Muskrats ♦ 81
The role of muskrats (Ondatra zibethicus) as ecosystem engineers in
created freashwater marshes
Cheri Higgins and William J. Mitsch
Environmentl Science Graduate Program and School of Natural Resources
The Ohio State University
Introduction
The role of muskrat (Ondatra zibethicus) populations
in wetland succession is widely recognized (Danell,
1979; Kangas, 1985; Berg and Kangas 1989; Nyman
et al., 1993; Clark, 1994). These resident herbivores
influence hydrology, substrate properties, and ecosystem
biota through grazing, burrowing, and lodge construction
(Wilcox and Meeker, 1992; Taylor and Grace, 1995; Hewitt
and Miyonshi, 1997; and Gough and Grace, 1998; Connors
et al., 2000). During winter, muskrats use vegetation to build
lodges for shelter. These lodges are commonly constructed
from above ground portions of sturdy vegetative species
such as Typha spp. and Phragmites spp. Berg and Kangas
(1989) found removal of biomass to be as high as 20% of
the cover in a Czechoslovakian marsh.
Muskrat populations function as part of wetland
ecosystems and can have profound effects. Mitsch and
Gosselink (2000) illustrate that muskrats play an important
role in the development of spatial patterns in wetlands.
Wilcox and Meeker (1992) argued that foraging and
lodge building increased interspersion in dense vegetated
stands, causing a positive effect on invertebrates and
avian populations. Taylor et al. (1997) challenged the
assumption that herbivory reduces neighbor competition,
and concluded that intense small mammal herbivory has
neither positive nor negative effects on neighboring species.
In this study we intend to examine the consequences of
muskrat eat-outs and herbivory on vegetative regrowth
in wetlands.
Muskrat activity and its effect on ecosystem
development were compared, over two years (2000 and
2001) in planted and naturally colonized created wetlands.
A significant immigration of muskrats was beginning in
these two wetlands as the study began. The goal of this
study was to determine what effects the removal of large
amounts of plant biomass by muskrat eat-out would have
on macrophyte recovery, productivity, and macrophyte
species diversity in freshwater marshes. The objectives
to meet these goals were: (1) to compare macrophyte
biomass in eat-out areas and non-eat-out areas in the
growing season following winter lodge building, and (2)
to isolate the effect of herbivory during the study period
by building exclosure structures, and comparing exclosure
areas to open areas. Meeting these objectives permitted
the testing of the following hypotheses.
Hypothesis 1: Areas where muskrats remove most of the
vegetation for lodge building will show a greater short-term
decrease in macrophyte biomass than undisturbed areas.
Hypothesis 2: Muskrat herbivory will not significantly
affect macrophyte biomass in eat-out areas during the
growing season following lodge building.
Methods
Vegetation recovery after muskrat eat-out
In early January 2000, areas where muskrats had removed
vegetation for lodge building were outlined on site maps.
Missing vegetation and the presence of chewed-off shoots
made eat-out areas easy to identify. Total eat-out area
was calculated in wetlands 1 and 2 using a planimeter to
measure the individual eat-out areas sketched on the scaled
map. Plots were randomly chosen throughout the eat-out
with larger areas containing more plots and smaller areas
containing fewer plots. This method of plot selection
was done to equally represent all eat-out areas throughout
the two marshes; 47 eat-out plots resulted (Table 1). The
southeast corner of each plot was marked with PVC pipe
and orange flagging. Control areas, where vegetation had
not been eaten-out, were established in a paired order with
study plots, resulting in 94 total, eat-out and control plots
(Figure 1). Each eat-out plot was paired with a non eat-out,
or control, plot in a corresponding vegetation zone.
Biomass estimation
During spring 2000, the map with previously recorded
eat-out areas was updated to show vegetation that had been
removed for lodge upkeep and repair. On June 24-26, 2000
biomass data were recorded, using nondestructive methods,
in all plots. A square 0.25 m2-pipe assembly was used to
define the plot boundary, so that only vegetation inside the
square was counted and recorded. Number of each species,
stem length, number of leaves, number of flowers, and water
Table 1. Number of plots established in two experimental
marshes
_____________________________________________
____
Eat-out Number of plots
Total plots
area, m2 Eat-out Control Exclosure
_____________________________________________
____
W1
1007
19
19
10
48
82 ♦ The Olentangy River Wetland Research Park
2000 into dry weights. Equation 3, from a study by Ahn
and Mitsch (2002) was used to calculate the dry weight for
Schenoplectus tabernaemontani.
