikes
-------
1500
1200
900
160m POWER
•PLANT STACK
CLASS 1
BNW-1 (1090m)
(4.0pphm)
HAY COULEE (907m)
(2.0pphm)
GREENLEAF
(0.7pphm)
J—I—L
J
' '
|
J—I—I—L
J—I
10 15
DISTANCE (km)
20
25
Figure 1.5. Southeast terrain cross section from the Colstrip power plants.
are rather common (376 individual plume strikes were observed throughout the
monitoring network in 1978, Ludwick et al., 1979). Daytime convective turbu-
lence as suggested by fluctuations in the solar data, appears often to mix the
plume downward to the stations in spite of the elevated plume injection height
and interfering terrain.
As field experience was gained concerning the probable location of plume
intercepts, emphases were placed on relocating the monitoring stations for
maximum effectiveness. The results of such reorientation are shown in Figure
1.5. Here, three co-linearly oriented stations (see Figure 1.1) exhibited the
effects from the same power plant plume. It is also a representation of the
topographical features southeastward from the Colstrip, Montana area. The low
peaks to the east are slightly south of the exact station orientation. Al-
though Class 1 land falls within the Greenleaf site distance, it does not
approach this site in its actual geographic fluctuations. .The maximum concen-
trations observed at each site are listed. This was one of two similar triple
strike situations that took place in ,early November 1978. At that time,
approximately 5,000 pphm (1150 mg/m3) of sulfur dioxide were measured in the
output from each of two power plant stacks. Thus, the concentration dilution
from the first sample location 3 km downwind was approximately 2.5 x 101*.
Previous estimates of the frequency and intensity of plume fumigations at
the more distant monitoring stations were negligible. This was not found to
be the case. These estimates were derived from diffusion models which at-
tempted modification of existing flat surface models to the mountainous
terrain environment. A brief attempt was made, under the auspices of this
program, to compare downwind exposure predictions from the diffusion models
that are in present usage, with those actually observed at the sites. Experi-
10
-------
ence with the EPA, PTMTP atmospheric model used for comparison purposes during
high sulfur dioxide exposures in February and March indicated a considerable
underestimation of the observed values. Use was made, therefore, of a com-
puter modeling program at the University of Montana relied upon for downwind
plume estimates (Williams and Tourangeau, unpublished work) in this type
terrain. Atmospheric stability data was incorporated into this model derived
from tower instrumentation located beside the power plant stack. Comparisons
between actual and predicted concentrations of sulfur dioxide at the sites in
Figure 1.5 are shown on Table 1.6. The model predictions were based upon
maximum exposures from full power operation of the coal-fired plants. This
was not the situation at this time; however, even these overestimates fall
short of reality at the greater distances. Also evident in this case history
is the variation of exposure with distance. Although the peak concentrations
observed at the sites varied by a factor of 5, the integrated exposures did
not change significantly with downwind travel. This produced unexpectedly
high exposures at longer distances. Meteorological situations that promote
such occurrences are evidently more frequent than expected.
TABLE 1.6. DIFFUSION MODEL PREDICTIONS vs. MEASURED HOURLY AVERAGED (S02)
Time BNW-1 HC Greenleaf
Modelt
Measured*
9
10
11
12
9
10
11
12
- 1000
- 1100
- 1200
- 1300
- 1000
- 1100
- 1200
- 1300
0 pphm
0.11
0.01
2.2
0.25
0.50
0.30
0.30
0 pphm
0.02
0
1.0
0.35
0.50
0.60
0.45
0 pphm
0.27
0.01
0.05
0.25
0.30
0.45
0.50
t See CWilliams and Tourangeau, unpublished work).
* All values / 0.05 pphm.
The 1978 data set was interrogated with respect to 3-hr and calendar day
sulfur dioxide exposures. These limits were used to allow comparisons with
the shorter term National Air Quality Standards (Table 1.5). No excesses were
observed for the Class 2 category on which the sites were located. The total
number of estimated excesses with respect to the nearby Class 1 land are shown
in Table 1.7 along with the other pertinent data. A specific compilation by
date, time and exposure is available for all the monitoring sites (Ludwick
et al., 1979). A significant number of these days are illustrated in the
monitoring information. Because of the close proximity of the sites to the
Class 1 area, their importance is enhanced.
In order to identify the location of probable sources of sulfur dioxide
to each major monitoring station, the direction of incoming sulfur dioxide was
correlated with its concentration isopleth. The entire year of data was
11
-------
TABLE 1.7. FREQUENCY OCCURRENCE OF ELEVATED S02 EXPOSURES DURING 1978
Site
BNW-1
Kluver West
Hay Coulee
Kluver North
Kluver East
Class 1 Land
Greenleaf
Distance
From Source (km)
4.1
11.2
11.6
14.1
18.2
22.2
22.5
3 hr AVE >^
10 pphm
0.9
0.5
1.8
0
0
0
24 hr AVE >
2 pphm
7.2
5.4
16.1
1.9
4.7
1.2
0.1
Observation
Period (mo)
5.5
4.5
10.8
1.8
6.3
12*
1.0
* Estimated number of excesses of air quality standards as interpolated from
graphic analysis of sulfur concentration vs. distance from the power plant.
360° 0°
PLANT
DIRECTION
5*319°
225'
135"
Figure 1.6.
Time averaged SOz concentration isopleth for 10° increments
around Hay Coulee for 1978.
12
-------
360° 0°
315'
Figure 1.7. Time aver-
aged SOz concentration
isopleth for 10° incre-
ments around KE for 1978.
225'
315'
Figure 1.8. Time aver-
aged SOa concentration
isopleth for 10° incre-
ments around KW for 1978.
225'
135"
13
-------
TABLE 1.8. FREQUENCY DISTRIBUTION OF S02 EXPOSURES IN THE AIR MONITORING NETWORK
Hourly Averaged
S02 (pphm)
*o
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
•
5.*0
Number of Times Air Concentration Exceeded or Equaled
Hay Coulee
4638
3035
2408
1625
944
577
320
194
125
90
43
28
19
15
12
9
7
6
1
0
0
0
0
0
0
0
0
0
0
0
KE
1987
1319
1006
660
368
180
57
23
11
4
3
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
KW
1926
1339
1040
676
274
103
39
22
15
11
5
5
4
2
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
•
0
KN
821
386
297
196
123
59
21
7
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
BNW-1
3041
1732
1291
846
537
297
168
94
66
49
37
27
23
15
11
8
8
6
6
6
5
4
3
3
3
3
3
2
1
•
1
Greenleaf
334
80
46
22
8
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
* Valid data collection times.
-------
utilized for this purpose in order to reflect all possible sources of sulfur
dioxide. Figures 1.6, 1.7, and 1.8 illustrate the results of these compari-
sons. Each display is a sulfur dioxide concentration isopleth on an annual
basis for 10° directional increments around the particular stations shown.
Several features are common to these illustrations. Each has a major concen-
tration lobe in the direction of the coal fired power plants. All illustrate
a rather low sulfur dioxide input from southeasterly winds opposed to the
power plant's direction. A lesser source term seems apparent to the northeast
between 15° and 30° true north with respect to each station. A comparison of
the topographical features of the area with this secondary source was helpful.
It appears that the apparent direction of this secondary source correlates
witH the contours of Rosebud Creek. Evidently the common westerly winds
fumigate the area to the north of our sites. Then sulfur dioxide becomes
trapped in the lower terrain and northerly winds create flow through the
valley bringing elevated concentrations of sulfur dioxide to the stations.
As the plant plume moves eastward over the station network, it will from
time to time, as shown earlier, mix downward to ground level. The frequency
of this occurrence as well as the magnitude of such events relate to the
instantaneous and integrated exposures observed from the emitted pollutants.
Table 1.8 illustrates the integrated hourly sulfur dioxide exposure frequency
actually observed at the sites. This data is presented in a manner comparable
with other field studies (Drufuca and Guigliano, 1977; Zimmer and Jutze,
1964). These studies suggest that the product of the sulfur dioxide concen-
tration and exposure time should be correlated with biological damage that
might occur from such fumigations. The period of operation varied consid-
erably from site to site as instrumentation was utilized in what appeared to
be the most expeditious manner at the time. Data omission also occurred
because of power failures, instrument breakdowns and data editing.
Aerosol Sampling Data
An aerosol sampling program was conducted at three of the sites in con-
junction with the air monitoring. Samples were collected at intervals of a
day or longer depending upon weather and personnel availability. The sampling
system consisted of a cyclone separator with special orifice and pump. The
systems were matched to provide an aerosol separation at slightly less than
one micron. Calibration was accomplished using standardized aerosol disper-
sions. Two fractions were collected from each sample with the fine portion
accumulating on Whatman 41 filter paper. The larger fraction was centrifu-
gally maintained in a stainless cup of the separator, A procedure was devel-
oped for slurrying this fraction and transferring it onto a polypropylene
film. Then both fractions underwent x-ray fluorescence analysis for a series
of trace elements. Thin film standards were routinely utilized in each batch
analyzed by this technique.
The entire 1978 data set was scanned for level sulfur dioxide concentra-
tions. These are .called plume "strike days" and are defined as those in which
the hourly averaged sulfur dioxide concentrations exceeded 1.0 pphm (23
yg/m3). It was also noted that this situation occurred only when the wind
direction was from the power plants. A second set of background days of "non-
strike days" were also derived from the data. These were days in which the
15
-------
winds blew from the power plant quadrant but the sulfur dioxide levels were
less than 0.2 pphm (4.6 yg/m3). This method provided a means of comparing the
influence of the plume upon the aerosol.
By comparing the elemental distributions of samples collected during
these two periods, one can look for changes in the Colstrip natural aerosol
levels with the addition of the power plant plume. Table 1.9 illustrates
results of this comparison and includes background and source term information
taken from earlier measurements in the Colstrip area (Crecelius et at., 1976).
TABLE 1.9. ELEMENTAL DISTRIBUTIONS IN COLSTRIP AREA AEROSOLS
Ratio in Sample/Ratio in Crustal*
Element Non Strike Days Stack Strike Days Colstrip Soil
Ca
Ti
Mn
Fe
Zn
Se
Pb
As
K
Br
1.3
1.0
0.80
1
19
1000
68
90
1.1
128
26
8
4.8
1
7
26,000
68
350
0.1
244
1.8 ± 6%
1.3 ± 17%
1.1 ± 32%
1 ± 2%
97 ± 11%
9000 ± 39%
450 ± 11%
825 ± 31%
1.6 ± 10%
990 ± 8%
1.4
1.4
1.0
1
3.4
600
-
9.3
1.7
3
* All values were normalized to Fe.
The ratios on the table were derived by first normalizing e.ach element to
the amount of iron found in the sample in question and then dividing that
number by the elemental ratio observed in normal crustal material. The stack
column illustrates the relative potency of the power plant emissions to change
the natural aerosol distribution. The Colstrip aerosol column illustrates the
aerosol differences that already exist between eastern Montana and average
crustal distributions of elements. It appears that the selenium ratios had
been modified in the aerosol by the stack; however, definitive explanations of
other elemental changes are not as evident. The elevated lead and bromine
values reflect the influence of Colstrip traffic upon the aerosols since the
stack is incapable of modifying the ratios to the extent shown. The effects
of mining operations in the plant direction versus those in other directions
would also leave their imprint upon the particulate data although no such
directional correlation was observed. In most situations, contributions to
the aerosol come from several sources including the natural crustal materials
and exact definition is complex. Significant elemental enrichment of the
aerosol versus local soil were found for arsenic (26x) and Zn (3.5x) in back-
ground measurements of the Colstrip area prior to power plant startup (Crece-
lius et al. , 1976).
16
-------
Utilizing the definition of strike and non-strike days given previously,
Table 1.10 illustrates the ratios found between the fine and coarse particu-
late for each element in the aerosols. Those elements usually associated with
crustal materials appear equally distributed between fractions. Elements
whose volatility or ability to fractionate to smaller particulate are found on
the smaller fraction.
TABLE 1.10. AEROSOL SIZE DISTRIBUTIONS IN S02 POLLUTED AIR ON STRIKE DAYS
Ratio of Fine/Coarse
Ni
Cr
As
Zn
Pb
Se
Br
8 ±
9 ±
17 ±
20 ±
23 ±
28 ±
136 ±
6
5
6
4
14
31
81
Tl
Ca
Mn
K
Fe
0.13
0.86
0.69
0.69
1.3
± 0.10
± 0.66
±0.31
± 0.21
± 1.1
CONCLUSION
The pristine environment of Montana State lent itself to identifying, in
part, the downwind impact from operation of the first major coal-fired power
plants in that region. The compexity of contributing sources in the indus-
trialized areas make specific plant impact relationships difficult to assess.
For example, results from a monitoring program of similar orientation around
two much larger plants (3000 MW) indicate that their combined effect repre-
sents only a 10% sulfur dioxide increase in the seasonal background levels
(Martin, 1973), whereas, our study showed that winds from the power plant
bring greater than twice the sulfur dioxide concentration to the monitoring
network as from other directions. Although the Colstrip power plant plumes
have been clearly identified, the quantitative levels of sulfur dioxide are
very low. With respect to both exposures and maximum observable concentra-
tions, we certainly are dealing in the parts per billion range.
Prior estimates of the extent of downwind sulfur dioxide fumigation were
negligible with respect to the data presented here. In all specific cases
studied beyond five kilometers, the actual exposures were at least twice that
predicted by field models in use, and generally more. Monthly averaged data
as well as individual case situations indicate that although peak exposures
may change dramatically with distance, the integrated exposures do not. The
atmospheric models do not suggest this and, therefore, diverge from the meas-
ured values with distance. The high monthly values measured 12 km from the
power plant at Hay Coulee may well represent approximate exposures further
downwind. These data strongly suggest longer-range monitoring information is
necessary to improve power plant siting criteria.
17
-------
The biological impact of the power plants has been under study for some
time. Results from such work are reported on a continuing basis from the
Corvallis, EPA (Preston and Gullett,1979). Three of the monitoring stations
were co-located within rangeland plots that were set aside specifically for
assessing biological changes in the power plant environment. Subtle effects
on the biota, with time, have been noted in the experimental data prior to and
during our monitoring program. These may now be compared with our data to
determine if correlations with long-term low-level chronic sulfur dioxide
exposures exist.
REFERENCES
Barnes, R, A. 1976. Long-Term Mean Concentrations of Atmospheric Smoke and
Sulfur Dioxide in Country Areas of England and Wales. Atmospheric En-
vironment, 10:619-631.
Crecelius, E. A., J. C. Laul, L. A. Rancitelli, and R. L. McKeever. 1976.
Baseline Air Quality Data from Colstrip, Montana. Pacific Northwest
Laboratory Annual Report for 1975 to the USERDA Division of Biomedical
and Environmental Research, Part 3, Atmospheric Sciences, BNWL-2000 PT3
Battelle-Northwest, Richland, WA. pp. 25-26.
Davis, D. D., G. Smith, and G. Klauber. 1974. Trace Gas Analysis of Power
Plant Plume via Aircraft Measurements: Os, NO , and SOa Chemistry.
Science, 186:733-736. X
Drufuca, G., and M. Guigliano. 1977. The Duration of High SC>2 Concentrations
in an Urban Atmosphere. Atmospheric Environment, 11:729-735.
Junge, C. E. 1963. Air Chemistry and Radioactivity. Academic Press, New
York.
Ludwick, J. D., T. D. Fox, and L. L. Wendell. 1976. Ozone and Radionuclide
Correlations in Air of Marine Trajectory at Quillayute, Washington. J.
Air Pollut. Control Assoc.., 26:565-569.
Ludwick, J. D., D. B. Weber, K. B. Olsen and S. R. Garcia. 1979. Air Quality
Measurements in the Coal-Fired Power Plant Environment of Colstrip,
Montana. Pacific Northwest Laboratory, PNL 2964, Battelle-Northwest,
Richland, WA.
Martin, A. and F. R. Barber. 1966. Investigation of Sulfur Dioxide Around
Modern Power Stations. J. Inst. Fuel, 39:294-307.
Martin, A. and F. R. Barber. -1967. Sulfur Dioxide Cocentrations Measured at
Various Distances from a Modern Power Station. Atmospheric Environment,
1:655-677.
Martin, A. and F. R. Barber. 1973. Further Measurements Around Modern Power
Stations—I-III. Atmospheric Environment, 7:17-37.
18
-------
Miller, A. 1974. Rural and Urban Ozone Relationships. J. Air Pollut. Con-
trol Assoc. , 24:819-821.
Moore, D. J. 1967. SOa Concentrations Measurements Near Tilbury Power Sta-
tion. Atmospheric Environment, 1:389-410.
Preston, E. M. and T. L. Gullett. 1979. The Bioenvironmental Impact of a
Coal-Fired Power Plant. Fourth Interim Report, Colstrip, Montana,
EPA-600/3-79-944, U.S. Environmental Protection Agency, Corvallis,
Oregon.
Raynor, G. S., M. E. Smith, and I. A. Singer. 1974. Meteorological Effects
of Sulfur Dioxide Concentrations on Suburban Long Island, New York.
Atmospheric Environment, 8:1305-1320.
Williams, M. D. and P. C. Tourangeau. Unpublished Work.
Zanetti, P., P. Melli, and E. Runca. 1977. Meteorological Factors Affecting
S02 Pollution Levels in Venice. Atmospheric Environment, 11:605-616.
Zeedijk, H., and C. A. Velds. 1973. The Transport of Sulfur Dioxide Over a
Long Distance. Atmospheric Environment, 7:849-862.
Zimmer, C. E. and G. A. Jutze. 1964. An Evaluation of Continuous Air Quality
Data. J. Air Pollut. Control Assoc., 14:262-266.
19
-------
SECTION 2
PLUME AND AEROSOL PROPERTIES NEAR COLSTRIP
C. C. Van Valin, R. F. Pueschel,
D. L. Wellman, and G. M. Williams
ABSTRACT
Analysis of plume aerosol samples with the trans-
mission electron microscope has shown that the small-
size fraction (~0.2 |Jm diameter) is dominated by
sulfate particles. The samples included a few nitrate-
containing particles of diameter 0.4 to 1.0 |Jm, but
most of the nitrate-containing particles were found in
the diameter range of 1-2 (Jm.
Measurement of the plume cross-sectional dimen-
sions with the research aircraft has shown that the
smoke plume is often wider than predicted by Gaussian
diffusion models. The vertical extent as measured was
less than the predicted value, however. Tracking of ,/
the plume with the aircraft to distances up to 50 km
downwind has shown that the plume may be diverted as
much as 20° from the prevailing wind direction by
terrain features.
INTRODUCTION
In May 1975 the Atmospheric Chemistry Program Area (ACPA), Atmospheric
Physics and Chemistry Laboratory, with the Atmospheric Spectroscopy Program
Area (ASPA), Wave Propagation Laboratory, began a research program at the Hay
Coulee site, 12 km southeast of Colstrip, Montana, The work was designed to
characterize the aerosol content of the Colstrip power plant plume and its
environs. The ACPA and ASPA mobile laboratories were operated at the Hay
Coulee site during May-June and August-September of 1975, 1976 and 1977; the
ASPA laboratory was moved to the Battelle No. 1 site for the last week of the
August-September 1977 period to take advantage of the line-of-sight view of the
plume at the point of exit from the power plant smoke stack. The information
from these field studies has been reported (Abshire et al., 1978; Van Valin
et al., 1979).
When its mobile laboratory was in use at Hay Coulee, ACPA also operated
its research aircraft in the vicinity of Colstrip, beginning in August 1975.
There were additional aircraft field trips in February 1977, June 1978, and
February and March 1979. Information from the aircraft operation during 1975
was reported in the Third Interim Report (Abshire et al., 1978); a partial
20
-------
report of the 1976 and 1977 projects was included in the Fourth Interim Report
(Van Valin et al., 1979). This report includes some unevaluated aspects of
the 1976 and 1977 projects and some information from the 1978 and 1979 pro-
jects, e.g., plume trajectory and diffusion, fine particle formation, gas
reactions, and secondary aerosol characteristics.
From 1975, the Colstrip research project was carried out concurrently
with other research projects related to energy production at the Four Corners
power plant near Farmington, New Mexico, at the Kennecott Copper Company near
Salt Lake City, Utah, in the vicinity of the petroleums-refining complex in the
south Los Angeles basin, and recently, in the vicinity of the Homer City,
Pennsylvania, power plant. The findings from the study of the plume of the Four
Corners and Colstrip power plant plumes, particularly in rate and type of secon-
dary particle formation, in spite of considerable differences in plant capacity,
emissions control methods, siting, and climate.
The Four Corners power plant production capacity is about 3 times that of
the Colstrip plant. Four Corners uses five generating units; emissions from
the two largest units, 800 MW each, are controlled by electrostatic precipita-
tors, and from the other three units by lime scrubbers. Research findings
from the Four Corners plant follow: (1) only 2 percent (by mass) of the
particulate emission was water soluble and contained the elements S, Na and
Ca, whereas the 98 percent that was insoluble contained the elements Si, Al,
and Fe (Parungo et al., 1978a); (2) sulfur is preferentially concentrated at
the surface of 40 percent to 60 percent (by number) of the fly ash particles
(Parungo et al. , 1978a); (3) the sulfur-containing layer on fly ash consists
of ammonium sulfate or bisulfate, and the thickness of the layer is inversely
related to the diameter of the fly ash particle (Mamane and Pueschel, 1979a);
(4) the number of secondary particles formed by a gas-to-particle conversion
process is far greater than the number of primary (fly ash) particles (Pueschel
and Van Valin, 1978), and (5) Mamane and Pueschel (1979a) concluded that long-
range transport of both sulfur dioxide and sulfates occurs downwind from the
power plant.
DISCUSSION AND RESULTS
Aerosol Composition
Nuclepore filter samples were used for the determination of ice nucleus
concentration and for elemental composition determination, by the scanning
electron microscope with energy dispersive X-ray analysis (SEM-EDX). The
histograms in figures 2.1a-2.1e, show elemental occurrence, shape, and mineral
type of samples collected during August-September 1977. Figures 2.2a-2.2c
represent samples collected during February and March 1979; they were analyzed
for 22 elements. Figures 2.2b and 2.2c include relative signal strength,
indicating elemental abundance as well as distribution. Thus, in Figure 2.2b,
where 47 percent of the emitting particles contain Si, the X-ray emissions due
to Si account for 34.5 percent of all X-ray emissions; 71.5 percent of emitt-
ing particles contain Ca, but Ca emissions account for only 14.7 percent of
total emissions; and 9.4 percent of emitting particles contain Fe, but Fe
21
-------
80 —
60 —
40 —
20 —
^ 0.1 < d <
^ 0.5 < d <
1
I
1
Na Mg Al Si
0.5 pm
T.O fj.m.
I
I
I
^^1
Colstrip
1
Ij
S Cl K Ca Ti
1
J
Cr Fe
27 Aug77
Power Plant
P771
i
i
0935-0950 MDT
Plume. 2 mi d.w.
5500' MSL
t<^^*^^m%m2m%*i
J
? ~™ y. S
.E o 2 —
§ s •= -S «
in "I" °-a! <£ c
S ° "5 3
— in o o
u "i -ex:
'r Jm
- £
1
0)
Q.
K
"ra
o
c
2
—
\
I
vt ^ C
0) SO
U en Q-
e ^ S
<£ '£ K
Q. z
Figure 2.la. SEM-EDX analysis of particles collected on Nuclepore filters and
classified according to size, for 27 August 1977, 0935-0950 MDT.
^ 0.1 < d <
UA3
__
^ 0.5 < d <
' J
Rx^ ^ \v
0.5 /im
1.0 Mm
I
%%m^m^%^<^
JL
^5 Colstrip
I
I
yy
^v
1
i
29Aug77 0713-0728 MDT
Background, 5500' MSL
1
,
^
^
i
mr///////^/////^ r
80
60
40
20
Na Mg Al Si S Cl K Ca Ti Cr Fe
to c
KO
15 «
S fe
CO.
.0 (T
Q.
CO
Figure 2.1b. SEM-EDX analysis of particles collected on Nuclepore filters and
classified according to size, for 29 August 1977, 0713-0728 MDT.
22
-------
80 —
60 —
40 —
20 —
_
^| 0.1 < d < 0.5 /im
^ 0.5 < d < 1.0 ^m
1
I
1
I
1
!
i
\
!
J
si
I
i
i
j
11
^7^^
!
^
1
|
|
f^riletrin
29 Aug 77 0738-0753 MDT
Power Plant Plume,
2-3 mi d.w.
4300' MSL
1
i
Na Mg M Si S Cl K Ca Ti Cr Fe °> 15 j> S
g ~ .0 .y
l| II 1
a ° ra
1
CD
C
i
o
I
"
Q>
_C
I
01
0
ro
0.
r 3> «
y/
y/ —
6 S
3 §
in Q.
^ CC
~ O
CD •^
Q. Z
Figure 2.1c.
SEM-EDX analysis of particles collected on Nuclepore filters and
classified according to size, for 29 August 1977, 0738-0753 MDT.
80 —
60 —
40 —
20 —
^ 0.1 < d < 0.5 ^m
1^^*
Colstrip
29 Aug 77
0802-0817
' Power Plant Plume
^ 0.5 < d < 1.0 /im
:
—
I
^
W//////////////////A
\
\
I
i
s
\
i
W ^
^
i
V///////////////////A
J
Na Mg Al Si S Cl K Ca Ti Cr
I
I
i
MDT
, 10 mi d.w.
4100' MSL
i
'mfm////mm
JL
Fe J> gs |
|
-------
100
80
60
40
20
Colstrip
2 Sep. 77 0731-0746 MDT
Power Plant Plume, 15 mi d.w.
4000' MSL
0.1 < d < 0.5/xm
0.5 < d < T.Ofj-m
J
1
Na Mg Al Si
Cl
Ca Ti Cr Fe
5 §
.y oc
C o
Q_ <^
Figure 2.1e.
SEM-EDX analysis of particles collected on Nuclepore filters and
classified according to size, for 2 September 1977, 0731-0746 MDT.
100
80
60
40
20
—
-
-
to
^f
Jl
I
^
1
I
^
/f
^
fv.
1 0.54/0.
ww
1
Colstrip
27 Feb79 1630 - 1723 MST
8 km Downwind in Plume
0.6fj.m Median Particle Diameter
410 X-Ray-Emitting Particles, 150 Non-Emitters
OH Percent of emitting particles containing the element
^2 Percent of all particles containing the element
-
P?! ra
I
^
\
«> - - ffl
h~ o> i~- i-- y —
m co in in ^
Na Mg Al Si P S Cl K Ca Ti V Cr Mn Fe Ni Cu Zn Au Hg Pb Br U Non-
Emitters
Figure 2.2a. SEM-EDX analysis of particles collected on Nuclepore filters,
27 February 1979, 1630-1723 MST. Numbers appearing above the
very short bars indicate the percent of particles containing
the element.
24
-------
100
80
60
40
20
Colstrip
21 Mar 79 0917 -'0944 MST
11 km Downwind in Plume
0.4/im Median Particle Diameter
53 X-Ray-Emitting Particles, 90 Non-Emitters
EU Percent of emitting particles containing the element
E Percent of all particles containing the element
O EDX peak area x relative peak height (S = 100%)
o o •-
^ h- p
do oo oo oo oo oo do oo
Na Mg Al Si PS
nm—Y ' + f' ^— J^ -g mjf t- (t mmtf i .... _^—^— tin
Cl K Ca Ti V Cr Mn Fe Ni Cu Zn Au Hg Pb Br U Non-
Emitters
Figure 2.2b.
SEM-EDX analysis of particles collected on Nuclepore filters,
21 March 1979, 0917-0944 MST. EDX peak area times relative peak
height indicates emission strength and thus indicates the rela-
tive concentration of the element in the sample.
100
80
60
40
20
Colstrip
21 Mar 79 1255- 1355 MST
24 km Downwind in Plume
0.35/im Median Particle Diameter
49 X-Ray-Emitting Particles, 135 Non-Emitters
EH Percent of emitting particles containing the element
EZ3 Percent of all particles containing the element
C3 EDX peak area x relative peak height (S= 100%)
Na Mg Al Si P S Cl K Ca Ti
V Cr Mn Fe Ni Cu Zn Au Hg Pb Br • U Non-
Emitters
Figure 2.2c.
SEM-EDX analysis of particles collected on Nuclepore filters,
21 March 1979, 1255-1355 MST. The numbers above very short bars
indicate the percent of particles containing the element. EDX
peak area times relative peak height indicates emission strength
and thus indicates relative concentration of the element in the
sample.
25
-------
emissions account for 39.3 percent of total emissions. We see that Ca is
rather widely distributed in the plume particles at this distance from the
stacks, but nevertheless is apparently only a minor constituent, whereas Fe
occurs in a few particles, but these few particles must be large and/or
composed primarily of Fe.
In all of the examples except Figure 2.1b, which is a background sample,
the elemental composition is like fly ash; Ca is prominent in all plume
samples, and more so at the shorter downwind distances, as one would expect
from this lime-scrubbed plume. Cr and Fe are prominent in the August-September
1977 samples, but much less so in the other samples. We think such variations
in the metallic elements may be due to differences in coal or soil composition.
Twelve samples collected during 1977 and 1979 were analyzed for the 22 ele-
ments shown in Figures 2.2a-c; of these, eight contained U in trace concentra-
tions. All samples consisted of >50 percent spherical particles; plume and
background samples were the same in this respect. While the spherical parti-
cles in the plume are mostly fly ash, those in background samples from other
geographic areas are primarily organic or are nitrates and therefore non-
emitting in the EDX analysis (containing no element heavier than F). Although
particles with no EDX response (i.e., the non-emitters) represented 20 percent
to 60 percent of all particles in the examples of Figures 2.1 and 2.2, analy-
sis of other samples from these time periods has indicated an even wider
variation from nearly zero to 75 percent or more. For two reasons the re-
ported incidence of non-emitting particles must be considered only an approx-
imation. First, the existence of electronic noise in the instrument requires
that an elemental emission exceed the noise level by a predetermined factor.
Second, compounds with an appreciable vapor pressure, such as organic sub-
stances, sulfuric acid, or to a lesser extent, the ammonium salts of sulfuric
acid, sometimes evaporate in the vacuum of the SEM, especially under the heat
of the electron beam. This also brings into question the accuracy of the
analysis for sulfur-containing particles. We show that large numbers of very
small sulfate particles exist in the plume. Obviously, the SEM-EDX analysis
for sulfur-containing particles should be regarded only as a lower limit. The
particles showing only sulfur are made up of a compound in which S is combined
with a cation that emits X-rays that are not detectable in our analyzer.
Likely cation candidates are ammonium ions and the particles are assumed to be
ammonium sulfate or ammonium bisulfate.
In addition to the Nuclepore filter samples analyzed with the SEM-EDX,
samples of plume aerosol particles of diameter 0.2 to 2.0 |Jm were collected on
electron microscope screens mounted in front of the impactor jet nozzles of a
four-stage Casella cascade impactor carried on the research aircraft. The
screens were vacuum coated with BaCl2 or with nitron (1,4-diphenyl-3,5-
endoanilino-4,5-dihydro-l,2,4-triazole) for analysis of the sample for the
presence of sulfate (Mamane and Pueschel, 1979a; Mamane and dePena, 1978) or
of nitrate (Bigg et al. , 1974; Mamane and Pueschel, 1979b) with the trans-
mission electron microscope (TEM). Figure 2.3a shows photomicrographs of
particles collected on a BaCl2~coated grid that was mounted in the fourth
(final) stage of the cascade impactor. The network of small, elongated
particles of uniform size and distribution is the BaCl2 substrate. The more
or less circular spots of a larger size than the BaCl2 are the reaction pro-
duct from the BaCl2 and the sulfate particles that impacted on the grid. When
26
-------
Figure 2.3a.
TEM photomicrographs of particles collected on the fourth stage
a Casella cascade impactor. Elongated particles of uniform
dispersion are the BaCl2 coating; round spots indicate sulfate
particles that have reacted with the BaCl2, and are about double
the diameter of the original particle. Left photomicrograph
includes several particles that include mineral (or fly ash)
centers.
viewed in the TEM, only four of the darkest spots in the left-hand photomicro-
graph have opaque centers that are fly ash particles. The diameters of the
BaS04 reaction spots are approximately twice those of the original sulfate
particles; the smallest particles that appear in these photomicrographs were
therefore about 0.2 |Jm initial diameter.
Figure 2.3b shows photomicrographs of particles collected on grids that
were coated with nitron before (right) or after (left) collection of the
sample. The nitron method is a specific test for nitrate by the formation of
needle-like crystals or masses of needles. There are no nitrate-containing
particles in the left-hand photomicrograph, and only two in the right that may
contain nitrate. These impactor samples from power plant plumes do not show
appreciable numbers of nitrate particles on the fourth stage impactor grids,
but do frequently show nitrate in the next larger (third stage) size fraction,
i.e., most of the nitrate is found in particles of 1 |Jm diameter and larger,
and there are few nitrate-containing particles smaller than 1 (Jm (Van Valin et
al., 1980). This finding is consistent with those of other studies (Parungo
et al. , 1980; Pueschel et al., 1979a) . The explanation is that nitrate and
27
-------
0 *
/im
Figure 2.3b.
TEM photomicrographs. Left: particles collected on an impactor
grid that was coated with nitron following collection. Right:
grid was coated with nitron before sample collection. Spots
with clear halos that shade into a darker center represent
sulfates and are about three times the diameter of the particle
they represent.
nitric acid particles are very hygroscopic and grow to above-micron-sized
droplets rapidly at slight supersaturations or even subsaturation. The atmos-
pheric residence times of these particles is limited by rapid fallout and
washout. The examples in Fig. 2.3b are typical of fourth-stage nitron-coated
samples, however; the spots with the clear halos and increasing density toward
the centers are typical of (though not specific for) sulfates, and the spots
with neither halos nor shading are probably fly ash. Fly ash or other mineral
particles with sulfate coatings would show as uniformly dense, sharp-edged
spots surrounded by the clear halo. In the left-hand photomicrograph there
are two particles that may be mineral, with very faint halos; most of the rest
are probably sulfates. The right-hand example includes eight spots that are
probably fly ash particles of ~0.5 |Jm diameter without the sulfate coating.
The diameters of the spots produced by sulfate on nitron are about three times
those of the original particles; thus, the sulfate particles in these two
nitron-coated samples were from 0.2 to 0.6 [Jm diameter, approximately.
Ice Nuclei
Coal contains a variety of impurities that, if occurring as constituents
of aerosol' particles, may be capable of acting as ice nuclei (IN). Because of
the importance of the ice phase in the formation of precipitation, the poten-
28
-------
tial for inadvertent weather modification as a consequence of increasing
energy production in the western states has been of concern to atmospheric
scientists. However, research has given an ambiguous picture of the impact of
power plant effluents on the atmospheric ice nucleus budget. Schnell et al.
(1976) and Pueschel et al. (1979b) found no difference in the number density
of IN between samples from the plumes of the Four Corners and Homer City,
Pennsylvania, power plants and samples of the surrounding clean air. Parungo et
al. (1978a) found similar results when samples were handled in a manner similar
to that described by Schnell et al.; however, Parungo et al, discovered that IN
.could be activated in the plume samples by subjecting the sample to .high water
supersaturation, high temperature, or high vacuum, or combinations of these
treatments. Research in our laboratory suggests that tests using different
modes of nucleation will produce different results, i.e., the filter method
used by Schnell et al. and Parungo et al. depends upon the deposition (reverse
of sublimation) or condensation-followed-by-freezing mechanisms, while the
acoustic counter method depends upon the contact nucleation mechanism (Langer
et al., 1967). Because of its ease of sample collection, the filter method of
Bigg et al. (1963), as modified to use Nuclepore filters and a static water
vapor source, was used exclusively for the airborne measurements at Colstrip,
and for the majority of samples at the mobile laboratory at Hay Coulee. The
acoustic counter at Hay Coulee was used to the extent possible. The first
report on NOAA/ERL activity at Hay Coulee (Abshire et al., 1978) presented ice
nucleus measurements by both methods, showing equivalent concentrations of IN
in the atmosphere prior to power plant operation. The second report (Van
Valin et al., 1979) showed that the ice nucleus concentration, as measured by
the acoustic counter method during a well-documented plume exposure at Hay
Coulee (23 August 1977, 0828-0854 MDT), was raised by nearly an order of
magnitude. Similar episodes of plume exposure at Hay Coulee, while not always
as striking, produced similar increases .in the IN-acoustic counter method.
During times of non-exposure to the plume -this method showed IN concentrations
in the range 1-4 £ 1, which is entirely consistent with the levels reported
earlier. However, filter-method samples collected during these times of plume
exposure were not different from those collected during non-exposure times.
Hay Coulee filter-method ice nucleus measurements for the four field periods
in 1976 and 1977 are listed in Table 2.1. In addition to the comparison of IN
concentrations between plume and non-plume exposures, the samples were segre-
gated by time of collection (morning, mid-day, and afternoon), but the variabil-
ity is so large that one cannot have confidence in a possible pattern. In like
manner, the plume and ambient atmosphere samples cannot be shown to be differ-
ent; the present methods are not sufficient to identify the plume as a source
of IN. The filter-method (modified from Bigg et al., 1963) IN measurements of
samples collected with the aircraft are presented in Table 2.2. Again, no
difference can be shown between the plume and ambient IN concentrations.
The indication that IN were not generated during operation of coal-fired
power plants has caused some puzzlement and skepticism. The measurement of
very much higher IN concentrations with the acoustic counter during plume
exposure periods and the observance of enhanced snowfall downwind from a power
plant (Parungo et al. , 1978b) lends weight to the belief that the filter
method used in this study is adversely affected by the surface absorption of
substances, probably soluble sulfates, that deactivate the deposition and
condensation-followed-by-freezing modes of nucleation. The contact mode of
nucleation is therefore affected to a significantly smaller degree. Further
research on the ice-nucleating characteristics of plume aerosols continues.
29
-------
TABLE 2.1. ICE NUCLEUS CONCENTRATIONS (£ i) AT HAY COULEE AS MEASURED BY THE
FILTER METHOD, EXPRESSED AS MEANS AND ONE-SIGMA STANDARD DEVIATIONS
Period
All samples
Before 1000 MDT
1000-1300 MDT After 1300 MDT
May-June 76 (n=34) 2.03±1.63 (n=9) 3.67±1.42 (n=13) 1.8011.64 (n=12) 1.0710.58
Aug-Sept 76 (n=45) 2.78+1.15 (n=17) 3.1211.15 (n=17) 2.5411.27 (n=ll) 2.6210.91
May-June 77 (n=44) 2.4311.56 (n=22) 2.8311.53 (n=ll) 2.5811.89 (n=ll) 1.5410.75
Aug-Sept 77 (n=44) 6.4313.59 (n=15) 7.7415.00 (n=ll) 5.8713.08 (n=18) 5.6812.02
Period
In Plume
(all times)
Ambient
(all times)
May-June 76
Aug-Sept 76
May-June 77 (n =9) 3.2112.16 (n=35) 2.2411,32
Aug-Sept 77 (n=10) 4.6412.54 (n=34) 6.9613'.71
Note: n is the number of samples.
Plume Dispersion
Plume dispersion rates determine the volume of the atmosphere in which
compositional changes take place and the land area that is affected by de-
position of pollutants. Cross-sectional characteristics, e.g., concentration
profiles of aerosols and gases at various distances downwind, are used to
measure plume dispersion. We present here several examples of plume profiles.
Figures 2.4 and 2.5 show the Aitken nucleus (AN) concentrations found in the
horizontal and vertical profiles on 1 September 77. Figures 2.6 and 2.7 are
similar profiles done on 31 August 77. Figures 2.8 and 2.9 are representa-
tions of the maximum change from background of the other constituent con-
centrations measured at the same time. Figures 2.10 and 2.11 are profiles of
the power plant plume representing AN and ozone concentrations, respectively,
as measured 25 km downwind from the power plant on 16 February 1977. The
inset in the upper left-hand corner of Figure 2.10 represents the plume
dimensions 0.5 km downwind from the stacks, according to Turner (1970); the
symbols on the x-axis 10 m on either side of the centerline indicate stack
spacing and diameter. Figure 2.12 shows the ozone profile a fraction of an
hour later and at a downwind distance of 8 km. Figure 2.12 represents only
the upper part of the plume following a reduction of wind velocity from ap-
proximately 13 m sec 1 to 4 m sec 1 and a decrease in atmospheric stability
resulting in either plume looping or fumigation.
30
-------
TABLE 2.2. ICE NUCLEUS CONCENTRATIONS (£~ ) FROM AIRBORNE SAMPLE COLLECTION
AS MEASURED BY THE FILTER METHOD, EXPRESSED AS MEANS AND ONE-SIGMA
STANDARD DEVIATIONS.
Period
In Plume
Ambient
May-June 76
Aug-Sept 76
Feb 77
May-June 77
21-25 Aug 77
27 Aug-2 Sept 77
Aug-Sept 77 (all flights)
(n=3) Ave =2.56
(n=5) 1.3610.31
(n=7) 1.5810.68
(n=2) Ave =3.04
(n=!4) 4.5614.06
(n=l6) 4.3713.82
(n=!4) 0.8810.60
(n=38) 1.1911.02
(n=5) 0.9310.34
(n=48) 1.6111.29
(n=19) 2.0711.96
(n=7) 7.3015.26
(n=26) 3.4813.87
Note: n is the number of samples.
I
East Side
Stacks
OH-
Over US. 212
between Ashland
and Ljme Deer
Southwest of
Ashland, between
Tongue River and Huffs
West Side
26 30
Distance Downwind, km
46
Figure 2.4.
Horizontal profile, based on AN measurements, of the Colstrip
power plant plume from 12 to 45 km downwind on 1 September 77,
1240-1500 MDT.
31
-------
Elevation,
AGL
53Jr
Aitken Nuclei (X 10 cm3)
Sept. I, 1977 1233-1249 MOT
Wind' MO" 7m sec'
Vertical Plume Profi
12 km Downwind
Near Hay Coulee
Figure 2.5. Vertical profile, based on
AN measurements, of the
Colstrip power plant plume
on 1 September 1977,
1233-1249 MDT.
These three sets of plume measurements demonstrate the inhomogeneities
along the x, y, and z axes that are typical for coal-fired power plant plumes.
These findings are in contrast to theoretical treatments of plume dispersion
that predict Gaussian distributions of plume properties along both the z and y
axes (e.g., Turner, 1970).
A similar discrepancy is found if one compares the measured with the
predicted plume cross sections. Table 2.3 compares the measured and predicted
plume dimensions (Turner, 1970) at distances from the stacks of l/2, 8 and 25 km
for stability classes D, C, and D respectively. It follows from Table 2.3 that
theory overestimates plume height z, but underestimates the plume width.
This discrepancy follows from at least three factors: (a) measurements
with the mobile laboratory at Hay Coulee show, on the very short term, wind
direction commonly varying over a range of 40 to 70 degrees, (b) terrain
factors in the vicinity of Colstrip contribute additional deflection (see
Influence of topography on plume transport), (c) Aitken particles and ozone
are far from ideal as plume tracers, especially because the profile dimensions
and measured constituent concentrations indicate a reactive mixture involving
a particle-generating process. The third conclusion follows from Figures 2.4
and 2.6, which show that the AN concentrations at 40-50 km downwind from the
stacks are equal, or nearly so, to those found 12 km downwind, in spite of an
observed dispersion of the plume. These findings conclusively show, however,
that the plume does not follow a Gaussian distribution, which is the basis for
all theoretical plume dispersion models.
Additional examples of plume dimensions and constituent concentrations
follow. On 30 August 1977 a vertical plume profile was obtained near Lame
Deer, Montana, which is 29 km on a straight line from Colstrip (Figure 2.13).
The plume's external profile and underlying topography are shown in Figure
2.14. The shape of the profile and the reasons for the plume's location near
Lame Deer are discussed below. Figure 2.15 summarizes the plume constituent
32
-------
Aitken Nuclei (xio3cnr3)
i i i I i i i
0 10
20 30
Distance downwind (km)
Figure 2.6. Horizontal profile, based on AN measurements, of the Colstrip
plume on 31 August 1977, 1024-1140 MDT.
i i i I i i i i I i i i
Aitken Nuclei (xio'cnT3)
3000
2000
2000
Left
Plume width ( m ), from centerline
Right
Figure 2.7. Vertical profile, based on AN measurements, of the Colstrip plume
on 31 August 1977, 0915-1024 MDT, 12 km downwind.
33
-------
100
12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
O
JD
Q.
Q.
50
O
100 •
50-
0 •
-------
(m) (ft)
5500
1600[-
0)
13
D
1500
1400
1300
1200
1100
10001—
5000
4500
4000
3500
3100
1 T
Surface
_ Elevation
I
I
I
I
J_
I
_L
103
10
m-3'
Aitken Nuclei
25 50
ppb
Ozone
0.5 1.0
x10~4m ~1
bscat
50 100
ppb
NO
50 100
ppb
NO,
Figure 2.9. Maximum deviation of plume constituent concentrations from
background during a vertical profile 12 km downwind, 31 August
1977, 0915-1024 MDT.
a consistent ratio in all cases. Figure 2.18 is the representation of plume
vertical profile limits at the same time, as determined by the point at which
the 03 concentration was reduced from background by approximately double the
instrument electronic noise level. Figure 2.19 and 2.20 are the corresponding
constituent concentrations and profile obtained about 12 km downwind. The
wind direction and velocity during this time period were SSW at 0-3 m sec 1
at plume altitude; stability index according to Turner (1970)could have been
D or C, considering the isothermal layer at plume elevation. Table 2.4 lists
the predicted and measured dimensions of the plume. Theory understates both
plume width and thickness at the shorter distance, but overestimates both
at the 12 km distance. Height estimates are rather uncertain and subject to
extreme variation according to the atmospheric stability. The measured plume
is thinner in overall vertical dimensions at 12 km than at 4 km; this appears
35
-------
211-
Plume dimensions est.occ.to Turner, 1970
1400-
1300-
_l
CO
1 1200-
-o
u
5 1100-
1000-
611 411 2(1 C 20 40 60
Plume width (m)
Aitken Nuclei (Xio3crrr3)
!6Feb77
0920-0946MST
Colstrip, Montana
25 km (15 mi) downwind from power plant
(ft)
-4500
-4000
-3500
-3100
Surface elev. (approx.)
I
I
j I
J I
J I
_|_
1500
Left
500 £ 500
Plume width (m),from centerline
1000 1500
Right
Figure 2.10. AN vertical contour profile of the Colstrip plume on 16 February
1977, 0920-0946 MST; the inset represents the theoretical plume
boundary and measured maximum concentration of AN.
CO
5
(m)
1400-
1300-
1200-
1100-
1000-
I I I | I I I I | I I I I | I I
Ozone
16 Feb 77
0920-0946 MST
Colstrip, Montana
25km (I5mi) downwind from power plant
1 I I
-4000
-3100
43 < ppb < 47
Surface elev. (approx.)
1500
1000
Left
500 £ 500
Plume width (m) from centerline
1000
1500
Right
Figure 2.11. Ozone vertical contour profile of the Golstrip plume on
16 February 1977, 0920-0946 MST, corresponding to AN concen-
trations shown in Figure 2.10.
36
-------
(m)
1400-
1300-
•g 1200-
_3
** 1100-
1000-
1—I I I
Ozone
!6Feb77
1002-1014 MST
Colstrip, Montana
8km (5mi) downwind from power plant
-3100
Surface elev. (approx.)
I I I
I I I I
I I I I I I I I I
J I L
1500 1000
Left
500 I 500
Plume width (m) from centerline
1000 1500
Right
Figure 2.12. Ozone vertical contour profile of the Colstrip plume.
TABLE 2.3. MEASURED AND PREDICTED DIMENSIONS FOR THE COLSTRIP POWER PLANT
PLUME ON 16 FEBRUARY 1977
Distance (km) •
^ (Stability Class D)
8 (Stability Class C)
25 (Stability Class D)
Width (m)
Measured
—
>1800
2800
Predicted
100
1400
2400
Height (m)
Measured Predicted
— 40
135
-------
Figure 2.13. Topographic map of the Colstrip area (U.S. Geological Survey)
38
-------
(m) (ft)
5500
leoor
1500
c7> 1400 -
o>
I 1300
1200
1100
1000
5000
4500
4000
3500
3000
Plume Width (km), from Centerline
Figure 2.14. Vertical boundary profile of.the Colstrip plume according to
measured AN concentrations, 30 Aug. 1977, 0645-0735 MDT, 29 km
downwind.
(m) (ft)
5500
Approximate Surface Elevation
I I I 1 1 1
1600r
1500
1400
CO
5
£ 1300
<
1200
1100
5000
4500
4000
3500
I I
West Leg (1st V2)
East leg (2nd Vi)
I
I
I I
I I
J I
1 I
\
J I
I I
I I
5 10 0
x103cm~3
Aitken Nuclei
25 50 75
ppb
Ozone
0.5 1.0
50 100
ppb
NO
50 100
PPb
NO,
25 50
ppb
SO,
Figure 2.15. Maximum deviation of the plume constituent concentrations from
background during the vertical profile of Figure 2.14.
39
-------
(m) (ft)
10500
3000
2500
,10000-
0)
-a
3
•S 2000
1800
1600
1400
1200
1000
4500
.4000|r
3700
3500
3100
;' i 1 1 1 ' • — Q
0 40 80 120 160 200 240 0 0.4 0.8 1.2 10J
104
NO, O3 (ppb)
bscat(x 10~4m~1)
AN (cm^3)
20 40 60 0 10 20
Relative Temp °C
Humidity, %
Figure 2.16. Constituent concentrations during a descending spiral from
3600 m MSL, 28 Aug. 1977, 0544-0621 MDT, 4 km downwind.
(m) (ft.)
5500
1500
1400
j
'- 1300h
5000-
< 1200
1100
1000
4500 -
4000-
3500-
3100
- Surface Elev. (Approx.) -
i i i i i
0 5 10 15 20 25 30 0 25 50 75 0 .5 1.01.52.0^50 | 150 ( 250 ( 350 , 450 500 0 25 50 75 100
(X103 cm~3)
Aitken Nuclei
ppb
O,
(x 10-4m~1)
bscat
0 100 200 300 400
ppb
NO
ppb
SO,
Figure 2.17. Maximum deviation of plume constituent concentrations from
background during a vertical profile immediately following the
profile of Figure 2.16 (0635-0723 MDT).
40
-------
(ft)
(m)
5000
1500r
1400
1300
1200
1100
1000L
4500
4000
-3600
3100
-Surface Elevation (Approx.)
i l i i i l i
Figure 2.18.
Vertical boundary profile
of the Colstrip plume
according to measured 03
concentrations, 28 Aug. 1977,
0635-0723 MDT, 4 km downwind.
500 400 300 200 100 £ 100 200 300 400 500
(m)
(m) (ft)
5500
1600r-
c>
•a
3
1500
1400
1300
1200
1100
1000
5000
4500
4000
3500
3100
1 T
Surface Elevation
' (Approx.)
I
I
I
I
I
"1 T
I
l
J_
"1 T
I
Q. —
Q —
I
5 10 15
x103crrr3
Aitken Nuclei
25 50 75
ppb
0.5 1.0 1.5
bscat
50 100 150
PPb
NO
25 50
ppb
SO,
Figure 2.19. Maximum deviation of plume constituent concentrations from
background during a vertical profile, 28 Aug. 1977, 0800-0834
MDT, 12 km downwind.
41
-------
(m) (ft.)
5000
1500 r-
1400
1300
1200
1100
1000L
4500-
4000-
3500
\ \ \ \ r
i i i r
Surface Elevation (Approx.)
J I I I L
I I l
900 800 700 600 500 400 300 200 100 (£_ 100 200 300 400 500 600 700 800 900 1000
(m)
Figure 2.20. Vertical boundary profile of the Colstrip plume according to
measured 03 concentrations, 28 Aug. 1977, 0800-0834 MDT, 12 km
downwind.
to be the consequence of a slight increase in wind velocity during the inter-
vening time of roughly 1 hour. The plume at 12 km also shows evidence, on its
top surface, of wind shear; our pibal sounding at Hay Coulee, as well as
aviation weather reports, indicated winds above plume altitude of 270°-300°.
On 29 August 1977, at 5 km and 16 km downwind, repeated penetrations of
the plume in its densest part were made by aircraft. The data are presented
in Table 2.5. Repeated passes at the same location (16 km downwind) and
altitude (1220 m MSL) permitted a look at the relationships among the gases
03, N02, and NO. Figure 2.21 shows the concentrations of N02 and 03 rela-
tive to NO. The two lines were obtained by least squares fit; the coeffi-
cients of determination (r2), were 0.66 and 0.58, indicating rather modest
fit. The background 03 concentration of 68 ppb where NO and N02 were near
zero was not included in the curve-fitting calculations. Both N02 and 03 are
higher in the less dense portions of the plume, representing both more dilu-
tion and a greater degree of reaction between 03 and NO, but the 03 decline
is greatly in excess of that required to produce the small amount of N02 actu-
ally measured at least in consideration of the plume expansion as indicated
by diminution of NO concentration. It is apparent that the reaction with NO
is not the primary loss mechanism for 03.
42
-------
TABLE 2.4. MEASURED AND PREDICTED DIMENSIONS FOR THE COLSTRIP POWER
PLANT PLUME ON 28 AUGUST 1977
Distance (km) Width (m) Height (m)
Measured Predicted Measured Predicted
(stability index) (stability index)
DC DC
4 900 500 740 445 160 440
12 1600(max.) 1300 2000 260 300 1200
1150(av.)
Influence of Topography Upon Plume Transport
It often appears, that plumes are guided to a degree by the topography of
the underlying terrain; they seem to be ducted along valleys rather than
across the parallel higher terrain, diverted to a noticeable extent from the
direction of the prevailing wind upon encountering a valley, or deflected
around higher terrain. We present here three examples of the apparent alter-
ation of plume travel direction by terrain.
(1) On 30 August 1977 at 0615 MDT our research aircraft encountered the
plume near Colstrip, followed it southeast about 16 km to the Rosebud
Creek valley, where it appeared to divert toward the southwest. A
vertical profile of the power plant plume near Lame Deer from 0645 to
0735 MDT was constructed, and then the plume was followed farther south-
west along the Rosebud Creek valley as far as the town of Busby. The
profiled plume edges and representation of approximate terrain contours
are presented in Figure 2.14. Maximum deflections of constituent concen-
trations are shown in Figure 2.15. At 0600 MDT, the Miles City airport
reported surface winds of 3.6 m sec 1 at 280° and 7 m sec * at 270° at
1830 m MSL. The record from our mobile laboratory at Hay Coulee shows
the 0600 MDT wind at 270° and less than 2 m sec'1. From approximately
2150 MDT, 29 August, to 0200 MDT, 30 August, the wind direction had been
between 330° and 360° at Hay Coulee, with an apparent frontal passage
about 2300. Maximum wind velocity when the front passed was about 8 m
sec 1, with 3-5 m sec * during the 1.5 hours following. Between 0200 and
0600 MDT the wind gradually veered to the west; velocity was less than
2 m sec 1. The plume found at Lame Deer and Busby probably left the stacks
between 2150 and 0200 MDT, being carried a few degrees east of due south.
After a few hours travel at an average heading of 165°, the plume was de-
flected to the west by the Rosebud Creek valley and by the higher eleva-
tions surrounding Badger and Garfield Peaks (elev. 1348 and 1315 m respec-
tively). Near Lame Deer, the shape of the plume also appears to have
been affected by the valley of Lame Deer Creek (Figure 2.15), which flows
into Rosebud Creek from the south.
43
-------
TABLE 2.5. MAXIMUM AND BACKGROUND CONCENTRATIONS OF PLUME CONSTITUENTS, EXPRESSED AS MEANS AND
ONE-SIGMA STANDARD DEVIATIONS
Location
Downwind
1310 m MSL
approx 5 km
Background
5 km
1220 m MSL
approx. 16 km
Background
16 km
Time
(MDT)
0738-0755
0738-0755
0800-0818
0800-0818
NO
(ppb)
2961112
(n=15)
~5 ppb
49.5120.0
(n=10)
~5 ppb
NO2
(ppb)
< 40
< 5 ppb
13.9+5.2
(n=9)
< 5 ppb
bscat
(xio"4 m"1)
1.3110.32
(n=15)
0.2110.05
0.50±0.08
(n=12)
0.20
O3
(ppb)
approx. 0
65.8+1.6
15 .4+5.9
67.8+1.2
AN
(xlO3 cm"3)
16.814.7
(n=13)
~ 1.0
5.2+4.1
(n=10)
~ 1.0
SO2
(ppb)
87+55
(n=13)
ND*
35+6
(n=5)
ND*
Not detectable
(2) On 31 August 1977, for several hours following 0800 MDT, the wind
direction at Hay Coulee was between 290° and 330°, with velocities
averaging 5 m sec 1, with gusts to 9 m sec 1. During this time the wind
recorded at stack height was 290°-330°, at 10 m sec"1, gusts to 15 m
sec 1. A vertical profile of the plume 12 km downwind was constructed for
the period 0915 to 1024 MDT; for the period 1024 to 1140 MDT a horizontal
profile of the plume from 12 to 48 km downwind was made (Figures 2.7 and
2.6). Figure 2.8 summarizes the maximum change of the plume constituents
at the various distances downwind, and gives the aircraft flight altitude
and surface elevation. Along Beaver Creek, 34 to 48 km downwind, the
elevations of the upland parallel to the valley are as much as 500 ft
(152 m) higher than the valley floor; the plume was neatly centered along
the valley. The Beaver Creek valley is oriented exactly so as to not
require an alteration of the plume's average direction of travel, but in
the situation where the wind direction varies over a 40° range, the
probability of the plume centering on Beaver Creek valley without any
ducting influence would have been very low. Thus, we consider.this to be
an instance of ducting.
(3) On 1 September_1977, winds aloft at 1300 MDT were 300°, 8 m sec"1,
gusting to 15 m sec ^ Beginning at 1030 MDT, winds at Hay Coulee were
320° ± 30° at 7 m sec 1 average, with a very gradual change in direction
to 360° by 1300 MDT. Wind direction stayed constant at 360° until 1540
MDT, when it moved to 340° for 45 min, then resumed the 360° for most of
the time until about 1900 MDT. The research aircraft constructed a
vertical plume profile near Hay Coulee from 1233 to 1240 MDT, then
commenced a horizontal plume profile out to a distance of 45 km downwind,
ending at 1500 MDT (Figures 2.5 and 2.4, respectively), several kilometers
southwest of Ashland in the Tongue River valley. It appears that the
plume was deflected to the east around the Badger and Garfield Peaks high-
land, then ducted slightly west of due south by the Tongue River valley,
while the prevailing wind was neither westerly enough in the first in-
stance, nor northeasterly enough in the second to have caused such a plume
44
-------
29 Aug. 1977
Colstrip
16kmd.w. 4000' MSI
0800-0818
60
50
40
30
20
10
_|_
I
• •• N02
o o o 0,
I
I
Figure 2.21.
10 20 30 40 50 60 70 80
NO, ppb
Least squares fit of NC>2
and 03 vs. NO concentrations
in the Colstrip plume in
regions of different plume
densities.
trajectory. In short, the evidence in this case indicates a deflection in
the direction of plume travel by as much as 20° from the prevailing wind
direction. This effect of terrain on plume dispersion introduces a
complexity that cannot be explained by existing models.
Meteorology
Wind direction, wind velocity, relative humidity, and temperature re-
corded at the mobile laboratory at Hay Coulee were previously reported for
the periods 27 August to 7 September 1976, 20 May to 4 June 1977, and 20 August
to 3 September 1977 (Van Valin et al., 1979). As a supplement to the ground-
based observations the same parameters were measured at certain times within
the given periods by aircraft, by radiosonde release, and by pilot balloon
(pibal) release.
As Van Valin et al., (1979) reported during the August to September
periods the surface winds were typically light and variable at night, with
daytime velocities of 2 tc 6 m sec"1, usually westerly. During the May-June
periods, surface wind flow was less obviously diurnal, there was more cloudiness
and more convective storm system activity. During all four periods nighttime
and early morning inversions between 600 and 1200 m AGL with stability classes
E and F were so common as to be almost the rule. Usually the change in wind
direction and velocity with height became less distinct as surface heating
took place during the day. Under the conditions of light and variable wind
with the temperature inversion, the power plant plume rises to the inversion
layer and then spreads out into a broad fan as it slowly moves away from the
source area. This condition usually does not result in ground impact by the
plume, save for the gravitational settling of particles. In fact, in the few
hours preceding sunrise one often sees the formation of another inversion well
below the stack height, which further insulates the surface from plume contact.
The daytime counterpart during this diurnal regime involves the mid-morning
45
-------
development of convective activity, with concomitant looping of the plume and
intermittent ground contact, followed by sufficient boundary layer heating to
overcome temperature inversions and allow the plume to mix upward.
During frontal passage, with the cooler air, greater cloudiness, more
persistent wind, and neutral temperature structure with "coning" of the plume
occurs. Plume contact with the ground does not occur for some appreciable
distance (on the order of several kilometers) but becomes rather constant with
increasing distance, and is confined within an arc of roughly 20°. During
stable temperature conditions with light-to-moderate wind, the plume may be
carried horizontally with little vertical movement for long distances. This
is the state that permits the best appraisal of plume dimensions and integrated
loading, and represents meteorological conditions that exist during most of
the airborne plume measurements dealt with in this report.
During winter, the fronts come more frequently, the average wind speed is
greater, there is more cloudiness, and convective activity is reduced. There
is also a greater frequency of long-lived low-level inversions; these are
often below the level of the effective stack height and thus have the effect
of shielding the surface from plume contact. Meteorological factors extant
during the examples of plume measurement are included with the discussion of
those measurements.
CONCLUSIONS
This research is a continuation of work reported previously (Abshire et
al., 1978; Van Valin et al., 1979). It is the nature of research that con-
tinuing effort in a given direction, with new equipment and methods, not only
goes forward using previous work and conclusions as a foundation, but also
produces new information that modifies previous conclusions. This report
includes new information that modifies the conclusions of the Fourth Interim
Report in respect to the composition and size range of the plume aerosol, the
formation of secondary aerosols, and the generation of IN. To the two types of
particles reported previously for downwind distances of 10 km or more, i.e., a
man-made aerosol consisting of spherical sulfate-containing particles in the
0.1-0.4 pm diameter range and a natural aerosol of irregularly shaped alumino-
silicates of > 0.6 (Jm mean particle diameter, must now be added a third man-
made aerosol consisting of nitrate-containing droplets mostly of 1-2 (Jm
diameter. The formation of nitrate-containing droplets constitutes a signifi-
cant sink for the NO formed in the power plant combustion chambers. In the
Fourth Interim Report we concluded that IN concentrations, as measured by the
acoustic counter method, were increased by an order of magnitude. Evaluation
of IN concentrations by the filter method, however, shows no increase above
background concentrations. We thus qualify the earlier conclusion by stating
that potential IN are deactivated in relation to sublimation and condensation-
followed-by-freezing modes of ice crystal formation, probably by surface
absorption of sulfates, but that the contact mode of nucleation is affected to
a much smaller extent.
Plume cross-section measurements at various distances downwind have
shown the plume, under conditions of C or D stability, to be usually wider
than predicted by Gaussian diffusion models. The necessary consequences
46
-------
of greater-than-predicted width is that a greater land area will be affected
by deposition or by attenuation of solar radiation.
Plume trajectories have been found to be influenced by topography to such
an extent that a plume's direction of travel is altered by as much as 20° from
the direction of the prevailing wind. This introduces too much complexity for
existing models of plume diffusion and transport.
REFERENCES
Abshire, N. L., V. E. Derr, G. T. McNice, R. Pueschel, and C. C. Van Valin.
1978. Integrated Aerosol Characterization Monitoring, Colstrip, Montana.
In: The Bioenvironmental Impact of a Coal-Fired Power Plant, Third
Interim Report, Colstrip, Montana, E. M. Preston and R. A. Lewis, eds.,
EPA-600/3-78-021, U. S. Environmental Protection Agency, Corvallis,
Oregon, pp. 291-321.
Bigg, E. K., S. C. Mossop, R. T. Read, and N. S. C. Thorndike. 1963. The
Measurement of Ice Nucleus Concentrations by Means of Millipore Filters.
J. of Appl. Meteor. 2:266-269.
Bigg, E. K., A. Ono, and J. A. Williams. 1974. Chemical Tests for Individual
Submicron Aerosol Particles. Atmos. Environ. 8:1-13.
Langer, G., J. Rosinski, and C. P. Edwards. 1967. A Continuous Ice Nucleus
Counter. J. of Appl. Meteor. 6:114-125.
Mamane, Y., and R. G. dePena. 1978. A Quantitative Method for the Detection
of Individual Submicrometer Size Sulfate Particles. Atmos. Environ.
12:69-82.
Mamane, Y., and R. F. Pueschel. 1979a. Oxidation of S02 on the Surface of
Fly Ash Particles under Low Relative Humidity Conditions. Geophys. Res.
Lett. 6, pp. 109-112.
Mamane, Y., and R. F. Pueschel. 1979b. A Method for the Detection of Individ-
ual Nitrate Particles. Atmos. Environ., accepted for publication.
Parungo, F., E. Ackerman, H. Proulx, and R. Pueschel. 1978a. Nucleation
Properties of Fly Ash in a Coal-Fired Power-Plant Plume. Atmos. Environ.
12:929-935.
Parungo, F. P., P. A. Allee, and H. K. Weickmann. 1978b. Snowfall Induced by
a Power Plant Plume. Geophys. Res. Lett. 5:515-517.
Parungo, F. P., R. F. Pueschel, and D. L. Wellman. 1980. Chemical Character-
istics of Oil Refinery Plumes in Los Angeles. Atmos. Environ., 14:509-
522.
Pueschel, R. F., F. P. Parungo, E. W. Barrett, D. L. Wellman, and H. Proulx.
1979a. Meteorological Effects of Oil Refinery Operations in Los Angeles.
NOAA Tech. Memo. ERL APCL-22, 62 pp.
47
-------
Pueschel, R. F. , R. C. Schnell, H. K. Weickmann, and D. L. Wellman. 1979b.
Aerosol and Ice Nuclei Measurements in the Plume of the Homer City, PA,
Power Plant. Geophys. Res. Lett. 6:371-374.
Pueschel, R. F., and C. C. Van Valin. 1978. Cloud Nucleus Formation in a
Power Plant Plume. Atmos. Environ. 12:307-312.
Schnell, R. C., C. C. Van Valin, and R. F. Pueschel. 1976. Atmospheric Ice
Nuclei: No Detectable Effects from a Coal-Fired Power Plant Plume.
Geophys. Res. Lett. 3:657-660.
Turner, D. B. 1970. Workbook of Atmospheric Dispersion Estimates. Public
Health Service, 999-AP-26, 84 pp.
Van Valin, C. C., R. F. Pueschel, and F. P. Parungo. 1980. Sulfate and
Nitrate in Plume Aerosols from a Power Plant Near Colstrip, MT. In:
Environmental and Climatic Impact of Coal Utilization, J. J. Singh and A.
Deepak, eds., Academic Press, New York, pp. 35-48.
Van Valin, C. C., R. F. Pueschel, D. L. Wellman, N. L. Abshire, G. M. Lerfald,
and G. T. McNice. 1979. Aerosol Characterization in the Vicinity of
Colstrip, MT. In: The Bioenvironmental Impact of a Coal-Fired Power
Plant, Fourth Interim Report, Colstrip, MT, E. M. Preston and T. L.
Gullett, eds., EPA-600/3-79-044, Section 1, U.S. Environmental Protection
Agency, Corvallis, Oregon, pp. 2-52.
48
-------
SECTION 3
OBSERVATIONS OF TWO LICHEN SPECIES
IN THE COLSTRIP POWER PLANT VICINITY
S. Eversman
ABSTRACT
Two lichen species, Usnea hirta (L.) Wigg.
and Parmelia chloroohroa Tayl. appear to be
exhibiting no adverse effects from 862 emissions
from the Colstrip coal-fired power plant.
Variations in percentages of algal plasmolysis,
respiration rates, and relative chlorophyll
extract absorbances do not have a pattern that
can be readily attributable to power plant
emissions.
INTRODUCTION
Sulfur dioxide is known to adversely affect lichens (LeBlanc and Rao,
1975; Eversman, 1978). Community analysis, generally documenting community
simplification in polluted areas, has been the most common method of
detecting or describing patterns of pollution (Gilbert, 1973; Nash, 1972).
Since there were insufficient numbers of species for community analysis on the
upper trunks and branches of ponderosa pines in the Colstrip area, the major
approach of this study has been to observe two lichen species for physical and
anatomical changes. Two species of lichens, Usnea hirta (L.) Wigg. and Parmelia
ohloroohroa Tayl., from the Colstrip power plant vicinity have been observed
in the laboratory for indications of stress from power plant emissions, since
coal-burning began in September, 1975.
MATERIALS AND METHODS
During 1978, U. hirta samples were collected from 15 ponderosa pine
(Pinus ponderosa L.) sites 1-60 km from the Colstrip power plant. Monitoring
sites within 10 km of Colstrip received U. hirta transplants from a generally
east-facing site in Custer National Forest 60 km ESE of Colstrip (Site P10,
Figure 3.1).
49
-------
Ponderosa pine branches containing U. hivta were wired onto appropriate
branches of ponderosa pines facing the power plant (Sites Pl-8, P17-19).
Sites P10-16 have abundant native U, hirta. Other transplants were made to
sites P15~.and P16 as transplant controls; i.e, , to observe possible effects
of transplanting.
Northern Cheyenne
Reservation
Figure 3.1. Map showing lichen collection sites. P1-P16 are
ponderosa pine sites with Usnea hirta; G1-G7 are
grassland where Paxmelia ohloYoehroa is collected.
Sites P8, 15, and 16 have both native and trans-
planted U. hirta (n,t). Site P7 has two buttes
(a, b); U. hirta is collected from the top and
bottom of a hill at P14. P10* = U. hirta trans-
plant source. TC = Taylor Creek fumigation
sites (ZAPS).
50
-------
Parmelia chlorockpoa was collected from four grassland sites near EPA
exclosures (G1-G4), three other sites near ponderosa pine sites (P3, P4, P18),
and the control grassland site (G7).
Respiration rates were determined manometrically for all samples.
Chlorophyll extracts were made with hot methanol and absorbance was recorded
at 665 nm. Percentages of algal cell plasmolysis were determined according to
previously described methods (Eversman, 1978).
RESULTS AND DISCUSSION
Neither U. hirta nor P. ahlorochroa exhibited consistent visible signs
of S02 injury in the Colstrip vicinity. Samples from sites closest to Col^
strip were examined for algal plasmolysis rates compared with control levels.
While some sites had a mean plasmolysis rate above control rates, differences
were not significant (Figure 3.2). Paxmelia chlorochpoa from site G5 (Kluver
50
45
ul 40
o
g 35
i *°
UJ
g. 25
o
« 20
I 5
I 0
5
I
I
I
PI06 10 18 18 2 5 1612 10
a b
May July August
Usnea hirta
G7 3 5 G3 5 67 3 5 I 4PI8P4
May June July
Pormelia chlorochroa
SITE, MONTH OF COLLECTION
Figure 3.2. Percentages of algal plasmolysis of U, hirta and P, chloroohroa
from monitoring sites. Percentages are means of three samples
± 0.95 confidence interval. Cross-hatched bars are control
sites (P10, G7). Solid bars are field collection sites.
Confidence intervals = t...05(standard error)/Jn for each mean.
51
-------
East) was exceptionally yellow and had an unusually high algal plasmolysis
rate (35%) in May, 1978. However, by June the population appeared normal
again; plasmolysis percentage dropped and remained as low as other samples
during 1978. The general level of plasmolysis of samples from the Colstrip
area was below the plasmolysis rates of samples exposed to known levels of
S02 in ZAPS fumigation plot B (See Section 9 ).
Respiration rates in 1978 showed no pattern associated with distance
from the power plant (Figure 3.3). Differences among sites appeared to be
due to natural variabilities, not air pollution.
1200
I 100
1000
900
800
i
_f 700
. 600
O
UJ
IE 500
C/5
Z 0
O
O
CM
O
_ 700
a.
600
500
400 -
Usneo hirto
I
P 10 6 P 10 7 7 18 16 16
ob n t
MAY JUNE
P 10 2 18 18 3 4 5
a b
JULY
12 13 M 14 IS IS PlO
o b n t
AUGUST
Parmelio chlorochroa
t
"Figure 3.3.
07 03 05
MAY
07 G3 05 04 01
JUNE
07 03 OS 04 01
JULY
P4 PIS P3
07
AUGUST
SITES, MONTH OF COLLECTION
Respiration rates of [7, h-lvta and P, chlopoohvoa from
monitoring sites in the Colstrip area, 1978. Rates are
given as means of five samples ± 0.95 confidence interval.
Cross-hatched bars are control samples (sites PlO, G7).
Solid bars are monitoring sites. Confidence intervals =
t. 05 (standard error) /j~n for each mean.
52
-------
Relative chlorophyll extract ajpsorbance values for both species were
generally not significantly less than control values (figure 3,4), and they
were higher than values obtained from samples from plot B on the ZAPS sites
(Section 9 ), Differences, however, were not significant.
1.50
1.25
1.00
0.75
gO.50
CD
o
O
0.25
CO
O
CO
CD
LJ
Usneo hirto
I
'hi
I
P 10 19 7 7 16 16
0 b n t
10 5 18 ,18 3 4 i II 12 13 14 15 16 16
ob o b n n t
LU 0.50
CC.
0.25
o.oo,
JUNE
Pormelio chlorochroo
JULY
AUGUST
I
I
G7 3 4 5 I
JUNE
07 3 4 5 I PI8 P3P4 G7
JULY AUGUST
SITES, MONTH OF COLLECTION
Figure 3.4. Relative chlorophyll extract absorbance at
665 nm for U. hirta and P. chloTOchroa. Values
are means of three samples ± 0.95 confidence
intervals. Cross-hatched bars are controls
(P10, G7). Solid bars are monitoring sites.
Confidence intervals = t.95(standard error/Jn
for each mean.
53
-------
Samples collected from the ZAPS site plots B, C, and D have consistently
been bleached, with yellowing of the thallus occurring in 30-60 days after
transplanting onto the fumigation plots (Eversman, 1978). No such yellowing
has been observed in samples from the Colstrip area (with the one temporary
exception in P. ehlorodhpoa in May, 1978). All collections made during 1978
exhibited the "normal" green thallus color.
CONCLUSIONS
If impact from power plant emissions were occurring on the lichens, one
would expect a gradient of respiration rate changes, decreased algal
plasmolysis, increased relative chlorophyll extract absorbance, and decreased
yellowing with increasing distance from the power plant. The two lichen
species showed no such trends, and therefore did not appear to be affected
by Colstrip power plant emissions in 1978.
REFERENCES
Eversman, S. 1978. Effects of Low-Level S02 on Usnea hirta and Parmelia
chloroehroa. Bryologist 81: 368-377.
Gilbert, 0. L. 1973. Lichens and Air Pollution. In: The Lichens,
Ahmadjian, V. and M. Hale, eds. Academic Press, New York. pp. 443-472.
LeBlanc, F. and D. Rao. 1975. Effects of Air Pollutants on Lichens and
Bryophytes, In: Responses of Plants to Air Pollution. Mudd, J, B,
and T. T, Kozlowski, eds. Academic Press, New York. pp. 237-272.
Nash, T. H. III. 1972. Simplification of the Blue Mountain Lichen
Communities Near a Zinc Factory. Bryologist 75: 314-324.
54
-------
SECTION 4
PLANT COMMUNITY MONITORING
IN THE VICINITY OF COLSTRIP, MONTANA
J.E. Taylor, W.C. Leininger and M.W. Hoard
ABSTRACT
Since 1974, plant community studies have been con-
ducted near Colstrip, Montana on four exclosures estab-
lished by the EPA and on three contiguous pristine
knolls along Rosebud Creek. All study locations were
within 20 kilometers of the coal-fired plants at Col-
strip. Comparisons of species cover, diversity, and
density from 1974 to 1978 indicated that a stabili-
zation of grazing deferment effects occurs over time.
This should allow better isolation of air pollution ef-
fects in the future. Canopy coverage varied sub-
stantially among sites due to the relative amounts of
graminoids and lichens. Forbs did not vary markedly.
Mosses and shrubs were not important. No trends in
lichen cover can be ascribed to pollution because
of the confounding effects of grazing deferment in
earlier years. Western wheatgrass density varied
among sites, but intra-site variation was slight,
suggesting this species is a good monitoring
organism. Species richness and evenness contributed
about equally to species diversity. Standard errors
for species diversity were mostly within 5% of mean
values. Photographic plots were charted in detail
and added to the long-term monitoring record.
INTRODUCTION
This research was begun July 15, 1974, to monitor bioenvironmental ef-
fects of coal-fired electric generating plants in southeastern Montana. The
objective is to monitor the effects of stack emissions on native rangeland
vegetation. This objective is being approached through canopy coverage and
plant density transects and by aerial and ground-level photography and
species diversity analysis.
55
-------
The data reported in this section cover only our research activities in
the vicinity of Colstrip, Montana. In addition to four exclosures estab-
lished and maintained by the EPA, three relatively pristine knolls are being
examined (Figure 4.1). Previous work and study site descriptions have been
summarized by Taylor and Leininger (1977 and 1979).
MATERIALS AND METHODS
Plant community analysis and photographic monitoring are discussed in
detail below.
Plant Community Analysis
Canopy Coverage
Canopy coverage estimates were made on all study sites near the middle
of August, 1978. Sample plots within study areas were located by placing a
cord with meter-spaced knots in a random meandering pattern through the sam-
ple area. Then 2 x 5 dm plot frames were placed at each knot and canopy
coverage estimated using the procedures of Daubenmire (1959) . Preliminary
studies showed that 2 lines of 25 frames each provided the optimum sampling
efficiency.
Western Wheatgrass Density
In the same samples which were used for canopy coverage, numbers of
western wheatgrass plant units were determined. Plant units were defined as
single or clustered vegetative and/or reproductive stems which appeared to
originate at a common basal node. Thus, a reproductive culm closely sur-
rounded by one or several vegetative culms was considered a plant unit, and
an isolated vegetative culm was a plant unit.
Species Diversity
The diversity index used was the Shannon-Weaver function (Shannon and
Weaver, 1949):
Where H' = Index of diversity
c „ • ,,. S = number of species
H' = - E^f1 I°g2 ~*T~~ Mi = canopy of ith species
l=l N = total canopy of all species
Percent canopy estimates were used to calculate diversity.
In addition to diversity, these data were used to calculate evenness
(equitability) and species richness.
Evenness was calculated with the equation of Pielou (1969).
H' Where J' = Evenness
Iog2 S H' = Shannon-Weaver function
S = total number of species
56
J' =
-------
MAP LOCATION
BRANDENBERG
POWDER
L^'i
xvxx>x77i_ ,
xxxxxxxxxx"
xxxxxxxxxx
xxxxxxxxx
xxxxxxxxxxxxxxx
xxxxxxxxxxxxxx
xxxxxxxxxxxxx
xxxxxxxx
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
xxxl/xxxxx
xxxxxxxx. .
xxxxxxxxxx
xxxxxx
xxxxxxxxxxxxxx
xxxxxx
xxxxxxxxxxxxxx.
xxxxxxxxxxxxxx
xxxxxxxxxxxxxxx.
. xxx
XXXXXXXI
• ' • ' ' A
1
Figure 4.1.
Study sites in the vicinity of Colstrip. (1 = Hay
Coulee, 2 = Kluver West, 3 = Kluver North, 4 = Kluxer
East, 5 = McRae Knolls)
57
-------
Species richness is the numerical sum of species which were used to
calculate H?.
In order to reduce variances and still be able to calculate standard er-
rors, data were grouped into five consecutive frame sums before H' or J' were
calculated. S was the total number of species encountered in each five-frame
group.
Photographic Studies
Ground-level photography provides a detailed record of plant species,
phenology, and pathologic signs. This assists in the interpretation of
aerial imagery. Also, vertical ground photo plots may be measured and
analyzed for cover, number, frequency, pattern, and plant volume. Plant
volume, which can be used to estimate biomass, may be obtained by combining
canopy coverage and height, the latter measured with a parallax wedge. Plant
density and pattern also can be studied from these pictures.
At the Hay Coulee and Kluver sites, two photo plots were established in
each exclosure, while one photo plot was placed on each of the three McRae
knolls.
The photo plots are 1 meter square in area, and are marked for reloca-
tion. Each is photographed in color and black-and-white film emulsions.
Stereoscopic photography is used for ease of plant identification. Most of
the plots have also been photographed with infrared color film. More details
are given by Taylor et at., 1976.
Aspect photographs are made from vantage points within and overlooking
study areas. These are taken with color and infrared color film, the latter
to compare with aerial coverage.
For each photo plot an index of species identification has been prepared.
During summer, 1978, precise maps were drawn of each photo plot. Each
square-meter plot was divided into a 10 x 10 grid, each unit a square deci-
meter. Within each grid cell, plant species with substantial area were de-
lineated; smaller plants were indicated symbolically, and numbers of western
wheatgrass plant units were recorded.
