United States
Environmental Protection
Agency
Environmental Research
Laboratory
Duluth IN/IN 55804
Research and Development
EPA-600/S3-84-051  May 1984
Project  Summary
Changes in Terrestrial  Ecology
Related  to  a  Coal-Fired  Power
Plant: Wisconsin  Power  Plant
Impact Study

Daniel E. Willard, William W. Jones, and Jay Benforado
  The full report summarizes the eco-
logical effects on terrestrial plants and
animals  from a  1054-MW coal-fired
power plant. Research was conducted
from 1971 through 1977 at the site of
the Columbia Generating Station, in the
eastern flood-plain of  the Wisconsin
River in south central Wisconsin.
  Initial studies were largely descriptive,
involving development of species lists
and documentation of habitats. Mapping
of plant communities documented
extensive changes, most notably heavy
losses of sedge meadow  habitats and
increases in open water. These impacts
on plant communities  brought about
concomitant changes  in animal com-
munities.
  Bird populations were observed and
analyzed in a sampling system consisting
of five permanent transects, each
having 17 or 18 stations and representing
the entire range of plant communities.
In mean analysis, a "t" test was applied
to identify significant differences in the
mean number of birds observed in each
station during each year of study. The
Shannon-Weaver function, which com-
bines  species richness and evenness
components into a single index, was
used  as a measure of diversity. Polar
ordination, a  type of multivariate
analysis based on a similarity coefficient
to measure "distances" between sta-
tions, revealed many  changes  in  the
composition  and structure of bird
communities. Observations of a nesting
colony of great blue herons showed that
the nesting population  decreased and
foraging habits changed as a result of
the impact of the cooling lake on the
wetlands.
  Drift fence-pitfall trapping is one
source of information on presence and
abundance of species, migration routes
and seasonal use of different habitats,
and occurrence of rare and endangered
species. This report summarizes the
effects of drift fence construction and
orientation, trap dimensions, cycles in
time,  and weather variables  on the
results  obtained  by drift fence-pitfall
trapping.
  This Project Summary was developed
by EPA 's Environmental Research
Laboratory, Duluth, MN, to announce
key findings of the research project that
is fully documented in a separate report
of the same title (see Project Report
ordering information at back).

Introduction
  This report summarizes several studies
on the effects of a 1054-MW coal-fired
power plant on  terrestrial  plants and
animals. The studies were conducted
from 1971 through 1977 at the site of the
Columbia Generating Station near Portage,
Wisconsin, in the eastern floodplam of
the Wisconsin River. Before construction
began, the dominant vegetation  of the
area was sedge meadow and -lowland
forest  Building and operating the power
plant had major consequences for the
surrounding communities of plants and
animals.

Approach
  The initial studies were primarily
descriptive. They involved developing

-------
lists of species, evaluating habitats, and
documenting elements in the  natural
history of important species. For popula-
tion studies, field censuses were made
on  a  regular  basis,  using visual  and
acoustical observations, traps, and  nets
to assess densities of animal populations.
At the community level, plant communities
and  habitat types were  mapped  and
analyzed by aerial photography, point-
quarter, and transect methods. Finally,
analytic studies were made to determine
the physiological and behavioral responses
to the impact  of the power plant  This
variety of approaches to impact analysis
provided a balance between long-term
qualitative observations and short-term
quantitative studies.

Studies  of Plant Communities
  Studies of plant communities are
important  both in their  own right and in
relation to  animal studies,  because
impacts on animal populations must be
understood in terms  of any  habitat
changes.  Eighteen natural plant com-
munities were identified, described, and
mapped Table 1 summarizes changes in
the sizes of plant communities between
1971  and  1979. The  net loss of 638 ha
represents the area occupied by the
generating station and its associated
facilities Four plant communities disap-
peared  completely, while  heavy losses
were  sustained by the sedge  meadow
community 1