A nested ANOVA was created using SAS (1996) to
statistically compare the amount of biomass in planted
verses naturally colonizing wetland and in eat-out versus
non eat-out plots.
Muskrat herbivory effects
Figure 1. Scaled diagram of the two experimental
marshes with eat-out areas caused by muskrats in
spring 2000 outlined in black. Eat-out plots (white) and
control plots (black) are shown. Open plots are circular,
exclosure plots are rectangular.
depth were recorded for each plot.
On August 20, 2000 representative samples of Typha
spp. and Sparganium spp., the main plant species observed
in the wetlands, were harvested in 16 random areas
throughout the two basins. Wet weight, dry weight, stem
length, number of leaves, and number of flowers for each
individual harvested plant were recorded. Regression
equations were computed for the dry weight of Typha spp.
(Equation 1) and Sparganium eurycarpum (Equation 2)
from the harvested sample.
BT = 0.0453* SL -1.37, n = 10, r2 = 0.94
(Eqn. 1)
BS = 0.0466* SL - 6.93 n = 10, r2 = 0.65
(Eqn. 2)
Ba = 0.0165* CSL - 134.87 n = 10, r2 = 0.91
(Eqn. 3)
where,
BT = above ground biomass for Typha, g-DW m-2
BS = above ground biomass Sparganium, g-DW m-2
Ba = above ground biomass for Schenoplectus, gDW
m2
SL = (stem length/2) * number of leaves, cm
CSL = cumulative stem length (average stem length *
number of stems), cm
Regression equations were used to non-destructively
convert the vegetation data collected in June and July of
Twenty plots were established with temporary exclosures
surrounding each plot. Ten of the plots were located in eatout areas, and ten of the plots were located in corresponding
control areas. The purpose of the exclosure structures was to
keep muskrats out of the plots so that vegetation could grow
unaltered by future muskrat herbivory, thereby determining
what potential effects muskrat herbivory may have on the
recovery of vegetation in eat-out and control plots.
Stakes made of PVC pipe were used to define the four
corners of each 0.25 m2 plot (Figure 2). The perimeter
of each plot was trenched and plastic coated wire fencing
(openings = 38 cm2) was inserted into the marsh soil. The
fencing extended approximately 1.2 m above ground, and
8.0 cm below ground. Each plot was completely surrounded
by fencing, and secured with plastic fasteners. Openings
in the wire mesh surrounding the exclosure plots were 6.0
cm2, and not likely to inhibit plant growth.
Biomass
Number of each species, number of leaves, stem length
and number of flowers were recorded in each exclosure and
open plot. Regression equations for Typha, Schenoplectus,
and Sparganium were used to calculate dry weight for
vegetation in the exclosure plots (Equation 3.1, 3.2 and 3.3).
Exclosures were closely monitored for signs of intrusion
during the data collection period.
Figure 2. Diagram of muskrat exclosure
Muskrats ♦ 83
A nested ANOVA was created using SAS (1996) to
compare the amount of biomass in planted verses naturally
colonizing wetland and eat-out versus non eat-out plots, and
grazed versus not grazed plots.
Results
Vegetation Recovery after Muskrats
In the 2000 growing season following muskrat invasion,
biomass recovery was significantly less in eat-out areas than
in control areas for both wetland 1 and wetland 2 (Figure
3). The areas where muskrats removed large amounts of
vegetation, for lodge building during winter 1999-2000,
grew 87% less biomass on average than control areas in
summer 2000. Control areas, where muskrats removed
little to no vegetation for mound building, recovered similar
amounts of biomass relative to pre-invasion years at the
same marsh. The results of a nested ANOVA support the
findings that biomass in eat-out versus control plots within
wetlands is significantly different (r = 0.0039) (Table
2). As hypothesized, the data show a reduced recovery
of macrophyte biomass in the growing season following
intense muskrat eat-out.
Muskrat herbivory after invasion
Biomass of macrophytes on exclosure plots was not
different from that on open plots in July 2000 for either
wetland (Figure 4). Isolated muskrat herbivory, independent
of eat-out effects, had no effect on biomass in either
wetland.