The combination of plot photographs and plot indices makes a permanent
record of species presence and distribution. Sequential records allow the
evaluation of temporal changes.
During 1978, no new aerial photography was acquired in the Colstrip
area. Color photography from the helicopter flight in 1977 was used to de-
velop detailed vegetation maps of the Hay Coulee and Kluver sites.
58
-------
RESULTS AND DISCUSSION
Plant Community Analysis
Canopy Coverage
The generalized plant composition among study sites is shown in Figure
4.2. Species vary, reflecting differences in sites, microclimates, and pre-
vious use. Therefore, plant group data are presented, which more clearly
portray inter-site variation.
Total canopy coverage varies among sites, primarily reflecting differen-
tial site potentials. More significant differences may be observed by con-
sidering the relative coverages of plant groups within sites.
The patterns of variation in total cover were similar to those of 1977.
This suggests that inherent site differences may be replacing deferment ef-
fects as the predominant influences on plant species composition. If this
is the case, 1978 and later observations should begin to show any pollution
effects more clearly than earlier data, since the relative confounding in-
fluence of deferment will be diminished.
As should be expected in these vegetation types, graminoids dominate on
all sites. The differences among sites which were observed in 1977 had es-
sentially disappeared in 1978. This reflects the more favorable growing
season in 1978 and the lesser influence of grasshoppers. The canopy cover
of graminoids ranges from 39 to 59 percent across the sites.
Forb cover is remarkedly uniform among sites. The lowest and highest
forb covers recorded were 9 and 14 percent, respectively. There were no
particularly conspicuous forb species associated with sites. Shrub and moss
cover likewise was considered unimportant. Mosses never were conspicuous,
and shrubs were clearly a function of pre-treatment distribution. 1978 can-
opy coverage values are presented in Table 4.1.
With the initiation of the Colstrip studies in 1974, pre-treatment base-
line vegetation data collecting was begun within newly established grazing
exclosures. Comparable grazed plots were not established with available re-
sources because there was no way to quantify grazing in ways in which
grazing would be the only variable inside and outside exclosures.
The lack of grazed controls has caused some problems in the interpreta-
tion of data from the exclosures. By 1975 it was clear that exclosure ef-
fects (ie., grazing deferment effects) were producing substantial increases
in plant growth. This was particularly marked on the Kluver East site, which
had previously been used as a feeding area for horses and bulls and as a
season-long pasture. Under protection from livestock grazing, the dominant
vegetational aspect changed to crested wheatgrass, apparently from resi-
dual seed which had come from grass hay and which had been trampled into
the soil.
59
-------
100
90-
Graminoids
Forbs
Lichens
eo
ra
i-,
oj
o
80-
70-
60.
50-
40-
30-
20-
10-
|
4-M
—
I
1
I
=
7?<
ID
^ M
I
Shr
M
Hay K.West K. North K.East Knoll Knoll Knoll
Coulee ABC
Figure 4.2. Canopy coverage on the Colstrip sites, August, 1978.
This rejuvenation of plant vigor greatly over-compensated for any air
pollution effects. Thus, yield, or even species composition changes have
been impossible to positively ascribe to either pollution or deferment,
since these factors are confounded. The vegetation which exists in the
exclosures is the product of deferment and of any air pollution influences
which may exist, as well as soil, climate, weather variation, previous
grazing, and all the other many interrelated variables which function in
rangeland vegetation communities.
To further cloud this issue, the research which has been conducted on
the ZAPS sites has only recently shown subtle signs of incipient effects of
SC>2 pollution, and there the levels of pollutants are higher and more con-
tinuous than those likely to be experienced in power plant situations.
In the long term, species composition changes should follow pollution
stresses, and each year's data collection will help to characterize the
natural amplitude of plant growth and vigor which respond to both climatic
variation and pollution. Meanwhile, more restrictive indicators must be
60
-------
TABLE 4.1. CANOPY COVERAGE (PERCENTAGE) FOR COLSTRIP STUDY SITES, 1978
SPECIES
GRAMINOIDS
Ajropyron cristatum
A. smithii
A. spicatwn
Aristida longisf-ta
Bouteloua gracilis
Bromus japonicus
B. teotorum
Qzlamovilfa longifolia
Car ex f Hi folia
Koeleria oristata
Poa pratensis
P. sandbergii
Schedonnardus paniculatus
Sporobolus cryptandrus
Stipa comata
S. viridula
Vulpia octoflora.
FORBS
fehillea millefoUwn
Alysstm desertorum
Anbrosia psilostaahya
Androsace occidentalis
Hay Coulee
20 Aug
11.50
27.95
2.50
.65
8.55
.60
4.55
1.30
.75
.55
.70
.30
Kluver West
17 Aug
4.20
.35
3.45
2.50
1.60
1.55
.40
3.50
.30
26.05
.30
.85
Kluver North
16 Aug
6.40
6.55
2.20
.30
.05
7.20
17.70
2.05
.25
.10
.40
Kluver East
18 Aug
29.50
6.25
.05
1.75
2.75
.30
.05
2.40
.10
.05
4.85
.30
1.30
McRae
Knoll A
19 Aug
2.45
1.15
2.75
1.05
13.85
1.00
.35
.40
15.10
.45
.65
.80
McRae
Knoll B
19 Aug
2.40
7.30
.45
6.00
1.70
4.40
4.55
10.10
5.15
.05
.10
McRae
Knoll C
18 Aug
4.35
2.40
4.35
1.75
2.75
23.45
.75
.30
13.25
.20
1.90
.25
Mtennaria rosea
Arabis holboellii
Aster sp,
Astragalus sp.
Qmelina microoarpa
.10
.10
.10
.10
.05
.40 .05
.05
-------
TABLE 4.1 (continued)
SPECIES
McRae McRae McRae
Hay Coulee Kluver West Kluver North Kluver East Knoll A Knoll B Knoll C
20 Aug 17 Aug 16 Aug 18 Aug 19 Aug 19 Aug 18 Aug
FORBS (continued)
Comandra wribellata.
Conyza oanadensis .05
Drdba sp. .05
D. nemevosa
Erysimum asperwn .05 ,15
Erigeron divevgens .05
Evolvulus pilosus .35
Gaura cocc-lnea
Gilia oongesta
Grindelia squarTosa
Heteicofheoa villosa .80 1.50
Laotuoa sevriola
Liatris punctata .10
Lygodesnria junoea 3.00
Mammi'llax"ia missouriensis .05
Mediaago sativa
Opuntia fragilis .35
0. polyacantha
Phlox hoodii .85 .10
Plantago patagonica .10 1.05
P. spinulosa 1.55 .05
Polygala alba
Polygonum vivipamm
Psoralea argophylla .15
Sphaeralcea coccinea .40 .90
Taraxacum offioinale .90
Tragopogon dubius 5.05 3.00
Miscellaneous forbs
.05
,30
.05
.20
.00
.05
10
.10
.30
.80
.05
.45
.30
.20
.05
.35
.05
.15
.35
.20
.15
.10
.05
.25
5.95
.05
.05
.15
.45
5.10
1.45
.05
1.15
.10
.40
.70
.60
.05
.50
.10
.30
.05
.05
3.70
.05
.60
.20
2.50
.50
.70
.10
1.90
.05
.30
.75
.05
3.00
1.20
.30
.05
.05
.25
.35
.70
-------
TABLE 4.1 (continued)
Co
SPECIES
HALF SHRUBS AND SHRUBS
Artemisia draounoulus
A . frigida
A . tridentata
Ceratoides lanata
Chrysothamnus nauseosus
Xanthocephalwrn sarothrae
OTHERS
Bare ground
Litter
Moss
Lichens
Rock
TOTAL VEGETATION
TOTAL GRAMINOIDS
TOTAL FORBS
TOTAL SHRUBS
Hay Coulee
2 Or Aug
V~ — : — ^
3.00
3.85
12.40
75.85
.05
3.50
.15
80.20
58.90
10.90
6.85
Kluver West
17 Aug
.60
16.50
60.85
10.40
.75
66.85
44.20
11.65
.60
McRae McRae McRae
Kluver North Kluver East Knoll A Knoll B Knoll C
16 Aug 18 Aug 19 Aug 19 Aug 18 Aug
.30
8.40
16.50
62.95
11.95
.20
77.70
42.70
14.35
8.70
2.90
13.90
69.55
9.15
.15
69.55
48.05
9.45
2.90
6.85
79.00
1.65
.25
52.15
38.55
11.95
3.90
1.50
.05
.30
25.10
57.50
.45
2.05
2.30
61.65
42.10
11.30
5.75
.30
1.10
13.00
69.55
.45
.95
1.70
65.60
53.55
9.25
1.40
-------
used to evaluate pollution effects. The most likely approaches to use in
this connection are to seek sensitive indicator groups or species and look
for physiological effects or direct pathological signs which can be defi-
nitely ascribed to pollution. Of these, the most useful in this work has
been the former, using lichens as indicator species.
Many species of lichens have long been considered to be sensitive indi-
cators of pollution stresses, both in the scientific literature and in the
CFPP study (LeBlanc and Rao, 1975; Eversman, 1978 and 1979; Taylor and
Leininger, 1979).
Figures 4.3 and 4.4 show the lichen cover for 1975 prior to the opera-
tion of the power plants. The data show the lichen response to seasonal
changes in the absence of air pollution. The trend is for lichen cover to
decrease as the season progresses. This corresponds to the dying or
maturing of vegetation, loss of soil moisture, and more difficult growing
conditions with the advancing season. Some sites have more lichen growth
than others. Sites that have recently been grazed (Figure 4.3) tend to
have short vegetation and abundant bare ground where lichen growth thrives,
especially in early season. On sites where recent grazing had not taken
place (Figure 4.4), lichen levels were lower in early season and the trend
toward reduction in plant population as the season progressed was less pro-
nounced. This is probably caused by the increased vegetative volume that
restricts available open space for lichen growth and shields lichens from
seasonal shifts in microclimate.
Figure 4.5 and 4.6 show the yearly lichen cover variations from 1974 to
1978. Since data were gathered only in late August or early September in
1976-1978, the late season samples for 1974 and 1975 were used in these com-
parisons.
The sites with recent grazing history (Figure 4.5) show variability in
cover between years. The variation between data collected in 1974 and 1975
(prior to power plant operation) probably reflects annual differences in
growing conditions. The significant increases in lichen cover during 1976
may be due to favorable growing conditions- In 1977 and 1978 canopy cover-
age appeared to be stabilizing except for the high-level cover observed at
the Kluver East site in 1977. This may have been caused by the atypical
vegetational composition on that site composed of crested wheatgrass rather
than a native mixture as on the other locations.
The sites with no recent grazing history (Figure 4.6) show less variabi-
lity in cover between years. Lichen cover may gradually stabilize following
grazing deferment.
Western Wheatgrass Density
Western wheatgrass density was measured in 1978 for several reasons. It
is one of the principal forage species in the Northern Great Plains for both
domestic and wild animals. It is found on numerous kinds of sites. Its
64
-------
16 -,
14 -
12 _
10 -
g 8 -J
6~S 4 -
2 -
Hay
Coulee
Kluver
West
Kluver
North
Kluver
East
16 "
14 ~
12 ~
10 ~
a
CO O
'^1
111
6 ~
a
o
1 4 -
2 ~
E=
1
//
Knoll
A
ED June
H July
fcj
PJ7J September
luuQ
=
i
1
i i
Knoll Knoll
B C
Figure 4.3. Seasonal changes in lichen coverage
on the EPA established sites near
Colstrip, 1975.
Figure A.4. Seasonal patterns in lichen coverage
for the McRae Knolls, 1975.
-------
24 -.
22 -
20 -
18 -
16 -
12 -
10 -
6 -
2 -
Hay Coulee
Kluver West.
Kluver North
Kluver East
20 -
18 -
16 -
14
12
10 -
6 -
2 -
1974
1975
1976
1977 1978
T~
1974
—r—
1975
1—
1976
1977
1978
Figure 4.5. Lichen coverage on the EPA sites
near Colstrip, 1974-1978.
Figure 4.6. Lichen coverage near Colstrip
taken late in August or early
September, 1974-1978.
-------
rhizomatous growth form allows direct comparisons among its density, cover,
and weight. Finally, it has been studied in considerable detail by other
project elements (Dodd et al., 1979; Rice et at., 1979; Heitschmidt, 1977).
In 1978, western wheatgrass density varied widely among sites (Figure 4.7)
ranging from 2 to 19 plant units per 1/10 square meter. Although inter-site
variation is high, intra-site variability is low, so temporal changes should
be easy to detect when the anayses are completed.
CN
O
CO
111
Q
to
CO
ra
u
on
4-1
nl
d)
(3
VJ
0)
4J
CO
22 -
20 -
18 -
16 -
14 -
12 -
10 -
6 -
2 -
I
Hay
Coulee
I I I
K.Wcst K.North K.East
Knoll
A
Knoll
B
Knoll
C
Figure 4.7. Density of Western wheatgrass on the Colstrip
sites, 1978. (Mean value + 1 standard error).
67
-------
Species Diversity
Plant species diversity, evenness, and species richness for 1978 are shown
in Figure 4-8 and 4-9.
As in past years, the diversity is highest at Hay Coulee and McRae Knoll
B, and lowest at Kluver East. With the exception of Kluver East and Knoll B
the values are quite comparable. The relationship between range condition
and diversity, which has been previously discussed by Taylor et al., (1975),
is evident with Knoll B being the most diverse and also having the highest
range condition and Kluver East being the least diverse, with the lowest
condition. Standard errors are within 5% of the mean for all locations but
Kluver East. The high variability in this site can be largely attributed to
the great difference in diversity between the north and south replicates,
caused by the uneven distribution of crested wheatgrass.
Species richness and evenness parallel diversity and appear to contri-
bute about evenly to it (Figure 4.9). The high evenness value for the
Kluver West site pulls up the diversity on the site even though it has a
very low richness.
Diversity, richness, and evenness continue to be sensitive measures of
community differences.
Photographic Studies
All permanent photo plots were photographed and mapped in August. These
photos and plot maps are on file and constitute an important record of
species composition, coverage, etc. Aspect photos of the sites were taken
at the same time.
Site maps for the Kluver plots were also constructed in August and will
be presented in the final report. McRae Knoll study sites were mapped the
previous year. No aerial photography in the vicinity of Colstrip was ob-
tained by our project in 1978.
CONCLUSIONS
Comparison of species cover, diversity, and density from 1974 to 1978
indicates that a stabilization of deferment effects occurred over time.
Study of air pollution effects at the plant community level is impractical
until stabilization is complete, but individual species and population
responses can be meaningful. Species diversity, richness, evenness, and
species composition all show sensitivity to site differences. Total canopy
coverage is largely determined by graminoid cover and varies substantially
among sites. Forbs do not vary markedly. Mosses were uncommon and shrub
cover was clearly a function of pre-treatinent distribution. Lichens con-
stitute a small percentage of the biomass. We have not shown lichen cover
to be a function of air pollution, perhaps because of confounding influences
of grazing deferment and weather. Western wheatgrass density varies little
68
-------
2.8-
VO
f
\
T
Hay
Coule
Kluver Kluvei Kluver Knoll
West North East A
Knoll
B
Knoll
C
.80-
.78-
.76-
.72-
.70-
.68-
.66-
.64-
.62-
.60-
Evenness
Richness
f
Hay
Coule<
Kluver
West
Kluver Kluver Knoll
North East A
Knoll
Knoll
C
-17
-16
-15
-14
(?
-13 S
u
-123
-llg
Figure 4.8. Diversity (H1) for seven monitoring
sites in the vicinity of Colstrip,
1978 (mean value + 1 standard error),
Figure 4.9. Evenness and species richness on
the Colstrip monitoring plots, 1978
(Mean value + 1 standard error).
-------
within sites, but showed significant inter-site variation. Since it is an
ubiquitous and important forage species in the Northern Great Plains, it may
be a key species for monitoring pollution effects.
REFERENCES
Daubenmire, R. F. 1959. A Canopy-Coverage Method of Vegetational Analysis.
Northw. Sci., 33(1):43-64.
Dodd, J. L., W. K. Lauenroth, G. L. Thor, and M. B. Coughenour. 1979.
Effects of Chronic Low Level S02 Exposure on Producers and Litter
Dynamics. In: The Bioenvironmental Impact of a Coal-Fired Power
Plant, Fourth Interim Report, Colstrip, Montana. E. M. Preston and
T. L. Gullett, eds. EPA 600/3-79-044. U. S. Environmental Protec-
tion Agency, Corvallis, Oregon, pp. 384-493.
Eversman, S. 1978. Effects of Low-Level S02 Stress on Two Lichen Species.
In: The Bioenvironmental Impact of a Coal-Fired Power Plant, Third
Interim Report, Colstrip, Montana. E. M. Preston and R. A. Lewis, eds.
EPA-600/3-78-021, U. S. Environmental Protection Agency, Corvallis,
Oregon, pp. 385-398.
Eversman, S. 1979. Effects of Low-Level S02 on Two Native Lichen Species.
In: The Bioenvironmental Impact of a Coal-Fired Power Plant, Fourth
Interim Report, Colstrip, Montana. E. M. Preston and T. L. Gullett,
eds., EPA-600/3-79-044, U. S. Environmental Protection Agency,
Corvallis, Oregon, pp. 642-672.
Heitschmidt, R. K., 1977. Chronic Effects of S02 on Western Wheatgrass
in a Montana Grassland. Ph.D. Thesis, Colorado State University,
Fort Collins, Colorado. 100 pp.
LeBlanc, F. and D. N. Rao. 1975. Effects of Air Pollutants on Lichens and
Bryophytes. In:Mudd, J. B. and T. T. Kozlowski (Eds). Response of
Plants to Air Pollution. Academic Press, New York. pp. 237-272.
Pielou, E. C. 1969. An Introduction to Mathematical Ecology. J. Wiley &
Sons, N.Y. 286 pp.
Rice, P. M., L. H. Pye, R. Boldi, J. O'Loughlin, P. C. Tourangeau, and
C. C. Gordon. 1979. The Effects of "Low Level S02" Exposures on
Sulphur Accumulation and Various Plant Life Responses of Some Major
Grassland Species on the ZAPS Sites. In: The Bioenvironmental
Impact of a Coal-Fired Power Plant, Fourth Interim Report, Colstrip,
Montana. E. M. Preston and T. L. Gullett, eds. EPA 600/3-79-044.
U. S. Environmental Protection Agency, Corvallis, Oregon, pp. 494-591,
Shannon C. and W. Weaver. 1949. Mathematical Theory of Communication.
Univ. Illinois Press, Urbana. 117 pp.
70
-------
Taylor, J. E. and W. C. Leininger. 1977. Baseline Vegetational Studies near
Colstrip, Montana. Montana State University Mimeo. 67 pp.
Taylor, J. E. and W. C. Leininger. 1979. Plant Community Changes Due to
Low Level SC-2 Exposures. In: The Bioenvironmental Impact of a Coal-
Fired Power Plant, Fourth Interim Report, Colstrip, Mbntana. E. M.
Preston and T. L. Gullett, eds. EPA 600/3-79-044. U. S. Environmental
Protection Agency, Corvallis, Oregon, pp. 610-641.
Taylor, J. E., W. C. Leininger, and R. J. Fuchs. 1975. Baseline
Vegetational Studies near Colstrip. In: Proc. Ft. Union Coal Field
Symp., Mont. Acad. Sci., Billings, Montana, pp. 537-551.
Taylor, J. E., W. C. Leininger, and R. J. Fuchs. 1976. Monitoring Plant
Community Changes due to Emission from Fossil Fuel Power Plants in
Eastern Montana. In: The Bioenvironmental Impact of a Coal-Fired
Power Plant, Second Interim Report, Colstrip, Montana, June, 1975.
R. A. Lewis, N. R. Glass, and A. S. Lefohn, eds. EPA/600-3-76-013,
U. S. Environmental Protection Agency, Corvallis, Oregon, pp. 14-40.
71
-------
SECTION 5
ACCUMULATION AND TRANSFER OF FLUORIDE
AND OTHER TRACE ELEMENTS IN HONEY BEES
NEAR THE COLSTRIP POWER PLANTS
J. J. Bromenshenk
ABSTRACT
Coal-fired power plants at Colstrip, Montana went
on line in 1975 and 1976. Fluoride levels in and on
the body tissues of honey bees collected in 1978 were
significantly higher at several locations than in 1974
and 1975. Mean fluoride increased at apiaries within
20 km downwind of Colstrip, especially in 1976 when
mean levels at some beeyards were more than double
baseline levels. Mean fluoride levels generally were
lower in 1977 than 1976, although not as low as' base-
line levels. Levels in 1978 were higher than those of
1977 but not as high as 1976. There were significantly
higher distributions of fluoride concentrations for
each post-operational year compared to the pre-operational
fluoride distributions in bee populations. Fluoride
levels in bees from pristine areas and from fluoride
exposed areas (via water and air) were highest in adult
foragers, lower in hive bees, and lowest in the pupae.
Preliminary data indicates that fluoride levels in or
on soft body tissues were not significantly different
from those of the hard exoskeletons of bees. Arsenic
content of the bodies of worker honey bees in most
cases was less than 0.8 ppm on a dry weight basis for
all years (1974 through 1977). At an apiary 13 km
northeast of Colstrip, 1.3 ppm arsenic was observed in
1977. This is twice the mean for 1977 and similar to
levels observed in bees exposed to arsenic from a
smelter near Butte, Montana. Accumulation as well
as marked seasonal and site differences of arsenic,
fluoride and lead in honey bees was clearly demonstrated
by data presented in a 1976 legal suit brought by a
Montana beekeeper against a copper smelter. These
findings are.summarized and discussed in terms of
relevance to the Colstrip biomonitoring.
72
-------
INTRODUCTION
In 1974, I initiated a program utilizing honey bees from commercial
apiaries as biomonitors of the bioenvironmental impacts of the coal-fired power
plants located at Colstrip, Montana. I hypothesized that honey bees could
serve as indicators of pollution impacts: (1) they provide an early warning
and a means of continuously monitoring for the presence and amounts of atmos-
pheric pollutants in the environment, and (2) honey bees could be used as
environmental bioassays to measure the effects of pollutant mixtures entering
ecosystems or to evaluate the suitability of the ambient environment to
support bees and other pollinators. Although attempts had been made by
investigators both in Europe and the United States to use honey bees in.this
manner, the Colstrip project was unique. It provided an opportunity to:
(1) conduct baseline studies in a relatively pristine area before power plants
or other industrial sources of anthropogenic pollutants began operation;
(2) monitor post-operational changes in the ecosystems of the Northern Great
Plains; and (3) follow (over a six-year period) the effects of long-term
exposure to low levels of air pollutants.
Honey bees were chosen for use as indicator species and as monitors for
several reasons. They are efficient foragers and gatherers of materials from
the environment. They may forage as far as 8 km or more from their hives,
canvas several thousand acres and visit innumerable flowers. Bees collect
pollen, nectar, honeydew, resin and water. Behaviorally and structurally,
honey bees are adapted for the efficient collection of these substances. They
seem unable to screen out potentially hazardous substances which settle on
plant surfaces or on the bees themselves or which become associated or
entrapped in honey, pollen or water. Bees are likely to inhale fine dusts,
acid mists or gases since the respiratory passages of bees open directly to the
atmosphere and lack protective linings such as mucosa. Hairs surrounding the
spiracles filter out larger particles.
Honey bees are known to be magnifiers of many contaminants. In general,
levels of air pollutants are highest in the tissues of the oldest adult bees,
lower in young bees, and much lower in pollen, nectar, floral parts or water.
Since honey bees are social insects, they can be managed. They are a useful
indicator organism because they can be transported to monitoring sites, they
collect copious quantities of pollen and honey that can be sampled, and there
are thousands of bees in a colony. Analysis of vegetation and water samples
collected in the vicinity of apiaries and physical and chemical air quality
data along with the bee studies will give a comprehensive profile of environ-
mental quality. This comprehensive data reveals the transportation routes of
contaminants through honey bee systems. Honey bees should serve as indirect
tracers of the passage of pollutants through pollination systems.
Honey bees act as accumulators or concentrators of noxious substances and
as such can be used for exposure monitoring. They tend to be more sensitive
than many other insects and higher animals such as livestock to harm from these
materials. This sensitivity can be used as a leading indicator of harmful
effects. These effects are manifested in several ways. For example, air
pollutants, pesticides and anesthetics have induced shortened life spans,
altered brood laying and rearing, reduced hatchability and fecundity, genetic
73
-------
alterations, toxicant avoidance, disorientation, memory loss, and both temporary
and permanent behavior modification. Ultimately, lack of colony vigor and
other effects such as reduced pollination, depressed wax and lessened honey
production may occur. Thus, not only bees but in turn beekeepers and growers
of crops reliant on insect pollination may be affected. More difficult to
detect but equally important are impacts to pollination of indigenous vegetation.
Many ground cover and soil holding plant species rely on insect pollination.
Thus, honey bees offer considerable potential as biological indicators and
monitors for environmental assessments and are ecologically and economically
important as organisms which need to be protected from significant harm.
There are more than 6,000 colonies located in the Colstrip area. Per
colony honey production by these Colstrip bees ranks among the highest in
the nation. Colstrip bees also pollinate alfalfa seed crops grown in south-
eastern Montana. Approximately one-third of these colonies are transported
yearly to California for rental as pollinators of almond groves, orchards
and truck gardens. Detailed discussions of the effects of air pollutants on
honey bees, pertinent agricultural statistics about honey production and the
beekeeping industry of Montana, and detailed descriptions of the materials and
methods incorporated in this project are presented by Bromenshenk (1976, 1978,
1979).
MATERIALS AND METHODS
In 1978, honey bees, pollen, air and water were sampled from 15 apiaries
during late June and again in mid-September. By June, contaminants which may
have been carried back with the "migrant" colonies that had been taken to
California to be rented for pollination should have been cleared out of the
colonies. Tests conducted in 1975 indicated that very little fluoride or
arsenic was found in any colonies returned from California. Food supplies
brought back from California with the bees were consumed and replenished within
a few weeks. By late June, all colonies had had several population turnovers
since the return from California in April and early May. Colonies distant
from Colstrip are not moved out of the region. As such, these can be used
as checks as can be the one to two colonies which were over-wintered at each
sampled apiary, September is the latest that most of the apiaries could be
sampled because most of the hives were taken back to California in early
October.
Each apiary contained at least 15 colonies, some as many as 50, while
most had 20 to 30 colonies. At each beeyard, approximately 300 bees (30 gm wet
weight) were obtained at the entrance of each sampled hive by using a high
velocity, battery-powered, acrylic vacuum apparatus. This drew bees directly
into plastic sample jars. All samples were fast frozen in sealed sample jars.
The jars containing bees were immediately placed under blocks of dry ice
held in heavy-duty cooler chests. Each chest was lined with 5 cm of styrofoam
and encased in 1.2 cm thick plywood sheathing coated with fiberglass, A
minimum of 20 to 25 kg of dry ice was maintained in the chests. After being
frozen, samples were transferred to Whirl Pacs ® and stored at -20°C until
analyzed. Pollen was obtained by removing pellets from honeycombs using a
plastic pick and was stored in plastic vials at room temperature. Water was
obtained from the closest supply, frozen immediately in the plastic sample
74
-------
bottles, and stored in a freezer. Sulfation and calcium formate plates
were mounted on posts in each apiary in June and collected in September
in order to measure ambient air concentrations of reactive sulfurs and
fluorides.
Bee and pollen samples were collected from each of ten hives at each
location. Also, a pooled sample of approximately 1,000 bees (100 gm wet
weight) was obtained by sampling bees at the entrance of every hive in
each beeyard. These samples provided a quick screening method of analyses,
produced an average sample from each location, and ensured sufficient quantities
of materials for tests which required large numbers of bees such as quality
assurance tests and determinations for materials such as pesticides.
Pesticide determinations were not carried out in 1977 or 1978 because
no appreciable levels of insecticides in Honey bees were detected in 1974
and 1975. In the event that an unusual mortality was observed or a report
of pesticide usage in the test areas was received, pesticide analyses could
be performed. In general, these analyses are time consuming and costly.
As such it was decided to conduct them based on a need for the information.
Sulfur levels in honey bees were not examined in 1978. Determinations
of the total bee tissue sulfur content were made in 1974 and 1975. In no
cases were significant differences in sulfurs in bee tissues discerned,
even in bees placed at the EPA sulfur dioxide zonal air pollution system.
Relatively high and variable background sulfur levels (3,500 to 4,800 ppm)
in bee tissues probably masked any changes attributable to small incremental
additions of anthropogenic sulfurs. Hillmann (1972) came to this conclusion
when he was unable to show increases of sulfurs in bees fumigated by several
thousand parts per million of sulfur dioxide injected into hives.
Based on the 1974 through 1977 results of investigations near Colstrip
for 1978, I decided to delete sulfur determinations, to continue to monitor
fluorides in honey bees and to include determinations of arsenic. Both
fluoride and arsenic are toxic to honey bees and accumulate in colonies. How-
ever, sample preparation and analyses procedures differed for these substances.
Fluoride Analyses
•*
Before being analyzed for fluoride, whole bees and pollen were oven-dried
at 50°C for seven days and ground in a Wiley-Mill ® to pass a 40-mesh screen..
For each sample, 0.5 gm of ground and dried material was placed in a metal
crucible and slurried with distilled water with 0.05 gm of reagent grade
calcium oxide. Next, each sample was charred under infrared lamps before
being ashed in a muffle furnace at 600°C for at least six hours (usually
overnight). The ashed samples were digested in 2 ml of perchorlic acid and
subsequently diluted to 100 ml total volume with Orion Tisab total ionic
strength activity buffer. Fluoride determinations were made using an Orion
specific ion probe inserted into the 150 ml beakers containing the dissolved
samples which were agitated constantly by a teflon-coated magnetic stirring
bar during the fluoride determinations.
75
-------
Arsenic
All arsenic determinations were made by A, R. Neuman, Research Chemist,
Animal and Range Science, Montana State University. Samples were oven-dried
in glass beakers covered with watch-glasses to minimize exposure to dust carried
on the air currents in the forced air ovens. One gram samples were weighed into
125 ml Erlynmeyer flasks and digested in 30 ml of a 3:2 mixture of nitric
perchloric acid. Samples were left overnight and then heated slowly to
solubilize. The heat was gradually increased to reduce sample volume to
one-half. The samples were cooled and 10 ml of a 1:1 mixture of nitric:sulfuric
acid added. Heat was increased until perchlorate fumes evolved out of the
flasks. Heating continued until dense white sulfate fumes occurred, Volume
was reduced by heating to 5 ml, and the samples allowed to cool. The cooled
samples were transferred to 50 ml volumetric flasks containing 15 ml of
concentrated hydrochloric acid, 10 ml of water and 1 ml of 1% (w/v) potassium
iodide. Cooling to ambient temperature and allowing the samples to stand for
one hour allowed reduction of As+ to As
Standards were prepared in 100 ml volumetric flasks which contained ^0
ml hydrochloric, 2 ml of potassium iodide and 5, 15 and 25 ng of arsenic. These
were allowed to stand for one hour prior to analyses.
All determinations were made using an atomic absorption spectrophotometer
connected to an arsine generator. Instrument parameters were:
Wavelength: 193.7 Mode: Absorbance/Peak
Slit Width: 1.0 nm p.m.: 495V with BC6
Hollow Cathode: 7.0 mA Atomized: N2/H2 7.5 @ 60 psi/3.0
(not to be exceeded) @ 15 psi
The analytical procedure consisted of placing a 20.0 ml aliquot of
standard/sample into the reaction flask. The sample was purged by bubbling 50%
N2 through the solution. Five ml of 5% (w/v) sodium borohydrate was added thru
the septum while stirring vigorously. Neuman presented the following example:
Standards Total Arsenic Absorbance Absorbance ng As/g
•>
5 ng As/ml 100 ng 0.060 0.058 0.25
(5 ng X 20 ml) 0.055
0.058
15 ng As/ml 300 ng 0.135 0,164 0.75
0.178
0.180
25 ng As/ml 500 ng 0.238 0.253 1.25
0.250
0.270
r2 value = 0.997
76
-------
As an additional check on the arsenic determinations and in order to
gain an insight into other trace elements which may be found in bee tissues,
samples from an apiary located near the EPA ambient air monitoring trailer at
Hay Coulee were sent to Ames Laboratory, Iowa State University, for neutron
activation determinations. At the time of this report, the samples had been
analyzed, but the final concentrations had not been computed.
Basic statistics have been completed for fluorides and arsenic, although
additional statistical treatments by a variety of statistical techniques are
ongoing. Previous statistical tests (1974 through 1977) have indicated that
post-operational fluoride levels in honey bees may not be normally distributed.
Therefore, in addition to parametric tests such as analyses of variance which
assume normality of distributions (Sokal and Rohlf, 1969), Kruskal-Wallis and
Wilcoxon two-sample tests were used. In this case, Kruskal-Wallis tests the
null hypothesis that all the populations have the same distribution of fluoride
concentrations. The alternative hypothesis is that some populations provide
higher distribution values than others. A significant value for the Kruskal-
Wallis statistic causes a rejection of the former hypothesis.
Fluoride concentrations in adult worker honey bees from individual hives
at given locations were considered to be the basic group for comparisons of
pre- and post-operational data. Using a Dec-20 computer, fluoride values
for honey bees from all colonies were ranked. First, all variates from all
groups were pooled. Then the variates were sorted from lowest to highest
(X0, Xj, X2 ...Xn). Next, the sorted observations were ranked from low to
high (Rj, R2, R3 ...Rjj), sorted back into the original groups, and average
ranks computed as:
ni
(ZR),
n
These calculations were conducted for all sites and all years (1975 through
1977) and for the principal sites by year (1975, 1977, 1978).
The Wilcoxon two-sample test will be used to determine whether there was
a significant difference in distributions of levels of fluoride in or on bees
at a given location (apiary). For example, if X designates 1975 values, Y,
1976 values, Z, 1978 values, and E expected, then the null hypotheses may be
stated as follows:
A
(one-tailed test)
B
(one-tailed test)
H0 : E(Y)
. H! : E(Y)
H0 : E(Z)
Hj : E(Z)
£ E(X)
> E(X)
1 E(Y)
< E(Y)
77
-------
RESULTS
All analyses of bees, pollen and water from 1978 were complete at the
time of this report. In addition, a back log of early season analyses for
1977 were finished.
Bees do not forage at random. They display preferences for particular
flowers and foraging areas (Doull, 1971). If an area is subjected to airborne
pollutants, one would expect that the ambient concentrations of contaminants
would be affected by factors such as topography and meteorology. As such,
some areas would be exposed to higher concentrations of pollutants. One would
expect that bees foraging in regions of greater pollution exposure would
contact and bring back more anthropogenic materials than bees foraging in
parts receiving less exposure. Bees from a colony may forage up to 8 km,
although they usually stay within 2 to 3 km (McGregor, 1976). Because
of this, bees from colonies in the same beeyard may gather materials from
different places and bring back different quantities of pollutants. Thus,
bees from a colony foraging an area of relatively high pollution impact could
be expected to demonstrate higher concentrations of contaminants than bees from
a colony that foraged in a part of the region exposed to lower impacts. In a
pristine area, regardless of where the bees foraged, the relative differences
of exposure should be slight, unless there is some localized source of pollu-
tants. Therefore, the levels of noxious substances under these conditions
should be essentially the same from colony to colony. Because of preferential
foraging and non-uniform exposures under pollution stress, variance of levels
of materials such as fluoride in bees should increase as well as mean concen-
tration. For this reason, whereas in 1974 through 1976 four individual
colonies were sampled, beginning in 1977 the number of individual colonies was
increased to ten. The application of Stein's two-stage sample test to
determine the number of observations necessary for a mean with a confidence
interval of a prescribed base is discussed by Bromenshenk (1979).
My previous reports (1979) did not include the fluoride analysis results
for June, 1977. These results are presented along with a map of the apiary
sites in Figure 5.1. Site NE 10 has a water supply containing as much as
20 times the fluoride content of most ground water supplies visited by honey
bees. Bees from this apiary have consistently demonstrated high concentrations
of fluoride. Baseline levels of fluoride of bees in 1974 and 1975 averaged
seven to eight ppm (dry weight). Sites SE 12 and SE 6 are the distant check
sites and displayed the lowest fluoride concentrations. Site SE 12 is located
45 km south of SE 6. Sites N 1 and S 1 are closest to Colstrip. Because of
an unusually dry summer in 1977, nectar and pollen sources diminished rapidly
at many southeastern Montana locations. Colonies from N 1 and S 1 were moved
to California in August in order for the bees to replenish marginal food
supplies. Thus, there were no beehives at these two locations in September.
However, the June 1977 fluoride levels at these sites were the highest
observed in southeastern Montana. These levels exceeded baseline by a
factor of fourteen. Figure 5.2 typifies baseline levels of fluoride in honey
bees and illustrates the biomagnification that occurs.
78
-------
PPM FLUORIDE IN ADULT HONEYBEES
JUNE 1977 COLLECTION
100-
80
o:
o
40-
O.
Q.
.•NEIO'
0 MILES
• COLSTRIP
si« ">.
\SWL-tftfs5
SW2** S4
/»SW3 /
'SLAME DEER
Figure 5.1.
Locations of apiaries with respect to Colstrip
and June 1977 levels of fluoride.
,10
FLUORIDE
BIOMAGNIFICATION
Figure 5.2.
Biomagnif ication of fluoride by honey bees at
baseline concentrations (1975 data collected
in the Colstrip area) .
79
-------
I have carried out investigations since 1974 at several Montana locations
subjected to fluoride from sources such as oil refineries, aluminum reduction
facilities and chemical plants. These studies demonstrate that fluoride
concentrations in honey bees may be as much as 20 times greater than the
baselines established in southeastern Montana. In addition, whenever fluoride
levels in or on honey bees exceeded 120 to 130 ppm, beekeepers reported poor
production and weak colonies. This is consistent with literature reports
concerning toxic doses of fluoride and case history levels of flouride reported
for dead bees in industrial zones where severe and sudden bee mortality
had occurred. Figure 5.3 summarizes these findings.
220n
180-
LJ
||40H
100-
S: eon
20-
HONEYBEES
INDICATORS
OF
FLUORIDE
POLLUTION
INCREASING FLUORIDE POLLUTION
Figure 5.3.
Levels of fluorides in bees from pristine and
industrial areas. LDso based on literature
reports.
The results of the fluoride determinations for adult honey bees collected
in southeastern Montana in June and September, 1978, are presented in Table
5.1. Fluoride levels in pollen are given in Table 5.2 and fluoride levels
in water presented in Table 5.3. Included in these tables are basic statis-
tics—mean, standard deviation and standard error.
Bees collected in September were exposed throughout the growing season
to emissions from the Colstrip power plants, Since many of the colonies near
Colstrip are transported to California in autumn and returned in the spring,
bees collected in June for the most part were exposed for a period of 60 days
or less, compared to an exposure period of approximately 135 days for bees
sampled in September. Therefore, the September sampling period should reflect
any long-term accumulation trends. Fluoride concentrations in bees in August
and September for all years are presented in Figure 5.4.
80
-------
TABLE 5.1. FLUORIDE CONTENT OF ADULT WORKER HONEY BEES, 1978
00
Site
NE 2
NE 3
NE 4
NE 10
SE 1
SE 2
SE-6
SE-12
S-l
S-4
S-5
SW-1
SW-2
SW-3
GB-3
Date
June
Sept
June
Sept
June
Sept
June
Sept
June
Sept
June
Sept
June
Sept
June
Sept
June
Sept
June
Sept
June
Sept
June
Sept
June
Sept
June
Sept
Sept
1
6.1
10.1
7.2
6.4
6.9
11.1
80.0
84.9
3.3
10.4
5.7
6.7
4.9
7.6
5.0
7.5
3.7
8.9
2.8
8.4
4.0
6.9
5.9
12.4
4.7
13.6
4.8
6.3
11.4
2
3.9
10.6
7.6
6.3
6.0
12.4
65.3
82.9
5.2
12.1
5.6
5.5
2.7
7.1
5.3
6.3
6.3
5.0
4.3
6.8
4.3
11.2
5.2
11.7
4.8
11.3
4.0
8.4
6.9
3
5.4
11.3
7.0
6.8
4.6
8.1
50.2
56.3
4.6
7.9
3.5
5.9
2.6
8.6
4.8
7.6
3.4
6.3
5.6
9.2
7.2
10.6
5.4
12.7
3.9
11.0
5.1
7.9
10.5
4
6.3
10.7
19.5
10.6
11.4
14.6
40.6
53.5
6.9
11.7
6.4
10.8
2.6
5.8
5.0
7.1
5.2
4.4
5.9
15.3
5.1
•11.7
5.6
12.0
6.4
7.3
3.1
8.6
8.5
5
5.1
6.8
11.3
11.0
3.6
11.6
37.8
51.8
5.0
6.5
4.1
5.1
5.5
9.0
5.7
7.8
6.8
5.1
6.6
12.0
5.4
10.9
7.3
7.0
3.9
10.2
3.1
6.4
7.2
6
7.3
13.3
5.2
8.4
3.4
6.8
48.4
44.7
5.8
11.4
5.0
8.1
5.4
7.9
2.7
6.1
3.8
4.3
3.4
6.8
5.0
12.0
6.9
12.7
5.7
11.9
5.1
8.1
7.5
7
4.2
7.1
18.3
11.3
4.5
7.4
105.4
72.1
5.5
12.2
7.2
11.3
4.9
8.8
4.8
5.7
3.9
4.0
7.2
13.5
4.4
11.3
7.2
8.0
5.4
12.5
5.0
10.3
7.7
8
6.2
10.0
11.5
15.5
4.8
7.8
105.8
30.9
4.4
11.4
8.3
8.8
5.9
6.1
5.6
5.9
7.0
6.1
7.0
13.4
3.3
6.4
6.5
16.3
3.5
6.6
5.9
10.7
7.6
9
3.6
5.7
21.6
8.2
4.3
11.8
59.1
86.4
5.3
9.4
5.8
12.3
3.2
9.4
4.0
7.9
4.1
8.3
6.3
11.9
3.5
10.4
8.2
8.0
4.6
11.4
5.1
6.7
11.2
10
4.5
10.3
12.1
14.0
5.5
12.8
122.9
90.5
4.4
15.0
12.0
10.2
5.3
4.3
5.5
6.1
4.7
.5.9
6.9
11.2
1.9
6.5
4.5
15.5
4.8
7.6
3.0
6.9
7.4
Combined
Samole *
5.9
12.6
10.7
11.0
4.5
11.0
73.0
57.9
4.5
13.6
6.5
6.6
4.1
6.5
4.2
8.1
4.8
4.9
4.1
12.5
4.5
8.0
6.2
11.3
6.1
11.7
4.8
7.9
7.5
X
5.3
9.6
12.1
9.8
5.5
10.4
71.6
65.4
5.0
10.8
6.4
8.5
4.3
7.5
4.8
6.8
4.9
5.8
5.6
10.9
4.4
9.8
6.3
11.6
4.8
10.3
4.4
8.0
8.6
s.n.
1.2
2.3
5.8
3.2
2.3
2.7
30.4
20.6
1.0
2.4
2.4
2.6
1.4
1.6
0.9
0.9
1.4
1.7
1.6
2.9
1.4
2.2
1.1
3.1
0.9
2.4
1.0
1.5
1.8
S.E.
0.38
0.74
1.83
1.01
0.74
0.85
9.62
6.53
0.31
0.76
0.77
0.83
0.43
0.52
0.29
0.27
0.43
0.53
0.50
0.93
0.45
0.71
0.36
0.99
0.28
0.75
0.33
0.49
0.55
95% Confidence
Interval**
4.4 -
7.9 -
8.0 -
7.6 -
3.8 -
8.5 -
49.8 -
50.6 -
4.3 -
9.1 -
4.6 -
6.6 -
3.3 -
6.3 -
4.2 -
6.2 -
3.9 -
4.6 -
4.5 -
8.8 -
3.4 -
8.2 -
5.5 -
9.4 -
4.1 -
8.6 -
3.7 -
6.9 -
7.4 -
7.5
11.3
16.2
12.1
7.2
12.4
93.3
80.2
5.7
12.5
8.1
10.4
5.3
8.6
5.5
7.4
5.9
7.0
6.7
13.0
5.4
11.4
7.1
13.9
5.4
12.0
5.2
9.1
9.8
Each value represents a sample combined in the field from 15-40 colonies.
u = x 1 t-05 Sx
-------
TABLE 5.2, FLUORIDE IN POLLEN, 1978
Site
NE 2
NE 3
NE 4
NE 10
SE 1
SE 2
SE 16
SE 12
S 1
S 4
S 5
SW 1
SW 2
SW 3
GB 3
X
2.29
2.43
2.01
2.32
2.96
2.65
2.26
2.29
2.42
1.62
2.15
2.55
2.11
1.80
4.20
S.D.
1.02
1.40
0.79
1.27
1.07
1.04
1.05
0.80
1.55
0.77
1.48
0.90
1.21
0.66
2.74
S.E.
0.323
0.468
0.264
0.424
0.377
0.330
0.348
0.253
0.498
0.257
0.467
0.286
0.382
0.219
0.866
N
10
9
9
9
9
10
10
10
10
9
10
10
10
9
10
TABLE 5.3. FLUORIDE IN APIARY WATER SUPPLIES, 1978
Sit-.e
1977
1978
NE 2
0.5
0.5
All Sites
(1977)
X =
S.D.=
S.E."
1.35
3.27
0.85
N =14
NE 3 NE 4 NE 10
0.5 0.
0.5 0.
5 12.7
3 6.8
All Sices Except
NE 10 (1977)
X = 0
S.E.= 0
S.E.= 0
N =13
.48
.11
.03
SE 1
0.5
0.6
Rosebud
Sites (i
X - 0
S.D.= 0
S.E.= 0
N =10
SE 2 SE 12 SI S
0.6 0.6
0.6 0.5
Creek
977)
.51
.03
.01
0.2 0.
0.2 0.
All Sites
(1978)
X
S
S
N
- 0.96
.D.= 1.6B
.£.- 0.45
= 14
4 S 5
5 0.5
6 0.6
SW 1 SW 2
0.5 0.5
0.6 0.6
All Sites Except
NE 10 (1978)
X - 0
S.D." 0
S.E.= 0
N =13
.52
.13
.04
SW
0.5
0.6
} GB 3
0.3
0.5
Rosebud Creek
Sites (1978)
lc
S.D.=
S.E.=
K
0.55
0.10
0.03
10
82
-------
oo
20
UJ
o
E 15
o
10-
E
CL
a.
1975
,.}-4--i--*--{4rl-H
NE-4 NE-2 SE-I S-4 SW-I SW-3 SW-12
NE-3 SE-2 S-5 S-l SW-2 SE-6
30-
cc
o
o
15
'o-
1976
NE-4 NE-2 SE-I S-4 SW-I SW-3
NE-3 SE-2 S-5 S-l SW-2 SE-II
zu
111
5 15
(E
0
^ 10-
01
Q. 5.
L-y i i
^
...
f
1 , T f | ,
T"'t"T t f
t '
NE-4 NE-2 S-5 SW-I SW-3 SE-12
NE-3 SE-2 S-4 SW-2 SE-6
UJ
Q 15
QC
O
uf 10
Q.
O.
5-
1978
NE-4 NE-2 SE-I S-4 SW-I SW-3 SE-12
NE-3 SE-2 S-5 S-l SW-2 SE-6
Figure 5.4. Mean fluoride content and 95% confidence intervals of worker honey bees collected in
autumn, 1975, 1976, 1977 and 1978.
-------
Fluoride levels were lowest throughout the region in 1975 and demonstrated
relatively small differences in mean fluoride. Fluoride levels were highest
at check sites 40 and 80 km distant from Colstrip. In 1976, fluoride levels
at apiaries within 20 km of Colstrip were highest, mean fluoride at some sites
downwind from the power plants were double the 1975 values, and site-to-site
variability of means was considerably greater than for 1975. Unlike 1975,
fluoride levels at the check sites were lower than at any of the Colstrip
sites. In 1977, trends of increased mean fluoride at sites near Colstrip and
of considerable site-to-site variability of means continued. However, the
levels were not as high as for 1976 and in several instances approximated the
1975 levels. Fluoride levels at the check sites tended to be lower than at
most of the sites near Colstrip. Fluoride levels in 1978 were generally
increased at most apiaries near Colstrip relative to levels in 1975 and in
1977. None of the 1978 values for specific sites were as high as peak values
observed in 1976 and 1977, although many sites demonstrated lower mean values
in 1977 than in 1978. As in other post-operational years, values at the
check sites were lower than for most of the Colstrip sites. Changes in levels
of fluoride in honey bees at sites near Colstrip were significant for 1976
and 1977 and were not normally distributed as compared to the pre-operational
levels of 1975 (Bromenshenk, 1978, 1979).
Kruskal-Wallis test results for fluorides in honey bees from 1975 through
1978 are presented in Table 5.4. Data for 1974 through 1977 have been
previously published by Bromenshenk, (1979), Based on these results, the null
hypothesis was accepted for 1975. Pre-operational fluoride distributions were
the same at all locations. For 1977 and 1978 (post-operational), honey bees
displayed a higher distribution of fluoride concentrations than during 1975.
TABLE 5.4. KRUSKAL-WALLIS TEST OF FLUORIDE IN ADULT WORKER HONEY BEES
Site
Year
Kruskal-Wallis
Statistic, H
df
Critical X'
All Sites
(n = 48)
Principal
Sites
(n = 14)
1975-77
1975
1977
1978
151.50
21.78
54.74
47.58
47
13
13
12
63.98
22.36
22.36
21.03
p <_ 0.005
n. s ,
p <_ 0.005
p <_ 0,005
Results of the Wilcoxon two-sample tests of year-to-year differences in
the distributions of fluoride levels in bees at each apiary were not available
for this report. They will be included in the Sixth Interim manuscript.
84
-------
Fluorides in Pupae Versus Adults. Soft Tissues Versus Exoskeletcm
In September, 1978, pupae were obtained from three of the sampled hives
at five locations. A section of comb containing 105 to 200 pupae was removed
from each colony at the time that the adults were vacuumed from the entrances.
Bees at site NE 10 were exposed to fluoride in the water supply. Site GB 3 is
located in Billings and has demonstrated at times levels of fluoride in adults
of 30 to 40 ppm, presumably related to fluoride in the ambient air from nearby
. refineries. Sites S 5 and SE 1 are near Colstrip. Site SE 12 is from a
check yard near Fort Howes, Montana. The pupae were extracted from the comb
in the laboratory using stainless steel forceps (Figure 5.5). The results
are given in Table 5.5.
Figure 5.5.
Honey bee pupae extracted from comb.
contain pupae.
Capped cells still
Although there was slightly more fluoride in pupae from colonies using
the fluoride-contaminated water supply than in pupae from the other apiaries,
it is clear that relatively little fluoride reaches the pupae even when adult
bees demonstrate relatively high fluoride concentrations. Levels observed in
pupae were one-third to one-half that in adults from pristine areas. Levels
of fluoride in pupae were similar to levels of fluoride in pollen from the
same colonies.
Preliminary results indicate that in bees exposed to fluoride in water
(NE 10) and in bees where water was a suspected source (N 1), levels of fluoride
in the soft internal tissues versus the hard exoskeleton were essentially the
Three hundred bees were dissected for each test. Bees from N 1 contained
same.
85
-------
40.1 ppm fluoride in the exoskeleton and 40.7 ppm in the internal tissues.
Bees from NE 10 contained 120.3 ppm of fluoride in the exoskeleton, 140.9 in
the internal tissues.
TABLE 5.5. FLUORIDE IN PUPAE VERSUS ADULT WORKER
HONEY BEES
Site
GB-3
S-5
SE-1
SE-12
NE-10
Hive No.
1
4
10
1
4
10
1
4
10
1
4
10
1
4
10
Fluoride -
1.8
1.6
1.0
0.3
0.5
1.0
1.0
2.4
1.4
0.8
2.0
1.7
2.8
1.4
2.4
X-
s -
X -
s -
x" -
s -
X -
s -
X -
s -
Pupae
1.47
0.42
0.60
0.36
1.60
0.72
1.50
0.62
2.20
0.72
Fluoride -
11,4
8.5
7.4
6.9
5.6
4.5
10.4
11.7
15.0
7.5
7.1
6.1
84. 9
53.5
90.5
X -
S -
X -
S -
S -
X -
s -
X-
s -
Adults
9.10
2.07
5.67
1.20
12.37
2.37
6.90
0.72
76.30
19.94
Pooled Sample versus Observations from Individual Hives
Figure 5.6 shows the fluoride content determined by analyses of pooled
samples versus the mean fluoride obtained from ten independent observations.
Table 5.1 presents mean fluoride levels and fluoride content of pooled
samples for all apiaries utilized in 1978.
U
0
£
o
3
-1
U.
a
o.
a
100-
90-
80
70-
60
50
• MEAN OF TEN HIVES
mm
• -
(
• POOLED SAMPLE
I
• •
20-
10-
A.
* *" 1 ' "
IB 1 H
JUNE SEPT. JUNE SEPT. JUNE SEPT.
Figure 5.6. Fluoride content of worker honey bees.
86
-------
Arsenic and Trace Metals
The results of the atomic absorption spectrophotometry analyses for arsenic
in or on adult worker honey bees sampled near Colstrip are listed in Table 5.6.
Mean arsenic in bees ranged from 0.37 to 0.55 ppm. Arsenic concentrations of
pooled samples for 1974, 1975, 1976 and 1977 were respectively 0,41 ± 0.25 (1
standard deviation), 0.36±0.13; 0.49±0.16; and 0.5710.27 ppm.
Minimum and maximum individual values ranged from 0.1 to 1,3 ppm. The 1.3 was
observed at a site northeast of Colstrip in 1977 and was the highest value
recorded at any site for any year in southeastern Montana. The next highest value
value was 0.8. If the 1,3 value is excluded from the 1977 data as an outlier,
the 1977 mean for the yards sampled becomes 0.50 ± 0.13, For the most part,
these samples represent pooled samples from the apiaries .obtained in the autumn
of each year, although the 1975 values are based on collections made throughout
the growing season. For comparison, samples obtained near Butte from areas
subjected to arsenic impacts from a copper smelter demonstrated levels of
arsenic ranging from 1,2 to 3.06 ppm (1975). Similarly, a bee sample taken from
a site near an aluminum reduction facility at Columbia Falls, Montana had
2.03 ppm arsenic (1975).
TABLE 5.6. ARSENIC CONTENT (PPM) OF HONEY BEES FROM
SOUTHEASTERN MONTANA
YEAR N 5? S.D, S.E.
1974 10 0,41 0.25 0.079
1975 10 0,36 0.13 0.041
1976 8 0.49 0.16 0.057
1977 11 0,57 0.27 0.081
1977* 10 0.50 0.13 0.039
1975** 4 0,55 0.17 0.085
* 0.38 0.15 0.075
4. 0,25 0.10 0.050
* 1.3 ppm value excluded as an outlier,
** Each set represents values from four individual hives at one site.
A legal suit brought by a Montana beekeeper against a copper, smelter
alleging arsenic poisoning of over 400 colonies of honey bees included among
the exhigits extensive tables of results of analyses of heavy metals in bees.
(Furlong et al. correspondence, 1973; 1975; 1976). The data covers lead, zinc,
copper, cadmium, arsenic and fluoride in bees from the Deer Lodge Valley dur-
ing 1973 through 1976. For the most part, analyses were performed via atomic
absorption spectrophotometry. Generally, the bees were washed prior to analy-
sis. Based on a test of metal content of bees "as received" versus bee washed
with water, the conclusion was drawn that slightly less arsenic was found in
the washed bees, but essentially the same concentration of other metals occurred
in the unwashed and washed..bees. Figures 5.7 to 5.9 summarize data obtained
from these exhibits. Seasonal trends of accumulation of arsenic, fluoride and
other metals at Deer Lodge are pertinent to my Colstrip observations.
87
-------
oo
CO
ZO-i
15-
10
UJ
OT
K
O.
Q.
5-
JULY
AUGUST
1973
A-2
SEPTEMBER
12
o 10
z
UJ
in
(C
o.
Q.
OT
IE
8: 2
FATAL
MAY/JUNE
JULY AUGUST SEPTEMBER
1974
Figure 5.7. Arsenic in adult honey bees, Deer Lodge study, 1973-74. Stars indicate loss of
colony being sampled.
-------
Zn PbCu F Cd As
00
VO
o
£
O
IBO
160
140-
120
IOO
80
6O
40
2O
MAY/JUNE JULY AUGUST SEPTEMBER
1974
Figure 5.8. Fluoride in adult honey bees, Deer
Lodge study.
ouu
200
100
IU
-
4
ou
40
20
<:UU
I ^rt
1OU
—
100
ert
OU
3
2
-
1
ou
25
20
IS
10
5-
-
I
it
Zn
Pb
Cu
( i
i
I
b
Cu
Zn
i
.Cd
f
4 I
1
F
\f
I i
F
C"..As
Pb
Cu
I
F
(
Zn
i
Cd
*
*A>
'b
i
Cu
F
, ,
.Cd
{
<' At
MAY/JUNE JULY
1974
AUGUST SEPTEMBER
Figure 5.9.
Heavy metal content of adult honey
bees, Deer Lodge study.
-------
The Deer Lodge study for 1973 and 1974 began each year with package bees
obtained from the southern United States, Samples of bees from the packages
were analyzed at the start of the tests, although it is not clear in the reports
which packages went to which beeyards. Mean values of arsenic, fluoride and
cadmium were much lower in the package bees than in any of the later season
samples. Levels of zinc and copper in bees sampled in autumn of 1974 were
similar to the levels in the package bees, while lead values became highly
variable. Arsenic and fluoride levels by July of 1974 had increased dramatically
compared to those of May/June. After July, concentrations leveled off,
continued to increase or declined somewhat depending on the site. Arsenic
concentrations in 1973 displayed similar patterns. In 1975, the fluoride
content of bees distinctly increased at only one location. At the other sites
late season levels were about the same or lower than the early season values.
Fluoride concentrations in bees from this area were substantially higher
than those from southeastern Montana with the exception of the southeastern
apiary exposed to fluoride in its water supply. Arsenic concentrations were
much higher in bees from the Deer Lodge Valley. One of the highest values
observed in the Deer Lodge bees was 36.3 ppm (dry weight) which was 27 times
greater than the highest level (1.3 ppm) found in Colstrip bees and 66 to 98
times the mean arsenic content of Colstrip bees.
DISCUSSION
Honey bees as biological tools for environmental assessments of air
pollutants are among several methods being developed under the eoal-fired
power plant project. The ultimate objective of these studies is to provide
a means of assessing, predicting and mitigating ecological and related socio-
economic impacts of siting coal-burning electric generators in the Northern
Great Plains.
Sulfur dioxide has been the anthropogenic substance of primary interest
to the overall coal-fired power plant project. However, as mentioned earlier,
honey bees do not appear to be useful organisms for use in exposure monitoring
of sulfur dioxide. If the analytical procedures were modified such as utiliz-
ing stable sulfur isotopes rather than total sulfurs, it might be possible to
distinguish anthropogenic sulfurs in .bee tissues from the naturally occurring
sulfurs that'.occur in animal proteins. In other words, my inability to detect
incremental increases of sulfurs in bees exposed to sulfur dioxide and other
sulfur compounds in the ambient air may be a function of analytical technique
rather than of non-accumulation in bees.
It is well known that bees accumulate and concentrate fluoride and arsenic.
These materials are toxic to bees. Analyses of stack fly ash from Colstrip
Unit 2 indicate that small particles emitted by the stack were enriched in
several elements including fluoride and arsenic. Measured stack concentrations
of fluoride were 2,130 ± 400 (SD) ppm and of arsenic were 221 ± 20 (SD) ppm
(Crecelius e~t aZ., 1978). In addition, Munshower (personal communication)
detected increased post-operational levels of arsenic in sheaths of pine needles
from trees near Colstrip, while investigators at this laboratory have shown
post-operational increases in fluoride in pine foliage and in mice sampled near
Colstrip. At the time of this report, wind data for 1978 had not been obtained
90
-------
for the Colstrip area. Data from 1976 and 1977 indicated that in the summer
winds most frequently blew from Colstrip towards the southeast and northeast.
Also, plume strikes have been measured several times a week at an EPA monitoring
station at Hay Coulee (See Section 1). This monitor is approximately 11 km
southeast of Colstrip, less than 0.5 km from the S 5 apiary location. There
were no air monitoring stations near the apiaries northeast of Colstrip. Thus,
plume strike data was not available for that vicinity. However, I frequently
have observed the plumes from the power plants descending near the ground and
traveling relatively intact for at least 10 km, following drainage routes which
lead towards the NE 2, 3 and 4 beeyards. I intend to attempt to correlate
levels of fluorides in bees with concentrations in the ambient air. Sulfation
and calcium formate plates were set out at each of the beeyards during 1978 and
1979. The plates have been sent to the Montana Department of Health for analyses,
but the results were not available for this report.
In June,1977, at N 1, levels of fluoride in bees which exceeded baselines
by fourteen-fold were unexpected and puzzling. Although it is possible that
the fluoride was carried over from California, this is unlikely. The bees
had been at N 1 for several weeks. No other concentrations of fluoride of
this magnitude appeared in any other hives sampled. All of the migratory
colonies are moved on large flatbed trailers. Colonies are not kept separate
on the truck by apiary location. As such, hives marked during the previous
summer usually show up at Colstrip distributed in a more or less random pattern
across many locations. Therefore, even if the colonies at N 1 had been kept
at a site exposed to a source of fluoride before being set in at Colstrip,
one would expect to find some of these colonies at other locations or to find
some low-fluoride content colonies at N 1.
Air diffusion models predicted that N 1 would be subjected to plume strikes
during the winter months , but concentrations of fluoride in bees at N 1 were
as high or higher than fluorides in bees sampled in Montana near aluminum
reduction facilities and chemical plants. These facilities emit much more
fluoride than the Colstrip power plants. Thus, while it is possible, it seems
unlikely that this much fluoride would have reached the bees at N 1 via the
air. An unanticipated potential fluoride source was discovered in 1978. The
ranch owner complained at a legal hearing that hay meadows adjacent to this
apiary were being destroyed by waste waters which were being pumped from the
coal strip mines into Arnell's Creek. When I visited this apiary in 1978,
large areas of the creek bank had been eroded away and cattails were growing
in parts of the meadow which previously had supported hay forage. Arnell's
Creek typically is dry except during the spring runoff. Water in Arnell's
Creek was sampled in 1979 as were bees at N 1. I am trying to obtain 1977
records of chemical analyses of the mine waste water, if any were performed.
Fluorides may concentrate in bees until toxic thresholds are exceeded.
The LD50 threshold depicted in Figure 5.3, which displays levels of fluorides
found in bees from pristine and industrialized zones, is based on literature
reviews (Bromenshenk, 1978) and on personal observations made in Montana. The
LD50 zone is approximate and based on acute toxicities. Apparently no serious
attempt has been made to discern what tissue residues are indicative of chronic
poisoning and sublethal effects such as shortened life spans, immunological
suppression or disruptions of foraging activities.
91
-------
Elevated fluoride content in bees sampled after the Colstrip power plants
began operation as compared to baselines and checks have consistently been
observed at apiaries southeast and northeast of Colstrip. In 1978, bees from
sites near Colstrip continued to demonstrate significantly higher distributions
of fluoride concentrations than baselines. In general, mean fluoride content
was higher in Colstrip bees in 1978 than in 1977 and considerably higher than
baseline levels, although concentrations were not as high at individual apiaries
as in 1976. In 1977, forage was unusually poor because of dry weather, and winds
were more variable than in 1976. Based on data available at the time of this
report, 1978 was a more typical year than 1977 in terms of moisture, forage
and honey production. In all post-operational years, the content of fluoride
of bees at.distant check points remained at or below baseline concentrations.
Fluoride concentrations of surface waters have remained relatively constant for
all years. Also, concentrations of fluoride in pollen are essentially the same
for all years. Measurements of fluoride and sulfur in the ambient air should
provide valuable insights as to the route by which the bees are obtaining these
contaminants. This information should be available for the final report.
In 1979 throughout the region, the honey crop was exceptional due to heavy
flowering of yellow sweet clover. Therefore, foraging activity should have
been great. This should ensure that the bees came into contact with any
anthropogenic materials in their surroundings. Since the honey crop was poor
in 1977, it will be informative to see the opposite situation.
Fluoride in pupae versus adults clearly demonstrate that pupae are not
accumulating fluoride. Either pupae are not being exposed to fluoride or they
somehow eliminate it. The former seems probable. In previous studies, foraging
adults were found to demonstrate almost twice as much fluoride as hive bees
(nurse bees, guards) (Bromenshenk, 1978). The latter probably are only exposed
to materials brought back by other bees or to substances that directly enter
the hive such as gases. By merit of extensive foraging and by being older,
foragers are exposed to more materials and for longer periods than hive bees.
If fluoride enters a colony via the pollen, the pupae would be directly exposed
since pollen is a major component of the diet of the brood. Fluoride detected
in pupae in 1978 tended to be equivalent to levels in pollen, but levels
in all pollen samples were low compared to levels in the adult bees and in
general not significantly different from site to site. Fluoride in pollen
fed to the pupae may be the source of low background levels in pupae. It does
not appear to be the source of elevated fluoride in adults. Although water
definitely appeared to be the source of high concentrations of fluoride in
bees at NE 10 and although bees gather water for diluting honey, for cooling
hives and for drinking, fluoride in the water did not appear to affect the
fluoride content of the brood. Dilution of food and evaporative cooling of
hives by spreading water across the combs seem to be likely routes by which
fluoride in water would be transported to the developing brood, but apparently
this is not the case. Thus, fluoride in water does not appear to constitute
a problem to the brood welfare, although levels in adults approached those
known to be definitely toxic. Over the last five years, bees at NE 10 have
consistently displayed high body content of fluoride and always have displayed
a lack of vigor, slow growth and moderate to low honey production (about 40
pounds per colony). In 1978, the bees were moved to a position on the bank
of a creek that averages 0.4 to 0.6 ppm fluoride compared to the 11 to 20 ppm
92
-------
of the stock tank that the bees had previously visited. In 1978, after being
moved, colonies at NE 10 made a dramatic turn-around, producing well over 100
pounds per colony (G. Simpson, Beekeeper, personal communication, 1979). They
did not display any lack of vigor or other symptoms of chronic fluoride poisoning
Preliminary results of levels of fluoride in or on soft versus hard body
tissues demonstrated that the fluoride was not just clinging to the surface
such as on the numerous branched hairs that cover the body of a bee. This is
consistent with previous observations that washing bees with distilled water
prior to analyses did not produce any significant change in the fluoride values
(Bromenshenk, 1978). There needs to be more study on fluoride effects and
transport through the tissues of bees.
Results of fluoride determinations made for pooled bee samples versus
independent samples at each apiary indicate that pooled samples could be
utilized in a rapid screening program but do not provide the amount of infor-
mation, sensitivity nor interpretability of values obtained from many separate
colonies.
Observation of 1.3 ppm of arsenic in bees at NE 2 in 1977 raises a warning.
Arsenic is released by the Colstrip power plants. It appears to be accumulating
in the sheaths of pine needles, it occurred at an apiary that displayed signif-
icant post-operational increases of fluoride .and it is very toxic to bees.
Additional arsenic determinations will be carried out on 1979, 1978 and 1977
samples in order to determine whether this observation was an anomaly or
whether arsenic is beginning to reach the bees.
Bees in the Deer Lodge Valley are accumulating and concentrating consid-
erably more arsenic than any bees at Colstrip. This clearly indicates that
bees are excellent concentrators of arsenic as well as fluoride. Also, the
Deer Lodge data demonstrates that honey bees accumulate and concentrate other
metals such as cadmium. This arsenic in bees appears to take its toll. Dashed
lines on the graphs of arsenic accumulation in Deer Lodge bees (Figures 5.7
through 5.9) indicate levels reported from case histories near smelters in
the United States and Europe to be "fatal" or satisfactory proof of arsenic
poisoning" (Knowlton et al., 1950; Debackere, 1972). Note that in most cases,
if arsenic in the bees exceeded these thresholds, the bees either were reported
to have been lost or in the next sample period showed a lower arsenic concentra-
tion. Following this decrease, arsenic levels often increased. This suggests
total death of colonies when toxic thresholds were surpassed. In less severe
cases, the oldest foragers may have reached toxic levels before younger bees and
brood. If the older bees died, this would leave younger bees which should
display lower levels of arsenic. As the average age of bees in the colony
again increased, longer exposure periods and the shifts to foraging activities
should result in increased arsenic content. Admittedly, the forgoing basically
is conjecture. However, in 1973 the Montana State Apiarist was called in to
examine Deer Lodge colonies because of a suspected pesticide kill. He concluded
that the colonies did not exhibit symptoms of poisoning by pesticides. They
did appear to be suffering from poisoning by a slower acting toxin such as in
air pollutant (W. Kissinger, personal communication, 1979).
93
-------
An apiary near a copper smelter sustained severe losses of bees near
Whitehall, Montana, in 1975. These colonies demonstrated lower levels of
arsenic in bees and pollen than Deer Lodge bees but considerably more arsenic
than any bees from Colstrip. This illustrates a need for more and better in-
formation about arsenic in honey bee systems both as regards the use of honey
bees as biological monitors and hazards to bees and the beekeeping industry.
• CONCLUSIONS
Honey bees from southeastern Montana continued in 1978 to demonstrate
significant post-operational changes in fluoride content compared to 1975
(pre-operational) levels. Extremely high levels of fluoride seen in adult
bees at an apiary north of Colstrip in June of 1977 may have been the result of
dumping of mine waste waters. Arsenic was found at a considerably higher level
in bees at one apiary (NE 2) in September of 1977 compared to levels for all
other apiaries and all other years. The arsenic appeared at an apiary that
has often demonstrated significant post-operational increases of fluoride in
worker honey bees. Preliminary tests indicate that fluoride content of soft
and hard body tissues of bees is essentially the same—the fluoride is
distributed throughout the body and does not occur just on the body surface.
On the other hand, pupae demonstrate relatively low tissue content of fluoride
even in instances where the adult bees from the same colonies demonstrate
concentrations five to ten times that of baseline values. Also, although
relatively high levels of fluoride in water are reflected by high levels of
fluoride in adult bees (both in soft and hard tissues), the fluoride does not
appear to be affecting or reaching the brood.
Arsenic, as revealed by data obtained from studies in the Deer Lodge Valley ,
may concentrate in adult bees to levels as great as approximately 100 times
that of the levels seen in bees sampled near Colstrip. These levels appear to
be harmful to bees causing problems ranging from poor honey production and
an inability to overwinter to complete death of the colony. Based on these
studies, other materials such as cadmium also are concentrated by bees.
Thus, the results to date reaffirm the belief that honey bees are excellent
detectors of contaminants in their surroundings and point to hazards if these
materials reach toxic thresholds—problems not only to the bees themselves
but also to beekeepers, pollination systems and agriculture.
REFERENCES
Bromenshenk, J. J. 1976. Investigations of the Effects of Coal-Fired Power
Plant Emissions Upon Insects, Report of Progress. In: The Bioenviron-
mental Impact of a Coal-Fired Power Plant, Second Interim Report, Colstrip,
Montana, R. A. Lewis, N. R. Glass, and A. S. Lefohn, eds. EPA-600/3-76-013,
U. S. Environmental Protection Agency, Corvallis, Or. pp. 112-129 and
286-312.
Bromenshenk, J. J. 1978. Investigations of the Impact of Coal-Fired Power
Plant Emissions Upon Insects. I. Entomological Studies in the Vicinity
of Colstrip, Montana. II. Entomological Studies at the Zonal Air Pollution
94
-------
System. In: The Bioenvironmental Impact of a Coal-Fired Power Plant,
Third Interim Report, Colstrip, Montana, E. M. Preston and R. A. Lewis,
eds. EPA-600/3-78-021, U. S. Environmental Protection Agency, Corvallis,
Or. pp. 146-312 and 473-507.
Bromenshenk, J. J. 1979. Honey Bees and Other Insects as Indicators of
Pollution Impact from the Colstrip Power Plants. In: The Bioenviron-
mental Impact of a Coal-Fired Power Plant, Fourth Interim Report,
Colstrip, Montana, E. M. Preston and T. L. Gullett, eds. EPA-600/3-79-044,
U.S. Environmental Protection Agency, Corvallis, Or. pp. 215-239.
Crecelius, E. A., L. A. Rancitelli, and S. Garcia. 1978. Power Plant
Emissions and Air Quality. In: Potential for Gaseous and Heavy Metal
Contamination from Energy Extraction Processes in the Northern Great
Plains and the Consequent Uptake and Turnover in Range Ecosystems, ERDA
Annual Report. Activity RX-02-03, Ames Laboratory, Iowa State University,
Ames, Iowa. pp. 9-33.
Debackere, M. 1972. Industriele Luchtvervuiling en Bijenteelt (Industrial
Pollution and Apiculture). Vlaams Imkersblad., 2(6):145-155.
Doull, K. M. 1971. An Analysis of Bee Behavior as it Relates to Pollination.
Am. BeeJ., Ill (7) :266-273; (8):302-303; (9):340-341.
Furlong T., R. Anderson, W. Unger, F. J. Laird Jr., and M. W. Bowman. 1973;
1975; 1976. Correspondence on file at the University of Montana, Missoula,
Montana.
Hillmann, R. C. 1972. Biological Effects of Air Pollution on Insects,
Emphasizing the Reactions of the Honey Bee (Apis melHfera L.) to Sulfur
Dioxide. Ph.D. Thesis, The Pennsylvania State University, University
Park, Pa. 159 pp.
Knowlton, G. F., A. P. Sturtevant, and C. J. Sorenson. 1950. Adult Honey Bee
Losses in Utah Related to Arsenic Poisoning. Utah Agric. Exptl. Stat.
Bull. 340. 16 pp.
McGregor, S. E. 1976. Insect Pollination of Cultivated Crop Plants. Agr/
Handbook No. 496. USDA Agricultural Research Service, Washington, D.C.
411 pp.
Sokal, R. R., and F. J. Rohlf. 1969. Biometry. The Principles and Practice
of Statistics in Biological Research. W. H. Freeman and Company,
San Francisco, Ca. 776 pp.
95
-------
FIELD AND LABORATORY EXPERIMENTS
SECTION 6
TEMPORAL VARIATION IN S02 CONCENTRATIONS
ON ZAPS DURING THE 1978 FIELD SEASON
E.M. Preston, T.L. Gullett, and D.B. Weber
ABSTRACT
Sulfur dioxide exposure regimes on the ZAPS plots
during 1978 are characterized. Seasonal mean concen-
trations are similar to those of previous years on
CONTROL, LOW, and MEDIUM plots but were somewhat
higher on the HIGH plot. Frequencies of higher con-
centrations were somewhat greater on most plots than
in previous years. The diel pattern of SOa concentra-
tions may resemble that of area sources. It does not
resemble the point source induced diel cycle near
Colstrip. Daytime geometric mean concentrations are
substantially less than nighttime concentrations.
Diel activity cycles of organisms on the plots must be
considered in defining dose-response relationships.
INTRODUCTION
The Zonal Air Pollution System (ZAPS) was designed to allow experimental
evaluation of the effects of long-term chronic sulfur dioxide exposure on
native grasslands of the northern Great Plains. Experiments are being
conducted within two 27-acre grassland exclosures in the Custer National
Forest in southeastern Montana. Livestock are excluded to protect them from
injury and to protect equipment from damage. Sulfur dioxide (SOa) fumigation
within the first exclosure was initiated in May 1975 (ZAPS I). Fumigation
within the second exclosure was started in April 1976 (ZAPS II).
The experiments are designed to test the effects of SOa upon biomass
dynamics (plant, arthropod, and small mammals), plant and animal community
structure, insect and fungal diseases of plants, pollination systems, lichens,
and upon a number of physiological and biochemical functions. Dominant plants
on the study plots are western wheatgrass (Agropyron smifhii"), prairie June-
grass (Koeleria aristata), and Sandberg bluegrass (Poa secunda).
96
-------
The design of the ZAPS has been described in detail elsewhere (Lee et
al., 1976; Lee and Lewis, 1978; Lee et at., 1978; Lee et al., 1979) and will
only be reviewed briefly here. The ZAPS is a network of 2.54 cm diameter
aluminum pipes suspended about 75 cm above 0.5 hectare plots of native grass-
land. Gas release ports (0.8 mm diameter) are positioned at 3 m intervals.
There are more than 250 ports per 0.5 hectare plot with no location more than
5.5 m from a port.
Each ZAPS consists of four plots, each receiving different ("CONTROL,"
LOW, MEDIUM, and HIGH) median concentrations of S02. By convention, plots
designated A,B,C,D refer to "CONTROL", LOW, MEDIUM, and HIGH S02 treatment
levels, respectively. The "CONTROL" plot receives no S02 from its pipe net-
work, but does receive S02 drifting from upwind plots. Consequently, this
plot is, in effect, a very low level S.02 treatment plot whose exposure regime
simulates more closely that from a point source than that from an area source.
Exposures are continuous throughout the growing season (April-October).
S02 is released at constant rates, but ambient concentrations vary with wind
dilution and other micrometerological conditions.
In this section, we characterize temporal variation in SOa concentration
on the ZAPS plots during 1978.
MATERIALS AND METHODS
Sulfur dioxide concentrations were continually measured at a common
.central point on all plots (location c, see Lee et at., 1979, Figure 1)
throughout most of the growing season. Ambient air samples were taken 35 cm
above the.ground (approximate canopy height).
In general, the procedures described in Lee et at. (1979) were followed
during 1978. Several minor changes were implemented in order to minimize
sources of uncertainty in the S02 measurements discussed by Lee et al. (1979).
First, teflon sample lines were made equal in length (500 m). This minimized
differences in response time and resistance to sample flow between lines.
Secondly, all dilution air for calibration gas was drawn from outside the
monitoring shed and filtered through charcoal before it was used as a carrier
for calibration gas. This eliminated C02 interference during calibration (Lee
et al., 1979).
As in 1977 (Lee et al., 1979), calibration curves relating analyzer
response to the logarithms of S02 concentration typically became non-linear at
S02 concentrations below roughly 2 pphm. Linear extrapolation leads to an
overestimate of S02 levels in the 0-2 pphm range. The true value is likely to
lie between .1 pphm and the value estimated from the calibration curve. These
values were used to specify upper and lower bounds for any S02 concentration
falling in the non-linear portion of the calibration curve. During data
analysis, each analyzer response was entered into two calibration equations.
In one, the HIGH run, the calibration equation was adjusted to yield a maximum
value that could result from the above sources of uncertainty. In the other,
97
-------
the LOW run, the calibration equation was adjusted to yield the corresponding
minimum estimate. Summary statistics were computed for both the HIGH run and
the LOW run.
RESULTS
Interseasonal Trends
Seasonal summaries are given in Tables 6.1 and 6.2. The first number for
each entry is the LOW run value, the second is the HIGH run value. If only
one number is presented, there was no difference between the LOW run and HIGH
run values.
Geometric mean concentrations were similar to previous years for Control,
Low and Medium plots. They were somewhat higher for HIGH plots. However, the
frequencies of higher concentrations were greater than in previous years on
most plots (Lee et al., 1979, Lee et at., 1978). This is reflected in higher
arithmetic averages and in larger standard geometric deviations.
Diel Patterns
As in previous years (Lee et at., 1979) SOa concentrations on the plots
were higher and more variable during the nighttime hours than during daylight
hours (Table 6.3). This is caused by generally poor mixing and low wind
speeds at night and relatively better mixing and higher wind speeds during the
daytime. The diel pattern of SOa concentrations is strongly correlated with
the reciprocal of wind speed (Figure 6.1). This relationship suggests that
wind speeds much above 8 mph reduce geometric mean SOa concentrations to
background on ZAPS II, HIGH plot. Frequency distributions of SOa concentra-
tions for day hours and night hours are compared in Figures 6.2 and 6.3. The
combined frequency distributions are also provided since this format has been
used in previous interim reports describing the 1975-1977 data. For ZAPS II,
concentration frequency distributions are compared with that for Hay Coulee
(Ludwick et al., 19.791 in Figure 6.4. Hay Coulee is located 7.5 miles south-
east of the coal-fired generating units at Colstrip.
Intraseasonal Trends
Trends in SQa exposure concentrations over the 1978 season are illus-
trated in Figure 6.5. On most plots, concentrations increased as the season
progressed. This trend was more evident on ZAPS II than on ZAPS I.