Descriptive Animal Studies
  Some 287 species of vertebrates use
the Columbia  Generating Station  site
This total is distributed as follows among
the classes of vertebrates-
  Fishes                        22
  Reptiles and amphibians        32
  Birds                        197
  Mammals                      36
Usage  by some species  (e.g., geese,
warblers, walleyes) is seasonal, and some
of the animals are rare (muskellunge,
sandhill cranes, otter). Many species are
seasonally very abundant (red-winged
blackbirds,  muskrats, chorus frogs,
walleyes).  The site contains a number of
game species (ducks, white-tailed deer,
northern pike)
  Although changes  were observed m
many  animal  populations,  determining
the significance of these  changes  was
difficult and  identifying  causal links was
sometimes impossible.  The difficulty in
observing  some animals, and the size of
the site made complete coverage difficult.
Long-term observations are needed to
distinguish  natural  fluctuations  from
those induced by human activity. Finally,
Table 1.    Changes in Size of Plant Communities (ha) from 1971 to 1979
Community type
Open water
Emergent Aquatics
Sedge Meadow
Shrub Car/-
Alder Thickets
Fen
Prairie
Southern Wet Forest
Southern Wet Mesic Forest
Southern Mesic Forest
Southern Dry Mesic Forest
Southern Dry Forest
Northern Wet Forest
Northern Wet Mesic Forest
Oak Barrens
Sand Communities
Cliff Face
Old Field
Total
Area
1971
236
358
662
34
12
6
7
137
270
15
93
70
2
3
19
23
18
81
2046
Area
1979
274
311
181
13
0
0
4
165
270
18
88
26
0
0
7
30
18
35
1408
Area lost
1971-197-9
47
68
481
28
12
6
3
63
0
1
18
54
2
3
12
4
0
71

Area gained
1971-1979
74
21
0
7
0
0
0
91
0
3
13
0
0
0
0
11
0
15

Net
change
+27
-47
-481
-21
-12
-6
-3
+28
0
+2
-5
-54
-2
-3
-12
+7
0
-56
-638
animals tend to move in response to their
needs, so that some areas which seem
but lightly used may in fact be temporary
refuges necessary for survival.

Mammals
  Mammals  were  studied by direct
observation, winter tracks, box traps, drift
fences, dropping boards, den counts, and
records of trappers.
  Estimates of deer populations show a
reduction from  80  to 100 in  1971 to
approximately 50 in 1977. Destruction of
habitat coupled with increased access for
hunters was  probably the cause for this
decline
  Based on an average of five muskrats
per den, populations of muskrats varied
from approximately 1000 in  1972 to 250
in 1973.  Few dens were observed in
1976.  Muskrats do not exhibit regular
population  cycles  common in some
smaller microtines but are  sensitive to
changes in water level and other environ-
mental factors. Because of environmental
changes generally favorable for muskrats,
it is  expected that populations  will
increase.

Reptiles and Amphibians
  Reptiles and amphibians were studied
by seining, netting and trapping; and by
direct observation (shed skins, road kills,
chance  encounters, resting  places,  and
habitats).
  All  amphibian species seemed to be
reduced in numbers. Results for reptiles
were inconclusive due to sparsity of data.
Three species of turtle  occupy waters
throughout the  site  and nest in several
onsite areas.  The six-lined race runner,
an  endangered species of  lizard,  may
have been lost through habitat destruc-
tion.  Sightings  of  several  species  of
snakes (mostly road  kills) were too
infrequent for any conclusions about
population trends. The garter snake
remains in wet areas throughout the site.

Birds
  Birds were studied  during  regular
walks along a  set of  five transects,
through observation of confined  habitats
such as the cooling lake, and by incidental
observation
  The total of 197 species observed from
1973  to  1977  includes  permanent
residents, seasonal  residents, migratory
visitors, and species utilizing temporary
transitional habitats created by construc-
tion activities.
  Effects of the cooling  lake exceeded all
other impacts  on birds  Numerous
species dependent on the wetlands for
nesting or foraging were displaced when
the 200-ha lake was  filled. Affected
species included the red-winged blackbird,
sora rail, Virginia rail, American bittern,
harrier, and great blue  heron  Sandhill
cranes, relatively intolerant of disturbance,
deserted adjacent wetland areas.
  The new cooling lake, however, attracted
thousands of  migratory birds. Twenty-
two species of waterfowl were observed
during the 1975 spring  migration. These
included the double-crested  cormorant,
an endangered  species  in Wisconsin.
Only 17 migrant species used the cooling
lake in 1976  Furthermore, although the
coot population increased sharply, num-
bers of nearly all  other species were
reduced. The  decrease  in  use of the
cooling lake by migratory waterfowl may
be due to changes in the temperature or
quality of the water, or to the generally
wet conditions which created ephemeral

-------
ponds on  nearby lands that may have
attracted the birds. Additional studies of
the lake's use by migrating waterfowl are
recommended 2
Population and Community
Studies