The data from plots in wetland 1 revealed more biomass
in open plots than in exclosure plots for June, 2000.
Significantly more vegetation returned in open plots for
June, and is probably because of a greater number of open
plots (n=38) than exclosure plots (n=10) (the open plots
encompassed more area for possible plant regrowth). In
July, as more vegetation returned the data show no significant
difference between open and exclosure plots in wetland 1.
No significant difference was found between open verses
Figure 3. Above ground biomass in mid-summer (July
2000) eat-out and control plots forwetlands 1 and 2
Table 2. The p-Values of parameters and nested
parameters measured by running a nested ANOVA.
_____________________________________________
_____
Source
DF
F Value
Pr > F
_____________________________________________
_____
Wetland
1
16.72
0.0009
Eat-out
(wetland)
1
11.32
0.003
Exclosure (eat-out) 2
1.70
0.2136
Exclosure
2
1.06
0.3684
(wetland* eat-out)
exclosure plots in wetland 2. The results of a nested ANOVA
show no significant difference in the comparison of exclosure
verses open plots (Table 2).
Discussion
The decrease in biomass is probably a short-term response
to muskrat eat-outs, and vegetative productivity will likely
recover for these areas when muskrat eat-out pressure has
lessened [this indeed did happen in 2002]. The majority of
studies conducted on aquatic rodent eat-outs have been in
coastal areas where tidal effects cause erosion and eventually
marsh loss (Taylor et al. 1997; Grace and Ford, 1996; Carter
et al. 2000). This study is unique in its attempt to quantify
vegetative loss due to muskrat eat-outs in inland marshes,
and to separate the effect of subsequent muskrat herbivory
from the effect of muskrat eat-outs. Figure 5 shows number
of muskrat lodges at the ORWRP for all winters since
muskrats invaded the site in 1996, macrophyte biomass
for July 2000 in eatout and control plots, and macrophyte
coverage for each year. The number of muskrat lodges
increased significantly in winter 1999-2000, 3 years after
muskrat invasion. Following the large muskrat pulse in
1999-2000, macrophyte cover was reduced in wetland 2 from
66% in 1999 to 48.6% in 2000. In wetland 1, macrophyte
Figure 4. Above ground biomass in mid-summer (July
2000) in open and exclosure plots for wetlands 1 and 2
84 ♦ The Olentangy River Wetland Research Park
cover decreased from 63% in 1999 to 45.7% in 2000. More
specifically, emergent vegetation in eat-out areas showed
less re-growth than vegetation in control areas during the
growing season 2000 following muskrat invasion (Figure 5
middle). In August 2001, the second growing season after
continued muskrat dominance, macrophyte cover reduced
to 17.3% of the Typha-dominated Wetland 2 and 27.6% in
the more diverse Wetland 1 (see Mitsch and Zhang, 2002,
this report). This precipitous decrease in vegetation in 2001,
essentially an “eatout” in terms of vegetation loss led to a
crash in the muskrat activity in the basins in the winter of
2001-02 (Figure 5 top) in both basins.
Muskrat Role
Effects of muskrat engineering do not end at decreased
biomass. Svenson (1995) reported that Sphagnum fuscum
acted like a small-scale engineer by capturing mineral
nutrients quickly, and limiting Drosera rotundifolia growth
in a Swedish bog. Svensonʼs (1995) results support the
idea that muskrat eat-out activity positively or negatively
influence other organisms in a marsh by changing the
abundance and state of marsh vegetation.
Muskrats in this study cleared dense Typha stands in a
naturally colonizing Wetland 2. The highly aggressive Typha
achieves dominance in many naturally colonizing wetlands
in the United States. Communities of dominant macrophytes
serve many direct and indirect functions: provide food for
animals, compete for resources, provide shelter for other
species, capture food, cast shade, reduce the impact of rain,
and oxygenate soil (Svenson, 1995).
Though muskrat herbivory in this study was not correlated
with an increase in emergent macrophyte species richness
several other studies have noted such relationships (Nyman
et al., 1997). Wilby et al. (2001) depict many interactions
between herbivores and plant communities occurring
simultaneously, and that ecosystem engineering and
tropic processes can be closely associated, resulting from
single interactions of herbivores. Through feeding and
digging, muskrats create open areas for seed and resource
accumulation, allowing invasive species establishment.
