Content uploaded by Richard Ostfeld
Author content
All content in this area was uploaded by Richard Ostfeld
Content may be subject to copyright.
53
Am. Midl. Nat. 143:53–63
Muskrat (Ondatra zibethicus) Disturbance to Vegetation and
Potential Net Nitrogen Mineralization and Nitrification Rates
in a Freshwater Tidal Marsh
LISA M. CONNORS
1
Graduate School of Environmental Studies, Bard College, P.O. Box 5000, Annandale, New York 12504
ERIK KIVIAT
Hudsonia Ltd., Bard College, P.O. Box 5000, Annandale, New York 12504
AND
PETER M. GROFFMAN and RICHARD S. OSTFELD
Institute of Ecosystem Studies, Box AB (Route 44A), Millbrook, New York 12545
A
BSTRACT
.—The muskrat (
Ondatra zibethicus
) is a wetland mammal whose disturbance
activities include grazing, burrowing and lodge construction. We evaluated the effects of
these disturbances on plant biomass, species richness and diversity, stem density and potential
net nitrogen mineralization and nitrification rates in a freshwater tidal marsh on the Hudson
River in New York. We hypothesized that muskrats increase floristic richness and diversity by
decreasing the biomass of narrowleaf cattail (
Typha angustifolia
) and that muskrats increase
potential net nitrogen mineralization and nitrification rates through aeration and reduced
plant uptake because of herbivory. Because muskrats commonly build lodges on or close to
creek banks, we separated the disturbance effects of muskrats from the disturbance effects
of the creek bank by sampling quadrats along transects placed perpendicular to creek banks
at lodge sites. Muskrats decreased biomass, particularly of cattail, but had no measurable
effect on stem count, species richness or species diversity. Muskrats increased potential net
nitrogen mineralization and nitrification rates; however, this effect was limited to active sites.
Creek bank disturbance increased stem count but had no effect on the other variables.
Although muskrats did not significantly affect floristic diversity in this study, their disturbance
activities did influence soil nitrogen dynamics, which is an important component of wetland
function.
I
NTRODUCTION
Disturbance influences all levels of ecological organization (Connell, 1978; Pickett and
White, 1985; Reice
et al.,
1990). Burrowing animals, as agents of disturbance, can profound-
ly affect their environment, principally through increases in soil aeration and water infil-
tration, and by altering development of plant and animal communities (Meadows and Mead-
ows, 1991; Hansell, 1993; Butler, 1995; Johnston, 1995). Much research has been conducted
on the impact of invertebrate burrows in marine and terrestrial ecosystems (Pickett and
White, 1985; Meadows and Meadows, 1991) and mammal burrows in terrestrial ecosystems
(Huntly and Inouye, 1988; Swihart, 1991; Swihart and Picone, 1991; English and Bowers,
1994), but little study of mammal burrows in wetland ecosystems has been conducted. The
muskrat (
Ondatra zibethicus
) is a common wetland mammal that influences soil, plant and
animal communities through grazing, burrowing and lodge construction, and possibly in-
creases the diversity of its habitat (Danell, 1977; Kangas, 1985; Berg and Kangas, 1989;
Nyman
et al.,
1993). We expected narrowleaf cattail (
Typha angustifolia
) marsh ecosystems
1
Present address: 3 Lexington Avenue, Lexington, Virginia 24450
54
143(1)
T
HE
A
MERICAN
M
IDLAND
N
ATURALIST
to be influenced by the muskrat at many levels, but this has been quantified only in a
limited manner. Thus, understanding ecological dynamics in freshwater wetlands may re-
quire quantifying the role of the muskrat. In this study we examined the effects of muskrat
disturbance on soil and vegetation in a freshwater tidal marsh on the Hudson River at
Tivoli, New York.