Transect Counts and Analyses
of Bird Populations
  The sampling system consisted of five
permanent transects, each having 17 or
18 observation stations 2500 m2 in size.
All major plant communities were
sampled open water, emergent aquatic,
sedge meadow, shrub carr, southern wet
forest,  southern wet mesic forest,
southern dry forest/southern dry mesic
forest,  sand barrens, and old  field. In
addition, a  number of stations represented
community edge types.
  To obtain a sample count, an observer
walked  for three minutes through  a
station on  its center line, using visual and
song cues to record all birds observed
within the boundaries of the station. Each
transect was sampled approximately
every 10 days from April to October and
every 20 days during the rest of the year,
from 1973  through 1976  Data were
analyzed by three methods: mean analysis,
a diversity index, and polar ordination.


Mean Analysis
  In mean analysis, a "f'testwas applied
to identify significant differences in the
mean number of birds observed  in each
station during each  year of the study.
Mean analysis  revealed  many  cases
where a specific construction activity was
followed by an immediate decrease in the
sample  mean  In several cases,  the
activity resulted in total destruction of the
habitat and the mean remained depressed.
Where construction caused only  partial
or temporary destruction of a habitat, the
mean tended to rebound. Stations inun-
dated by  the cooling  lake showed  a
decrease in mean number of birds except
when occasional rafts of waterfowl
appeared

Diversity Index
  Although there is no concensus on how
to measure diversity, species diversity is
generally thought to be associated with
the stability of a habitat. This study
employed the Shannon-Weaver function,
H',  which  uses information theory for
analysis of community organization,
combining richness  and evenness com-
ponents into a single index of diversity.
The equation is.
              IM
        H' = -  I ( *J In  *',
              i-1   T       T
where
   N=total number of species
   T=total number of individuals
   x,=number of  individuals in  the ith
     species.
  Although the diversity index was lower
for impacted  stations in most habitats,
the decrease was usually small. More
significantly, diversity values tended to
decrease with time at impacted stations,
while diversity at nonimpacted stations
did not show this trend. The decrease in
H' was most marked in impacted sedge
meadow habitats where, following a peak
in H' during  filling  of  the lake, values
dropped  sharply. A similar  decrease
occurred  in impacted old field stations.
  Values of H' are more  closely related to
changes  in  species richness than to
changes  in  species  abundance  Thus,
while values for  the sample mean  may
bounce back following construction,
values for diversity do not  Construction
activities  therefore have a larger impact
at the community level  than is indicated
by changes in species abundance alone.

Polar Ordination
  Polar ordination, a type of multivariate
analysis,  was used to study changes in
composition  and structure of bird com-
munities through detection of patterns in
the  data. This  technique expresses
variation  by measuring the distance
between  each pair of stands with com-
munities  by  means of a  dissimilarity
coefficient, equal to
               1 - 2w
                 a + b
where
   a=sum of abundance  values of all  spe-
     cies in one stand
   b=sum of abundance  values of all  spe-
     cies in another stand
  w=sum of  abundance values that the
     two stands have in  common for each
     species.
These differences are then used to place
stands on axes
  For ordination studies, data were taken
from observations during June and  July
(the  breeding season) 1973  through
1976. The most  common species  (e.g.,
song  sparrow,  blue jays, red-winged
blackbirds) were relatively unaffected by
construction  activities The dominance of
these species is evident in the ordination
plots, in which it overshadows the effects
of construction on rarer  species and thus
decreases the separation of impacted vs.
nonimpacted stations. Nevertheless,
ordination studies revealed major changes
in bird communities at stations affected by
construction activities  Many of these
changes could be related to alterations in
the water regime,  with consequent
effects on plant communities.
  Clearly, construction of the Columbia
Generating  Station  has had a major
impact on local bird communities.  Impacts
range  from  immediate  responses to
specific construction activities, to changes
in diversity  and  evenness  in  various
habitats, to alteration in overall commun-
ity composition One widespread result is
the proliferation of common species at
the expense of rarer ones. This can have
serious  implications  for community
productivity, stability,  and succession

