United States
Environmental Protection
Agency
Environmental Research
Laboratory
Ouluth MN 55804
Research and Development
EPA-600/S3-83-081 June 1984
&EBA Project Summary
Response of Carex-
Dominated Wetlands to
Altered Temperature
and Flooding Patterns
Barbara Bedford and Orie Loucks
This report describes the effects of
construction and operation of an 1100-
MW coal-fired power plant on surrounding
wetland plant communities. Preliminary
studies began in 1971, and data
collection continued after the second
generating unit went into operation in
1978.
In its preconstruction state, about
82% of the 1900-acre site was wetland.
Almost half of it was an extensive
marsh, in which the cooling lake was
sited.
Existing paradigms for ecosystem
response gave no consistent answers
to what type and extent of impact the
cooling lake would have. Therefore, in
addition to the monitoring of changes in
vegetation which began in 1974. a new
basis was sought for predicting ecosystem
response. The research included four
phases: field inventory and classification
of plant communities, monitoring of
vegetation, field and laboratory experi-
ments to test hypotheses regarding
mechanisms regulating population
changes, and assessment of field and
theoretical approaches.
The wetland species were not equally
sensitive to flooding, heat stress, or
increased flow of surface water. Extreme-
ly sensitive species, such as the sedges
Carex lacustris and C. rostrata, declined
rapidly in all communities. Other
species such as C. str/cta and Calamagro-
st/s canadensis responded more slowly.
In contrast, the cattail Typha latifolia
increased continuously in the emergent
community and invaded all other
communities. Other hydrophytic but
non-persistent species such as Sagfttaria
latifolia increased in some communities
but not in others. Lemna minor (duckweed,
a floating annual) increased sharply in
expanding areas of open water. Annual
species which had been insignificant or
even absent before the disturbance,
e.g., Bidens cornua and Pilea pumila,
increased where perennial dominants
declined. Thus, changes occurred in the
structure of the plant communities as
well as in populations of individual
species.
This study revealed no consistent
relationship between species diversity
and the intensity of environmental
disturbance. Differences in measures of
diversity between communities were
related to the differential sensitivity of
spatially dominant species to disturbance.
Where dominant species were sensitive,
diversity and equitability increased as
new or competitively inferior species
colonized or spread to space made
available by the decline of dominant
populations. Diversity and equitability
decreased where disturbance favored
one or two dominant species.
It appears that the timing of phonolog-
ical events (such as the reproductive
cycle or the emergence of shoots) in the
life history of a plant may be important
determinants of a population's re-
sponse to heat stress. Physiological and
phonological data from laboratory and
field experiments related the response
of 7. latifolia and C. lacustris to altered
seasonal patterns of carbohydrate
storage and shoot phenology induced
primarily by groundwater temperatures
-------
out of phase with normal cycles of plant
growth. These data established the
causal mechanisms relating changes in
vegetation to leakage from the cooling
lake, identified field signs of heat stress,
and suggested general species charac-
teristics that may be useful indicators
of potential sensitivity to stress.
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).
Background
This research on wetland plant responses
was pan of a larger study of environmental
changes caused by construction and
operation of a large coal-fired electric
power plant in Columbia County north of
Madison, Wisconsin. The research objec-
tives of the program as a whole were
formulated in 1970 in the context of
Wisconsin's Shoreline Zoning Act of
1965, imminent passage of the Wisconsin
Environmental Policy Act (1971), and a
long history of protection of navigable
waters under the Public Trust Doctrine of
the State Constitution. Regulatory respon-
sibilities established under these legislative
and constitutional mandates required
evaluations, in advance, of the prospective
effects of the proposed development. The
Columbia study, undertaken initially with
support from the three Wisconsin utilities
building the power plant, was intended to
provide information for government and
industry in future environmental protection
decisions involving coal-fired steam
electric plants. Additional support from
the U.S. Environmental Protection Agency
permitted the study to expand after 1975
to take advantage of the baseline studies
and obtain more definitive answers to a
wide range of questions.