Muskrat lodges are conical or irregular mounds of plant material 0.3–1.2 m above water
and 1–2 m in diameter with internal chambers used for shelter from weather, flooding and
predators (Alexander, 1956). Burrows are tunnels 13–15 cm in diameter and up to 13 m
long that have underwater entrances (Rezendes, 1992). Burrows below the water table often
provide access to lodges whereas burrows above the water table serve as a substitute for
lodges. Runways are trails of trampled vegetation and soil, and canals are underwater trails
created for efficient swimming. Feeding platforms are small piles of plant material on which
muskrats eat, groom and sleep in the open. Push-ups are structures intermediate between
feeding platforms and lodges and serve as covered feeding stations in winter. Young are
reared in lodges and bank burrows as well as in open nests. In freshwater tidal marshes
muskrats typically build all the structures described (Smith, 1938).
Tidal marshes, including freshwater tidal marshes such as Tivoli North Bay, have a char-
acteristic plant zonation pattern along the creek bank (Odum
et al.,
1978; Mitsch and
Gosselink, 1993). This zonation pattern is a result of many environmental influences in-
cluding substrate elevation and duration of tidal inundation, shade, soil chemistry and ice
scouring. A collection of environmental influences causing a gradual spatial change in veg-
etation can be considered an
environmental gradient
(Gauch, 1982). In tidal marshes one
key environmental gradient is the lateral gradient of vegetational diversity with respect to
distance from the tidal creek, the creek-bank effect (Fig. 1a).
In contrast to the creek-bank effect, a muskrat lodge is potentially a point disturbance
that may produce a radial gradient of biotic diversity around the lodge location (Fig. 1b).
Because muskrats typically build lodges near banks, this point disturbance is commonly
superimposed on the edge disturbance of the creek-bank effect. To quantify any disturbance
gradient of the muskrat it is necessary to distinguish effects of muskrats from the environ-
mental gradient of the creek bank.
We quantified the effects of muskrat disturbance on the marsh ecosystem by measuring
plant species richness, plant species diversity, aboveground total biomass, stem density and
potential net nitrogen mineralization and nitrification rates of soil. These measurements
encompass interactions spanning three trophic levels (herbivore, vegetation and nutrient
remineralizers) and are of value because most studies generally address interactions within
one trophic level. Narrowleaf cattail is the dominant species in this marsh and is favored
by muskrats for food and lodge building. Therefore, we expected muskrats to decrease
Typha
biomass and, hence, total biomass. Consequently, we hypothesized that muskrat ac-
tivities increase vegetational species richness and diversity due to the decrease in the dom-
inant species. We also hypothesized that muskrats increase potential net nitrogen mineral-
ization rates by (1) increasing aeration of the soil, which increases microbial activity and
(2) reducing plant uptake of nitrogen through herbivory. With increased microbial activity
and reduced plant uptake of nitrogen there would be more ammonium available in the
soil for transformation to nitrite. Therefore, we further hypothesized that muskrat activity
also increases nitrification rates, as there would be more nitrite available for the transfor-
mation of nitrite to nitrate.
M
ETHODS
Tivoli North Bay, a component of the Hudson River National Estuarine Research Reserve,
is a ca. 150-ha freshwater tidal marsh on the east bank of the Hudson River in New York
2000
55C
ONNORS ET AL
.: M
USKRAT
D
ISTURBANCE
F
IG
. 1.—Diagrammatic representations of: a. Environmental gradient due to the edge disturbance of
a tidal creek in a freshwater tidal marsh and b. Disturbance gradient due to muskrat activities along
the bank of a freshwater tidal marsh. The marsh is completely covered at high tide
State (738559300W, 42829550Nto73855900W, 42829300N). Two alternating low and high tides
occur every 25 h. Water levels fluctuate approximately 1.2 m vertically, with water covering
the entire marsh at high tide (E. Kiviat, pers. obs.).
In Tivoli North Bay muskrats build lodges and feeding stations primarily from narrowleaf
cattail, usually in stands of cattail or cattail mixed with other plants. Lodges are built or
rebuilt in late summer or fall and often destroyed in Feb. or Mar. by floods (E. Kiviat, pers.
obs.). Consequently, lodges typically last only 6–18 mo. All lodges sampled were located on
creek banks, not in the creek itself, and were inundated at high tide. At low tide lodges are
completely drained.