Effects on Great Blue Herons
  A great blue heron nesting colony is
located about 2 km west of the wetlands
on the generating station site In 1974,
the colony contained 113 active nests at a
density of 94/ha. The herons foraged in
groups of up to 1 2, in the diked area of the
wetlands During the winter of 1974, the
dikes were joined and the cooling  lake
was filled.  Few herons were observed
foraging on the site during the following
summer,  and the density of nests in the
colony dropped to 69/ha. In 1976, herons
again began foraging along the dikes, but
the nest  density dropped to 33/ha. The
herons no longer foraged in groups, but
rather, fed individually along the  dikes at
regularly  maintained  intervals of about
150m  By 1977, the number of nests had
fallen  to  16 (9/ha),  even  though the
nesting site itself was unchanged from its
preconstruction condition.
  The response of great blue herons to
construction of the Columbia Generating
Station was clearly a response to  impacts
on the feeding grounds ratherthan onthe
nesting site  Although  the cooling  lake
contains  abundant food, the distribution
of food and its availability to great  blue
herons have  changed  With foraging
possible only in limited areas along the
lake shore, the herons changed from a
pattern of feeding in  groups to one of
solitary feeding in individual territories.
With food supplies curtailed, the  nesting
population dropped sharply.

Drift Fence-Pitfall Trapping  for
Data  on Populations of Small
Animals
  Baseline studies of populations of
small animals are often concerned  with
the presence and abundance of species,
migration routes  and seasonal  use of

-------
different habitats, and the occurrence of
rare or  endangered species  Drift fence
trapping  can  provide some  of this
information.
  This report summarizes the effects of
various  factors on  results  obtained  by
drift fence-pitfall trapping.  The  factors
include:
  1. Construction of drift fences (sheet
     material vs screen),
  2  Size and shape of traps,
  3. Orientation of  drift fences  (parallel
     or  perpendicular to habitat bounda-
     ries),
  4  Cycles in time  (daily and seasonal),
     and
  5  Weather (e g., precipitation, cloud
     cover, temperature, wind).
  Trapping took place between June  11
and October31,1976. Daily visits to traps
insured sampling of populations without
undue mortality. Altogether, 1227 indivi-
duals representing  28 species of small
mammals, reptiles, andamphibianswere
captured (Table 2). The species composi-
tion of captured animals varied spatially
and temporally,  with some  species
captured throughout the study and others
captured  discontmuously. It  would  be
desirable to collect data for an entire year,
to  provide  information on seasonal
changes in populations and their use of
various habitats.
  In general, more amphibians were
caught on fences running parallel to land-
water boundaries, while more mammals
were caught on fences running perpendic-
ular to these boundaries. This is because
amphibians tend to travel toward or away
from water, and mammals tend to travel
   parallel  to  water boundaries. With a
   longer period for collecting data, observa-
   tions on  relationships between animal
   captures and drift fence orientation could
   provide information on migration  routes
   and changes in habitats with time
    The study showed also that temperature
   and moisture are important in controlling
   movements of small  animals, and  that
   there is  interaction among weather
   variables. The effect  of weather on
   trapping results could be incorporated
   into a predictive model that would yield
   overall population estimates from actual
   numbers of  animals captured  under
   specific weather conditions
   Conclusions

   Plant Communities
     The construction  of the Columbia
   Generating Station resulted in the net
   loss of 638 ha of native plant communities
   on the 2046 ha site. The largest single
   loss was in sedge meadow (481 ha). Four
   minor communities  (alder thicket,  fen,
   northern  wet forest,  and northern  wet
   mesic forest) were lost completely.
   Seepage  and upwellmg of cooling  lake
   water under the dikes is dislodging peat
   mats which continues to affect the sedge
   meadow community west of the cooling
   lake.

   Mammals
     White-tailed deer populations on the site
   declined with the onset of construction
   due  to increased hunting pressure.
                                     Habitat destruction  and plant  facilities,
                                     which tend to restrict deer movements
                                     along the Wisconsin River, have continued
                                     to  keep deer populations depressed.
                                     Increases  in the amount and  depth of
                                     water in the sedge meadow should favor
                                     muskrats but may limit the range of voles
                                     and shrews.
                                       Drift fence-pitfall trapping  proved  a
                                     suitable method for capturing a variety of
                                     small mammals, reptiles, and amphibians.
                                     Capture rates depend on trap size and
                                     shape, orientation,  timing, and weather
                                     conditions. The method was useful for
                                     gathering baseline data.