The project provided a unique opportun-
ity to examine fundamental ecological
questions about the response of wetland
ecosystems to the external influences of
nearby development. The Columbia
station and its associated facilities are
located in the f loodplam of the Wisconsin
River. In its preconstruction state, about
82% of the site was wetland. Almost 50%
was an extensive marsh (31% sedge
meadow, 17% emergent aquatic vegeta-
tion), considered to be important as a
spawning habitat for walleye and northern
pike. These areas represented a major
portion of the nonwooded wetlands
remaining on the Wisconsin River in
Columbia County.
The cooling lake for the generating
station was sited in the marsh.The
utilities and the project research staff
predicted that the lake would leak
substantial amounts of warm water.
Although the effect of the leakage on the
remaining marsh was not examined fully
during preconstruction evaluation, it
became a major question during later
assessment of the effects of the power
station.
The monitoring of changes in vegetation
associated with leakage from the cooling
lake began in 1974. Although it was
known that the remaining wetlands
would be affected by elevated groundwater
levels and temperatures due to leakage
from the cooling lake, the type, magnitude,
and rate of change were unknown. Based
on existing paradigms for ecosystem
response, the potential effects cited by
the Environmental Impact Statement
(Wisconsin Department of Natural Re-
sources, 1973) ranged from a gradual
decrease of intolerant species, increase
of early-blooming plants, and decreased
diversity to disrupted competitive interac-
tions, changed rates of succession, and
shifts to types of vegetation (such as
cattail) unsuitable for spawning habitat.
In response to concern expressed by
several state and federal agencies, the
U.S. Army Corps of Engineers predicted
that there would be no significant effect
on the marsh vegetation. The environmen-
tal assessment was left in that form.
Research Objectives
The principal objective of the wetland
vegetation study, to determine how the
wetland plant communities west of the
cooling lake might change in response to
mild heat and water stress, required
answers to five questions:
1. What types of change would
occur? Would these changes be at
the level of species responses only,
or would the entire community
change? Would the composition or
the structure of the vegetation
change?
2. What would be the magnitude of
change? Would changes be within
the range of natural variability in
wetland ecosystems? How large an
area would be affected?
3. At what rate would the changes
occur? Would they be gradual or
would they take place within a
short time and be more noticeable
to observers?
4. Would a new equilibrium commun-
ity become established or might
the vegetation continue to change
throughout almost the entire dura-
tion of the stress?
5. Would changes be irreversible, or
might the communities recover
after a period of complete abatement?
Wetland vegetation is dynamic, and
methods for determining the response of
vegetation to waste heat and elevated
groundwater levels are not widely agreed
upon. Wetlands in the upper midwest
sometimes undergo natural shifts in
species composition and structure with-
out being subject to anthropogenic
environmental influences. Thus, this
study examines whether changes induced
by leakage from the cooling lake could be
detected within the limits of existing
techniques for field monitoring. It was
also necessary to determine whether the
theoretical paradigms for analysis provided
criteria by which significant impact could
be recognized.
The research objectives went beyond a
simple documenting of changes in
wetland plant communities to an under-
standing of the process of change itself.
Research goals designed to elucidate this
process fit into three categories:
1. Documenting the response of
wetland plant species and commun-
ities to mild heat and water stress.
2. Determining causal mechanisms
potentially relating changes in
vegetation to environmental varibles
that, in turn, could be related to
leakage from the cooling lake.
3. Examining available field techniques
and paradigms for recognizing,
monitoring, and — if possible —
projecting the probable effects of
waste heat and elevated water
levels on wetland vegetation.
The wetland plant ecology research
included four phases: field inventory and
classification of plant communities,
monitoring of vegetation, field and
laboratory experiments to test hypotheses
regarding mechanisms regulating popula-
tion changes, and assessment of field and
theoretical approaches.
The Study Site
The Columbia Electric Generating
Station is located in southcentral Wisconsin,
6.4 km south of Portage, on a 1900-ha
site which is almost entirely in the
floodplain of the Wisconsin River. Before
construction, an extensive and diverse
wetland system covered more than 80%
of 1100 ha declared for utility purposes. 1
Dominant types of vegetation in the|
wetland system included marsh (emergent
-------
ml
Sedge Meadow
Wet Forest
Dry Forest
Sampling Area
\
Figure 1.