We sampled three active (A, B, C) and three abandoned (X, Y, Z) muskrat lodge sites
56
143(1)
T
HE
A
MERICAN
M
IDLAND
N
ATURALIST
F
IG
. 2.—Two sampling designs were used so that the bank effect and muskrat effect could be ade-
quately represented. The sampling design was fixed with respect to distance from bank (at low tide
level). Lodges near the creek bank (B, C, X) lie between quadrats 1 and 2, whereas lodges farther
from the bank (A, Y, Z) lie between quadrats 2 and 3. Control5reference in text
along creek banks in the fall of 1994. Active lodges were built the previous fall and were
approximately 10–12 mo old. Abandoned lodge sites had a flattened mound of detritus
from the lodge. These lodges were probably 12–18 mo old. Similar areas of apparently
older muskrat activity existed, but it became difficult to verify that the area was a lodge
without knowing its history, thus areas such as these were not included in the study.
We sampled two transects at each site (Fig. 2). Sample units were five 0.5 30.5 m square
quadrats (de la Cruz, 1978) spaced 2 m apart on-center along the transects. Transect one
(T1) intersected the lodge perpendicular to the creek bank and represented muskrat dis-
turbance combined with the creek-bank effect. We placed quadrats on both sides of the
lodge, each 1 m from lodge center along the transect. Lodges varied slightly in distance
from the bank. Lodges close to the bank had one quadrat between the lodge and creek
and four quadrats extending from the lodge into the intercreek marsh. Lodges inland from
the bank had two quadrats between the lodge and the creek and three quadrats extending
from the lodge into the intercreek marsh.
Transect two (T2) was placed parallel to, and 8 m from, transect one and represented a
reference with respect to muskrat disturbance, sampling only the bank effect (Fig. 2). Ob-
servations of muskrat activity (scat, tracks and cuttings), along with preliminary sampling
of two lodges, showed current muskrat activity limited to a zone within 4 to 6 m of the
lodge. Placement of the reference transect to the right or left of the lodge transect was
randomized when possible. Three sites with lodges close to the junction of two tidal creeks
2000
57C
ONNORS ET AL
.: M
USKRAT
D
ISTURBANCE
did not have space for a randomly placed reference transect. In these cases we placed the
reference transect away from the channel junction.
Vegetation was sampled from 29 Aug. through 26 Sept. 1994. In each quadrat we iden-
tified plants, counted stems and clipped them at ground level. Samples were dried at 93 C
to constant weight and weighed to the nearest 0.1 g by species. Plants are senescing in Sept.,
but we found no correlation between sampling date and biomass.
Soil samples (15 cm deep and 15 cm across) were taken 27 and 28 Sept. 1994 from the
edge of each quadrat and stored in sealed plastic bags at room temperature. Within 2 d of
collection percent moisture and potential net nitrogen mineralization and nitrification rates
were quantified as the amounts of ammonium-nitrogen plus nitrate-nitrogen (NH
4
1
and
NO
3
2
, respectively) and nitrate alone, produced during a 10 d incubation at 25 C and
ambient moisture content (Duncan and Groffman, 1994). Ammonium and nitrate were
measured with an Alpkem TFS 300 rapid flow analyzer.
We performed a separate three-way factorial analysis of variance (ANOVA), setting a5
0.0071 (
see
discussion of multiplicity below), and examined all two-way interactions for each
of seven dependent variables: stem count, species richness, species diversity, total plant
biomass,
Typha
biomass and potential net nitrogen mineralization and nitrification. Species
diversity was evaluated by computing a Shannon index of diversity based on stem count for
each quadrat along each transect. The independent variables were quadrat (distance from
the bank), transect (lodge or reference) and age of lodge (active or abandoned). We treated
all independent variables as discrete fixed variables (Bennington and Thayne, 1994). Any
statistically significant difference among quadrats should represent the bank-imposed gra-
dient, whereas any difference among transects should represent muskrat activity. The seven
ANOVAs present a potential multiplicity effect. To control this Type I error we used the
Bonferroni correction (Samuels, 1989) and therefore adjusted ato a50.05/7 50.0071.
We analyzed all data using Data Desk (Velleman, 1988).