DISCUSSION
Biological Significance of the Diel Patterns of SOa Exposure on ZAPS
There is a need to convey a general feeling for the SOa exposure regime
that organisms on the ZAPS plots experience. In previous interim reports sea-
sonal means, seasonal frequency distributions, and aspects of intraseasonal
98
-------
TABLE 6.1 ZAPS I SEASONAL SUMMARY, 1978*
Control Geometric Mean = Standard Geometric Arithmetic Mean = 1.4
1.0-1.1 Deviation = 1.7, 2.1
1-Hour Ave Exceeded 25 pphm 1 Times: 0.0 Percent
3-Hour Ave Exceeded 50 pphm 0 Times: 0.0 Percent
24-Hour Ave Exceeded 10 pphm 0 Times: 0.0 Percent
24-Hour Ave Exceeded 14 pphm 0 Times: 0.0 Percent
1-Hour Peak 26 3-Hour Peak 11 24-Hour Peak 5
Low GM"= 2.8 SGD =2.7 Arithmetic Mean =5.4
1-Hour Ave Exceeded 25 pphm 125-127 Times: 3.7 Percent
3-Hour Ave Exceeded 50 pphm 3 Times: 0.3 Percent(b)
24-Hour Ave Exceeded 10 pphm 19-20 Times: 12.5-13.2 Percent(a)
24-Hour Ave Exceeded 14 pphm 3-5 Times: 2.0-3.3 Percent(b)
1-Hour Peak 110 3-Hour Peak 70 24-Hour Peak 17
Medium GM = 5 .1 SGD =3.1 Arithmetic Mean =11.6
1-Hour Ave Exceeded 25 pphm 426-433 Times: 12.5-12.6 Percent(a)
3-Hour Ave Exceeded 50 pphm 25-26 Times: 2.2 Percent(b)
24-Hour Ave Exceeded 10 pphm 76-77 Times: 50.0-50.6 Percent(a)
24-Hour Ave Exceeded 14 pphm 38 Times: 25.0 Percent(b)
1-Hour Peak 171-172 3-Hour Peak 109 24-Hour Peak 27-31
High GM = 8.6-8.9 SGD = 3.2,3.4 Arithmetic Mean =19.2
1-Hour Ave Exceeded 25 pphm 728-734 Times: 21.3-21.5 Percent(a)
3-Hour Ave Exceeded 50 pphm 106 Times: 9.2 Percent(b)
24-Hour Ave Exceeded 10 pphm 121-124 Times: 79.6-81.6 Percent(a)
24-Hour Ave Exceeded 14 pphm 99-101 Times: 65.1-66.5 Percent(b)
1-Hour Peak 461-465 3-Hour Peak 294-295 24-Hour Peak 72
*pphm S02 monitored near plot center at canopy height
(a) violates Montana standards
(b) violates Federal standards
99
-------
TABLE 6.2. ZAPS II SEASONAL SUMMARY, 1978*
Control Geometric Mean = Standard Geometric Arithemtic Mean =
0.9-1.2 Deviation = 1.3,2.3 1.1-1.2
1-Hour Ave Exceeded 25 pphm 0 Times: 0 Percent
3-Hour Ave Exceeded 50 pphm 0 Times: 0 Percent
24-Hour Ave Exceeded 10 pphm 0 Times: 0 Percent
24-Hour Ave Exceeded 14 pphm 0 Times: 0 Percent
1-Hour Peak 6 3-Hour Peak 4 24-Hour Peak 2.2
Low GM = 2.5 SGD = 2.3 Arithmetic Mean =4.0
1-Hour Ave Exceeded 25 pphm 5 Times: .2 Percent
3-Hour Ave Exceeded 50 pphm 0 Times: 0 Percent
24-Hour Ave Exceeded 10 pphm 2 Times: 1.3 Percent(a)
24-Hour Ave Exceeded 14 pphm 0 Times: 0 Percent
1-Hour Peak 30 3-Hour Peak 22 24-Hour Peak 11
Medium GM = 4.6 SGD =2.6 Arithmetic Mean =8.0
1-Hour Ave Exceeded 25 pphm 240-243 Times: 7.2 Percent(a)
3-Hour Ave Exceeded 50 pphm 0 Times: 0 Percent
24-Hour Ave Exceeded 10 pphm 44-45 Times: 28.0-28.6 Percent(a)
24-Hour Ave Exceeded 14 pphm 10-11 Times: 6.4-7.0 Percent(b)
1-Hour Peak 57 3-Hour Peak 47 24-Hour Peak 20
High GM = 9.2-9.3 SGD = 2.8,2.9 Arithmetic Mean =17.8
1-Hour Ave Exceeded 25 pphm 648 Times: 19.3 Percent(a)
3-Hour Ave Exceeded 50 pphm 65-67 Times: 5.8-5.9 Percent(b)
24-Hour Ave Exceeded 10 pphm 119 Times: 75.8 Percent(a)
24-Hour Ave Exceeded 14 pphm 97-98 Times: 61.8-62.4 Percent(b)
1-Hour Peak 125 3-Hour Peak 89 24-Hour Peak 44
*pphm S02 monitored near plot center at canopy height
(a) violates Montana standards
(b) violates Federal standards
100
-------
TABLE 6.3. COMPARISON OF WHOLE SEASON SUMMARY STATISTICS FOR S02 CONCENTRA-
TIONS DURING DAYLIGHT HOURS AND NIGHTTIME HOURS ON ZAPS*
Plot
Control
Geometric Mean
Std. Geo. Dev.
Arithmetic Mean
1 hour peak
Low
Geometric Mean
Std. Geo. Dev.
Arithmetic mean
1 hour peak
Medium
Geometric Mean
Std. Geo. Dev.
Arithmetic Mean
1 hour peak
High
Geometric Mean
Std. Geo. Dev.
Arithmetic Mean
1 hour peak
Day
1.0
1.4
1.1
26.0
1.9
2.0
3.0
98.5
3.1
2.3
5.9
145.6
6.3
2.6
11.4
418.0
ZAPS I
Night
1.4
2.0
1.9
17.8
4.7
2.9
8.4
109.3
9.6
3.2
19.2
181.8
14.0
3.4
29.3
442.9
ZAPS
Day
1.2
1.2
1.2
6.0
1.8
1.7
2.28
24.38
3.2
2.0
4.5
44.5
6.5
2.3
10.2
92.7
II
Night
1.3
1.3
1.4
6.1
4.1
2.5
6.4
30.7
8.0
2.7
13.1
62.9
15.5
3.1
28.5
147.9
-pphm S02, HIGH run
101
-------
Figure 6.1.
IE
.c
OL
a.
c
o
o
f— •
0)
(J
(2
o
o
CM
0
c
o
0)
*£
0
0)
E
o
0)
o
£O
2 20
Q.
E
T3
O)
0) 1 s
Q. ' °
C/)
•C,
c
'>
o> 10
0<
o
^
d)
<
o 5
n
1 1 1 1 1 1 1 1 1 1 1 1
ZAPS n, 1978
High Plot
A pphm S02= I70(I/WS)-20
/ \ p
/ \ r = 0.93
C5
_ i
\ c>r\ f\
\ 2 / ^^
. /
\ <
\ /
^ J -
O
\ /
\ /
t W S /
yt<— • — * *""3*v>»^
— 9 / Xv_ ^o"^ ^ — •-
*
i i i i i i I I i i I I
w 2 4 6 8 10 12 14 16 18 20 22 24
Hour of Day
Diel Cycle of wind speed and SC>2 levels (seasonal averages)
100
c 10
o
T)
O)
E
o>
"3
c
'E 100
CO
OJ
O
CO
10
I I I
ZAPS I, 1978
\ \N. Medium
O.I I 10 50 95 O.I I 10 50 95 99
Cumulative Frequency Above Given Concentration (%)
Figure 6.2. Frequency distributions of S02 concentrations on ZAPS I, 1978,
102
-------
lOOr
o
T3
c
E 1007
oo
CM
O
V)
a.
o.
O.I I 10 50 95 O.I I 10 50 95
Cumulative Frequency Above Given Concentration (%)
Figure 6.3. Frequency distributions of S02 concentrations on ZAPS II, 1978.
and diel variation have been presented. The potential significance of the
substantial difference between day and night exposure regimes has not been
emphasized.
Most organisms have diel activity cycles and consequently are likely to
be affected more at certain times of day than at others. The biologically ef-
fective S02 exposure for green plants must be radically different from that
for nocturnal rodents. In developing relationships betweeen exposure concen-
trations and effects (dose-response), differences in biologically effective
exposures among various organisms must be taken into consideration. In
general, day active organisms received substantially lower exposure at canopy
height than did night active organisms. The biologically effective exposures
of day-active organisms may be substantially less than implied by the overall
seasonal summary data. (Figures 6.2 and 6.3, Table 6.3)
The Diel Pattern of SQ2 Concentrations Near "Real World" Sources
Area sources, such as urban centers, tend to have a bimodal pattern of
daytime SOa concentrations with peaks at mid-morning and late afternoon fol-
lowed by low, constant night time levels (Holzworth, 1973; Munn and Katz,
1959). The diurnal S02 concentration patterns are primarily determined by
variations in mixing, wind speed and source strength. Concentrations in the
vicinity of point source are in addition to these factors, highly dependent
upon wind direction. Plume strikes are likely to be more frequent during the
103
-------
100
10
CVJ
o
o.
o.
O.I
ZAPS H, 1978
A = Control
B = Low Treatment
C = Medium Treatment
D = High Treatment
Day and night measure-
ments combined for entire
season. Taken 35cm
above ground at canopy
surface.
O.I I 10 50 95
Cumulative Frequency Above Given Concentration (%)
Figure 6.4. Frequency distributions of S02 concentrations monitored on ZAPS
II, 1978 compared with those monitored in Chicago, San Francisco,
(1962-67) (HEW, 1970), and Hay Coulee (7.5 Miles S.E. of Colstrip
power plants, 1977-79).
104
-------
e
.c
Q.
O.
OJ
O
CO
c
O
0>
E
O
o>
O
12
I I
10
9
8
7
6
5
4
3
2
I
0
-o-
Medium
Low
I I
•u——o—-^J—
Control
I I I
I I I I I I
High
Medium.
Control
I I I I I I
May J J A S 0 May J J A S 0
Figure 6.5. Monthly variation in geometric mean S02 concentrations for 1978
(HIGH run).
daytime because conditions often favor mixing while at night stable conditions
tend to prevent mixing of the plume to the ground (Smith et al., 1979). The
most probable concentration near a point source is zero. High S02 levels are
rare and unpredictable.
The diel pattern of the frequency of elevated S02 concentrations measured
at Hay Coulee, 7.5 miles southeast of Colstrip, is unimodal with plume strikes
occurring most frequently near midday and gradually decreasing in frequency to
very low levels at night (Figure 6.6) (Ludwick et at., section 1). This is the
inverse of the diel pattern of concentrations observed on the ZAPS plots. At
Hay Coulee, an elevated S02 concentration is dependent upon wind transport
from the stacks at Colstrip. Plume strikes tend to occur during the day when
wind speeds are elevated and turbulent mixing brings the plume to the ground.
On ZAPS, on the other hand, dilution at elevated wind speeds causes SOa
concentrations to decrease during the day. Here the canopy lies within 35 cm
of the S02 sources and the importance of wind in plume transport is less than
at Hay Coulee.
105
-------
80
o
£ 70
E
= 50
o_
o 40
o
I 30
| 20
IL
• 10
6 8 10 12 14 16 18 20 22 24
Hour of Day
Figure 6.6. Frequency of intercepts of the Colstrip plume at Hay Coulee
(7.5 miles S.E. of the stacks).
Though the ZAPS exposure regime appears to simulate the shape of the
concentration frequency distribution curve at Hay Coulee (Figure 6.4) reason-
ably well, it does not replicate the diel cycle of this point source. The
ZAPS exposure regime is expected to simulate an area source with greater
fidelity. Sulfur dioxide concentrations in an area-wide plume could be
expected to have the same relationship with wind speed as those observed on
ZAPS.
CONCLUSIONS
At canopy height, daytime S02 concentrations on ZAPS are substantially
lower than nighttime concentrations. The diel pattern of SO? concentrations
does not resemble that induced by the Colstrip power plants but should
resemble patterns generated by area sources. Diel activity cycles of
organisms on the plots must be considered in defining dose-response relation-
ships.
106
-------
REFERENCES
Holzworth, G.C. 1973. Variations of Meteorology, Pollutant Emissions, and Air
Quality. In: 2nd Joint Conf. Sensing Environ. Pollut., American Chemical
Society, Washington, B.C. pp. 247-255.
Lee, J.J., R.A. Lewis and D.E. Body, 1976. The Field Experimental Component:
Evaluation of the Zonal Air Pollution System. In: The Bioenvironmental
Impact of a Coal-fired Power Plant, Second Interim Report, Colstrip,
Montana, R.A. Lewis, N.R. Glass and A.S. Lefohn, eds. EPA 600/3-76-013,
U.S. Environmental Protection Agency, Corvallis, Oregon, pp. 188-202.
Lee, J.J, and R.A. Lewis. 1978. Zonal Air Pollution System. In: The Bio-
environmental Impact of a Coal-fired Power Plant, Third Interim Report,
Colstrip, Montana, E.M. Preston, and R.A. Lewis eds. EPA-600/3-78-021,
U.S. Environmental Protection Agency, Corvallis, Oregon, pp. 322-344.
Lee, J.J., E.M. Preston and R.A. Lewis. 1978. A System for the Experimental
Evaluation of the Ecological Effects of Sulfur Dioxide. In: 4th Joint
Conf. Sensing Environ. Pollut., American Chemical Society, Washington,
D.C. pp. 49-53.
Lee, J.J., E.M, Preston, and D.B. Weber. 1979. Temporal Variation in S02
Concentration on ZAPS. In: The Bioenvironmental Impact of a Coal-fired
Power Plant, Fourth Interim Report, Colstrip, Montana, E.M. Preston and
T.L. Gullett. eds. EPA 600/3-79-044, U.S. Environmental Protection
Agency, Corvallis, Oregon, pp. 284-305.
Munn, R.E. and M. Katz. 1959. Daily and Seasonal Pollution Cycles in the
Detroit-Windsor Area. Internat. J. Air Poll., 2:51.
Smith, T.B., D.L. Blumenthal, J.A. Anderson, and A.H. Vanderpol. 1979.
Transport of SQa in Power Plant Plumes: Day and Night. Atmos. Environ.,
12:605-611.
107
-------
SECTION 7
VERTICAL S02 CONCENTRATION PROFILE ON
ZAPS DURING THE 1978 FIELD SEASON
L. T. K. Wong, E. M. Preston, and T. L. Gullett
ABSTRACT
During July and August of 1978, the vertical
profile of S02 concentrations within the vegetation
canopy was'studied. At a central point on each treat-
ment plot, S02 concentrations were measured hourly at
approximate ground level, mid-canopy and canopy
surface with a flame photometric S02 analyzer. Vert-
ical stratification in S02 concentration was
pronounced at night and minimal during .daytime.
Geometric mean S02 concentration at median vegetation
height during daylight was approximately 1.4, 1.8, and
5.8 pphm on ZAPS I and 1.4, 1.8, and 2.8 pphm on ZAPS
II Low, Medium, and High plots respectively. These
concentrations are substantially lower (25%-70%) than
the canopy surface exposures reported in Section 6.
The S02 vertical profiles are reasonably similar to
those which might be expected during fumigation from a
coal-fired power plant plume.
INTRODUCTION
A detailed description of ZAPS can be found in Lee and Lewis (1978), and
in the preceding section. Because the ZAPS is an open system covering a large
land area, variations in S02 concentration in vertical and horizontal space
and time may occur. In Section 6, horizontal distribution of S02 on the ZAPS
plots was discussed. In this section, vertical distribution of S02 within the
plant canopy is characterized.
During the 1977 field season, the vertical profile of concentration on
the ZAPS plots was estimated using Huey sulfation plates (Preston and Gullett,
1979). The plates absorb ambient S02 and retain it as lead sulfate. The
amount of sulfate recovered after a given exposure period allows estimation of
relative exposure rates among sample locations. Results indicated that S02
concentrations were highest near the top of the plant canopy and decreased
with increasing depth within the canopy (Preston and Gullett, 1979).
108
-------
This vertical stratification of S02 has important implications for quant-
ification of exposure dose since organism activity is also stratified within
the canopy and this stratification may vary temporally. Therefore, during the
1978 field season, diel variation in the pattern of vertical S02 stratifica-
tion was investigated.
MATERIALS AND METHODS
Real time sulfur dioxide concentrations at 1, 15 and 31 cm above the
ground were measured at a common central point on all plots during the 1978
growing season. With minor changes (see Section 6), the procedures described
in Lee et al • (1979) were followed. Measurement periods are shown in Table
7.1.
Air temperature, relative humidity, and wind speed and direction were
also monitored with procedures described in Dodd et al. (1979).
TABLE 7.1. S02 CONCENTRATION MEASUREMENT PERIODS ON ZAPS, 1978
Treatment Plots
ZAPS
High Treatment
Medium Treatment
Low Treatment
I
II
7/26/78 •» 8/29/78 /
(34 days)
7/14/78 -* 7/27/78
(14 days)
7/13/78 -»• 7/26/78
(14 days)
7/27/78 -» 8/08/78
(13 days)
6/27/78 -> 7/13/78
(17 days)
8/09/78 -> 8/28/78
(19 days)
RESULTS AND DISCUSSION
Diel Variation in the S02 Profile
Sulfur dioxide concentrations averaged over the entire monitoring period
for each hour for each plot at 1, 15 and 31 cm above the ground are shown in
Figures 7.1 and 7.2. In general, S02 concentrations were highest near the top
of the plant canopy and decreased with height. However, the vertical concen-
tration profile varied over the diel cycle. Significant vertical stratifica-
tion of S02 only occurred at night when S02 concentrations at the canopy
surface were high. During daytime, concentrations were lower and vertical
differences were small.
The observed diel variation in vertical S02 concentration profile is
consistent with theoretical expectations based on diel changes in micrometeor-
ological conditions on the ZAPS plots. S02 concentration is expected to
decrease towards the ground because the ground and vegetation are sinks for
the pollutant. The magnitude of the concentration gradient should be
inversely related to the degree of air turbulence near the ground. Wind speed
109
-------
10
ZAPS I, 1978
3lcm
e
.C
3 10
c
o
c
0)
o
c
o
0 0
CJ
o
en
S 20
o
E
o
15
10
i I 1
Low Treatment
V * l
/ \ k. j. -r ^ , / \ /
/ \ Medium Treatment / »./
' V ' N
\ / ^3lcm
"- \
15cm
'-..-• I cm
High Treatment
I5cm
• 31cm
'• I cm
10 15 20 24
Time of Day (hour)
Figure 7.1. Diel variation in the S02 profile on ZAPS I, 1978.
110
-------
10
ZAPS E, 1978
E
.c
0.
c.
o
0
10
2 5
c
0)
o
c
o
O
00
O
c
o
(D
O
i_
o>
e
o
CD
O
0
25
20
15
10
Low Treatment
Medium Treatment
3lcm
15cm
High Treatment
/ \ A
V N3lcm
I5cm
5 10 15 20 24
Time of Day (hour)
Figure 7.2. Diel variation in the S02 vertical profile on ZAPS II, 1979.
Ill
-------
8 12 16
Time of Day (hour)
Figure 7.3. Diel variation in wind speed on ZAPS I and II, 1978.
at the ZAPS sites was higher during daytime (Figure 7.3). The greater winds
during the day increased the dilution of S02 and resulted in lower daytime
concentration.
The correlation between hourly average wind speed and hourly average S02
concentration on the ZAPS plots was estimated. Because the only wind speed
measurement was taken 210 cm above the ground, it was necessary to extrapolate
wind speed from 210 cm to 1, 15 and 31 cm above the ground. This was done
using equations developed by Ripley and Redmann (1976) for a grassland similar
to the ZAPS plots. The equations were generated by curve fitting wind speed
measured at various heights above the ground.
Wind speed above the canopy:
u(Z) = u(h) {In [(Z-d)/ZQ] /ln[(h-d)/ZQ] } Z 3: h
(1)
Wind speed within the canopy:
112
-------
u(Z) = u(h) exp{-[0.385 (Z/h) + 0.72] L(Z)} Z S h (2)
where u = wind speed, ms-1
Z = height above ground
d = zero plane displacement, ^ 10 cm
Z = aerodynamic roughness length, ^ 2.7 cm
h = canopy height, ~ 17.5 cm
L(Z) = total leaf area index above height Z
If the wind speed at one height above the canopy is known, the vertical wind
profile can be estimated from these two equations.
Results of correlation analysis for representative treatment plots IB
(low treatment) and IID (high treatment) are summarized in Table 7.2. As
expected, S02 concentration was roughly inversely proportional to the wind
speed.
Condition of the vegetation surface may also influence the vertical S02
gradient. Wetness of the vegetation surface influences S02 uptake rates.
Because of higher relative humidity and dew formation at night, greater S02
uptake rates than during daytime may occur even though stomatal resistance is
less during daytime. Greater uptake rates at night should enhance the vert-
ical concentration gradient.
TABLE 7.2. WIND SPEED AND S02 CONCENTRATION CORRELATION
Correlation Coefficients
Wind speed (u) and S02 concentration (c)
for height (z), u(z) c(z)
Plot
ZAPS IB
ZAPS IID
c(31)
u(210)
-0.855
-0.823
c(15)
u(210)
-0.862
-0.891
c(l)
u(210)
-0.901
-0.867
c(31)
u(31)
-0.851
c(15)
u(15)
-0.867
c(l)
u(l)
-0.883
Thus the observed diel variation in the vertical profile in S02 concen-
tration can be explained from meteorological factors alone. However, the
vertical concentration gradient should also be enhanced by the ZAPS exposure
system. Sulfur dioxide concentration is dependent upon distance from the
source and thus should decrease between the exposure pipe network and the
ground unless wind speed is sufficient to keep the air mass between the
network and the canopy surface well mixed. This artifact of ZAPS design would
also enhance the vertical S02 gradient at night when winds are calm.
113
-------
Effective Exposure for Grasses
Figure 7.4 shows the mean profiles of daytime (6 a.m.-6 p.m.) and night-
time (7 p.m.-5 a.m.) S02 concentrations for the ZAPS plots. Since the uptake
of S02 by plants is greatest during daytime when the stomata are open, their
effective exposure dose is more nearly characterized by the day curves.
Effective exposure dose in grasses is also influenced by the vertical
distribution of leaf area. Vertical differences in S02 concentration results
in non-uniform exposure to leaves at different heights. Cumulative percentage
of total leaf area as a function of height above the ground for a grassland
similar to the ZAPS plots is given in Figure 7.5 (Ripley and Redmann, 1976).
Fifty percent of total leaf area is found below 6.5 cm above the ground and
95% below 20 cm. The median effective daytime exposure for grasses on the
ZAPS plots was estimated by the daytime geometric mean S02 concentration at
6.5 cm for each treatment plot (Table 7.3). These are substantially lower
(25-70%) than the canopy surface concentrations reported in Section 6.
TABLE 7.3. EFFECTIVE DAYTIME S02 EXPOSURE (PPHM)
Low Treatment Medium Treatment High Treatment
ZAPS I
ZAPS II
1.40 (0.48)*
1.45 (0.39)
1.85 (0.64)
1.75 (0.56)
5.75 (0.25)
2.80 (0.70)
* Proportional decrease from combined day-night canopy surface concentrations
computed as reported in Section 6.
Comparison with other Area and Point Sources
It is of interest to know if the pattern of vertical stratification in
S02 concentrations within the vegetation canopy observed on ZAPS is reasonably
similar to that which might be expected in "real world" pollution fumigations.
Unfortunately, appropriate field data for comparison are lacking. The
major interest has been the estimation of S02 flux to the canopy surface
(Shepherd, 1974; Owers and Powell, 1974). Extrapolation of above canopy S02
concentration profile into the canopy is not possible because the turbulent
structure of the air flow changes inside the canopy (Inoue, 1963). Moreover,
the effect of S02 uptake by plants can be significant (Bennett and Hill,
1973).
In the absence of field data, vertical S02 concentration profile within a
grass canopy was estimated theoretically. Based on the assumption that S02
concentration within canopy depends on wind speed and leaf surface abundance,
a semi-black box model was constructed.
114
-------
40
30
e20
o
c
O
O
10
°
20
10
ZAPSI
CD
»•
\
/ B = low treatment
/ C= medium treatment
D= high treatment
o - Daytime
• -- Nighttime
+
+
+
ZAPS n
0 2 4 6 8 10 12 14 16 18 20
Geometric Mean S02 Concentration (pphm)
Figure 7.4. Mean S02 vertical profiles for daytime (6 a.m. - 6 p.m.) and
nighttime during the study periods.
C(Z) = C(h)
4 LAI(h) - LAI(Z)
LAI(h)
WS(fa) - WS(Z)
WS(h)
(3)
where C = S02 concentration (pphm)
Z = height above ground (cm)
h = height of canopy (cm)
LAI = leaf area index (Ripley and Redmann, 1976)
WS = wind speed
115
-------
30
E
o
TJ
§ 20
o
l_
o
0>
o
.a
a>
-------
E
o
30
I 20
CD
O
.0
cu
X
10
0
/ 2
1234
S02 Concentration (pphm)
/ Chamber measurement (Bennett & Hill, 1973)
2 ZAPS HB (low treatment)
J Intermediate fumigation zone of coal-fired power plant
4 ZAPS EC (medium treatment)
Figure 7.6. Comparison of the vertical S02 concentration profiles on grass-
land canopies measured on selected ZAPS treaments during chamber
study, and expected in the vicinity of a coal-fired power plant.
Fumigation Zone
Annual Average S02 Concentration (pphm)
Inner
Intermediate
Outer
5.0
3.0
1.0
Wind speed profile was estimated as described earlier. With these data,
the vertical S02 concentration profile for the inner, intermediate and outer
fumigation zones of a coal-fired power plant was computed from Equation 3.
The result is compared with the vertical 50% concentration profiles observed
on the ZAPS plots in Figures 7.6 and 7.7. There is a general similarity
117
-------
30
20
10
0 30
3
I 20
O)
>
o
J2
< 10
Coal-Fired Power
Plant Impact Zones
X= outer fumigation
zone
Y= intermediate
fumigation zone
Z= inner fumigation
zone
2 4 6 8 10 12
Geometric Mean S02 Concentration (pphm)
Figure 7.71 Comparison of vertical S02 concentration profiles observed on
ZAPS with estimates of those expected in the vicinity of a coal-
fired power plant.
between the shape of the vertical S02 concentration curves on ZAPS, chamber
experiment, and coal-fired power plant fumigation zones. The low treatment
and medium treatment plot on ZAPS are comparable respectively to the outer and
intermediate fumigation zone. The high treatment plot has a higher S02
concentration than the inner fumigation zone.
118
-------
CONCLUSIONS
Vertical stratification of S02 concentrations within the grassland canopy
on treatment' plots is pronounced at night but small during daytime when wind
speed is elevated. The vertical profiles of S02 concentration are reasonably
similar to those which might be expected during fumigation from a coal-fired
power plant plume.
REFERENCES
Bennett, J. H. and A. C. Hill. 1973. Absorption of Gaseous Air Pollutants by
a Standardized Plant Canopy. J. Air Pollut. Control Assoc., 23:203-
206.
Dodd, J. L. , W. K. Lauenroth, R. G. Woodmansee, G. L. Thor, and J. Chilgren.
1979. Soil, Chemical, and Meteorological Characteristics at ZAPS. In:
The Bioenvironmental Impact of a Coal-Fired Power Plant, Fourth Interim
Report, Colstrip, Montana, E. M. Preston and T. L. Gullett, eds. EPA-
600/3-79-044. U.S. Environmental Protection Agency, Corvallis, Oregon.
pp. 332-390.
Inoue, E. 1963. On the Turbulent Structure of Airflow within Crop Canopies.
J. Meteorol. Soc. Jap. 41(6):317-325.
Lee, J. J. and R. A. Lewis. 1978. Zonal Air Pollution System. In: The
Bioenvironmental Impact of a Coal-Fired Power Plant, Third Interim
Report, Colstrip, Montana. E. M. Preston and R. A. Lewis, eds. EPA-600/
3-78-021. U.S. Environmental Prot.ection Agency, Corvallis, Oregon. pp.
322-344.
Lee, J. J. , E. M. Preston, and D. B. Weber. 1979. Temporal Variation in S02
Concentration on ZAPS. In: The Bioenvironmental Impact of a Coal-Fired
Power Plant, Fourth Interim Report, Colstrip, Montana. E. M. Preston and
T. L. Gullett, eds. EPA-600/3-79-044. U.S. Environmental Protection
Agency, Corvallis, Oregon, pp. 284-305.
Owers, M. J. and A. W. Powell. 1974. Deposition Velocity of Sulphur Dioxide
on Land and Water Surfaces Using a ^5S Tracer Method. Atmos.
Environ., 8:63-67.
Preston, E. M. and T. Gullett. 1979. Spatial Variation of Sulfur Dioxide
Concentration on ZAPS During the 1977 Field Season. In: The Bioenviron-
mental Impact of a Coal-Fired Power Plant, Fourth Interim Report,
Colstrip, Montana, E. M. Preston and T. L. Gullett, eds. EPA-600/
3-79-044. U.S. Environmental Protection Agency, Corvallis, Oregon, pp.
306-331.
Ripley, E. A. and R. E. Redmann. 1976. Grassland. In: Vegetation and the
Atmosphere, Vol. 2. Case Studies, J. L. Monteith, ed. Academic Press,
London, pp. 349-398.
Shepherd, J. G. 1974. Measurements of the Direct Deposition of Sulphur
Dioxide onto Grass and Water by the Profile Method. Atmos. Environ.,
8:69-74.
119
-------
SECTION 8
MYCORRHIZAL ASSOCIATION AND ROOT
CHARACTERISTICS IN WESTERN WHEATGRASS
FUMIGATED WITH SULFUR DIOXIDE
P. M. Rice, C, C. Gordon, P. C. Tourangeau and L. Pye
ABSTRACT
The degree of Agropyron smithii (western wheatgrass)
root tissue infection by beneficial symbionic fungi
belonging to the group (Endogonaceae) of vesicular-
arbuscular endomycorrhizae declines as leaf sulfur
content increases with sulfur dioxide fumigation. The
proportion of vesicles increases while arbuscules, the
nutrient transfer stage of the endophyte, decline as a
function of leaf sulfur load, suggesting the normal balance
between the symbionts may be altered. Increased root
senescence results with increases in leaf sulfur and root
sulfur. A. smithii- culms are larger, roots more vigorous
and infection is higher on two control plots (OPC 1 and
OPC 2) than a third control (ZAPS IIA) possibly as a
consequence of the latter still recovering from grazing
pressure and receiving some sulfur dioxide fumigation.
The quantification of infection levels by absorbance
determination cannot be applied with consistent reliability
to long term field studies.
INTRODUCTION
Endomycorrhizae are a fungal group which form a beneficial,, symbiosis
with the root systems of many higher plants, including grasses, forest trees
and many food crops (Baylis, 1974). The host plant provides carbohydrates to
sustain the fungus. The primary advantage to the host is an increased
ability to extract mineral nutrients from the soil, especially phosphorous
(Sanders and Tinker, 1973). Increased nodulation and nitrogen fixation
(Richards and Voigt, 1964), disease deterrence (Zak, 1964) and more efficient
soil water extraction (Safir et al., 1971) are among the numerous additional
benefits. Ruehle and Marx (1979) have summarized the importance of fungal
root symbionts to plants of economic importance.
The hyphae of endomycorrhizae actually penetrate the feeder roots and
enter the cortex of the host plant roots. This is a controlled entry and does
not normally disrupt the morphological structure of the host. In addition to
120
-------
the hypae, two distinct structures, vesicles and arbuscules, are formed within
the host cells. Vesicles, which are found externally as well as within the
roots, are a spherical enlargement of a hyphal strand. The vesicles are
considered to serve as both food storage organs and infectious spores (Kinden
and Brown, 1975). Arbuscules are formed by a multitudinous branching of the
hyphae in a localized area and are believed to be the mechanism which transfers
mineral nutrient from the fungus to the host (Cox and Tinker, 1976). The
vesicles, arbuscules and internal growth habit of these root-associated fungi
are the basis for referring to this group as vesicular-arbuscular (VA)
endomycorrhizae (EDM). We have reviewed the biology of these organisms as it
relates to this study in an earlier report (Rice et al., 1979).
There has been little study of the responses of mycorrhizae to air pollu-
tants. Ectomycorrhizae provide an analogous function in terms of mineral
nutrition for their host, but do not actually penetrate the roots as do the
endomycorrhizae, Grzywacz (1964) is cited in McCool et al, 5 (1979) as reporting
a decline in the frequency of fungal association with the roots of forest trees
subjected to sulfur dioxide and fluoride pollution. An investigation of the
ectomycorrhizae on spruce (Sobotka, 1964 in Heagle, 1973) stressed by sulfur
dioxide also reports a detrimental impact on the symbiosis. Recent work at
the Statewide Air Pollution Research Center, Riverside, California, has
demonstrated that the endomycorrhiza Glomus faseiculatus associated with
tomato plants is sensitive to both sulfur dioxide and nitrogen dioxide (Granett,
1978). Also working at Riverside, McCool et al., (1979) reported reductions
in infection and spore production of G. fasciculatus associated with citrange
when the host plant was fumigated with ozone.
In 1977, a study was conducted at the Zonal Air Pollution Systems (ZAPS)
site to determine whether endomycorrhizae were associated with prairie
grasses occupying the sites and whether the association was influenced by
sulfur dioxide treatment. The results of the investigation are reported by
Rice et al., (1979) and are briefly summarized here:
(1) Vesicular-arbuscular endomycorrhizae are present with several
prairie grass species including Agropyron smithii (western wheat-
grass) .
(2) Infection levels declined with increasing ambient sulfur dioxide
treatments and leaf sulfur concentrations.
(3) Swollen interior vesicles and excessive numbers of exterior vesicles *
were observed on the low treatment plots, suggesting a disruption of
the normal balance between the symbionts.
(4) A colorimetric method for determining infection levels (Becker and
Gerdemann, 1977) was correlated with microscopically determined
levels of infection and thus, potentially could be used as a much
less labor-intensive assay technique.
This study was continued through 1978 to further refine the methods and confirm
the 1977 observations. The collection and preparation of A. smithii root samples
for mycorrhizae assays also provided an opportunity to investigate a number of
other root and shoot characteristics.
121
-------
MATERIALS AND METHODS
The ZAPS II site consists of four .5 hectare plots fumigated with sulfur
dioxide for the third growing season in 1978. Gas is delivered at different
but fixed rates to three of the plots by a perforated pipe network supported
75 cm above the ground. Meteorological factors induce significant temporal
variation in ambient concentrations of each of the plots, and the fourth plot
receives sulfur dioxide as a consequence of wind drift from the three plots
which are directly fumigated. Two additional grassland plots (Off Plot Controls,
OPC), well removed from the ZAPS II site, were also sampled. This was done
primarily to provide control plots which did not receive sulfur dioxide as
a consequence of wind drift. The ambient sulfur dioxide concentrations of the
four plots at the ZAPS II site were continuously measured at the top of the
plant canopy (30 cm) by a flame photometric analyzer (See section 6, Table 6.2).
In addition to temporal variation in sulfur dioxide concentrations, significant
horizontal and vertical concentration gradients exist. Daytime concentrations
are less than the stated mean, and the average concentration declines with
descent into the plant canopy. Detailed description of the fumigation patterns
are provided by Preston et al., (Section 6) and Preston and Gullett (1979).
Rice et al., (1979) have demonstrated that sulfur accumulation by plant tissues
reflects these gradients within the individual treatment plots. Therefore,
stated concentrations reflect upper limits of dosage delivered.
Twenty subplots on each treatment plot were randomly chosen for sampling
in June, July and September of 1978. One soil core (15 cm diameter by 20 cm
deep) with intact roots and shoots was taken from each subplot. The cores were
bagged in plastic and shipped to the Environmental Studies Laboratory in
Missoula, Montana, for processing. Cores were refrigerated at the laboratory
and processed by blocks in a random manner within one week.
The cores were trimmed to a uniform size (10 cm diameter x 15 cm deep),
softened in water for one hour, and the entire ^4. smith-Li- plants separated
from the soil with a hose. The number of live culms of A. smi-thii per core
and the weight of ten current year culms per core were recorded. Fresh root
weight was determined after the recovered root mass was placed in a humidifier
to obtain full turgor and then blotted dry. Sub-samples (.5 g) of the root
mass were autoclaved to obtain a hot water extract for colorimetric determination
(400 NM), to be used as an index of mycorrhizal infection, with an adaptation
(Rice et al., 1979) of a method proposed by Becker and Gerdemann (1977). An
additional modification of previous techniques was the filtration of the
extracts to remove any debris.
After the absorbance readings were obtained, the same root mass was cleared
in KOH and treated with trypan blue in lactophenol (Phillips and Hayman, 1970)
to stain the fungal elements in and on the root tissues. Infection levels
were then estimated by a modification (Rice et al., 1979) of Newman's (1966)
point intersection method as done by Sparling and Tinker (1975). The method
consists of spreading the stained root sample in a petri dish, randomly
selecting ten fields of visions under 21X binocular magnification, and scoring
the number of mycorrhizal or non-mycorrhizal roots intersected by a rotated
ocular graticule. Mycorrhizal presence was further stratified as to arbuscules,
122
-------
vesicles and other signs of infection (distinctive hypae, entry points, etc.).
Both the absorbance readings and visual determinations were conducted as blind
tests.
The point intersection method was also used to determine total relative
root length (// of intersections per .5 g sample), to assign all roots to one
of three size classes (less than .5 mm, .5 mm to 1.5 mm and greater than 1.5
mm), and to establish the ratio of browning and senescent roots.
The sulfur content of live shoots and root tissues were determined for
each sample item by the Leco induction method (Gordon, 1976). The variables
and collections for which they were determined are summarized in Table 8.1.
The basic experimental design employed for the biological response variable
was the randomized complete block form of the two-way anova. Samples were
collected, processed and recorded under the format of this model. The limited
size of some individual sample items and the labor-intensive nature of the
scoring restricted the generation of replicate data. When replication was
available, these points were pooled, and consequently interaction was assumed
to be non-significant. Various transformations were employed where suitable.
Zero value residuals were calculated for missing sample items, adjusted sum
of squares computed (Sokal and Rohlf, 1969) and degrees of freedom reduced
appropriately. The ZAPS HA plot was considered to be the "control," and
all other plots including the OPC's were contrasted pair-wise to ZIIA (Dunnett,
1955).
TABLE 8.1. SUMMARY OF THE VARIABLES QUANTIFIED IN THE ZAPS MYCORRHIZAL STUDY
VARIABLE UNIT COLLECTION
Sulfur in live tops PPM July
Sulfur in roots PPM July
Absorbance (400 NM) % June, July
Visual Infection % July
Vesicles % July
Arbuscules % July
Total relative root length // of intersections July
Small roots (<.5 mm dia.) % July
Medium roots (.5 mm to 1.5 mm dia.) % July
Large roots (> 1.5 mm dia.) % July
Browning and senescent roots % July
Culms per core # July
Dry Weight Per Culm mg September
Fresh root weight per culm mg July
123
-------
These treatments of the data constituted a preliminary analysis. It had
been anticipated that tighter control of experimental procedures based on
experience gained by the 1977 work would allow meeting the assumptions of
analysis of variance. Homoscedasticity was obtained for most data sets,
but uncertainies about normality remain. Accordingly, non-parametric
methods (Spearmann's Rank Correlation and Mann-Whitney (Wilcoxon) Rank Sum
Test) were applied to the 1978 data where appropriate, as was done with the
1977 data.
RESULTS
Colorimetric Absorbance Determination of Mycorrhizal Infection
The previous study of the relationship between endomycorrhizal infection
levels as estimated by microscopic inspection of roots and infection level
determined by root extract absorbance at 400 NM had suggested that although
absorbance was not a precise predictor of infection levels, it did have
potential as an index of relative levels of infection. A significant
positive correlation between infection and absorbance was not re-established
for the 1978 field work (p > .95). The procedure had been altered to include
micropore filtration of the root extract to remove plant debris. The sample
processing time was also increased as a consequence of larger cores and more
numerous samples. This may have affected the absorbent or increased potential
light degradation of the absorbance capacity of the molecule/ The absorbent
has not been isolated or chemically identified, so its role in the host-fungus
relationship is unknown. Also, the percentage of browning and senescent roots
of A. smithii increased strongly with the increase in sulfur dioxide treatment.
Becker and Gerdemann (1977) had suggested "This method is not recommended for
long term experiments where phenolic compounds of dying roots could interfere
with the absorbance of the water extract." With present methods and without
identification of the chemical nature and biological role of the absorbent,
the colorimetric assay appears unreliable for determining mycorrhizal infection
levels in stressed field populations.
Sulfur Accumulation in Tissues
Sulfur content in live leaves of A. smithi-i from the July collection was
significantly (p <^ .01) elevated on each of the ZAPS II treatment plots
relative to the ZAPS IIA "control" plot (Table. 8.2). Although OPC 1 and
ZAPS IIA had similar sulfur levels in live tissue, the OPC 2 plot had a higher
sulfur content (p <_ .05) than the site II "control." Chemical analysis of the
roots indicated a trend of increasing sulfur content with increasing ambient
sulfur dioxide levels, but the difference relative to the ZAPS IT. "control"
was significant (p <_ .01) only for the ZAPS IID plot (Table 8.2). However,
the rank correlation between sulfur in roots and live leaves indicated a
very highly significant positive relationship (Spearman Rho = .4073, n = 88,
p _<_ .001). These patterns of sulfur accumulation are consistant with those
observed at ZAPS in previous years.
124
-------
TABLE 8.2. SULFUR CONTENT (PPM1") OF Agropyron smithii FOR THE JULY 1978
COLLECTION
Treatment Plot OPC 1 OPC 2 ZAPS IIA ZAPS IIB ZAPS IIC ZAPS IID
Live Leaves 1008 1106* 931 1166** 1156** 2498**
Roots 640 595 641 703 719 843*
*p <^ .05, Dunnett's procedure, each treatment plot is compared with the ZAPS
IIA plot.
**p <_ .01, Dunnett's procedure, each treatment plot is compared with the ZAPS
IIA plot.
Antilog of It-
log 10
Mean plot differences separate well by treatment, but substantial differences
exist in dosage accumulated within individual treatment plots (Rice et al.,
1979). Individual variates have substantial overlap with those of other, plots.
These plant-to-plant differences in sulfur accumulation reflect within-plot
heterogeneity in edaphic factors, micrometeorlogical features, gas delivery
system patterns and inherent plant factors. Tissue sulfur accumulation
results from a physiological integration of those factors, and sulfur content
should provide a more precise estimate of low level chronic sulfur dioxide
loading of individual plants than stratification of biological response
variables by treatment plot. The comparison by plot stratification is summarized
in Table 8.3. Table 8.4 presents a rank correlation (Spearman's) matrix
which frees the analysis from the constraint of plot assignment and pairs the
biological responses variables with the specific sulfur tissue content of the
A. smithii from the individual cores. The ranking also causes total linear-
ization of the dose-response curve, assuming it does not contain multiple
inflection points.
Treatment Plots Compared to ZAPS IIA
Stratification by plot reduces the number of items per sample to 20
or less. This limited sample size weakens the test for normality, making it
difficult to detect deviations from the normal distribution. The non-parametric
Mann-Whitney (Wilcoxon) test was used to contrast the treatment plot to the
ZAPS IIA plot (Table 8.3). The best estimate of the sample standard error
as derived from the preliminary two-way anova under appropriate transformations
is also presented in Table 8.3. The pattern of significant differences based
on this standard error and using Dunnett's procedure was quite similar, but
the Mann-Whitney test is preferred because of the aforementioned uncertainies,
including occasional heteroscedasticity.
125
-------
TABLE 8.3. OPC AND SITE II PLOTS COMPARED TO THE ZAPS II A PLOT
Response Variable
Visual Infection
Vesiclesf
Arbuscules"
Total Relative Root
Length
Small Koots
Medium Roots'
Large Roots
Browning and Senescent
Roots'"
Culms/Core''
3 9
Fresh Root
Weight/Culm
Culm Dry Weight55
UPC 1
55.7
(68.3)
IS. 7
(9.9)
25.3*
(18.3)
44.2
42.2*
(45.5)
35.8***
(34.3)
23.9
(16.4)
14.3*
(6.1)
2.73***
(7.46)
1 .94
(87)
2.275**
(188)
OPC 2
57.8**
(71.6)
25.2
(18.2)
30.9***
(26.4)
54.9
49.6***
(57.9)
35.4***
(32.0)
16.5
(8.1)
9.7***
(2.8)
2.96***
(8.77)
1.99
(9ft)
2.306***
(202)
ZAPS IIA
47.3
(54. 0)
20.8
(12.6)
17.6
(9.2)
44.4
34.3
31.8
47.1
(53.6)
20.5
(12.2)
23.3
(15.5)
4.52
(20.48)
1.89
(78)
2.144
(139)
ZAPS I IB
53.4
(64.5)
25.9
(19.1)
11.8
(4.2)
36.0*
28.0
T*. 0
47.2
(53.9)
26.7
(20.2)
25.5
(20.0)
4.50
(20.26)
1 .85
(71)
2.218*
(165)
ZAPS II C
47.6
(58.0)
27.8
(21.8)
14.1
(5.9)
44.6
34.7
32.3
46.5
(52.6)
20.3
(12.0)
31.2**
(26.9)
3.87
(14.99)
1.80
(63)
2.129
(135)
ZAPS I ID
46.6
(52.9)
25.2
(18.2)
13.9
(5.7)
43.0
35.8
34.1
46.0
(51.9)
19.8
(11.4)
34.1***
(31.4)
4.46
(19.91)
1.91
(81)
2.252**
(179)
S.E.t
+3.55
+3.32
+3.02
+4.55
+3.12
+2.52
+2.89
+2.94
+0.303
+0.055
+0.0377
'Significant, p £ .05, Mann-Whltney (Wtlcoxon) test, individual plots contrasted to the ZAPS IIA plot.
"Highly significant, p <_ .01, Mann-Whitney (Wllcoxon) test, IndiviJuul plots contrasted to the ZAPS IIA plot.
•••Highly significant, p £ .001, Mann-Whitney (Wllcoxon) test. Individual plots contrasted to the ZAPS IIA plot.
Best estimate of sample standard error, derived from two-way anova; sample size is 20 except OPC 1 is 19.
Angular transformation, back transformation in parenthesis.
JSquare root transformation, back transformation In parenthesis.
''"Log transformation, original units were milligrams, back transformation in parenthesis.
126
-------
TABLE 8.4. RANK CORRELATION MATRIX
rO
Root Sulfur °*
Infection
Vesicles
Arbusculos
TRRI,1'
Brown Ing
Small Roots
Medium Roots
Large Roots
Root Weight /Culm
Culms/Core
Vesicles/
Infection
Arbiisrul cs/
Infect Ion
Vesicles/
Arbuscules
Culm
Sulfur"
.4073***
-.1786*
. 1 990*
-.2702**
ns
.4705***
ns
.2082*
ns
ns
ns
.3118***
-.2703**
.3201***
Root
'Sulfur Infect Inn
ns -
ns .40r>!***
ns . /ifWl***
ns -.4225***
.4300*** -.5166***
ns ns
ns -.2079**
ns .?6lr>**
ns .2867**
.3895*** ns
ns
ns
ns
Vesicles Arbtiscules TRRI.*
-.2183* .1834*
•ns -.4794*** ns
ns .6005*** .6312***
ns -.5448*** -.3555***
ns ns -.5442***
ns .4286*** ns
ns ns ns
Small Medlun 1-arge Root Welfclit/
Browning Roots Roots Roots Culm
-
- - - - -
-.2461** -
.2880** -.792M** - - -
ns -.5002*** ns
-.2479*** .2703** -.1113*** ns
.3108*** .2051* .1998* ns ns
*P <_ .05, Spe;irm.in'H Rnnk Corn-1 at (on.
**p <_ -01, Spearman's Rank Uirrelat Ion.
***p < .001, Spenrman's Rank CorrelntInn.
Casns where a one-tailed .-t I ternat I vo W.-IK clearly siiKRt'JHi'il by previous work are underlined.
Sample size Is 11.9 except wort1 superscript to the variable indicates otherwise.
TRRL = Total relative rout Irnp.ih.
-------
The OPC plots are separated from each other by about 35 kilometers;
however, they tend to differ similarly from the ZAPS HA plot. Mycorrhizal
infection levels are higher (the exact probability for OPC 1 versus ZAPS HA
was .08), arbuscules are more numerous and the small root fraction is larger
while the medium size root fraction is reduced. Browning and senescent roots
are less numerous, and the average dry weight of culms is greater. There are
few significant differences detected within site II between the intentionally
fumigated plots and ZAPS HA. Total relative root length appears to be less
on ZAPS IIB, culm dry weight is elevated on ZAPS IIB and ZAPS IID, and the
percentage of browning and senescent roots is greater on both ZAPS IIC and
ZAPS IID than the ZAPS HA plot (Table 8.3).
Culm weights can be used only for a plot-to-plot index of relative culm
size and should not be used to estimate yield or shoot/root ratios. The dry
weights appear to be biased upward. The fall 1978 standing crop yield for
site II plots, based on culm density x culm dry weight, ranges from 257 to
453 g/m2, while Dodd et al., (1979) measured peak current year production
ranging from 80 to 120 g/m2 for 1975 through 1977. Unless yield was
unusually high for 1978, it appears that the subsampling of the September
cores for culm weight determination was biased towards the larger culms.
Comparisons of ZAPS IIA with all other plots demonstrate that infection
level, including arbuscular frequency, are greater for the OPC's and that
although culm density is lower, plant growth is more rank and vigorous. The
increase in root necrosis is the only response that clearly parallels the
average sulfur dioxide treatment applied to. the plots.
Biological Responses Compared to Actual Tissue Sulfur Load
The simple rank correlation matrix is arrayed in Table 8.4. Response
variables are paired with the tissue sulfur content of the same plant material
from which the response variable was estimated. Three mycorrhizal component
ratios, vesicles/infection, arbuscules/infection and vesicles/arbuscule, are
also compared with sulfur in tissues. The examination of these ratios permits
the detection of changes in the status of the proportion of vesicles and
arbuscules comprising the total infection.
Infection declines with increasing sulfur content of live culms (p < .05).
This relationship remains when only the intentionally fumigated plots (ZAPS
IIB, C and D) are considered (rs = .2572, n= 57, p < .05). The frequency of
arbuscules in the sample and the proportion of the total infection due to
arbuscular presence (arbuscules/infection) decline with increasing leaf sulfur
content (p < .01). Both the frequency of vesicles and the proportion of the
total infection due to vesicles increase with leaf sulfur (p < .05 and p < .001
respectively). The net result is an overall decline in infection (p < .05)
and an increase (p < .001) in the ratio of vesicles to arbuscules (vesicles/
arbuscle) as a function of leaf sulfur. Significant correlation between these
fungal endophyte variables and root sulfur were not observed (p > .95).
The results support the decline in infection with increasing leaf sulfur
concentration observed in the 1977 study. Excessive numbers of exterior
vesicles were noted on the low treatment plots during 1977, but their observation
1*28
-------
was not quantified in that study. Arbuscules are considered typical of young or
new infections, while vesicles are viewed as representing the mature or
reproductive state of the endophyte (Nicolson, 1959). The increase in vesicles
and yesicles/arbuscule ratio may indicate an alternation of the normal symbiotic
balance between the host grass and the endomycorrhizae in response to sulfur
dioxide-induced stress.
The browning and senescence of roots was positively correlated with both
sulfur content of leaves and roots (p < .001). Culm density is positively
correlated (p < .001) with root sulfur but not sulfur in leaves.
Infection is negatively correlated with total relative root length
(p < .001), browning and senescent roots (p < .001), and medium roots (p < .01),
but positively correlated with large roots and root weight per culm, suggesting
the endophyte is maintained at highest levels in stable A. smith-Li populations
which have sufficient photosynthate to form larger root systems, Root
enlargement is a function of cortex development and concurrent storage of
photosynthate reserves in roots and rhizomes of this perennial grass where
vegetative reproduction dominates. Declining infection levels and altered
vesicle and arbuscule ratios may reflect limited photosynthate availability or
altered host-plant metabolism rather than direct sulfur toxicity to the
endomycorrhizae.
Vesicles decline with increasing total relative root length; arbuscules
increase with total relative root length (p < .05), the frequency of small
size roots (p < .001), and root weight per culm (p < .001), while the correlation
between arbuscules and both browning and medium size roots is negative
(p < .001). This appears to provide support for the concept that arbuscules
are a sign of young infection found in association with active, functional
root systems.
Total relative root length determined by point intersection is a reasonable
index of root growth as is evidenced by its strong correlation with small roots,
and negative correlation with both medium and large roots (p < .001) in these
samples of fixed root weight (.5 g).
Browning and senescence is positively correlated with medium roots (p < .01),
and high culm densities (p < .001), while being negatively correlated with small
roots and high root weight per culm, again suggesting that root necrosis is
greatest in plants with declining belowground growth vigor.
Small roots are positively correlated with root weight (p < .01) and
culms per core (p < .05) in addition to their association with total relative
root length and arbuscule frequency (p < .001). Medium roots are negatively
correlated with root weight per culm (p < .001) but positively correlated with
culms per core (p < .05).
DISCUSSION
The method of estimating the degree of host plant (AgTOpyron smithii)
root infection resulting from the presence of benefical vesicular-arbuscular
endomycorrhizae by measuring absorbance (400 MM) of root extracts from field
129
-------
populations is unreliable as employed. A significant positive correlation
could not be established between absorbance and percent infection determined
by microscopic inspection of 1978 root samples. This is at variance with
the correlation observed for the 1977 work. Procedural changes from 1977 to
1978 may have resulted in absorbent adherence to filters, light degradation
or alterations resulting from increased sample processing time. Phenolic
compounds, released by dying roots, are a recognized interference with absorp-
tion determinations (Becker and Gerdemann, 1977). The proportion of browning
and senescent roots increased strongly with sulfur dioxide treatment at
ZAPS II and may have confounded absorbance readings. The absorbance method
should be restricted to short-term potted plant trials until the biochemical
nature of the absorbing compound is determined and more sensitive field plot
methods develop.
The endomycorrhizal infection and the root dynamics of the two Off Plot
Controls (OPC 1 and OPC 2) are similar in their distinction from the site
II control (ZAPS IIA). They exhibit more extensive endomycorrhizal infection
and lower culm density but decidedly more vigorous growth. The OPC exclosures
have been maintained for 15 years. The ZAPS II exclosure was in place for the
third year in 1978. The OPC communities should be stable in terms of release
from grazing pressure while recovery at site II is still progressing. The
ZAPS IIA plot also receives significant sulfur dioxide fumigation, with
approximately 10 percent of the eight-minute median readings ranging from 2
to 6 pphm, Fumigations less than 2 pphm were below the detection limit of the
analyzer (Preston e± al., Section 6). Sulfur dioxide treatment as low as 3.5 pphm
for one to two hours has been demonstrated to decrease net photosynthesis
and accelerate dark respiration in Vioia faba relative to clean air (Black
and Unsworth, 1979). Prior to power plant operation or outside areas impacted
by Colstrip power plant plume, the identified background level of sulfur dioxide
in southeastern Montana is reported as approximately .1 pphm (Ludwick et at.,
1979). Thus, the possibility that A plot fumigation levels elicit physiological
response in A. smithii should not be excluded. Tissue sulfurs on ZAPS IIA were
not elevated relative to the OPC's in the July 1978 sample, although the 1977
data suggested that the A plots were accumulating ambient sulfur. Mean
differences between the A plot and OPC's are likely to be small, and effects
including accumulation are not a linear function of concentration x duration
when fumigations are acute (Van Haut in Horsman et al., 1979) in the sense
of being relatively higher than background but of short duration. An answer
to the question of elevated tissue sulfur derived from ambient concentrations
on ZAPS IIA must await completion of the entire 1978 data base. Thus,
differential grazing history and sulfur dioxide fumigation may contribute to
the similar disparities of both OPC's relative to ZAPS IIA.
Although it does not integrate the consequences of acute fumigations well,
the sulfur content of leaf tissue directly measures accumulated dosage. Leaf
tissue sulfur means clearly segregate by plot, increasing with ambient level,
on the intentionally fumigated ZAPS IIB, IIC and IID plots. However, individual
leaf tissue sulfur variates within plots overlap substantially with those of
other plots. The mean coefficient of variation for plots in the July collection
was 16 percent. As is typical of living organisms, especially field populations,
the biological responses are substantially more variable. Plot means do not
segregate well except for the increase in browning and senescent roots. The
130
-------
biological responses determined in this study are certainly influenced by
factors other than sulfur loading of tissues, but the correlation analysis
estimates the degree of association between the biological responses and
accumulated dosage without the additional variance of plot assignment.
The correlation between the decline in endomycorrhizal infection and
increasing leaf tissue sulfur, which was also significant for the intentionally
fumigated plots only, may reflect alterations in host plant metabolism rather
than direct sulfur toxicity to the endophyte. Long-term sulfur deposition
can lead to greater relative sulfur increases in belowground organs than in
shoots (Rice et aZ., 1979) and then chemical species metabolized from sulfur
dioxide may cause direct toxicity. However, after fumigation for three grow-
ing seasons at ZAPS II, the increase in root sulfur is still small compared
with that of tops, and mycorrhizal characteristics do not yet significantly
correlate with root sulfur. Several workers have argued that sulfur dioxide
and its higher oxidation states reduce photosynthetic electron transfer (Horsman
and Welburn, 1976), and Ziegler (1975) has shown that ribulose - 1,5 - diphos-
phate carboxylase is inhibited by sulphite, reducing carbon dioxide binding.
Sulfur dioxide-induced decreases in photosynthate availability in the root
system of the host plant could alter the status of the fungal symbiont by
decreasing this principle advantage it derives from the association. Total
vesicles and the proportion of vesicles increase with leaf sulfur. The
endophyte may be accelerating its entry into the sporulation stage or increasing
its own stored food reserves as excess fixed carbon from the host declines
(Hawker, 1966). The decline in total arbuscules and the proportion of
arbuscules as leaf sulfur increases suggest the host plant is receiving a
reduced mineral nutrition benefit from the endophyte. The vesicles/arbuscule.
ratio is positively correlated with leaf sulfur loading. The fungus may be
diverting a larger proportion of its host-derived photosynthate from structures
which directly benefit the host to those which maintain the endophyte. Harley
(1959) has argued that the mycorrhizal relationship, although usually benefical,
is potentially pathogenic. This is not to suggest that sulfur dioxide has
completely shifted the host-fungus relationship from symbiosis to pathogency,
but the symbiosis may be altered towards avirulence.
Potential changes in photosynthate flux to root systems are not the only
changes in host biochemistry which could be affecting the host fungus balance.
The observed increase in root necrosis with leaf and root tissue sulfur could
be increasing phenolic releases to the rhizosphere. Phenols are typically
involved in restricting fungal activity on roots (Ling-Lee et aZ., 1977a,
1977b; Miles, 1968). A single biochemical mechanism is unlikely.
CONCLUSIONS
The following results were derived from the 1978 investigation of endo-
mycorrhizae associated with Agropyron smithii (western wheatgrass) subjected
to sulfur dioxide fumigation and provide significant confirmation of similar
observations made in the 1977 study (Rice et al., 1979).
(1) Sulfur content of leaf and root tissues increase with ambient sulfur
dioxide treatment and leaf sulfur is positively correlated with root
sulfur.
131
-------
(2) The percentage of roots infected by endomycorrhizae decline with
increasing leaf tissue sulfur.
(3) The number of vesicles increases with leaf sulfur content.
New significant findings were:
(1) Arbuscules decline with increasing leaf sulfur accumulation.
(2) The vesicles/arbuscule ratio increases with leaf sulfur load.
(3) The percentage of roots declining into a stage of browning and
senescence increases with sulfur dioxide treatment.
In opposition to earlier results, a significant positive correlation could
not be established between infection determined by visual scoring of roots and
infection estimated from absorbance readings of root extracts. The colorimetric
absorbance method as employed is unreliable for long-term studies involving
field populations.
REFERENCES
Baylis, G. T. S. 1974. The Evolutionary Significance of Phycomycetous
Mycorrhizas. In: Mechanisms of Regulation of Plant Growth, R. L.
Bieleski, A. R. Ferguson, and M. M. Cresswell, eds. Bull. 12. The
Royal Society of New Zealand, Wellington, pp. 191-193.
Becker, W. N., and J. W. Gerdemann. 1977. Colorimetric Quantification of
Vesicular-Arbuscular Mycorrhizal Infection in Onion. New Phytol.,
78:289-295.
Black, V. J., and M. H. Unsworth. 1979. Effects of Low Concentrations of
Sulphur Dioxide on Net Photosynthesis and Dark Respiration of Vioia faba.
J. Exp. Bot., 30(116):473-483.
Cox, G., and P. B. Tinker. 1976. Translocations and Transfer of Nutrients
in Vesicular-Arbuscular Mycorrhizas. I. The Arbuscule and Phosphorous
Transfer: A Quantitative Ultrastrutural Study. New Phytol., 77:371-378.
Dodd, J. L., W. K. Lauenroth, G. L. Thor, and M. B. Coughenour. 1979. Effects
of Chronic Low Level S02 Exposure on Producers and Litter Dynamics. In:
The Bioenvironmental Impact of a Coal-Fired Power Plant, Fourth Interim
Report, Colstrip, Montana, E. M. Preston and T. L. Gullett, eds.
EPA-600/3-79-044, U.S. Environmental Protection Agency, Corvallis, Or.
pp. 384-493.
Dunnett, C. W. 1955. A Multiple Comparison Procedure for Comparing Several
Treatments with a Control. J. Amer. Stat. Assn., 50:1096-1121.
132
-------
Gordon, C. C. 1976. Investigations of the Impact of Coal-Fired Power Plant
Emissions Upon Plant Disease and Upon Plant-Fungus and Plant-Insect
Systems. In: The Bioenvironmental Impact of a Coal-Fired Power Plant,
Second Interim Report, Colstrip, Montana, R. A. Lewis, N. R. Glass, and
A. S. Lefohn, eds. EPA 600/3-76-013, U.S. Environmental Protection
Agency, Corvallis, Or. pp. 61-90.
Granett, A. L. 1978. Effects on Mycorrhizal Fungi Associated with Tomato
Plants Exposed to S02 or N02 Gas. Phytopath. News, 12(9):126.
Grzywacz, A. 1964. The Influence of Industrial Air Pollution on Pathological
Fungi of Forest Trees. Sylwan., 155:55-62.
Harley, J. L. 1959. The Biology of Mycorrhiza. Interscience Publications,
Inc., New York. 233 pp.
Hawker, L. E. 1966. Environmental Influences on Reproduction. In: . The
Fungi. Volume II. The Fungal Organisms, G. C. Ainsworth and A. S.
Sussman, eds. Academic Press, New York. pp. 435-469.
Heagle, A. S. 1973. Interactions Between Air Pollutants and Plant Parasites.
Ann. Rev. Phytopath., 11:365-388.
Horsman, D. C., T. M. Roberts, M. Lambert, and A. D. Bradshaw. 1979. Studies
on the Effect of Sulphur Dioxide on Perennial Ryegrass (Loliwn perenne L.).
I. Characteristics of Fumigation System and Preliminary Experiments.
J. Exp. Bot., 30(116):485-493.
Horsman, D. C., and A. R. Wellburn. 1976. Appendix II. Guide to the Metabolic
and Biochemical Effects of Air Pollutants on Higher Plants. In: Effects
of Air Pollutants on Plants, T. A. Mansfield, ed. Cambridge University
Press, Cambridge, pp. 185-199.
Kinden, D. A., and M. F. Brown. 1975. Electron Microscopy of Vesicular-
Arbuscular Mycorrhizae of Yellow Poplar. II. Intracellular Hyphae and
Vesicles. Can J. Microbiol., 21:1768-1780.
Ling-Lee, M., A. E. Ashford, and G. A. Chilvers. 1977a. A Histochemical
Study of Phenolic Materials in Mycorrhizal and Uninfected Roots of
Eucalyptus fastigata Deane and Maiden. New Phytol., 78:313-318.
Ling-Lee, M., A. E. Ashford, and G. A. Chilvers. 1977b. A Histochemical
Study of Polysaccharide Distribution in Eucalyptus Mycorrhizas.
New Phytol., 80:135-141.
Ludwick, J. D., D. B. Weber, K. B. Olsen, and S. R. Garcia. 1979. Air
Quality Measurements in the Coal-Fired Power Plant Environment of
Colstrip, Montana. PNL-2964, UC90i, Battelle Pacific Northwest
Laboratory, Richland, Wa. 42 pp.
133
-------
McCool, P. M., J. A. Menge, and 0. C. Taylor. 1979. Effects of Ozone and
HC1 Gas on the Development of the Mycorrhizal Fungus Glomus fascioulatus
and Growth of 'Troyer' Citrange. J. Amer. Soc. Hort. Sci., 104(2):151-154.
Miles, P. W. 1968. Insect Secretations in Plants. Ann. Rev. Phytopath.,
6:137-164.
Newman, E. I. 1966. A Method of Estimating the Total Length of Root in a
Sample. J. Appl. Ecol., 3:139-145.
Nicolson, T. H. 1959. Mycorrhizae in the Gramineae. I. Vesicular-Arbuscular
Endophytes, with Special Reference to the External Phase. Trans. Brit.
Mycol. Soc., 42:421-438.
Preston, E. M., and T. L. Gullett. 1979. Spatial Variation of Sulfur
Dioxide Concentrations on ZAPS During the 1977 Field Season. In: The
Bioenvironmental Impact of a Coal-Fired Power Plant, Fourth Interim
Report, Colstrip, Montana, E. M. Preston and T. L. Gullett, eds.
EPA-600/3-79-044, U.S. Environmental Protection Agency, Corvallis, Or.
pp. 306-330.
Rice, P. M., L. H. Pye, R. Boldi, J. O'Loughlin, P. C. Tourangeau, and C. C.
Gordon. 1979. The Effects of "Low Level S02 Exposure on Sulfur Accumu-
lation and Various Plant Life Responses of Some Major Grassland Species
on the ZAPS Sites. In: The Bioenvironmental Impact of a Coal-Fired
Power Plant, Fourth Interim Report, Colstrip, Montana, E. M. Preston and
T. L. Gullett, eds. EPA-600/3-79-044, U.S. Environmental Protection
Agency, Corvallis, Or. pp. 494-591.
Richards, B. N., and G. K. Voigt. 1964. Role of Mycorrhiza in Nitrogen
Fixation. Nature, 201:310-311.
Ruehle, J. L., and D. H. Marx. 1979. Fiber, Food, Fuel and Fungal Symbionts.
Science, 206(4417):419-422.
Safir, G. R., J. S. Boyer, and J. W. Gerdemann. 1971. Mycorrhizal Enhancement
of Water Transport in Soybean. Science, 172:581-583.
Sanders, F. E., and P. B. Tinker. 1973. Phosphate Flows into Mycorrhizal
Roots. Pest. Sci., 4:385.
Sobotka, A. 1964. Effects of Industrial Emissions on Soil Biology of Norway
Spruce Stands in the Ore Mountains. Lesnicky Casopis, 10:987-1002.
Sokal, R. R., and F. J. Rohlf. 1969. Biometry. The Principles and Practice
of Statistics in Biological Research. W. H. Freeman and Company,
San Francisco, Ca. 776-1- pp.
Sparling, G. P., and P. B. Tinker. 1975. Mycorrhizas in Penine Grassland.
In: Endomycorrhizas, F. E. Sanders, B. Mosse, and B. B. Tinker, eds.
Academic Press, London, pp. 545-560.
134
-------
Van Haut, H. 1961. Staub., 21:52-56.
Zak, B. 1964. Role of Mycorrhizae in Root Disease. Ann. Rev. Phytopath.,
2:377-392.
Ziegler, I. 1975. The Effect of S02 Pollution on Plant Metabolism. Residue
Reviews—Residues of Pesticides and Other Contaminants in the Total
Environment, 56:79-105.
135
-------
SECTION 9
RESPONSES OF TWO LICHEN SPECIES
TO LOW-LEVEL S02
S . Eversman
ABSTRACT
Usnea hirta (L.) Wigg. and Parmelia chloTOchroa Tayl.
exhibited reduced photosynthesis rate, disrupted
chloroplast structure, and increased algal plasmolysis
in 27-92 days in ZAPS I and II plot B. Differences
in respiration rates and relative chlorophyll extract
absorbances were not significant among control U. hirta
samples and U. hivta samples from fumigation plots A
and B.
INTRODUCTION
The ZAPS (Zonal Air Pollution System) SC>2 fumigation systems were
designed to investigate effects of low-level chronic 862 stress on grass-
land ecosystem components. Two lichen species, Usnea hivta (L.) Wigg, and
Pcamel-ia chlorockpoa Tayl., have been transplanted into each ZAPS treatment
plot each spring. Samples have been collected monthly and analyzed for
anatomical and physiological injuries due to S02 stress.
Responses in the lichens collected from medium and high treatment plots
have been acute and well-defined: significantly reduced respiration rates
and relative chlorophyll extract absorbance, and significantly increased
algal plasmolysis (P < .05) within 30-60 days in these plots (Eversman,
1978). In the low treatment it has been somewhat difficult to precisely
identify injuries in lichen samples that are definitely attributable
to S0£ effects. The primary purpose of the 1978 field season activities
was to better distinguish between transplant and S02 effects, and
possible combinations of the two.
136
-------
MATERIALS AND METHODS
Ponderosa pine branches containing U. hivta were transplanted onto one
fencepost in each ZAPS plot. Posts were in the northeast portion of each
plot adjacent to EPA sulfation plates. Three vertical levels of branches were
on each post corresponding to the sulfation plates: 15, 35, and 70 cm above.
the ground. Pannel-La chlovoohvoa thalli were tied onto the branches adjacent
U. hirta thalli.
Transplants were made May 3, June 6, and July 6, 1978, Collections for
analysis were made August 3 from each transplant, and from controls.
Respiration rates were determined manometrically. Algal plasmolysis was
determined with a light microscope. Chlorophyll extracts were made with hot
methanol, and read at 665 nm in a spectrophotometer. These methods have been
described previously (Eversman, 1977, 1978).
Relative photosynthetic rates were determined by exposing a small piece
of thallus (oa. 0.3 mm2) to 1'*C02 for 20 seconds in a field porometer
(Shimshi, 1969; Moser and Nash, 1978). The lichen fragments were combusted,
C02 was collected in trapping fluids, and the radioactivity was determined
with a scintillation counter.
Preparation for transmission electron microscopy included fixation with
glutaraldehyde-osmium tetroxide, embedding in Spurr's resin, and post-
staining with uranyl acetate and lead citrate. A Phillips 201 transmission
electron microscope at Arizona State University was used for electron-
micrographs. All lichen tissue from control, plot A and plot B samples was
taken from green, non-bleached thallus tips.
RESULTS AND DISCUSSION
Differences in respiration rates between U. hirta samples in plots A
and B were not significant after 92 days of S02 treatment (Figure 9.1).
These results are similar to those obtained in previous years (Eversman,
1978). Respiration rates were not determined for P. ohloroohroa in the
ZAPS sites in 1978.
Algal plasmolysis percentages were significantly higher (non-overlapping
0.95 confidence intervals) in U. hirta and P. chloroehroa samples from ZAPS
I, plot B, than in control samples (Figure 9.2). This tissue analysis has
given the most consistent significant differences between plot A and plot B
samples.
Because of large amounts of material required, chlorophyll extract
absorbances were not determined for P. chlorochroa from ZAPS sites. Differ-
ences in U, hirta samples from plots A and B were not distinct (Table 9.1).
Relative chlorophyll extract absorbance values were generally lower in
samples from the ZAPS sites than in samples from native habitats or from
137
-------
'o.
Q"
Ul
ID
Z
0
o
M
0
3.
1200
1100
1000
900
800
700
600
500
400
300
200
100
-
o-t
\ ©=/
n^ A-J
1^^. W~t
\ NW ^ =
; l^tt^^^^
' ~*"' Nw^
X
+
1 -L- 1 1 1 1 1 1 1 1
o=Bo(70cm)
H=Bb(35cm)
Ac •=Bc(l5cm)
- Control (EOC)
X=C
+=D
10 20 30 40 50 60 70 80 90 100
JULY JUNE MAY
Figure 9.1.
Respiration rates of U. hirta from ZAPS I
plots and control samples (EOC = Site P10,
in Section 3, Figure 3.1.). Values are
means of three samples expressed as yl oxy
gen consumed g" h~ . Abscissa = days S02-
transplant sites in the Colstrip area (Section 3). It is not yet clear if
differences are due entirely to SC>2 or to some sun bleaching since the ZAPS
sites are more exposed than the ponderosa pine forests from which U. hirta
was transplanted.
Mean relative photosynthesis rates were less in samples from plot B
than in samples from plot A (Figure 9.3). Comparisons with control sites
P10 (EOC, transplant source) and Pll (SEAM 1) were inconclusive. Large
variances about the means resulted in no statistical differences between any
of the plot A, plot B, or control samples. Photosynthesis rates in plot C
were nearly zero, and they were zero for samples of both species from plot D.
138
-------
O)
o
100
90
80
70
60
O. 50
< 40
_J
< 30
20
I 0
A = Pormelio chlorochroa
A = CONTROL P: chlorochroa
• = Usnea hirlo
O = CONTROL LL hirto
B/
B:
Figure 9.2.
10 20 30 40 50 60 70 80 90 100
DAYS S02
Percentage of algal plasmolysis in U, hirta and P.
chloYodhroa in ZAPS I and II, plots A and B.
A, B = ZAPS I. A', B' = ZAPS II. Control samples
are from transplant sources (Sites P10, G7).
Electron Microscopy
Figures 9.4 and 9.8 illustrate control samples of i/, hirta and P.
chlorockpoaf respectively. Large chloroplasts and small amounts of cytoplasm
are present (Chervin, et al.f 1968; Peveling, 1973). Exposure to S02 on the
ZAPS plots resulted in increasing plasmolysis and internal disorganization:
thylakoid swelling, visible in Figures 9.5, 9.6, and 9.9. Starch grains
in Figures 9.5 and 9.6 may be due to transplant effects on humidity
(Ascaso, 1978), light differences, or S02 (Ikonen and Karenlampi, 1967). It
is not yet clear which factors are causing these changes in the chloroplasts.
Slight bleaching was present in the samples from plot B although only.
green thallus tips were used for electron microscopy. Not all the algal
cells exhibited fine structural injury or starch grains visible in the
chloroplasts as in Figures 9.5 and 9.6; some were nearly identical with
control cells. Some intact and some injured cells would be expected when
compared with plasmolysis, chlorophyll extract, and photosynthesis observa-
tions with similar samples. An increased number of starch grains may
139
-------
Table 9.1. RELATIVE CHLOROPHYLL EXTRACT ABSORBANCE VALUES (665 ran) FOR
USNEA EIRTA, ZAPS I and II, 1978.
June August
Days S02 Mean ± 0.95 CI Days S02 Mean ± 0.95 CI
Control (EOC) 0
II
II
II
II
II
II
I
I
Aa
Ba
Ab
Bb
Ac
Be
Ac
Be
30
30
30
30
30
30
30
30
0.
0.
0.
0.
0.
0.
0.
0.
0.
630 ±
213 ±
173 ±
793 ±
352 ±
586 ±
280 ±
552 ±
590 ±
0.493 Control (EOC) 0
0.102
0.069
0.791
0.162
0.096
0.037
0.291
0.424
I
I
I
I
I
1
Aa
Bb
Aa
Bb-
Aa
Ba
27
27
57
57
92
92
1.24 ± 0.132
0.365
0.445
0.673
0.475
0.241
0.310
± 0.
± 0.
± 0.
± 0.
± 0.
± 0,
221
181
544
587
336
233
Pairs are matched according to height above ground and days of S02 exposure.
a = 70 cm, b = 35 cm, c = 15 cm above ground. Values are given as means of
three samples ± 0.95 confidence intervals. Confidence intervals = t,Q5 x
standard error/ n for each mean.
explain respiration rates considerably higher than control rates. More
detailed comparisons are currently being undertaken with control samples,
other transplants, plot A, and plot B samples to distinguish transplant and
S02 effects. Preliminary electronmicrographs of samples from plot A and
off-ZAPS-plot grassland transplant samples showed less disturbance of
chloroplast and cytoplasm, and fewer starch grains than in samples from
ZAPS plot B. These results are being examined further.
All samples from ZAPS plots C and D have been bleached yellow in 30-60
days of exposure to the medium and higher levels of S02. The complete loss
of protoplasmic structure visible in Figure 9.7 is presumably caused by
S02- The reduced respiration and photosynthesis rates and disordered
internal structure are probably related.
CONCLUSIONS
The levels of S02 in ZAPS plots C and D have been high enough to cause
acute visible and physiological tissue damage in U. hivta and P. ohloTodhroa.
Samples from ZAPS plot B (both I and II) are visibly more bleached than
samples from ZAPS A and samples collected in their native habitats. This
indicates definite S02 injury to the lichens in plot B. The most sensitive
methods of defining the amount of damage appear to be examining the algal
140
-------
cell condition (plasmolysis), bleaching of the thallus, and observing the
fine structure with transmission electron microscopy. Respiration rates and
chlorophyll extract absorbance have been used, and will continue to be used,
but appear to be less sensitive techniques. Respiration is a measure of the
thallus as a whole, which is mostly fungal tissue and presumably not as
sensitive as algal tissue.
3.0
2.5
(0
O
I- H SEAM I
2.0
1.5
0.
I EOC
Usneo hirto
M SEAM I
EOC
28
57
.50
.25
92
Parmelio chlorochroo
28
57
DAYS S02
92
Figure 9.3.
Relative photosynthesis rates of U. hirta and
P. Ghlorochroa from ZAPS I and II. Rates are
expressed as means of three samples (decays per
minute), EOC (Site P10) and SEAM I (Site Pll)
were controls .
141
-------
FIGURE CAPTIONS
Figure 9.4. Algal cell of Usnea hirta from Site P10. W = cell wall.
Cv= chloroplast with thylakoids. Cy = cytoplasm. P =
pyrenoid in chloroplast with dark bodies assumed to be
lipids. Magnification 34.250X.
Figure 9.5. Algal cell of U. hirta after 27 days S02 in ZAPS I plot B,
Increased plasmolysis and starch grains (S) in chloroplast are
visible. N = nucleus with nucleolus. F = fungal hypha.
Lipid bodies in this electronmicrograph stain light-colored;
they are dark in the cytoplasm in Figure 11.4. This is not
a treatment-related phenomenon. Magnification 14,OOOX.
Figure 9.6. U. hirta algal cells after 27 days in ZAPS I plot B. The
thylakoids in the cell on the left are swollen; the cytoplasm
in the cell on the right has become denser and disorganized.
Magnification 40,391X.
Figure 9.7. U. hirto. algal cell after 92 days in ZAPS I plot C. No
internal structure is distinguishable. All algal cells are
bleached and plasmolyzed in the lichen thallus. Magnification
38,400X.
Figure 9.8. Algal cell of Pormelio. ahloroohroo. control sample. P =
pyrenoid with dark lipid bodies. L = light-staining lipid
bodies in cytoplasm. ER = rough endoplasmic reticulum.
F = fungal hypha. M = mitochondrion. Magnification 21,OOOX.
Figure 9.9. P. dhloTodkroa algal cell after 28 days in ZAPS I plot B.
Thylakoids in chloroplast are disrupted. Magnification
41,100X.
142
-------
Figure 9.4
Figure 9.5
143
-------
Figure 9.6.
Figure 9.7.
144
-------
Figure 9.8.
Figure 9.9.
145
-------
REFERENCES
Ascaso, C. 1978. Ultrastructural Modifications in Lichens Induced by
Environmental Humidity. Lichenologist 10: 209-219.
Chervin, R. E., G. E. Baker, and H. R. Hohl. 1968. The Ultrastructure of
Phycobiont and Mycobiont in Two Species of Usnea. Can. J. Bot. 46:
241-245.
Eversman, S. 1977. Effects of Low-Level S0£ Stress on Two Lichen Species.
In: The Bioenvironmental Impact of a Coal-Fired Power Plant, Third
Interim Report, Colstrip, Montana. EPA 600/3-78-021. Environmental
Protection Agency, Corvallis. pp. 385-398.
Eversman, S. 1978. Effects of Low-Level S02 on Usnea hirta and Parmelia
chlorochroa, Bryologist 81(3) :368^-377.
Ikonen, S. and L, Karenlampi, 1967. Physiological and Structural
Changes in Reindeer Lichens Transplanted Around a Sulphite Pulp Mill.
In: Proceedings of the Kuopio Meeting on Plant Damages Caused by Air
Pollution, L. Karenlampi, ed. Kuopio.
Moser, T. J. and T. H. Nash III. 1978. Photosynthetic Patterns of Cetraria
cuoullata (Bell.) Ach. at Anaktuvuk Pass, Alaska. Oecologia (Berl.)
34: 37-43.
Peveling, E. 1973. Fine Structure of Lichens. In: The Lichens,
Ahmadjian, V. and M. Hale, eds. Academic Press, New York. pp. 147-184.
Shimshi, D. 1969. A Rapid Field Method for Measuring Photosynthesis with
Labelled Carbon Dioxide. J. Expt. Bot. 20: 381-401.