Distribution of major plant communities before construction, showing future sites
of generators and associated facilities.
aquatics and southern sedge meadow),
southern lowland forest, and lowland
shrub communities. In 1971, more than
50% of the area was classed as nonf orested
wetland. Figure 1 shows the plant
communities at the site before construc-
tion. The future locations of the main
generating units and associated facilities
are indicated, and the area of sedge
meadow sampled in this study is shown.
The wetlands on the Columbia site
occupy a former channel of the Wisconsin
River in an area of regional groundwater
discharge. A layer of peat varying in
thickness from 1 to 3m covers a thin layer
of organic clay and silt, which overlies
alluvial sands with clay lenses. Bedrock,
composed of Upper Cambrian sandstones
and Pre-Cambrian granites, occurs at a
depth of 125 m below the surface.
Before the cooling lake was filled,
groundwater discharge, precipitation,
and floodwaters supplied water to the
wetland communities on the site. The
most important source of water was
precipitation. This provided about 76
cm/yr, more than half of it falling
between May and September. Inflow
from groundwater averaged 22 cm/yr.
Floodwaters from the Wisconsin River
and Duck Creek frequently inundated the
site, usually in early spring. However,
their residence time in the wetlands was
short, generally less than 3 days, due to
the gentle slope of the area toward the
Wisconsin River. Other inflows of surface
water were minor.
The levels of surface water in the
wetlands fluctuated widely before the
lake was filled. Generally, levels were
high during spring floods, decreased in
summer, and rose again in the fall as plant
growth and evapotranspiration ceased.
Frequently, surface water disappeared
entirely during the summer.
The climate around the site is continental.
Weather is seasonably variable with a
wide annual temperature range. Mean
annual temperature is 6.5°C (standard
deviation = 0.9°), with a mean summer
maximum of 25.7°C, a mean winter
maximum of -3.3°C, and a mean minimum
of -12.9°C. The first killing frost usually
occurs around September 30, and the last
frost, around May 1.
The Columbia station consists of two
527-MW coal-fired electric power generat-
ing units, a 200-ha cooling lake, a 28-ha
ashpit, a 16-ha coal pile, and other
associated facilities. Construction began
in 1971. Pumping to fill the cooling lake
took place during June and July 1974. By
September, the lake was almost empty
due to high leakage rates. Consequently,
the dikes were sealed with bentonite and
the lake was again filled during November
and December 1974. The first generating
unit began operating in May 1975; the
second went into operation in the spring
of 1978.
Preliminary studies of wetland vegetation
on the site were carried out from 1971 to
1973. Collection of data for the study
reported here began in June 1974 and
has continued to the present time.
Studies of the groundwater system and
surface flows began in 1972.
Findings
Data collected from 1974 through
1977 showed that changes in water
levels and temperatures caused by
seepage from the cooling lake led to
significant changes in populations and
communities of wetland plants.
-------
Types of Change
Population studies revealed that different
species were not equally sensitive to
flooding, heat stress, or increased flow of
surface water. Some species which were
extremely sensitive, e.g., Carex lacustris
and C. rostrata, declined rapidly and
dramatically in all communities. Other
perennial species such as C. stricta and
Calamagrostis canadensis responded
more slowly, decreasing only after 1976.
Typha latifolia, in constrast, increased
continuously in the emergent community
and invaded all other communities as
well. Other hydrophytic but nonpersistent
species such as Sagittaria latifolia
increased in some communities but not in
others. Lemna minor, a floating annual,
increased sharply in expanding areas of
open water. Certain annual species, e.g.,
Bidens cernua and Pilea pumila, which
had been insignificant or even absent
before the disturbance, increased markedly
where perennial dominants declined.
Taken together, these observations show
that significant changes occurred in the
structure of the plant communities as
well as in populations of individual
species.
Species diversity frequently seems to
be related to the stability of the environ-
ment. This study revealed no uniform
relationship between species diversity
and the intensity of environmental
disturbance. Neither the richness of
species nor the distribution of their
abundances was consistently affected.