All data except the Shannon diversity index violated the ANOVA assumption of normality.
Stem count and species richness were transformed to log
e
(x 11) before analysis, and total
plant biomass,
Typha
biomass and potential net nitrogen mineralization and nitrification
were transformed to square root (x 10.5) to reduce deviations from normality.
We measured the height of water above ground surface at high tide (relative elevation)
of each quadrat to determine any within-site variation in elevation between the muskrat
and reference transects. With our ANOVA design, any difference among transects should
represent muskrat activity. However, we could not attribute this difference to muskrats if
elevations between transects were not equal. A paired
t
-test comparing each matched quad-
rat along the two transects showed no difference (P .0.05) in relative elevation between
matched quadrats at each site. Therefore, differences between T1 and T2 at each site were
not due to elevation. We also normalized the relative elevation measurements to a water
depth sensor in Tivoli South Bay (Paul Barten, School of Forestr y and Environmental Stud-
ies, Yale University, pers. comm.) to compare elevations among sites.
R
ESULTS
Muskrat activity had a highly significant effect on total biomass (F 513.24, df 51, P 5
0.0007) and
Typha
biomass (F 526.10, df 51, P ,0.0001), both being significantly lower
in transects through muskrat lodges than in reference transects (Table 1). The effect of
distance from the bank was not significant for total biomass, but there was a significant
increase in
Typha
biomass with distance from the bank (F 510.84, df 54, P ,0.001). Age
of lodge was also a significant determinant of total biomass (F 512.67, df 51, P 50.0009)
and
Typha
biomass (F 515.95, df 51, P 50.0002), with greater biomass in active sites
58
143(1)
T
HE
A
MERICAN
M
IDLAND
N
ATURALIST
T
ABLE
1.—Summary of P-values from 3-way ANOVAs for each dependent variable. Significant values
are in bold type (P ,0.0071 with Bonferroni correction)
Inde-
pendent
variable
Log
stem
count
Log
species
richness
Square
root
biomass
Square root
Typha
biomass
Square root
mineral-
ization
Square
root
nitrification
Species
diversity
Shannon, H9
Transect 0.3204 0.0899 0.0007 #0.0001 0.9686 0.6438 0.4721
Quadrat 0.0010 0.0118 0.0093 #0.0001 0.9290 0.0130 0.0927
Age 0.8720 0.2747 0.0009 0.0002 0.0001 #0.0001 0.2972
Trt∗Qdr 0.3284 0.6411 0.3877 0.4371 0.7529 0.5558 0.4024
Trt∗Age 0.2353 0.1191 0.9311 0.3389 0.1153 0.1654 0.0793
Qdr∗Age 0.2565 0.0046 0.3134 0.3831 0.1042 0.3520 0.0117
T
ABLE
2.—Raw data summary showing means-standard errors for each dependent variable. Each
transect, quadrat and whether active or abandoned is shown separately for efficient comparison of the
raw data
Independent
variable
Stem
count
(per
0.25 m
2
)
Species
rich-
ness
(per
0.25 m
2
) Biomass (g)
Typha
biomass (g)
Mineralization
(mgNg
2
1
d
2
1
)
Nitrification
(mgNg
2
1
d
2
1
)
Relative
elevation (m)
Transect 1 39/41 3/1 123.9/141.4 65.2/100.0 9.42/6.09 0.45/0.56 0.31/0.12
Transect 2 39/43 4/2 201.1/116.1 149.0/124.4 10.05/9.45 0.55/0.86 0.31/0.17
Quadrat 1 89/69 5/1 110.7/60.8 40.4/48.2 10.74/12.37 0.84/1.10 0.22/0.05
Quadrat 2 28/19 3/1 149.8/179.0 63.5/98.8 8.91/6.09 0.72/0.66 0.33/0.16
Quadrat 3 20/16 3/1 108.1/113.9 74.0/109.4 10.89/0.47 0.47/0.47 0.34/0.17
Quadrat 4 32/19 3/2 239.4/168.2 196.9/170.6 8.66/5.92 0.14/0.22 0.33/0.15
Quadrat 5 27/11 3/1 204.5/68.0 160.9/59.1 9.50/7.25 0.33/0.73 0.33/0.16
Active sites 36/31 3/2 214.8/151.2 147.6/139.8 13.74/8.76 0.86/0.83 0.33/0.20
Abandoned 41/51 4/1 110.2/89.6 66.7/78.6 5.74/4.10 0.14/0.32 0.29/0.05
(Table 1). For ease in interpreting the statistical results, please refer to Table 2 which gives
means and standard errors of the raw data.