146
-------
SECTION 10
SULFUR ACCUMULATION IN WESTERN WHEATGRASS
EXPOSED TO CONTROLLED S02 CONCENTRATIONS
VI. K. Lauenroth, C. J, Bicak, and J. L, Dodd
ABSTRACT
Sulfur concentrations of western wheatgrass tillers
and individual leaves were measured from plants exposed to
four S02 concentrations on ZAPS. Sulfur concentration of
plants was a linear function of either time of exposure or
concentration. Young leaves and the youngest portion of
leaves contained less sulfur than their older counterparts
regardless of whether or not they had been exposed to S00.
Current hypotheses that relate plant sensitivity to amount
of sulfur taken up do not apply for western wheatgrass.
INTRODUCTION
An increase in the plant tissue concentration of sulfur often serves as
an indication of sulfur dioxide exposure. Exposures to high concentrations of
sulfur dioxide often result in low sulfur uptake because of irreversible
damage to the stomatal mechanisms. In contrast, low sulfur dioxide concentra-
tions can result in the accumulation of large amounts of sulfur (Guderian,
1977). Even in severely polluted areas, low atmospheric concentrations of S02.
predominate levels that comprise the measurable S02 concentrations. Conse-
quently, the accumulation of sulfur in plant organs is an easily determined
impact of S02 pollution. The physiological importance of increased sulfur
concentrations and the balance between positive and negative effects will
depend upon the particular species and its rate of growth and metabolism.
Cowling and Locker (1978) demonstrated that low concentrations of S02 (approx-
imately 2 pphm) can correct sulfur deficiency in perennial ryegrass (Lolium
perenne L.)CV S23. In contrast, Bell and Clouth (1973) reported a 50 per cent
reduction in the yield of the same species at a concentration of approximately
7 pphm.
Recently it has been hypothesized that plant sensitivity to S02 is depen-
dent, in part, upon rate of S02 uptake. Garsed and Read (1977) interpreted
the total quantity of S02 taken up by different plant species under identical
147
-------
exposures as an indication of the plant's capacities to avoid the deleterious
effects of S02. Klein et ol. (1978) and Kondo and Sugahara (1978) found that
plant species known to be resistant took up smaller quantities of S02 than
plants known to be susceptible, where sensitive species are characterized by
visible injury criteria.
Sulfur accumulation rates for western wheatgrass (Agropyron smithii
Rydb.)> a species considered resistant to S02 using visible injury criteria
(Heitschmidt et aZ.,1978), are reported in this paper. The data were
collected from both ZAPS I and ZAPS II.
MATERIALS AND METHODS
Plant Analysis
Western wheatgrass plants consist of a number of individual tillers
connected by. a rhizome located at a depth of approximately 5-10 cm in the
soil. Sulfur content of entire plants was assessed by collecting a large
number of individual live tillers from each S02 treatment approximately once
each month during the growing season. Small amounts of dead tissue associated
with each tiller were also included in the analyses. Each sulfur value
reported is the mean of two samples. Sulfur content was measured with a Leco
Induction Furnace (Laboratory Equipment Corporation, St. Joseph, Michigan,
USA) with the inorganic sulfate ion (S0it=)2 the primary form of sulfur recov-
ered.
The sulfur content of western wheatgrass leaves was assessed by col-
lecting the oldest and youngest fully expanded green leaf from a large number
of tillers in the control and high concentration plots in late June 1977.
Each leaf was further separated into a younger and older half to investigate
gradients in sulfur content in the same leaf. Older leaves are located at the
base of the plant, while younger leaves are near the top. The oldest portion
of the leaf is near the tip, while the youngest portion is near the base.
Triplicate samples of each leaf and age were analyzed for sulfur content.
RESULTS AND DISCUSSION
Sulfur content of western wheatgrass plants was positively correlated
with both time of exposure (Table 10.1) and concentration of S02 (Table 10.2).
Sulfur accumulation rates for the May through August growing periods in the
medium treatment ranged from 193-460 yg • g"1 • month"1. Corresponding values
for the high concentration treatment ranged from 385-650 yg • g"1 • month"1.
Accumulation of sulfur by plants in the low treatment was very weakly related
to time of exposure and as expected the slopes of the regression lines were
much smaller than those for plants from either the medium or high treatments.
The differences in the accumulation rates among years as illustrated by the
slopes of the regressions (Table 10.1) may be at least partially explained by
differences in growing season precipitation which largely controls growth rate
and productivity of these grasslands. Since sulfur is accumulated both from
the soil solution and the atmosphere when S02 is present, the concentration
measured within the plant will be determined by the growth rate, the duration
of favorable growing conditions, the concentration of S02, the duration of
148
-------
TABLE 10.1. LINEAR REGRESSION PARAMETERS FOR THE RELATIONSHIP OF
SULFUR ACCUMULATION IN AGROPYRON SMITHII AND TIME OF
EXPOSURE TO THREE S02 CONCENTRATIONS AT TWO SITES IN
THREE GROWING SEASONS
Treatment
Low
Medium
High
Year
1975
1976
1976
1977
1977
1975
1976
1976
1977
1977
1975
1976
1976
1977
1977
Site
I
I
II
I
II
I
I
II
I
II
I
I
II-
I
II
Y intercept*
1350
1000
1300
1388
1523
1150
1200
1050
1487
1607
800
1450
2550
1668
2458
Slopef
0
120
50
18
37
390
400
460
237
193
650
620
560
385
441
r2
0
0.90
0.09
0.03
0.02
0.90
0.92
0.86
0.85
0.72
0.96
0.92
0.98
0.85
0.74
* i
yg • g"1
yg • g"1
• month -1
TABLE 10.2. LINEAR REGRESSION PARAMETERS FOR THE
RELATIONSHIP OF SULFUR CONTENT IN WESTERN
WHEATGRASS EXPOSED TO THREE S02 TREATMENTS
AND CONCENTRATIONS
Site
I
II
Year
1975
1976
1977
1976
1977
Y intercept*
1110
835
1123
799
961
Slope1"
185
277
208
370
277
r2
0.95
0.98
0.97
0.94
0.96
-i
yg • g
yg * pphnr-]
149
-------
exposure, and the concentration in the soil sulfur pool. In both 1975 and
1976 precipitation amounts and distributions favored sulfur uptake compared to
1977.
Sulfur accumulation was also linearly related to treatment concentra-
tion. Average sulfur contents of the control plants ranged from 800 to 1100
yg ' g"1. Thomas et dl. (1950) surveyed 29 trees, shrubs, and herbs throughout
the western United States and found total sulfur contents ranging from 400
yg • g"1 in conifer needles to 35100 yg • g"1 in a halophytic herb (Sueda
spp.). Sulfur accumulation rates increased in the presence of S02 at the rate
of 185 to 370 g • g for each pphm of SC>2 to which the plants were exposed
during the growing season. The largest differences in incremental accumula-
tion rates occurred between sites. At Site I the rates varied by 92 g • g~ •
pphm and indicated no relationship to the wet years (1975 and 1976).
Rates were higher at Site II but demonstrated the same trend between years in
that 1976 had the largest rate.
Thomas et al. (1950) hypothesized that after entry into the leaf, S02 is
rapidly dissolved on most surfaces to form sulfite, which is then slowly
oxidized to sulfate. Thomas and Hendricks (1956) found sulfate was much less
toxic to plants than sulfite. It has been assummed that this oxidation step
was an important determinant of whether a plant would be damaged by a given
exposure to S02. However, Malhotra and Hocking (1976) have found that toxic
levels of sulfate are eventually attained, leading to injury symptoms. This
explanation was used by Heitschmidt et at. (1978) to explain increased leaf
senescence in western wheatgrass as a result of low-concentration exposures.
Because senescence begins at the leaf tip and spreads downward, this implies
that sulfate concentrations accumulate with the same pattern. The data in
Table 10.3 verify that there are substantial gradients between the tips and
bases of individual leaves and between leaves of different ages. JSger (1976)
reported the same trends in spruce needles exposed to S02. In addition, he
TABLE 10.3. SULFUR CONTENTS* OF TWO DIFFERENT LEAF AND TISSUE
AGES OF PLANTS GROWING IN THE CONTROL AND HIGH-
CONCENTRATION TREATMENT PLOTS
Leaf age
Young Old
Treatment Base Tip Base Tip
Control 830 ± 60 1117 + 25 980 ± 60 1197 ± 95
High 1810 ± 121 2643 ± 186 3883 ± 6 5010 ± 144
__
X ± SD in yg • g 1
150
-------
found gradients between tips and bases of control plants, which agrees with
our results. Guderian (1977) reported that secondary transport of sulfur
taken up as S02 results in highest concentrations in leaf tips and margins.
Both Jager (1976) and Guderian (1977) cite Halbwachs (1963) to explain that
this transport is affected by sulfate moving with water and the existence of a
suction gradient favoring movement towards leaf tips and margins.
CONCLUSIONS
The degree of S02 uptake has been suggested as an important factor in
determining sensitivity (Garsed and Read, 1977; Klein et al., 1978). The
limitations imposed by so simplistic an explanation relate largely to the
meaning of sensitivity and its interactions with dose. The often-used
synonymy of sensitivity and visible leaf injury places restrictions upon
interpretations of responses to low-concentration exposures that may result in
important physiological responses with no visible injury. Western wheatgrass
is shown to accumulate significant quantities of sulfur when exposed to S02,
yet we have never found symptoms of S02 injury to leaves. We have found
increased leaf senescence (Heitschmidt et at., 1978) and chlorophyll destruc-
tion (Lauenroth and Dodd, in prep.) as a result of exposures to low-S02
concentrations.
REFERENCES
Bell, J. S., and W. S. dough. 1973. Depression of Yield in Ryegrass Ex-
posed to Sulphur Dioxide. Nature (Lond.)> 241:47-49.
Cowling, D. W., and D. R. Lockyer. 1978. The Effect of S02 on Lolium perenne
L. Grown at Different Levels of Sulphur and Nitrogen Nutrition. J. Exp.
Bot., 29:257-265.
Garsed, S. G., and D. J. Read. 1977. The Uptake and Metabolism of 35S02 in
Plants of Differing Sensitivity to Sulphur Dioxide. Environ. Pollut.,
13:173-186.
Guderian, R. 1977. Air Pollution. Springer-Verlag, Berlin. 127 pp.
Halbwachs, G. 1963. Utersuchnunger uber gerichete aktive strbmungen and
stofftransporte in blatt. Flora, 153:333-357.
Heitschmidt, R. K., W. K. Lauenroth, and J. L. Dodd. 1978. Effects of
Controlled Sulphur Dioxide Fumigation on Western Wheatgrass in a South-
eastern Montana Grassland. J. Appl. Ecol., 14:693-702.
Jager, H. J. 1976. S-Lokalisation in S02-begasten Fichtennadeln. Eur. J.
For. Pathol., 6:25-29.
Klein, H., H. J. Jager, W. Domes, and C. H. Wong. 1978. Mechanisms Contri-
buting to Differential Sensitivities of Plants to S02. Oecologia,
33:203-208.
151
-------
Kondo, N., and K. Sugahara. 1978. Changes in Transpiration Rate of S02
Fumigation and the Participation of Abscisic Acid. Plant Cell Physiol.,
19:365-373.
Lauenroth, W. K., and J. L. Dodd. (In prep.) Chlorophyll Reduction in
Western Wheatgrass Exposed to Sulfur Dioxide.
Malhotra, S. S., and D. Hocking. 1976. Biochemical and Cytological Effects
of Sulphur Dioxide on Plant Metabolism. New Phytol., 76:227-237.
Thomas, M. D., and R. H. Hendricks. 1956. Effects of Air Pollution on
Plants, In: Air Pollution Handbook, P. L. Magill, F. Holden and C.
Ackley, eds. McGraw-Hill Book Co., Inc., New York. pp. 9.1-9.44.
Thomas, M. D., R. H. Hendricks, and G. R. Hill. 1950. Sulfur Metabolism of
Plants. Ind. Eng. Chem., 42:2231-2235.
152
-------
SECTION 11
WEIGHT AND GERMINATION RESPONSES
OF GRASS SEEDS FROM PARENTAL STOCK
SUBJECTED TO SULFUR DIOXIDE FUMIGATION
P. M. Rice, C. C. Gordon, and P. C. Tourangeau
ABSTRACT
In 1977, seeds were collected from three perennial
grass species and one weedy biennial forb which had
been subjected to chronic sulfur dioxide fumigation at
ZAPS I and II. These seeds were then incubated without
further sulfur dioxide treatment in growth chambers to
assess the consequences of stressing the parental stock
on various germination responses and seed weight.
Edaphic factors at the site-I control (ZAPS IA) may be
obscuring potential treatment effects. Results from
site II provide evidence of a depression of seed
responses in three of the four species studied.
Responses do not change incrementally with changes in
fumigation levels, thus other factors or experimental
procedures may be influencing experimental results.
Germination trials to be conducted on 1978-79 seed
collections will incorporate procedural alterations to
reduce the potential effect of these other factors.
INTRODUCTION
Community structure and composition can be altered by air pollution-
induced stress (Winner and Bewley, 1978; Farrar et al., 1977; Westwood, 1979)
at ambient concentrations which do not directly cause mortality to established
plants. These changes in community composition can be viewed as secondary
effects resulting from air pollution influencing plant competitive relation-
ships (Bennett and Runeckles, 1977). The importance of seed factors and
responses in maintaining.community diversity and stability were recognized by
early grassland ecologists (Davies, 1927; Blake, 1935; McAlister, 1943).
That seed parameters such as weight and germination capacity can be altered by
air pollutants has been demonstrated by several workers (Houston and Dochinger,
1977; Kohut et al. , 1978).
This portion of the ZAPS work concerns the development of methods and the
assessment of sulfur dioxide influence on the seed characteristics of four
153
-------
major Northern Great Plains grassland species, Poa sandbergii (native bluegrass),
Koeleria eristata (prairie junegrass), Stipa viridula (green needlegrass), and
Tragopogon dubius (salsify). The major parameters examined were:
Germination Capacity^—the percentage of "experimental seeds" exhibiting
2 mm or more of root or shoot growth under growth chamber conditions
adjudged suitable for that species.
Seedfill—the percentage of "experimental seeds" which contain an endo-
sperm.
Germination Capacity of Filled Seeds—the percentage of filled seeds
exhibiting successful (>2 mm) root or shoot emergence.
Weight per Seed—average weight per "experimental seed" equilibrated
to ambient laboratory conditions.
Coefficient of Rate of Germination (CRG)—an index of how rapidly seeds
that exhibit successful germination did germinate.
The CRG was adopted from Maguire (1962) and Young and Evans (1977). It
can be expressed as:
n
CRG = Z gn " g(n-l)
i n
where: gn = accumulated germination percentage on a given day
S(n-l) = accumulated germination percentage on the previous day
n = number of days incubation
The larger the CRG, the faster the germination response. Additionally, it is
of interest to examine certain characteristics of the germination response
curve as a function of time. These include the number of days to 50% total
germination and the day of peak germination. Earlier work (Rice et al, , 1979)
suggested that these various seed responses were potentially influenced by
sulfur dioxide treatment.
MATERIALS AND METHODS
In 1977, seed heads were clipped from a minimum of ten subplots per treat-
ment plot and combined to form a composite for the respective plots. The
material was stored for one year to allow after-ripening. Seeds were then
hand-stripped from the heads. Replicates (ten per plot) were obtained by
apportioning the samples by weight. This is done to standardize the total
tissue mass in the trial, and it is intended to reduce within sample variance
and bias arising from visual selection of seed if direct counts are made from
a limited total sample. When the total weight of seeds from any treatment
plot was limited, the available material was apportioned among ten replicates.
When larger seed volumes were available, enough seeds to obtain a standard
154
-------
weight was randomly selected from the sample. In both cases, the actual
counted number of "experimental seeds" for each replicate was determined after
completion of the incubation period.
Germination tests were continued for 30- or 45-day periods, depending on
the species. Fungal and bacterial growth were restricted by dry dusting seeds
with 50% Ortho Captan and using an Ortho solution (two teaspoons per liter) as
the water supply for germination. The petri dishes and cellulose pads which
supported the seeds were autoclaved (15 pounds-15 minutes). Growth chamber
conditions are described in an earlier report (Rice et al. , 1979). Replicate
positions in the chambers were randomized daily after the germination counts
were made. The number of seeds germinating each day was recorded, and at the
end of the trial period the number of ungerminated seeds was determined.
Ungerminated seeds were tabulated as those containing an endosperm and those
with an empty palea. Seeds which germinated or did not germinate but contained
an endosperm were considered "filled seeds."
Growth chamber size and the daily counting task precluded running all ten
replicates at the same time. A randomized complete block design was used to
remove the contribution to error mean square resulting from staggering the
replicates within the trial. Dunnett's (1955) procedure was used to compare
the Off Plot Control 1 (OPC 1), Fifteen Mile Creek Control (15 Mi), and site
I or site II treatment plots with the respective site I or II A plots. The
daily germination counts were also used to compute the relative and cumulative
relative frequency distributions of germination as a function of time. These
curves were used to estimate the number of days for 50% germination and the day
of peak germination. These two responses were examined by Friedman's (1937)
nonparametric two-way analysis of variance. If the Friedman statistic indi-
cated significance (p £ .05), the plot with the most extreme rank sum was
removed from the data set and another Friedman's test conducted until the set
was nonsignificant. Conover (1971) cautions that the a calculated with
repeated application of Friedman's test is not exact.
RESULTS
The error term for seed weight and each major response parameter deter-
mined in the germination trials is presented in Table 11.1. The comparisons
of the various treatment plots with the respective A plots are portrayed in
Tables 11.2 to 11.5 for each of the four species.
All five response variables for Poa sandbergii (Table 11.2) from ZAPS IA
were lower than those from OPC 1. ZAPS IA germination responses were also
lower than the ZAPS IB, 1C and ID treatment. Seed weight between site I plots
and ZAPS IA did not vary significantly. At site II, Germination Capacity,
Seedfill, and Germination Capacity of Filled Seeds were lower on ZAPS IIB and
IID than on the ZAPS IIA plot. Weight per Seed was also significantly lower
on the ZAPS IIB and IID plots.
All four germination responses in Koeleria or>istata from ZAPS IA were
depressed relative to the OPC 1. This pattern of a depressed ZAPS IA plot is
similar to that observed in P. sandbevgii', however, the K. cr-istata experi-
mental seeds were heavier on ZAPS IA, with weight declining on the higher
155
-------
TABLE 11.1. 1977 SEED RESPONSE ERROR MEAN SQUARES*
Poa
aandbergii
Kaeleria
aristata
Stipa
viridula
Tragopogon
df
Error Mean
8,70 Square
Error Mean
8,72 Square
Error Mean
5,42 Square
Error Mean
8,68 Square
Coefficient
of Rate of
Germination
1.377
0.734
3.304
0.965
Germi-
nation
Capacityt
3.481
3.212
10.534
6.617
Germination
Capacity of
Seedflllt Filled Seedt
3.964 57.117
3.501 25.695
N.D.t N.D.
N.D. N.D.
Weight per
Seed
(mg)
5.00 x 10-*
6.63 x 10"s
2.64 x ID'2
2.69 x 10"2
'Derived from a randomized complete block design. All treatment effects measured were highly
significant (p < .001).
'Values expressed as angular (arcsin) transform of percentage.
TN.D. » Not Determined.
TABLE 11.2. POA SANDBERGII SEED RESPONSES COMPARED TO THE
A PLOTt AT EACH SITE
Response
Coefficient of 'Rate of Germination
Germination Capacltyt
Seedflllf
Germination Capacity of Filled Seeds:):
Weight per Seed (mg)
Coefficient of Rate of Germination
Germination Capacity^
Seefill^
Germination Capacity of Filled Seed 8^
Weight per Seed (mg)
OPC 1
9.468*
20.8 **
24.1 **
61.1, **§
0.265*
OPC 1
9.468
20.8
24.1
61.1
0.265
ZAPS LA
8.073
5.2
9.0
34.3
0.236
ZAPS IIA
10.421
20.0
22.9
61.5
0.251
ZAPS IB
8.551
7.8 *
11.4 *
44.8 *
0.225
ZAPS IIB
9.234, §
6.1, **§
8.5 **
47.9, **§
0.204**
ZAPS 1C
10.485**
19.1 **
23.2 **
56.3 **
0.255
ZAPS IIC
10.530
18.5
21.5
60.2
0.233
ZAPS ID
10.683**
18.3 **
22.0 **
57.6 **
0.247
ZAPS IID
10.646
8.5 **
12.5 **
43.9 **
0.192**
*Signlflcant, p <. .05.
**Hlghly significant, p <. .01.
tDunnett's procedure for comparing several treatments to a "control."
^Values are expressed as the angular (arcsin) transform of the percentage.
§Ten items per sample, except where subscript indicates otherwise (X9),
156
-------
TABLE 11.3. KOELERIA CRISTATA SEED RESPONSES COMPARED TO
THE A PLOTt AT EACH SITE
Response
Coefficient of Rate of Germination
Germination Capacity*
Seedfill*
Germination Capacity of Filled Seeds*
Weight per Seed (mg)
Coefficient of Rate of Germination
Germination Capacity*
Seed fill*
Germination Capacity of Filled Seeds*
Weight per Seed (mg)
OPC 1
15.145**
19.5 **
20.8 **
70.2 **
0.133**
OPC 1
15.145
19.5
20.8
70.2
0.133*
ZAPS IA
13.904
11.0
12.8
61.0
0.161
ZAPS IIA
16.212
29.3
30.5
75.0
0.151
ZAPS IB
14.177
8.1 **
10.8
49.6 **
0.144**
ZAPS IIB
15.927
23.0
24.5
70.6
0.150
ZAPS 1C
14.052
12.5
14.0
64.1
0.124**
ZAPS IIC
15.993
32.3
33.2
77.9
0.153
ZAPS ID
14.816
22.5 **
23.8 **
71.4 **
0.146**
ZAPS IID
16.180
33.5
34.8
75.4
0.176**
*Signlficantf p <_ .05.
**Hlghly significant, p £ .01.
tDunnett's procedure for comparing several treatments to a "control."
tValues are expressed as the angular (arcsln) transform of the percentage.
TABLE 11.4. STIPA VIRIDULA SEED RESPONSES COMPARED TO THE
A PLOTt AT EACH SITE
Response 15 Hi OPC 1 ZAPS IIA ZAPS IIB ZAPS IIC ZAPS IID
Coefficient of Rate of Germination 13.892 13.179 11.868 11.574 10.669,5 8.876,**§
Germination Capacity^ 14.7 15.1 17.3 17.4 10.1 ** 9.1 **
Weight per Seed (mg) 2.801 2.477* 2.681 2.885* 2.875 * 2.550
*Signifleant, p £ .05.
**Highly significant, p £ .01.
tDunnett's procedure for comparing several treatments to a "control."
rvalues are expressed as the angular (arcsin) transform of the percentage.
§Ten items per sample, except where subscript Indicates otherwise (X»).
157
-------
TABLE 11.5. TMGOPOGON DUBIUS SEED RESPONSES COMPARED TO
THE A PLOTt AT EACH SITE
Response 15 Ml ZAPS IA ZAPS IB ZAPS 1C ZAPS ID
Coefficient of Rate of Germination 31.735»*§ 30.3749§ 30.480 30.318 29.903
Germination Capacity? 80.9 ** 86.8 82.6 ** 84.6 81.8 **
Weight per Seed (mg) 6.74 ** 6.14,§ 5.52,**§ 5.62 ** 5.38 **
15 Mi ZAPS IIA ZAPS IIB ZAPS IIC ZAPS IID
Coefficient of Rate of Germination 31.7359§ 32.58995 32.670 31.696 31.309*
Germination CapacityT 80.9 ** 86.2 83.2 * 83.4 86.0
Weight per Seed (mg) 6.74 * 6.32 6.19 6.00 ** 5.98 *
*Signiflcant, p < .05.
"Highly significant, p <_ .01.
tDunnett's procedure for comparing several treatments to a "control."
rvalues are expressed as the angular (arcsin) transform of the percentage.
§Ten items per sample, except where subscript Indicates otherwise (X9).
treatment levels on site I. Germination Capacity and Germination Capacity of
Filled Seeds were depressed on ZAPS IB but elevated relative to ZAPS IA on the
ID plot. Seedfill was also higher on ZAPS ID. Germination parameters were
similar on all plots at site II. The low K. cvistata. seed weight found in the
OPC 1 sample was still significant compared with the site IIA plot. Addition-
ally, the heaviest experimental seed was derived from the ZAPS IID plot sample.
Stipa viridula was collected from site II and two plots (OPC 1 and 15 Mi)
located away from either of the fumigation sites. The Coefficient of Rate of
Germination was lowest for seeds from ZAPS IID. Germination Capacity was
depressed for both the ZAPS IIC and IID seeds. Seeds from OPC 1 were lightest,
while those from ZAPS IIB and IIC were heavier than the site II control.
Tragopogon dubius from 15 Mi germinated faster and was heavier than that
from the ZAPS IA plot (Table 11.5). However, the Germination Capacity was
slightly higher (99...7% vs. 97.5%, back transform of the arcsin) for the ZAPS IA
sample. In seeds from ZAPS IB and ID, Germination Capacity was reduced, while
seed weight was lower on site I plots relative to the control. The slight
depression in Germination Capacity observed for 15 Mi T. dubius seed was also
significant relative to the ZAPS IIA plot (97.5% on 15 Mi vs. 99.6% on ZAPS IIA
back transform). Seed weight remains highest for the 15 Mi seeds. Weight per
Seed and the Coefficient of Rate of Germination were depressed on ZAPS IID
relative to ZAPS IIA. Seed weight was also depressed on ZAPS IIC and germi-
nation capacity on ZAPS IIB.
158
-------
The relative frequency distributions of total germination success for
each day of incubation are portrayed graphically for each species by treatment
plot in Figures 11.1 through 11.4. These graphs are formed from the pooled
data for all replicate trials. The day of peak germination can be seen on
these histograms and days to 50% germination is approximately indicated by the
arrow. Day of peak germination of the individual replicates was analyzed by
Friedman's test (Table 11.6). T. dubius was not treated in this analysis, as
86 of the 90 cases peaked on day three. The only significant difference
detected was the high rank sum representing the delayed peak day for the
Poa sandbergii sample from ZAPS IIB. The pooled responses for this sample
suggest a bimodality (Figure 11.1) which may have led to an overestimate of
peak for this species.
The relative cumulative frequency distributions as a function of the
natural log of the number of days of incubation are portrayed in Figures 11.5
through 11.8. These graphs are also derived from the pooled data for each
species and treatment plot. The time required for germination of 50% of the
seeds was estimated by linear interpolation of the counts for the day preceed-
ing and following this event for each individual replicate. This variable
was analyzed by Friedman's test and the results are indicated in Table 11.7.
Comparisons for a species are made within a site and include either OPC 1 or
15 Mi when the sample was available. The higher rank sums within a column in-
dicate longer time periods for 50% germination response. For Poa sandbergii
at site I, the ZAPS IA and IB samples were delayed in reaching 50% germination
TABLE 11.6. DAY OF PEAK GERMINATIONt
Plot Rank Sum
Poa eandbergii
OPC 1 31.5
ZAPS IA 38.5
ZAPS IB 32.5
ZAPS 1C 23.0
ZAPS ID 24.5
Koeleria cristata
OPC 1 21.0
ZAPS IA 29.0
ZAPS IB 31.5
ZAPS 1C 36.0
ZAPS ID 32.5
Stipa viridula
Mean
9.4
11.4
9.9
8.8
8.5
5.8
6.3
6.4
6.4
6.3
Plot
OPC 1
ZAPS IIA
ZAPS IIB
ZAPS IIC
ZAPS I ID
OPC 1
ZAPS IIA
ZAPS IIB
ZAPS IIC
ZAPS I ID
15 Mi
OPC 1
ZAPS IIA
ZAPS IIB
ZAPS IIC
ZAPS I ID
Rank Sum
31.5
24.0
43.0
28.5
23.0
32.0
27.0
29.5
27.0
34.5
25.5
29.0
38.0
42.5
43.5
31.5
Mean
9.4
8.6
10.6*
8.8
9.4
5.8
5.6
5.7
5.6
5.9
5.7
6.7
9.2
12.0
8.1
10.1
•Significant, p <_ .05.
tFriedman's test.
159
-------
>20
z
Ul
3
O
Ul
IE
U.
Ul
IE
IO 20 3O
DAYS INCUBATION
ZAPS I B
x-zoi
u
z
FREQUE
o
Ul
>
k—
S
UJ
" o-
/I
1
ZAPS 1 C
'~*n
LI
S
L
1 ,_
. 20
10 20 3O
DAYS INCUBATION
V201
U
z
Ul
3
O
UJ
X
* 10.
Ul
>
t_
5
_l
UJ
' 0-
1
Jl
L
-
ZAPS II B
•^
I
Till
»* 10
n
20 30
DAYS INCUBATION
ZAPS II C
IO 20
DAYS INCUBATION
3O
IO 20 30
DAYS INCUBATION
10 20
DAYS INCUBATION
DAYS INCUBATION
Figure 11.1.
Histogram of successfully germinating Poa sandbevgii, seeds on OPC 1 and ZAPS I and II,
\ indicates days to 50% germination.
-------
40
£jo
o
u
(r
K>
OPC 1
ZAPS 1 A
10 20 30 0 K>
DAYS INCUBATION
n n
ZAPS IB
20 30 0 10 20
DAYS INCUBATION
_P
X
40-
D30-
o
IO-
ZAPS 1C
„ n
ZAPS 1 D
20 30 0 10
DAYS INCUBATION
ZAPS II A
20 30 0 10
DAYS INCUBATION
K 30
40
10
ZAPS II B
ZAPS II C
ZAPS II D
to 30 0 10
DAYS INCUBATION
10 30 0 10
DAYS INCUBATION
tO 30
Figure 11.2.
Histogram of successfully germinating Koeleria. oristata seeds on
OPC 1 and ZAPS I and II. 4- indicates days to 50% germination.
161
-------
o
z
IU
IE
U.
IU
>
§
10
15 Mi.
in
S
IE
U.
10
lu
(C
20 30
DAYS INCUBATION
OP C I
20i
lu
3
S
IE
U.
IU
>
IU
IE
IO
j^n. n
4O
20 30 40
DAYS INCUBATION
ZAPS II A
50
50
p n n
Finn
2O 30 40
DAYS INCUBATION
50
10
10
10
ZAPS II B
in." n n\[
40 SO
20 30 40
DAYS INCUBATION
ZAPS II C
10 20 30 40
DAYS INCUBATION
20 30 40
DAYS INCUBATION
nq
50
2 On
5
Z
111
8
IE
"- io-
UJ
>
f—
4
_l
IE
- oJ
I ZAPS II D
r
j
_r
nr
ir
n n
SO
Figure 11.3. Histogram of successfully germinating Stipa viridula seeds on 15 Mi, OPC 1
and ZAPS II. -I- indicates days to 50% germination.
-------
[
lu
O
u
K
Ik.
ui 20-
_l
IJ
c
** 10
•V
rp
0""
IS Mi ZAPS 1 A ZAPS IB
50
20-
T
10-
1
30
20-
K>-
I
6 10 2O 6 10 20 (
L
l
> 10 20
DAYS INCUBATION DAYS INCUBATION
ZAPS 1 C _.. ZAPS 1 0
O
ZJO-
bj
3
O
U
K
h.
U20
_f
u
or
" ,0-
O
1
20
10'
I.
L
30-
2O
10
.S . n O
- 92%
ZAPS II A
s
1.
10 20 0
DAYS INCUBATION
10
20 0 10
DAYS INCUBATION
20
U20
ZAPS II B
SO
20
ZAPS II C
10 20 O
DAYS INCUBATION
10
30
2O
10
ZAPS II D
20 0 10
DAYS INCUBATION
20
Figure 11.4.
Histogram of successfully germinating Tragopogon
dubius seeds on 15 Mi and ZAPS I and II.
163
-------
98
093-
z
UJ
o90
UJ
c
u.
_
§60'
030-
£40-
30-
10
00
OPC I
o ZAPS I A
a ZAPS I B
A ZAPS I C
O ZAPS I D
—I 1 1
9 10 II
LN DAYS
-\ 1 1 1—
13 IS
INCUBATION
20
98-
95-
o
Z90
in
70
UJ
reo
z ^
o
30-
_i
bi
20-
10-
3 -
X OPC I
o ZAPS II A
O ZAPS II B
A ZAPS IIC
O ZAPS II D
8 9 10
LN DAYS INCUBATION
13
20
23
Figure 11.5.
Cumulative germination of Poo. sandbergii seeds
from ZAPS I and II as a function of time.
164
-------
98
Sao-
cr
"•70
u.60
>SO
440
20
X OPC I
o ZAPS I A
O ZAPS I B
A ZAPS I C
O ZAPS I 0
6 7
LN DAYS
6 9
INCUBATION
10
12 13 14 15
99
98
O 95
UJ
3 90
O
UJ
£80^
70-
UJ
> 60
?30
o
20
ui
>
10
5-
2
H
X OPC I
o ZAPS II A
a ZAPS II B
A ZAPS II C
ZAPS II 0
9 10
INCUBATION
12 13 14 IS
Figure 11.6.
Cumulative germination of Koeleria cristata seeds
from ZAPS I and II as a function of time.
165
-------
96
93-
>•
o
5 9°
3
a
E 80
70
ui
5-H
30-
o-o—o-o-o-o-o
X—x-A-A _^o—o ,0-0-0
8^VX e^F
IS Mi
OPC I
ZAPS II A
ZAPS II B
ZAPS II C
ZAPS II D
7 8 9 10
LN DAYS INCUBATION
IS
20
Figure 11.7. Cumulative germination of S'tipa viridula seeds
from ZAPS II as a function of time.
166
-------
15 Mi
O ZAPS I A
D ZAPS I B
A ZAPS I C
O ZAPS I D
99.95
99.9
99.8
99.5
>- =
O
§98
u
F95
=>
u
90
b>
>
80
TO
IS Mi
ZAPS II A
ZAPS II B
ZAPS II C
ZAPSII D
INCUBATION
4 5
LN DAYS INCUBATION
Figure 11.8. Cumulative germination of Tragopogon dubius seeds from ZAPS I and II as a function of time.
-------
response. In the site II contrast, the OPC 1 sample was slower than the
site II sample. Koeleria cristata from site I exhibited an earlier 50%
response in the OPC 1 and ZAPS ID samples. Site II seeds were faster than the
OPC 1 seeds but exhibited no difference between site II samples themselves.
Stipa viridula reached 50% germination most rapidly in seeds from 15 Mi and
OPC 1, while ZAPS IID was significantly slower. Tragopogon dubius seeds from
OPC 1 reached 50% germination before any of the site I seeds. From site II,
the ZAPS IIA and IIB seeds were half done before seeds from any of the other
treatments.
TABLE 11.7. DAYS TO 50% GERMINATIONt
Plot Rank Sum
Poa sandbergii
OPC 1 29.5
ZAPS IA 40.5
ZAPS IB 38.0
ZAPS 1C 20.0
ZAPS ID 22.0
Koeleria aristata
OPC 1 18.0
ZAPS IA 38.0
ZAPS IB 34.0
ZAPS 1C 39.0
ZAPS ID 21.0
Trajopoycn dubius
15 Mi 18.5
ZAPS IA 35.0
ZAPS IB 28.0
ZAPS 1C 29.5
ZAPS ID 39.0
Stipa vindula
Mean
10.16
12.25*
11.48*
9.48
8.87
5.85*
6.60
6.41
6.27
5.92*
2.62*
2.82
2.74
2.81
2.84
Plot
OPC 1
ZAPS IIA
ZAPS IIB
ZAPS IIC
ZAPS IID
OPC 1
ZAPS IIA
ZAPS IIB
ZAPS IIC
ZAPS IID
15 Ml
ZAPS IIA
ZAPS IIB
ZAPS IIC
ZAPS IID
15 Ml
OPC 1
ZAPS IIA
ZAPS IIB
ZAPS IIC
ZAPS IID
Rank Sum
39.0
24.5
37.0
28.5
21.0
44.5
23.5
24.5
30.5
27.0
38.5
17.5
17.5
37.5
39.0
17.5
23.0
35.5
40.5
39.5
54.0
Mean
10.16*
9.03
10.32
9.12
8.76
5.85*
5.50
5.52
5.55
5.55
2.62
2.55*
2.55*
2.62
2.66
6.69*
7.21*
8.53
9.18
8.75
16.48*
*Slgniflcant, p <_ .05.
tFrledman's test.
DISCUSSION
Seed parameters for all significant responses except seed weight in
K. cristata and a slight difference in germination capacity for T. dubius
demonstrate that seed viability and vigor on the ZAPS IA "control" is
depressed relative to the control plots (OPC 1 and 15 Mi) removed from the
ZAPS I site. Increases in germination response parameters for P. sandbergii
and K. cristata at ZAPS IB, 1C and ID thus may well reflect a relative depres-
sion in vigor of plants on the ZAPS IA plot. We concluded that edaphic condi-
tions at ZAPS IA are relatively harsh, and this is also reflected in the
apparent difference in community structure as evident by the presence of
Festuca idahoensis on the plot and the proximity of the Artemisia tridentata
stands. This discontinuity may extend into the ZAPS IB treatment plot.
168
-------
The weedy biennial, T. dubius, responded with reduced germination capacity
as the sulfur dioxide level increased on site I. Although this reduction in
Germination Capacity was significant, the magnitude of the back transformation
difference (<2%) is quite small compared to total germination success which
ranges from 97% to 99%, and shifts of this relative magnitude may be too minor
to have much effect on seedling competition between species.
At site II, Germination Capacity, Seedfill, and Germination Capacity for
P. sandbergii were reduced at ZAPS IIB and IID levels. The ZAPS IIC plot was
similar to IIA. This relative vitality of P. sandbergii seed produced on the
ZAPS IIC plot was also evident in the 1976 trials (Rice et at., 1979) and
appears to be sufficient to overwhelm any treatment effect. For K. cristate,
at site II, germination responses were unaffected. The reduction in Germi-
nation Capacity for 5. viridula on site II plots C and D parallels a similar
but nonsignificant trend observed in 1976. There was decreased Germination
Capacity response with increasing treatment level and a reduction in the
Coefficient of Rate of Germination on ZAPS IID. The 50% germination point was
delayed for S. viridula from ZAPS IID. Germination success for 5. viridula
from all plots (untransformed % = 6.4) remains an order of magnitude below
that possible (83%) (Eddleman, 1977) for this species under the germination
conditions employed. The characteristic dormancy reported of this species
(McAlister, 1943; Rogler, 1960) is not being broken by the procedures currently
employed. Cold water stratification will be employed on the 1978 collections
to bring the total germination closer to its potential. For T. dubius at
site II, the Coefficient of Rate of Germination tends to decline with
increasing sulfur dioxide treatment, the decline being significant on the
ZAPS IID plot, 50% of the germination response being reached sooner on ZAPS IIA
and IIB than IIC and IID. The CRG's are so high (30+) for T. dubius compared
to the perennial grasses (8 to 16) that it is unlikely the small decrease on
the ZAPS IID plot would alone alter the competitive relationship between this
weed and the grasses.
If the sulfur dioxide treatments are stressing the entire plant, it is
reasonable to assume that available photosynthate would be prioritized for
maintaining the vegetative top growth. Depressed fruit and/or seed yields for
gas-stressed plants are frequently reported (Houston and Dochinger, 1977; Pack
and Sulzback, 1976; Oshima et al., 1977). Average seed weights were contrasted
to their respective site I or II controls (A plots) for 21 cases, each case
being a specific combination of species and intentional sulfur dioxide fumi-
gation level (B, C or D plot treatment). No difference could be discerned in
8 cases, 3 cases suggested heavier seeds, while 10 cases indicated a. signi-
ficant (p _< .05) decline in seed weight with sulfur dioxide treatment.
It was not determined what proportion of the depressed seed weights are
a reflection of smaller seeds or due to lack of fill or seed set. Seed weight
was derived from the sample weight divided by the total number of experimental
seeds. This included both filled and unfilled "seed." Among the perennial
grass seed collection for 1977, the average weight from each of the plots is
less than the lower range of seed weights for the 1976 collections (Rice et al.,
1979). For the biennial T. dubius, the 1977 average weight range (5.38 to
6.76 mg/seed) overlaps the lower values for the 1976 weight range (6.04 to
7.58 mg/seed). Two factors potentially contribute to this general decline in
seed weight. First, available soil moisture and precipitation was lower
(Dodd et al., 1979) during the seed set and development period (May, June,
July) in 1977 than 1976. The moisture conditions for this period in 1977 were
169
-------
below the long-term averages. Second, the differences in seed head cleaning
procedures may have increased the proportion of unfilled seed in the total
experimental seed for the 1977 trials. Cleaning procedures for the 1978-79
collections will be more closely supervised and the experimental procedure
altered to determine seed fill and seed weights prior to the actual germination
trial.
CONCLUSIONS
The ZAPS IA plot appears to provide a poor habitat for seed production.
This makes interpretation of site I seed responses difficult in terms of the
single factor effect of sulfur dioxide treatment. A form of multi-variate
analysis would be of value here but is beyond the sampling resources of the
experimenter. Site II provides evidence of the depressing effect of sulfur
dioxide on germination in three of the four species studied. Seed weights
generally decline with increasing fumigation levels at both site I and II.
Although responses are generally depressed with increasing gas treatment, the
data does not suggest a response linear to the treatment gradient. Thus, the
depressions may reflect edaphic factors or reflect inherent differences in
the plant population across the sites. Procedural progress was made with the
1977 trials and will be adapted for testing the 1978-79 seed collections.
REFERENCES
Bennett, J. P., and V. C. Runeckles. 1977. Effects of Low Levels of Ozone
on Plant Competition. J. Appl. Ecol., 14(3):877-880.
Blake, A. K. 1935. Viability and Germination of Seeds and Early Life History
of Prairie Plants. Ecol. Monog., 5:407-460.
Conover, W. J. 1971. Practical Nonparametric Statistics. John Wiley and
Sons, Inc., New York. 462 pp.
Davies. W. 1927. Seed Mixture Problems: Soil Germination, Seedling and
Plant Establishment with Particular Reference to the Effects of Environ-
mental and Agronomic Factors II. Field Trials. Welsh Plant Breeding
Station. Bui. Series H. No. 6. pp. 39-60.
Dodd, J. L., W. K. Lauenroth, R. G. Woodmansee, G. L. Thor, and J. Chilgren.
1979. Soil, Chemical, and Meteorological Characteristics at ZAPS. In:
The Bioenvironmental Impact of a Coal-Fired Power Plant, Fourth Interim
Report, Colstrip, Montana, E. M. Preston and T. L. Gullett, eds.
EPA-600/3-79-044, U.S. Environmental Protection Agency, Corvallis, Or.
pp. 331-376.
Dunnett, C. W. 1955. A Multiple Comparison Procedure for Comparing Several
Treatments with a Control. J. Am. Stat. Assoc., 50:1096-1121.
Eddleman, L. E. 1977. Indigenous Plants of Southeastern Montana. I. Viability
and Suitability for Reclamation in the Fort Union Basin. Special Publi-
cation Four. Montana Forest and Conservation Experiment Station, School
of Forestry, University of Montana, Missoula, Mt. 122 pp.
170
-------
Farrar, J. F., J. Ralton, and A. J. Rutter. 1977. Sulphur Dioxide and the
Scarcity of Pinus sylvestris in the Industrial Pennines. Environ.
Pollut., 14: 63-68.
Friedman, M. 1937. The Use of Ranks to Avoid the Assumption of Normality
Implicit in the Analysis of Variance. J. Amer. Statis. Assoc., 32:675-701.
Houston, D. B., and L. S. Dochinger. 1977. Effects of Ambient Air Pollution
on Cone, Seed and Pollen Characteristics in Eastern White and Red Pines.
Environ. Pollut., 12:1-5.
Kohut, R. J., S. V. Krupa, and F. Russo. 1978. An Open-Top Chamber Study to
Evaluate the Effects of Air Pollution on Soybean Yield. In: Proc. of
4th Joint Conference on Sensing of Environmental Pollutants, American
Chemical Society, Washington, D.C. pp. 71-73.
Maguire, J. D. 1962. Speed of Germination-Aid in Selection and Evaluation
for Seedling Emergence and Vigor. Crop Science, 2:176-177.
McAlister, D. F. 1943. The Effect of Maturity on the Viability and Longevity
of the Seeds of Western Range and Pasture Grasses. J. Amer. Soc. Agron.,
35:442-453.
Oshima, R. J., P. K. Braegelmann, D. W. Baldwin, V. Van Way, and 0. C. Taylor.
1977. Reduction of Tomato Fruit Size and Yield by Ozone. J. Amer. Soc.
Hort. Sci.s 102: 289- 293.
Pack, M. R., and C. W. Sulzback. 1976. Response of Plant Fruiting to Hydrogen
Fluoride Fumigation. Atmos. Environ., 10:73-81.
Rice, P. M., L. H. Pye, R. Boldi, J. O'Loughlin, P. C. Tourangeau, and C. C.
Gordon. 1979. The Effects of "Low Level S02" Exposure on Sulfur Accumu-
lation and Various Plant Life Responses of Some Major Grassland Species
on the ZAPS Sites. In: The Bioenvironmental Impact of a Coal-Fired
Power Plant, Fourth Interim Report, Colstrip, Montana, E. M. Preston and
T. L. Gullett, eds. EPA-600/3-79-044, U.S. Environmental Protection
Agency, Corvallis, Or. pp. 494-591.
Rogler, G. A. 1960. Relation of Seed Dormancy of Green Needlegrass (Stipa
viridula) to Age and Treatment. Agron. J., 52:467-469.
Westwood, W. E. 1979. Oxidant Effects on Californian Coastal Sage Scrub.
Science, 205: 1001-1003.
Winner, W. E., and J. D. Bewley. 1978. Contrasts Between Bryophyte and
Vascular Plant Synecological Responses in an SOa-Stressed White Spruce
Association in Central Alberta. Oecologia (Berl.), 33:311-325.
Young, J. A., and.R. A. Evans. 1977. Squirrel Tail Seed Germination.
J. Range Manag., 30: 33-36.
171
-------
SECTION 12
EFFECTS OF CONTROLLED S02 EXPOSURE ON NET PRIMARY
PRODUCTION AND PLANT BIOMASS DYNAMICS
J. L. Dodd, W. K. Lauenroth, and R. K. Heitschmidt
ABSTRACT
Objectives of this study were to determine effects
of low level S02 fumigation on above- and belowground
plant biomass dynamics and aboveground net primary
production. With one exception we were unable to
measure significant effects on these parameters.
Rates of increase in rhizome biomass over the four-
year period of the study were reduced by S02 and in-
dicate that long term exposure of this grassland to
low level S02 will ultimately reduce the vigor of the
perennating. organs to a point where above- and below-
ground biomass dynamics will be altered, net primary
production will be reduced, and species composition
will be modified.
INTRODUCTION
Because of the increase in electrical power production in the Northern
Great Plains, air pollution is expected to increase (Northern Great Plains
Resource Program, 1975). Aside from the power industry itself, the main
industry of the region is the range livestock industry. It is quite possible
that the altered air quality will affect the livestock industry. As with
other rangelands, these rangelands are native ecosystems that are complex
combinations of indigenous plant and animal species that interact with each
other and their physical environment in capturing, storing, and utilizing
energy in a self-sustaining manner. Under normal circumstances, domestic
livestock production can be fit into this process quite profitably without
interfering with the self-sustaining character of the ecosystem processes
(Lewis, 1969). Further, livestock production on rangelands is closely associ-
ated with the quantity and quality of aboveground forage production. The
threat that air pollution poses to a livestock industry based on native range-
land is threefold: direct effect of air pollutants on the livestock, direct
effect on quantity and quality of forage production, and effects on the self-
sustaining properties of the ecosystem. The last two of these questions are
considered for one grassland type in this study.
172
-------
Several investigators have reported biomass dynamics and net primary
production estimates for the northern mixed grass prairie under pollution-
free conditions (Coupland, 1973; Lauenroth and Whitman, 1977; Lauenroth et
al., 1975; Dodd et at., 1974). These studies indicate rapid increases of
plant biomass in late May, June, and early July and aboveground net primary
productivity estimates ranging from 100 to 200 g • m~2 • yr"1. Although the
species composition of these grasslands is dependent on specific site fac-
tors, such as soils, rainfall, and grazing history, these grasslands are
nearly always dominated by cool season grass species, and in most cases
Agropyron smifhii Rydb. is the main species. Other cool season grass or
grasslike species of importance throughout the region are St-Lpa oomata Trin.
& Rupr., S. viridula Trin., Koeleria oristata (L.) Pers., Calamagrostis
montanensis Scribn., Carex spp., Bromus spp., and other species of Agropyron.
Although much research has been done on effects of air pollution on
plants of ornamental or agronomic importance (see reviews by Daines, 1968;
Mudd and Kozlowski, 1975; Treshow, 1970), comparatively little has been
conducted on effects on productivity of the native plant species of the
semi-arid rangelands of the western United States. Recent studies indicate
that grasses of semi-arid rangelands are more resistant to air pollution than
are agronomic species (Hill et al., 1974; Bennet and Hill, 1973; Davis et
al., 1966; Tingey et al., 1976). However, Bell and Clough (1973) have found
significant reduction in yield of a cool season grass, Lol-iwn pepenne, exposed
to low-level S02 (12 pphm) for sevexal weeks.
The objectives of this study were to determine the short- and near-term
effects of exposure to low-level S02 concentration on intraseasonal biomass
dynamics and aboveground net primary productivity (ANPP) of a northern mixed
grass prairie. We hypothesized that low-level exposures of S02 throughout
the growing season would reduce ANPP and change species and species group
contributions to total ANPP.
MATERIALS AND METHODS
Precipitation was measured on each study site between May and September
in each year. Soil water content to a depth of 105 cm was determined gravi-
metrically on each treatment plot throughout each growing season. Sampling
frequency varied from weekly to monthly during the study.
Aboveground plant biomass was sampled on six dates in 1975 and 1976, in
mid-July in 1977 and in mid-May and mid-July in 1978. Quadrats were located
randomly, and vegetation was clipped and separated by species into current
live, recent dead (current year's dead) and old dead (previous year's dead).
Samples were oven dried at 60°C to constant weight and weighed. In 1975, ten
circular 0.5-m2 quadrats were sampled in each treatment on each date. This
was changed to 20 circular 0.25-m2 quadrats in 1976, 1977, and 1978 to in-
crease precision.
Although numerous computational procedures exist for estimating ANPP from
harvest data (Singh et al., 1975), only one procedure was appropriate for all
treatments in all years of this study: the total standing crop of current
production as measured in mid-July. The more accurate procedure (summation of
173
-------
species and species group peaks of current production from frequent harvest
data) could be used only in 1975 and 1976, when we sampled monthly throughout
the growing season. We computed ANPP estimates by both procedures for 1975
and 1976. The summation of peaks procedure resulted in an 18% (SE = 3%)
greater estimate of ANPP than did the July standing crop estimate. A compar-
able difference between the two procedures was reported for a South Dakota
grassland of similar botanical character (Dodd et at., 1974). Therefore, our
estimates of ANPP are July standing crop estimates increased by 18%.
Belowground plant biomass to a depth of 10 cm was sampled monthly (May-
September) in 1975, 1976, and 1978 and in mid-July in 1977. On each sampling
date two belowground samples were taken from each herbage sample location
after harvest of the aboveground material. Samples were secured with a 7.5 cm
diameter x 10 cm long steel cylinder.
Plant material was separated from soil by the washing-flotation procedure
described by Lauenroth and Whitman (1971) and separated into roots, rhizomes,
and crowns (basal parts of shoots that are subterranean but positioned above
the transition zone of the plant). This washing-flotation procedure retains
all live and dead belowground plant material that will not pass through a
60-mesh • in"1 screen. Separated materials were oven dried to a constant
weight, ashed at 600°C, and reweighed; results are reported on an ash-free
organic matter basis.
RESULTS AND DISCUSSION
Precipitation and Soil Water
Early growing season precipitation (May-July) was near normal in 1977,
40% greater than normal in 1975, and more than twice normal in 1978 (Table
12.1). Although early season precipitation in 1976 was nearly 40% above
normal on Site I, it was only slightly above normal on Site II. August and
September precipitation was low and near normal in all years. Soil water
storage was high in early season and rapidly depleted in July and August in
all years (Figure 12.1).
Aboveground Biomass Dynamics
In 1975 and 1976 seasonal biomass dynamics for the total plant community
consisted of an early period of rapid increase (late April to early July),
followed by a period of slight increase (1975) or a slight decrease (1976) in
standing crop of current seasons production (Figure 12.2). The cessation of
rapid growth was associated with diminished rainfall and rapid exhaustion of
soil water (Figure 12.1). The S02 treatments did not appear to alter the timing
or rates of standing crop increases during the rapid growth period, nor did the
treatment alter the post-peak biomass dynamics.
Community biomass dynamics were largely a reflection of the biomass
dynamics of western wheatgrass (Figure 12.3) and to a lesser extent of prairie
Junegrass (Figure 12.4). Although the low frequency of standing crop estimates
prohibits precise evaluation, western wheatgrass appeared to have its most
174
-------
TABLE 12.1. GROWING SEASON PRECIPITATION* FOR SITES I AND II, 1975-1978,
COMPARED WITH LONG-TERM AVERAGE (1932-1977) AT BROADUS,
MONTANAt
v
Month
May
June
July
August
September
1975
Site
I
104
109
29
14
6
1976
Site
I
97
102
38
10
14
II
44
94
41
17
14
1977
Site
I
45
91
18
18
36
II
46
98
14
18
40
1978
Site
I
263
57
28
13
No
II
276
56
30
11
data
Long-term
average
57
80
35
27
30
May-July
242
237 179
154 158
348 362
172
" TTffll
t Broadus, Montana, is located 40 km northeast of the study sites.
rapid aboveground growth period later in the season than did prairie June-
grass. Also, western wheatgrass standing crop appeared to persist longer
after the rapid growth period than did prairie Junegrass. Treatment-induced
changes in biomass dynamics of these two species were not detected.
Belowground Biomass Dynamics
Belowground plant biomass dynamics were dominated by changes in root
biomass. Even though root standing crops were estimated monthly during the
growing season on all treatments in 1975, 1976, and 1978, we found consistent
dynamics only in 1978 (Table 12.2). From May to July, root biomass decreased
by about 25% and from July to September increased by about the same amount.
This demonstration of a mid-season low is generally consistent with findings
of other studies conducted in the northern mixed grass priarie (Lauenroth and
Whitmann, 1977; Dodd et al., 1974; Lewis et al., 1971) and simulations of
belowground dynamics. (Bartog and Jameson, 1974; Detling &t al.31979).
The S02 treatments did not alter seasonal changes in root biomass in
1978. Because the 1975 and 1976 monthly estimates of root biomass were more
erratic, we could not test the effects of S02 treatments on seasonal changes
in root biomass in these years.
Total belowground plant biomass increased over the course of the experi-
ment on ,the controls and all treatments on both sites (Table 12.3). On Site I
175
-------
Control
E
u
UJ
I
O
CO
Low
High
Mar Apr May Jun Jul Aug Sep Mar Apr May Jun Jul Aug Sep
E
o
(£.
UJ
i
o
CO
SITE IE
32
28
24
20
16
12
8
4
28
24
20
16
12
8
4
Control
Low
Medium
High
Mar ' Apr ' May Jun Jul Aug Sep Mar Apr May Jun Jul Aug Sep
Figure 12.1.
Soil water dynamics within 0-105 cm depth for Sites I and
II, 1975-1978.
176
-------
ZAPS-I
Figure 12.2.
Apr May Jun Jul Aug Sep
1975
- Control
Low
Medium
High
Mar Apr May Jun Jul Aug Sep
1976
Seasonal change in current years production for all plant
species combined, ZAPS I and II, 1975 and 1976.
M
E
CO
CO
<
o
00
100
80
60
40
20
ZAPS -I
Apr May Jun Jul Aug Sep
1975
Control _ 140
Low «*
Medium
High
120
100
co 80
| 60
ffl 40
20
0
zAPS-n
Mar Apr May Jun Jut Aug Sep
1976
Figure 12.3. Seasonal change in current year's production for western
wheatgrass, ZAPS I and II, 1975 and 1976.
177
-------
1
§
80
60
40
20
0
ZAPS -I
Apr May Jun Jul Aug Sep
1975
ST 120
Control
Low
Medium
High
E
^
o»
O
CD
100
80
60
40
20
0
ZAPS-
Mar Apr May Jun Jul
1976
Aug Sep
Figure 12.4. Seasonal change in current year's production for prairie
Junegrass, ZAPS I and II, 1975 and 1976.
TABLE 12.2. SEASONAL ROOT DYNAMICS FOR SITES I AND II, 1978.
Date
20 May 1978
20 July 1978
17 July 1978
16 August 1978
15 September 1978
Average
Site I
Control
709127
650150
549±30
598+35
709137
643
732128
614126
501127
713142
771143
676
Low
759131
624124
521120
741144
724131
674
Site II
Medium High Control
822138
737141
491116
720132
801129
714
613143
643+25
649127
738133
646133
658
766128
755138
585135
765144
795+46
733
Low
688134
624134
546143
827137
788135
695
Medium High
791133
608+29
434122
708+27
665+38
641
* X 1 SE, g • m 2, ash-free biomass, 0-10 cm depth.
178
-------
TABLE 12. 3. SEASONAL AVERAGE OF BELOWGROUND PLANT BIOMASS (TOTAL)* FOR
CONTROL AND S02 TREATMENTS, SITES I AND II, 1975-1978
Site I
Site II
Year
Control Low Medium High
Control Low Medium High
1975
1976
1978
664 601 623 610
652 619 627 619
776 807 797 832
616 709 785 717
780 873 834 732
* g • m~2 ash-free, 0-10 cm depth.
the increases between the first and fourth year were greater with increasing
dosages of S02, but the reverse was true on Site II by the third year of
exposure.
Belowground plant biomass consists of roots, crowns, and rhizomes. The
storage organs—crowns and rhizomes—accounted for 6-13% and 3-6% of total
belowground plant biomass, respectively, while roots accounted for 83-90% of
belowground plant biomass (Table 12.4).
The proportion of total belcwground plant biomass in crowns increased
during the experiment on both test sites, while that in roots decreased
slightly. The increase in crowns appeared to be enhanced by the S02 treat-
ments, with the greatest increase taking place on the high treatment of Site I
in the fourth year of treatment. A similar trend was exhibited on Site II by
the third year of treatment. In addition to the increase of crowns as a
proportion of total belowground plant parts, crowns also increased in absolute
abundance. By the fourth year on Site I the increases were by factors of 1.5,
1.9, 1.9 and 2.4 on the control, low, medium, and high treatments, respec-
tively. And by the third year on Site II the increases were by factors of
1.4, 1.4, 1.5, and 1.3 for the control, low, medium, and high treatments,
respectively.
The proportion of belowground plant biomass existing as rhizomes on
Site I increased slightly on the control and decreased on all treated plots
(Table 12.4). Site II exhibited no changes in this parameter by the third
year of treatment.
In spite of the decrease in rhizomes as a proportion of total belowground
organs over the course of the experiment, they did increase in abundance
(Table 12.6). By 1978 rhizome biomass had increased by 30%, 20%, 10%, and
less than 10% on the control, low, medium, and high treatments, respectively,
179
-------
TABLE 12.4. MORPHOLOGICAL COMPOSITION OF BELOWGROUND PLANT
BIOMASS (BASED ON SEASONAL AVERAGE ASH-FREE
BIOMASS)
Year
1975
1976
1978
1975
1976
1978
1975
1976
1978
Site I Site II
Control Low Medium High Control Low Medium High
Rhizomes
4 444
6555 4444
5333 4444
Crowns
9 986
10 11 12 11 11 10 9 11
12 13 12 ' 11 12 12 13 13
Roots
86 87 88 90
83 84 83 84 86 85 87 85
83 84 85 86 84 84 83 83
* %
TABLE 12.5.
SEASONAL AVERAGE CROWN BIOMASS* FOR CONTROL AND S02
TREATMENTS, SITES I AND II, 1975-1978
Year
1975
1976
1978
Site I Site II
Control Low Medium High Control Low Medium High
61 56 50 38
67 68 75 68 65 73 70 76
94 104 95 92 93 103 97 101
ash-free, 0-10 cm depth.
on both study sites. Thus, the reduction in proportional contribution of rhi-
zomes to total belowground plant biomass with increasing S02 concentrations
resulted largely from a reduction in rates of increase of the rhizome popula-
tion.
180
-------
TABLE 12.6. SEASONAL AVERAGE RHIZOME BIOMASS* FOR CONTROL AND S02
TREATMENTS, SITES I AND II, 1975-1978
Site I
Site II
Year
Control Low Medium High
Control Low Medium High
1975
1976
1978
29
42
39
22
30
27
25
29
28
24
29
26
23
29
32
37
28
32
29
30
g • m~2 ash-free, 0-10 cm depth.
Aboveground Net Primary Production (ANPP)
ANPP was not estimated for the test sites before exposure to S02- There-
fore, our assumption that the treatment areas within each of the test sites
were equal in production potential to the control areas before fumigation can
only be tested with less appropriate information. Harvest estimates of stand-
ing crop residual from production occurring in the year prior to fumigation
were made in the spring of the first season of exposure and indicated only
slight differences among the control and treatment areas within either Site I
or Site'II (Table 12.7). The low and medium areas on Site I had slightly
higher residual material than did the control and high areas. And on Site II
the residual material was slightly greater on the control than on the treat-
ment areas. Since slight differences in residuals could also arise from
differential weathering during the winter months we conclude that the residual
estimates indicate slight, if any differences in production potential within
either test site.
TABLE 12.7. RESIDUAL STANDING CROP* OF PREVIOUS YEARS PRODUCTION FOR
SITE I AND SITE II IN SPRING BEFORE FIRST SEASON OF S02
TREATMENT
Control
Low
Medium
High
Site I, April 1975
Site II, March 1976
77 ± 7
86 ± 4
97 ± 8
75 ± 6
93 ± 7
66 ± 6
81 ± 10
64 + 4
* X ± SE.
181
-------
ANPP ranged from 119 to 268 g • m~2 throughout the study period and
Site II was more productive than Site I (Table 12.8). The greatest differences
in ANPP were between years. While 1977, the driest year of the study (Table
12.1), was clearly the least productive year on both sites, 1978, the wettest
year, was the most productive only on Site II. In fact, ANPP in 1978 on
Site I was not significantly greater than in 1977 and was lower than ANPP in
1975 and 1976. Since most of the increase in precipitation in 1978 arose from
May precipitation being nearly 5 times normal (with 130 mm on May 20), we must
conclude that this amount was excessive and reduced productivity via leaching
of required nutrients on Site I. Since ANPP was not similarly depressed on
Site II we suggest that nutrients were not leached beyond the effective root-
ing zone because of differences in soil characteristics. Clay contents in the
upper 30 cm of the profile range from 33 to 42% on Site II and from 16 to 32%
on Site I (Dodd et al., 1979). Fertility factors associated with these tex-
tural differences between Site I and Site II are presumed to account for the
differential productivity noted between Sites I and II.
TABLE 12-. 8. ABOVEGROUND NET PRIMARY PRODUCTION* FOR SITES I AND II,
1975-19781"
ZAPS I
ZAPS II
Year Control Low Medium High Control Low Medium High
X ± SE X ± SE X ± SE X ± SE X ± SE X ± SE X ± SE X ± SE
1975
1976
1977
1978
150114 149117 165113 156H4
199113 1861 9 195111 1991 8 1771 8 218H7 205H4 227H5
126112 131113 1311 7 1191 5 1371 7 210H3 1691 9 190H1
1361 8 156114 1231 6 1291 6 268H8 219118 214+13 2531 8
* g • m~2.
t ANPP = July current production x 1.18
Treatment differences in ANPP were not detected on Site I. And, although
statistically significant treatment differences were noted on Site II, the
patterns of the differences are not suggestive of responses to S02 exposure.
CONCLUSIONS
Considering the biological and sampling variability associated with our
estimates of biomass parameters, we cautiously draw the following conclusions.
Aboveground plant biomass dynamics, aboveground net primary productivity, and
182
-------
total belowground biomass dynamics of northern mixed-grass prairie are not
immediately sensitive to low level S02 exposure regimes. However, short term
reductions in rhizome biomass suggest that aboveground biomass dynamics and
primary productivity will eventually be altered because of loss of capability
of the rhizomes to support aboveground processes in western wheatgrass, the
dominant plant species in this grassland type. We assume that similar dis-
ruptions of carbon allocation to storage organs of other perennial species
also occurs and will ultimately contribute to significant changes in primary
productivity, biomass dynamics, and species composition.
REFERENCES
Bartos, D. L., and D. A. Jameson. 1974. A Dynamic Root Model. Am. Midi. Nat.,
91:499-504.
Bell, J. N. B., and W. S. Clough. 1973. Depression of Yield in Ryegrass Ex-
posed to Sulfur Dioxide. Nature (London)., 241:47-49.
Bennet, J. H., and A. C. Hill. 1973. Inhibition of Apparent Photosynthesis
by Air Pollutants. J. Environ. Qual., 2(4) -.526-530.
Coupland, R. T. 1973. Producers: I. Dynamics of Aboveground Standing Crop.
Canadian IBP Matador Project Tech. Rep. No. 27. Univ. of Saskatchewan,
Saskatoon. 159 pp.
Daines, R. H. 1968. Sulfur Dioxide and Plant Responses. J. Occup. Med.,
10:216-224.
Davis, C. R., D. R. Howell, and G. W. Morgan. 1966. Sulfur Dioxide Fumigations
of Range Grasses Native to Southeastern Arizona. J. Range Manage.,
19:60-64.
Detling, J. K., W. P. Parton, and H. W. Hunt. 1979. A Simulation Model of
Bouteloua graoilis Biomass Dynamics on the North American Shortgrass
Prairie. Oecologia, 38:167-191.
Dodd, J. L., W. K. Lauenroth, G. L. Thor, and M. B. Coughenour. 1979. Effects
of Chronic Low Level S02 Exposures on Producers and Litter Dynamics. In:
The Bioenvironmental Impact of a Coal-fired Power Plant, E. M. Preston
and T. L. Gullett, eds. 4th Interim Rep., OS/EPA, Corvallis, Oregon.
pp.
Dodd, J. L., J. K. Lewis, H. L. Hutcheson, and C. L. Hanson. 1974. Abiotic
and Herbage Dynamics Studies on the Cottonwood Site, 1971. US/IBP Grass-
land Biome Tech. Rep. No. 250. Colorado State Univ., Fort Collins. 195
pp.
Hill, A. C., S. Hill, C. Lamb, and T. W. Barrett. 1974. Sensitivity of Native
Desert Vegetation to S02 and to S02 and N02 Combined. J. Air Poll. Cont.
. Assoc., 24:153-157.
183
-------
Lauenroth, W. K., K. J. Dodd, R. K. Heitschmidt, and G. R. Woodmansee. 1975.
Biomass Dynamics and Primary Production in Mixed Prairie Grasslands in
Southeastern Montana: Baseline Data for Air Pollution Studies. Fort
Union Coal Field Symposium, Eastern Montana College, Billings, pp. 559-
578.
Lauenroth, W. K., and W. C. Whitman. 1971. A Rapid Method for Washing Roots.
J. Range Manage., 24:308-309.
Lauenroth, W. K., and W. C. Whitman. 1977. Dynamics of Dry Matter Production
in a Mixed-grass Prairie in Western North Dakota. Oecologia, 27:339-351.
Lewis, J. K. 1969. Range Management Viewed in the Ecosystem Framework. In:
The Ecosystem Concept in Natural Resource Management, G. M. Van Dyne, ed.
Academic Press, New York. pp. 97-187.
Lewis, J. K., J. L. Dodd, H. L. Hutcheson, and C. L. Hanson. 1971. Abiotic
and Herbage Dynamics Studies on the Cottonwood Site, 1970. US/IBP Grass-
land Biome Tech Rep. No. 111. Colorado State Univ., Fort Collins.
147 pp.
Mudd, J. B., and T. T. Kozlowski, eds. 1975. Responses of Plants to Air Pol-
lution. Academic Press, New York. 383 pp.
Northern Great Plains Resource Program. 1975. Effects of Coal Development in
the Northern Great Plains. Northern Great Plains Resource Program.
(Special report prepared as part of NGPRP in cooperation with federal,
state, regional, local and private organizations). 165 pp. '
Singh, J. S., W. K. Lauenroth, and R. K. Steinhorst. 1975. Review and Assess-
ment of Various Techniques for Estimating Net Aerial Primary Production
in Grasslands from Harvest Data. Bot. Rev., 41:181-230.
Tingey, D. T., R. W. Field, and L. Baird. 1976. Physiological Responses of
Vegetation to Coal-fired Power Plant Emissions. In: The Bioenvironmen-
tal Impact of a Coal-fired Power Plant: 2nd Interim Rep. Colstrip, Mon-
tana, February 1976, R. A. Lewis, N. R. Glass, and A. S. Lefohn, eds.
EPA 600/3-76-013, Environmental Protection Agency, Corvallis, Oregon, pp.
14-40.
Treshow, M. 1970. Environment and Plant Response. McGraw-Hill, New York.
184
-------
SECTION 13
FORAGE QUALITY IN A GRASSLAND EXPOSED TO SULFUR DIOXIDE
D. G. Milchunas, W. K. Lauenroth, and J. L. Dodd
ABSTRACT
Effects of three exposure levels of S02 (see
section 6) on the nutritive value of western
wheatgrass were investigated. Significant increases
in plant sulfur content were observed, both with time
and level of S02 exposure. Plant ash content paralleled
the trends observed for sulfur concentrations. Nitrogen
levels in western wheatgrass were not affected by S02
treatment. The increasing plant sulfur content and
decreasing N:S ratios across treatments did not signifi-
cantly affect forage digestibility as measured by in
vitro digestible dry matter.
INTRODUCTION
The Northern Great Plains possess a combination of fertile soils and a
climate conducive to the production of cereal grains and beef. Gaseous and
particulate air pollutants from coal burning pose a threat to traditional re-
source use because they are phytotoxic and because they disperse over large
areas. Data from permit applications (Durran et al., 1979) project an increase
in electric generating capacity for the Northern Great Plains from approximately
3000 MW in 1976 to 17000 MW in 1986.
Sulfur dioxide (S02) is one product of coal combustion that is of concern
with regard to its effects upon plants and animals. Sulfur is an important
constituent of many organic molecules, including methionine, cystine, gluta-
thione, thiamine, biotine, and coenzyme-A. At high levels of sulfur exposure,
plant yields can be reduced (Bell and Clough, 1973) and increased leaf senes-
cence (Heitschmidt et al., 1977) has been observed. Reports of changes in the
amount of plant protein or amino acids in response to S02 exposure are contro-
versial (Ziegler, 1975). Microorganisms of the ruminant animal are capable of
utilizing N and S in the synthesis of amino acids (Thomas et al., 1951; Starks
et al., 1954; Hale and Garrigus, 1953; Abdo et al., 1964). Sulfate supplemen-
tation has been shown to increase in vitro cellulose digestion (Hunt et al.,
1954; Martin et al., 1964; Barton et al., 1971). Sulfur has also been reported
185
-------
to be toxic to rumen microorganisms at a concentration of 100 ppm from sodium
sulfate (Hubbert et at., 1958) and at 30 ppm from sodium sulfite (Trinkle et
al., 1958). Because S02 may be metabolized by the plant and rumen microorgan-
isms and used as a source of sulfur, the balance between positive and negative
effects of exposure will depend upon concentration, duration of exposure, and
rate of growth and metabolism of S02. This study investigates the effects of
three exposure levels of S02 on the nutritive quality of an important Northern
Great Plains forage species, western wheatgrass (Agropyron smithii Rydb.).
MATERIALS AND METHODS
Nitrogen and ash were determined by procedures in A.O.A.C. (1965). Total
sulfur concentrations were analysed with a Leco Induction Furnace (Laboratory
Equipment Corp., St. Joseph, Missouri). In vitro digestible dry matter (IVDDM)
was determined with techniques described by Tilley and Terry (1964), as modified
by Pearson (1970), with buffer solution mixed according to McDougall (1948:
Table 11). Inoculum was obtained from a fistulated cow maintained on grass hay
for 2 weeks before collection.
Data were analyzed with a repeated measure analysis of variance repeating
across time (Winter, 1971). Tukey's Q values were used to identify significant
differences between means (Snedecor and Cochran, 1967).
RESULTS AND DISCUSSION
Analysis of sulfur data for western wheatgrass indicated significant across-
treatment within-month and across-month within-treatment effects (Table 13.1).
No significant differences were observed between the control and low-S02 treat-
ments, although a trend of higher sulfur accumulation was evident. The number
of significant differences across treatments increased with time of exposure,
indicating greater sulfur accumulation across time with higher levels of S02
exposure. Significant differences were observed between years on the high
treatment, with the order 1975 < 1977 < 1976 (Figure 13.1 ). We cannot deter-
mine, therefore, to what degree sulfur accumulated with each additional year of
treatment. Different yearly abiotic factors and plant growth responses appear
to outweigh the effect of sulfur accumulation from previous years.
A trend of higher ash concentration with increased level of S02 was ob-
served for all year-site combinations (Table 13.2). There also appeared to be
a general trend of increasing ash concentrations as the growing season pro-
gressed. The nutritional implications of the increased ash content is uncer-
tain. Materna (1961) and Matsushima and Harada (1966) observed increasing
potassium content in leaves exposed to S02. Incoming S02 is rapidly converted
to S04 by the plant. The formation of sulfate salts for storage of excess
sulfur may be increasing the requirement of the plant for the cations used in
formation of these salts. At the same time, increased mineral nutrition for
consumers may result.
Nitrogen concentrations within-month across-treatments showed no signifi-
cant differences and no trends were observed (Table 13.3).. Values differed
significantly on a seasonal basis, with highest levels occurring in May and
lowest levels in August or September. Nitrogen levels for site I 1975 and 1977
186
-------
TABLE 13.1. PERCENT SULFUR CONTENT OF WESTERN WHEATGRASS FOR 5 DATES, 4
TREATMENTS, 2 SITES, AND 3 YEARS OF S02 FUMIGATION
Site and
treatment Year
Site I 1975
Control
Low
Medium
High
Date mean
Site I 1976
Control
Low
Medium
High
Date mean
Site I 1977
Control
Low
Medium
High
Date mean
Site II 1976
Control
Low
Medium
High
Date mean
Site II 1977
Control
Low
Medium
High
Date mean
Date
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
May
.180a
.128a
.140a
.131a
.140
.071a.
.114ab
.162°
.235°
.146
.127a
.152a
o
.163a
V.
.237°
.170
.129\
Q K
.142a°
.213
.322°
.201
.135a
• 169a.
f\ n
.194a°
.258b
.189
June
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.098a,
.142ab
.207b°
.228bc
.169
.091a
.135a
K
.216°
.242b
.171
__
___
—
—
—
.107a
Q
.124a
K
.233
.351°
.204
__.
__
~"^
July
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
191ab
137a
241bc
276°
211
094\
153ab
b
218°
321°
197
113a
133a
K
215°
h
270
183
095a
Q
167a
h
271°
433°
241
086\
134ab
h
177°
376°
195
Aug. Sept.
0.115a
0.134a
0.262^
0.331
0.210
0.086a
0.151a
v>
0.293°
0.412°
0.235
0.121a
0.131a
K
0.213°
>>
0.324
0.197
o.ioia
Q
0.147a
K
0.277
0.482°
0.252
0.1093,
0.176ab
K
0.222°
0.389°
0.224
-
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
-
104a
174a
b
252
364°
223
121a
158a
h
243
p
371°
223
—
—
—
—
—
—
__
Treatment
mean*
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.141a
.135a
.223b
.241b
.185
.089a
.145b
p
.288°
.315d
.194
.146a
.152a
K
.218°
p
.290
.201
.108a
a
.145a
h
.249
.397°
.225
.117a
.176b
K
.189
.329C
.203
abed
Treatment means for site I 1977 include date for April and an
additional August date. Treatment means for site II 1977 include
data for April.
Any two means not sharing common superscripts within a column and
year-site are significantly different (P < 0.05).
187
-------
were not significantly different, while site I 1976 was significantly lower.
Schwartz et al. (1978) reported a decrease in crude protein after 2 years of
fumigation. Analysis of 3 years of data indicates that the drop in crude
protein at site I 1976 was in response to different yearly plant growth patterns
as affected by abiotic factors, rather than in response to 2 years of S02 ex-
posure. Increases in total nitrogen have been observed when sulfur beyond
optimum levels was applied both as fertilizer (Randig, 1956) and S02 (Leone
and Brennan, 1972). Sulfur accumulation on our high-S02 treatment indicates
well above optimum levels with no increase in plant total nitrogen.
0.4
ID
u.
_I
ID
05
Lul
O
(T
UJ
Q_
0.3
0.2
O.I
II-76
1 2345
S02 CONCENTRATION
8
Figure 13.1.
Percent sulfur mean across all sampling dates at control, low',
medium, and high S02 treatments of Site I—1975, 1976, 1977—
and Site 11—1976, 1977.
188
-------
TABLE 13.2. PERCENT ASH CONTENT OF WESTERN WHEATGRASS FOR 5 DATES, 4
TREATMENTS, 2 SITES, AND 3 YEARS OF S02 FUMIGATION
Site and
treatment Year
Site I 1975
Control
Low
Medium
High
Date mean
Site I 1976
Control
Low
Medium
High
Date Mean
Site I 1977
Control
Low
Medium
High
Date mean
Site II 1976
Control
Low
Medium
High.
Date mean
Site II 1977
Control
Low
Medium
High
Date mean
May
__
—
—
—
—
6.940
7.080
7.180
8.250
7.363
6.223
6.487
6.742
7.832
6.822
9.020
8.265
9.575
9.425
9.071
7.532
6.466
7.147
7.250
7.099
June
7.660
7.830
8.070
8.210
7.904
6.485
7.385
7.130
7.890
7.222
__
—
—
—
7.620
6.420
8.035
8.205-
7.570
__
—
__
—
""•"
Date
July
6.782
6.380
7.685
8.310
7.289
6.570
6.935
7.015
8.265
7.196
7.904
7.619
7.551
8.137
7.803
8.605
8.400
9.020
9.160
8.796
8.298
6.944
8.307
8.256
7.977
Aug.
7.565
7.270
7.830
8.605
7.818
6.925
7.900
8.175
9.475
8.119
8.461
8.773
8.350
9.209
8.698
8.265
7.440
9.045
9.325
8.519
8.632
7.651
8.533
8.843
8.415
Sept.
—
—
—
—
7.040
7.795
8.295
9.280
8.103
7.894
9.474
9.260
9.615
9.061
7.780
8.675
8.610
10.480
8.667
__
—
—
—
«_
Treatment
mean*
7.386a
7'160ab
7.862ab
K
8.408
6.972a.
7.419ab
7'559K
8.632b
7.9413,
8.410ab
8.221ab
8.828
8.258ab
7.840a
8.884bc
9.190C
8.088ab
7.258a
7.787ab
8.374
* Treatment means for site I 1977 include April and an additional
August sampling date.
abc Any two means not sharing common superscripts within a column and
year-site are significantly different (P < 0.05).
189
-------
TABLE 13.3. PERCENT NITROGEN CONTENT OF WESTERN WHEATGRASS FOR 5 DATES,
4 TREATMENTS, 2 SITES, AND 3 YEARS OF S02 FUMIGATION
Site and
treatment Year
Site I 1975
Control
Low
Medium
High
Date mean
Site I 1976
Control
Low
Medium
High
Date mean
Site I 1977
Control
Low
Medium
High
Date mean
Site II 1976
Control
Low
Medium
High
Date mean
Site II 1977
Control
Low
Medium
High
Date mean . , n
All
All sites years
Control
Low
Medium
High
Date mean
May
2.545
2.513
2.694
2.662
2.590a
2.230
2.177
2.047
1.891
Q
2.086a
2.208
2.313
2.094
2.527
2.285a
2.285
2.446
2.303
2.498
2.383a
2.587
2.950
2.770
2.815
2.781a
2.371
2.480
2.347
2.478
2.421
June
1.821
1.721
1.677
1.944,
1.791b
1.544
1.378
1.461
1.351
1.434b
1.979
1.837
1.897
1.960
1.918
1.099
1.908
1.644
1.915,
1.641b
2.212 '
2.406
2.279
2.459
2.339
1.731
1.850
1.791
1.925
1.824
Date
July
1.333
1.269
1.361
1.183
1.286°
1.237
1.059
1.094
1.092
/•>
1.120
1.384
1.087
1.186
1.218
1.219°
1.377
1.599
1.322
1.476
1.443b
1.580
1.590
1.570
1.603
1.588°
1.382
1.321
1.277
1.314
1.325
Aug.
1.107
1.082
1.124
1.000
1.078d
0.974
0.799
0.936
0.890
0.900
1.270
1.096
1.092
1.237
1.174°
1.035
1.153
1.012
1.175
(-*
1.094
1.729
1.699
1.498
1.907
1.708°
1.223
1.165
1.132
1.242
1.191
Sept.