References
Ahn, C. and Mitsch, W. J., 2002. Scaling considerations
of mesocosm wetlands in simulating full-scale
wetlands. Ecological Engineering 18: 327-342
Berg, K.M. and Kangas, P., 1989. Effects of Muskrat
mounds on decomposition in a wetland ecosystem.
Freshwater Wetlands and Wildlife 61: 255-264.
Carter, J., Foote, A. L., and Johnson-Randall, L. A., 1999.
Modeling the effects of nutria (Myocastor coypus) on
wetland loss. Wetlands 9: 209-219
Clark, W.R., 1994. Habitat selection by muskrats
in experimental marshes undergoing succession.
Canadian Journal of Zoology 72: 675-680.
Connors, L.M., Kiviat, E., Groffman, P.M., and Ostfeld,
Figure 5. top: number of mustrat lodges over time, 19942002; middle: biomass in eat-out and control areas in
July 2000 for wetland 1 and 2, and bottom: macrophyte
coverage in both experimental basins, 1994-2001
R.S., 2000. Muskrat (Ondatra zibethicus) disturbance
to vegetation and potential net nitrogen mineralization
and nitrification rates in a freshwater tidal marsh.
American Midland Naturalist 143: 53-63
Danell, K., 1979. Reduction of aquatic vegetation
following the colonization of a northern Swedish lake
by the muskrat, Ondatra zibethica. Oecologia 38:
101-106.
Gough, L. and Grace, J. B., 1998. Herbivore effects on
plant species density at varying productivity levels.
Ecology 79: 1586-1594.
Hewitt, N. and Miyanishi, K., 1997. The role of
mammals in maintaining plant species richness in a
floating Typha marsh in southern Ontario. Biodiversity
and Conservation 6: 1085-1102.
Jones, C. G., Lawton, J. H., and Shachak, M., 1994.
Organisms as ecosystem engineers. Oikos 69: 373-386
Kangas, P. and Hannan, G. L., 1985. Vegetation on
muskrat mounds in a Michigan marsh. American
Midland Naturalist 113: 392-396.
Kangas, P., 1988. A Chess Analogy: Teaching the role
of animals in ecosystems. The American Biology
Teacher 50: 160-162.
Metzker, K. D. and Mitsch, W. J., 1997. Modeling self-
Muskrats ♦ 85
design of the aquatic community in a newly created
freshwater wetland. Ecological Modelling 100: 61-86.
fascum and Drosera rotundifolia: a case of ecosystem
engineering. Oikos 74: 205-212
Mitsch, W.J., X. Wu, R.W. Nairn, P.E. Weihe, N. Wang,
R. Deal and C.E. Boucher. 1998. Creating and
restoring wetlands – an ecosystem experiment in selfdesign. BioScience 48: 1019-1030.
Taylor, K. L. and Grace, J. B., 1995. The effects of
vertebrate herbivory on plant community structure in
the costal marshes of the Pearl River, Louisiana, USA.
Wetlands 15: 68-73.
Mitsch, W. J. and Gosselink, J.G., 2000. Wetlands, 3rd
ed. John Wiley and Sons, New York.
Wilby, A., Shachak, M., and Boeken, B., 2001.
Integration of ecosystem engineering and trophic
effects of herbivores. Oikos 92: 436-444.
Mitsch, W.J. and L. Zhang. 2002. Plant community
development after eight growing seasons in the two
experimental wetland basins. In: Mitsch, W.J. and L.
Zhang, eds. Olentangy River Wetland Research Park
Annual Report 2001, Columbus, OH, pp. 35-42.
Nyman, J.A., Chabreck, R.H., Kinler, N.W., 1993. Some
effects of herbivory and 30 years of weir management
on emergent vegetation in brackish marsh. Wetlands
13: 165-175.
Svensson, B. M., 1995. Competition between Sphagnum
Wilcox, D. A. and Meeker, J. E. 1992. Implications for
faunal habitat related to altered macrophyte structure
in regulated lakes in northern Minnesota. Wetlands 12:
192-203.
Wilson, S. D., and Keddy, P. A., 1991. Competition,
survivorship and growth in macrophyte communities.
Freshwater Biology 25: 331-337.
86 ♦ The Olentangy River Wetland Research Park