Although obvious changes in community
structure took place during the study
period, the ecological significance of
different patterns of response in the
various communities was not evident
from the diversity data alone. Disturbance
did not simply increase or decrease
diversity. Species richness, diversity, and
equitability sometimes varied independent-
ly of each other. All of the measures of
diversity employed decreased in some
cases, increased in others, and frequently
did both during the period from 1974
through 1977 as the spatial distribution
and intensity of disturbance increased.
Observed differences in measures of
diversity between communities were
related to the differential sensitivity of
spatially dominant species to disturbance.
Where disturbance was of the type to
which dominant species were sensitive,
diversity and equitability increased as
new or competitively inferior species
colonized or spread to space made
available by the decline of previously
dominant populations. Diversity and
equitability decreased where disturbance
favored one or two dominant species.
Carex lacustris showed an almost
uniformly negative response to disturbance.
In contrast, Typha latifolia exhibited both
positive and negative responses. It
showed a clear positive response to
elevated water levels, increasing its
density where it had been abundant and
expanding its distribution substantially by
1977. Adverse effects on Typha were
limited to areas where temperature
increases of more than 7°C occurred
during the winter.
Phenological changes (associated with
the relationships between climatic variables
and periodic biological phenomena such
as flowering) appear to be reliable
indicators of heat stress. Although
changes in population density occurred
throughout the study area, obvious
phenological changes were not observed
outside the area affected by altered
temperature patterns. In areas receiving
waste heat, individual plants showed
visible signs of stress before the population
collapsed. Where the species grew in
dense stands, the symptoms were visible
to the naked eye at distances up to 300 m.
In general, phenological changes became
evident 1 year before population density
declined and 2 years before the population
collapsed.
Observable and easily measured charac-
teristics of plants experiencing heat
stress included unseasonable chlorosis,
reduced height of mature shoots, increased
height of new spring and fall shoots, and
early or delayed shoot emergence at
reduced density.
Magnitude of Change
The magnitude of effect produced at
both the population and the community
levels appears to be outside the range of
natural varibility for this particular
wetland. Many records and lines of
evidence indicate that densely vegetated
perennial wetland plant communities
have occupied the site for at least the last
150 years. Thus, although shifts toward
more hydrophytic species and decreased
vegetative cover are not uncommon for
some wetland systems, evidence for the
Columbia site suggests that the vegetation
mapped in 1974 represented a fairly
long-lived community type under histori-
cally prevailing environmental conditions.
By 1977, changed water levels significant-
ly affected about 75% of the study area.
The area significantly affected by waste
heat is more difficult to define. Physiological
and phenological data for C. lacustris and
T. latifolia documented adverse effects
from waste heat within 100 m of the wet
dike in 1977 and apparent subsidy effects
on Typha at distances up to 300 m from
the dike.
Rate of Change
In addition to rapid shifts in dominance
and diversity patterns in each plant
community, a continuing trend toward
decreased vegetative cover occurred
from 1974 through 1977. Open water
and exposed mudflats replaced the
previously closed and densely vegetated
perennial plant communities over an
increasing portion of the study area.
Annuals colonized some of the habitat
opened by the removal of perennial
species, but large areas remained unvege-
tated. By 1977, 19% of the quadrats
sampled in the area of major impact had
no rooted vegetation. Another 2% con-
tained only annual vegetation.
Relationships Between
Laboratory Investigations and
Field Observations
Laboratory and field experiments
began in 1977. Measurements were
made of alterations in plant phenology
and associated changes in amounts of
nonstructural carbohydrates stored in
underground organs of two species of
perennial plants. These measurements
were used to test hypotheses regarding
the causes of population decline and to
identify ecological criteria by which
significant impact on the plant communi-
ties could be recognized and monitored in
the field.
Physiological and phenological data from
laboratory and field experiments related the
response of T. latifolia and C. lacustris to
altered seasonal patterns of carbohydrate
storage and shoot phenology induced
primarily by groundwater temperatures
out of phase with normal cycles of plant
growth. In addition to establishing the
causal mechanism relating changes in
vegetation to leakage from the cooling
lake, this data set identified field signs of
heat stress and suggested general
species characteristics that may be
useful indicators of potential sensitivity to
stress.