—
—
—
—
—
0.760
0.640
0.646
0.685
0.683
0.944
0.700
0.862
0.862
0.842
0.984
0.949
0.811
1.298
(-•
0.996
—
—
—
—
—
—
—
—
—
— —
Treatment
mean*
1.701
1.646
1.574
1.697
1.657
1.497
1.353
1.385
1.306
1.385
1.710
1.583
1.567
1.736
1.649
1.449
1.776
1.570
1.766
1.640
2.027
2.161
2.095
2.196
2.121
1.677
1.704
1.628
1.740
1.688
* September data are not included in analysis of variance or treatment
mean.
abed Any two means not sharing a common superscript within a row are
significantly different (P < 0.05).
190
-------
Cattle and sheep require a N:S ratio of 11-13:1 and rumen microorganisms
of 12-14:1 (Leibholz and Naylor, 1971; Moir et aZ., 1967; Whanger et al., 1978).
This is similar to the ratio found in many proteins (Block and Weiss, 1956).
N:S ratios (Table 13.4) suggest that sulfur may be limiting protein production
on control and low-S02 fumigation plots in early spring. Whether this occurs
cannot be ascertained from total nitrogen (crude protein) analysis. Plants
contain nitrogen in various forms; therefore, true protein nitrogen to non-
protein nitrogen ratios would be necessary before conclusions could be drawn
about a possible fertilizing effect of S02. Plant senescence observed from S02
exposure (Heitschmidt, 1977) may in part be earlier maturation in re-
sponse to increased spring protein production made possible by the increase in
available sulfur, and this may occur irrespective of soil sulfur levels. Al-
though roots apparently have the ability to reduce sufficient sulfate to provide
the sulfide for their own needs, they do not translocate appreciable amounts of
reduced sulfur forms to the shoots (Salesbury and Ross, 1969). Therefore, the
differential requirements for reduced sulfur forms and thereafter ability of
leaves to reduce incoming sulfur could be compensated for by stomatal entry of S02.
Although S02 exposure in early spring may have a fertilizing effect, there
are data indicating a reduction in soluble protein (Rabe and Kreeb, 1979) and
changes of free amino acid concentrations (Godzik and Linskens, 1974) with S02
fumigation when soil sulfur levels are adequate. Longer exposure periods always
resulted in a decrease in true protein concentration. These studies were con-
ducted with crop species and the length of S02 treatment was relatively short
compared with our May through October exposure schedule. Our data indicated
that the effect of S02 exposure on protein production in native range plants
cannot be determined from total nitrogen analysis. True protein and non-protein
forms of nitrogen need to be distinguished.
The high N:S ratios in May and the high sulfur concentrations in August and
September did not significantly affect digestibility of western wheatgrass when
measured by the in vitro digestible-dry-matter technique (Table 13.5). However,
one cannot conclude that the S02 treatments do not affect ruminant metabolism of
protein or energy. The disappearance of dry matter or cellulose is only of
indirect importance to the energy metabolism of ruminants. The products pro-
duced from this disappearance supply the ruminant's energy requirements and are
of more practical importance. The proportion and quantity of these products
vary with the array of substrate components fermented. Huisingh et od. (1974)
described high concentrations of sulfate-reducing bacteria in the rumen of sheep
given sulfate as the only S source. Sulfur in the ruminant diet can alter
microbial population dynamics and, because of the different end products sup-
plied to the host by various species of microorganisms, may thereby affect
energy as well as nitrogen metabolism.
SO2 effects on ruminant nutrition may be manifested in structural fiber,
carbohydrate, protein-nitrogen, and sulfur status of the plant and thus in-
fluence the rumen microbial ecosystem and the supply of energy and protein to
the host. The content of a particular element or dietary constituent in a
forage has little significance unless it is qualified by a factor indicating
the biological utilization of that component. Because of the complexity of
191
-------
TABLE 13.4. NITROGEN TO SULFUR RATIOS OF WESTERN WHEATGRASS FOR 5 DATES,
4 TREATMENTS, 2 SITES, AND 3 YEARS OF S02 FUMIGATION
Site and
treatment Year
Site I 1975
Control
Low
Medium
High
Date mean
Site I 1976
Control
Low
Medium
High
Date
Site I 1977
Control
Low
Medium
High
Date mean
Site II 1976
Control
Low
Medium
High
Date mean
Site II 1977
Control
Low
Medium
High
Date mean
May
14.14
19.63
19.24
20.32
18.50
31.41
19.10
12.64
8.05
14.29
17.39
15.22
12.85
10.66
13.44
17.71
17.23
10.81
7.76
11.36
19.16
10.77
14.28
10.91
14.71
June
18.58
12.12
8.10
8.53
10.60
16.97
10.21
6.76
5.58
8.39
—
—
—
—
—
10.27
15.39
7.06
5.46
8.04
—
—
—
—
•^ ^
Date
July
6.98
9.26
5.65
4.29
6.09
13.16
6.92
5.02
3.40
5.69
12.25
8.17
5.52
4.51
6.66
14.49
9.57
4.88
3.41
5.99
18.37
11.87
8.87
4.26
8.14
Aug.
9.63
8.07
4.29
3.02
5.13
11.33
5.29
3.19
2.16
3.83
10.50
8.37
5.13
3.82
5.96
10.25
7.84
3.65
2.44
4.34
15.86
9.65
6.75
4.90
7.63
Sept.
—
—
—
—
—
7.31
3.68
2.56
1.88
3.06
7.80
4.43
3.55
2.32
3.78
—
—
—
—
—
—
—
—
—
— ~
Treatment
mean*
12.06
12.19
7.06
7.04
8.96
16.82
9.33
6.07
4.15
4.14
11.71
10.41
7.19
5.99
8.20
13.42
12.25
6.31
4.45
7.29
17.32
12.28
11.08
6.67
10.45
For May, June, July, and August data.
192
-------
TABLE 13.5. PERCENT IN VITRO DIGESTIBLE DRY MATTER OF WESTERN WHEATGRASS FOR
7 DATES, 4 TREATMENTS, 2 SITES, AND 3 YEARS OF S02 FUMIGATION
Site and
treatment Year
Site I 1975
Control
. Low
Medium
High
Date mean
Site I 1976
Control
Low
Medium
High
Date mean
Site I 1977
Control
Low
Medium
High
Date mean
Site II 1976
Control
Low
Medium
High
Date mean
Site II 1977
Control
Low
Medium
High
Date mean
Date
April
83
83
83
83
83
84
84
82
82
83
__
__
__
.17
.70
.60
.08
•a
.38a
—
—
.75
.20
.30
.14
.29a
May
71.84
64.40
69.20
69.04
67.86a
62.58
62.08
59.97
58.20
60.71a
72.37
70.89
69.00
74.08
71.46b
60.52
66.89
66.70
69.18
a
65.82
73.46
74.88
71.43
72. 63^
73.10b
June
54.18
53.92
55.53
60.04
55.92
49.40
48.50
50.46
50.63,
49.75b
71.20
—
—
67.09
69.15D
56.06
54.34
52.72
54.61
54.43°
—
—
—
—
—
July
50
47
50
51
50
47
44
44
44
45
60
58
57
58
58
48
50
50
47
49
65
63
64
63
64
.57
.05
.72
.92
.07°
.81
.05
.44
.96
.32°
.95
.31
.98
.12
r«
.86°
.53
.74
.15
.90
f>
.33°
.93
.41
.50
.82
.40C
T-
J. 1
Aug. 6 Aug. 30 Sept.
45
44
44
45
45
41
41
40
38
40
55
55
53
56
55
42
45
43
46
44
64
61
62
61
62
.73
.41
.24
•91^
.07d
.75
.75
.83
.68,
.75b
.70
.51
.15
•92d
.23d
.84
.05
.32
•A9d
.43°
.36
.23
.03
.65
.32°
40.
34.
39.
37.
38.
36.
36.
37.
35.
36.
51.41 54.
52.03 47.
53.65 53.
50.98 52.
52.05 51.
38.
40.
34.
41.
39.
—
—
—
—
—
67
10
43
32
05e
97
23
30
54
51e
10
76
12
82d
86
99
02
98
12
p
03
—
—
—
—
—
reatment
mean*
50.46a
49.46a
49.90a
51.91a
50.42
47.70a
46.52a
46.60a
45.60a
46.61
64.73a
59.003
61.13a
61.57a
61.75
49.39a
51.41a
49.57a
52.06a
50.61
70.32a
70.93a
69.22a
70.06a
70.20
* For May, June, July, and August data.
abcde Any two means not sharing a common superscript within a row are
significantly different (P < 0.05).
193
-------
interactions and possible masking effects, the study of the effect of S02 on
forage nutritive value simply by qualifying plant components and digestible dry
matter or cellulose disappearance is inadequate.
CONCLUSIONS
Increased sulfur accumulation was observed in western wheatgrass both
across time and level of S02 exposure. Trends in ash concentration paralleled
those of sulfur. Nitrogen concentrations decreased during the growing season
but were not affected by S02 treatments. Nitrogen to sulfur ratios were higher
than optimum in May on the control and low S02 treatments and very low on the
high S02 treatments in August and September. These ratios suggest a possible
fertilizing effect of S02 in early spring, with toxic levels being reached in
fall. S02 fumigation, and the high forage sulfur concentration, did not offset
digestibility for ruminants as measured -In vitro.
REFERENCES
Abdo, K. M., K. W. King, and R. W. Engel. 1964. Protein Quality of Rumen
Microorganisms. J. Anim. Sci., 23:734.
A.O.A.C. 1965. Official Methods of Analysis, 10th ed. Ass. Offic. Agr. Chem'. ,
Washington, D. C.
Barton, J. S., L. W. Bull, and R. W. Henten. 1971. Effects of Various Levels
of Sulfur upon Cellulose Digestion in Purified Diets and Lignocellulose
Digestion in Corn Fodder Pellets in vitro. J. Anim. Sci., 33:682.
Bell, J. N. B., and W. S. Clough. 1973. Depression of Yield in Ryegrass Ex-
posed to Sulfur Dioxide. Nature, 241:47.
Block, R. J., and K. W. Weiss. 1956. Amino Acid Handbook. Thomas, Spring-
field, Illinois.
Durran, D. R., M. J. Meldgin, M. K. Liu, T. Thoem, and D. Henderson. 1979. A
Study of Long Range Air Pollution Problems Related to Coal Development in
the Northern Great Plains. Atmos. Environ., 13:1021-1037.
Godzik, S., and H. F. Linskens. 1974. Concentration Changes of Free Amino
Acids in Primary Bean Leaves After Continuous and Interrupted S02 Fumiga-
tion and Recovery. Environ. Pollut., 7:25.
Hale, W. H., and U. S. Garrigus. 1953. Synthesis of Cystine in Wool from
Elemental Sulfur and Sulfate Sulfur. J. Anim. Sci., 12:492.
Heitschmidt, R. K. 1977. Chronic Effects of S02 on Western Wheatgrass in a
Montana Grassland. Ph.D. Thesis. Colorado State Univ., Fort Collins.
100 pp.
Hubbert, F., Jr., E. Cheng, and W. Burroughs. 1958. Mineral Requirements of
Rumen Microorganisms for Cellulose Digestion in vitro. J. Anim. Sci.,
17:559.
194
-------
Huisingh, J., J. J. MeNeill, and G. Matrone. 1974. Sulfate Reduction by a
Desulforibrio Species Isolated from Sheep Rumen. Appl. Microbiol., 28:489.
Hunt, C. H., 0. G. Bentley, T. V. Hershberger, and J. H. Cline. 1954. The
Effect of Carbohydrates and Sulfur on B-vitamin Synthesis, Cellulose Di-
gestion, and Urea Utilization by Rumen Microorganism in vitro. J. Anim.
Sci., 13:570.
Leibholz, J., and R. W. Naylor. 1971. The Effect of Urea in the Diet of the
Early-weaned Calf on Weight Gain, Nitrogen and Sulfur Balance, and Plasma
Urea and Free Amino Acid Concentrations. Aust. J. Agric. Res., 22:655.
Martin, J. E., L. R. Arrington, J. E. Moore, C. B. Ammerman, G. K. Davis, and
R. L. Shirley. 1964. Effect of Magnesium and Sulfur upon Cellulose Di-
gestion of Purified Rations by Cattle and Sheep. J. Nutr., 83:60.
Materna, J. 1961. Einfluss des schwefeldioxyds auf die mineralische
zusammensetzung der fichtennadeln. Naturwissenschaften, 48:723-724.
Matsushima, J., and M. Harada. 1966. Sulfur Dioxide Gas Injury to Fruit Trees.
V. Absorption of Sulfur Dioxide by Citrus Trees and Its Relation to Leaf
Fall and Mineral Contents of Leaves. J. Jap. Soc. Hort. Sci., 35:242-246.
McDougall, E. I. 1948. Studies on Ruminant Saliva. I. The Composition and
Output of Sheep's Saliva. Biochem. J., 43:99.
Moir, R. J., M. Somers, and A. C. Bray. 1967. Utilization of Dietary Sulfur
and Nitrogen. Sulfur Institute J., 3:15.
Pearson, H. A. 1970. Digestibility Trials: In vitro Techniques. In: Range
and Wildlife Habitat Evaulation—A Research Symposium, USDA Misc. Pub. No.
1147. pp. 85-92.
Rabe, R., and K. H. Kreeb. 1979. Enzyme Activities and Chlorophyll and Protein
Content in Plants as Indicators of Air Pollution. Environ. Pollut., 19:119.
Salesbury, F. B., and C. Ross. 1969. Metabolism and Functions of Nitrogen and
Sulfur. In: Plant Physiology. Wadsworth Publ. Co., Calif. 747 pp.
Schwartz, C. C., W. K. Lauenroth, R. K. Heitschmidt, and J. L. Dodd. 1978.
Effects of Controlled Levels of Sulphur Dioxide on the Nutrient Quality
of Western Wheatgrass. J. Appl. Ecol., 15:869.
Snedecor, G. W., and W. G. Cochran. 1967. Statistical Methods. Iowa State
Univ. Press, Ames. 593 pp.
Starks, P. B., W. H. Hale, U. S. Garrigus, R. B. Forbes, and M. F. James. 1954.
Response of Lambs Fed Various Levels of Elemental Sulfur, Sulfate Sulfur
and Methionine. J. Anim. Sci., 13:249.
Thomas, W. E., J. K. Loosli, H. H. Williams, and L. A. Maynard. 1951. The Util-
ization of Inorganic Sulfate and Urea Nitrogen by Lambs. J. Nitr., 43:515.
195
-------
Tilley, J. M. A., and A. Terry. 1963. A Two-stage Technique for the in vitro
Digestion of Forage Crops. J. Br. Grassland Soc., 18:104.
Trinkle, A., E. Cheng, and W. Burroughs. 1958. Availability of Different Sul-
fur Sources for Rumen Micro-organisms in in vitro Cellulose Digestion. J.
Anim. Sci., 17:1191 (abstr.).
Whanger, P. D., P. H. Weswig, and J. E. Oldfield. 1978. Selenium, Sulfur and
Nitrogen Levels in Rumen Microorganisms. J. Anim. Sci., 46:515.
Winter, B. J. 1971. Statistical Principles in Experimental Design, 2nd ed.
McGraw-Hill, New York.
Ziegler, I. 1975. The Effect of S02 Pollution on Plant Metabolism. Residue
Rev., 56:79.
196
-------
SECTION 14
CHLOROPHYLL REDUCTION IN WESTERN WHEATGRASS
EXPOSED TO SULFUR DIOXIDE
W. K. Lauenroth and J. L. Dodd
ABSTRACT
Chlorophyll concentrations in the dominant species in
the grassland, western wheatgrass, were examined under ex-
posure to controlled S0£ concentrations. Significant de-
creases in chlorophylls a and b were observed in the absence
of visible necrosis. Chlorophyll a was moire sensitive than
chlorophyll b. Sensitivity of chlorophylls to S02 changed as
the growing season progressed indicating cumulative effects
and interactions with normal senescence.
INTRODUCTION
Early recognition of chlorosis as a symptom of sulfur dioxide exposure led
investigators to suspect interactions of the gas with chlorophyll metabolism.
Experimental work with lichens (Rao and LeBlanc, 1965) and bryophytes (Coker,
1967) confirmed that exposure to high concentrations of S02 results in the
breakdown of chlorophylls to phaeophytin. Gilbert (1968), using a concentra-
tion which one may expect in polluted areas (1.0 pphm 802), found that sensi-
tive lichen species underwent a continual decline in chlorophyll content over
40 days, while resistant species were unaffected. Malhotra (1977) exposed
pine needles to aqueous S02 and found significant decreases in chlorophyll
content with S02 concentrations between 10,000 and 50,000 pphm. Chlorophyll
a was more sensitive than chlorophyll b.
S02 exposure can cause quantitative changes in chlorophyll content. Most
studies rely upon leaf injury to express the amount of chlorosis or necrosis
caused by S02 as well as other phytotoxic pollutants. Knudson et at. (1977)
found a good relationship between chlorophyll reduction and necrosis and
chlorosis for pinto beans exposed to ozone. They concluded that chlorophyll
reduction measured the same aspect of leaf injury as visual estimation but
with more precision and without observer bias. Using chlorophyll content
provides the potential for developing relationships with photosynthesis,
photosynthetic efficiency, and production.
197
-------
We investigated the impact of three controlled S02 treatments, under
field conditions, on the chlorophyll content of western wheatgrass (Agropyron
smithO- Rydb.). Our objective was to determine if chlorophyll destruction
was occurring in the absence of visible necrosis.
MATERIALS AND METHODS
Plant material was collected to allow us to measure chlorophyll contents
of individual leaves as well as entire tillers (live plus dead leaves). Leaf
chlorophyll content was measured for leaves of three different developmental
ages on each date. A sample for leaf chlorophyll measurement consisted of two
leaves representing the same developmental age that had been cut, at the
ligule, from two different tillers. On each measurement date, 10 such samples
were collected for each of the three leaf ages from each S02 treatment. Tiller
chlorophyll content was estimated by collecting and analyzing 10 tillers from
each treatment on 9 July and 20 tillers per treatment on 8 and 18 August.
Chlorophyll analysis procedures followed Knudson, et al, (1977), with the
following modifications. Immediately after collection, leaves were cut in
approximately 2-mm segments and placed in a scintillation vial (approximately
30 ml) containing 100% ethanol. The vials were then kept in the dark for 48
hrs. The contents of each vial were then placed in a blender with an addition-
al 30 ml of ethanol (one rinse) and blended for 1 min. The contents of the
blender were poured into a 250-ml flask, capped, and stored in the dark for
an additional 24 hr. to complete the extraction. The ethanol-chlorophyll solu-
tion was filtered through Whatman No. 540 filter paper, which had been oven-
dried at 60°C for 12 hr. and weighed. Each filter paper was then placed back
in the oven for 12 hr. and the volume of the ethanol-chlorophyll solution
standardized and absorbances were measured at 665 and 649 nm with a spectro-
photometer. Chlorophyll a and b contents were calculated using the equations
presented in Knudson, et at. (1977). Residual dry weights were determined by
reweighing the filter papers, and chlorophyll content is expressed as yg
chlorophyll • mg dry weight"1. Analysis of the change in dry weight of western
wheatgrass leaves after the ethanol extraction indicated losses of 16 to 19
percent with no differential influence of S02 exposure. Chlorophyll accounted
for less than 1 percent of this loss.
Statistical analysis of the results was by analysis of variance with
Tukey's Q values used to compare means (Snedecor and Cochran, 1967). All
references to differences in chlorophyll concentrations in the results section
are statistically significant at the 5 percent probability level.
RESULTS AND DISCUSSION
Analysis of chlorophyll a and b concentrations in leaves of western wheat-
grass exposed to three controlled S02 concentrations resulted in significant
differences due to date, treatment, and leaf age, as well as the two- and
three-way interactions of these factors.
Maximum chlorophyll concentrations ranging from 4 to 9 yg • mg"1 were
measured on the May collection date (Figures 14.1 and 14.2). June through
August chlorophyll concentrations were in the 1 to 3 yg • mg"1 range. Rauzi
198
-------
£
Q.
O
| •
O
Q -value
(P=0.05)
18 June
Control
Low
Medium
High
9 July
I 2 3
LEAF AGE
7 August
2 3
18 August
2 3
Figure 14.1. Chlorophyll a concentration of western wheatgrass leaves
exposed to three controlled S02 treatments.
20 May
Q -value
(P = 0.05)
19 June
2 3
I 2 3
—:— Control
Low
Medium
High
9 July
LEAF AGE
7 August
18 August
Figure 14.2. Chlorophyll b concentration of western wheatgrass leaves
exposed to three controlled S02 treatments.
199
-------
and Dobrenz (1970) reported peak chlorophyll a and b concentrations of oven-
dried western wheatgrass tillers of 1.5 and 1.1 pg • mg"1. Bokhari (1976)
found chlorophyll contents of western wheatgrass plants grown in a growth
chamber to vary from 4 to 6 yg • mg-1 for chlorophyll a and from 1 to 2 yg •
mg"1 for chlorophyll b. We did not encounter differences between chlorophyll
a and b until mid-June, when they amounted to only 1 yg • rag"1. Bokhari (1976)
reported a chlorophyll a:b ratio of three for western wheatgrass over a 100-day
developmental period. We found the ratio to be near one in May. It increased
to a maximum of approximately three in early August and dropped back to two in
mid-August. Sanger (1971) reported similar dynamics for three deciduous trees.
Concentrations of chlorophylls a and b were significantly lower on the May
date in the high 862 treatment for the two oldest leaves than in the control
or other treatments (Figures 14.1 and 14.2). In June, the chlorophylls had
decreased to approximately 20% of the May values. While chlorophyll b con-
centrations were not significantly reduced by S02 after May, significant
differences in chlorophyll a concentrations were found on all except the 18
August sample date. That chlorophyll a was more sensitive to S02 than chloro-
phyll b is in agreement with the findings of Malhotra (1977). He demonstrated
that the decrease in chlorophyll a in pine needles treated with aqueous S0£
was the result of the conversion of chlorophyll to phaeophytin.
On 18 June, chlorophyll a concentration in leaves from control plants was
greater than that measured for leaves from either the medium or high treatment,
regardless of leaf age (Figure 14.1). Chlorophyll a concentration in leaves
from the control was also greater than that for leaves from the low treatment
except for the oldest leaf. In July, leaves from the control treatment have
significantly greater chlorophyll a concentration than leaves from the other
treatments, regardless of age* In early August the only significant differ-
ence that remained was between the youngest leaves from the control and low
treatment. By 18 August chlorophyll a concentrations were similar for all
treatments.
The data for both chlorophylls indicated a pattern that began in May with
similarity among the control and low and medium treatments. By June the medium
treatment was much more similar to the high treatment. In July all three 862
treatments were similar and, in the case of chlorophyll a, much lower in con-
centration than the control. By early August chlorophyll a concentrations had
reached a minimum. In addition, there was a pattern of maximum differences in
chlorophyll concentrations among the oldest leaves at the beginning of the
growing season, indicating a cumulative effect of SC>2. By the middle of the
growing season the differences were found among all leaves with similar magni-
tudes, indicating no cumulative effect. By late in the growing season the
final significant difference is between youngest leaves, suggesting perhaps
inhibition of chlorophyll synthesis.
Analysis of total tiller chlorophyll content (live.plus dead leaves) re-
sulted in no significant effects due to treatments. Combination of these data
with estimates of western wheatgrass tiller density and biomass for the July
sample date after correction for dry matter losses in the extraction procedure
made it possible for us to calculate chlorophyll standing crop (Table 14.1).
Western wheatgrass tiller density (tillers per m2) and biomass (g • m~2) were
200
-------
available from a concurrent investigation of the impact of 862 on grassland
primary production (Dodd et at., 1979). Chlorophyll content on an individual
tiller basis was not very different among treatments, although plants from the
high treatment had the lowest amount of both chlorophylls. Estimates of chlo-
rophyll standing crop indicated 10 to 20 mg • m~2 differences in chlorophyll a
between the control and treatment plots and 7 to 10 mg • m~2 differences in
chlorophyll b (Table 14.1). Taylor and Leininger (1978) reported a gradient
in infra-red reflectance of the SC>2 treated plots that corresponded to in-
creasing S02 concentrations. The observed gradients in chlorophyll concen-
trations per tiller and chlorophyll standing crop were in agreement with the
photographic data.
Heitschmidt et at. (1978) found no visible injury to western wheatgrass
exposed to S02- The only detectable responses of western wheatgrass to 862
were increased sulfur (Lauenroth et at,, in press) and decreased functional
leaf life (Heitschmidt et at., 1978). Chlorophyll analyses from the same
experimental area support those results and demonstrate that biochemical inter-
actions involving SC>2 are occurring within the leaf before their expression as
tissue death (Ziegler,1975). These results complicate the widely accepted re-
lationship between classification of a plant sensitive to S02 and visible leaf
injury (Thomas et at., 1950; Kondo and Sugahara, 1978).
TABLE 14.1. STANDING CROP OF CHLOROPHYLLS A AND B IN WESTERN
WHEATGRASS
Chlorophyll
Concentration a Concentration
Treatment (mg • plant"1) (mg • m~2) (mg • plant"1) (mg • m~2)
Control
Low
Medium
High
0.10
0.11
0.11
0.09
48
43
31
41
0.06
0.06
0.05
0.05
28
22
15
21
Does a reduction in chlorophyll concentration necessarily result in a
reduction in photosynthesis? Malhotra (1977) found that the breakdown in chlo-
rophyll molecules by S02 was reflected in a decreased ability of pine needles
to photosynthesize, as measured by H^COs fixation. This decline in photosyn-
thesis may be related to an inactivation of the functional sulf-hydrol groups
of enzymes required for fatty acid biosynthesis. Glycero-lipids have been
reported as important constituents of chloroplasts, mitochondria, and other
organelles (Macher et at., 1975).
Brougham (1960) found a significant relationship between maximum rates
of dry-matter accumulation of several species and chlorophyll content of the
201
-------
portion of the stand canopy exposed to light intensities greater than 500 f .c.
at noon. Hesketh (1963) reported that chlorophyll content was not an impor-
tant explanatory variable for differences in photosynthetic rates among species.
Ruetz (1973) and Buttery and Buzzel (1977) both found significant positive re-
lationships between leaf chlorophyll content and photosynthesis for a single
plant species. Although one would expect many environmental controls oper-
ating to modify photosynthetic rates of western wheatgrass, it is possible that
a reduction in chlorophyll concentration as a result of SC>2 exposure could re-
duce their photosynthetic capacity.
CONCLUSIONS
Chlorophyll concentrations in western wheatgrass leaves is a sensitive in-
dicator of exposure to S02. Reductions in chlorophyll concentrations may re-
sult in undetectable decreases in net primary production.
REFERENCES
Bokhari, U. G. 1976. Influence of Temperatures, Water Stress, and Nitrogen
Treatments on Chlorophyll and Dry Matter of Western Wheatgrass. J. Range
Manage., 29:127-131.
Brougham, R. W. 1960. The Relationship Between the Critical Leaf Area, Total
Chlorophyll Content, and Maximum Growth-rate of Some Pasture and Crop
Plants. Ann. Bot. N.S., 24:463-474.
Buttery, B. R., and R. I. Buzzel. 1977. The Relationship Between Chlorophyll
Content and Rate of Photosynthesis in Soybeans. Can. J. Plant Sci.,
57:1-5.
Coker, P. D. 1967. The Effects of S02 Pollution on Bark Epiphytes. Trans.
Br. Bryo. Soc., 5:341.
Dodd, J. L., W. K. Lauenroth, G. L. Thor, and M. B. Coughenour. 1979; Effects
of Chronic Low Level S02 Exposure on Producers and Litter Dynamics. In:
The Bioenvironmental Impact of a Coal-fired Power Plant, Fourth Interim
Report, Colstrip, Montana, E, M. Preston and T. L. Gullett, eds.
EPA-600/3-79-049. U. S. Environmental Protection Agency, Corvallis,
Oregon, pp. 384-493.
Gilbert, 0. L. 1968. Biological Indications of Air Pollution. Ph. D. Thesis,
University of Newcastle-on-Tyne.
Heitschmidt, R. K., W. K. Laurenroth, and J. L. Dodd. 1978. Effects of Con-
trolled Levels of Sulphur Dioxide on Western Wheatgrass in a Southeastern
Montana Grassland. J. Appl. Ecol., 14:693-702.
Hesketh, J. D. 1963. Limitations to Photosynthesis Responsible for Differ-
ences Among Species. Crop. Sci., 3:493r-496.
Knudson, L. L., T. W. Tibbits, and G. E. Edwards. 1977. Measurement of Ozone
Injury by Determination of Leaf Chlorophyll Concentration. Plant Physiol.,
60:606-608.
202
-------
Kondo, N., and K. Sugahara. 1978. Changes in Transpiration Rate of SC>2-
Resistant and -Sensitive Plants with S02 Fumigation and the Participation
of Abscisic Acid. Plant Cell Physiol., 19:365-373.
Lauenroth, W. K., D. Bicak, and J. L. Dodd. In press. Sulfur Accumulation in
Western Wheatgrass Exposed to Three Sulfur Concentrations. Plant and Soil.
Macher, B. A., C. P. Brown, T. T. McManus, and J. B. Mudd. 1975. Studies on
Phospholipid-synthesizing Enzyme Activities During the Growth of Etiolated
Cucumber Cotyledons. Plant Physiol., 55:130-136.
Malhotra, S. S. 1977. Effects of Aqueous Sulphur Dioxide on Chlorophyll De-
struction in Pinus aontorta. New Phytol., 78:101-109.
Rao, D. N., and F. LeBlanc. 1965. Effects of S02 on the Lichen Algae with
Special Reference to Chlorophyll. Bryologist, 69:69.
Rauzi, F., and A. K. Dobrenz. 1970. Seasonal Variation of Chlorophyll in
Western Wheatgrass and Blue Grama. J. Range Manage., 23:372-373.
Ruetz, W. F. 1973. The Seasonal Pattern of C02 Exchange of Festuoa rubra L.
in a Montane Meadow Community in Northern Germany. Oecologia (Berl.),
13:247-269.
Sanger, J. E. 1971. Quantitative Investigations of Leaf Pigments from their
Inception in Buds Through Autumn Coloration to Decomposition in Falling
Leaves. Ecology, 52:1075-1089.
Snedecor, G. W., and W. G. Cochran. 1967. Statistical Methods. Iowa State
Univ. Press, Ames.
Taylor, J. E., and W. C. Leininger. 1978. Remote Sensing of the Bioenviron-
mental Effects of Stack Emissions in the Colstrip Vicinity. In: The
Bioenvironmental Impact of a Coal-fired Power Plant, Fourth Interim Re-
port, Colstrip, Montana, E. 11. Preston and R. A. Lewis, eds.
EPA-600/3-78-201, U. S. Environmental Protection Agency, Corvallis,
Oregon, pp. 280-290.
Thomas, M. D., R. H. Hendricks, and G. R. Hill. 1950. Sulfur Metabolism in
Plants: Effects of S02 on Vegetation. Ind. Eng. Chem., 42:2231-2235.
Ziegler, I. 1975. The Effect of S02 Pollution on Plant Metabolism. Residue
Rev., 56:79-105.
203
-------
SECTION 15
RESPONSE OF A GRASSLAND TO SULFUR AND
NITROGEN TREATMENTS UNDER CONTROLLED S02 EXPOSURE
W. K. Lauenroth and J. L. Dodd
ABSTRACT
Interactions of nitrogen and sulfur fertilization and
S02 exposure were examined in a native grassland. Nitrogen
fertilization and its interactions with low-concentration
S02 significantly increased herbage yield. Sulfur and
nitrogen contents, as well as N:S ratios in the dominant
species were altered by the treatments.
INTRODUCTION
In a previous paper we reported both positive and negative impacts of S02
on western wheatgrass (Agropyron smithii Rydb.) (Heitschmidt et al., 1978).
The positive response was an increase in the average number of leaves per
plant as a result of S02 exposure. That plants can utilize S02 as a sulfur
source is widely recognized for plants growing in sulfur-deficient soils
(Faller, 1971; Cowling et al., 1973; Cowling and Lockyer, 1978). Although the
soils in our study area-are not regarded as sulfur deficient (Wight, 1976),
N:S ratios ranging from 15 to 30 were measured on the control plots early in
the growing season.
The negative effect of S02 on western wheatgrass was an increase in leaf
senescence. Other investigators have observed increased leaf senescence as a
result of exposing plants to S02 (Bleasdale, 1973; Bell and Clough, 1973),
which has been explained as the result of toxic concentrations of sulfate
(Malhotra and Hocking, 1976).
This paper reports results of an experiment designed to test hypotheses
about both sulfur deficiency and sulfur toxicity. This was accomplished by
manipulating the soil sulfur and nitrogen concentrations on small plots within
the large plots exposed to S02 treatments (Heitschmidt et at., 1978).
Experimental Design
A split-plot design with replication was used. The main plots were four
S02 treatments (Heitschmidt et al., 1978). The split plots were nitrogen,
204
-------
sulfur, nitrogen plus sulfur, and glucose. The objective of the nitrogen and
sulfur fertilizer treatments was to provide 2 percent nitrogen, 0.2 percent
sulfur, or both in live tissue at peak standing crop. To accomplish this, we
applied the equivalent of 150 kg ha"1 nitrogen (ammonium nitrate) and/or 15 kg
ha"1 sulfur (magnesium sulfate) on 15 April. The glucose treatment consisted
of applying the equivalent of 850 kg ha"1 of ^-glucose. The objective of this
was to reduce the availability of nitrogen to plants by increasing microbial
populations (Behera and Wagner, 1974).
The nitrogen and sulfur treatments were replicated twice within each S02
treatment. Each treatment plot was 1 x 2 m. The glucose treatment was repli-
cated twice in the control and high-S02 treatment only.
METHODS
Standing crop biomass above and belowground was measured at the end of
the growing season by the harvest method. Aboveground biomass was measured by
clipping one 0.25-m2 quadrat in each of the fertilizer or glucose treatment
plots. Plants were clipped at the soil surface and separated by species.
Belowground biomass was measured by removing three soil cores, 7.5 cm diameter
x 10 cm deep, from each harvested quadrat. Crowns, rhizomes, and roots were
separated from the mineral soil by the method of Lauenroth and Whitman (1971).
All plant material was oven dried at 60°C to a constant weight before weighing.
Western wheatgrass was analyzed for total sulfur with a Leco Induction
Furnace (Laboratory Equipment Corp., St. Joseph, MO, USA), and total nitrogen
(Kjeldahl N) by procedures in A.O.A.C. (1965).
Statistical analyses of the results were by analyses of variance with
Tukey's Q values used to compare means (Snedecor and Cochran, 1967). Because
the sulfur, nitrogen, and sulfur plus nitrogen treatments were applied across
all S02 treatments and the glucose treatment was applied only to the control
and high concentration treatment, two separate analyses were performed. The
first analyses included four S02 and four, sulfur/nitrogen treatments. The
second considered two S02 treatments and five sulfur/nitrogen treatments.
RESULTS
Aboveground yield was significantly altered only by the addition of
nitrogen fertilizer (Table 15.1, Figure 15.1). Yields of plots receiving
nitrogen were significantly greater than those of plots not receiving nitrogen,
regardless of S02 concentration. Combinations of nitrogen and S02 produced a
significant interaction, the nitrogen/low S02 treatment producing a significant
increase in yield over all other nitrogen/S02 combinations. Increased S02
concentrations above the low treatment caused significant decreases in yield
compared with the low treatment but no distinguishable differences from the
control. Yields for treatments not receiving nitrogen fertilizer were similar,
regardless of S02 concentration (Table 15.1, Figure 15.1). Results from
analysis of the glucose treatment produced similar conclusions. The reductions
in yield observed on the glucose treatments were not statistically significant.
205
-------
TABLE 15.1. TOTAL ABOVEGROUND YIELD* ON 15 AUGUST (g ' m 2)
Fertilizer Treatment
S02 Treatment Control Sulfur Nitrogen Nitrogen Glucose
plus
sulfur
Control 177 ± 48 120 ± 44 240 ± 64 368 ±4 100 ± 24
Low 156 ±52 152+8 448 ± 32 480 ±40
Medium 104 ±8 120 + 24 352 + 28 328 +32
High 124 ± 24 140 + 32 272 ±72 260 ± 4 72+4
* ± 1 S.E.
400
M
1
o>
o 300
LJ
Q
o 200
o:
O
UJ
I
o:
- A\
/ \^
/ ^^^\
^^^^^
^^
tii i i
<26 52 105 183
S02 CONCENTRATION
m
~3)
Figure 15.1.
Aboveground yield as a function of S02 treatment and nitrogen
fertilization. (Solid line is with added nitrogen, dashed
line is without added nitrogen.)
206
-------
Belowground standing crop was not significantly changed by any of the
treatments. This result was probably as much the result of our inability to
distinguish live roots as of the reduced responsiveness of belowground organs
to perturbations compared with aboveground organs (Dodd and Lauenroth, 1979).
Sulfur concentrations in the aboveground biomass of western wheatgrass
(Figure 15.2) were significantly related to the interaction of sample date,
S02 treatment, and nitrogen treatment. Under nominal nitrogen conditions
sulfur concentrations of plants growing on the medium and high concentrations
treatments increased with both treatment and time (Figure 15.2a). Addition of
mineral nitrogen reversed this trend, resulting in significant decreases in
sulfur content of western wheatgrass between June and mid-August (Figure
15.2b). Treatment with glucose had no effect on the sulfur content of western
wheatgrass under control conditions (Figure 15.2c,d) but significantly
increased sulfur content on the July and August sample dates in the high SC>2
treatment, compared with other amendments.
Nitrogen concentrations were significantly related to interactions of
sample date and nitrogen treatment, as one would expect (Figure 15.3). June
nitrogen concentrations were significantly increased by N fertilization. This
effect diminished throughout the growing season. Mid-August N contents were
the same, regardless of treatment. Glucose had no effect on nitrogen content.
Significant changes in N:S ratios were largely related to time (Figure
15.4). SC>2 treatment decreased N:S ratios as a result of its influence on S
concentration. Nitrogen additions significantly increased N:S ratios on the
final sample date (Figure 15.4b), and S fertilization significantly decreased
N:S ratios early in the growing season (Figure 15.Ac). Additions of glucose
had no influence upon N:S ratios.
DISCUSSION
Sulfur dioxide is widely recognized as a phytotoxic air pollutant
(Ziegler, 1975) but is less widely understood as a potential source of an im-
portant plant nutrient (Cowling and Lockyer, 1978). Because of its unique
position among phytotoxic gases as both a nutrient and toxic agent, any
assessment of the responses of plants to sulfur dioxide must consider both
positive and negative impacts. In a very general sense the relationship
between sulfur dioxide concentration and a measure of plant performance is
expected to be non-linear with a single optimum. This is the subsidy-stress
gradient described by Odum et al. (1979). For example, a plant grown in a
sulfur-free rooting medium and exposed to a range of concentrations of sulfur
dioxide may be expected to achieve a maximum growth rate at a sulfur dioxide
concentration supplying sulfur at a rate very close to the plant's demand. At
higher concentrations toxicity may occur, and at lower concentrations sulfur
deficiency will limit growth. The situation for a plant growing as an individ-
ual in a community is complicated by the large number of variables that can be
expected to modify its response to sulfur dioxide.
The responses illustrated in Figure 15.1 clearly demonstrate a gradient
in responses of aboveground yield of a grassland to a range of S02 treatments.
The important determinant of this response is the presence of adequate mineral
207
-------
2500
2000
T 1500
o>
1000
o:
H
Z
LJ
O
Z
3 3000
15
L_
_l
Z>
2000
1000
3
June
2 15
July
18
Aug
3
June
2 15
July
18
Aug
Figure 15.2.
Sulfur content of western wheatgrass: (a) without additions
of N, (b) with additions of N (control — , low -•- , medium --- ,
high -- ), (c) in the absence of 862 > (d) under high concentration
SC>2 (control — , sulfur added --- , nitrogen added -- —
nitrogen and sulfur added — • — , glucose -- ).
208
-------
2500r
o>
0>
~ 2000
g
<
o 1500
z
o
o
UJ
g 1000
\
\
\
\
3
June
2 15
July
18
Aug
Figure 15.3. Nitrogen content of western wheatgrass over the growing season.
(Solid line is with added nitrogen, dashed line without added
nitrogen)
20
15
a: 10
tn
3
June
2 15
July
16
Aug
3
June
2 15
July
18
Aug
3
June
2 15
July
18
Aug
Figure 15.4. Nitrogen to sulfur ratios in western wheatgrass:
SC>2 treatment (control — , low -•- , medium
(a) related to
high - - ) ,
(b) with (solid line) and without (dashed line) the addition of N,
(c) with (solid line) and without (dashed line) the addition of S.
209
-------
nitrogen. Noggle and Jones (1979) suggested that the presence of S02 in the
atmosphere in the southeastern U.S. has prevented the expression of sulfur-
deficiency symptoms in agricultural crops to which N, P, and K fertilizers are
commonly applied. Our results indicate that, although their interpretation
may be correct, the response for grasslands is sensitive to SC>2 concentration
and is only positive over a narrow range of concentrations (Figure 15.1).
Fertilization studies conducted in the Northern Great Plains with the
objective of increasing range forage production have not shown responses to
sulfur fertilization with or without nitrogen (Wight, 1976) . Our results
document significant yield response to atmospheric sulfur inputs and large but
non-significant responses to sulfur fertilizer when applied with nitrogen
(Table 15.1). Evidence from N:S ratios suggests that, during rapid growth,
sulfur cannot be taken up rapidly enough to meet demands. Apparently,
atmospheric sulfur inputs as S02 can compensate for slow uptake.
Wight and Black (1979) reported average increased forage yields of 43
percent over 8 years of experimentation with the addition of 112 kg ha"1
nitrogen. Yield increases for individual years ranged from less than 25
percent to more than 100 percent. We found a 35 percent increase in yield as
a result of N fertilization alone.
CONCLUSIONS
Because sulfur is an important mineral nutrient in grasslands its poten-
tial actions with nitrogen must be considered in assessing impacts of atmos-
pheric sulfur imputs.
REFERENCES
A.O.A.C. 1965. Official Methods of Analysis, 10th ed. Association of
Official Agricultural Chemists, Washington, D.C.
Behera, B., and G. H. Wagner. 1974. Microbial Growth Rate in Glucose Amended
Soil. Soil Sci. Soc. Am. Proc., 38:591-594.
Bell, J. N. B., and W. S. Clough. 1973. Depression of Yield in Ryegrass Ex-
posed to Sulfur Dioxide. Nature (London), 241:47-49.
Bleasdale, J. K. A. 1973. Effects of Coal Smoke Pollution Gases on the Growth
of Ryegrass (Lolium perenne L.). Environ. Pollut., 5:275-285.
Cowling, D. W., L. H. P. Jones, and D. R. Lockyer. 1973. Increased Yield
Through Correction of Sulfur Deficiency in Ryegrass Exposed to Sulfur
Dioxide. Nature (London), 243:479-480.
Cowling, D. W., and D. R. Lockyer. 1978. Effect of S02 on Lolium perenne
Grown at Different Levels of Sulfur and Nitrogen Nutrition. J. Exp. Bot.,
29:257-266.
Dodd, J. L., and W. K. Lauenroth. 1979. Analysis of the Response of a Grass-
land Ecosystem to Stress, pp. 43-58. In: N. R. French, ed. Perspec-
tives in Grassland Ecology. Springer-Verlag, New York, 204 pp.
210
-------
Faller, N. 1971. Plant Nutrient Sulfur-S02 VS. S0k. Sulfur Inst. J., 7:5-6.
Heitschmidt, R. K., W. K. Lauenroth, and J. L. Dodd. 1978. Effects of Con-
trolled Levels of Sulfur Dioxide on Western Wheatgrass in a Southestern
Montana Grassland. J. Appl. Ecol., 14:859-868.
Lauenroth, W. K., and W. C. Whitman. 1971. A Rapid Method for Washing Roots.
J. Range Manage., 24:308-309.
Malhotra, S. S., and D. Hocking. 1976. Biochemical and Cytological Effects
of Sulfur Dioxide on Plant Metabolism. New Phytol, 76:227-237.
Noggle, J. C., and H. C. Jones. 1979. Accumulation of Atmospheric Sulfur by
Plants and Sulfur-supplying Capacity of Soils. Interagency Energy/
Environment R & D Program Rep. TVA, EPA EPA-600/7-79-109. 37 pp.
Odum, E. P., J. f. Finn, and E. H. Franz. 1979. Perturbation Theory and the
Subsidy-stress Gradient. Bio. Sci., 29:349-352.
Snedecor, G. W., and W. G. Cochran. 1967. Statistical Methods. Iowa State
Univ. Press, Ames.
Wight, J. R. 1976. Range Fertilization in the Northern Great Plains. J.
Range Manage., 29:180-185.
Wight, J. R., and A.L. Black. 1979. Range Fertilization: Plant Response
and Water Use. J. Range Manage,, 32:345^49.
Ziegler, I. 1975. The Effect of S02 Pollution on Plant Metabolism. Residue
Rev., 56:79-105.
211
-------
SECTION 16
EFFECTS OF LOW-LEVEL S02 FUMIGATION ON DECOMPOSITION OF
WESTERN WHEATGRASS LITTER IN A MIXED-GRASS PRAIRIE
J. L. Dodd and W. K. Lauenroth
ABSTRACT
Litter bags were used to determine the effects of S02
and substrate-S on field decomposition rates of western
wheatgrass leaves. In our study atmospheric S02 decreased
seasonal decomposition rates by about 10% while increased
S content of substrate (.098% vs .059% in control) had no
measurable effect. The observed inhibition is probably
due to reduced pH and/or accumulation of toxic derivatives
of S02 in the microenvironment of decay organisms on the
decaying leaf surfaces.
INTRODUCTION
Most aboveground net primary production in native grasslands passes
through the decomposition process at ground level. In ungrazed grasslands
nearly all the aboveground production eventually decays on or near the sur-
face, while in most properly grazed grasslands half or more of the aboveground
production is decayed by surface processes. Volatile products from this
process are lost from the system. Other inorganic and organic products become
directly involved in subsurface nutrient cycling and energy transformation
processes that are critical in maintaining ecosystem stability and primary
productivity. Strojan (1978) reported that decomposition rates for tree leaves
were greatly reduced near a zinc smelter in Pennsylvania. He attributed this
to less biological activity in the leaf litter brought about by increased
litter concentrations of Cd, Cu, Fe, Pb, Zn, and S originating from the
smelting process.
The purpose of this study was to determine the effects of atmospheric S02
on decomposition rates of leaves of western wheatgrass (Agropyvon smithii), the:
dominant native grass of the Northern Great Plains, and whether the rates were
altered by sulfur content of the leaf tissue.
212
-------
MATERIALS AND METHODS
In March of 1978, before initiation of fumigation on the treatment plots,
standing dead western wheatgrass leaves were collected from one location in
the control and high-S02 treatment plots. The leaf tissue collected was grown
in 1977 under control and high 862 treatment conditions. Each source of leaf
material was oven dried (24 h @ 60°C) and subsampled for determinations of
total S (Leco Induction Furnace, Laboratory Equipment Corp., St. Joseph,
Missouri) and for ash content (A.O.A.C, 1965).
A sample of intact leaves weighing approximately 250 mg (ash-free), were
weighed and placed in litterbags (30 x 15 cm) made of 1-mm opening nylon mesh.
One-hundred litterbags of each of three series of samples were prepared.
Series A(A), material grown on the control plot, and series A(D), material
grown on the high-S02 treatment plot, were placed in the control plot. Series
D(D), material grown on the high-S02 treatment plot, was placed in the high-S02
treatment. All litterbags were placed in the field on April 15, 1978. Litter-
bags were laid flat on the ground between clumps of grass. Each bag was
attached to the soil surface with nails to assure close contact of the con-
tents with the soil.
Twenty litterbags from each series were retrieved on each of five dates
(18 May, 16 June, 15 July, 15 August, and 16 September). Contaminating debris
was removed and the contents of each bag were oven dried, weighed, ashed, and
reweighed to determine organic weight. Decomposition rates for the three test
series were expressed as percent of organic weight remaining since April 15.
Daily rates of loss (mg • g"1 • d"1) were computed for each = 30-day interval
of the season from the average percent of material remaining at each date.
Estimates of variance for this statistic could not be computed, because organic
weights of the individual samples at the beginning of each interval were not
measurable (except for the first interval).
Precipitation was measured on the control and S02-treated test areas from
May through August. Differences between test plots were not significant and
averaged 263, 57, 28, and 13 mm for May, June, July, and August, respectively.
RESULTS AND DISCUSSION
The March 1978 sulfur contents (X ± SE) of dead western wheatgrass leaves
grown on the control and S02-treated plots in 1977 were 0.059 ± 0.0035% and
.098 ± .0015%, respectively.
Significant differences in rates of organic weight losses were found
between sample material decayed on the control plot and on the S02~treated
plot (Table 16.1). By the end of the 154-d test period, 76.3% of the leaf
material remained in the S02~treated samples, compared with 66.5% in the
control samples. Differences in organic weight loss due to substrate sulfur
contents at the beginning of decay period were not detected.
Average daily loss rates were greatest between 18 May and 15 July and
least between 15 August and 16 September (Table 16.2). The daily loss rates
for leaf material decayed in the S02 treatment ranged from 56 to 88% of the
rates for materials decayed in the control plot.
213
-------
TABLE 16.1. PERCENT* OF AGROPYRON SMITHII LEAF LITTER WEIGHT
(ASH-FREE, DRY WEIGHT) REMAINING SINCE APRIL 15, 1978
A(A)1" A(D)J D(D)§
18 May
16 June
15 July
15 August
16 September
91.9 ± .7 94.2 ± .4 95.4 ± .6
83.4 ± .9 84.8 ± .5 89.3 ± .8
75.4 ± .9 77.7 ± 1.4 83.7 ± 1.0
70.4 ± 1.3 72.1 ± 1.7 80.4 ± 1.3
66.5 ± 1.3 67.2 ± 1.6 76.3 ± 1.2
* X ± SE.
T A(D) = Tissue grown on high 862 concentration plots and decomposed
on control.
§
D(D) = Tissue grown on and decomposed on high-S02 concentration
plots.
TABLE 16.2. RATES OF LOSS* OF AGROPYRON SMITHII FOR FIVE
INTERVALS DURING 1978
15 April
18 May -
16 June -
15 July -
15 August
- 18 May
16 June
15 July
15 August
- 16 September
A (A)
2.5
3.1
3.2
2.1
1.7
A(D)
1.8
3.3
2.8
2.3
2.1
D(D)
1.4
2.1
2.1
1.2
1.5
mg • g
214
-------
CONCLUSIONS
Our results indicate that low levels of atmospheric S02 inhibits decompo-
sition rates of western wheatgrass during, at least, the first year in the
decomposition cycle. The results also suggest that the inhibition is least
during the latter part of the season, when surface conditions are quite dry.
The decomposition rate probably slows because of reductions in pH and/or
accumulations of toxic S02 derivatives (803, SO^) in the microenvironment of
decay organisms on the decaying leaf surfaces. These changes likely result
from sorption of S02- Since both sorption of S02 on the surfaces and decom-
position rates would normally be greatest under wet conditions, the prob-
ability of inhibition would be greatest early in the season.
REFERENCES
A.O.A.C. 1965. Official Methods of Analysis. 10th edn. Assoc. of Official
Agr. Chemists, Washington, D.C.
Strojan, C. L. 1978. Forest Leaf Litter Decomposition in the Vicinity of a
Zinc Smelter. Oecologia, (Berl.) 32:203-212.
215
-------
SECTION 17
PLANT COMMUNITY STRUCTURE ON ZAPS
J. E. Taylor, W. C. Leininger, and M. W. Hoard
ABSTRACT
Canopy coverage, species diversity, frequency, and
western wheatgrass density were sampled periodically
during the 1978 growing season in an attempt to assess
the effects of sulfur dioxide fumigation on native
rangeland vegetation. The two study plots (ZAPS I and
ZAPS II) have been fumigated during the growing
seasons since 1975 and 1976, respectively. Dif-
ferences in vegetational cover are associated with
inherent inter- and intra-plot variation. Other
differences are demonstrably due to sulfur dioxide
stress. Total living vegetation, total graminoids,
and western wheatgrass show decreasing cover with in-
creased S02 rates. Lichen cover likewise has de-
/ creased significantly. Litter cover has increased on
all treatments for the first three years, probably due
to grazing deferment. ZAPS I shows a significant
litter reduction on the D plot the fourth year.
Neither diversity nor evenness are consistent in-
dicators of stress, although some depression has
been observed. Species richness was decreased with
fumigation when some species were lost and not re-
placed in the composition. Western wheatgrass den-
sities on ZAPS I are significantly lowered by fumiga-
tion.
INTRODUCTION
The objective of this research is to determine whether low-level S02 con-
centration causes changes in vegetation on native rangeland. This question
is being approached through studies of canopy coverage, species diversity,
species frequency, and western wheatgrass (Agropyron smithii) density.
216
-------
MATERIALS AND METHODS
Methodological details are presented in Section 4. The following dis-
cussion is restricted to modification and additions to those procedures.
Plant Community Analysis
Canopy Coverage
Coverage data were obtained following the method of Daubenmire (1959).
Data were collected on the ZAPS sites at four times during the 1978 growing
season: 3-5 June, 16-19 June, 14-15 July, and 9-11 September. These dates
approximated the stages of early growth, peak of cool-season green, peak of
warm season green, and summer dormancy, respectively.
Each of the ZAPS plots was subdivided for sampling, with equal observa-
tions collected from the northern and sourthern halves. These were handled
as "replicates" in the statistical analyses, although they were not randomly
designated.
No samples were taken in microtopographic depressions (run-in moisture
areas) nor in areas less than one meter from a SC>2 delivery pipe.
To select plant community components for detailed analysis, a prelim-
inary data review was accomplished. Several functional groups (perennials
and annuals; grasses, forbs, and shrubs; dominant species) were evaluated as
to their responses to treatment. Total living vegetation, total graminoids,
western wheatgrass, lichens, and litter were found to show significant re-
sponses. Each of these categories was subjected to an analysis of variance.
Significant treatments were examined with Duncan's multiple range test.
These analyses were made for all treatment years.
Diversity
Diversity, evenness, and species richness were calculated at every
sampling date on every plot. The Shannon-Weaver function (H1) was used for
diversity (Shannon Weaver, 1949). Evenness (equitability) was calculated
with Pielou's Index (Jf) (Pielou, 1969). Species richness was taken as the
numerical sum of plant species present in each five-frame group.
Frequency
Frequency data by species were extracted from the plot data used in
cover analysis. These were grouped into five equal frequency classes. The
number of species in each frequency class was plotted in an attempt to
further characterize the evenness component of diversity. Number rather
than percent data were used so that species richness also was accounted for.
June, July, and September data from 1975 and 1978 were compared on ZAPS I
only.
217
-------
Western Wheatgrass Density
At the same time that cover estimates were being made on the 2 x 5 dm
sample plots, numbers of western wheatgrass plants were counted on ZAPS I
only. Plant units as described in Section 4 were used.
Photographic Monitoring
Ground Level Photography
Photoplots were recorded twice during 1978 using color film. Detailed
vegetational charts were developed for each plot. Procedures are explained
more fully in Section 4. New photoplots using both color and color infrared
film were recorded on the fertilizer study plots being conducted by Colorado
State University.
Aerial Photography
The aerial photographic activity conducted on the ZAPS sites is de-
scribed in Section 22.
RESULTS AND DISCUSSION
Plant Community Analysis
In the following discussion, ZAPS I and II are considered separate
studies rather than true replicates. They were established in different
years, received different intensities of grazing pressure before fencing, are
located on different sites, support different plant communities, and have
received different patterns of precipitation. Also, SC>2 distribution charac-
teristics are different because of design modifications in the ZAP system.
Further, the D plot on ZAPS II is not comparable to the other plots. It
falls on an area with numerous microdepressions of dendritic rill form. This
pattern was inconspicuous in the first years of study because pre-treatment
grazing had reduced the plot aspect to a uniformly short sod. Thus, no
sample stratification was employed. With deferment, it soon appeared that
very different plant communities occupied the depressions and the intervening
sites. Therefore, beginning with the 1978 samples, the microdepressions were
avoided in sampling. For this reason, 1978 D plot data are not comparable to
either earlier D plot data or to data from the. other treatment plots. For
one important example, western wheatgrass is less .abundant, in the micro-
depressions. This species is the dominant species and major contributor to
both graminoid and total vegetation categories. Therefore, the D plot cover
data were excluded from the statistical analyses of these groups.
Canopy Cover
Figure 17.1 shows canopy coverage by plant type for ZAPS I in 1978.
Canopy coverage can exceed 100% due to overlapping crowns of different
218
-------
species. The seasonal differences in total vegetation were primarily due to
normal phenologic patterns of vegetational development and to the loss of
annuals at the later sampling dates. The ratio of forbs to grasses did not
change much during the year. Lichens decreased with S02 fumigation. Shrubs
were present in low amounts, and no trends in their cover were noted. Obser-
vation suggested that their distribution was largely a reflection of pre-
treatment patterns. Mosses varied with season, with no correlation between
moss cover and S02-
The overall trends on ZAPS II were similar to those on ZAPS I (Figure
17.2). The period of maximum plant growth appeared earlier, reflecting the
drier and warmer nature of the ZAPS II site. More ephemeral forbs were pre-
sent at the first sampling date. The trend toward decreased coverage as the
season progresses is seen clearly. Graminoid coverage was comparable on all
treatments. Forbs became relatively less conspicuous with the advancing
season, and may also have shown a reduction due to fumigation. Dandelion
(Taraxacum offiainale) was the predominant forb species in the early season.
Table 17.1 summarizes the statistical significance of treatment means
as determined by analysis of variance. Where significant means were en-
countered, Duncan's multiple range tests were made.
TABLE 17.1. SIGNIFICANCE OF TREATMENT MEANS, ZAPS I AND II, 1975-78
Components of ZAPS I ZAPS II
Vegetation 1975 76 77 78 1976 77 78
Western wheatgrass
Graminoids
Lichens
Total living vegetation
Litter
*
AA
A
A*
NS
NS
A
A
A
A
AA
AA
AA
A
A
AA
AA
A
AA
A
AA
NS
A
NS
NS
AA
AA
AA
NS
NS
AA
AA
AA
A
AA
NS = not significant
* = significant (P=.05)
** = highly significant (P=0.01)
Total living vegetational cover is shown in Figures 17.3 and 17.4.
Initially the D plot on ZAPS I had the highest total cover, but by the second
year the control plot had increased about 25%, while the fumigated plots re-
mained at about 1975 levels. This may be a negative treatment response
whereby S02 is inhibiting vigor enhancement from grazing deferment. The
tendency for increased cover on the control relative to fumigation treatments
continues in subsequent years.
219
-------
— - GRAMINOIDS
160
140-
120-
w
g 100-
I 80-
w 60-
40 -
20-
Q FORBS
=
1
=
—
X"!
i§
s
z^s
^
^
3&
§1
00
S
S
=
|
1
?g
S /.
=
=
•-•-
*
:•:•:• LICHENS
. . •
s
=
=
=
^
^
jl""
)8c
§E
^
^
ft
goo MOSS
(1
=
We
i
$^ SHRUBS
=
=
^
^
....
M
=
^
^
S
s
n
'•"•*
M
p. ,
=
=
=
' • •
w
t
^BH
=
M
...
M
=
S
M
=
=
—
W
S
^
^
—
-------
WO-i
130-
120-
UJ
I no-
I 10°-
j* 90-
o
80-
70-
* ZAPS J
--o ZAPS II
"•o b
1975
1976
1977
1978
I ( I I Y I I I ) f I I I I Y I I I*
ABCDABCDABCDABCD
TREATMENT
Figure 17.3. Total living vegetational coverage on ZAPS I and II, 1975-
1978. Means within years and ZAPS not followed by the same
letter are significantly different at the .05 level.
WO-i
130-
g 120-
<->
| 110-
| 100 »
LJ_
fe 90-
i—
a so-
70-
» «ZAPS I
Q---OZAPS !!
T
IST YEAR
2ND YEAR
3RD YEAR
ITH YEAR
A B C D A
CDABCDABCD
TREATMENT
Figure 17.4. Total living cover expressed as a percentage of first year
cover, ZAPS I and II. (Years refer to longevity of SC>2
fumigation.)
221
-------
Total living vegetational cover on ZAPS II showed similar trends, with
less variable treatment responses within years. By the third fumigation
season the B and C plots had significantly less total living vegetational
cover than the control.
The graminoids are the plant component of highest significance to range-
land utilization by livestock and wildlife in the study area. Their cover is
illustrated in Figures 17.5 and 17.6. There is a clear depression in cover
with increased SC>2 on ZAPS I. By the second year, all levels of S02 fumiga-
tion .showed less graminoid cover than the control. For the D plot this dif-
ference was statistically significant every year. By 1978 a significant neg-
ative linear relationship was observed between fumigation rates (Table 17.1)
arid graminoid cover. No clear trends are evident on ZAPS II.
Western wheatgrass is a large contributor to the graminoid trends
Figures 17.7 and 17.8). In the first year of fumigation there was signifi-
cantly more western wheatgrass on the D than on the A plot; B and C were
intermediate. In 1976 there were no significant treatment effects, but there
was a noticeable increase in coverage on the control. By 1978 the A plot
cover was significantly greater than that of any of the fumigated plots.
This was produced by concommitant increases in control plot cover and de-
creases in fumigated plot cover, especially at the higher rates. (Figure
17.8). On ZAPS II the C plot has had significantly less western wheatgrass
cover than A and B every year. This disparity is increasing with time, pri-
marily due to a reduction in C plot. This has potentially serious economic
implications since this species is the predominant forage plant of the area,
and indeed of much of the entire mixed prairie region.
Forbs are the most dynamic component of the vegetation system on the
ZAPS sites. They are highly responsive to annual and seasonal differences
in climate, and are usually the first group of plant species to change in
composition with disturbance, drought, or other ecological factors. When
they are considered as a group rather than as individual species, the situa-
tion is made more difficult because many are relatively short-lived or short-
season plants. Thus, seasonal trends represent the responses of a succession
of species, even when year-to-year composition does not vary. Even so, there
is a tendency for a forb cover to increase less on the higher S02 plots, even
under favorable growing conditions. For example, on ZAPS I, when compared
with the control plot, forb values in 1978 were well below their relative
values in 1975 on the D plot, even though 1978 was the more favorable mois-
ture year.
Annual and biennial species increased dramatically in 1978. This seems
more due to weather variability than to treatment. This group never is a
large part of the composition, and is so variable that it is difficult to
analyze statistically. Also, some are very early ephemeral species and com-
plete their life cycles before fumigation begins each year, so the only S0£
effects they would show would be from sulfur remaining in the soil from pre-
vious years' fumigation. Therefore, the use of these groups as biomonitors
is limited.
222
-------
. 75-
70-
65-
5 60-
O
LJ
= 55-
1 5°-
.D
45-
40-
35-
30-
1975
1976
1977
1978
-rV-, , , rV*-, , , rV~i r
ABCDABCDABCDABCO
TREATMENT
Figure 17.5. Graminoid cover on ZAPS I and II, 1975-1978. Means within
years and ZAPS not followed by the same letter are
significantly different at the .05 level.
150,
140
130
120
§ 110
90]
O
1—
LU
80
70
60
50
40-
^
» "ZAPS I
o---eZAPS II
IST YEAR 2ND YEAR
3RD YEAR ITH YEAR
./i . . . . ^ . .„,.,.
ABCDABCDABCDABCD
TREATMENT
Figure 17.6. Graminoid cover expressed as a percentage of first year
cover, ZAPS I and II. (Years refer to longevity of SC>2
fumigation.)
223
-------
45-
40-
35-
30-
25-
20-
10-
5-
—« ZAPS I
.--eZAPS II
1975
1976
1977
1978
r /(
C
TREATTENT
PI 5
Figure 17.7. Western wheatgrass cover on ZAPS I and II, 1975-1978. Means
within years and ZAPS not followed by the same letter are
significantly different at the .05 level.
140-,
130-
120-
£ 110-
o
<_J
§ 100-
LU
5 90-
80-
70-
60-
50-
40-
» "ZAPS I
<>---«ZAPS II
1:
D
2ND YEAR
—r
5RDYEAR
4THYFAR
B
D A
D
C D
TREATTEJT
Figure 17.8. Western wheatgrass cover expressed as a percentage of first
year cover, ZAPS I and II. (Years refer to longevity of SC>2
fumigation.)
224
-------
One conspicuous phenomenon on both ZAPS sites is an accumulation of
litter (Figures 17.9 and 17.10). This is a product of decreased rates of
decomposition under sulfur dioxide fumigation (see Section 16 ) and grazing
deferment. Even though initial litter cover differed between the ZAPS sites,
both areas showed highly significant linear increases (r = 0.88 and r = 0.94
for ZAPS I and II, respectively) over the first three years of treatment
(Figure 17.9). In the fourth year litter build-up on ZAPS I apparently has
stabilized. Analysis of variance revealed no significant treatment dif-
ferences for either ZAPS. By the third season of fumigation, both ZAPS
showed significantly more litter cover on the D plots than on the controls.
In the fourth year (attained only by ZAPS I), the opposite result was seen.
The decreased litter on the D plot probably is a function of reduced vegeta-
tional production under fumigation.
In 1978 lichen reductions with S02 again were observed (Figures 17.11
and 17.12). This agrees with earlier findings (Eversman, 1978, 1979; Taylor
and Leininger, 1979). Some yearly variations exist, but the trend clearly
remains for lichen cover to decrease under fumigation. After the first
season, the D plot has significantly less lichen cover than the A plot in
every case. On ZAPS II there is a linear decrease. Pretreatment values on
ZAPS I and their subsequent decreases are particularly striking.
This trend is shown more clearly when the lichen cover of 1978 is ex- .
pressed as a percentage of the first year value (Figure 17.13). On ZAPS I
all levels are less than those observed in 1975, with a down-trend associated
with fumigation. The ZAPS II A plot lichen cover still is comparable with
that of 1976, but all others are depressed, though not in as linear a fashion.
It may be that since these plots have received one year less fumigation than
ZAPS I, there has not been enough cumulative effect to stabilize the curve,
or it may be due to the different growth conditions on the two ZAPS sites,
as discussed earlier.
Another way of looking at lichen cover changes is shown in Figure 17.14.
Here, each year's cover is expressed as a percentage of the cover observed in
the first year of fumigation. In 1975, all ZAPS I plots were of course at
100% of that year's levels, yielding a horizontal curve. ZAPS II was not yet
established. In 1976, lichen cover on ZAPS I was below the first year values
on all plots. Except for the anomolous relative increase on the B plot, a
substantial reduction was observed with S02 fumigation. ZAPS II was in its
first year. In 1977, very strong depressions in lichen cover were observed
on both sites. The response curves were very similar. In 1978, ZAPS II
closely traced its curve of the previous year, but on ZAPS I, all plots in-
cluding the A plot were well below first-year levels. It is not known
whether the depression in control cover is due to cumulative low S02 ex-
posures or to different growth conditions than the ZAPS II site experienced
in 1978. Even so, the C and D plots continued to support less lichen cover
than the A plot.
During the past two years, the biological significance of this lichen
loss has been presumed to be as an early indication of ecosystem disruption
which might foreshadow similar changes in more economically important species.
225
-------
90
85
80
75
£70
!
UJ
§60-
55-
50-
45-
40-
» « ZAPS I
o---e ZAPS I!
i 5 i
YEARS OF S02 FUMIGATION
Figure 17.9. Litter cover (means and standard errors) for ZAPS I and II.
85
80
75-
oc.
LU
60
55
50
x ZAPS I
- -• ZAPS 11
1st YEAR
e
2ND YEAR
ab
^a.
abx
•*
ab
3RD YEAR
ITH YEAR
A B C D A B C
D A
TREATMENT
C D A B C D
Figure 17.10.
Litter cover on ZAPS I and II. First year on ZAPS I, 1975,
and first year on ZAPS II, 1976. Means within years and
ZAPS not followed by the same letter are significantly
different at the .05 level.
226
-------
9-
8
7
6
3
2
1
1975
1976
A B C D A
1 A
1977
1978
C D
7APS I
A B C D A
C D
Figure 17.11. Mean lichen coverage for ZAPS I, 1975-1978. Means within
years not followed by the same letter are significantly
different at the .05 level.
81
7-
6-
5-
1-
3-
2
1
1976
ab
1977
\
1978
A B C D A B C D A
ZAPS II
C D
Figure 17.12.
Mean lichen coverage for ZAPS II, 1976-1978. Means within
years not followed by the same letter are significantly
different at the .05 level.
227
-------
o
H
O W
Pi P>
w o
PH O
-------
Such changes now have been observed in several plant groups, as discussed
above. If this trend continues, it will support the use of lichens as impor-
tant bio-monitoring organisms.
Diversity
Diversity Index (H!) values for 1978 are given in Figure 17.15. On
ZAPS I at the first two observations all fumigated plots were significantly
less diverse than the A plot. The low diversity on the B plot traces the
pattern discussed earlier with respect to interplot variation. There is no
obvious trend across the seasons, although the second observation, which co-
incides with the time of maximum plant growth, is a little higher. No con-
sistent SC>2 effects are apparent.
The ZAPS II values are consistently lower than those of ZAPS I, and are
more variable among treatments and dates. If any SC>2 effects are present
they are masked by the inherent intrasite variation.
The evenness and richness values which contribute to the diversity of
ZAPS I discussed above are shown in Figure 17.16. Even though diversity is
more or less level in several instances, evenness and species richness can be
seen to be changing in mutually compensatory ways. Evenness shows no clear
seasonal or treatment trends. Richness tends to peak in mid-June, as is ex-
pected because of the active plant growth at that season. There are no con-
sistent treatment effects, although mid-June and July may exhibit a down-
trend with fumigation. All fumigated plots are significantly lower than the
A plot at these times.
On ZAPS II evenness is much more variable than richness (Figure 17.17).
The high evenness on C plot contributes most of the high diversity observed.
Evenness is significantly higher than any other level in two of the four
cases on C. Richness on D plot is significantly lower than all other levels
at every observation. The B & C levels do not always differ significantly
from the A plot.
Diversity on ZAPS I in 1978 compared to 1975 showed down-trends with
fumigation in June and July, but not in September. September is a poor time
to interpret diversity samples because many species present earlier have dis-
appeared from the composition, leaving mostly the perennial dominants. Sev-
eral of these earlier growing plants could be S02~sensitive. Also, late
season composition varies with years depending upon the late season precipi-
tation patterns. Thus, variation in fall vegetation is more a function of
years than anything else.
Frequency
In a further attempt to characterize the nature of diversity changes
among treatment plots, a frequency analysis was made (Figure 17.18). The
theoretical basis of this analysis is derived from the work of Raunkaier
(1934), who demonstrated a "J-shaped curve" of frequency distribution classes
to be characteristic of many natural communities. Within each plot on ZAPS I,
229
-------
3.8-,
3.6-
3.2-
3.0-
2.8-
2.6-
2.4-
2.2-
3JLNE
i > ZAPS I
e oZAPS II
*l
IS JUHE
M JULY
1DSEPT
BCFA
C D A B C
TREATMENT
A B C D
Figure 17.15. Diversity (means and standard errors) for ZAPS I and II
for 4 sample dates, 1978.
.84
.82-
.80-
.78-
.76-
.74-
.72-
.70-
.68-
.66-
.64-
.62-
.60-
i
r
» *EVEWESS
*---* Rioiess
3 JUTE
18 JLNE
14 JULY
"H4
10 SEPT
21
20
-19
IS
CO
•IBS
or
"1
•14
-13
J2
•11
•ID
-rV-
A B C D" A B C DA B C IT A B O D
TREATMENT
Figure 17.16. Evenness and species.richness (means and standard errors)
for ZAP I for 4 sample dates, 1978.
230
-------
.84-
.82.
.80.
.78.
.76-
.74-
.72-
,70-
.68.
.66-
.64-
.62-
.60-
3 JUNE
18 JUNE
14 JULY
ID SEPT
22
•21
.20
-19
.]&
•Vc-
A B C IK A B C IK A B C DA
TOATOff
C D
Figure 17.17. Evenness and species richness (means and standard errors)
for ZAPS II for 4 sample dates, 1978.
25-
20-
15-
fS 10
* 1-1975
o---01978
OTE FOR FREQUENCY CLASSES:
1= 1-20J
4 = 61-8CK
5 = 81-10CR
12345 12315 12315 12315
A B C D
TREATMENT
Figure 17.18. Number of species within each frequency class, ZAPS I,
July, 1978.
231
-------
the July data were used to construct frequency curves by plotting numbers of
species as functions of percentage of the 2 x 5 dm canopy cover plots in
which they appeared. Frequency percentages were divided into five equal
classes, giving the curves presented.
If diversity changes are affected through either changes in species
numbers or evenness, the shape of the frequency distribution should be
altered. If dominant species are lost and not replaced, class 5 (81-100%
frequency) should lose numbers. If rare or contagiously distributed species
are removed from the composition, class 1 (1-20% frequency) will be depressed.
Finally, if either or both of these circumstances should occur, the inter-
mediate frequency classes may gain numbers. The net result would be a flat-
tening of the whole curve, produced by the diminuation of end points and aug-
mentation of the center.
The analysis is inconclusive, although some of the predicted curve
changes may be happening. In particular, the B plot appears to be responding
as suggested above. In 1979, both ZAPS I and II data from every sampling
date will be examined in this way.
Western Wheatgrass Density
Western wheatgrass densities for 1975 and 1978 on ZAPS I A plot show
little difference. With fumigation, though, almost all levels are signifi-
cantly lower in 1978 (Figure 17.19). This should be expected after the sim-
ilar pattern reported for cover. The ecological and economic significance of
40
38
36
34
32
30
28
26
24
22
20
18
Figure 17.19.
•« H975
o-- -01978
^
1975
1978
ABCDABCDABCD
Western wheatgrass density (mean and standard errors) for
ZAPS I, 1975 and 1978.
232
-------
this change is substantial, since western wheatgrass is the dominant native
forage species in the area, and is similarly important through most of the
Northern Great Plains.
Photographic Monitoring
The ground photo plots were continued in 1978. They were photographed
twice with color film, and charted in detail. The fertilization plots
established by the Colorado State University team were photographed in
stereoscopic color infrared once and color twice.
The aerial photography collected and analyzed over the ZAPS sites is
discussed in Section 22.
CONCLUSIONS
Significant plant community changes have occurred in response to sulfur
dioxide fumigation. Total living vegetation canopy cover on control plots
generally increased, especially during the first three years. This probably
is due to grazing deferment. Concurrently, fumigated plots, particularly the
higher levels, decreased in cover. When 1978 data were compared with first
year values, the control increased and the higher S02 rates decreased in
cover. Graminoid cover was depressed at all fumigation rates. Much of this
change was made by western wheatgrass, one of the most important forage
species for livestock and wildlife in the region. Forbs and shrubs were too
variable to show clear trends, but lichens were reduced by S02• Litter accum-
ulated linearly for the first three years, apparently due to deferment and
perhaps inhibited decomposition under fumigation. Litter cover leveled off
the fourth year.
Neither diversity nor evenness were consistently sensitive indicators
of fumigation stress, although some observations showed depressed values, at
least at the higher S02 rates. Species richness was almost always lower on
S02 plots.
Densities of western wheatgrass on ZAPS I in 1978 were significantly
lower than 1975 levels on fumigated plots. This has considerable economic
and ecological significance because of the importance of this plant as a
forage species.
REFERENCES
Daubenmire, R. F. 1959. A Canopy-Coverage Method of Vegetational Analysis.
Northw. Sci., 33(l):43-64.
Eversman, S. 1978. Effects of Low-Level S02 Stress on Two Lichen Species.
In: The Bioenvironmental Impact of a Coal-fired Power Plant, Third
Interim Report, Colstrip, Montana. E. M. Preston and R. A. Lewis, eds.
EPA-600/3-78-021, U. S. Environmental Protection Agency, Corvallis,
Oregon, pp. 385-398.
233
-------
Eversman, S. 1979. Effects of Low-Level SC>2 on Two Native Lichen Species.
In: The Bioenvironmental Impact of a Coal-fired Power Plant, Fourth
Interim Report, Colstrip, Montana. E. M. Preston and T. L. Gullett, eds.
EPA-600/3-79-044, U. S. Environmental Protection Agency, Corvallis,
Oregon, pp. 642-672.
Pielou, E. C. 1969. An Introduction to Mathematical Ecology. J. Wiley &
Sons, N. Y. 286 pp.
Raunkaier, C. 1934. The Life Forms of Plants and Statistical Plant
Geography, Being the Collected Papers of C. Raunkiaer. Clarendon Press,
Oxford.
Shannon, C. and W. Weaver. 1949. Mathematical Theory of Communication.
Univ. Illinois Press, Urbana. 117 pp.
Taylor, J. E. and W. C. Leininger. 1979. Plant Community Changes Due to
Low Level S02 Exposures. In: The Bioenvironmental Impact of a Coal-
fired Power Plant, Fourth Interim Report, Colstrip, Montana. E. M.
Preston and T. L. Gullett, eds. EPA-600/3-79-044, U. S. Environmental
Protection Agency, Corvallis, Oregon, pp. 610-641.
234
-------
SECTION 18
BEHAVIORAL RESPONSES OF SAPROPHAGOUS
AND NECROPHAGOUS BEETLES
EXPOSED TO SULFUR DIOXIDE
J. J. Bromenshenk
ABSTRACT
In 1976 and 1977, investigations at the Zonal Air
Pollution System field study sites repeatedly demonstrated
reduced numbers of saprophagous beetles captured in pitfall
traps on plots treated with low level sulfur dioxide
fumigation. This was particularly true for the tumble
bug Canthon, mostly laevis Drury (Scarabaeidae). Pitfall
trappings were again conducted in 1978 and 1979. In
addition, experiments were initiated to determine whether
the decreased abundance of beetles on the sulfur dioxide-
treated plots was due to mortality, behavioral responses,
or a combination of factors. Furthermore, quantification
of the kinds and amounts of organic matter shredded and
consumed by Canthon laevis was undertaken to estimate
the effect of reduced numbers of these beetles on rates of
decomposition, nutrient cycling, and regulation of dung
and carrion-inhabiting pest insects. Preliminary evidence
indicates that the observed low numbers of beetles on the
sulfur-treated plots are due primarily to a behavioral
reaction rather than to mortality. These scavanger beetles
appear to be important removers of carrion and contribute
significantly to the rapid removal of dung from grasslands
of the Northern Great Plains.
INTRODUCTION
Data obtained in 1976 and 1977 at the Zonal Air Pollution System
(ZAPS) study plots demonstrated that chronic low levels of sulfur dioxide
(S02) significantly affected distributions of several different species and
families of grassland beetles (Bromenshenk, 1979, 1978; Leetham et al.,
1979). Necrophagous, coprophagous, and predatory beetles reacted to sulfur
dioxide, and tumble bugs or dung beetles of the Canthon group seemed to be
particularly responsive. Repeatedly, more Canfhon, mostly laevis, beetles
were captured on the control plots than on any of the sulfur dioxide-fumigated
plots. This depressed abundance may have been due to: (1) disorientation,
235
-------
avoidance, or otherwise altered direction or rate of locomotor activity;
(2) natality or mortality of resident or transient populations, or (3) a
combination of life table and behavioral factors. In 1978 and 1979, the
primary objective of my research was to discover why fewer captures were made
on the sulfur dioxide-treated plots.
Canthon laevis is considered to be a dung feeder, but observations at
ZAPS indicated that these beetles also are voracious carrion feeders.
Furthermore, the effective density of these insects in prairie grasslands is
considerable. Typically, during ten-day trapping intervals from June to
early August, hundreds to thousands of these beetles were captured on each of
of the eight 0.62 ha plots.
Dung and carrion beetles are common in the grasslands of the Northern Great
Plains, yet there is little information on their ecological roles in prairies.
The major role of saprophagous invertebrates is assumed to be their contribu-
tion to the decomposition of organic material which in turn effects nutrient
cycling. These organisms appear to function as regulators or catalysts of
rates of decomposition by consuming, shredding, and burying organic matter, by
altering chemical processes associated with decomposition, and by promoting
the activities of microbial decomposers. However, few investigators have
attempted to quantify the influence of these insects on decomposition rates.
Coprophagous and necrophagous beetles also compete for food and space with
insects harmful to livestock and game animals, such as flesh flies and horn
flies, and with potential transmitters of human disease such as muscid flies.
In addition, dung and carrion-inhabiting beetles often carry mites and other
insects that attack fly eggs. Other functions served by saprophagous beetles
include less wastage of pasturage because of dung or carrion fouling and
improving soil structure. They are also a food source for other organisms.
Birds and amphibians may be their major predators. Australian investigators
report that frogs, toads, magpies, starlings, plovers, crows, and other
insectivorous birds feed on dung beetles (MacQueen, 1975; Rowley and Vestjens,
1973; Vestjens and Carrick, 1974; Green, 1966).
Verification of these functions, identification of other functions (if
any), and quantification of arthropod-mediated rates of decomposition became
major objectives of my 1978 and 1979 ZAPS research. Although these insects
exhibit a response to sulfur dioxide, the significance of this response to
ecosystems over the short and the long term remains to be determined. Also,
beetles identified as being sensitive to sulfur dioxide represent only a few
members of the .invertebrate fauna associated with the biocenology of dung
and carrion. Many other members of these communities have not been examined
with regard to the effects of sulfur dioxide.