                                     Birds
                                       Construction activities directly affected
                                     bird abundance; however, if the  habitat
                                     destruction was modest, abundance
                                     levels in the affected  areas returned to
                                     predisturbance levels with the completion
                                     of the construction activity. In the sedge
                                     meadow  and old  field communities,
                                     construction  activities kept species
                                     abundance below preconstruction levels.
                                       Species  diversity indices  generally
                                     decreased over time in those  sampling
                                     stations impacted by construction activi-
                                     ties. This suggests that  construction
                                     activities had a larger community  level
                                     impact than species abundance changes
                                     alone  indicated.
                                       The disturbances on the site seemed to
                                     favor the more common bird species over
                                     the rarer ones. Species which declined in
                                     number during the study period  include
                                     harriers, sandhill cranes, and great blue
                                     herons. A  great blue heron colony which
                                     used the  Columbia sedge meadow for
Table 2.    Summary of Pitfall Trap Captures of Small Mammals. Reptiles, and Amphibians
Mammals
Sorex arcticus
Saddlebacked shrew

S cinereus
Cinereous shrew
B/arina brevicauda
Short tailed shrew
Microtus pennsylvanicus
Meadow vole
No. Reptiles
Ophisaurus attenautus
Western glass lizard
446
S tor en a dekayi
Dekay's snake
12 S occipitomaculata
Red-bellied snake
Thamnophis sirtalis
Eastern garter snake
No,
3


34

82

54

Amphibians
Bufo americanus
American toad

Rana ciamitans
Green frog
R pipiens
Leopard frog
R. sylvatica
Wood frog
No.
124


52

185

2

M. Ochrogaster
  Prairie vole
Zapus hudsonius
  Hudsonian jumping mouse
Peromyscus bairdi
  Prairie deer mouse

P gracilis
  Woodland deer mouse
Eutamias jacksoni
  Wisconsin least chipmunk
                                      79
44
Opheodrys vernalis
  Smooth green snake
Heterodon platyrhmos
  Eastern hognose snake
Chrysemys picta
  Painted turtle

Emydndea blandingi
  Blandmg's turtle
Graptemys geographica
  Map turtle
Trionyx spmiferus
  Spiny soft shell turtle
25

 6

20


 7

 2

 1
Pseudacns trisenata
  Chorus frog
Hyla crueller
  Spring peeper
H. versicolor
  Grey tree frog

Ambystoma laterale
  Blue-spotted salamander
A  tigrinum
  Tiger salamander
  3

83

  8


23

  2

-------
feeding declined from 113 active nests in
1974 to 16 nests in 1977
  Polar ordination was useful in consider-
ing how certain environmental param-
eters affect bird species distributions. The
results  suggest that polar ordination
could be used to predict how changes in
these parameters  might  affect species
distribution and abundance.

Recommendations
  The baseline data and, consequently,
the results of this study would have been
more useful had  they been  obtained
before construction began. Unfortunately,
the initial field sampling  occurred after
the start of construction.
  Methods have recently been developed
to provide before-and-after assessments
in a shorter time frame and at less cost
than the present study. Most notable of
these  is  the  U.S  Fish and Wildlife
Services Habitat Evaluation Procedures
(HEP). However, such methods are  not
designed  to consider  impacts to  rare
species or unique habitats.  In these
cases, direct surveys and measurements
must be used.
  Drift fence-pitfall trapping and transect
counts  are two conventional methods
that can be easily and cheaply used to
gather baseline data. When  conducted
over sufficient time, such methods
effectively monitor and document impacts
to terrestrial communities.

References
   1. Bedford,  B. and 0.  Loucks, 1982,
Response of Carex-Dommated Wetlands
to Altered Temperatures and Flooding
Patterns,  Water  Resources Center,
University  of Wisconsin, Madison, Wis-
consin,  135 p.
  2. Jaeger, M J. Predicting Effects of an
Electric Station on Wetland Passerine
Birds, EPA-600/3-80-020, U.S. Environ-
mental  Protection Agency, Duluth, Min-
nesota,  25 p.
D. E. Willard, W. W. Jones, and J, Benforado are with Institute of Environmental
  Studies, University of Wisconsin, Madison, Wl 53706.
Gary E. Glass /s the EPA Project Officer (see below).
The complete report, entitled "Changes in Terrestrial Ecology Related to a Coal-
  Fired Po wer Plant: Wisconsin Power Plant Impact Study," (Order No. PB 84-171
  701; Cost: $13.00, subject to change) will be available only from:
        National Technical Information Service
        5285 Port Royal Road
        Springfield, VA22161
        Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
        Environmental Research Laboratory
        U.S. Environmental Protection Agency
        Duluth, MN 55804

-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
        PS
        U  i>  t"I
        RtbiU
-------