Recommendations
The extreme sensitivity of Carex
lacustris appears to be a consequence of
its particular life history cycle and the
timing, as well as the magnitude, of the
disturbance. Characteristics of a species
that may serve as indicators of potential
sensitivity to stress include life span,
seasonal schedule of various phenophases, I
seasonal distribution of biomass between 1
above- and below-ground parts, ability to
-------
regenerate or colonize, and schedules of
reproduction and mortality. Physiological
and phenological data for both C. lacustris
and T. latifolia suggest that the likelihood
of predicting the response of dominant
species would be enhanced by knowing
the key features of their life cycles. It is
evident that,
• The concept of life history strategies
warrants further consideration as
a theoretical tool for predicting
species sensitivity to disturbance
and assessing probable effects on
wetland plant communities.
The 1974-1977 data address but
cannot answer questions regarding the
establishment of an equilibrium commun-
ity or recovery of the community following
disturbance. Temporal patterns in both
population and community data reveal
that these wetlands are still in a phase of
transient behavior. Further changes can
be expected. In terms of environmental
factors, the wetlands are only one-third
through a predicted transient. Water
levels have more or less stabilized, but
temperature changes are expected to
increase in both magnitude and spatial
extent after 1978 as the heat load of the
second generating unit is added to the
cooling lake. Results of population and
physiological data through 1977 indicate
that dominant perennial populations of
rhizomatous sedges and grasses will
continue to decline. As they do, the
structure of the community will continue
to change. In areas subject to minimal
temperature increases, tolerant hydrophy-
tic species such as T. latifolia probably
will become dominant. Areas subject to
maximum temperature increase may
support only annual vegetation or remain
as unvegetated muck or open water. For
better understanding of these trends,
• Research should continue on the
Columbia site in order to take
advantage of the long-term data
base and to monitor the eventual
outcome of a sustained perturbation
of wetland vegetation.
The development of effective tools for
assessment and management depends
on the identification and understanding
of critical system components and
interactions. Field observations on the
Columbia site indicate that increased
erodibility of the peat mass may be
associated with the transition in dominants
from perennial to annual and from
persistent to nonpersistent species.
Changes in numbers of species and
species diversity values appear to be less
relevant than differences in the life
histories of spatially dominant species.
The grouping of species with similar sets
of life history traits may identify meaningful
aggregate variables between the level of
the individual species and the community.
Therefore,
• Future research should address
the system-level consequences of
change in the vegetation of the
wetland, and
• Further research should be directed
toward identifying appropriate
analytical categories and the compo-
nents and processes of wetland
systems critical to assessment of
environmental impact.
Activities that alter physical inputs to
wetlands are likely to continue as energy
resources are developed and used. Better
tools and paradigms for predicting short-
and long-term effects of environmental
change on wetland plant communities
are more likely to come from demographic
studies than from existing theories of
succession or diversity in plant commu-
nities. This study showed no evidence of
community replacements predicted by
the succession model and no fixed
relationship between disturbance and
measures of diversity.
To improve the theoretical basis for
prediction, future studies should investigate:
• The population biology of important
wetland species including the
temporal sequence of events in the
life history, seasonal changes in
population age/size structure, and
requirements for germination.
• The relationships between charac-
teristics of a population and the
type, severity, and periodicity of
natural fluctuations in the physical
environment.
• The relationship of aspects of the
life history to the probability of
extinction and to the capacity for
regeneration or recolonization
under human-induced fluctuations
in the physical environment.
-------
Barbara Bedford and Orie Loucks are with Institute for Environmental Studies and
Water Resources Center, University of Wisconsin-Madison. Madison. Wl
53706.
Gary E. Glass is the EPA Project Officer (see below).
The complete report, entitled "Response of Carex-Dominated Wetlands to Altered
Temperature and Flooding Patterns: Wisconsin Power Plant Impact Study,"
(Order No. PB 84-198 944; Cost: $14.50. subject to change) will be available
only from:
National Technical Information Service
5285 Port Royal Road
Springfield. VA 22161
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
------- |