Feces and Carrion as Ecological Units
Feces and carrion each support a large and varied faunal community. Hayes
(1927, 1929) commented, "One of the most interesting groups of prairie
insects is the savenger group which originally derived its sustenance from
droppings of the buffalo, but now thrives on the excrement of domestic cattle
and horses which have replaced the buffalo on the prairies." Allee et al.s
236
-------
(1949) remarked:
"A neglected microsere is that of carrion. Decomposing bodies
of fishes washed ashore, and the remains of dead reptiles, birds,
and mammals, are especially well suited for research in this
connection. Associated with changes in the chemistry of the flesh
are numerous problems involving bacterial activities, carrion
biocoenses, and the micoseral succession of the carrion fauna."
Hayes could not have known of the rapid spread and colonization success in
dung of Hydrophilids of the genus Sphaeridiion which are now very common in
dung throughout the United States and Canada. These beetles apparently were
introduced to the east coast of North America from Europe during the late
1800's or early 1900's (Brown, 1940). The concerns of Allee et al. have been
addressed by Payne (1965) and others who have examined the carrion micro-
community.
Hammer (1941) cited Portchinsky (1885), a Russian scientist, as the first
to consider cow droppings as ecological units, characterized by sharp competi-
tion among the insects inhabiting the dung and by predation of many insect
species upon others. Hammer (1941) carried out extensive studies in Denmark
concerning the biology and ecology of insects associated with cattle excre-
ment, focusing his attention on Diptera (flies) but also examining Coleoptera
(beetles), Hymenoptera (wasps), Collembola (springtails), and some non-insect
arthropods.
Mohr (1943) seems to have been the first United States scientist to
seriously consider cattle dung as an ecological unit. He conducted a study
which still stands as one of the most detailed investigations of the entire
insect complex of cattle droppings. Mohr documented the species of insects
present and their succession, studied the physical changes that occurred in
droppings from weathering and drying, examined competition and other relations
among the flies, beetles, and parasitic wasps, and compared these phenomena at
sites on sunny dry hillside pastures and on moist land in shady swamps in
Illinois. He regarded bovine dung as a structural unit of prairie ecosystems
and concluded that dung inhabitors apparently are "...not as common as general
feeders on grass, weeds or shrubs, but are probably more common than many
insects more often caught, but dependent on some specific grass, weed or shrub."
More recently, Sanders and Dobson (1966) determined the species, habits,
and inter-relationships of insects in manure pads in a blue-grass pasture in
Indiana. Coffey (1966) employed sites in Idaho and Washington to examine the
potential for flies capable of carrying human disease to develop in cow, horse,
sheep, chicken, swine, dog, mink, and human feces. Several studies in the
United States, Canada, and Australia have examined supression of horn and
flesh flies by populations of competitive or predatory insects inhabiting
dung (Harris and Oliver, 1979; Blume, 1970; Blume et al., 1970; MacQueen and
Beirne, 1975; Thomas, 1967; Thomas and Morgan, 1972). Interest in dung beetles
was engendered in Australia when cattle but not dung fauna were introduced.
Problems developed such as excessive build-up of dung, epidemic populations of
flies and parasitic worms, and excessive pasture wastage because of fecal
fouling. In an attempt to alleviate these problems, dung beetles were
237
-------
introduced to Australia and intensively studied (Gillard, 1967; MacQueen, 1975;
McKinney and Morley, 1975; Durie, 1975).
Currently, researchers interested in saprophagous invertebrates have
reached the point where attempts are being made to quantitatively model the
role of coprophagous fauna in the decomposition of organic matter. McKinney
and Morley (1975) modelled dung and pasture interactions to assess the effect
of dung beetles on the amount of area of unfouled pasture accessible to
grazing, on soil moisture, and on nutrient storage, especially of phosphorus
and nitrogen. Studies by Olechowicz (1974, 1976, 1977) and Andrzejewska (1974)
on decomposition rates of dung, the contribution of invertebrate fauna to
this process, the amounts of dung deposited, green biomass production, and
sheep consumption have been used to construct a diagram of energy flow through
the main food chain of sheep grazing a mountain pasture. The conclusion of
these studies was that coprophagous animals play an important part in the
rapid decomposition of sheep feces and its "fast recycling," After 30-day
exposure periods, an average of 35.3% of the energy in dung dissipated.
Invertebrates accounted for 61.4% of the dung loss, while the microorganisms
processed 38.6%. In other words, dung accessible to invertebrates had a
decomposition rate twice as high as that of dung subject only to the action
of microorganisms. Whether the same was true of cattle droppings in grass-
lands of the Northern Great Plains became the pragmatic basis for my 1978-79
ZAPS studies.
MATERIALS AND METHODS
Pitfall Traps
In 1978, I continued population studies using pitfall traps arranged in
grid arrays across the treatment plots. Each consisted of a cone-shaped,
disposable, plastic coffee cup set into 14 oz. plastic beverage cups and dug
into the ground so that the lip of the coffee cup was flush with the soil
surface. The traps are based on the design of Merrill, (1975) and modifications
by Olson e~t al., (1976) and Bromenshenk, (1978). An inch of water in the
bottom of each trap drowned captured beetles, prevented predaceous species from
devouring each other and other captured insects, and kept the bait moist. All
traps were baited with five grams of meat. Baiting increased trapping success
for Scarabaeids and Silphids and resolved an over-trapping problem caused
by traps containing beetles acting as attractants as compared to unbaited traps
without beetles. Each grid formed a 7 x 7 square, the perimeter traps located
outside the pipeline system of each treatment plot (Figure 18.1). Thus, at
each plot there were 49 traps, 25 within the gas delivery lines, 24 outside
of the delivery lines or 392 traps in all. The main purposes of using these
traps in 1978 were verification of the 1976 and 1977 findings and capture-
recapture studies. Tending the traps and conducting counts of captures was
very time consuming. Therefore, in 1978 traps were used only in mid-summer
(July) at a time when the beetle populations, based on our previous findings,
would be at peak levels of abundance.
The use of pitfall traps and the baiting of pitfall traps is controversial
among investigators. Ttie main concern is that differences of species suscep-
tibility to capture, attraction of animals from a distance, and trap shielding
238
-------
by dense vegetation seriously limit the ability to obtain accurate estimates
of absolute population density. However, pitfalls are generally thought to
be useful for studies of relative abundance, movements, spatial patterns of
distribution and diel activity patterns, if the plots occur in similar vege-
tation. This sample method is simple, easy to use, relatively inexpensive,
and may not have any more problems than other methods such as direct quadrant
counts, total population per area, or mark-capture (Kulman, 1974). More
detailed discussions of the advantages, disadvantages, and limitations of
pitfall trapping can be found in Southwood (1975), Kulman (1974), Thomas and
Sleeper (1977), and Bromenshenk (1979).
X7 X8 2. X22 35 X36 X49
/\ ^
^mm
~x
10
~x
II
13
X
•MM
X9
tm^tmmm
19
*
12
X_
x20
•^H
X
24
X|Q
•••
17 X
'•x 27x
X23
^•M*
33V
X
25
X_
26
X
•^^
•^^•i
x34
•••••M
X
38
•^H*
^32
^^^
Jx3' 4,x
X37
^I^B
x
39
40
X
^^•M
14
28
42
15
29
48
.47
46
45
44
43
Figure 18.1. Trap locations for each treatment plot.
239
-------
Shubeck Traps
In 1979, a new trap design was utilized. Separating beetles from a
slurry of decomposing meat in water was putrid, and mark-recapture tests
conducted in 1978 indicated that markings with paint peeled off of the beetles
in the traps. Shubeck's (1968) trap was a ground-surface trap containing the
bait in a cup set into a one-gallon can which was capped by a wide mesh hard-
ware cloth and housed in a wooden box with a canopy roof to keep out rain.
Thus, the trap was relatively clean, dry, and efficient to use. Shubeck (1976)
reported good success in collecting beetles representing several families:
Silphidae, Scarabaeidae, Staphylinidae, Histeridae, Leiodidae, and Nitidulidae.
But for my purposes, this trap was considerably more complex and expensive
than the pitfalls (@ $5.00/trap versus 4.5c/trap), especially when large
numbers of traps were to be used. . Rain was a major problem to Shubeck but of
little importance at the ZAPS, since during mid-summer generally the only
precipitation was from infrequent thunder-showers. Therefore, I utilized a
modification of Shubeck's original design (Figure 18.2). A plastic food
container (pail) was fitted with a wire mesh screen nailed to a wood frame.
The frame was attached by an L-shaped, V1 bolt and wing nut to a steel rod,
in order to secure the trap from being knocked over or removed by coyotes or
wind. A beverage cup set into the pail held the bait.
CLAMP DETAIL
-BENT BOLT
HARDWARE CLOTH
WOOD FRAME
PLASTIC BUCKET
6.35mm (1/4") WING NUT CLAMP
4.76mm (3/16") STEEL ROD
CUP WITH BAIT
Figure 18.2. Modified Shubeck trap.
One of these traps was placed beside each of the poles used by EPA to hold
sulfation plates for an extensive study of the general relative concentrations
of sulfur dioxide at different horizontal and vertical locations on the plots
240
-------
(Preston and Gullett, 1979). Each pole held sulfation plates in clips at
four heights. One pole was located in the center of each quadrant of each
treatment plot and each internal plot. A fifth pole was placed in the geo-
metric center of each internal plot. The fifth trap on each treatment plot
was placed 2 m from the geometric center of the plot to avoid placement
directly under a gas delivery line (Figure 18.3). This differed from the
location of the fifth pole which was placed next to the real-time sulfur
analysis sample position c. Two posts and traps were located in the buffer
zones at each end of the series of plots.
A-B B
B-C
2.
•
«
2. .3
5.
C
1. .4
2. .3
5.
1. .4
2. .3
5.
C
1. .4
2. .3
5.
1. .4
2. .3
5.
c
1. .4
C-D
2. .3
5.
I. .4
2.
5.
.3
2.
I.
Figure 18.3
Location of Shubeck traps for beetle capture on the
extensive grid.
Mark-Recapture Tests
During 1978 and 1979, beetles of Canthon laevis were marked and released
on the ZAPS plots. The main objective of the 1978 test was to monitor move-
ment as regards direction and distance. The intent for 1979 was to conduct
life-table tests using mark-recapture techniques,but until late in the summer
population levels were so low that both total numbers and recapture success
were too small for meaningful interpretation of the results. Beetles were
marked in three ways: (1) dots of acrylic paint on the thorax, (2) tarsal
clipping, and (3) branding of the thorax with a hot needle.
Paint markings were rapidly lost probably because the beetles are strong
diggers, descending below the soil surface to avoid temperature and other
climate extremes and to bury their dung balls. Thus, the paint is literally
scraped off. In pitfall traps, water dissolved the paint. Tarsal clipping
was permanent, but extremely difficult to see, especially in traps containing
hundreds of individuals. Branding also was permanent but has a greater chance
of injury. Paint appeared to suffice for short-term movement evaluations, if
dry traps such as the Shubeck design were utilized.
Mouse Carrion Study
In 1978, a study of the decomposition of mice was conducted at ZAPS.
Mice were placed in two types of cages on each of the ZAPS treatment plots
and at an additional control plot near the Rocky Boy Reservoir about two miles
from either ZAPS site. The first type cage consisted of a bottomless box
241
-------
of galvanized hardware cloth (1.2 cm spacing of the mesh wire) with long tabs
at the sides which were dug into the ground to secure the cage. Each cage was
rectangular and just large enough to contain a mouse. Thus, each mouse carcass
rested on the surface of the soil with hardware cloth covering it. Mice in
these cages were more or less protected from mammals and birds but still
accessible to the beetles which could easily pass through the mesh. The
second type of cage consisted of a metal window screening (2 mm spacing of the
mesh) formed into a rectangular box just large enough to hold a mouse. These
cages totally enclosed the carcasses and all seams were folded over and
stapled for added security. The mesh size was small enough to exclude
macroarthropods. After an early season test at ZAPS, it became apparent that
many small arthropods still had access to these carcasses. In later tests,
each carcass was wrapped in four layers of cheesecloth before being put into
the screen cage.
Two tests of carrion decomposition were carried out in 1978. Each
utilized 90 mouse carcasses. In the first test, the mice were obtained in
June from the Western Energy's mine reclamation areas near Colstrip. Two
nights of snap-trap captures were carried out. Captured mice were removed
from the traps just after sunrise, and the bodies frozen until used. Ninety
deer mice (Peromysaus maniaulatus) comprised of 45 females and 45 males were
selected for the ZAPS test. Each frozen mouse was weighed and the fresh
weights recorded. Specimens of each sex were arranged in order by weight,
then segregated and ranked into five weight categories. Using a table of
random numbers, nine male and nine female mice were drawn from each weight
grouping and assigned to a treatment plot. This provided five females and
five males for each plot. Sets of ten mice were separated into male/female
pairs, each individual of a pair of similar weight. Thus, for each treatment
there was one pair of small mice, one pair of somewhat larger size, and so
forth. The series ended with a pair of the largest mice. One individual of
each pair was placed into a screen cage, the other into a hardware cloth cage.
The choice of which sex to place in which cage was again randomized using the
table of random numbers.
As mentioned before, on each treatment plot, there are five posts for
holding sulfation plates arranged in an x-shaped pattern. One pair of caged
mice was placed as near as possible to each post. The selection of which of
the five mouse pairs to put by a post was made from the random numbers table.
Thus, at each post was placed a male/female pair of mice of similar weight,
one individual in a screen cage and the other in a hardware cloth cage,
Thus, weight and sex factors were thoroughly randomized for each treatment.
The cages were examined frequently for several days following placement on
the plots and observations made concerning insect activity and decomposition.
After 60 days the carcasses or whatever remained of the mice were retrieved
from the plots, the debris cleaned away, the remains oven dried for 97 hours
at 50°C, and weights recorded. The soil under each hardware cloth cage was
removed to a depth of 30 cm and sifted through a fine mesh screen in order
to remove any buried material. A second set of 90 mice was utilized in a
test begun in September, 1978. It differed from the June test in four ways:
(1) white laboratory mice instead of deer mice were utilized, (2) mice were
not segregated on the basis of sex, (3) carcasses placed in the screen cages
were wrapped in cheesecloth, and (4) carcasses were exposed for 12 months.
242
-------
White mice were used because they were much more readily obtained than deer
mice. The mice were not separated by sex because 45 of each sex of mice were
not available and because no sex-related effects had been observed during the
June test.
Bovine Feces Decomposition
In 1979, a study of the decomposition of cattle feces was begun at ZAPS.
Based on the experiences with the 1978-79 carrion tests, it became apparent
that it is difficult if not almost impossible to effectively exclude scavenger
arthropods. Powerful shredding and biting mouthparts, digestive enzymes, and
efficient strategies of obtaining their food when difficult to reach lead
eventually to at least the smaller arthropod species gaining access. Watching
Canthon laevis on cow pats made it apparent that their feeding and ball rolling
activites could be easily recognized by a visual examination of the dropping.
Dung loss attributable to these beetles can be estimated by the area of the
pat disturbed 'or removed. Since the primary objective was to determine if the
observed reduction in the numbers of Canthon laevis on the sulfur dioxide
treated plots would alter rates of decomposition and since virtually nothing
was known about the responses of the total decomposer communities of dung to
sulfur dioxide or whether a change in the numbers of Canthon laevis might be
offset by concurrent changes in the population dynamics of their competitors,
I elected to concentrate on the responses of the entire decomposer community
to sulfur dioxide. To accomplish this, 16 standardized cow pats were placed
on each plot at the ZAPS in early and late summer, 1979.
Fresh bovine droppings were obtained from a concrete pad of a holding pen
of a dairy operation at Billings, Montana. The concrete was washed down before
the cattle were admitted, the cattle held for two-hour periods and then released,
The fresh dung was shoveled into 30-gallon garbage cans lined with plastic bags.
The cans were then sealed and the manure transported by truck to the ZAPS. At
ZAPS, each can was thoroughly stirred using a long-handled shovel. Then 14 oz.
disposable, plastic beverage cups were filled level with the top with dung. A
trowel was used for the dung transfer to the cups. Sets of 18 cups were filled.
From these,two were randomly selected for immediate weighing (fresh weight)
and subsequent oven-drying and weight determinations. From the remaining 16
cups, groups of four cups were drawn randomly and assigned to one of the four
treatment plots at each ZAPS. While one individual filled and weighed cups,
another carried the groups of four to the plots and set them on 4 x 4 grids,
laying a line of four cups in an east-west direction on each plot. After one
set of four had been placed on each plot, the procedure was repeated until 16
per plot had been delivered. The grids were marked off before filling the cups
by placing a 24.2 cm paper plate at each grid intersection point. The contents
of the cups were emptied onto the plates and spread with a trowel into a
symmetrical pat approximately 3.5 x 20 cm (Figure 18.4).
Insect activity was observed immediately following placement of the dung
on the ZAPS. In July, the number of insect-produced holes in the top and
bottom of each pat, the area removed, and the field dry weights were recorded
for the two early season trials. The "dry weights" were obtained in the field
in late July on hot, sunny days with the temperature over 24°C and the relative
humidity less than 30%. All dung pats were replaced in the original positions
243
-------
after performing the counts, area determinations, and weighings. A third set
of dung pats was placed on the ZAPS in September. A subset of these pats was
enclosed in mesh cages to exclude the larger invertebrates. All pats will be
removed following shutdown of the gas delivery in October, 1979. These will
be transported to the EVST Laboratory, oven-dried, and final measurements
conducted. Two-way ANOVA (analysis of variance) of the field dry weights,
emergence and entry holes, and area of dung removed will be carried out.
Figure 18.4. Standard cow pat on paper plate.
Besides investigating sulfur dioxide effects on rates of dung decomposi-
tion, it was hypothesized that the dung pats might provide a means of assessing
the activities of insect population by the number of feeding and emergence holes
in the pats. It was hoped that this might provide a useful tool analagous to
moss bags.
Tube Test of Beetle Movements and Orientation
Fewer numbers of beetles on sulfur treated plots may be due to a behavioral
response such as disorientation, altered rates of locomotor activity, impairment
of ability to find dung or carrion, or directed movements such as caused by an
irritant or a repellant. In order to examine this in a field setting, 12
cylindrical wire mesh tubes (0.3 cm spacing of wires) 9.5 x 4.57 cm were con-
structed. An entry was provided at the middle of each tube. The tubes were
marked off into five equal-length sections.
244
-------
The use of the tubes was as follows : Sets of tubes were placed on the
A and D plots at either ZAPS I or ZAPS II. For each test, the tubes on A
and D were similarly placed with respect to the gas delivery lines, sun,
slope, and wind. An observer was situated near the tubes on each of the two
treatments. At precisely the same time, each observer would release 20
beetles into each tube and then record at five minute intervals for 30 minutes
the position of each b.eetle. At the end of 30 minutes the beetles were
removed from the tubes. Tests in which sulfur dioxide was excluded as a
factor were conducted off the plots. By various arrangements and orientations
of the tubes, factors such as slope, light, food odors, wind, and sulfur
dioxide could be examined individually and in combinations. Obviously, this
procedure was best suited to investigating tactic responses rather than kinetic.
Kinetic experiments were also carried out at ZAPS in 1979. They will be
described in the Sixth Interim Report, and only preliminary results of the
tube tests will be given here.
RESULTS
The results of the 1978 investigations are presented here, as are some
of the 1979 results which either are a continuation of the 1978 work or
• preliminary information from the 1979 season which facilitates discussion
of the hypotheses and objectives.
Summaries of the captures of all species of beetles in pitfalls at
ZAPS I and II for June-July, 1978 appear in Appendices 18.1 through 18.9.
Captures of Canfkon laevis for the ten-day period totalled 28,076 beetles,
2.7 times greater than for any prior sample period. Total capture for ZAPS I
exceeded peak season captures for 1977 by 4.76 times, while total captures
at ZAPS II were 1.25 times that of 1977. In 1976 and 1977, captures at ZAPS I
were always lower than at ZAPS II. As in previous years, capture success on
all of the sulfur dioxide treated plots as compared to the "controls"
(A plots) was significantly lower (p £ 0.005) (Figure 18.5) by Chi-Square
Goodness of Fit Analysis (Table 18.1). This was true for total captures,
perimeter captures, and interior captures. On ZAPS I, there was no signi-
ficant difference for interior captures on the B or C plots as compared to D.
1000 2000 3000 4000 5000
NUMBER OF BEETLES / 196 TRAP DAYS
6000 7000
Figure 18.5. Pitfall captures of Canthon laevis
at ZAPS I and II, June-July, 1978.
245
-------
TABLE 18.1. CHI-SQUARE GOODNESS OF FIT FOR PITFALL CAPTURES OF
CANTHON SP. AT ZAPS I AND II, 1978
Total
Plots p £ .05 p £ 0.005
ABCD 958.4
AB 415.8
AC 677.1
BC 16. 4
BD 14.1
CD 3.9
AD 580.3
BCD 34.3
ABCD 1045.4
AB 427.6
AC 904.1
BC 101.0
BD 25.8
CD 226.0
AD 248.0
BCD 225.6
Perimeter
N.S. p £ 0.05 p £ 0.005
ZAPS I
703.0
222.5
549.5
75.4
27.6
11.9
403.5
77.4
ZAPS II
741.3
190.8
729.4
191.4
17.6
319.6
93.6
324.8
Interior
N.S. p < 0.05 p £ 0.005
299.1
207.8
148.7
6.6
1.3
2.03
176.9
6.6
629.5
398.4
176.6
64.0
15.0
.18.3
289.4
39.8
Although there were significant differences in the numbers of other specie'.s
of beetles for plots at ZAPS I and II as indicated by Chi-Square Goodness of
Fit Analysis, none of the differences correlated with any discernible sulfur
dioxide treatment effect. However, for the most part, captures of beetles
other than Canthon species were relatively low.
Preliminary tests of the Shubeck traps in June, 1979 produced numerous
captures of Silphids of the genera Nicrophorus and Silpha but almost no other
insects. Setting the traps into the ground flush with the soil surface
immediately resulted in captures of Canthon species and of other species of
beetles which frequently .had been caught in the 14 oz, pitfalls. Results of
the 1979 trapping will be presented in the Sixth Interim Report, 1980.
Mark-Recapture
Mark-recapture tests were conducted in 1978 during the June-July trapping
interval. Additional tests were conducted in 1979, the results of which will
be presented in the Sixth Interim Report. In 1978, 528 Canthon laevis were
marked with acrylic paint and released on ZAPS I and II—400 at ZAPS I and 128
at ZAPS II. Beetles were obtained from the pitfalls on the second day of the
ten-day trap interval, marked, released on the third day, and total recaptures
246
-------
after seven days (tenth day of the trapping interval) determined. The released
beetles were color-coded and tarsal-clipped by plot. Forty-eight hours after
capture, two sets of beetles were released on each plot at two release points.
Each set was designated by a distinctive pattern of the markings. Fifty
beetles were released at each point on ZAPS I and 16 at each point on ZAPS II,
or a total of 100 per plot at ZAPS I and 32 per plot at ZAPS II.
Recaptures, as indicated by colored markings,were very low—45 beetles or
11.3% for ZAPS I, and 17 beetles or 13.3% at ZAPS II. However, most of the
markings on the beetles in the traps dissolved, so that the recapture numbers
indicated the number of traps with paint chips in them. It is possible to
obtain a better estimate from the clipped tarsi, but detecting these on a
small number of beetles out of more than 28,000 beetles is extremely difficult,
although it is currently being attempted.
Marking tests performed on 400 Canthon beetles kept in two cages at the
University of Montana's Botany Gardens, Missoula, Montana, revealed that the
digging activities of these beetles effectively removed any type of paint or
tags within hours to a few days. Clipping of any part of the body other than
the tarsi either resulted in hindrance of beetle activities or injured the
beetles. For example, the thorax is so hard that attempting to clip the
margin often results in shattering the carapace. Clipping the elytra
interfered with flight. During the winter of 1978, branding the prothorax
with a hot needle was found to result in a permanent mark, easier to discern
than clipped tarsi, and both survival and activity did not seem to be apprec-
iably altered. Trapping in June and July of 1979 resulted in relatively
low numbers of Canthon- In August, trapping success increased and some
mark-recapture tests were performed, the results of which will appear in the
next report.
Although the mark-recapture data for 1978 is of little or no value for
interpretation of population dynamics and life-table factors due to loss of
markings and due to apparent low recapture numbers, useful information was
obtained concerning movements. Release and recapture positions are indicated
in Figure 18.6. As expected, the greatest number of recaptures as indicated
by the presence of paint in the traps were at the traps nearest the release
points. However, beetles released on the A plot of ZAPS I were captured as
far away as D plot and vice versa. Similar movements were observed on
ZAPS II. Movements of distances greater than 430 meters occurred. Movements
of 250 to 300 meters were common. Upon release, some beetles ran away on
the ground, but many took to the air and flew off. Frequent observations of
beetles in flight at ZAPS in 1978 and 1979, both upon release and when
approaching bait or dung, have shown Canthon beetles to be strong and swift
fliers, capable of traversing a plot in a few seconds, able to hover in place,
and able to land with almost pinpoint accuracy. Canthon beetles have been
observed flying behind an investigator carrying bait. They have been seen
to hover approximately 0.5 m above a pitfall trap, suddenly fold in their
wings, and drop straight into the trap without touching the sides of the
coffee cup. More often, they fly in a zigzag course towards dung or bait,
land nearby, and scramble across the ground to their objective. When taking
to flight, most crawl to the top of the vegetation canopy on leaves or stems
before launching. Attempts to take off from the ground usually end by crashing
247
-------
into the vegetation. Occasionally, beetles took to flight from areas of bare
ground or from dung pats. Flight activity appears to be affected by meteoro-
logical conditions. When strong winds occur, these beetles are seldom seen in
flight but continue to move about on the ground. Cool temperatures and very
hot temperatures seem to depress flight activity. Most flight activity seems
to occur at air temperatures from 16°-29°C, based on field observations.
Caged beetles generally burrowed under vegetation or soil when temperatures
exceeded 29°C or dropped below 10°C. These temperature limits to kinesis are
being investigated in detail, the figures obtained to date are approximations,
and the thresholds have not been determined accurately. It is apparent that
my first presumption, based on trap data, that these beetles are active at
night may not be correct. They are very active during the daytime and demon-
strate actions suggesting temperature and possibly photic responses.
7 8 21 • D O49
6 9 X • • O •
X
4 • • • ODD
3 ......
2
I • A- 29 -43
ZAPS - I , A
• O ' • •
O • O -D
D
. . O • •
ZAPS- I, B
• OD • •
. . . . Q
n • n • -
. . n • •
on
n • n • •
ZAPS- I , C
n
- • • • A-
••AAAA
• -O• • •
• o • -An
ZAPS-I , D
o n .A
RELEASE POINT LOCATIONS AND SYMBOLS
ZAPS-II , A ZAPS-II , B
O D
. . Q . . .
• • • -D • • D
ZAPS-II , C
•D• • • •
•DO• • •
. . . o • •
. . . o . •
.-A--D
ZAPS-II, D
o
Figure 18.6, Recapture points of painted beetles,
248
-------
Although most beetles lost their paint markings rapidly, a few did not.
In July of 1979, the body of a painted beetle was found in one of the dung
pats used in the dung decomposition tests. This beetle had been marked and
released a full year prior to the setting out of the dung. Although it is
possible that something deposited the body of the beetle on the dung such
as a bird, small animal, or wind, the position of the body suggests that
the beetle was alive and active and attempted to burrow into the pat before
expiring. If so, then the beetle had lived for more than a year at ZAPS.
Mouse Carrion Study
The first carrion study was conducted from June 26, 1978, to September 16,
1978; the second from September 16, 1978, to September 16, 1979. Thus, only
the data from the first test was completed for this report. Fresh and
dried weights are given in Table 18.2. The remainder will be in the Sixth
Interim Report.
Carrion exposed to macroarthropods in the hardware cloth cages was
rapidly found by carrion beetles (mostly Nicrophorus species) , tumble bugs
(mostly Canthon laevis), and ants (mainly harvester ants). The carrion
beetles buried several carcasses within two to three hours of placement on
the plots. The tumble bugs rolled away flesh balls, tearing and shredding
the flesh, especially in the abdominal region. On the Off Plot Control, ants
arrived at the carcasses almost immediately and attacked any other invaders
such as carrion beetles and tumble bugs. Within 48 hours, tufts of hair,
bones, and a few dried pieces of flesh were all that remained of 43 out of 45
specimens. The ants took several days to thoroughly strip a carcass. Of the
45 carcasses, 82.2% were buried by carrion beetles, 15.6% stripped by ants
(6.7%) or tumble bugs (8.9%), and 2.2% stripped by other invertebrates. By
the end of the trial, occasional bones and tufts of hair of the buried car-
casses were found. In most cases, nothing was found even after sifting the
soil through a fine mesh screen. The carcasses in the screen cages underwent
a slow drying out process, and after two to three weeks, it became apparent
that microarthropods such as mites and small ants had invaded the abdominal
regions. By the end of the trial, these carcasses had mummified with bones,
hair, and skin still more or less intact. Carrion beetles, tumble bugs, and
ants removed the majority of organic material; other invertebrates such as
skin and hair beetles, collembolans, flies, and mites consumed smaller amounts.
Assuming 67% water content for the body of a mouse, 94.37% of the carrion
(dry weight basis) was removed when large insects were present. Microorganisms
decomposed 63.97% of the organic material. The larger insects removed major
masses of tissue within a few hours, while microorganisms took several weeks
before discernible changes in the carcasses occurred. In addition, at the
end of the trial, bones, skin, and hair remained where microorganisms were
the only scavengers and decomposers, but in over 80% of the cases even the
bones were gone when large insects had access.
In terms of total biomass, six times as much organic material remained
of the carcasses exposed only to microorganisms compared to those accessible
to the large insects.
249
-------
TABLE 18.2. FRESH AND DRY WEIGHTS OF MOUSE CARCASSES BY PLOT, SEX, CAGE AND POLE POSITION
to
Ui
o
Pole
Off Plot
1
2
3
4
5
ZAPS I,
1 -
2
3
4
5
ZAPS I,
1
2
3
4
5
ZAPS I,
1
2
3
4
5
ZAPS I,
1
2
3
4
5
Hardware
(Sex)
Control
Male
Male
Female
Female
Female
Plot A
Male
Male
Female
Female
Female
Plot B
Female
Male
Female
Female
Male
Plot C
Female
Male
Female
Female
Male
Plot D
Male
Male
Male
Female
Female
Fresh
Weight
Cloth Cage Screen
(Weight)
22.2
17.4
22.6
23.0
13.4
24.8
20.5
21.3
22.9
13.7
25.4
21.5
22.6
17.0
15.8
27.1
15.8
20.2
25.2
19.3
25.7
21.2
19.8
21.0
15.5
(Sex)
Female
Female
Male
Male
Male
Female
Female
Male
Male
Male
Male
Female
Male
Male
Female
Male
Female
Male
Male
Female
Female
Female
Female
Male
Male
Cage
(Weight)
31.2
20.6
19.9
21.5
10.8
26.8
23.1
16.9
18.3
U.3
23.3
24.6
19.0
19.0
15.1
23.4
17.5
16.2
21.8
23.0
29.1
25.4
21.6
17.7
15.7
Dry Weight
Hardware
(Sex)
Cloth Cage
(Weight)
Screen
(Sex)
Cage
(Weight)
Male
Male
Female
Female
Female
0.60
0.13
0.88
0.00
1.80
Female
Female
Male
Male
Male
3.47
1.93
3.30
1.82
4.27
• Male
Male
Female
Female
Female
0.00
0.00
0.00
0.00
0.00
Female
Female
Male
Male
Male
2.80
2.73
3.91
1.57
2.20
Female
Male
Female
Female
Male
0.00
0.00
0.00
0.00
0.00
Male
Female
Male
Male
Female
1.91
2.02
2.31
3.14
3.48
Female
Male
Female
Female
Male
0.00
0.00
0.60
0.00
0.00
Male
Female
Male
Male
Female
1.40
2.17
2.06
3.27
1.84
Male
Male
Male
Female
Female
0.50
0.00
0.00
0.00
0.80
Female
Female
Female
Male
Male
2.42
2.91
2.53
2.16
2.49
Fresh Weight
Pole
ZAPS
1
2
3
4
5
ZAPS
1
2
3
4
5
ZAPS
1
2
3
4
5
ZAPS
,
2
3
4
5
All
Hardware
(Sex)
II, Plot A
Female
Female
Female
Male
Male
II, Plot B
Female
Female
Male
Female
Male
II, Plot C
Female
Male
Hale
Male
Female
II, Plot D
Female
Male
Female
Female
Male
Plots
I -
X =
sx -
Si"'
Cloth Cage
(Weight)
35.5
25.3
23.3
16.8
16.0
29.0
24.0
18.7
20.9
15.7
31.3
20.7
18.3
17.8
14.4
21.9
15.8
18.7
24.0
26.0
949.05
21.09
4.70
0.70
Screen Cage
(Sex)
Male
Male
Male
Female
Female
Male
Male
Female
Male
Female
Male
Female
Female
Female
Male
Male
Female
Male
Male
Female
I -
X -
sx -
%-
(Weight)
22.1
21.4
18.3
19.2
15.8
23.8
22.0
23.0
17.2
13.0
23.3
25.0
22.3
21.2
13.2
19.4
14.0
16.9
21.1
26.7
914.70
20.33
4.38
0.65
Hardware
(Sex)
Dry
Weight
Cloth Cage Screen
(Weight)
(Sex)
Cage
(Weight)
Female
Female
Female
Male
Male
0.40
0.30
1.20
1.35
0.40
Male
Male
Male
Female
Female
2.33
1.97
1.54
1.55
1.66
Female
Female
Male
Female
Male
2.70
0.30
0.00
0.70
0.00
Male
Male
Female
Male
Female
2.58
2.43
2.69
1.83
2.68
Female
Male
Male
Male
Female
0.50
0.80
0.00
0.00
0.00
Male
Female
Female
Female
Male
2.84
2.39
1.33
1.79
1.77
Female
Male
Female
Female
Male
0.40
0.00
0.75
1.30
0.60
Male
Female
Male
Male
Female
2.04
2.88
1.37
3.67
2.51
I =
X «•
sx =
Sx-
17.61
0.37
0.57
0.08
I - 108.76
X -
Sx -
**'
2.42
0.69
0.10
* Grams
-------
In September, a second trial was performed. Care was taken to exclude
microarthropods more effectively by wrapping the mice in layers of cheesecloth.
The tests were run for an entire year. As before, beetles and ants began
working on the bodies within a matter of minutes or a few hours. On ZAPS I, a
large animal, probably a coyote or fox, dug up and removed the bodies from 12
of the 20 hardware cloth cages. It appeared that the animal worked across the
plots taking four mice at A, four at B, three at C, and one at D. Final
results for the trial will be presented in the Sixth Interim Report.
Bovine Feces Decomposition
Fresh and oven-dried weights of 415 cm of bovine feces are presented in
Table 18.3. Both wet and dry weight means demonstrated a low degree of
variability, coefficients of variation ranging from 1.3%-5.1%.
Soon after placement on the plots, the cow pats began to attract insects.
Canthon beetles arrived as soon as 5 minutes after setting out a plate,
although most arrived 20 minutes to a few hours later. Canthon continued to
arrive for 2-3 days, if the interior of the pats remained moist and semi-liquid,
One of the first and most common visitors to the dung were Sphaeridiwn
serabaeiodes Linnaeus which arrived at the dung almost as soon as it was set
out and continued to arrive for 3-4 hours, most appearing within the first
hour. Mohr's description (1943) of hydrophillid activity is germaine: "Most
of them arrive during the first half hour, run excitedly over the surface,
plunge inside and come out again to repeat the performance many times."
Flies and early visitors were abundant. Horn flies (Ea&matobia), false
bottle flies (Cryptolucia), sarcophagid flies (Sarcophaga), minute dung flies
(Leptoceva), and others were common on the fresh dung, and fly larva were
numerous in the dung by the second or third day.
Other scarabaeid beetles, particularly Aphodius species arrived at the
dung and immediately burrowed into the interior. These beetles were easiest
to detect by breaking open one-day or older pats. Many mites were observed
on the surface of the droppings, many arriving on the beetles.
The competition for the dung resource appears to be sharp. Time of day,
temperature, wind, moisture content of the pat, and time of year were but a
few factors which seemed to affect the succession of insects at the feces.
During the summer of 1979, T. J. McNary, Colorado State University, performed
tests on the kinds and numbers of insects in the microseral stages, and on
the effect of time of day on insect succession. This information should be
available for incorporation into the Sixth Interim Report.
During mid-July, some of the pats set out in early summer had the entire
middle area removed so that only a ring of dung remained. Some were exten-
sively hollowed out from underneath, while others had large portions missing
out of the sides of the pat.
Preliminary examination of the data shows distinct seasonal changes in
insect activity, as indicated by holes in the tops and bottoms of the pats.
251
-------
The greatest average loss-observed by mid-summer was 52% of the dung on a
treatment plot which occurred over approximately 45 days from early June
through mid-July, 1979. Field observations indicated that the dry pats were
sustaining renewed visitation and consumption in July and August, primarily
by grasshoppers. At the time of this report, the data had not been examined
for sulfur dioxide treatment effects. Final measurements will be carried out
in the EVST Laboratory and the results presented in the Sixth Interim Report,
1980.
TABLE 18.3. MEAN WEIGHTS OF 415 CM3 PATS OF BOVINE MANURE
Experiment
No.
A79I
A79II
A79I4II
B79I
B79II
B79I&II
n
8
8
16
8
8
16
Wet
X
451.0
439.1
445.1
456.0
447.6
451.8
Weieht
S.D.
8.11
5.89
9.19
8.49
6.37
8.44
(Grams)
S.E.
2.87
2.08
2.30
3.00
2.25
2.11
C.V.
1.8
1.3
2.1
1.9
1.4
1.9
n
7
8
15
8
8
16
X
69.00
69.36
69.19
94.30
93.07
93.69
Dry Weieht
S.D.
3.045
1.635
2.312
4.801
2.611
3.786
(Grains)
S.E.
1.151
0.578
0.597
1.697
0.923
1.339
C.V.
4.4
2.4
3.3
5.1
2.8
4.0
Moisture
(Percent)
84.7
84.2
84.5
79.3
79.2
79.3
Tube Tests of Beetle Movements and Orientation
The initial tube tests at ZAPS produced contradictory results. At times
Canthon laevis appeared to demonstrate a distinct adversion to sulfur dioxide
and at other times just the opposite response was observed. It became
apparent that factors such as temperature, slope, wind direction, and orienta-
tion to the sun were interacting. Therefore, an intensive study of these fac-
tors was initiated before continuing tests of response to sulfur dioxide.
These experiments were being carried out at the time of this report. The
results will appear at a later time.
The technique seemed to be useful for tests involving Canthon laevis -
It proved to be of little value with Nicrophorus species. Upon release,
Canthon laevis soon began to move about, walking easily and usually oriented
with the axis of the tube. On the other hand, Nicrophorus beetles spent
most of the test period trying to squeeze through the mesh. Many of the
smaller individuals eventually escaped. Movements within the tube by these
beetles were extremely limited. Temperatures in excess of 32°C suppressed
movement by Canthon beetles. At these temperatures the beetles would either
remain sedentary at the point of release or would walk until encountering
252
-------
shade from the vegetation and then stop all but very limited movements. At
these temperatures, most Nterophorus beetles perished, presumably from heat
prostration.
DISCUSSION
As in previous years, the pitfall trapping at both ZAPS I and II demon-
strated more captures of Canthon laevis on the controls than on any of the
sulfur dioxide-treated plots. This difference was significant (p <_ 0.005)
for total captures, exterior captures, and interior captures. Unlike 1976
and 1977, the population abundance of C. laevis, as indicated by pitfall
captures, was not similar at ZAPS I and II. There were 2.7 times as many
beetles captured at ZAPS I as at ZAPS II. In 1976 at both sites, captures
of C. laevis were inversely correlated with sulfur dioxide concentrations.
In 1977, an inverse correlation was observed for ZAPS II, but at ZAPS I there
was a sharp drop-off in numbers on the B plot as compared to the control plot
(A plot), and captures on the B, C, and D plots were similar. The 1978
captures at both ZAPS I and II displayed a pattern like that seen at ZAPS I
in 1977. Based on the arithmetic mean for sulfur dioxide fumigation levels
(Lee et al., '1979), the treatment levels were greater in 1978 than 1977 at
both ZAPS—1.4-19.2 pphm at ZAPS I and 1.1-17.8 pphm at ZAPS II versus 1.1-
15.3 pphm at ZAPS I and 1.0-16.9 pphm at ZAPS II. This suggests that higher
fumigation levels in 1978 may have induced a more pronounced response by the
beetles.
Total captures of Canthon laevis at both ZAPS I and II exceeded peak
population abundance for any previous year. In 1978, rainfall at Colstrip,
Montana, exceeded that of 1977 by a factor of 1.48 and that of 1976 by a
factor of 1.42. May and June precipitation was 20.22 cm for 1976, 12.04 cm
for 1977, and 27.05 cm .for 1978 (Munshower, personal communication, 1979).
Based on trapping experience at ZAPS, captures increase following rain and
are lowest during extended hot, dry periods. From my own observations and
frequent comments by other investigators working with invertebrate scavengers
of dung and carrion, rainfall extends the period of resource utilization by
slowing the drying process and by rehydrating the organic material. Thus,
the large number of captures in 1978 may have been due to greater precipi-
tation, which affects beetle activity and may promote population growth by
maintaining the suitability of the food resource.
The 1978 pitfall data clearly indicated that sulfur dioxide affected
populations of Canthon even at the lowest fumigation level. Since the
reduction in capture success was similar on all of the treated plots as
compared to the controls, it appears that the threshold for the response to
sulfur dioxide occurs at or below the lowest fumigation level.
No distinct trends in captures as related to sulfur dioxide for any
other species of beetles were noted for 1978, although some other species
have been seen to respond to sulfur dioxide in previous years. However,
capture numbers of species other than Canthon in 1978 were small and the data
characterized by high variability. Thus, even if responses occurred, the
"noise" in the data would have obscured any trends.
253
-------
It appears unlikely that sulfur dioxide induced changes in population
dynamics such as altered rates of natality or mortality would explain the
Canthon responses observed at ZAPS. If the insects had a restricted range,
this would be an acceptable hypothesis. But the behavioral observations at
ZAPS clearly indicate that these beetles are far ranging and capable of
traveling considerable distances in a short period of time. Since a
beetle released on a plot may be recaptured on any other plot, this indicates
a range greater than that of the entire series of plots, and as such, natality
or mortality of a resident population could affect total captures at a
ZAPS site but not necessarily captures on individual plots.
Whether this behavioral response to sulfur dioxide takes the form of a
taxes (directed reactions or movements) or kineses (undirected reactions or
movements)'remains to be discerned. Anemotaxic, geotaxic, phototaxic, telo-
taxic, and klinotaxic reactions are being investigated via the tube tests.
Observations made at ZAPS in 1979 of Canthon arriving at dung on the control plot
(A) versus the highest treatment plot (D). suggest an orthokinetic response.
The fecal and carrion decomposition studies clearly demonstrate the speed
with which scavenger beetles arrive; the substantial amounts of organic
material that is shredded, consumed, and buried; and the increased area of
dispersal of this material via activities such as ball rolling. A large and
varied dung and carrion community, microseral succession, and competition
are obvious characteristic's. Although decomposition, nutrient recycling, and
soil improvement are obvious benefits, other benefits such as competition for
the resource with pest or disease-carrying insects, clean-up on rangelands
and reduction of wastage of fouled forage, and immediate recycling of the
organic material to food chains via insectivorous animals may be as important
if not more so.
Figure 18.7 attempts to model some of these factors. Quantification of
the main pathways has not been accomplished to date, but the results of the
1979 field research and incorporation of published information should provide
a means for developing a more specific model for inclusion in the final project
reports.
Other Observations
A number of behavioral observations for Canthon laevis will be covered in
the next interim report. A few are of interest with regard to the studies des-
cribed here.
Dung removal by Canthon is accomplished by: (1) feeding, (2) ball rolling,
and (3) shredding usually in conjunction with feeding or ball rolling. The
distance a dung ball is rolled from the dung pat is variable. Dung balls have
been found immediately under the pat, presumably to be buried in the moist
soil there. Generally, the balls are rolled away from the immediate proximity
of the dropping. For example, in one trial in which 19 dung balls were formed,
the average distance of deposition from the edge of the pat (burial or abandon-
ment) was 72.1 ± 52.2 cm. Three balls were left by the edge of the dropping
and two were moved twice. Of those moved twice,one was rolled off by two
beetles. At a resting point, part of the ball was consumed by the two.
254
-------
ASSIMILATION
OF ANIMALS
ANIMAL
CARRYING
CAPACITY
CONSUMPTION
(TOTAL)
OF PLANTS
CONSUMPTIO
OF ANIMALS
BY PREDATORS
DEAD PARTS
OF PLANTS
TOTAL PLANT
BIOMASS
PLANT GROWTH!
MORTALITY
SOIL NUTRIENTS
ASSIMILATION
OF
MICRO-ORGANISMS
DETRITUS
ASSIMILATION
OF
INVERTEBRATES
FECES AND/OR
CARRION
ON SURFACE
DUNG/CARCASS
REMOVAL
BURIAL
MICRO-ORGANISMS
WEATHER
LEACHING
INVERTEBRATES
(INSECTS)
LOST TO
SYSTEM
Figure 18.7. Interactions of Coprophagous and Necrophagous beetles
(based on McKinney and Morley, 1975, and Olechowicz, 1977)
255
-------
Then, one beetle reformed the remainder of the ball into a smooth sphere and
continued to roll it off. In the second case, a beetle abandoned a ball it
was rolling. About 30 minutes later either the same beetle or another beetle
wandered off of a nearby dung pat pursuing a rather erratic course through
the vegetation until coming within sight of the ball, at which time the beetle
dashed over to the ball, examined it from all sides, and then continued to
roll it off.
Balls were only formed from moist dung. Usually the dung used to build
the ball was pulled from the top or sides of the pat. When feeding and
shredding, Canfhon usually worked around the edge of the pat or from under-
neath. On hot, sunny days with temperatures over 32°C, only feeding and
shredding activities were observed. At these times, all of the beetles would
either be under the pat or in the shaded area under the edge of the pat oppo-
site the position of the sun. When temperatures were high, newly arrived
beetles quickly slid under the edge of the dropping. At times, so many beetles
were under a pat that the entire surface undulated almost as if it were pulsing.
Sudden movements by the observer disturbed the beetles, although sounds
did not, as far as could be determined. Beetles .were particularly likely to
cease activities and fly or run off if disturbed soon after arriving at the
dung or carcass. When feeding, or building or rolling balls, they seemed to
pay little attention to the observer. When disturbed, they either fled by
flight, running away, or burrowing rapidly into the dung or under vegetation.
If cornered or captured, such as by being picked up by hand, they often
"played dead," remaining totally motionless for periods up to several minutes
even after being released.. When in this state of suspended animation, they
can be moved around, picked up, prodded, and still show no indication of being
alive and active. The appendages are held rigid. If left alone, they even-
tually begin to make some very hesitant and limited movements, such as moving
a leg or antenna. If not disturbed at this point, they shortly resume activity,
usually not leaving the area where they were disturbed.
It is almost impossible to hold a moving beetle. Invariably, the beetle
crawls into the space between the base of the fingers and begins digging and
prying movements with its powerful prothoracic legs which are broadened and
strengthened for burrowing. The sensation is uncomfortable to the holder
because the legs are hard, ridged, and have rather sharp tubercles. Even if
the discomfort is ignored, the beetle normally succeeds in shoving its way
out between the fingers.
The responses of Canthon laevLs in response to the presence and movements
of an observer demonstrate several strategies of self-defense. This points
to an evolutionary fitness selection in response to predation. The Australian
observations that birds and amphibians prey on these beetles are important
in that for the most part these predators rely heavily upon movement to detect
their quarry. The cryptic (dull brown, greenish brown, or black) colors of
these beetles, in addition to the tactic of remaining motionless would be
distinct advantages in escaping notice by these predators. Fleeing and
hiding would be of advantage if the predator were too far away to reach the
beetle before it could make its escape, or if the predator suddenly encountered
the beetle. It is even possible that the digging and prying actions taken by
256
-------
a beetle held in a hand could enable the beetle to wrench itself free of some
captors. Swallowing a scrabbling beetle might be less than a pleasant exper-
ience and might serve to dissuade the predator from eating any others.
Given the complexity of the insect communities of dung and carrion, the
significance of an alteration in the abundance of one or more species is a
difficult question. On the one hand, it is possible that any reduction in
numbers of beetles such as Canthon laevis could be offset by concurrent shifts
in other species. Thus, a process such as decomposition might not demonstrate
any alteration if a new equilibrium were established. On the other hand, one
or more species may serve a unique or critical function. For example, suppres-
sion of populations of harmful flies such as flesh flies, horn flies, and
disease vectors appears to be a function in which certain beetles play a
particularly important role. Although there are a number of predators and
parasites of these flies, the dung beetles appear to be particularly efficient
competitors for food and space resources in the following ways: (1) removal
and consumption of quantities of dung and carrion, (2) transportation of mites
that prey on flies, (3) dislodging and breaking up egg masses by feeding and
shredding, (4) aeration of the dung and carrion increasing the drying rate,
and (5) riddling with entry holes, exit holes, and galleries which provide
avenues of access for predators such as staphylinid beetles which do not
construct their own corridors.
The use of a standardized bovine dung pat represents an attempt to examine
in a more holistic manner an ecological unit and the effects of sulfur dioxide
on that unit. The finding that Canthon laevis is particularly susceptible to
the influence of sulfur dioxide.provides an indicator organism that has con-
siderable potential as an early warning mechanism concerning the presence of
low levels of sulfur dioxide. An understanding of the mechanism involved
should provide a more refined means of utilizing the indicator and should
provide some insights into the potential consequences of sulfur dioxide induced
reactions both to the indicator organism and to the ecological structural
and functional units of which it is a component. Hopefully, approaches such
as the use of a standardized dung unit could provide a microcosm for use
which would be much more information rich concerning pollution impacts. An
obvious disadvantage is interpretation of large amounts of information and
separation and identification of the interacting driving variables. At this
point, it is impossible to say if dung pats could serve this type of purpose,
but there is a need for these types of bioassay tools. Attempts already have
been made to develop a soil microcosm test in order to address questions such
as: (1) the mobility of a pollutant through the system, (2) the form in
which the pollutant is exported from the soil,(3) disruption or alterations of
nutrient cycling, (4) effects on soil biota, and (5) dose-responses to the
pollutant with respect to mobility, export, nutrient cycling, and population
dynamics (Duke et al. , 1977).
CONCLUSIONS
1. Beetles of the species Canthon, mostly laevis,respond to low level sulfur
dioxide fumigation. Captures on the control plots (A) were significantly
higher than for any of the sulfur dioxide treatment plots at both ZAPS
in 1976, 1977, and 1978.
257
-------
2. The reaction of Canthon laevis to sulfur dioxide appears to be primarily
behavioral and not due to altered life table factors such as changes in
natality or mortality.
3. Whether the response is a kinesis or taxis or both has not been determined,
but preliminary evidence suggests an orthokinesis.
4. Scavenger beetles make significant contributions to the rapid breakdown
and removal of dung and carrion in rangelands.
5. These scavenging activities contribute not only to the quick turnover of
nutrients but also result in the cleaning up of fouled herbage and strong
competition for food and space resources with pest insects such as parasites
and carriers of disease.
6. Canthon laevis employs several defensive strategies when disturbed, in
addition to being cryptically colored. This indicates that they either
are or have been a food resource for some,other organism(s).
7. The beetle studies at ZAPS have not only revealed information concerning
the responses to sulfur dioxide but have provided a considerable amount of
information concerning the biology of these scavengers or decomposers in
grasslands of the Northern Great Plains.
REFERENCES
Allee, W. C., A. E. Emerson, 0. Park, T. Park, and K. P. Schmidt. 1949.
Principles of Animal Ecology. W. B. Saunders Co., Philadelphia, Pa.
837 pp.
Andrzejewska, L. 1974. Analysis of a Sheep Pasture Ecosystem in the Pieniny
Mountains (The Carpathians). V. Herbivores and Their Effect on Plant
Production. Ekol. Pol., 22:527-534.
Blume, R. R. 1970. Insects Associated with Bovine Droppings in Kerr and
Becar Counties. Texas J. Econ. Entomol., 63:1023-1024.
Blume, R. R., S. E. Kunz, B. F. Hogan, and J. J. Matter. 1970. Biological
and Ecological Investigations of Horn Flies in Central Texas: Influences
of Other Insects in Cattle Manure. J. Econ. Entomol., 63:1121-1123.
Bromenshenk, J. J. 1978. Investigation of the Impact of Coal-Fired Power
Plant Emissions Upon Insects: Entomological Studies at the Zonal Air
Pollution System. In: The Bioenvironmental Impact of a Coal-Fired
Power Plant, Third Interim Report, Colstrip, Montana, E. M. Preston and
R. A. Lewis, eds. EPA-600/3-78-021, U.S. Environmental Protection Agency,
Corvallis, Or. pp. 473-507.
Bromenshenk, J. J. 1979. Responses of Ground-Dwelling Insects to Sulfur
Dioxide. In: The Bioenvironmental Impact of a Coal-Fired Power Plant,
Fourth Interim Report, Colstrip, Montana, E. M. Preston and T. L. Gullett,
eds. EPA-600/3-79-044, U.S. Environmental Protection Agency, Corvallis,
Or. pp. 673-722.
258
-------
Brown, W. J. 1940. Notes on the American Distribution of Some Species of
Coleoptera Common to the European and North American Continents. Can.
Entomol., 72(4):70-72.
Coffey, M. D. 1966. Studies on the Association of Flies (Diptera) with Dung
in Southeastern Washington. Ann. Entomol. Soc. Am., 59(1):207-217.
Duke, K. M., M. E. Davis, and A. J. Dennis. 1977. IERL-RTP Procedures Manual:
Level 1 Environmental Assessment Biological Tests for Pilot Studies.
EPA-600/7-77-043, U.S. Environmental Protection Agency, Research Triangle
Park, N.C. 105 pp.
Durie, P. 1975. Some Possible Effects of Dung Beetle Activity on the Infes-
tation of Pastures by Intestinal Worm Larvae of Cattle. J. Appl. Ecol.,
12(3):831.
Gillard, P. 1967. Coprophagous Beetles in Pasture Ecosystems. J. Aust.
Inst. Agr. Sci., 33:3-34.
Green, R. 1966. Gut Contents of Some Tasmanian Birds. The EMU, 66:105-110.
Hammer, 0. 1941. Biological and Ecological Observations on Flies Associated
with Pasturing Cattle and Their Excrement. Videsk. Medd. Dansk.
Naturhist. Foren. Kjobenhaun, 105:141-393.
Harris, R. L., and L. M. Oliver. 1979. Predation of Philonthus flavolirribatus
on the Horn Fly. Environ. Entomol., 8:250-260.
Hayes, W. P. 1927. Prairie Insects. Ecology, 8:238-250.
Hayes. W. P. 1929. Morphology, Taxonomy, and Biology of Larval Scarabaeoides.
111. Biol. Mono., 12:1-119.
Kulman, H. M. 1974. Comparative Ecology of North American Carabidae with
Special Reference to Biological Control. Entomophaga Mem. H. S.,
7:61-70.
Lee, J. J., E. M. Preston, and D. Weber. 1979. Temporal Variation in Sulfur
Dioxide Concentration on ZAPS. In: The Bioenvironmental Impact of a
Coal-Fired Power Plant, Fourth Interim Report, Colstrip, Montana, E. M.
Preston and T. L. Gullett, eds. EPA-600/3-79-044, U.S. Environmental
Protection Agency, Corvallis, Or. pp. 284-305.
Leetham. J. W., T. J. McNary, and J. L. Dodd. 1979. Effects of Controlled
Levels of Sulfur Dioxide on Invertebrate Consumers. In: The Bioenviron-
mental Impact of a Coal-Fired Power Plant, Fourth Interim Report, E. M.
Preston and T. L. Gullett, eds. EPA-600/3-79-044, U.S. Environmental
Protection Agency, Corvallis, Or. pp. 723-763.
MacQueen, A. 1975. Dung as an Insect Food Source: Dung Beetles as Compet-
itors of Other Coprophagous Fauna and as Targets for Predators. J. Appl.
Ecol., 12(3):821-827.
259
-------
MacQueen, A., and R. P. Beirne. 1975. Influence of Other Insects on Produc-
tion of Horn Fly, Haematobia irritans (Diptera: Muscidae), From Cattle
Dung in South-Central British Columbia. Can Entomol., 107:1255-1264.
McKinney, G. T., and F. H. W. Morley. 1975. The Agronomic Role of Introduced
Dung Beetles in Grazing Systems. J. Appl. Ecol., 12(3):831-836.
Mohr, C. 0. 1943. Cattle Droppings as Ecological Units. Ecol. Mono.,
13(3):275-298.
Morrill, W. L. 1975. Plastic Pitfall Trap. Environ. Entomol., 4:596.
Olechowicz, E. 1974. Analysis of a Sheep Pasture Ecosystem in the Pieniny
Mountains (The Carpathians). X. Sheep Dung and the Fauna Colonizing It.
Ekol. Pol., 22:589-616.
Olechowicz, E. 1976. The Role of Coprophagous Dipterans in a Mountain
Pasture Ecosystem. Ekol. Pol., 24:124-165.
Olechowicz, E. 1977. The Effect of Coprophages on Feces Decomposition in
a Pasture Ecosystem. Ecol. Bull. (Stockholm), 25:553-556.
Olson, Elliott, and Associates. 1976. Effects of Spruce Budworm Control on
Pollinating Insects. USDA Forest Service, Region 1, Missoula, Mt. 56 pp.
Payne, J. A. 1965. A Summer Carrion Study of the Baby Pig Sus serofa
Linnaeus. Ecology, 46(5):592-602.
Portchinsky, I. A. 1885. Muscarum Cadaverinarum Stercoriarumque Biologia
Comparata. Horae Soc. Entomol. Ross., 19:210-244.
Preston, E. M., and T. Gullett. 1979. Spatial Variation of Sulfur Dioxide
Concentrations on ZAPS during the 1977 Field Season. In: The Bioenviron-
mental Impact of a Coal-Fired Power Plant, Fourth Interim Report,
Colstrip, Montana, E. M. Preston and T. L. Gullett, eds. EPA-600/3-79-044,
U.S. Environmental Protection Agency, Corvallis, Or. pp. 306-330.
Rowley, I., and W. J. M. Vestjens. 1973. The Comparative Ecology of
Australian Corvids. V. Food. Aust. Wildl. Res., 18:131-156.
Sanders, D. P., and R. C. Dobson. 1966. The Insect Complex Associated with
Bovine Manure in Indiana. Ann. Entomol. Soc. Am., 59(5):955-959.
Shubeck, P. P. 1968. Oreientation of Carrion Beetles to Carrion: Random or
Non-random? J. New York Entomol. Soc., 76(4):253-265.
Shubeck, P. P. 1976. An Alternative to Pitfall Traps in Carrion Beetle
Studies (Coleoptera). Entomol. News, 87(5-6):176-178.
Southwood, T. R. E. 1975. Ecological Methods with Particular Reference to
the Study of Insect Populations. Chapman and Hall Ltd., London. 391 pp.
260
-------
Thomas, D. B., Jr., and E. L. Sleeper. 1977. The Use of Pitfall Traps for
Estimating the Abundance of Arthropods, with Special Reference to the
Tenebrionidae (Coleoptera). Ann. Entomol. Soc. Am., 70(2):242-248.
Thomas, G. D. 1967. Natural Enemies of the Face Fly Musca autwnnalis De
Geer in Missouri. Ph.D. Thesis, University of Missouri, Columbia, Mo.
131 pp.
Thomas, G. D., and C. E. Morgan. 1972. Field-Mortality Studies of the
Immature Stages of the Horn Fly in Missouri. Environ. Entomol., 1:453-459.
Vestjens, W. J. M., and R. Carrick. 1974. Food of the Black-Backed Magpie,
Gymnorh-ina T, Tibicen, at Canberra. Aust. Wildl. Res., 1:71-83.
261
-------
APPENDIX TABLE 18,1. 7 .X 7 GRID CAPTURE OF CANTRON SP, AT ZAPS I AND II? 1978
PLOT
PLOT
PLOT
PLOT
A (Control)
Total Traps
I 7,001
X 142.9
S 120.4
S* 17.2
B (Low)
Total Traps
1 4.787
X 97.7
S 89.4
S- 12.8
C (Medium)
Total Traps
£ 4,242
X 86.6
S 102.0
Sx 14-6
D (High)
Total Traps
£ 4,426
X 90.3
S 75.7
S£ 10.456
ZAPS I
Perimeter Traps
£ 4,777
X" 199.0
S 120.2
Sx 24. S
Perimeter Traps
£ 3,426
X 142.8
S 94.2
S- 19.2
Perimeter Traps
£ 2,744
X 114.3
S 115.9
Sj 23.7
Perimeter Traps
I 3,005
X 125.2
S 80.2
Sj 16.4
Grand Total - 20,456
Trap Days - 196
June 30 - July 10, 1978
Interior Traps
I 2,224
X 89.0
S 94.7
S- 18.9
June 30 - July 10, 1978
Interior Traps
1 1,361
X 54.5
S 59.0
sx 11'8
June 30 - July 10, 1978
Interior Traps
I 1,498
X 59.9
S 80.0
Sj 16.0
June 30 - July 10, 1978
Interior Traps
I 1,421
X" 56.8
S 54.1
S£ 10.8
PLOT A (Control)
Total Traps
I 2,887
7 58.9
S 86.6
S- 12.4
PLOT B (Low)
Total Traps
£ 1,515
X 30.9
S 63.2
Sj 9.0
PLOT C (Medium)
Total Traps
£ 1,010
X 20.6
S 31.1
Sj, 4.4
PLOT D (High)
Total Traps
£ 1,808
X" 36.9
S 56.3
Sj 8.0
ZAPS II
Grand Total - 7,620
Trap Days - 196
June 30 - July 10, 1978
Perimeter Traps
I 2.264
X 94.3
S 99.6
Sj 20.3
Perimeter Traps
£ 1,425
X 59.4
S 81.3
Sj 16.6
Perimeter Traps
I 776
X 32.3
S 38.8
S, 7.9
Perimeter Traps
£ 1.658
X 69.1
S 65.7
Sj 13.4
Interior Traps
£ 623
X" 24.9
S 54.9
S, 11.0
June 30 - July 10,
Interior Traps
£ 90
X 3.6
S 8.2
Sj 1.6
June 30 - July 10,
Interior Traps
I 234
X 9.4
S 14.9
S* 3.0
June 30 - July 10.
Interior Traps
£ 150
7 6.0
S 13.1
Sj 2.6
1978
1978
1978
-------
APPENDIX TABLE 18.2. 7X7 GRID CAPTURE OF NICRO1HORIS SP. AT ZAPS I AND II, 1978
to
ON
Co
PLOT A (Control)
Total Traps
I 83
* 1.7
S 2.4
85, 0.4
PLOT B (Low)
Total Traps
I 83
X 1.7
S 2.0
S* 2-3
PLOT C (Medium)
Total Traps
I 125
X 2.6
S 8.4
S- 1.2
PLOT D (High)
Total Traps
1 58
X 1.2
S 1.9
S- 0.3
ZAPS 1
Perimeter Traps
r si
X 2.1
S 2.8
Sj, 0.6
Perimeter Traps
I 47
X 2.0
S 2.5
Sj 0.5
Perimeter Traps
I 31
X 1.3
S 1.5
s; -3
Perimeter Traps
I 35
X 1.5
S 2.1
Sj 0.4
Grand Total - 349
Trap Days - 196
June 30 - July 10, 1978
Interior Traps
£ 32
X 1.3
S 2.0
sx °'*
June 30 - July 10, 1978
I 36
X 1.4
S 1.5
Sj, 0.3
June 30 - July 10, 1978
Interior Trape
I 94
X 3.8
S 11.7
S;2.3
June 30 - July 10, 1978
Interior Traps
.1 23
X 0.9
S 1.5
Sj 0.3
PLOT A (Control)
Total Traps
I 152
T 3.1
S 3.7
Sj 0.5
PLOT B (Low)
Total Traps
I 139
X 2.8
S 3.0
sx °'A
PLOT C (Medium)
Total Traps
I 117
X 2.4
S 3.0
S- 0.4
PLOT D (High)
Total Traps
I 208
X 4.2
S 4.3
Sj 0.6
ZAPS II
Perimeter Traps
I 55
X 2.3
S 2.7
Sj 0.6
Perimeter Traps
I 82
X 3.4
S 3.6
S- 0.7
Perimeter Traps
I 82
X 3.4
S 3.7
sx °'8
Perimeter Traps
£ 153
X 6.4
S 4.9
Sj 1.0
Grand Total - 616
Trap Days - 196
June 30 - July 10,
Interior Traps
E 97
X 3.9
S 4.4
Sj 0.9
June 30 - July 10,
Interior Traps
I 57
X 2.3
S 2.2
Sx °-5
June 30 - July 10,
Interior Traps
I 35
y 1.4
S 1.8
Sg 0.4
June 30 - July 10,
Interior Traps
I 55
X" 2.2
S 2.0
Sj 0.4
1978
1978
1978
1978
-------
APPENDIX TABLE 18.3.
7X7 GRID CAPTURE OF m ANAEW
SP. AT ZAPS I AND II, 1978
APPENDIX TABLE 18.4.
Total Traps
PLOT A
I 18
5f 0.37
S 0.07
S- 0.10
Total Traps
PLOT A
I 5
T 0.10
S 0.31
S- 0.04
PLOT B
E 7
X" 0.14
S 0.41
S- 0.06
PLOT B
I 2
X 0.04
S 0.29
S- 0.04
ZAPS I
PLOT C
I 14
X" 1.56
S 1.13
Sj 0.16
ZAPS II
PLOT C
I 5
X 0.10
S 0.37
S- 0.05
Grand Total - 49
Trap Days - 196
June 30 - July 10, 1978
PLOT D
I 10
X 1.11
S 0.33
S- 0.05
Grand Total - 17
Trap Days - 196
June 23 - July 3. 1978
PLOT D
I 5
X 0.10
S 0.37
Ss 0.05
7X7 GRID CAPTURE OF ONW OSIAGU5
SP. AT ZAPS I AND II, 1978
Total Traps
PLOT A
I 4
X 0.08
S 0.28
Sj 0.04
Total Traps
PLOT A
£ 1
X 0.02
S 0.14
SJ 0.02
ZAPS I
PLOT B
£ 11
X" 0.22
S 0.55
S- 0.08
ZAPS II
PLOT B
£ 2
X 0.04
S 0.29
Sg 0.04
PLOT C
£ 2
X 0.04
S -.29
Sj 0.04
PLOT C
£ 1
X 0.02
S 0.14
Sg 0.02
Grand Total - 18
Trap Days - 192
June 30 - July 10. 1978
PLOT D
£ 1
X 0.02
S 0.14
S- 0.02
Grand Total - 8
Trap Days - 192
June 23 - July 3, 1978
PLOT D
£ 4
X 0.08
S 0.28
Sj 0.04
-------
APPENDIX TABLE 18.5.
7X7 GRID CAPTURE OF PPB IMAOi IS
SP. AT ZAPS I AND II, 1978
APPENDIX TABLE 18.6.
ZAPS I
7X7 GRID CAPTURE OF CINCINDELLA
SP. AT ZAPS I AND II, 1978
ZAPS I
NJ
Grand Total - 56
Trap Days - 196
Total Traps
PLOT A
I 5
X 0.10
S 0.47
S- 0.07
PLOT B
£ 19
X 0.39
S 0.76
S- 0.11
PLOT C
£,14
X 0.29
S 0.58
Sj 0.08
June 30 - July 11
PLOT D
I 18
X 0.37
S 0.81
Sx °-12
Total Traps
PLOT A
PLOT B
PLOT C
Grand Total - 21
Trap Days - 196
June 30 - July 10. 1978
PLOT D
£ 12
X 0.24
S 0.60
S£ 0.09
£ 0
X -
S -
S«-
£ 0
x -
s -
si-
£ 9
X 0.18
S 0.57
Sj 0.08
ZAPS 11
ZAPS II
Grand Total - 24
Trap Days - 196
Total Traps
PLOT A
£ 1
f 0.02
S 0.14
Sj 0.02
PLOT B
£ 8
X 0.16
S 0.43
S- 0.06
June 30 - July 11
PLOT C PLOT D
£ 8
X 0.16
S 0.37
S; 0.05
£ 7
X 0.14
S 0.41
Sj 0.06
Grand Total - 7
Trap Days - 196
Total Traps
PLOT A
£ 3
X 0.06
S 0.32
Sj 0.05
June 30 - July 1(
PLOT B
£ 2
X 0.04
S 0.29
S£ 0.04
PLOT C
I 1
X 0.02
S 0.14
Sjj 0.02
PLOT D
£ 1
X 0.02
S 0.14
Sj 0.02
-------
APPENDIX TABLE 18.7.
N3
7X7 GRID CAPTURE OF WEEVIL
SP, AT ZAPS I AND II, 1978
APPENDIX TABLE 18.8.
Total Traps
PLOT A
£ 0
X -
S -
Total Traps
PLOT A
I 3
X 0.06
S 0.32
S; 0.05
ZAPS I
PLOT B
t 0
X -
S -
ZAPS II
PLOT B
I 1
X 0.02
S 0.14
S- 0.02
PLOT C
I 1
X 0.02
S 0.14
Sj 0.02
PLOT C
£ 0
X 0
S -
S; -
Grand Total - 2
Trap Days - 196
June 30 - July 10. 1978
PLOT D
£ 1
X 0.02
S 0.14
Sj 0.02
Grand Total - 8
Trap Days - 196
June 30 - July 10, 1978
PLOT D
£ 4
X 0.08
S 0.34
S; 0.05
7X7 GRID CAPTURE OF TROX
SP. AT ZAPS I AND II, 1978
Total Traps
PLOT A
I 6
X 0.12
S 0.44
S- 0.06
ZAPS I
Grand Total - 59
Trap Days - 196
June 30 - July 10, 1978
PLOT B PLOT C PLOT D
I 16
X 0.33
S 0.59
S- 0.08
I 16
X 0.33
S 0.75
S- 0.11
Z 21
X 0.43
S 0.71
Sj 0.10
ZAPS II
Grand Total - 98
Trap Days - 196
Total Traps
PLOT A
I 13
X 0.27
S 0.64
S, 0.09
PLOT B
I 36
X 0.73
S 1.00
S5 0.14
PLOT C
£ 28
y 0.57
S 0.94
Ss 0.13
June 30 - July 11
PLOT D
Z 21
X 0.43
S 0.76
S- 0.11
-------
APPENDIX TABLE 18.9.
S3
7X7 GRID CAPTURE OF OTHER
BEETLES AT ZAPS I AND II, 1978
ZAPS I
Total Traps
PLOT A
I 36
X 0.73
S 1.24
Sg 0.18
PLOT B
Z 97
X 1.98
S 2.02
Sj 0.29
PLOT C
I 121
X 2.47
S 2.20
Si °'31
Grand Total - 293
Trap Days - 196
June 30 - July 10, 1978
PLOT D
I 39
X 0.80
S 1.32
S- 0.19
ZAPS II
Total
Traps
PLOT
I 34
X 0.
S 1.
SjO.
A
69
23
18
PLOT
I 82
X 1.
S 1.
SjO.
B
67
66
24
PLOT
I 48
5f 0
S 1
Sj 0
C
.98
.07
.15
Grand
Trap
June
Total - 220
Days - 196
30 - July 10,
PLOT
I 56
if i.
S 1.
S- 0.
D
14
87
27
1978
-------
SECTION 19
RESPONSE OF FIELD POPULATIONS OF TARDIGRADES TO VARIOUS
LEVELS OF CHRONIC, LOW-LEVEL SULFUR DIOXIDE EXPOSURE
J. W. Leetham, T. J. McNary, J. L. Dodd, and W. K. Lauenroth
ABSTRACT
The soil populations of Tardigrades of a northern
mixed grass prairie were sampled after having been ex-
posed to various low levels of sulfur dioxide. The
populations of ZAPS I and ZAPS II sites showed substan-
tial reduction, although it was difficult to show the
reductions were significant, due to high variability of
sample counts. Substantial reduction in the frequency
of occurrence in the samples helped to substantiate the
population decreases. Three genera of tardigrades were
identified from sample material: Maorobiotus3
Hexapodibuis, and Diphasoon.
INTRODUCTION
Tardigrades are small (<1.0 mm) arthropod-like animals commonly called
"water bears" because most known species are aquatic or semiaquatic. Because
their phylogeny is unclear (Riggin 1962), they have been classified with the
arthropods by some and given separate phylum status by others. Their func-
tional or trophic status is also unknown, although population estimates in
some soil systems range as high as 300,000 • m~2 (Franz 1941, 1950; Ramazzotti
1959). Tardigrades are often overlooked in soil faunal studies because most
commonly used extraction methods are inefficient in retrieving them. Although
they may be comparatively less important to ecosystem function than say nema-
todes, tardigrades may be very sensitive indicators of air pollution effects
on soil ecosystem processes.
MATERIALS AND METHODS
Each treatment plot of the ZAPS I and II sites was divided into replicate
halves and each of the replicates was subdivided into a grid of sampling lo-
cations. A random numbers table was used to choose a set of sampling locations
at each of three sampling dates. In July, 1977 and July, 1978, five samples
were taken from each replicate of each treatment in both ZAPS I and ZAPS II.
In September, 1978, 10 samples were collected from each replicate in ZAPS I
268
-------
only. A sample consisted of a 4.8 cm diameter by 10.0 cm deep soil core from
which the tardigrades were extracted by a Baermann funnel process, a commonly
used live extraction technique. Specimens were preserved in 70% ethanol and
representatives were sent out for identification.
RESULTS AND DISCUSSION
Three tardigrade genera were identified including Macrobiotus Schultze,
Hexapodibius,, and Diphascon Plate (=Hypsibius Ehrenberg) . The Diphasoon is
probably undescribed and further work is needed to determine if the others
fit recognized species. The mean counts for all sample dates showed sub-
stantially reduced populations in the high treatment plots and variable popula-
tion reductions in the intermediate treatment plots (Figures 19.1 and 19.2).
The raw and log-transformed data [log (X+l)] were evaluated with a repeated
measures ANOVA. The log transformation seemed appropriate because sample
variability tended to increase with the mean and many zero counts occurred in
high treatment plots. Because of high variability in the data, significant
treatment differences were not found on all dates. For the July dates on ZAPS
I, the high treatment plot had significantly lower tardigrade populations than
the control plot (P=.05) for the raw data only. All other treatment comparisons
were not significantly different (Pi.10). No significant treatment differences
were found in either the raw or transformed data from the September, 1978 ZAPS
I sampling (Pi .10). For the July dates on ZAPS II, a significant treatment
effect was found in comparing the control and high treatment raw data (P=.03)
and transformed data (P=.09).
Considering the frequency of occurrence of the tardigrades in the samples,
there were substantial reductions with increased S02 concentration on all dates
on both sites (Table 19.1). The average reduction was greater on ZAPS I (90%
to 16.7) than on ZAPS II (60% to 15%).
A majority of the tardigrade specimens that were identified was of the
genus Maerobtotus. However, since only total counts were made in each sample,
no attempt was made to determine S02 treatment effect by genus. Whether or not
there is differential sensitivity among the three identified genera remains un-
known.
We could not show significantly reduced tardigrade populations on the high
treatment plots on all dates and ZAPS sites because of high sample variability.
However, since some significant differences were found and there was a sub-
stantial reduction in frequency of occurrence in the samples in the high treat-
ment plots on all dates, we are confident that the tardigrade populations were
reduced by the S02 exposure. The high sample variability was probably the re-
sult of clumped spatial distribution in tardigrades and the small sample core
size (18.1 cm2 surface area). From recent sampling (1979 season) the tardi-
grades have been shown to be restricted to the surface 1-2 cm of the soil pro-
file. Additional sampling with larger diameter cores has been done in an
attempt to confirm the effect of S02 that is apparently reducing the tardigrade
populations.
269
-------
/
/
\
Y~
/
/
a
0 14 July 77
D 8 July 78
\
-| P
'
/
pi
M
b
h
1 I/In
Figure 19.1.
Mean counts of Tardigrades on two field sites in southeastern
Montana in 1977 and 1978 (a = Site I and b = Site II).
20r
16
in
O
x 12
(T
S 8
5
z
4
O
-
-
1 — 1
S02 CONCENTRATION (yq • m"3)
Figure 19.2.
Mean counts of Tardigrades on Site I in September, 1978 in
southeastern Montana.
270
-------
TABLE 19.1. FREQUENCY OF OCCURRENCE OF TARDIGRADES
IN SAMPLES
Treatments
Dates Control Low Medium High
14 July 77
8 July 78
16 Sept. 78
X
14 July 77
8 July 78
X
X
S.E.
80
100
90
90
60
60
60
78.0
8.00
Site I
70
90
75
78.33
Site II
20
30
25
Total
57.0
13.56
60
30
80
56.67
30
20
25
44.0
11.22
20
0
30
16.67
20
10
15
16.00
5.10
CONCLUSIONS
Since the tardigrades reside on the soil surface, they may, as a group,
be more vulnerable to S02 and other air pollutants than nematodes and micro-
arthropods which are distributed deeper into the soil profile.
REFERENCES
Franz, H. 1941. Unterschungen liber die Bodenbiologie alpiner Grlindland und
AckerbBden. Forschungsdienst, 11:355-368.
Franz, H. 1950. Etat de nos connaissances sur da microfaune du sol.
Ecologie (Colloques int. cent. natn. Rech. scient., Paris 1950), pp.
81-92.
Ramazzotti, G. 1959. Tardigradi in terreni prativi. Atti. Soc. ital. Sci.
nat., 98:199-210.
Riggin, G. T., Jr. 1962. Tardigrada of Southwest Virginia: With the
Addition of a Description of a New Marine Species from Florida. Va.
Agric. Exp. Stn. Tech. Bull., 152. 145 pp.
271
-------
SECTION 20
RESPONSE OF RANGELAND GRASSHOPPER POPULATIONS
TO CONTROLLED LEVELS OF S02
T. J. McNary, J. W. Leetham, W. K. Lauenroth, and J. L. Dodd
ABSTRACT
The grasshoppers (Acrididae) on two mixed-grass
prairie sites in southeastern Montana were subjected
to low-level chronic fumigation with S02• A greater
reduction in the grasshopper population occurred with
greater exposure time and with higher S02 concentra-
tions. Possible mechanisms for this population reduc-
tion are discussed.
INTRODUCTION
Past studies show that arthropods respond to air pollution in different
ways. They adapt by industrial melanism (Kettlewell, 1955, 1956, 1961; Creed,
1975; Bishop et al., 1975; and Lees et al., 1973) and/or they modify their be-
havior by avoiding the pollution (Bromenshenk and Gordon, 1978). Population
changes may occur caused by direct pollution toxicity to the insects (Kanaga,
1956; Hillman and Benton, 1972; and Freitag et al., 1973a).
The purpose of this study is to determine the effects of controlled S02
exposure on grasshopper population dynamics and species composition in the
mixed-grass prairie of Montana. Grasshoppers, being herbivores, play an im-
portant role in energy transfer and nutrient cycling.
MATERIALS AND METHODS
The grasshopper population was estimated by a flush census (Pfadt, 1977).
The grasshoppers were flushed by manually brushing the vegetation on a oa.
0.1 • m quadrat. Twenty-five temporary quadrats were systematically located
3-4 m apart along a transect across the treatments. Species and age classifi-
cation (instar) of each individual was noted. Size, shape, color patterns,
songs, and behavior were used for field identification. Collected voucher
specimens were verified by Dr. Robert Pfadt (University of Wyoming). Instar
identifications were based on size of nymph (Brusven, 1967, 1972; Scoggan and
272
-------
Brusven, 1972). Sex of each individual was recorded when possible. Species
that were not positively identified were grouped in the lowest taxonomic level
in the summary.
Censusing was performed at the ZAPS sites in southeast Montana on 10 May,
24 May, 7 June, 15 June, 24 June, 7 July, 22 July, and 7 August in 1977. In
1978 the censusing dates were 26 May, 4 June, 14 June, 1 July, 7 July, 26 July,
3 August, and 17 August. Population estimates were made by replicate for each
sample date. In 1977 and the first three dates in 1978, the population esti-
mates for a replicate were based on one transect per replicate. To obtain a
more precise estimate of population, two transects were followed on each rep-
licate on 1 July 1978 and thereafter. The analysis of variance used was for
repeated measures on a randomized block.
RESULTS AND DISCUSSION
Melanoplus sanguinipes (Fab) was the most common grasshopper found on
both sites in 1977 and 1978. This species overwinters as eggs and matures by
August. The most common spring maturing species was Eritettix simplex
(Scudder). Fourteen other species were observed with fewer species observed
in 1978 than 1977 (Table 20.1).
Analysis of variance of the total grasshopper population and
also that of M. sanguinipes showed a significant interaction of treatment x
date within each year (p < .01). Early in the season the population decreased
as the overwintering species died. Then hatching caused the population to
rise rapidly. This rise was faster on the control and low treatments than on
the high treatments in 1977 and 1978. This results in lower mid and late
season populations in the higher treatments (Figure 20.lb and 20.2b).
A significant interaction also.occurred in the analysis of site x date
within year (p < 0.01) for total and the M. sanguinipes population. In 1977,
at both sites, both the total population (Figure 20.la) and that of M.
sanguinipes (Figure 20.2a) start at similar levels, but the Site II popula-
tion increases faster to a higher level than Site I, before the population
declines in the fall. Differences between years occur mainly as a result of
hatching being delayed in 1978 because of the wet, cold spring. Here the
populations on the two sites are not significantly different until later in
the season.
The trend on each site shows decreased population with increased concen-
tration of S02. This was even more noticeable in 1978 than in 1977. There
were significantly less grasshoppers on the medium and high treatments than on
the control on Site II in 1977, and on both sites in 1978 for the medium and
high treatments (p = 0.0672). These differences also occur in the M.
sanguinipes population for the medium and high treatments on Site II in 1977,
and for all fumigated treatments on Site II in 1978 (p = 0.0457).
We can only speculate on how the S02 reduced the populations. Increased
emigration and decreased immigration rates may be the cause. Canthon spp.
273
-------
NJ
TABLE 20.1. PERCENT SPECIES COMPOSITION OF GRASSHOPPERS ON .52 HA TREATMENTS FUMIGATED WITH DIFFERENT
LEVELS OF S02
Control
Grasshopper species
Aeropedellus clavatus (Thomas)
Ageneotettix deorwn (Scudder)
Amphitornus coloradus (McNeil!)
Auloaara elliotti (Thomas)
Drepanopterna femoratwn (Scudder)
Eritettix simplex (Scudder)
Opeia obscura (Thomas)
Unidentified Gomphoceri nae
Melanoplus aonfusus (Scudder)
M. infantilis (Scudder)
M. packardii ' (Scudder)
M. sanguinipes (Fabricius)
Phoetaliotes nebrascensis (Thomas)
Unidentified Crytacanthacridinae
Arphia pseudonietana (Thomas)
Spharagemon equate (Say)
Trachyrhachys kiaaa (Thomas)
Xanthippus corallipes (Haldeman)
Unidentified Oedipodinae
Unidentified Acrididae
Actual number observed
1977
5.0
1.2
7.5
2.5
3.8
1.2
2.5
3.8
57.5
1 .2
1.2
1 .2
11.2
80.0
1978
0.8
11.5
1.6
18.8
8.2
*t.1
36.9
3.3
*t.1
0.8
1.6
0.8
7.
-------
5r
eg
I
E
QL
LU
CD
>N o r-
1 5 J §
^ 3 ^ ?
,n ** W
i2 CVJ N CVJ N
1977
Control
Low
• Medium
•High
ID
CVJ
•=5 = 3
< <
CD is.
cvj rO !_
1978
Figure 20.1. Seasonal dynamics of total population of grasshoppers fumigated
with four different levels of S02, (a) 1977, (b) 1978.
CM
I
cc.
LJ
CO
I -
Control
• Low
Medium
- — High
c
3
—3
ID
CM
3
3 ~3
—3
CD
h- CVJ
O>
13
1977
1978
Figure 20.2.
Seasonal dynamics of Melanoplus sanguinipes (Fab.) population
fumigated with four different levels of S02, (a) 1977, (b) 1978.
275
-------
beetles avoided immigrating into the S02 treatments at the ZAPS sites (Bromen-
shenk and Gordon, 1978). A relative increase in emigration of the grasshoppers
may result from direct olfactory avoidance or from SC>2 making the forage im-
palatable.
Seasonal displacement rates for female A. ettiotti averaged ca. 20 m
(Mussgnug, 1972). Edwards (1961) concluded that M. sanguinipes moved only
very short distances from a release point. Baldwin et at. (1958) found that
the most actively tagged M. sanguinipes did not migrate outside a 10 m radius
and none migrated more than 30 m from the point of release. Thus the treat-
ment areas are larger than the normal displacement range of the grasshoppers.
Therefore, the reduction in the populations could be attributed to direct con-
tact with the pollutant or from indirect effects of the pollutant on other
components of the ecosystem. Weedon et at. (1939) and Kanaga (1956) have es-
tablished the direct toxicity of high S02 concentrations to insects. The con-
centrations used were higher than the ZAPS studies. No research has been done
on direct toxicity from a long-term (chronic) pollution source.
The SC>2 could affect the grasshopper population indirectly by its effects
on the grasshopper's predators and parasites in other insect populations.
Hillman and Benton (1972) suggest that S02 disrupts the host-parasite balance
resulting in decreased parasitic wasps and increased aphids. Freitag et at.
(1973b) showed a reduction in predatory carabid beetles near a Kraft mill.
This may cause a rapid increase in prey populations. But it has been suggested
that grasshopper populations are not limited by predators and parasites (White,
1976). Therefore, it is possible that the population decrease resulted from
the effects of S02 on the forage of the grasshoppers. Decreases in food qua-
lity and quantity consumed has been shown to decrease egg production in M.
sanguinipes (F.) (Pfadt and Smith, 1972).
Since many chemicals have been shown to control feeding behavior and
survival (Harley and Thorsteinson, 1967), the changes in forage chemistry be-
cause of air pollution could result in insect population changes. The total
sulfur in western wheatgrass was four times greater on the high treatment than
on the control at the end of the growing season. There was also a reduction
in crude protein in the treated plots at Site I in 1976 (Schwartz et at., 1978)
and a reduction in the chlorophyll content (Lauenroth and Dodd, in prep.).
These and other changes in the forage chemistry may be responsible for the
decreases in grasshopper population.
CONCLUSIONS
The fumigation by S02 on the ZAPS sites caused a decrease in the popula-
tion of grasshoppers. The non-statistically significant trend was lower popu-
lation with higher concentration of S02 and length of fumigation. The specific
action of S02 reducing the population was not determined.
REFERENCES
Baldwin, W. F., F. D. Riordan, and R. W. Smith. 1958. Note on Dispersal of
Radioactive Grasshoppers. Can. Entomo.i., 90:374-376.
276
-------
Bishop, J. A., L. M. Cook, J. Muggleton, and M. R. D. Seaward. 1975. Moths,
Lichens and Air Pollution Along a Transect from Manchester to North Wales.
J. Appl. Ecol., 12: 83-93.
Bromenshenk, J. J., and C. C. Gordon. 1978. Terrestrial Insects Sense Air
Pollutants. In: Conference Proceeding, Forest Joint Conference on the
Sensing of Environmental Pollutants, 1977. Am. Chem. Soc., Wash., D. C.
pp. 66-70.
Brusven, M. A. 1967. Differentiation, Ecology and Distribution of Immature
Slant-face Grasshoppers (Acridinae) in Kansas. Tech. Bull. 149. Agr.
Exp. Sta. Kansas State Univ., 59 pp.
Brusven, M. A. 1972. Differentiation and Ecology of Common Catantopinae and
Crytacanthacridinae Nymphs (Orthoptera: Acrididae) of Idado and Adjacent
Areas. Melanderia, 9:32 pp.
Creed, E. R. 1975. Melanism in the Two Spot Ladybird: The Nature and Inten-
sity of Selection. Proc. R. Soc. Lond. Ser. B., 190:135-148.
Edwards, R. L. 1961. Limited Movement of Individuals in a Population of the
Migratory Grasshopper Melanoplus bilituratus (Walker) (Acrididae) at
Kamloops, British Columbia. Can. Entomol., 93:628-631.
Freitag, R., and L. Hastings. 1973a. Kraft Mill Fallout and Ground Beetle
Populations. Atmos. Environ., 7:587-588.
Freitag, R., L. Hastings, W. R. Mercer, and A. Smith. 1973b. Ground Beetle
Populations near a Kraft Mill. Can. Entomol., 105:299-310.
Harley, K. L. S., and J. J. Thorsteinson. 1967. The Influence of Plant Chemi-
cals on the Feeding Behavior, Development, and Survival of the Two Striped
Grasshopper, Melanoplus bivittatus (Say) (Acrididae: Orthoptera). Can. J.
Zool., 45:305-318.
Hillman, R. C., and A. W. Benton. 1972. Biological Effects of Air Pollution
on Insects, Emphasizing the Reactions of the Honey Bee (Apis mellifera
L.) to Sulfur Dioxide. Elisha Mitchell. Sci. Soc., 88:195.
Kanaga, E. E. 1956. An Evaluation of the Use of Sulfur Dioxide in Fumigant
Mixtures for Grain Treatment. J. Econ. Entomol., 49:723-729..
Kettlewell, H. B. D. 1955. Selection Experiments on Industrial Melanism in
the Lepidoptera. Heredity, 9:323-342.
277
-------
Kettlewell, H. B. D. 1956. A Resume of Investigations of the Evaluation of
Melanism in the Lepidoptera. Proc. R. Soc. Lond. Ser. B., 145:297-303.
Kettlewell, H. B. D. 1961. The Phenomenon of Industrial Melanism in Lepidop-
tera. Ann. Rev. Entomol., 6:245-262.
Lauenroth, W. K., and J. L. Dodd. Chlorophyll Reduction in Western Wheatgrass
Exposed to Sulfur Dioxide. (In prep.).
Lees, D. R., E. R. Creed, and J. R. Duckett. 1973. Atmospheric Pollution and
Industrial Melanism. Heredity, 30:227-232.
Mussgnug, G. L. 1972. The Structure and Performance of an Adult Population of
Aulooara elliott-i (Thomas) (Orthoptera, Acrididae) near Billings, Montana.
M. S. Thesis, Montana St. Univ., 97 pp.
Pfadt, R. E. 1977. Some Aspects of the Ecology of Grasshopper Populations In-
habiting the Shortgrass Plains. In: Kulman, H. M., and H. C. Chiang, eds.
1977. Insect Ecology—papers presented in the A. C. Hodson Lectures.
U. Mn. Agr. Expt. Sta. Tech. Bui. 310, 107 pp.
Pfadt, R. E., and D. S. Smith. 1972. Net Reproductive Rate and Capacity for
Increase of the Migratory Grasshopper, Malanoplus sangui-n-ipes
(F.). Acrida, 1:149-165.
Schwartz, C. C., W. K. Lauenroth, R. K. Heitschmidt, and J. L. Dodd. 1978.
Effect of Controlled Levels of Sulfur Dioxide in the Nutrient Quality of
Western Wheatgrass. J. Appl. Ecol., 15:869-874.
Scoggan, A. C., and M. A. Brusven. 1972. Differentiation and Ecology of
Common Immature Gomphocerinae and Oedipodinae (Orthoptera: Acrididae) of
Idaho and Adjacent Areas. Melanderia, 8:77 pp.
Weedon, F. R., A. Hartzell, and C. Setterstrom. 1939. Effects on Animals of
Prolonged Exposure to Sulfur Dioxide. Contrib. Boyce Thompson Inst.,
10:281-324.
White, T. C. R. 1976. Weather, Food and Plagues of Locusts. Oecologia,
22:129-134.
278
-------
SECTION 21
THE EFFECTS OF LOW CONCENTRATIONS OF SULFUR DIOXIDE ON THE
GROWTH, QUALITY AND NITROGEN FIXATION OF ALFALFA
D. T. Tingey, G. E. Neely and M. L. Gumpertz
ABSTRACT
The effects of ambient concentrations of sulfur
dioxide on the growth, yield, quality and nitrogen
fixation of alfalfa .(Medicago sativa L.) were deter-
mined. Plants were exposed throughout the growing
season to median sulfur dioxide concentrations of 0,
2, 6, 15 or 20 pphm. Sulfur dioxide at 15 and 20 pphm
significantly reduced plant growth, yield and nitrogen
fixation. Total sulfur in the forage increased
proportionally to the median sulfur dioxide concentra-
tion. No signficant effects in forage quality were
observed.
INTRODUCTION
Increased sulfur dioxide emissions from power generation and other indus-
trial processes pose an increased threat to the production of forage. Bell
and Clough (1973) reported that continuous exposures of 6.7 and 12 pphm sulfur
dioxide for 26 and 9 weeks, respectively, decreased the yield of ryegrass
approximately 50%. Dactylis (Ashenden, 1978) exposed to 11 pphm sulfur
dioxide for 4 weeks exhibited significant reductions in the number of tillers,
leaf area and dry weights of leaves and roots. Katz and Ledingham (1939),
summarizing a series of experiments, concluded that chronic exposures of 10 to
30 pphm sulfur dioxide for extended periods would have no effect on the yield
of alfalfa under normal conditions. Intermittent exposures to 10 pphm sulfur
dioxide for 7 hr per day 6 days per week through the growing season had no
effect on alfalfa forage yield, but there was a tendency toward reduced root
weight (Thomas et at,, 1943). More recently, however, Tingey and Reinert
(1975) reported significant reductions in alfalfa forage and root growth by
exposure to 5 pphm S02 (8 hr/day, 5 days/week) from seeding to harvest.
Sulfur dioxide-induced reduction in root dry weight in grass (Ashenden, 1978)
and alfalfa (Thomas et al., 1943; Tingey and Reinert, 1975) indicates that the
roots received less photosynthate. This suggests that sulfur dioxide expo-
sures could affect processes that require an input of photosynthate, such as
symbiotic nitrogen fixation.
279
-------
Howell (1975) exposed alfalfa to 150 pphm sulfur dioxide for 2 hr and
analyzed the forage 36 hr later for indicators of foliage quality. He
reported significant reductions in the protein and fat content and increases
in the total carbohydrate and fiber content. However, at lower concentrations
(10 to 40 pphm for 1 to 14 days) there were no effects on either the nitrogen
fraction or the carbohydrates in alfalfa (Katz and Pasternak, 1939).
The objectives of this study were (1) to assess the effects of continu-
ously varying sulfur dioxide concentrations, which simulate ambient exposures,
on the growth and yield of alfalfa; (2) to determine the effects of sulfur
dioxide on indicators of forage quality; and (3) determine the impact of
sulfur dioxide on symbiotic nitrogen fixation.
MATERIALS AND METHODS
Plant Growth and Exposure.
Alfalfa (Medicago sativa L.) cv Mesa-Sirsa seeds were inoculated with a
commercial Rhisobium inoculum and planted in 20 mesh silica sand in 22 cm dia.
plastic pots and watered daily with a modified Hoagland nutrient solution,
minus nitrogen. To promote rapid plant growth, the nutrient solution was
supplemented with 220 ppm N during the first two weeks and during the third
week with 110 ppm N. After emergence, the plants were thinned to two individ-
uals per pot.
Plants were placed in field exposure chambers (10 pots/chamber) (Heagle
et al.3 1973) when approximately 10 cm tall and exposed to a simulated sulfur
dioxide time series for 68 days. A gradient in median sulfur dioxide con-
centration was designed with a control and four concentrations of sulfur
dioxide (Table 21.1). The exposure simulation duplicated the correlation
structure of ambient air monitoring data (Air Quality Criteria for Sulfur
Oxides, 1970) and had a standard geometric deviation of 2.44 (Male et al.t
1978; Male, 1980). The sulfur dioxide concentrations used in the time series
exposure were generated using a lognormal stochastic model. The same exposure
pattern (sulfur dioxide concentrations increased and decreased at the same
time in all the treatment chambers) was used in all treatment chambers except
the median sulfur dioxide concentrations were different. The sulfur dioxide
levels were varied daily at hourly intervals between 8:00 a.m. and 4:00 p.m.
PDT. During the remaining 16 hr period the sulfur dioxide concentration
remained constant at the median concentration calculated from the stochastic
model for that 16 hr period. The median sulfur dioxide concentrations for
each harvest are shown in Table 21.1. Sulfur dioxide was dispensed and
controlled in the field exposure chambers as described by Heagle e~t at. 1974.
It was monitored using flame photometric detectors which were calibrated daily
using sulfur dioxide permeation tubes.
The plants were exposed for 35 and 33 days, respectively, prior to the
first and second harvests. All plants were harvested when the plants in the
control chamber reached the 1/10 bloom growth stage. Tops were clipped 8 cm
above the growth medium; this is subsequently referred to as the top or forage
portion. At the second harvest, the forage was once again clipped at the 8 cm
280
-------
level and the remaining plant material was separated into stubble and roots.
After harvest the plant parts were dried at 70°C until the material reached a
constant weight.
TABLE 21.1. MEDIAN SULFUR DIOXIDE CONCENTRATIONS (PPHM).
Desired
(nominal)
0
2
6
15
20
Actual Median
1st Harvest
0.0
2.1
6.0
17.2
13.9
Actual Median
2nd Harvest
0.0
2.4
5.6
17.9
22.8
Actual Median
Seasonal
0.0
2.3
5.8
17.6
18.6
Chemical Analysis.
The dried plant material was ground in a Wiley Mill (40 mesh screen);
aliquots were used for the various metabolite analyses. To determine Kjeldahl
nitrogen, samples were digested (Taras Q± oil. , 1971) and the resulting ammonia
was measured using the phenol-hypochlorite reaction (U.S. Environmental
Protection Agency, 1974). Total nonstructural carbohydrates were extracted
and analyzed using the method of Smith (1969). Total sulfur was estimated by
digesting the plant material in 70% nitric and perchloric acid; the resulting
sulfate was determined turbidimetrically (Rand et al.3 1976). Dry matter
digestibility of the forage was estimated using the -in vitvo method of Tilley
and Terry (1963).
Data Analysis.
Preliminary data analyses indicated that the variances of the response
variables increased with the means for all responses except digestibility.
Therefore, the data were transformed to their respective logarithms to stabi-
lize the variances. All statistical analyses except those for digestibility
were performed on the transformed data. The anti-logarithms of the data were
computed to simplify the presentation of results. Data from each harvest were
subjected to individual statistical analysis because the median sulfur dioxide
concentrations for the two harvests differed (Table 21.1).
The data from the first harvest were subjected to analyses of variance to
determine if there was an effect of sulfur dioxide on any of the response
variables. Data from the second harvest were subjected to analyses of vari-
ance to determine if the sulfur dioxide treatment affected the response vari-
ables or the plant parts differently. When an analysis of variance yielded a
significant F-test, at a = 0.05, the Bonferroni method of multiple comparisons
was used to determine which treatment means differed significantly from the
control mean. For all response variables except digestibility, the test for
determining if a particular treatment mean was significantly different from
the control mean consisted of comparing the ratio x ,/x with a
control treatment
281
-------
critical value. For digestibility, the differences x n - x,_
control treatment
were compared with a critical value. The critical values from the Bonferroni
multiple comparison procedures are shown in the appropriate figure legends.
RESULTS
Forage dry weight of the control plants for harvest 2 was approximately
twice that of harvest 1 (Figure 21.1A) indicating better plant growth. For
both harvests, forage dry weight for the lowest two sulfur dioxide concen-
trations was similar to or slightly larger (not significant) than the
controls. However, higher concentrations of sulfur dioxide elicited signifi-
cant reductions in forage dry weight (yield). Foliar injury was observed in
both harvests at the higher sulfur dioxide concentrations (15 and 20 pphm,
nominal). Visually, the plants in the 6 pphm treatment appeared to have fewer
leaves and more spindly stems. However, these visual symptoms were not
reflected in a significant effect on plant yield.
At harvest 2 the sulfur dioxide treatment impacts were measured on the
forage, stubble, roots and total plant dry weight (Figure 21.IB). There were
no significant decreases in the growth of any plant part at the 2 or 6 pphm
sulfur dioxide treatment but significant growth reductions for all plant parts
occurred at the higher sulfur dioxide concentrations.
In harvest 2, the proportion of dry weight in the root, stubble and
forage was determined. Across treatments, the proportion of the plant dry
weight in the roots was essentially constant (approximately 42%). The propor-
tion of the dry weight in the stubble increased from approximately 18% for the
control plants to approximately 30% at the highest sulfur dioxide treatment.
The proportion of dry weight in the forage decreased from approximately 39% in
the control plants to approximately 25% at the highest sulfur dioxide treat-
ment. This decrease in dry matter partitioning into the forage may, in part,
account for the visual observation of weaker stems and less leaf matter in the
6 pphm treatment.
The concentration of nitrogen, total non-structural carbohydrates, total
sulfur and in vitro dry matter digestibility were determined to estimate the
impact of the sulfur dioxide on forage quality (Figure 21.2). In the control
plants, for both harvests 1 and 2, the nitrogen content was 3.2 and 3.4%,
respectively (Figure 21.2A) which is equivalent to protein contents of approx-
imately 20 and 21%. There was no significant change in the nitrogen content
of the forage with respect to sulfur dioxide concentration for harvest 1.
However, for harvest 2 there was an approximately linear increase in nitrogen
content with sulfur dioxide concentration.
Total nonstructural carbohydrates (Figure 21.2B) tended to be higher for
harvest 2 than for harvest 1. Although no significant differences in carbo-
hydrate content were detected between treatments, the carbohydrate contents
tended to increase with increasing sulfur dioxide concentrations in harvest 1
and were variable in harvest 2 with a tendency to decrease with increasing
sulfur dioxide concentration.
282
-------
0 5 10 15 20 25
Median Sulfur Dioxide
Concentration pphm
5 10 15 20
Median Sulfur Dioxide Concentration
pphm
Figure 21.1.
The influence of sulfur dioxide on the growth and yield of
alfalfa. Each mean is based on 10 observations. Open symbols
indicate that the treatment mean was not significantly differ-
ent from the control (a = 0.05). (A) The yield of alfafa for
each harvest. The critical values are 1.66 and 1.62 for harvest
1 and 2, respectively. (B) The growth of the individual plant
components and total plant growth for harvest 2, The top line
describes the summed weights of the individual plant components
and is equal to total plant growth. The weights of the indiv-
idual plant parts are indicated by the appropriate cross-hatched
areas. The critical value for the individual components is 1.62
and total growth is 1.42.
283
-------
5.0
B. Total Nonstructural
Carbohydrates
\ Harvest 2
10 15 20 25
D. in vitro Digest-
Harvest 2
0.0
5 10 15 20 25 05 10 15 20 25
Median Sulfur Dioxide Concentration pphm
Figure 21.2.
The influence of sulfur dioxide on indicators of forage quality.
Each mean is based on 10 observations. Open symbols indicate
that the treatment mean was not significantly different than
the control while solid symbols indicate that the mean was sig-
nificantly different from the control (a = 0.05). (A) Nitrogen
content of the forage. The critical values for harvests 1 and
2 are 1.12 and 1.31. (B) Forage content of total nonstructural
carbohydrates. The critical values for harvest 1 and 2 are 1.29
and 1.37. (C) The total sulfur content of the forage. The crit-
ical values for harvests 1 and 2 are 1.19 and 1.40. (D) The
disappearance of dry matter or digestibility. For harvest 1,
the analysis of variance indicated no significant differences.
For harvest 2 the critical value is 3,41.
284
-------
Total sulfur content in the forage increased in an approximately linear
manner with increasing sulfur dioxide concentration (Figure 21.2C) for both
harvests. The sulfur content ranged from approximately 0.46 and 0.47% for the
controls to a high of approximately 1.8% for the highest sulfur dioxide
concentration for each harvest. The concentration of sulfur in the stubble
showed a similar trend to that of the foliage but the content was approx-
imately half that in the forage. The sulfur concentration in the control
plant roots was approximately 0.69%, but when the sulfur dioxide concentration
exceeded approximately 6 pphm, the concentration in the roots decreased to
approximately 0.45%.
In vitro dry matter digestibility (Figure 21.2D) for harvest 1 was not
significantly different among treatments and averaged 73.7%. The digesti-
bility for harvest 2 was similar to harvest 1 for the control and the lowest
two sulfur dioxide treatments, however, there were significant increases in
digestibility to 81.4 and 78.7%, respectively, for the highest two sulfur
dioxide treatments.
To estimate symbiotic nitrogen fixation, giants were inoculated with a
Phisobium culture and grown in a nitrogen-free medium. Using this method the
nitrogen content of the plant depends on nitrogen derived from the atmosphere
via symbiotic nitrogen fixation. The total nitrogen fixed was estimated by
determining the nitrogen content of the plant tissue and the total amount of
plant tissue produced. The control plants fixed approximately 1050 mg/pot of
nitrogen for the season (Figure 21.3). There was a nonsignificant increase in
nitrogen fixation at the 2 and 6 pphm levels. In the 15 and 20 pphm treat-
ments large decreases in nitrogen fixation occurred.
The total nonstructural carbohydrate content of the stubble and roots
were analyzed to determine if sulfur dioxide altered the concentration
(Figure 21.4). Total nonstructural carbohydrates in the forage tended to
increase slightly in harvest 1 and decrease in harvest 2 although the results
for both harvests were somewhat variable (Figure 21.2B) and the differences
were not significant. In contrast, the concentration of total nonstructural
carbohydrates in the stubble was similar to the control except at the highest
sulfur dioxide concentration, which was significantly less. The nonstructural
carbohydrate content of the roots was similar to or slightly less than the
control at the lowest two sulfur dioxide treatments. Higher concentrations of
sulfur dioxide elicited significant reductions in the carbohydrate content in
the roots. These results indicate that sulfur dioxide significantly reduced
the partitioning of carbohydrate into the roots.
DISCUSSION
Plant growth and yield in the low concentration sulfur dioxide treatments
tended to be greater than the control but not significant statistically.
Thomas et al. (1943) also reported a nonsignificant stimulation of alfalfa
growth from low concentrations of sulfur dioxide. The trend toward increased
growth observed in our study was not the result of the atmospheric sulfur
dioxide overcoming a sulfur deficiency. Foliar analysis (Figure 21.2C) indi-
cated that sulfur content exceeded 0.45% which is much greater than 0.20 to
0.22% established as the critical level for adequate sulfur nutrition (Bickoff
285
-------
1400
200
0 5 10 15 20 25
Median Sulfur Dioxide Concentration
pphm
Figure 21.3.
The effect of sulfur dioxide on total nitrogen fixation by
fa plants for the growing season. Open symbols indicate that
the treatment -mean was not significantly different than the
control while solid symbols indicate that the mean was signif
icantly different from the control (a = 0.05). Each mean is
based on 10 observations and critical value is 1.49.
ei> al.j 1972). Low concentrations of sulfite, an oxidation product of sulfur
dioxide in plant tissue, will stimulate photosynthesis (Libera et al. 3 1973);
this stimulation may explain the trend toward increased growth observed in the
low concentrations of sulfur dioxide.
286
-------
35.0
5.0
Roots
I
0 5 10 15 20 25
Median Sulfur Dioxide Concentration
pphm
Figure 21.4.
The influence of sulfur dioxide on the total nonstructural carbo-
hydrate content of the stubble and roots. Open symbols indicate
that the treatment mean was not significantly different from the
control while solid symbols indicate that the mean was signif-
icantly different from the control (a = 0.05). Each mean is
based on 10 observations and the critical value is 1.37. Data
were collected at harvest 2.
Plant growth was reduced at lower sulfur dioxide concentrations in our
study than has been reported for other chronic field studies with alfalfa
(Thomas et at., 1943; Katz and Ledingham, 1939). A more sensitive cultivar
may have been used in our study; or the continuous exposures to fluctuating
sulfur dioxide concentrations, rather than a constant level as in previous
studies may have caused effects at lower concentrations.
287
-------
The reduced partitioning of carbohydrates into the roots is similar to
previous reports for cocksfoot (Ashenden, 1978) and alfalfa (Thomas et at.3
1943). Short term exposures to high sulfur dioxide concentrations have been
reported to reduce both photosynthesis and the translocation of photosynthate
from the leaves to other sinks (Teh and Swanson, 1979). A reduction in photo-
synthate translocation could account, in part, for the reduced levels of
photosynthate in the roots of the alfalfa plants. The roots accumulate
carbohydrate only in excess of that needed to maintain foliar growth (Brown
et al., 1972; Smith and Marten, 1970). The reduction in storage carbohydrates
in the roots may have long-range significance for the vigor and the production
of the crop. Root carbohydrate reserves provide the energy necessary for
foliar growth of alfalfa following defoliation (harvest).
The forage quality parameters of the control plants were similar to those
reported for field grown alfalfa crops (Bickoff et al., 1972). In our study,
sulfur dioxide had little effect or caused a slight increase in protein
content, and caused no distinct change in the carbohydrate content. The
increased nitrogen content in harvest 2 is of little significance because the
forage production was significantly less so the total amount of protein avail-
able per unit of ground area was also reduced significantly.
Katz and Pasternak (1939) reported no change in the nitrogen fractions or
carbohydrates of alfalfa exposed to 10 to 40 pphm sulfur dioxide for 1 to 14
days. Howell (1975) reported that short term exposures to high concentrations
of sulfur dioxide reduced the protein content and increased the carbohydrate
content of forage. The digestibility of all our samples was high and the
slight but significant increase with the high sulfur dioxide treatments would
have no practical value. The lack of significant effects on forage quality in
our study is similar to results reported by Cunningham et al. (1937); they
reported no difference in performance of dairy cows fed either control alfalfa
or alfalfa with less than 25% of the leaf area showing sulfur dioxide injury.
The total sulfur content of the alfalfa foliage increased proportionally to
the sulfur dioxide concentration. Faller et al. 3 (1970) reported that
inorganic sulfur was the major sulfur component accumulated in plant foliage
from sulfur dioxide fumigations. Inorganic sulfur presumably accounted for
the large total sulfur increase observed in our alfalfa forage. The majority
of the sulfur dioxide absorbed from the atmosphere was retained in the forage
and lesser amounts were exported to stems and roots (Faller et al.s 1970;
Yamazoe, 1973). Also, plants fumigated with sulfur dioxide absorbed less
sulfur through the roots (Thomas et al. 3 1943, Scharer et al., 1975). The
suppression of sulfur uptake from the soil and the low export of sulfur from
the leaves to the roots could explain the decreased sulfur content of the
roots with increased sulfur dioxide exposure observed in our study.
Our results indicate that symbiotic nitrogen fixation was suppressed at
the higher sulfur dioxide concentrations. The reduction in nitrogen fixation
is expected since metabolites from leaves and stems are translocated to the
roots and the carbohydrate levels of the roots were reduced. Carbohydrates
provide the energy needed to reduce and incorporate atmospheric nitrogen into
organic forms (Gibson, 1976). This suggests that the reduction in carbo-
hydrate translocated to the alfalfa roots decreased the amount of nitrogen
fixed and thus reduced, in part, plant growth and yield.
288
-------
CONCLUSIONS
Sulfur dioxide suppressed alfalfa growth, yield and nitrogen fixation
only at the highest two concentrations (15 and 20 pphm). Total sulfur in the
forage increased proportionally to the median sulfur dioxide concentration.
Forage quality was not impaired by the sulfur dioxide exposures.
REFERENCES
Air Quality Criteria for Sulfur Oxides. 1970. National Air Pollution Control
Administration. Washington, DC. AP-50.
Ashenden, T. W. 1978. Growth Reductions in Cocksfoot (Dactylis glomerata L.)
as a Result of S02 Pollution. Environ. Pollut. 15:161-166.
Bell, J. N. B. and W. S. Clough. 1973. Depression of Yield in Ryegrass to
Sulfur Dioxide. Nature, 241:47-49.
Bickoff, E. M., G. 0. Kohler and D. Smith. 1972. Chemical Composition of
Herbage. In: C. H. Hanson, ed. Alfalfa Science and Technology.
American Society of Agronomy, Inc. Madison, Wisconsin. pp. 247-282.
Brown, R. H., R. B. Pearce, D. D. Wolf and R. E. Blaser. 1972. Energy Accum-
ulation and Utilization. In: C. H. Hanson, ed. Alfalfa Science and
Technology. American Society of Agronomy, Inc. Madison, Wisconsin, pp.
143-166.
Cunningham, C. C., L. H. Addington and L. T. Elliot. 1937. Nutritive Value
for Dairy Cows of Alfalfa Hay Injured by Sulfur Dioxide. J. Agri. Res.,
55:381-391.
Faller, N., K. Herwig and H. Kuehn. 1970. The Absorption of Sulfur Dioxide
(35S02) From the Air. II. Absorption, metabolism and distribution in
plants. Plant and Soil, 33:283-295.
Gibson, A. H. 1976. Limitation to Dinitrogen Fixation by Legumes. In:
Nitrogen Fixation. W. E. Newton and C. J. Nyman, eds. Washington State
University Press, Pullman, pp. 400-428.
Heagle, A. S., D. E. Body and W. W. Heck. 1973. An Open-Top Field Chamber to
Assess the Impact of Air Pollution on Plants. J. Environ. Quality,
2:365-368.
Heagle, A. S., D. E. Body and G. E. Neely. 1974. Injury and Yield Responses
of Soybean to Chronic Doses of Ozone and Sulfur Dioxide in the Field.
Phytopathology, 64:132-13.
Howell, R. K. 1975. Ozone and Sulfur Dioxide—Their Effect on Nutrient
Composition of Forages. Reg. Conf. Rumen Funct. 1975:4-5.
289
-------
Katz, M. and G. A. Ledingham. 1939. Experiments on Yield of Barley and
Alfalfa. Chapter XIII. Effect of Sulfur Dioxide on Vegetation. National
Research Council of Canada, pp. 322-368.
Katz, M. and D. S. Pasternak. 1939. Effect of Fumigation on Some Chemical
Constituents of Barley, Wheat and Alfalfa. Chapter XIV. Effect of Sulfur
Dioxide on Vegetation. National Research Council of Canada. pp.
369-392.
Libera, W., H. Ziegler and I. Ziegler. 1973. Forderung der Hill-Reaktion und
der C02-Fixierung in Isolierten Spinatchloroplasten durch Niedere Sulfit-
konzentrationen. Planta, 109:269-279.
Male, L. 1980. Time Series Experiments for Predicting Plant Growth Response
to Pollution. Part I.—Time Series Model and Experimental Design.
Atmospheric Environment (Submitted).
Male, L. , J. Van Sickle and R. Wilhour. 1978. Time Series Experiments for
Studying Plant Growth Response to Pollution. EPA-600/3-78-038. U.S.
Environmental Protection Agency, Corvallis, OR.
Rand, M. C., A. E. Greenberg and M. J. Taras. 1976. 427 C. Turbidimetric
Method. In: Standard Methods for the Examination of Water and Waste-
water, 14th edition. American Public Health Association, Washington, DC.
pp. 496-498.
Scharer, M. , C. Brunold, and K. H. Erismann. 1975. Hemmung der Sulfatauf-
nahme durch Lemna minor L. durch S02 in Subletalen Konzentrationen,
Experientia, 31:1414-1415.
Smith, D. 1969. Removing and Analyzing Total Nonstructural Carbohydrates
from Plant Tissue. Research Report 41. University of Wisconsin, Madison
Wisconsin.
Smith, L. H. and G. C. Marten. 1970. Foliar Regrowth of Alfalfa Utilizing
14C-labeled Carbohydrates Stored in Roots. Crop Science, 10:146-150.
Taras, M. J. , A. E. Greengerg, R. D. Hoak and M. C. Rand, eds. 1971. Nitro-
gen (organic). In: Standard Methods of the Examination of Water and
Wastewater, 13th edition. American Public Health Association, Wash-
ington, DC. pp. 244-248
Teh, K. H. and C. A. Swanson. 1979. Comparative Inhibition of Photosynthesis
and Translocation by Sulfur Dioxide in Bush Bean. Plant Physiology (sup-
plement), 63:34.
Thomas, M. D. , R. E. Hendricks, T. R. Collier and G. R. Hill. 1943. The
Utilization of Sulphate and Sulfur Dioxide for the Sulfur Nutrition of
Alfalfa. Plant Physiol., 18:345-371.
Tilley, J. M. A. and R. A. Terry. 1963. A Two Stage Technique for the in
vitro Digestion of Forage Crops. J. British Grassl. Soc., 18:104-111.
290
-------
Tingey, D. T. and R. A. Reinert. 1975. The Effect of Ozone and Sulfur Diox-
ide Singly and in Combination on Plant Growth. Environ. Pollut.,
9:117-125.
U.S. Environmental Protection Agency. 1974. Methods for Chemical Analysis of
Water and Wastes, Washington DC. 298 pp.
Yamazoe, F. 1973. Distribution and Reaction of Sulfur Dioxide after Absorp-
tion by Plants. Jap. Agric. Res. Q., 7:243-247.
291
-------
SECTION 22
REMOTE SENSING OF S02
EFFECTS ON ZAPS
J.E. Taylor, W.C. Leininger, and M.W. Hoard
ABSTRACT
Color infrared aerial photographs were taken during
June, August and October 1978 on the ZAPS sites.
1:15,000 photos were taken as synoptic records of the
third and fourth years of sulfur dioxide fumigation ef-
fects of ZAPS II and I, respectively. Larger scale pho-
tography (1:3,700) was analyzed visually and densito-
metrically for color and density differences associated
with S02. Visual bleaching with increasing fumigation
rates was observed on both sites at all dates; this
effect was most conspicuous in the August sample.
Densitometric data indicated a general tendency for de-
creased color intensity in red, green, and blue wave
lengths. Total density likewise decreased across treat-
ments, but red/green ratios were insensitive. Although
intermediate levels varied, significant density reduc-
tions were seen between control and high S02 rates at
every sample date. Significant linear regressions
related stress and response in several comparisons,
especially in the red wave lengths. This indicates a
loss of actively growing plant material. Ground-
level photographs were taken on the ZAPS photo plots
and on the fertilizer study conducted by Colorado
State University. Aspect views for visual compari-
sons among treatments and years also were taken.
INTRODUCTION
We have been using remote sensing to monitor the effects of S02 on veg-
etation on the ZAPS and in the vicinity of Colstrip since 1974. Ground-
level and aerial photography have been used in our monitoring efforts.
Ground-level photo techniques and data for the Colstrip area are discussed
in section 4 ; only data on the ZAPS are reported herein.
292
-------
MATERIALS AND METHODS
Aerial Photography
Procedures
In our aerial photography we have concentrated on developing procedures
which use standard, readily available equipment, since if a system is to be
applied in a variety of situations it must not necessitate highly specialized,
costly investments.
For our overflights we use a Cessna 182 airplane which can easily handle
the required elevation range of 500 to 7,000 feet above the ground. The plane
we use is leased from Miles City Aero Services, Miles City, Montana. This
aircraft has been modified by the addition of a 30.5 cm diameter belly port,
which accepts a special camera mount, designed and manufactured by W.E.
Woodcock of Miles City (Woodcock, 1976). The camera is fitted with a 50 or
80 mm lens depending on the desired photo scale.
In addition to the mounted camera, a second camera is used for making
oblique photographs from the air. This kind of photography supplements the
more traditional vertical imagery, in that it is more representative of
familiar aspects of scenes for interpretation, display, etc.
We use various photo scales, depending upon the amount of ground detail
(resolution) and the area per frame (coverage) required for different pur-
poses. All of our vertical aerial photography is flown with 60% end lap so
it can be viewed stereoscopically.
Our largest scale is 1:3,700, which is obtained by flying at 305 meters
above ground with an 80 mm lens. Larger scales are effectively precluded by
the minimum 1/500 second shutter speed of the camera. At 195 km per hour,
motion causes a slight displacement in the direction of flight. This is
negligible at smaller scales, but becomes noticeable at lower elevations or
longer lens focal lengths.
Even with image displacement, photography at this large scale shows
excellent ground detail, discriminating individuals of many plant species
and even 10 x 15 cm plastic marking flags and the wires supporting them.
When more coverage is desired and ground detail is less critical, flying
altitude is increased or lenses changed to shorter focal lengths. For in-
stance, to get each ZAPS location in one frame we use a scale of 1:15,000.
This still allows good species discrimination and also reveals more of the
surrounding areas for topographic analysis, drift detection, etc.
For optimum discrimination of infrared reflectance, CIR film requires a
minus blue filter, which removes some of the shorter wave length radiation
to which the film is also sensitive. The manufacturer's recommendation is a
yellow (Wratten 12) filter. We prefer a Wratten 15, which is a little more
orange colored, and which cuts off slightly more of the yellow-orange portion
293
-------
of the spectrum. In our situation, this improves the separation of infrared
reflectances, which are rendered on CIR in tones of red. Also, it produces
ground tones which are less blue than those obtained with the 12 filter, and
we find these more visually pleasing.
In 1978, aerial photo missions were conducted over the ZAPS on June 28,
August 8, and October 9. No aerial photography was taken of the Colstrip
sites. Color I. R. film (Kodak #2443) was exposed at scales of 1:3,700 and
1:14,000 for all dates. In addition, color negative film (Kodak Vericolor
VPS-120) was used at the 1:3,700 scale in June and both scales in October,
Film Processing
All CIR film is sent to Precision Photo— in Dayton, Ohio for processing.
An optical density step wedge is exposed onto the film end before processing
for density and color calibration. The color negative film is commercially
processed within the State.
Printing
We use the Cibachrome — process for all infrared color printing be-
cause of its sharpness, good color saturation, and resistance to fading. We
can produce these color prints in our laboratory. Prints from color negative
film are made by us or by commercial firms.
Aerial Photo Analysis
Visual Observations
The simplest and easiest kind of photo interpretation (PI) of S02
effects is identification on the photographs of ground abnormalities. Shape,
size, shadows, texture, and proximity .to other objects all are clues for PI,
but in color systems (including CIR), color is the most conspicuous and use-
ful characteristic in detecting temporal and spatial changes. When we re-
ceive the photos from processing, a visual comparison is made between fumiga-
tion plots and adjacent vegetation to the plots. We examine color and density
differences for plant communities and inspect the boundaries of the plots for
possible drift patterns.
Densitometric Analysis
We also sample optical density within the film areas which represent
the fumigation plots. This produces numerical data which can be used in
regression and correlation expressions with other data types. It allows
rigorous characterization of image colors.
The densitometer used in these analyses is a Macbeth TD 524 (Status A),
with a 1 mm spot. This instrument reads the density of aerial transparencies
in the red, blue, and green spectral ranges, and also gives a total density
— No endorsement of specific brand names is implied.
294
-------
value for each observation. Transmission density, as analyzed here, refers
to the proportion of light projected onto the transparency which passes
through. Density is expressed as the common logarithm of opacity, which is
defined as the reciprocal of transmittance. Thus, if 10 percent of the
light striking the film is transmitted, opacity is 10 (1/0.1), and the den-
sity is 1.00 (log 10).
The density values which are yielded in this analysis are therefore re-
lated to the color saturation of the wave length being discussed. Thus,
a high density for red indicates a high red content in the image.
In reading, 10 samples (of 1 mm diam.) are selected from each treatment
area recorded on the film (5/north replication and 5/south replication).
Red, green, blue and total density values are determined at each sample plot.
At the scale of 1:3,700, a 1 mm spot represents a ground area of 10.75
m . Thus, each reading integrates the reflectance exhibited by all parts of
the scene. Gas delivery pipes, with their high reflectances, are avoided in
placing the densitometer spot.
For the regression analyses, we calculated mean S02 rates from the start
of the fumigation season until the time of photography (Table 22.1). We
feel that since we are photographing current annual growth in the sites, the
fumigation rates for current year should best represent the treatment
levels influencing photographic reflectance.
TABLE 22.1. 1978 MEAN S02 RATES (PPHM) USED FOR REGRESSION ANALYSES*
T.V, *- i, ZAPS !
Photography
date
June 28
August 8
October 9
A B C D
1.05 2.10 3.60 7.05
1.07 2.30 3.90 7.60
1.06 2.46 4.40 7.80
ZAPS II
A B C D
1.15 2.40 4.25 8.4
1.13 2.30 4.10 8.56
1.18 2.38 4.32 8.94
* Data from Dr. Eric Preston.
Ground-Level Photography
Permanently marked photo plots on the ZAPS were stereoscopically photo-
graphed with color film on June 18, August 3, and September 10, 1978.
Details on photo plot monitoring procedures are given by Taylor e~t al., 1976.
Color aspect photos of the site from adjacent vantage points were also made.
Detailed vegetation maps for the photo plots were constructed (see section 4
for details) with species identification on the photos to obtain quantitative
295
-------
data on species canopy coverage and frequency. These data are to be compared
with similar data taken from the plot photographs and will be used to eval-
uate the feasibility of estimating these parameters from photo plots.
Photo plots were established and stereoscopically photographed (June 4
and July 16) on Colorado State University's fertilization study on ZAPS I.
Plots were established on the southern replication only. Both color and
color I. R. film was exposed in June, while only color I. R. film was taken
in July. These photos were sampled densitometrically in order to provide
digital data that can be used by C.S.U. in assessing the effects of SC>2 on
grassland under varying regimes of sulfur and nitrogen.
RESULTS AND DISCUSSION
Aerial Photo Analysis
Visual Observations
Visual differences continue to be observed across fumigation plots at
all seasons. With increasing S02 rates, treatment plots appear increasingly
lighter in tone. This is most apparent at the peak of growth and is less
noticeable as the vegetation matures. This is evident in the field as well
as in both color and infrared color photography.
Densitometric Analysis
Photographic density values for the three 1978 aerial missions using
CIR film are shown in Figures 22.1, 22.2 and 22.3 (June, August and October
flights, respectively.)
The colors quantified by the densitometer are those produced in CIR
film. The relationships among original and CIR image colors and some
examples used in interpretations are given in Table 22.2.
June
ZAPS I shows a general decrease in density with fumigation for all
colors. There was a slight rise in blue and green on the C (medium) plot,
but the red content of the images, re.cording infrared reflectance, decreased
linearly. Comparing the extreme rates, the high fumigation was signifi-
cantly less dense than the control in most cases.
On ZAPS II, the decreased color densities with fumigation were more
striking, especially between control and high rates.
August
The color content of the film was higher than in June, which could be
due to different proportions of plant and substrate reflectance, or to dif-
ferences in film exposure, and/or processing. Thus, comparisons of density
296
-------
1.2 -
+
RED DENSITY
O- O GREEN DENSITY
O O BLUE DENSITY
1.0-
0.8-
\ T
p
O
I
Pu
0.6-
0.4-
L
OAB
D
\
\
\
i^
A
ZAPS I
i
C
ZAPS II
D
Figure 22.1. Densitometric readings on ZAPS plots; June 1978
(1:3,700 color infrared film)
297
-------
2.0-
1.8-
1.6-
1.4-
GO
!r; 1.2-
1.0-
0.8-
0.6.
0.4_
-f RED DENSITY
--0 GREEN DENSITY
-O BLUE DENSITY
Si
k
0 A B C
ZAPS I
0 A B C
ZAPS II
Figure 22.2. Densitometric readings on ZAPS plots;
August 1978 (1:3,7000 CIR film)
298
-------
1.2-,
1.0-
0.8-
Q
U
g 0.6-
o
E-i
O
IE
P-i.
0.4-
•H
RED DENSITY
GREEN DENSITY
BLUE DENSITY
I
0 A
T
B
I
D
0 A
ZAPS I
B C
ZAPS II
D
Figure 22.3. Densttometrlc readings on ZAPS plots; October
1978 (1:3,700 CIR film).
299.
-------
TABLE 22.2. RENDITION OF SUBJECT COLORS IN CIR (COLOR INFRARED) FILM
Original (natural) color
CIR rendition
Wave lengths
(nanometers)
630-700
500-600
400-500
Color
rendition
red
green
blue
Infrared
rendition
green
blue
Predominant
wave length
in densitometer
646
545
460
Examples of CIR records:
Healthy, vigorous vegetation
Mature or diseased vegetation
Litter
Bare ground
Turbid water
Clear water
Red
Greenish
Green to yellow
White or green
Light blue
Black
values among dates is not appropriate. When comparing within-date reflec-
tances, trends observed in June were even more pronounced in August. Both
a loss in color and a slight rebound at the C plot were again measurable
across the fumigation rates. ZAPS II densities differed somewhat from the
earlier date, particularly at the low (B) S02 level, where substantial color
was lost relative to the control. Again, within-treatment variation was
greater than at the June observation. This probably was caused by the more
pronounced difference in maturity between warm and cool-season plant species
at the latter time. In the field, the D (high) plot on ZAPS II looked even
more chlorotic than the data suggest because of the highly bleached "hot
spots" directly under the apertures in the gas delivery pipes.
October
Densitometric curves at the end of the growing season are very similar
to those obtained from August data. Sample variability was dampened some-
what due to the more uniform maturity stages of plant groups. Again, the
color content, especially red, was significantly reduced by S02-
Regression Analysis
Simple linear regression equations and coefficients were calculated for
each color and for total density for all dates and treatments (Table 22.3).
The ratios between red and green densities also were computed. The SOa rates
used in these calculations were cumulative means from the beginning of the
300
-------
growing season until the time of photography. Values are presented in
Table 22.1.
In June, a highly significant linear Lepression of red density was ob-
served across treatments; no other trends were significant. On the other
hand, ZAPS II data were significantly or highly significantly linear for all
colors and for total density.
In August, red density changes were not significant on ZAPS I, but blue,
green, and total were. No significance was obtained on ZAPS II, clearly
because of the high variation associated with samples.
By October, the only significant linear relationship between color and
fumigation on ZAPS I was in the red wave lengths. ZAPS II showed a signifi-
cant red/green ratio. This was due to the relatively faster loss of green
than red color with fumigation. The biological significance of this obser-
vation is unclear. In general, the insensitivity of the red/green ratio to
vegetational changes was disappointing. Research by Schott and others
(Schott et a1>, ND) has suggested that this ratio is a more sensitive indi-
cator of vegetational change than any single wave length range. Our data,
admittedly derived somewhat differently, did not support this observation.
The densitometric analyses show that losses of color on infrared film
occur due to SC>2 fumigation. Interplot variability causes some insensitivity
of statistical tests, but at least high and control rates are significantly
different in most cases.
The detection of such differences does nothing to explain them. We are
therefore conducting field studies in 1979 to account for these reflectance
differences on cellular and physiological levels. Chlorophyll contents of
western wheatgrass leaves and microscopic sections for pathological signs
will be collected and analyzed coincidentally with aerial photo missions.
Ground-Level Photography
The square-meter photo plots were mapped by individual species. These
maps are filed for future comparisons, both between times of observation
and scales of photography (i.e., aerial vs. ground imagery).
Densitometric readings from photo plots on Colorado State University's
fertilizer study were provided to CSU for record and further analysis.
Aspect photographs recorded the visual differences in plant species
reflectance across fumigation plots. Abrupt density changes at treatment
boundaries also were conspicuous. The principal value of these photographs
is as visual references. They are not suitable for either densitometric
analyses or mapping.
301
-------
TABLE 22.3.
LINEAR REGRESSION EQUATIONS ESTIMATING PHOTOMETRIC DENSITY (Y) AS
A FUNCTION OF SULFUR DIOXIDE FUMIGATION (X)
o
to
DATE
JUNE
197S
AUGUST
OCTOBER
COLOR
REJ)
GREEN
BLUE
TOTAL
R/G
RED
GREEN
BLUE
TOTAL
R/G
RED
GREEN
BLUE
TOTAL
R/G
ZAPS I
EQUATION
Y =
Y =
•y
V =
Y =
Y =
Y =
Y =
Y =
Y =
Y =
Y =
Y =
Y =
Y =
.5100-.
.5865-.
.9486-.
.5111-.
.8636-.
1.2538-
.7853-.
1.7056-
.8709-.
1.5924+
.8505-.
.6190-.
1.0682-
.6318-.
0097X
0204X
027X
0165X
019X
.0341X
0217X
.0433X
0273X
.0019X
0277X
0197X
.0332X
0216X
1. 3711-8. 5X10~5
r2 DATE
.99** JUNE
.80 1978
.86
.85
.49
. 82 AUGUST
.93*
.91*
.92*
.009
.95* OCTOBER
.78
.73
.87
2.3X10~5
ZAPS II
COLOR
RED
GREEN
BLUE
TOTAL
R/G
RED
GREEN
BLUE
TOTAL
R/G
RED
GREEN
BLUE
TOTAL
R/G
EQUATION
Y =
Y =
Y =
Y =
Y =
Y =
Y =
Y =
Y =
Y =
Y =
Y =
Y =
Y =
Y =
.4529-.
.8051-.
1.2283-
.6061-.
.5584-.
1.0522-
.7754-.
1.6810-
.8023-.
1.3651-
.6788-.
.6043-.
1.0025-
.5650-.
1.1158-
0168X
034X
.056X
0254X
0054X
.0309X
0161X
.0431X
0195X
.0144
0222X
0294X
.0438X
0253X
.0297
2
.93*
.95*
.98**
.99**
.16
.79
.45
.71
.61
.63
.76
.79
.72
.78
.96*
* P < .
** P < .
05
01
-------
CONCLUSIONS
The large scale (1:3,700) aerial photographs were subjected to densito-
metric analysis. A general reduction in density with increased S02 was
observed with red, green, and blue wave lengths. Significant differences
were observed between control and high S02 levels at most dates and for most
colors. Total density changes were less consistent, as were red/green den-
sity ratios.
Absolute density values differed among seasons and between locations due
to several confounded variables. These included inherent inter- and intra-
plot variation, differences in species composition and cover, and differing
phenologic rates. They were further influenced by differences in photo-
graphic conditions, exposure, and processing.
Within dates and treatments, photographic density differences, both ob-
served and measured, were clearly related to S02 fumigation.
Linear regression analyses quantified the direct relationship between
S02 and color loss in vegetation. The red density, representing near-infra-
red reflecting species and objects, was depressed by S02, especially during
the active growth season.
REFERENCES
Schott, J. R., D. W. Gaucher, and J. E. Walker, ND. Aerial Photographic
Technique for Measuring Vegetation Stress from Sulfur Dioxide. Calspan
Corp. Rep. YB-5967-M-1.
Taylor, J. E., W. C. Leininger, and R. J. Fuchs. 1976. Monitoring Plant
Community Changes due to Emission from Fossil Fuel Power Plants in
Eastern Montana. In: The Bioenvironmental Impact of a Coal-Fired
Power Plant, Second Interim Report, Colstrip, Montana, June, 1975.
R. A. Lewis, N. R. Glass, and A. S. LeFohn, eds. EPA/600-3-76-013,
U. S. Environmental Protection Agency, Corvallis, Oregon, pp 14-40.
Woodcock, W.E. 1976. Aerial Reconnaissance and Photogrammetry with Small
Cameras. Photogrammetrie Eng. and Remote Sensing. 42:503^511.
303
-------
PRELIMINARY SYNTHESIS
SECTION 23
SIMULATION OF A GRASSLAND SULFUR CYCLE
M. B. Coughenour, W. J. Parton, W. K. Lauenroth,
J. L. Dodd, and R. G. Woodmansee
ABSTRACT
A simulation model was developed as a conceptual
tool to organize information on sulfur cycling in a
grassland ecosystem. The model successfully simulated
the concentrations of sulfur in shoots of two plant
functional groups over the course of two growing seasons
and seemed to realistically represent such processes as
microbial mineralization-immobilization, sulfate adsorp-
tion by soil, and nematode consumption. The greatest
transfers of sulfur in the system were between soil sul-
fate and microorganisms. Various turnover rates were
calculated, ranging from 3 x 10~5 yr"1, for the most
resistant components of humus, to 10 yr"1, for the highly
labile metabolic litter component. Only 2-4% of the
total system flux of sulfur was reapportioned among
system components, indicating that the model system was
near equilibrium.
INTRODUCTION
Man's activities seemingly play an important role in the availability of
sulfur in both natural and agricultural ecosystems. Fossil-fuel combustion
adds millions of tons of sulfur to the biosphere annually (Kellog et at., 1972),
yet sulfur deficiencies in agro-ecosystems are becoming increasingly
widespread because of the use of low-sulfur NPK fertilizers in high-yield
agricultural practices (Bell, 1975) and the increasing emphasis on air-
pollution control. We will present here a method to assess the effects of
anthropogenic activities on nutrient cycling, using simulation modeling to
integrate information, concepts, and assumptions about sulfur transformations
in ecosystems.
The cycling of sulfur in grazed pastures was first elucidated by
Walker (1957), who showed that the dynamics of the sulfur cycle are parti-
cularly important in pastures where nitrogen is continually added through
fixation and in ungrazed harvested pastures where sulfur is not returned to
304
-------
soil in urine. Because responses in animal productivity to sulfur ferti-
lizers depend on the integrated processes of plant uptake, animal utiliza-
tion, animal return, and microbial release, subsequent studies were ini-
tiated, which led to the formulation of a kinetic mathematical model of the
sulfur cycle. These studies revealed that the productive potential of the
system depends on the rates of sulfur transfer through, rather than the
quantities present in, any given system component (May et at., 1968; Till
and May, 1970, 1971; May et al., 1972).
Our model provides a similar dynamic framework for understanding the
nature of sulfur cycling in a grassland ecosystem dominated by western
wheatgrass (Agropyron smith-ii,) in southeastern Montana (Heitschmidt et at.,
1978) . We feel the model is conceptually sound because the information it
contains was either drawn from the literature or based on experience when
published information was lacking or inappropriate. The model in its present
form will be more useful as a conceptual than as a predictive tool. The model
was tested by comparing its behavior with limited field data from the
actual ecosystem.
Model Overview
Notation
f. jth flow of sulfur (g S • m~2 • day"1) associated with an F.
flow of carbon (biomass)
S sulfur state variable (g S • m~2)
C carbon state variable (g C • m~2)
CS C/S ratio
SC S/C ratio
a. rate coefficient used to calculate sulfur flow f.
E(x) flow rate reduction factor (effect) as a function of variable x
V maximum rate
The sulfur-cycle submodel (Figure 23.1) should not be viewed as a set
of independent calculations, because it relies extensively on information
supplied by other simultaneously executed submodels describing primary
producer-carbon cycling (Detling et al., 1979), decomposer-carbon and nitrogen
cycling (McGill et al., 1978), and abiotic processes (Parton, 1978). Conse-
quently, sulfur state variables have names similar or identical to those of
the other submodels. The assumption here is that there is a functional corre-
spondence between elements (C,N,S) in all organism-mediated flows of biomass
or substrates.
Although the producer-carbon submodel individually simulates roots and
rhizomes (or crowns), these have been grouped into the single category
305
-------
Saprophagic
Nematodes
35-38
Figure 23.1.
Flow diagram showing structure of sulfur cycle model. Large
circles with numbers are duplicate representations of metabolic
and structural litter. Small circles represent the partitioning
of a flow. Flows to the atmospheric sink (violatilization) are
assumed to return immediately to metabolic litter.
"roots" for the sulfur submodel. Likewise, young leaves, mature leaves,
and stem components of the producer model have been grouped as "shoots" for
the sulfur submodel. Roots are simulated in three soil strata: 0-5, 5-20, and
20-60 cm; and producers are simulated in two functional groups: Agropyron
smithii and "other" species.
A considerable degree of correspondence exists between the decomposer-
nitrogen and sulfur submodels. All three elements exactly correspond for
bacteria, fungi, litter, humads, and resistant organic materials. The decom-
poser model is especially useful for calculating sulfur flows because the microbes
themselves are the decomposers. The relationship between microbial demands
and substrate C/S ratios controls mineralization-immobilization. Further-
more, in the model the substrate can be "altered" in quality and nutrient
content during decomposition by simulating humads (discussed below) and the
306
-------
metabolic and structural components of litter, roughly analagous to the
labile and resistant components of Hunt (1977). Thus carbon, nitrogen, and
sulfur mineralizations are not strictly parallel processes. Sulfur miner-
alized early can be recycled to microbes for later use (Stotzky and Norman,
1961), such as during decomposition of more resistant materials. If there
is an imbalance between N and S relative to microbial needs, the deficient
element is immobilized.
The metabolic fraction of litter is soluble, high in nitrogen, and
rapidly decomposable; while the structural fraction is insoluble, low in
nitrogen and slowly decomposable. Only decomposition of structural mater-
ials is affected by microbial C/N and C/S ratios. The term humads was
coined by McGill et-al* (1978) by blending the words humus and adsorb.
Humads are that component of soil organic material having a relatively short
turnover time and are formed in a little-understood fusion-adsorption
relationship between soluble and insoluble litter fractions and clay miner-
als. Resistant soil organic matter is humus having an extremely slow
turnover rate. Bacteria and fungi were treated separately because they have
different responses to abiotic variables and abilities to decompose soluble
and insoluble substrates. All belowground components except roots were
simulated at four depth strata: 0-1, 1-5, 5-20, and 20-60 cm. The top
layer is a receptacle for litter from aboveground plant materials. Some
physical mixing with soil is assumed.
The model was programmed in the SIMCOMP simulation language (Gustafson
and Innis, 1973) version 4.0 (Gustafson, 1977). Daily maximum and minimum
air temperatures, precipitation, humidity, and cloud cover are driving
variables for the abiotic submodel, which in turn calculates temperatures
and moisture levels at any point in the soil profile. Simulation proceeds
over a daily time step. Model dynamics were tested against field data from
the Taylor Creek Site in southeastern Montana (Dodd et dl., 1978) for the
period 20 April 1975 through 23 September 1976.
Soil Sulfur
Inorganic sulfur is either completely oxidized as sulfate or found in
a number of lower oxidation states, the most important of which is sulfide.
Sulfate levels may range from 1 to 45 ppm in soils (Evans and Rost, 1945).
With adequate aeration the sulfide form is usually present at only 1-3 ppm
(Freney, 1961). Sulfate is found in water-soluble, adsorbed, or insoluble
forms in soils. Free gypsum (CaSO^) in the lower parts of semi-arid soils
is common (Williams, 1975).
The model distinguishes sulfate in all three forms, but only one state
variable is shown for clarity (Figure 23.1). Although inorganic sulfur has
numerous oxidation states below that of sulfate, for our purposes we can
consider them all as "reduced" and assume that the predominant form is S~2.
Organic sulfur can be extracted in two distinguishable fractions from
soil: hydriodic acid reducible (HI) and carbon bonded (Freney, 1961;
Freney et al., 1971). Hydriodic acid-reducible sulfur are sulfate esters
307
-------
and usually constitutes about 50% of total organic sulfur. The origin of
these esters is uncertain, but they may be simply the products of the
metabolism of 3'-phosphoadenylyl sulfate (PAPS) by plants and microorganisms:
XOH + PAPS sulfokinase pAp + XOSQ- (1)
where XOH is any of a number of acceptor molecules and PAP is the desul-
fated phosphoadenylyl (Peck, 1969). Carbon-bonded sulfur probably origi-
nates from sulfur-containing amino acids. Because they are rapidly decom-
posed, free amino acids do not normally accumulate in soils; but the oxidized
forms, cysteic acid and methionine sulfone, accounted for 21% of the total
sulfur and 45% of the carbon-bonded sulfur in a virgin grassland soil
(Freney et al., 1972).
Ideally, the sulfur submodel would include separate state variables
for carbon-bonded and esterified sulfur and the appropriate transformations
for each. However, lack of information prohibits including such complexity
in the model, so HI sulfur can be calculated best with Eq. 10. The model
assumes that both organic sulfur forms are present in humads and resistant
soil organic material.
Microbial Processes
Sulfur transformations in soils are mainly microbiological (Freney,
1967) and are of four types: biosynthetic sulfate reduction, respiratory
sulfate reduction, respiratory sulfur oxidation, and photosynthetic sulfur
oxidation (Peck, 1975). Respiratory sulfate reduction is unimportant in
aerobic environments (Starkey, 1950), as it is inhibited by oxygen (Peck,
1975). Photosynthetic oxidation can be disregarded except in consistently
anaerobic conditions. Obligate autotrophic respiratory oxidizers are
important only in soils with added elemental sulfur, since sulfur can be
oxidized in most soils by common heterotrophic microorganisms (Starkey,
1950).
The normal range of sulfur concentrations in microorganisms is 0.1% to
1.0% of the dry weight (Alexander, 1971). Additional information about
microbial chemical composition was used in a detailed calculation of the
C/S ratios of structural and metabolic components of bacteria and fungi
(Coughenour, 1978). The structural component includes cell walls and
associated membrane lipoproteins, while the metabolic component includes
all soluble intra-cellular material. If structural composition is fixed,
metabolic sulfur concentration is free to vary according to the variable
cell sulfur status. The calculated C/S ratios for structural components
were 335 for fungi and 210 for bacteria. Bacterial metabolic C/S ratios
range from 35 to 75, yielding total cell C/S ratios from 50 to 100. Fungal
metabolic C/S ranges from 120 to 250, yielding total cell C/S ratios from
150 to 280.
When microbial death is calculated in the decomposer submodel, sulfur
flow is controlled by carbon flow:
308
-------
f =
cs
m
where fj and F} are the flows (g • m~2 • day"1) of structural sulfur and
carbon due to death, respectively, and CST is the fixed C/S ratio of the
microbial structural component. The flow from metabolic components varies
with the C/S ratio of the microbes:
f2 =
m
where f2 and F2 are tne flows of metabolic sulfur and carbon due to death,
and CS is the C/S ratio of the microbes.
m
Sulfate in solution is available for microbial uptake, but little is
known about microbial uptake kinetics. For a number of aerobic aquatic
bacteria, uptake increases with sulfate concentration up to a maximum rate.
Increasing concentrations of glucose in the media may increase uptake,
while increasing cysteine or methionine in the media decreases sulfate
uptake (Monheimer, 1975). Binding of labeled sulfate by Thiobac'illus
ferroxidans follows typical saturation kinetics, but cysteine and methionine
do not affect binding (Tuovinen et at., 1975), as would be expected in view
of the role of sulfate in autotrophic metabolism. Thus, we assumed that
sulfate is taken up from solution according to Michaelis-Menten enzyme
kinetics: uptake is curtailed if sulfur demands are met by organic sub-
strates, and uptake is enhanced by greater rates of carbon utilization.
Sulfate uptake rate is therefore modified by the effects of soil water
and temperature on general microbial metabolism and by the demand for
sulfur as expressed in the microbial C/S ratio:
a. = E (T) • E (W) • E (CS ) (4)
, 3 m m u m '
and
V • [SOiJ
.. _ max H
f = a ' C
3 3 m K + [S0/J
m 4
where 33 is the combined effect of temperature IE (T) , Appendix Figure 23.1-a],
water [E (W) , Appendix Figure 23.1-b], and C/S ratio [E (CS ), Appendix Figure
23. 1-c] ; V is the maximum uptake rate; ISO^] is the concentration of sulfate
in soil solution (ppm S); C is the microbial biomass (g C • m~2) ; and K is
m m
the half-saturation constant. First approximations of V and K values
max m
were calculated from published data (Monheimer, 1975; Tuovinen et al . , 1975).
309
-------
However, uptake rates were too great and microbial sulfur levels rose to
excessively high levels and fluctuated erratically. This instability may
have been because the simulation interval (1 day) was large relative to the
much faster turnover rate of microbial sulfur implied by the literature
rate constants. To maintain reasonable C/S ratios, we gave K a value of
m
250 ppm for bacteria and fungi and V values of 0.056 and 0.01 g S • day"1
max J
• g"1 microbial C for bacteria and fungi, respectively.
During decomposition, some substrate sulfur is used to construct
microbial biomass, while the excess is converted to inorganic forms (Freney,
1967). Thus, in the model the extent of mineralization-immobilization is
affected by the relationship between the sulfur content of the substrate
and the sulfur requirements of the microbes. Attempts to predict mineral-
ization from substrate sulfur concentrations only have been variously
successful (Barrow, 1960a, 1960b; Kowalenko and Lowe, 1975b; Freney and
Spencer, 1960; Williams, 1967). Also, substrate C/S ratios may mean little
unless the nature of the substrate molecule is considered (Barrow, 1961;
Frederick et al., 1957; Stewart et al., 1966b).
During decomposition, a correlation between C, N, and S mineralizations
is sometimes observed, but the strength of this correlation may vary during
decomposition (Nelson, 1964; Kowalenko and Lowe, 1975b). Sulfur release is
often better correlated with nitrogen mineralization than with C02 evolution
(Kowalenko and Lowe, 1975b, Stewart et al., 1966a). However, the mineralized
N/S ratio varies, depending on which element is limiting (Stewart et al.,
1966a), and may not correspond to the substrate N/S ratio, simply because N
and S do not occur in the same compounds or substrate fractions (Freney and
Stevenson, 1966). To avoid these difficulties, we considered each element
separately in the model, microbial activity explicitly, and four substrate
types.
It is difficult to ascertain whether sulfate or sulfide is the form
directly mineralized from decomposing organic materials. When aeration was
varied by changing soil moisture, no detectable levels of sulfide were
found until 80% field capacity was reached (Chaudry and Cornfield, 1967).
Very little sulfide was mineralized from cysteine, suggesting either that
oxidation of sulfide to sulfate is very.rapid or that sulfide is indeed not
an intermediate in sulfate production (Freney, 1960). Decomposition of
cysteine or cystine by single-species cultures produces either sulfate or
reduced forms, depending on the species. Because of species interactions,
mixed populations generally produce only sulfate under aerobic conditions
(Freney, 1967). Either sulfate or methyl mercaptan and dimethyl disulfide
are formed when methionine is decomposed aerobically by mixed populations
(Frederick et at., 1957; Freney, 1967). The last two compounds are volatile
and may escape from the soil. Soils treated with a variety of sulfur-
containing organics evolve methyl mercaptan, dimethyl sulfide, dimethyl
disulfide, carbonyl sulfide, and carbon disulfide. Under aerobic conditions,
1.9% and 3.0% of total added sulfur are volatilized from dried and fresh
plant materials, while under anaerobic conditions the percentages rise to
2.5% and 3.3% (Banwart and Bremner, 1976).
310
-------
In the model, mineralization and sulfate uptake are affected by the
same temperature and moisture functions. The model also includes an
effect of microbial C/S that mirrors the C/S effect on sulfate uptake:
34 = E (T) • E (W) • E (CS ) (6)
m m mn m
fit = a4 • S • V (7)
H H m mn
where E (T) and E (W) are as in Eq. 4, E (CS ) is the effect of C/S
m m mn m
(Appendix Figure 23.1-d), a^ is the combination of these effects, f^ is the
mineralization rate (g S • m"2 • day"1), V is the maximum rate (day"1), and
mn
S is microbial sulfur (g S • m"2) . Maximum rates chosen were 0.8 day"1 for
bacteria and 0.165 day"1 for fungi. The product of mineralization is
sulfate or sulfide, depending on aeration. Sulfate is the sole product
until 80% field capacity is reached, then the sulfate percentage decreases
linearly to 0% at 133% field capacity (Appendix Figure 23.1-e).
The oxidation of sulfide to sulfate is controlled by soil aeration and
bacterial metabolism. Bacterial metabolism is expressed as a function of
temperature and water. The oxidation relationship is:
a5 = E (T) • E (W) • E (W) (8)
m m ox
f5 = a5 • VQx • SR (9)
where E (T) and E (W) are the effects of temperature and water in bacteria;
E (W) is the effect of water on aeration (Appendix Figure 23.1-e); 35 is the
OX
combined effect; f5 the oxidation rate (g S • m2 • day"1); S is the quantity
R
of reduced forms (g S • m"2); and V is the maximum rate (day"1), based on
OX
results from a silt loam soil in which 55% of the reduced sulfur was oxidized
in 3 weeks (Attoe and Olsen, 1966). This expression thus assumes that
oxidation is proportional to S .
K
Volatilization of sulfur during decomposition is represented as a
proportion of metabolic litter decomposition, increasing slightly as aeration
declines:
f6 = F6 • PV(W) • scme
where fg is volatilization (g S • m"2 • day -1); Fg is metabolic litter
decomposition (g C • m"2 • day"1); P (W) is the variable proportion as a
function of soil water (Appendix Figure 23.1-h), based on data of Banwart and
Bremner (1976); and SC is the S/C of metabolic litter. We assume that this
me
equation applies to all soil layers.
311
-------
Because of the uncertainty of their origin, sulfate esters are not
explicitly treated but are assumed to be associated with humads and resis-
tant soil organic matter. A regression equation based on data of Cooper
(1972) and Bettany et al. (1973, 1974) predicts that HI sulfur (SUT, g S •
HI
m~2) is a function of total sulfur (S , g S • m~2) and total carbon
1U JlAJ_j
(CTOTAL' 8 C * m"2):
SHI=0'71 ' STOTAL-°-°029' CTOTAL (10)
This equation can be used to calculate HI sulfur at any point in the
simulation.
Sulfate may be released through the action of sulfatase enzymes on
soil sulfate esters:
+ H20 ->• ROH + H+ + SO" (11)
Sulfatase has been detected in plants, animals, and microorganisms.
Arylsulfatase activity in soils can be measured by the method of Tabatabai
and Bremner (1970), while the Houghton and Rose (1976) method can be
applied to a variety of sulfatases. Arylsulfatase activity correlates well
with total carbon and total sulfur and strongly with HI sulfur (Cooper,
1972). Activity is related to temperature (Tabatabai and Bremner, 1970)
and soil water (Cooper, 1972) . Sulfate may not be liberated directly from
humic acids; rather, a sequential attack by depolymerizing and desulfating
enzymes may be required (Houghton and Rose, 1976) .
A regression equation, formulated with data of Tabatabai and Bremner
(1970), predicts sulfatase activity as a function of total sulfur:
A = 2.026 • (ppm total soil sulfur) (12)
where the zero-intercept was forced as a simplified assumption.
The units of A are yg p-nitrophenol • g"1 soil • hr"1 , since p-nitrophenol
was the aryl product assayed. Knowing that one mole of sulfate is released
for each mole of p-nitrophenol, we can make the conversion:
a7 = 0.0011 - S (13)
where ay is activity (g S • m~2 • day"1), and S . is total soil sulfur
_LU -LA.J_i
(g S • m~2) . This basic rate is then modified to reflect the hypothesis
(Houghton and Rose, 1976) that depolymerizing enzymes must act on humic
materials before sulfatase can take effect. Thus, the effects of water and
temperature on microbial activity averaged for bacteria and fungi was
included, assuming depolymerizing action is proportional to microbial
activity. Furthermore, an effect of temperature on extracellular enzyme
312
-------
activity is incorporated, based on data of Tabatabai and Bremner (1970) .
The effect equals 1 (no effect) at 37°C, the standard temperature of their
experimental procedure. The final expression is:
R = a7 • E (1) • E (T) • E (W) (14)
e m m
where a-j is defined in equation 13, R is sulfur released (g S • m~2 •
day"1), E (T) is the effect of temperature on extracellular enzymes (Appendix
Figure 23.1-f), E (T) and E (W) are the average effects of temperature and
mm
water for the two microbial groups.
Because roots and microbes are the presumed source of the sulfatase
enzyme, activity in deeper layers is scaled to that calculated for the
surface layer according to differences in root and microbial biomass
densities. If B = sum of root and microbial biomass per cm of soil in the
surface layer, and B = the same measure in a deeper layer, then
B = - (15)
is the proportional decrease in activity for layer 1. Because sulfate
esters are assumed to be associated with both humads and resistant soil
organic matter, the total flow is divided between them in proportion to
their relative sulfur concentrations:
SCR (16)
• S
f7 = R • 0 -a (18)
f8 = R • |3 • (1 - a) (19)
where A is the lability of humad-associated esters compared to that of
resistant organic-associated esters, SC and SO are the S/C ratios of
H R
humads and resistant organic material, a is the proportion of sulfate
originating from humads, S is humad sulfur (g S • m~2) , S is resistant
H. K
organic sulfur (g S • m~2) , f 7 is the sulfate release from humads (g S •
m~2 • day"1), and f.% is the release from resistant organic material.
Decomposition rates of metabolic and structural litter, humads, and
resistant organic material are calculated in the decomposer submodel,
313
-------
sulfur merely following carbon according to the C/S ratio of the substrate.
As postulated, only the decomposition of structural litter components is
influenced by sulfur. Because humads and metabolic components are considered
sufficiently high in sulfur, their decomposers are not sulfur limited. When
bacterial C/S is more than 68 and fungal C/S is more than 202, structural
litter decomposition becomes limited; complete limitation is reached before
the microbial sulfur content reaches minimum (Appendix Figure 23.1-g).
Plant Processes
Sulfate uptake by roots seems to follow classical Michaelis-Menten kine-
tics (Leggett and Epstein, 1956; Persson, 1969; Nissen, 1971). First, the
sulfate undergoes a labile binding to positive sites in the apparent free
space. Maintenance of these positive sites is likely related to root meta-
bolism (Persson, 1969), but the quantity of ions reaching the sites seems
to be affected also by a process mediated in the shoots, since decapitated
plants bind less than nondecapitated plants (Ingelsten, 1966) . The trans-
fer of ions from the apparent free space to the xylem appears to be linked
to the supply of energy-rich substances from the shoots rather than to a
root-retention mechanism (Pettersson, 1966) . Over a narrow range of media
concentrations, uptake obeys a single set of Michaelis-Menten parameters
(Leggett and Epstein, 1956). However, over a much broader range, uptake is
mediated by a single multiphasic isotherm (Nissen, 1971) and passes through
a series of phases, each separable from the next by a sharp concentration
level. Michaelis-Menten kinetics apply within each phase, but V and K
m m
increase between phases as concentration rises. This mechanism is different
from the dual high- and low-affinity mechanism postulated from nitrogen
(Reuss and Innis, 1977) and phosphorous (Cole et al . , 1977).
Root sulfate uptake is calculated:
V [SOiJ
max •
K
m
fg = CR - ER(T) • ER(W) • Eu(CSp) - U (22)
where U is the Michaelis-Menten basic uptake rate (g S • m~2 • day1 • g"1
root carbon), V is the maximum rate, K the half-saturation constant,
max m
[SOiJ is the solution sulfate concentration (ppm S), fg is uptake (g S •
m~2 • day"1), C is root carbon (g C • m~2), E (W) and E (T) are the effects
of water and temperature on root metabolism (Appendix Figures 23.1-i and 23.1-j),
and E (CS ) is the effect of whole-plant S/C on uptake (Appendix Figure 23.1-k).
Table 23.1 shows the V and K values used. Although K and switchpoint
max m m
concentrations are those of Nissen (1971), his V values were reduced by
nicix
0.004 to match observed shoot-sulfur concentrations.
314
-------
TABLE 23.1. V AND K PARAMETER VALUES USED IN THE MULTI-PHASE
m
MICHAELIS-MENTEN EXPRESSION FOR ROOT UPTAKE*
SO, concentration (ppm S)
<
1.
3
12
32
80
201
>
1.088
088/3.2
.2/12.8
.8/32.0
.0/80.0
.0/201.6
.6/768.0
768.0
V (g S
max
g root
3.
5.
7.
10.
17.
29.
77.
120.
-2 , -1
• m • day
carbon )
04 x
00 x
48 x
40 x
80 x
96 x
80 x
18 x
io~6
ID"6
io-6
io-6
io-6
10~6
io-6
10~5
• K (ppm S)
0
1
3
9
41
124
640
20480
.54
.60
.84
.92
.60
.80
.00
.00
&
V values are 0.4% those of Nissen's (1971), while K values are
max m
unmodified.
Movement from roots to shoots is also a metabolically mediated
process (Pettersson, 1966; Ingelsten, 1966). Sulfur is freely mobile in
the plant but is captured quickly in plant parts undergoing active cell
division (Biddulph et at., 1958). From the site of active uptake,
sulfate moves rapidly to the xylem and thence to shoots. Inhibition by
selenate on translocation therefore implies that inhibition is of whole-
plant sulfate absorption rather than of root and shoot uptake separately,
which corresponds to the inhibition of root uptake in decapitated plants
(Ingelsten, 1966). Likewise, root uptake of nitrogen is not related to
root-nitrogen content (Edwards and Barber, 1976). We therefore propose
that sulfate uptake is related to whole-plant sulfur demand.
Sulfur is translocated to shoots in the model according to a predefined
ratio of shoot S/C to root S/C, which is assumed in turn to be a function
of plant phenology. This formulation has worked well in a model of
nitrogen translocation (Reuss and Innis, 1977) and reflects observations
of higher shoot-sulfur than root-sulfur contents in early phenophases.
315
-------
Thus,
DSC = SCRT - I(P
RT
DSC = SCRT - I(Phen) (23)
S
DSC • <24>
where DSC is the determined shoot S/C, SCRT is actual root S/C, I(Phen) is
the predefined ratio of shoot S/C to root S/C as a function of phenology
(Appendix Figure 23.1-1), f^g is translocation from one root layer (g S
• m~2 • day"1), F^g is tne daily shoot carbon growth increment (g C • m~2
_ RT1
• day"1), SCJ is the root sulfur in a layer and S n is root sulfur summed
over all layers (g S • m~2) .
During death of plant parts, sulfur may be remobilized from the dying
organs and thus conserved within the plant. We assume, however, that shoot
death induced by freezing or drought will not conserve sulfur and postulate
here that dying roots do not conserve sulfur. Sulfur follows carbon during
non-sulfur-conserving death, with a metabolic/structural split determined
by:
SM - s - ^T (25)
CSM = (1 - ' C (26)
where SM is root or shoot metabolic sulfur (g S • m~2) , S is root or shoot
total sulfur, C is root or shoot carbon (g C • m~2) , a is the fraction of
structural carbon in dying tissue, CST is the fixed C/S ratio of structural
components, and CSM is the variable C/S of dying metabolic components. As
calculated from available information on plant chemical composition, CST was
63 for shoots and 1030 for roots (Coughenour, 1978) .
During shoot senescence, sulfur is remobilized in the plant by protein
hydrolysis in senescing cells (Williams, 1955) . This remobilized sulfur is
probably transferred to belowground organs, where it may be utilized the
following season during shoot regrowth, as is nitrogen (Clark, 1977).
However, net export of sulfur from Sudan grass leaves is delayed and de-
pressed more than that of nitrogen (Barrien and Wood, 1939, cited in Williams,
1955). Evidence suggests that sulfur is conserved in these grasses, as
shoot sulfur in A. smithii- declined from 0.12%, in live shoots, on 1 August
1975 to 0.08%, in recent dead, on 1 September to 0.06%, in old dead, the
following year.
For sulfur -conserving shoot death,
316
-------
CSg = CCNS +6) • NSg (27)
(28)
:ss CSTS
(1 - a) • Fn
(29)
a fn
where CS is the C/S of senescing shoots after conservation, CN is the C/N
o o
ratio of shoots, 6 is the increase in C/N in tissues where protein hydroly-
sis has occurred, a is the fraction of structural shoot carbon, NS is the
O
N/S of shoots, F}i is total shoot-carbon death (g C • m~2 • day"1), fji is
total metabolic sulfur in dying shoots (g S • m~2 • day"1), CST is the
fixed shoot structural C/S, and CSM is the C/S of dying metabolic compo-
O
nents of dying shoots. Thus, structural sulfur in dying shoots is:
fl2 = 7^" fl1 (30)
S
and total remobilization of sulfur is:
i i
) (31)
which is distributed among roots in all layers; CSMD is the C/S of the
metabolic component before protein hydrolysis.
The effect of sulfur on shoot growth and photosynthesis is critical to
any sulfur cycling submodel; however, scant information on the process which
is distributed among roots in all layers; CSMD is the C/S of the metabolic
component before protein hydrolysis.
The effect of sulfur on shoot growth and photosynthesis is critical to
any sulfur-cycling submodel; however, scant information on the process
exists, especially for these specific grasses. Even though concentrations
in A. smithii shoots are typically about 0.1% S, well below the critical
range of 0.3-0.35% supposedly required for maximum yield (Metson, 1973),
growth does not seem impaired (Dodd et at. 1978). Another index of sulfur
status is the N/S ratio. If a constant protein N/S ratio is assumed, non-
protein nitrogen should accumulate, yielding a high N/S ratio when sulfur is
deficient. The critical N/S ratio for grass is thought to be 15-20 (Metson,
1973). Typically, however, A. sm-ithii N/S is 20-25 in early, and 10-15 in
late, phenophases. This phenological change makes interpretation of N/S
ratios difficult. When nitrogen is limiting, as is likely here, N/S ratios
may mean little. Until more satisfactory information is provided, we
propose that shoot metabolism is affected only if nitrogen is non-limiting
317
-------
(above 2.5% N for photosynthesis and 2.0% N for growth). When N is non-
limiting, photosynthesis and growth are not sulfur limited at N/S < 25,
but will decline linearly to complete restriction at N/S = 35.
Nematodes
A model of plant parasitic nematode population dynamics (Ferris, 1976)
was adapted as a driving model for carbon, nitrogen, and sulfur flows.
Although the model was intended for Meloi-dogyne nematodes, which host on
grapevine, functions of soil water and temperature seemed comparable to
observations for nematode groups found in South Dakota (Smolik, 1975) .
Therefore, general principles of population growth should be roughly appli-
cable here. Population estimates and biomass conversion values determined
by J. Smolik for the Taylor Creek Site (Coughenour, 1978) were used to
calculate initial conditions and to roughly verify model performance.
The primary departure from Ferris' (1976) model is a consequence of
dormancy, which allows nematodes in Montana to withstand conditions of
severe stress not found in the grapevine system. To represent this process,
we allowed nematodes to pass between active and dormant stages at specified
critical soil moisture (8% field capacity) and temperature (2.5°C) levels.
To calculate consumption as a function of active population size, we
used Santmeyer's (1956) data to determine metabolic rate as a function of
temperature. The first approximation to intake is:
I = M (T) • 4.8 -!=%L_ * 4500 cal
c rvi/ "'" ml 02 ~"" g food
2.0 activity requirement * 0.5 assimilation efficiency •
0.25 £_|H£ . 24 ^- ^ 35 & ^ = M (T) • 0.02925 (32)
g fwt day g dwt r
(Smolik, 1974), where I is intake (g C • g"1 nematode C • day"1), fwt and
dwt are fresh and dry weights, and M (T) is the temperature-dependent
metabolic rate (ml 0_ • g"1 fwt • hr"1), shown in Appendix Figure 23.1-m.
Respiration is found:
R = AE -I (33)
c
where R is respiration (g C • g"1 nematode C • day~^), AE is assimilation
efficiency (0.5), and I is intake. The population model calculates growth
in terms of numbers but not of .biomass, hence the following algorithm. If N
is nematode biomass before growth and death increments, if N_ is nematode
i>
biomass after death and growth increments, and if N is nematode biomass
X
corrected for death (D) but not growth, then
318
-------
N = N - D (34a)
x
N = N - D + G (34b)
B
G = N - N (35)
B x
where G is the biomass growth increment (g C • m~2 • day ) caused by such
processes as hatching, moulting, and egg laying. Intake is reapproximated
to account for the growth increments :
(36)
c AE
and feces production follows as
. E = I • (1 - AE) (37)
(Petrusewicz and Macfadyen, 1970).
Intake of nitrogen and sulfur follows carbon according to the ratios in
the food. The N and S requirements for growth are:
RQ = G • NC , (38)
n nem
RQ = G • SC (39)
s nem
where RQ and RQ are the growth requirements (g N • m~2 • day"1, g S •
il S
m~2 • day"1), G is carbon growth (g C • m~2 • day"1), and NC and
SC are the N/C and S/C ratios of nematodes. Nematode composition was
nem
calculated as C/N/S = 46/14/1 (Coughenour, 1978). Thus, N and S in excess
of requirements are determined:
E =1 - RQ (40)
n n xn
E =1 - RQ (41)
s s s
where I and I are total N and S consumptions. Ammonium accounts for 27-
71% of total nitrogen excretion (Rogers, 1969) and will be assumed here to
be constant at 50%. Excreted carbon and sulfur are assumed to flow from
microbivores to metabolic and structural litter in a 1/3 ratio, while all C,
N, and S excreted by herbivores will flow to metabolic litter, since they
consume only the metabolic fraction of root cells.
319
-------
Abiotic Processes
Soil sulfate adsorption-desorption is formulated with the Freundlich
isotherm:
AS = b • [SOj 1/n (42)
where [SO^] is solution sulfate concentration (ppm S), AS is the .quantity
adsorbed (g S • g"1 soil), b and n are empirical constants. The Freundlich
isotherm is preferred to the Langmuir isotherm for sulfate because no ad-
sorption maxima is found (Harward et at., 1962; Chao et al., 1962) and in-
creasing solution sulfate concentrations seem to activate additional adsorp-
tion sites (Barrow, 1967).
To determine b and n, we divided a 60-cm soil core into four strata
(0.5, 5-10, 10-20, and 20-60 cm) for the following analysis. Five grams of
soil were shaken for 24 hours in a 10/1 solution to soil ratio. Solution
sulfate concentrations were 0, 5, 10, 25, 50, 100, 200, 300, and 400 ppm S.
The quantity adsorbed was the difference between solution concentrations
before addition to soil and those after the shaking process. Results were
fitted to the Freundlich equation by nonlinear least squares; the resulting
parameters are shown in Table 23.2.
Successive approximation is used to implement the isotherm in the model.
Initially, all sulfate is assumed to be in solution and the isotherm is used
to determine AS for that sulfate concentration. The AS is used to determine
a ratio of total adsorbed (g S • m~2) to total soil (g S • m~2) sulfate,
which is then used to calculate a new value for the proportion in solution.
This proportion yields a new approximation of solution concentration, and
the cycle is repeated. Five or six iterations are sufficient for convergence
on the equilibrium.
The solubility-product concept is applied to determine gypsum precipita-
tion:
K = [Ca+2] - [SCV2] • y2 (43)
sp
where K is the solubility product (2.4 x 10~5 at 25°C; Tanji, 1969),
sp
and Y is the mean activity coefficient. When K is reached, gypsum may
precipitate (Dutt, 1964). Although K is actually temperature dependent,
sp
this feature is not included in the model. The concentration of gypsum
can be estimated:
f\
[Ca+ ] '
320
-------
TABLE 23.2. FREUDLICH ISOTHERM PARAMETERS FOR TAYLOR
CREEK SOIL*
Depth (cm)
1_ Error mean
n Square of fit
0-5
5-10
10-20
20-60
4.88
4.50
5.74
1.74
0.361
0.281
0.365
0.634
187.8
106.7
354.0
51.4
* 1/n
Data were fit to AP = b C ' , where AP equals yg S
adsorbed per gram soil and C equals ppm S in initial
solution (yg S/ml H20).
where K is the dissociation constant (4.9 x 10.~3, Dutt 1964)
mean activity coefficient for the soil solution is estimated:
The
In y =
0.509 z v/u~
(45)
where z is the valence of the ion and u is ionic strength given by:
u = 1/2 I C. z?
1=1 X X
(46)
where C. is the concentration of one of the n ions in solution (Dutt, 1962).
-t- + +2 +2 2
The ions considered here include Na , K , Ca , Mg , and S04 . The solution
concentration of cations are calculated with subroutine CATEX (Reiniger et al.,
1971; Frissel and Reiniger, 1974), which uses Vaneslow-like exchange equa-
tions to describe competition of the four cations for sites on the soil ex-
change complex. Cation exchange capacity and cation concentrations were
obtained from a thorough chemical analysis of Taylor Creek soils (Coughenour,
1978).
The top soil layer is the receptacle of large quantities of above-ground
litter, yet is only 1 cm thick. Soluble decomposition products are therefore
assumed to move by mass flow to the next lower layer; for, as in a mixing
cell, the solution concentration of a layer is uniform. The flow from the
surface layer, then, is the product of total water drainage (supplied by the
321
-------
abiotic submodel) times the solution concentration of sulfate, metabolic
litter, or desorbed humads. Because of the large solution interval (1 day),
mass flow is restricted such that solution concentrations in the lower layer
do not exceed those in the surface layer.
Diffusion of sulfate is described by Pick's Law, where the diffusion
rate is proportional to the concentration gradient. Thus,
A[SOiJ
Df = D - Ax • 6 (47)
where Df is diffusion rate (g S • cm"2 • day"1), A[SOiJ is the concentration
difference between layers, AX is the distance (cm) between midpoints of
adjacent layers, D is diffusivity (cm"2 • day"1), and 0 is the volumetric
soil water content, necessary here to express diffusion rate per cross-
sectional area of soil.
Model Verification
Model performance was tested continually throughout its development
with a 2-year data set from the Taylor Creek Site (Dodd et at., 1978). The
same data set was used for the final verification presented here. Because
the sulfur submodel depends on the other submodels, their behaviors must be
verified first.
The abiotic submodel was verified with soil-water data from each of
four 15-cm layers to a 60-cm depth. Agreement with observed values was
favorable in all layers over the 1975 season. Typically, soil water.de-
clines through the summer until replenished by fall rains. The most serious
discrepancy was the underestimation of soil water at 0-15 cm in the spring
of 1976 (Appendix Figure 23.2-a). This and other minor discrepancies could
be related to variation in snowmelt and runoff or to field sampling problems
(Parton et at., 1978). Comparisons were more favorable by the end of the
1976 growing season in all layers, indicating a good approximation of the
yearly water budget.
Comparisons between observed and simulated shoot-biomass dynamics were
also satisfactory (Appendix Figures 23,2-c and ,2-d). The timing and the
magnitude of the peaks were well represented. The rapid growth phase from
early to mid-season proved most difficult to simulate, probably because of
the difficulty of distinguishing live from dead shoots during harvest mea-
surements. Discrepancies during the death phase were less serious. The sim-
ulated dynamics for rhizomes (Appendix Figure 23.2-e). showed fewer fluctua-
tions than the data. However, the May minimum-July maximum relationship was
supported by the data. Strict verification of intraseasonal root dynamics was
very difficult, if not impossible, again because of the lack of a reliable
method for distinguishing live from dead roots in the field. The comparison
here was between the sum of simulated live plus dead roots (belowground litter)
322
-------
and total root-biomass measurements from 0-10 cm soil cores (Appendix Figure
23.2-f). In view of this problem and the problem of large spatial sample
variability, these comparisons seemed adequate.
The decomposer model was tested in part by the verification of root dy-
namics, since dead roots were included in belowground litter. A better test
was to compare observed and simulated aboveground litter dynamics (Appendix
Figure 23.2-g). Litter biomass is also affected by the transfer from stand-
ing dead to litter, a poorly understood process. This lack of understanding,
along with the large variances in data, contributed to discrepancies between
observed and predicted litter dynamics.
The simulated shoot-nitrogen concentrations compared favorably with
observed values (Appendix Figure 23.2-h). The concentration dilution during
growth showed the correct timing and magnitude except late in 1975, when di-
lution or remobilization was overestimated. Simulated live-shoot sulfur con-
centrations (Figures 23.2 and 23.3) agreed with data in 1975 for A. smithii
and in 1976 for "other." The cause of the observed peak and decline in early
August 1975 for A. smithii is unknown and was not matched by the model. This
verification of shoot-sulfur concentrations establishes confidence in the
formulation of translocation and root uptake. Simulation of standing dead-
shoot sulfur concentrations (not shown) showed overestimations toward the ends
of the seasons, indicating an underestimation of sulfur remobilization and
retranslocatlon to roots.
RESULTS AND DISCUSSION
Total soil-sulfate levels (g S • m~2) are not presented, because they
remained essentially constant throughout the simulation, reflecting a near
balance of inputs and outputs to this central component. However, one
important deviation should be noted. In late June 1975, sulfate levels
dropped temporarily in the model by about 0.5 g S • m~2 at 1-5 cm, and by
approximately 1.5 g S • m~2 from July through August 1975 at 5-20 cm. This
behavior, caused by simulated precipitation of gypsum, suggested that the
combined concentration of calcium and gypsum in the real system may be
sufficiently high for this to occur. Further research is required to test
this prediction, including a representation of soil-calcium removal from,
and return to, the soil. The range of simulated sulfate-solution concen-
trations (Figure 23.4) was 16-328 ppm S. Highest levels coincided with periods
of lowest soil moisture, towards the end of the growing seasons.
Sulfide or reduced inorganic sulfur (Figure 23.5) rose to maxima of 0.015
g S • m~2 in the 0-1 cm layer, 0.06 g S • m~2 in the 1-5 cm layer, 0.055 g S •
m~2 in the 5-20 cm layer, and 0.09 g S • m~2 in the 20-60 cm layer. These
levels correspond to 1.48, 1.10, 0.27, and 0.14 ppm S (soil basis). Al-
though the initial conditions for this component were not known, levels
later in the simulation indicated a balance between oxidation and reduction
processes, implicit in the model formulation. The resulting low quantities
of reduced forms suggested that autotrophic sulfur oxidizers may be insigni-
ficant in this ecosystem.
323
-------
z:
o
<
0.4
^/) f—
o
o
CO
0
MJJASONDJ
1975
F M A M J
1976
J A
Figure 23-2. A. smithii live-shoot sulfur concentration verification.
o:
1.2
o -o 0.8
o _
i
cr CT
_i
h-
o
o
0.4
0
-------
CO
E
Q.
UJ
I-
u.
_l
CO
z
O
O
CO
320 -
240
160
80
0
Surface plus 0-1 cm
I- 5 cm ,
5-20 cm /
--- 20 -60 cm
MJJASOND.JFMAMJJA
1975 I 1976
Figure 23.4. Sulfate-solution concentrations.
0.08
\ 0.06
CO
Q 0.04
0.02
Surface plus 0-1 cm
1-5 cm
5-20 cm
20-60 cm
MJJASOND.JFMAMJJA
1975 I 1976
Figure 23.5. Sulfide or reduced inorganic sulfur.
325
-------
The range of C/S ratios of bacteria and actinomycetes (Figure 23.6) was
57-85, while that of fungi (Figure 23.7) was 182-230, which were reasonable.
Because bacterial C/S ratios exceeded 68 and fungal C/S ratios exceeded 202,
sulfur limited the decomposition of structural litter components formulated
in the model (Appendix Figure 23.2-g) The consistently wide C/S ratios of
both groups at 20-60 cm were the result of low solution concentrations of sul-
fate in that layer. The narrowest C/S ratios generally coincided with high
solution-sulfate concentrations. In the surface soil layer, widest C/S
ratios were simulated in the spring of 1975, when conditions for microbial
activity were most favorable, suggesting that sulfur recycling did not keep
pace with rapid microbial growth arid decomposition rates. Fungal C/S ratios
fluctuated less than bacterial C/S ratios, probably because of the lower
mineralization rate postulated for the fungi.
The range of sulfur concentrations of A. smith-Li live roots (Figure 23.8)
was 0.13-0.25%, with concentrations declining during periods of root growth
in both years, indicating a dilution effect. If the assumption that shoot-
sulfur concentrations are coupled to root concentrations according to a
phenologically dependent ratio holds, then the observed decline in shoot
concentrations between years implies a similar decline in root concentra-
tions. Because root uptake rates were adjusted to match this decline, root
concentrations necessarily dropped between years.
Accumulated flows of sulfur over two periods of the simulation are
presented in the form of input-output matrices (Tables 23.3 and 23.4). The
336-day period approximates the outcome of yearly processes, while the 182-
day period reveals the outcome of growing-season processes. Because sulfur
flows are highly dependent upon biomass (carbon) flows, key flows of carbon
are presented for comparison (Appendix Table 23.1).
The largest single accumulated flows were the uptake and mineralization
of sulfate by microbes: about 5.3 g S • m~2 followed both routes over the
336-day period and 4.7 g S • m~2 over the 182-day period. Although only a
small fraction of total sulfate turnover was the result of the action of
sulfatase on soil organic esters (3.0% and 3.5% over the two periods),
sulfatase release was comparable in magnitude to total root uptake (94% and
77% of root uptake), indicating that this could be a significant source for
plant growth.
Total root uptake over the two periods was 0.188 and 0.231 g S • m~ ,
of which 0.158 and 0.156 g S • nr2 was translocated to shoots. Of the
quantity translocated, 0.01 g S • m~2 (only 6% of total shoot uptake) was
remobilized during senescence and retranslocated to roots. As noted, sulfur
remobilization was significantly underestimated, perhaps by 2-fold.
For the two periods, microbes assimilated 0.462 and 0.191 g S • m~2
of metabolic-litter sulfur, 0.583 and 0.415 g S • m~2 of structural-litter
sulfur, 0.122 and 0.106 g S • m~2 of humad sulfur, and 0.002 and 0.003
326
-------
TABLE 23.3. ACCUMULATED SULFUR FLOWS, 20 APRIL 1975 TO 21 MARCH 1976 (336 DAYS, g S'nf2)
_. .. Standing Metabolic Structural
Shoots dead Roots Utter Utter
Shoots 0.147 0.01
Standing dead 0.052 0.069
Roots 0.158 0.196 0.043
Metabolic litter
Structural Utter
Sulfate 0.188
Reduced Inorganic
Grazing nematodes 0.029 0.013
Microbes 0.437 0.560
Saprophaglc nematodes 0.033 0.0033
Humads
Atmosphere 0.012
Total inputs 0.158 0.147 0.198 0.729 0.688
Flows to
Sulfate Redu"d Dazing Mlcrobes Saprophagic Resistant A nere
inorganic nematodes nematodes organic
0.157
0.121
0.004 0.015 0.416
0.462 0.431 0.012 0.905
0.583 0.015 0.598
5.36 5.55
0.250 0.250
0.042
5.34 0.387 0. 002/0. 028f 0.006 6.75
0.0066
0.014 0.122 0.122 0.258
0.012
5.77 0.387 0.045 6.53 0.006 0.446 0.122 0.012 15.23 system flux
* Each matrix element Indicates the accumulated flow from the row heading to the column heading, thus
row sums show total outputs and column sums show total inputs for individual state variables.
f fungi/bacteria
-------
TABLE 23.4. ACCUMULATED SULFUR FLOWS, 21 MARCH 1976 TO 20 SEPTEMBER 1976 (182 DAYS, g S • m 2
w
NJ
00
Standing Metabolic
Sho°CS dead Roots litter
Shoots 0.124 0.009
Standing dead 0.041
Roots 0.156 0.120
Metabolic litter
Sulfate 0.231
Reduced inorganic
Grazing nematodes 0.026
Microbes 0. 245
Saprophagic nematodes 0.0014
Humads
Resistant organic
Atmosphere 0.005
Total inputs 0.156 0.124 0.240 0.438
Flows to
Structural „ 1E Reduced Grazing , Saprophaeic , Resistant
,. Sulfate . . ,6 Microbes v vuaim- Humads . Atmosphere
litter inorganic nematodes nematodes organic
0.133
0.034 0.075
0.031 0.001 0.012 0.320
0.191 0.238 0.005 0.434
4.68 . 4.911
0.224 0.224
0.011 0.037
0.247 4.67' 0.133 0.0015/0.032* 0.003 5.331
0.0014 0.0028
0.015 0.106 0.106 0.227
0.164 0.003 0.167
0.005
0.324 5.074 0.133 0.045 5.395 0.003 0.249 0.106 0.005 12.29 system flux
fungi/bacteria
-------
o
H
CO
\
o
82
74
66
58
50
Surface plus 0-1 cm
1-5 cm
5-20 cm
20-60 cm
MJJASONDJFMAMJJA
1975 1976
Figure 23.6. C/S ratio of bacteria and actinomycetes.
CO
\
o
230
210
190
170
150
Surface plus 0-1 cm
1-5 cm
-5-20 cm
20-60 cm
J L I
M J J A S 0 N D
1975
Figure 23.7. C/S ratio of fungi.
J F M A M J J A
1976
329
-------
cn
O
\-
<
cr.
UJ
o
2
o
o
cr
z>
u_
_J
Z)
c/)
o
o
(X
2.6
2.2
1.8
1.4
1.0
Rhizomes plus 0-5 cm roots
5-20 cm roots
20-60 cm roots
i i
MJJ ASOND
1975
J F M A M J J A
1976
Figure 23.8.
Sulfur concentration in A. smithii live roots, by three
rooting layers.
g S • m 2 of resistant-organic-matter sulfur. During decomposition of
metabolic litter, only 0.012 and 0.005 g S • m~2 flowed to volatile forms.
Microbial death returned 0.437 and 0.245 g S • m~2 of metabolic litter,
and 0.560 and 0.247 g S • m~2 of structural litter. The total quantity of
sulfur gained by microbes during organic substrate decomposition was 1.169
and 0.715 g S • m~2. Because these amounts were greater than microbial
growth demands, net mineralization took place during decomposition. Because
net mineralization was of organic substrate sulfur, the question arises:
Why are the uptake levels of inorganic sulfur so high?
Grazing nematodes consumed 0.015 and 0.012 g S • m~2 of root sulfur, or
8% and 5% of total root uptake. Bacterial consumption by microbivorous
nematodes (0.006 and 0.003 g S • m~2) was surpassed by the bacterial con-
sumption of soil larvae forms of the herbaceous nematodes simply because of
the difference in their population sizes. These small quantities of
accumulated-sulfur flows show that the role of microbivorous nematodes in
sulfur cycling is quite small.
330
-------
Because it was assumed that there were no inputs or outputs to the
system as a whole, the sum of inputs to all individual components neces-
sarily equals the sum of outputs to all individual components. This quan-
tity is the total flux of sulfur within the ecosystem: 15.2 g S • m"2-
over the 336-day period and 12.3 g S • m~2 over the 182-day period. There
was a total of 96 g S • m~2 of sulfur in the model system, of which 78 g S
• m"2 were in humads and resistant organic material and therefore not in-
volved in rapid recycling. Hence, the turnover rate of the active pool was
approximately 15/18, or 0.83 yr"1. The calculated turnover rate of sulfate
was 0.36 yr-1; of metabolic litter, 10.0 yr"1; of structural litter, 1.2
yr"1; and of roots, 0.2 yr"1. In comparison, the turnover rate of humad
sulfur by decomposition was 0.015 yr"1, the turnover rate of resistant
organic sulfur by decomposition was 3 x 10~5 yr"1, and the turnover rate by
sulfatase activity of HI sulfur in humads and resistant organic components
was 0.004 yr"1.
Net gains or losses to individual components were calculated from the
difference between total inputs and outputs (Table 23.5). There were signi-
ficant net losses from roots, metabolic litter, and microbes. Significant
net gains accrued to sulfate, reduced inorganic sulfur, and humads. Because
there were no sulfur fluxes across the boundaries of the system, the total
reapportionment of sulfur among intra-system components could be calculated
as the sum of either all net gains or all net losses. This reapportionment
was 0.661 and 0.334 g S • nT2 over the 336- and 182-day periods. Because
only 4% and 2% of the total system fluxes were reapportioned, flows between
state variables were consistent with one another, the whole sulfur cycling
system was in equilibrium, and individual system components were maintained
near steady-state.
An assumption used in model formulation was that flows of sulfur
paralleled flows of biomass during growth, death, and heterotrophic consump-
tion. These flows could be calculated only in conjunction with submodels
describing the activities of the pertinent organisms. However, decomposi-
tion and mineralization decoupled individual elements, which reduced the
dependence of mineral flows on biomass or carbon flows. Because of this
decoupling, nutrient cycles must be considered individually. The behavior
of each element in its inorganic form is unique; for example, flows into and
out of the sulfate pool (such as soil sulfate adsorption, root and microbial
uptake, mineralization, sulfatase activity, and gypsum precipitation) are
specific to the sulfur cycle. Decoupling also affects the relationship
between elements in the organisms; for example, plant C/S or N/S ratios
could not be calculated without considering the availability of inorganic
sulfur in the soil.
Likewise, C, N, and S mineralizations from organic materials were not
formulated as strictly parallel processes. Microbial release and uptake of
331
-------
TABLE 23.5. NET INCREMENTS OR LOSSES IN INDIVIDUAL COMPART-
MENTS (g S • m~2) OVER TWO PERIODS OF SIMULA-
TION. TOTAL REAPPORTIONED SULFUR IN THE SYSTEM
IS THE SUM OF ALL NET LOSSES OR GAINS
Compartment
Shoots
Standing dead
Roots
Metabolic litter
Structural litter
Sulfate
Reduced inorganic
Grazing nematodes
Microbes
Saprophagic nematodes
Humads
Resistant organic
Atmosphere
Total reapportioned
336 days
0.001
0.026
-0.218
-0.176
0.09
0.22
0.137
0.003
-0.22
-0.0006
0.188
-0.042
0.0
~0.6610
182 days
0.023
0.049
-1.08
0.004
-0.102
0.163
-0.091
0.008
0.064
0.0002
0.022
-0.061
0.0
~0.3345
each element depends on the relationships among microbial requirements, the
nutrient concentrations of the substrates, and the ability of the microbe to
draw from the inorganic pool. Thus, nitrogen could be mineralized while
sulfur was immobilized. . By explicitly considering the sulfur requirements
of the microbes in relation to the sulfur gained through sulfate uptake and
heterotrophic decomposition, we did not need to use substrate C/S ratios to
predict mineralization.
332
-------
The dynamic framework provided by such a simulation approach should be
useful in assessing the impact of additions and losses of sulfur to the
system. The model-building process revealed several subjects for further
research, particularly on (a) microbial and root sulfate uptake kinetics,
(b) sulfatase activity, (c) the origin of soil sulfate esters, (d) the effects
of sulfur on plant and microbial metabolism, and (e) maximum microbial min-
eralization rates.
The model has been invaluable as a tool for integrating data, concepts,
and assumptions and, for now, seems to be the best way to deal with the many
interacting processes involved in nutrient cycles.
CONCLUSIONS
The simulation model described here proved to be a valuable conceptual
and experimental tool. Much was learned while constructing the model and is
currently being applied to second generation sulfur deposition and cycling
models.
REFERENCES
Alexander, M. 1971. Introduction to Soil Microbiology. John Wiley and
Sons, Inc., New York. 472 pp.
Attoe, V. J., and R. A. Olson. 1966. Factors Affecting Rate of Oxidation
in Soils of Elemental Sulfur and that Added in Rock Phosphate-sulfur
Fusions. Soil Sci., 101:317-325.
Banwart, W. L., and J. M. Bremner. 1976. Evolution of Volatile Sulfur
Compounds from Soils Treated with Sulfur-containing Organic Materials.
Soil Biol. Biochem., 8:439-443.
Barrien, B. S., and J. G. Wood. 1939. Studies on the Sulfur Metabolism of
Plants. II. The Effect of Nitrogen Supply on the Amounts of Protein
Sulphur and on the Value of the Ratio of Protein Nitrogen to Protein
Sulphur in Leaves at Different Stages During the Life Cycle of the
Plant. NewPhytol., 38:257-264.
Barrow, N. J. 1960a. The Effects of Varying the Nitrogen, Sulphur, and
Phosphorus Content of Organic Matter on its Decomposition. Aust. J.
Agric. Res., 11:317-330.
Barrow, N. J. 1960b. A Comparison of the Mineralization of Nitrogen and of
Sulphur from Decomposing Organic Materials. Aust. J. Agric. Res.,
11:961-969.
Barrow, N. J. 1961. Mineralization of Nitrogen and Sulphur from Sheep
Feces. Aust. J. Agric. Res., 11:645-650.
Barrow, N. J. 1967. Studies on the Adsorption of Sulfate by Soils. Soil
Sci., 104:342-349.
333
-------
Bell, G. K. 1975. Types of Sulphur Fertilizer Made in Australia and Their
Place in Agriculture. In: K. D. McLachlan, ed. Sulphur in Australian
Agriculture. Sydney University Press, Sidney. 261 pp.
Bettany, J. R., J. W. B. Stewart, and E. H. Halstead. 1973. Sulfur Frac-
tions and Carbon, Nitrogen, and Sulfur Relationships in Grassland,
Forest, and Associated Transitional Soils. Soil Sci. Soc. Am. Proc.,
37:915-918.
Bettany, J. R., J. W. B. Stewart, and E. H. Halstead. 1974. Assessment of
Available Soil Sulphur in an 35S Growth Chamber Experiment. Can. J.
Soil Sci., 54:309-315.
Biddulph, 0., S. Biddulph, R. Cory, arid H. Koontz. 1958. Circulation
Patterns for Phosphorus, Sulfur and Calcium in the Bean Plant. Plant
Physiol., 33:293-300.
Chaudry, I. A., and A. H. Cornfield. 1967. Effect of Moisture Content
During Incubation of Soil Treated with Organic Materials on Changes in
Sulphate and Sulphide Levels. J. Sci. Fd. Agric., 18:38-40.
Chao, T. T., M. E. Harward, and S. C. Fang. 1962. Adsorption and Desorption
Phenomena of Sulfate Ions in Soils. Soil Sci. Soc. Am. Proc., 26:234-
237.
Clark, F. 1977. Internal Cycling of 15Nitrogen in Shortgrass Prairie.
Ecology, 58:1322-1333.
Cole, C. V., G. S. Innis, and J. W. B. Stewart. 1977. Simulation of Phos-
phorus Cycling in Semiarid Grasslands. Ecology, 58:1-15.
Cooper, P. J. M. 1972. Aryl Sulphatase Activity in Northern Nigerian
Soils. Soil Biol. Biochem., 4:333-337.
Coughenour, M. B. 1978. Grassland Sulfur Cycle and Ecosystem Responses to
Low-level S02. Ph.D. Diss. Colorado State Univ., Fort Collins, Colorado.
353 pp.
Detling, J. K., W. J. Parton, and H. W. Hunt. 1979. A Simulation Model of
Bouteloua grac-Llis Biomass Dynamics on the North American Shortgrass
Prairie. Oecologia (Berl.), 38:167-191.
Dodd, J. L., W. K. Lauenroth, G. L. Thor, and M. B. Coughenour. 1978. Effects
of Chronic Low Level S02 Exposure on Producers and Litter Dynamics. In:
E. M. Preston and T. L. Gullett, eds., Bioenvironmental Impact of a Coal-
fired Power Plant, Fourth Interim Report, Colstrip, Montana. EPA-600/3-79-044.
U.S. Environmental Protection Agency, Corvallis, Oregon, pp. 384-493.
Dutt, G. R. 1962. Prediction of the Concentration of Solutes in Soil
Solutions for Soil Systems Containing Gypsum and Exchangeable Ca and
Mg. Soil Sci. Soc. Am. Proc., 26:341-343.
334
-------
Dutt, G. R. 1964. Effects of Small Amounts of Gypsum in Soils on the
Solutes in Effluents. Soil Sci. Soc. Am. Proc., 28:754-757.
Edwards, J. H., and S. A. Barber. 1976. Nitrogen Flux into Corn Roots as
Influenced by Shoot Requirement. Agron. J., 68:471-473.
Evans, C., and C. 0. Rost. 1945. Total Organic Sulfur and Humus Sulfur of
Minnesota Soils. Soil Sci., 59:125-137.
Ferris, H. 1976. Development of a Computer-simulation Model for a Plant-
nematode System. J. Nematol., 8:255-263.
Frederick, L. R., R. L. Starkey, and W. Segal. 1957. Decomposability of
some Organic Sulfur Compounds in Soil. Soil Sci. Soc. Am. Proc.,
21:287-292.
Freney, J. R. 1960. The Oxidation of Cysteine to Sulphate in Soil. Aust.
J. Agric. Res., 11:387-392.
Freney, J. R. 1961. Some Observations on the Nature of Organic Sulphur
Compounds in Soils. Aust. J. Agric. Res., 12:424-432.
Freney, J. R. 1967. Sulfur-containing Organics, In: A. P. McLaren and
C. H. Peterson, eds. Soil Biochemistry. Vol. 1. Marcel Dekker, Inc.,
New.York. pp. 229-259.
Freney, J. R., and K. Spencer. 1960. Soil Sulphate Changes in the Presence
and Absence of Growing Plants. Aust. J. Agric. Res., 11:339-345.
Freney, J. R., and F. S. Stevenson. 1966. Organic Sulfur Transformations
in Soils. Soil Sci., 100:307-316.
Freney, J. R., G. E. Melville, and C. H. Williams. 1971. Organic Sulphur
Fractions Labelled by Addition of 35S-sulphate to Soil. Soil Biol.
Biochem., 3:133-141.
Freney, J. R., F. S. Stevenson, and A. H. Beavers. 1972. Sulfur-containing
Amino Acids in Soil Hydrolysates. Soil Sci., 114:468-476.
Frissel, M. J., and P. Reiniger. 1974. Simulation of Accumulation and
Leaching in Soils. Centre for Agricultural Publ. and Documentation,
Wageningen, Netherlands. 116 pp.
Gustafson, J. D. 1977. SIMCOMP 4.0 Reference Manual. Natural Resource
Ecology Laboratory, Fort Collins, Colorado.
Gustafson, J. D., and G. S. Innis. 1973. SIMCOMP Version 3.0 User's Manual.
US/IBP Grassland Biome Tech. Rep. No. 218. Colorado State Univ., Fort
Collins.. 149 pp.
335
-------
Harvard, M. E., T. T. Chao, S. C. Fang. 1962. Soil Properties and Con-
stituents in Relation to Mechanisms of Sulphate Adsorption, In: Radio
Isotopes in Soil-plant Nutrition Studies. Int. Atomic Energy Agency,
Vienna, pp. 93-114.
Heitschmidt, R. K., W. K. Lauenroth, and J. L. Dodd. 1978. Effects of
Controlled Levels of Sulphur Dioxide on Western Wheatgrass in a South-
eastern Montana Grassland. J. Appl. Ecol. (submitted).
Houghton, C., and F. A. Rose. 1976. Liberation of Sulfate from Sulfate
Esters by Soils. Appl. Environ. Microb., 31:969-976.
Hunt, H. W. 1977. A Simulation Model for Decomposition in Grasslands.
Ecology, 58:469-484.
Ingelsten, B. 1966. Absorption and Transport of Sulfate by Wheat at Vary-
ing Mannitol Concentration in the Medium. Physiol. Plant, 19:563-579.
Kellog, W. W., R. D. Cadle, E. R. Allen, A. L. Lazrus, and E. A. Martell.
1972. The Sulfur Cycle. Science, 175:587-596.
Kowalenko, C. G., and L. E. Lowe. 1975b. Mineralization of Sulfur from
Four Soils and its Relationship to Soil Carbon, Nitrogen and Phos-
phorous. Can. J. Soil Sci., 55:9-14.
Leggett, J. E., and E. Epstein. 1956. Kinetics of Sulfate Absorption by
Barley Roots. Plant Physiol., 31:222-226.
May, P. F., A. R. Till, and A. M. Downes. 1968. Nutrient Cycling in Grazed
Pastures, I. A Preliminary Investigation of the Used of [35S] gypsum.
Aust. J. Agric. Res., 19:531-543.
May, P. F., A. R. Till, and M. J. Gumming. 1972. Systems Analysis of
35Sulphur Kinetics in Pastures Grazed by Sheep. J. Appl. Ecol., 9:25-
49.
McGill, W. B., H. W. Hunt, R. G. Woodmansee, and J. 0. Reuss. 1978.
Dynamics of C and N in Grassland Soils. Chapter 4. In: J. E. Ellis and
W. J. Parton, eds. Impact of Strip-mine Reclamation Practices A Simula-
tion Study. A report to the Western Energy and Land Use Team, Office
of Biological Services, U.S. Fish and Wildlife Service.
Metson, A. S. 1973. Sulphur in Forage Crops; Plant Analysis as a Guide to
the Sulphur Status of Forage Grasses and Legumes. Tech. Bull. No. 20,
The Sulphur Institute, Washington, D.C. 24 pp.
Monheimer, R. H. 1975. Effects of Three Environmental Variables on Sulfate
Uptake by Aerobic Bacteria. Appl. Microb., 30:975-981.
Nelson, L. E. 1964. Status and Transformation of Sulfur in Mississippi
Soils. Soil Sci., 97:300-306.
336
-------
Nissen, P. 1971. Uptake of Sulfate by Roots and Leaf Slices of Barley:
Mediated by Single, Multiphasic Mechanisms. Physiol. Plant, 24:315-324.
Parton, W. J. 1978. Abiotic Submodel. Chapter 2. In: J. E. Ellis and W.
J. Parton, eds. Impact of Stripmine Reclamation Practices: A Simula-
tion Study. A Report to the Western Energy and Land Use Team, Office
of Biological Services, U.S. Fish and Wildlife Service.
Parton, W. J., D. M. Swift, and J. E. Ellis. 1978. Chapter 6. In: J. E.
Ellis and W. J. Parton, eds. Impact of Strip-mine Reclamation Prac-
tices: A Simulation Study. A report to the Western Energy and Land
Use Team, Office of Biological Services, U.S. Fish and Wildlife Service,
Fort Collins, Colorado.
Peck, H. D., Jr. 1969. Sulfur Requirements and Metabolism of Microorganisms.
In: 0. H. Muth and J. E. Oldfield, eds. Symposium on Sulfur in Nutri-
tion. Oregon State Univ., Corvallis.
Peck, H. D. 1975. The Microbial Sulfur Cycle. In: Sulfur in the Environ-
ment. Missouri Botanical Gardens, St. Louis. 189 pp.
Persson, L. 1969. Labile-bound Sulfate in Wheat-roots: Localization,
Nature and Possible Connection to the Active Absorption Mechanism.
Physiol. Plant, 22:959-976.
Petrusewicz, K., and A. Macfadyen. 1970. Productivity of Terrestrial
Animals, Principles and Methods. IBP Handbook No. 13, International
Biological Program. F. A. Davis Company, Philadelphia. 190 pp.
Pettersson, J. 1966. Active and Passive Components of Sulfate Uptake in
Sunflower Plants. Physiol. Plant, 19:459-492.
Reiniger, P., M. J. Frissel, and P. Poelstra. 1971. Simulation of Cation
Exchange and Movement in Soils. Proc. Symp. on the Use of Isotopes
and Radiation in Research on Soil-plant Relationships Including
Applications in Forestry. Vienna, pp. 251-259.
Reuss, J. 0., and G. S. Innis. 1977. A Grassland Nitrogen Flow Simulation
Model. Ecology, 58(2):379-389.
Rogers, W. P. 1969. Nitrogenous Components and Their Metabolism: Acantho-
cephala and Nematoda. In: M. Forkin and B. T. Scheer, eds. Chemical
Zoology. Vol. II. Academic Press, New York. 687 pp.
Santmeyer, P. H. 1956. Studies on the Metabolism of PanagvetZ-Ls redivivus
(Nematoda, Cephalobidae). Proc. Helminthol. Soc., 23:30-36.
Smolik, J. D. 1974. Nematode Studies at the Cottonwood Site. US/IBP
Grassland Biome Tech. Rep. No. 251. Colorado State Univ., Fort Collins.
80 pp.
337
-------
Smolik, J. D. 1975. Effect of Soil Temperature and Soil Water on Preprod-
uction of Nematode Populations Indigenous to the Cottonwood Site.
US/IBP Grassland Biome Tech. Rep. No. 292. Colorado State Univ., Fort
Collins. 25 pp.
Starkey, R. L. 1950. Relations of Microorganisms to the Transformations of
Sulfur in Soils. Soil Sci., 70:55-65.
Stewart, B. A., L. K. Porter, and F. G. Viets, Jr. 1966a. Effect of Sulfur
Content of Straws on Rates of Decomposition and Plant Growth. Soil
Sci. Soc. Am. Proc., 30:355-358.
Stewart, B. A., L. K. Porter, and F. G. Viets. 1966b. Sulfur Requirements
for Decomposition of Cellulose and Glucose in Soil. Soil Sci. Soc. Am.
Proc., 30:453-456.
Stotzky, G., and A. G. Norman. 1961. Factors Limiting Microbial Activities
in Soil. II. The Effect of Sulfur. Archiv fur Mikrobiol., 40:370-382.
Tabatabai, M. A., and J. M. Bremner. 1970. Arylsulfatase Activity of
Soils. Soil Sci. Soc. Am. Proc., 34:225-229.
Tanji, K. K. 1969. Solubility of Gypsum in Aqueous Electrolytes as Affected
by Ion Association and Ionic Strengths up to 0.15 m and at 25°C.
Environ. Sci. Tech., 3:656-661.
Till, A. R., and P. F. May. 1970. Nutrient Cycling in Grazed Pastures, II.
Further Observations With [35S] Gypsum. Aust. J. Agric. Res., 21:253-260.
Till, A. R., and P. F. May. 1971. Nutrient Cycling in Grazed Pastures, IV.
The Fate of Sulphur-35 Following its Application to a Small Area in a
Grazed Pasture. Aust. J. Agric. Res., 22:391-400.
Tuovinen, 0. H., B. C. Kelley, and D. S. Nicholas. 1975. The Uptake and
Assimilation of Sulphate by Ih-ldbaoillus ferroxidans. Arch. Microbiol.,
105:123-127.
Walker, T. W. 1957. The Sulphur Cycle in Grassland Soils. J. Br. Grassl.
Soc., 12:10-18.
Williams, R. F. 1955. Redistribution of Mineral Elements During Develop-
ment. Ann. Rev. Plant Physiol., 6:25-42.
Williams, C. H. 1967. Some Factors Affecting the Mineralization of Organic
Sulphur in Soils. Plant Soil, 26:205-223.
Williams, C. H. 1975. The Chemical Nature of Sulphur Compounds in Soils,
In: K. D. McLahlan, ed. Sulphur in Australian Agriculture. Sydney
Univ. Press, Sydney. 261 pp.
338
-------
APPENDIX 23.1
Graphical Form of Functions Used in the Model
APPENDIX 23.1 FIGURE TITLES
a. Effect of soil temperature on microbial activity, E (T) (taken from
McGill et al., 1978). m
b. Effect of soil moisture on microbial activity, E (W);
(a) surface soil layer (b) subsurface soil layers (taken from McGill et al.,
1978).
c. Effect of microbial C/S ratio on uptake of solution sulfate, E (CS )
(based on calculated range of microbiol C/S ratios possible).
d. Effect of microbial C/S ratio on mineralization of sulfate, E (CS )
(based on calculated range of microbiol C/S ratios possible).
e. Effect of soil water on proportion of mineralized sulfur which is
oxidized (as sulfate) E (W) (based on Chaudry and Cornfield, 1967).
OX
f. Effect of soil temperature on sulfatase activity, E (T); dashed
lines indicate the temperature at which the unmodified rate was determined
experimentally (based on Tabatabai and Bremner, 1970).
g. Effect of microbial C/S ratio on decomposition of structural litter
(hypothesized).
h. Effect of soil moisture on the proportion of decomposing metabolic
litter sulfur which is volatilized, P (W) (based on Banwart and Bremner,
1976). V
i. Effect of soil moisture on root uptake of sulfate, E (W) (modified
from McGill et al., 1978). r
j. Effect of temperature (°- 1 on root uptake of sulfate, E (T) (modi-
fied from McGill et al., 1978). . r
k. Effect of whole plant C/S ratio on root uptake of sulfate, E (CS )
(hypothesized). u
1. The ratio of shoot sulfur concentration to live root sulfur con-
centration as a function of phenology, I(Ph). Plant phenophases are: (1)
early vegetative growth (2) middle vegetative growth (3) full vegetative
growth (4) reproductive growth (5) ripe fruit (6) senescence (hypothesized) .
m. Temperature dependent metabolic rate of nematodes, M (T) (adapted
from Santmeyer, 1956).
339
-------
Bacteria and actinomycetes
1.0
0.8
0'6
0.4
0.2
Bacteria and actinomycetes
Fungi
0 5 10 15 20 25 30 35 40
SOIL TEMPERATURE (°C)
Appendix Figure 23.1-a
Bacteria ~" 25 ~ 50 75 100 125
Fungi 150 175 200 225 250 275
C/S RATIO
Appendix Figure 23.1-d
' Bacteria and actinomycetes
v Fungi
0 10 20 30 40 50 60 70 80 90 100
SOIL MOISTURE TENSION (-bars)
Appendix Figure 23.1-b
Bacteria and actinomycetes
1.0
5 0.6
UJ
u.
LU 0.4
0.2-
Bacteria 25 50 75 100 125
Fungi 150 175 200 225 250 275
C/S RATIO
. 1.0
0.8
9 0.4
g 0.2
I I I I I IN 1 1
70 80 100 110 120 130 140 150
SOIL WATER (% field capacity)
Appendix Figure 23.1-e
2.0
1.5
1.0
0.5
20 40 ' 60
SOIL TEMPERATURE (°C)
80
Appendix Figure 23.1-f
Appendix Figure 23.1-c
340
-------
1.0
0.8
t
LLJ 0.4
0.2
Bacteria and actinomycete?
Bacteria 60
Fungi 200
70
225
80
250
90
275
C/S RATiO
Appendix Figure 23.1-g
100
300
10 20
SOIL TEMPERATURE
Appendix Figure 23.1-j
0.035
n
UJ
'- . r,.o30
0.020
a:
£
o
cc
o.
1.0 2.0 3.0 4.0
SOIL MOISTURE (-bars)
5.0
Appendix Figure 23.1-h
0.002 0.004 0.006 0.008
C/S RATIO OF WHOLE PLANT
0.01
Appendix Figure 23.1-k
10 20
SOIL MOISTURE (-bars)
Appendix Figure 23.1-i
J 2
23
PHENOLOGICAL STAGE
Appendix Figure 23.1-1
10 20 30 40
TEMPERATURE <°C)
Appendix Figure 23.1-m
50
341
-------
APPENDIX 23.2
Comparison Between Observed and Simulated Values for
Non-Sulfur Ecosystem Variables
APPENDIX 23.2 FIGURE TITLES
a and b. Observed vs. simulated 0-15 cm soil water (a) and 15-30 cm
soil water (b).
c and d. Observed vs. simulated total live shoot biomass (c) and for
Agropyron smithii live shoot biomass (d).
e and f. Observed vs. simulated rhizome biomass (e) and total root
biomass (f).
g. Observed vs. simulated litter biomass.
h. Observed vs. simulated nitrogen content of Agropyron smith-it live
shoots.
342
-------
40r
cc
O
•Oosarvad Simulated
r\ K-"\."\
1 \ r
\! \ \
\
"N
VJ
\.
,-, ;! I I L ! ' I I ' I ! I I I ! I I i
i,< J J A S 0 N 0 J F M A M J J A 3
1975
1976
M J J A S 0 N D J r M A M J o A S
_ „!
;
— y
d 100 'f
UJ
j~
li 50|
\ rw
Oil—! I J_i I I I '. ! 1 i 111 ! ! 1
i'vi J J A 3 0 N D J r ;A A i'.-i J J A S
I97G
i'. i '.-. i i i i Jd i i i i i
M J J A S 0 N 0 J F M A M J 0 A S
1975 1978
APPENDIX FIGURES 23.2-a,b,c,d,e,f,g,h.
343
^ 0.30.-
Tn> L
_ 0.24 P^
o.is
z O.i2r
o
o
., O.Coh
\
\
\
O
0 |j _ ! I I ! Ill I I ! I I I I I I t
M J J A S 0 N D J r M A M J J A S
IS75
I97G
-------
APPENDIX TABLE 23.1. ACCUMULATED CARBON FLOWS (g C
m
w
A. Primary production
P.. shoot
N
P plant (NPP)
Shoot growth
Rhizome or crown growth
Root growth
Crown respiration
Root respiration
B. Belowground
1 . Decomposition
Structural
Metabolic
Humad
Resistant organic
2. Respiration
Microbial
Nematode
3. Nematodes
Root consumption
Fungi consumption
20 April 1975 to 21 March 1976 to Total
21 March 1976 (336 days) 20 September 1976 (182 days) (518 days)
A. smith-Li Other A. smithii Other A. smithii Other
90.13 84.41 80.23 112.38 170.36 196.79
61.23 55.45 54.90 86.09 116.13 141.54
30.21 25.76 31.64 45.37 61.85 71.13
3.17 8.05 2.21 11.52 6.19 19.57
27.85 21.64 21.05 29.20 48.90 50.84
2.61 7.51 2.29 6.44 4,09 13.95
25.80 20.91 22.78 19.64 48.58 40.55
232.54 176.86 409.04
18.26 9.45 27.71
14.88 12.88 27.76
3.37 2.78 6.15
118.34 91.29 209.63
1.95 1.76 3.72
3.43 3.09 6.52
0.358 0.316 0.674
Bacterial consumption
(herbivore soil larvae)
Bacterial consumption
(saprophages)
2.02
0.448
2.25
0.199
ft.27
0.647
-------
SECTION 24
RELATIONSHIP OF S02 DEPOSITION TO A GRASSLAND SULFUR CYCLE
M. B. Coughenour
ABSTRACT
A simulation model representing deposition of S02 to
a grassland ecosystem was constructed. This model was
executed simultaneously with other models describing sul-
fur cycling, primary production, decomposition, and
abiotic processes. The model provided fair representa-
tion of the accumulation of sulfur on live and dead
leaves and soil. A deposition velocity of V =0.09 cm
• sec"1 was calculated from the verified simulation
results. As a result of continuous fumigation at 10
pphm S02, flows of sulfur through all system compon-
ents increased by approximately 1.21 g S • m~2 • yr"1.
Many flows of sulfur in the system increased, espe-
cially outflows from live and dead shoots. Total reap-
portionment of sulfur among system components was 3.8
times greater under the S02 exposure, indicating the
degree of departure of the system from equilibrium.
INTRODUCTION
•Effects of sulfur dioxide pollution have traditionally been limited to
assessment of toxic effects on metabolism at the organism level. However,
more subtle changes in the functioning of ecosystems may arise from an inter-
action of the pollutant with the nutrient cycle of which it is a part.
Analyses of the global sulfur budget reveal a major deviation from the
pre-industrial state. Man now contributes about half as much sulfur to the
atmosphere as nature (Kellogg et al. 1972). Globally, 20 x 106 metric tons of
sulfur (as S) are dry deposited on land (Almqvist 1974). Dry deposition over
land areas is probably about one-fourth of that brought down in precipitation,
while anthropogenic sources are wet and dry deposited on land at a rate of 24
x 106 tons • yr"1 each (Granat et al. 1976). Moss (1976) estimated that 100 x
105 tons are absorbed (as S02) by land surfaces, 75 x 105 tons are absorbed by
vegetation, and 25 x 106 tons by soil. Soil absorption downwind from a power
plant has been estimated at 49 to 98 kg S • ha"1 (Nyborg et al. 1973).
345
-------
These imbalances in the global sulfur cycle must consequently carry over
to terrestrial ecosystems, which remove portions of the atmospheric load. The
significance of an imbalanced terrestrial sulfur cycle is unknown. The stoi-
chiometry of life itself results in a certain relationship between macro-
nutrients in the biosphere (e.g., H2goo Ol'tSO ^1480 NIS Pl.8 ^ (Payrissat,
1974) . It is also conceivable that there exists a homeostatic control over
the entire planetary environment, including the chemical compositon of the
atmosphere, by and for the biosphere (Lovelock and Margulis 1974). Such a
homeostasis would be analogous to the homeostasis of biota with the physical
environment that arises during ecosystem development in which there is a closing,
or tightening, of nutrient cycles (Odum 1969). Ecosystem homeostasis may be
achieved by diversion of excess nutrients into unavailable pools or reduction
of losses from such pools so that depleted nutrients may later be restored
(Hutchinson 1948).
These hypotheses imply that the specific form of a nutrient cycle is re-
lated to the structure and function of the ecosystem. Thus, in assessing the
total environmental consequences of SC>2 pollution, all environmental components
must be drawn together and links between them should be made explicit. Refer-
ence to the sulfur cycle and the tracing of paths excess sulfur must follow
would facilitate such an approach (Moss 1976).
—precipitation
Figure 24.1.
Schematic of flows of SC^-sulfur to various sinks. Resistances
to S02 transfer are atmospheric boundary layer (ra), leaf bound-
ary layer (rb), leaf cuticular (re), stomatal (rs), leaf surface
(rsu), sub-canopy atmospheric (re), and ground surface (rsg.)
346
-------
MATERIALS AND METHODS
I developed a submodel (Figure 24.1) that could simulate the SC^-deposition
process and interface with a sulfur-cycling submodel (Figure 24.2) described
elsewhere in greater detail (Coughenour et al,, 1979). This model relies ex-
tensively on established micrometerological relationships presented in the
literature. Consequently, this paper does not document the derivations of
equations, and the reader is encouraged to examine relevant citations for de-
velopment and analysis.
To calculate the aerodynamic diffusion resistances within and above the
plant canopy, the windspeed profile must first be evaluated. An empirical
relation is used to calculate a zero plane displacement
d = 0.63 • h (1)
where h is the height (cm) of the plant canopy, and roughness length (z )
follows as
z k(h - d) (2)
o
where k is Von Karman's constant (0.41) (Monteith 1973).
Because the windspeed profile is affected by atmospheric stability, a
mean Richardson number is introduced
- S 2 (I - V (3,
U2T
r
where g is the gravitational constant (cm2 • sec"1), T is reference height
temperature ( K) , T is the temperature of the surface losing heat (canopy
temperature), U is reference height windspeed (cm • sec"1), and z is height
(Monteith 1963,rWhelpdale and Shaw 1974).
Friction velocity (U^) is then estimated with the stability-corrected
logarithmic windspeed profile relationship
^r = I [In (Zr-d)3 (1 - a R.)>2 (4)
U* K
where z is reference height (cm), a is an empirical constant (= 10, Monteith
1963), and U is reference height windspeed. The windspeed at canopy height
h is then
Uh - £ In (^) (1 - a I.)>2 (5)
o
The air layer within and above a canopy can be decomposed into three
regions (Inoue, 1963):
1. The lowest part of the atmospheric surface layer (atmospheric bound-
ary layer), usually represented by the logarithmic wind profile.
347
-------
[Atmospheric ^
Saprophagic
Nematodes
35-38
Figure 24.2.
Flow diagram showing structure of sulfur cycle model. Large
circles with numbers are duplicate representations of metabolic
and structural litter. Small circles represent the partitioning
of a flow. Flows to the atmospheric sink (volatilization) are
assumed to return immediately to metabolic litter. Litter com-
ponents occur at the soil surface and belowground. Metabolic
litter is labile, with variable sulfur composition, while struc-
tural litter is more resistant, with invariable sulfur composi-
tion. Reduced inorganic refers principally to the sulfide form.
Grazing nematodes consume roots and fungi in the adult stage,
bacteria in larval stages; saprophagic forms consume only
bacteria. Humads are relatively labile and represent soil
organic matter adsorbed and complexed with clay minerals.
Resistant organic if the highly recalcitrant fraction of soil
organic matter. Flows between atmosphere and metabolic litter
represent flows of organic volatile sulfur compounds rather
than S02.
348
-------
2. The canopy eddy layer, separated from layer 1 at canopy height h,
in which the wind profile is more complex.
3. The lowest part of the plant air layer, in which both plants and
ground surface influence the profile.
The necessary characteristics of region 1 have been calculated by Equa-
tions (1-5). The transition from region 1 to 2 may not be sharp, giving rise
to a transitonal region called the Honami-air layer. From this region down
into the canopy eddy layer (region 2) the logarithmic profile is replaced by
an exponential profile of eddy speeds. The depth of the canopy eddy is
Iw = 0.473(h -d) (6)
and extends down from approximately h (Inoue, 1963) . An estimate of the center
of the canopy layer is
i- Iw / -, v
zc - h - ~2 (7)
Within this layer, friction velocity varies with height; therefore, at z
M.c> - „
and, similarly, eddy velocity at z is
U(zc) -
where a is defined by the relationship
. k h
Iw = —
a h-d (10)
z
(Inoue, 1963). °
Over the logarithmic profile of region 1, the resistance to diffusion is
a ^2 (ID
where r is aerodynamic resistance (Monteith, 1973) . Diffusion through regions
3.
2 and 3 is more complex. A logarithmic profile is resumed in region 3 and the
atmospheric resistance from z to zero ground displacement may be approximated
by ° u(.c)
349
-------
An additional atmospheric diffusion resistance is due to a turbulent boundary
layer surrounding individual leaves. An empirical correlation has been dis-
covered that incorporates the effect of leaf shape on this resistance as
rb = 1.3(±-)a (13)
c
(Monteith, 1965), where 1 is the characteristic length of the leaf and U is
-2/3 °
canopy windspeed. However, r, varies with D (D is the diffusivity of the
gas). This leads to the relationship
c
(Kabel et aZ., 1976), which is the one used in this model. The characteristic
length (1) is
where A is planar area (cm2) of the leaf and P the perimeter (cm) (Kabel et al . ,
1976).
The final step of diffusion into the interior of the leaf requires evalu-
ating diffusive resistance through stomatal pores. Stomatal opening depends
upon light, plant water status, and atmospheric C02 concentration (Meidner and
Mansfield, 1968). Plant water potential was calculated empirically in equations
representing responses of Agropyron dasystaohyum to light and soil moisture
(Redman, 1973)
^ = -(10.59 +— ), if G < 40 (16)
P m —
Y0 = -(0.74 + — ), if G > 40 (17)
P m
*P = - [4.7 + 0.14(|g)], if m > 24 (18)
where 4* is plant water potential (bars) , m is soil moisture level (% volume) ,
and G is global radiation (ly/min) . Stomatal resistance is subsequently a
function of Y and G
l ?
r = - =^-= - ' if V > -IS
rs _ (0.2-0. 2O P
1 2 e
r = - - - if w < -
rs ,,(0.2-0. 02V) ' " Tp _
G r
which are empirical relationships derived from data for Agropyron dasystachyum
and A. smithii (Ripley and Saugier, 1975) . Since this resistance is applicable
to water vapor diffusion, the value is converted by
350
-------
DH20
where r is the resistance and D the diffusivity of the subscripted gas species.
An alternate pathway into the leaf interior is through the cuticle itself.
Although this resistance is large, it is incorporated here for completeness.
Cuticular resistance (r ) will be assumed at 80 sec • cm"1. When resistances
c
are in parallel as for r and r , the total resistance over the circuit is
s c
r • r
r = S . C (22)
r + r
s c
Resistance in series such as r and r, combine linearly as
a b
r = r + ru (23)
a b
The total resistance of the path into the leaf may then be written
r • r
r = r + ru + S . C (24)
a b r + r
s c
One further resistance (r ) may be postulated that expresses the capacity
m
of the mesophyllic cell surfaces to remove the gas for the substomatal cavity.
S02 is very highly soluble in water and is probably rapidly oxidized to S03=
on the mesophyllic surface. Thus, r may be negligible when the plants are
m
able to remove S03= from the mesophyll solution at a greater rate than it is
formed . Only when rate of entry exceeds this removal rate might r become
m
significant (Unsworth e~t at. 1976, Bennet et al. 1973). However, no measure-
ments of r are available (Unsworth et al. 1976). Since r, and r are known,
m b s
r must be adjusted to match observed values for accumulation of sulfur on
m
live leaves (after r is found) .
su
Passive deposition onto leaf surfaces is governed by the sink strength
of the surface. This sink strength is represented here as a constant leaf
surface resistance (r ) . The value of r was found by trial and error to
SU SU
most closely match observed and predicted values of sulfur accumulation on
dead leaves. The total resistance for deposition onto live or dead leaf
surfaces is then
r = r + r, + r (25)
a b su
In this model, flux to the soil is found in a similar fashion by use of
351
-------
r = r + r + r (26)
a e sg v '
where r and r are the atmospheric resistances to diffusion above and below
a e
the canopy, and r expresses the surface resistance of the ground. Estimates
sg
of soil surface resistance range from 0.24 sec • cm"1 (Payrissat and Beilke,
1975) to 5 sec • cm"1.(Seim, 1970). This suggests that no single value is
applicable in all circumstances. A surface resistance of 20 sec • cm"1 was
necessary to approximate the observed values of sulfate accumulation in the
surface soil layer.
The resistances of diffusion to each of the various sinks calculated
above are finally implemented to yield an absolute flux occurring over the
solution interval of the simulation (one day). The relationship
F = £ (27)
is used (Garland et at., 1973; Shepard, 1974) to facilitate this calculation,
where F is flux rate (yg SC>2 • cm"2 • sec"1), C is concentration (yg S02 • cm"3)
and r is total resistance from the point at which C was measured to the sink
(sec • cm"1). When C is measured at z , r must be included in r, but when C
IT 3.
is measured in the canopy (as is the case here) r need not be included in r.
3.
Flux to leaf surfaces or interiors is
F = £ • LAI (28)
where LAI is leaf area index. Flux through stomates is evaluated two times
for each daily period of sunlight as r changes in response to diurnal varia-
s
tion in light (G). Fluxes over each of the two time increments are summed to
yield daytime flux. This is then added to a nighttime flux, when only the
cuticular path is assumed operative.
Inorganic sulfur is leached from leaf surfaces as a function of the in-
tensity and quantity of rain (Tukey, 1966). Recent dead Agropyron smithii at
the end of the fumigation year in September 1975 averaged about 0.44% sulfur.
Over winter this level declined in the absence of fumigation to about 0.11%
sulfur In old dead leaves the following March. Over this period there was 16.6
cm precipitation yielding the empirical concentration loss rate of 0.02% sul-
fur per cm of precipitation used in the model. This is an approximation be-
cause a significant portion of the precipitation fell as snow and the intensity
of the precipitation was not considered.
RESULTS AND DISCUSSION
Model performance was verified in a simulation that included interactions
with sulfur cycling and ecosystem level models (Coughenour et at., 1979). This
simulation represented the continuous field fumigation at 10 pphm 862 (canopy
concentration) beginning on 10 May 1975 and ending 31 October 1975. Verifica-
tion of the S0£ deposition submodel was accomplished primarily through compar-
ison between observed and simulated values of live and dead A. smithi-i shoot
352
-------
sulfur concentrations (Figure 24.3). The comparison for live A. smithii shoots
displayed fairly acceptable agreement, except for the underestimation of depo-
sition in the early part of the season. The overall rate of accumulation,
however, seemed correct. The model satisfactorily represented the changes,in
concentration of sulfur on old dead leaf surfaces except in the latter part of
the season, when simulated values rose above observed. The model simulated a
2.5-fold increase in soil sulfate in the top 5-cm soil layer due to S02 depo-
sition over two consecutive growing seasons, in agreement with field deter-
minations.
The total flux of S02 to the ecosystem amounted to 2.03 g S • m~2 for the
174 days of fumigation at 10 pphm S02. This deposition rate is transformed in-
to a deposition velocity (V ) for the period.
O
cm
. _ flux rate (yg SO?
cr
concentration (pg S02
= 0.09 cm • sec'1
cm 3)
(29)
The accumulated flows of sulfur over the simulation period (Table 24 .1)
may be compared to those from the nonfumigated simulation over the same period
5.0r
% 4-°
o
E
•1 3.0
o>
o>
2.0
1.0
Data Simulated
Live shoots
Dead shoots
_^ S
__ s
May
June
July
August September
Figure 24.3.
Comparison of observed and simulated Agropyron smithii shoot
sulfur concentrations under 10 pphm S02
353
-------
TABLE 24.1. ACCUMULATED SULFUR FLOWS (g S • m~2) FROM 20 APRIL 1975 TO 21 MARCH 1976 (336 DAYS)
UNDER 10 pphm S02 (CANOPY CONCENTRATION) FUMIGATION
U)
.p-
Flows to
Shoots s"ndi"« Roots Metabolic Structural .Reduced Grazing Mlcrobes Saprophagic Resistant Ac here
dead litter litter inorganic nematodes nematodes organxc
Shoots 0.181 0.03 0.246
0.481T
Standing dead 0.251 0.070 0.377
0.074**
Roots 0.052 0.205 0.045 0.005 0.015
Metabolic, litter 0.582 0.513 0.016
Structural litter 0.587 0.015
Sulfate 0.188 5.260
Reduced inorganic 0.252
Grazing nematodes 0.030 0.014
Microbes 0.442 0.579 5.216 0.418 0.002 0.006
0.030}
Saprophagic nematodes 0.003 0.003
Humads 0.006 0.046 0.046
Resistant organic 0.056 0.001
Atmosphere 0.427 0.422 0.0221 0.016 0.713
0.484
Total inputs 0.963 1.084V 0.218 0.947 0.711 6.945 0.423 0.047 6.476 0.006 0.528 0.046 0.016
* Total outputs or inputs to each compartment are obtained by summing across the row and down the column respectively.
surface/internal
**leaching/death
T fungi/bacteria
§ translocation
Total outputs
0.938
0.772
0.322
1.111
0.602
5.448
0.252
0.044
6.692
0.007
0.981
0.057
2.062
18.40 system flux
-------
to assess the impacts of S02~sulfur inputs on the sulfur cycle. First, the
total system flux (TSF) is defined as
n n
TSF = Z Z FLOW.. (30)
i J 1J
where FLOW., is the accumulated flow from state variable i to state variable j,
and n is the total number of state variables. The increase in TSF due to S02
sulfur inputs for the 336-day simulation period is
ATSF = TSF^ - TSFcontrol = 3.17 g S • m* (31)
A significant portion of ATSF is a direct result of the atmospheric flows. If
these flows are subtracted from ATSF, the indirect effects of S02 on TSF may
be found,
m
ATSF. ,. = ATSF - Z FLOW. . (32)
indirect . kj
where k is the index of the atmospheric compartment, and m is the total num-
ber of flows from the atmosphere. This indirect effect was 1.12 g S • m~2 for
the 336-day period.
Perturbations to individual 'lows are expressed by the ratio of the accu-
mulated flow in the S02~fumigated simulation to that in the control simulation
(Table .2). The greatest positive perturbations occurred in flows from shoots
to standing dead, shoots to roots, standing dead to metabolic litter, and
leaching from live and dead shoots to soil sulfate. As a result of greater
inflows to metabolic litter, there were also greater outflows from metabolic
litter to microbes, humads, and volatile organic forms. As flows from shoots
to roots increased, so did flows from roots to litter and sulfate. Because of
greater flows from litter and sulfate to microbes, there were greater flows
from microbes to litter and nematodes. Finally, increased inputs to nematodes
from microbes resulted in greater outflows from nematodes to litter. No'per-
turbations occurred in outflows from structural litter because this component
was assumed to have an invariable sulfur composition. Although there was no
net change in root sulfate uptake over the whole period, this flow was in fact
greater in the fumigated simulation for most of the growing season. Root up-
take was, however, accelerated in the control simulation in the period follow-
ing the growing season. This compensated for the sulfur used in shoot growth.
The same phenomena held true for microbial sulfate uptake. Fumigation stimu-
lated microbial uptake during the growing season, but uptake in the control
simulation was accelerated thereafter and eventually surpassed that in the
fumigated simulation.
Only one flow was consistently suppressed due to fumigation, that from
roots to shoots. This was the result of a mechanism in the model that de-
presses upward translocation as shoot sulfur demands are progressively ful-
filled. Uptake from the atmosphere reduced shoot demands. Flows from humads
355
-------
TABLE 24.2. THE RATIO OF ACCUMULATED SULFUR FLOWS IN THE 10 pphm FUMIGATION SIMULATION (TABLE 24.1)
TO SULFUR FLOWS IN THE NON-FUMIGATED CONTROL SIMULATION FOR THE 336 DAY PERIOD
Co
Ui
0\
Flows from
Flows to
Standing „ Metabolic Structural „ ., Reduced Grazing „. , Saprophagic .. . Resistant
Shoocs . . Roots ,. , Sulfate . . ~ Microbes r r a Humads Atmosphere
dead litter litter inorganic nematodes nematodes organic
Shoots
Standing dead
Roots
4.50 3.00
4.82
0.33 1.04
„
1.01
1 . 05 1 . 30
1.00
Metabolic litter
Structural litter
Sulfate
Reduced inorganic
Grazing nematodes
Microbes
Saprophagic nematodes
Humads
Resistant organic
Atmosphere
1.00
1.26
1.00
1.04
1.19
1.00
1.33
1.03
1.01
1.06
1.33
1.08
1.03
1.06
1.00
0.98
0.43
0.34
1.08
1.07
1.07
0.38
0.50
0.38
* The ratio is infinite when the denominator is assumed to be zero. Pertubations to flows were progressively damped
or they became more removed from the sinks directly removing S0_ from the atmosphere.
-------
and resistant organic matter did not actually decrease in response to added
sulfur but were lower in the fumigated simulation because organic soil sulfur
was lower in the fumigated than in the control field plots.
The impact of S02 on sulfur cycling is further displayed in the differ-
ence between total inputs and outputs to each compartment, which is equivalent
to the net change in the compartment (Table 24.3). When inputs and outputs to
a compartment are equal, that compartment is at steady state. Soil sulfate
was perturbed from steady state by 862 inputs more than by any other state
variable. The relative S02~induced perturbations from steady-state behavior
for each compartment were SO^ > standing dead > humads > roots > reduced or-
ganic > resistant organic > shoots > structural litter > metabolic litter >
microbes > nematodes.
When either all net gains or all net losses are summed, a total reappor-
tionment among system components is derived (Table 24.3). This may be taken
as an aggregate measure of all compartment steady states or of system equili-
brium, since if all components were in balance, there would be zero reapportion-
ment. Significantly, this measure was 38 times greater under the continuous
S02 exposure. This reapportionment principle also would apply to the biosphere
and to the "Gaia" (Loveland and Margulis 1974), in that there has been a shift
in the equilibrium between sulfur locked away in fossil fuels, sulfur in the
atmosphere, and sulfur in terrestrial ecosystems.
CONCLUSIONS
A deposition velocity to the ecosystem as a whole was derived from 'simu-
lation modeling. The value was approximately .1 cm sec"1. At this rate, with
10 pphm S02 maintained continuously, 48 kg S ha"1 yr"1 could be removed from
the atmosphere. At 2 pphm about 10 kg S ha"1 yr"1 could be removed.
Sulfur absorbed by the ecosystem is a contaminant in the sense that the
mean biogeochemical compositions are significantly altered, as are internal
recycling rates, and reapportionments. Whether this contamination is harmful
or adverse, depends on the relationships of other ecosystem processes to the
sulfur cycle. In a system where sulfur is a non-limiting nutrient, the effects
would be subtle, or non-detectable in the short term, but could accumulate over
long periods such as decades. Nutrient cycles are, however, related to overall
ecosystem homeostasis. In mature ecosystems cycling rates are adjusted to
maintain near steady states on an annual basis. Sulfur inputs significantly
perturb these rates, thus perturbing the ecosystem beyond its normal homeo-
static domain. Whether this perturbation is harmful or not is in large part a
value judgement in the context of the benefits of generating electrical energy.
357
-------
TABLE 24.3. NET INCREMENTS (POSITIVE) OR LOSSES (NEGATIVE) OF
SULFUR* IN INDIVIDUAL COMPARTMENTS OVER THE 336-DAY
PERIOD IN THE CONTROL AND THE 10 PPHM S02 SIMULATIONS1"
Compartment
Shoots
Standing dead
Roots y
Metabolic litter
•s
Structural litter
Sulfate
Reduced inorganic
Grazing nematodes
\
Microbes
Saprophagic nematodes
Humads
Resistant organic
Atmosphere
Total reapportioned
Control
0.001
0.026
-0.218
-0.176
0.090
0.220
0.137
0.003
-0.220
-0.001
0.188
-0.042
0.000
,0.661
10 pphm S02
0.025
0.312
-0.104
-0.164
0.109
1.497
0.171
0.003
-0.216
-0.001
0.429
-0.011
-2.030
,2.536
g s
-2
t Total reapportionment among compartments is the sum of either all
losses or all gains and is therefore a measure of system
equilibrium.
358
-------
REFERENCES
Almqvist, E. 1974. An Analysis of Global Air Pollution. Ambio, 3:161-167.
Bennet, J. H., A. C. Hill, and D. M. Gates. 1973. A Model for Gaseous Pollu-
tant Sorption by Leaves. Air Pollut. Cont. Assoc. J., 23:957-962.
Coughenour, M. B., W. J. Parton, W, K. Lauenroth, J. L. Dodd, R. G. Woodmansee.
1979. Simulation of Sulfur Cycling in Grasslands. Ecol. Model, (accepted).
Garland, J. A., W. S. Clough, and D. Fowler. 1973. Deposition of Sulphur
Dioxide on Grass. Nature, 242:256-257.
Granat, L., R. 0. Hallberg, and H. Rodhe. 1976. The Global Sulphur Cycle, In:
B. H. Swensson and R. Soderlund, eds. Nitrogen, Phosphorus and Sulphur-
global Cycles. SCORE Report 7. Ecol. Bull. 22, Stockholm, pp'. 89-134.
Hutchinson, G. E. 1948. Circular Causal Systems in Ecology. Ann. New York
Acad. Sci., 50:221-246. (
Inoue, E. 1963. On the Turbulent Structure of Airflow Within Crop Canopies.
J. Meteorol. Soc. Japan, 41:317-325.
Kabel, R. L., R. A. O'Dell, M. Taheri, and D. D. Davis. 1976. A Preliminary
Model of Gaseous Pollutant Uptake by Vegetation. Center of Air Environ-
ment Studies. Publ. No. 455-476. Pennsylvania State Univ. 96 pp.'
Kellogg, W. W., R. D. Cadle, E. R. Allen, A. L. Lazrus, and E. A. Martell.;
1972. The Sulfur Cycle. Science, 175(11):587-596.
Lovelock, J. F., and L. Margulis. 1974. Atmospheric Homeostasis by and for
the Biosphere: The Gaia Hypothesis. Tellus, 26:1-9.
Meidner, H., and T. A. Mansfield. 1968. Physiology of Stomata. McGraw-Hill
Book Co., New York. 179 pp.
Monteith, J. L. 1963. Gas Exchange in Plant Communities. Chaper 7. In: L.
T. Evans, ed. Environmental Control of Plant Growth. Academic Press,
Inc., New York.
Monteith, J. L. 1965. Evaporation and Environment. Symp. Soc. Exp. Biol.,
Vol. 19, pp. 205-233.
Monteith, J. L. 1973. Principles of Environmental Physics. American Else-
vier Publ. Co., Inc., New York. 241 pp.
Moss, M. R. 1976. Biogeochemical Cycles as Integrative and Spatial Models
for the Study of Environmental Pollution (The Examples of the Sulphur
Cycle). Int. J. Environ. Studies, 9:209-216.
359
-------
Nyborg, M., and McKinnon, Allen and Associates Ltd. 1973. Atmospheric Sulphur
Dioxide: Effects on the pH and Sulphur Content of Rain and Snow; Addition
of Sulphur to Surface Waters, Soil, and Crops; and Acidification of Soils.
Proc. Workshop on Sulphur Gas Research in Alberta. Infor. Rep. NOR-X-72,
Northern Forest Research Centre, Edmonton, pp. 79-97.
Odum, E. P. 1969. The Strategy of Ecosystem Development. Science, 164:262-
270.
Payrissat, M. 1974. The Cycle of Sulphur Dioxide in the Atmosphere. Euro-
spectra, 13(l):16-22.
Payrissat, M., and S. Beilke. 1975. Laboratory Measurements of the Uptake of
Sulphur Dioxide by Different European Soils. Atmos. Environ., 9:211-217.
/' ,
Redman, R.-E. 1973. Plant Water Relationships. Tech. Rep. No. 29. Canadian
Committee for the International Biological Program, Matador Project. 84 pp.
Ripley, E., and B. Saugier. 1975. Energy and Mass Exchange of a Native Grass-
land in Saskatchewan, In: D. A. deVries and N. H. Afgan, eds. Heat and
Mass Transfer in the Biosphere. Part 1. Transfer Processes in the Plant
Environment. John Wiley and Sons, Inc., New York. pp. 311-325.
Seim,VE. 1970. Sulfur Dioxide Absorption by Soil. Ph.D. Diss. Univ. Minne-
sota. 138 pp.
>
Shepard, J. G. 1974. Measurements of the Direct Deposition of Sulphur Dioxide
onto Grass and Water by the Profile Method. Atmos. Environ., 8:69-74.
>
Tukey, H. B., Jr. 1966. Leaching of Metabolites from Above-ground Plant Parts
and its Implications. Bull. Torey Bot. Club, 93:385-401.
Unsworth, M. H., P. V. Biscoe, and V. Black. 1976. Analysis of Gas Exchange
Between Plants and Polluted Atmospheres, In: Experimental Studies of the
Biological Effects of Environmental Pollutants. Cambridge Univ. Press.
pp. 5-16.
Whelpdale, D. M., and R. W. Shaw. 1974. Sulphur Dioxide Removal by Turbulent
Transfer over Grass, Snow, and Water Surfaces. Tellus, 26:196-204.
360
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
REPORT NO.
EPA-600/3-80-052
I. RECIPIENT'S ACCESSION NO.
TITLE AND SUBTITLE
The Bioenvironmental Impact of a Coal-Fired Power Plant;
Fifth Interim Report, Colstrip, Montana. April, 1980.
5. REPORT DATE
June 1980 issuing date
6. PERFORMING ORGANIZATION CODE
. AUTHOR(S)
Edited by Eric M. Preston and David W. O'Guinn
8. PERFORMING ORGANIZATION REPORT NO.
I. PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. EPA
Corvallis Environmental Research Laboratory
200 S.W. 35th St.
Corvallis, OR 97330
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
SAME
13. TYPE OF REPORT AND PERIOD COVERED
Interim 12/77-712/78
14. SPONSORING AGENCY CODE
EPA/600/02
^.SUPPLEMENTARY NOTES In t^±& serieS) the lst interim Rept. is EPA number EPA-600/3-76-002
the 2nd Interim Rept. is EPA number EPA-600/3-013; the 3rd Interim Rept. is EPA number
EPA-6QO/3-78-021; the 4th Interim Rept. is EPA number EPA-6QQ/3-79-Q44.
16. ABSTRACT
The EPA has recognized the need for a rational approach to the incorporation of
ecological impact information into power facility siting decisions in the northern
great plains. Research funded by the Colstrip, Coal-Fired Power Plant project is a
first attempt to generate methods to predict the bioenvironmental effects of air
pollution before damage is sustained. Pre-construction documentation of the environ-
mental characteristics of the grassland ecosystem in the vicinity of Colstrip, Montana
began in the summer of 1974. Since then, key characteristics of the eco-system have
been monitored regularly to detect possibe pollution impacts upon plant and animal
community structure.
In the summer of 1975, field stressing experiments were begun to provide the data
necessary to develop dose-response models for 862 stress on a grassland ecosystem.
These experiments involve continuous stressing of one acre grassland plots with
measured doses of S02 during the growing season (usually April through October).
Results of the 1978 field season's investigations are summarized in this
publication. The six-year project will terminate in 1980 and a final report will be
published after data analyses are complete.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
plant and animal response to pollution
coal-fired power plant
air pollutants
grassland ecosystems
mathematical modeling
remote sensing
micrometeorological investigation
coal-fired power plant
emissions
air quality monitoring
aerosol characterization
51
18. DISTRIBUTION STATEMENT
release to public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
378
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
361
------- |