We identified 13 plant species in the sampling units using nomenclature from Gleason
and Cronquist (1991):
Bidens
sp.,
Echinochloa walteri
(Pursh) Heller,
Iris pseudacorus
L. or
I. versicolor
L.,
Impatiens capensis
Meerb.,
Leersia oryzoides
(L.) Swartz,
Lythrum salicaria
L.,
Peltandra virginica
(L.) Schott & Endl.,
Pilea pumila
(L.) A. Gray,
Polygonum punctatum
Elliott,
Sagittaria latifolia
Willd.,
Scirpus fluviatilis
(Torr.) A. Gray,
Sparganium eurycarpum
Engelm. and
Typha angustifolia
L. The number of species per lodge site ranged from 5 to
10. There was no significant difference in the number of species between transects with
muskrat activity and reference transects (F 53.01, df 51, P 50.09; Table 1). With the
Bonferroni correction the effect of distance from the bank on species richness also was not
significant (F 53.65, df 54, P 50.01), however there was a significant interaction between
age of lodge and distance from the bank affecting species richness (F 54.37, df 54, P 5
0.005).
Neither muskrat activity nor bank effect had a significant effect on species diversity (H9).
The interaction between quadrat and age (F 53.66, df 54, P 50.01), while not significant
2000
59C
ONNORS ET AL
.: M
USKRAT
D
ISTURBANCE
with the Bonferroni correction, suggests a gradient in diversity in either occupied or aban-
doned lodges, but not both.
Stem count was quite variable in Quadrat 1, but significantly higher than other quadrats
(F 55.58, df 54, P 50.001). Muskrat activity did not affect stem count, as there was no
difference between transects (F 51.01, df 51, P 50.320) or age (F 50.03, df 51, P 5
0.872).
There was no difference in potential net nitrogen mineralization or nitrification rates
between muskrat and reference transects or among quadrats. However, potential net nitro-
gen mineralization and nitrification rates were significantly higher in soil at active than
abandoned lodges (Table 1). Because both the transect factor (T1 vs. T2) and the age
factor (current vs. former muskrat activity) represent contrasts between presence and ab-
sence of muskrats, but gave conflicting results, we explored whether the significant age
effect may have been caused by differences in elevation between sites. However, differences
in elevation did not cause the difference in mineralizaton and nitrification between the
active and abandoned transects.
D
ISCUSSION
As we expected, muskrat activity reduced total plant biomass and
Typha
biomass. The
reduction was seen in both active and abandoned sites and was greatest closest to the lodge.
Typha
is the dominant plant in this freshwater marsh, so we expected a reduction of its
biomass to provide space for competitively inferior, or secondar y, species thereby increasing
total stem count and plant species richness and diversity. Muskrat effects on stem count and
plant species richness and diversity, however, were not statistically significant. Muskrat re-
duction of cattail biomass was concentrated near the lodge (quadrats two and three). Musk-
rats feed on aboveground stalks and often pull the rhizomes out for food. This disturbance
activity creates open spaces around the lodge, but there was no significant increase in
species richness as a result of these open spaces. The environmental conditions around the
lodge may not be as favorable to secondary species as the conditions on the bank despite
decreased competition.
We observed a significant interaction between quadrat and age of site affecting species
richness; yet, neither age (hence muskrat presence) nor distance from the bank separately
affected species richness. There was no gradient of increasing or decreasing species richness
along the quadrats when analyzing the total data set. However, when we separated the data
set into active and abandoned sites the active sites had higher species richness at quadrats
one, two and three, which are the quadrats closest to the bank and to the lodges on transect
one. Therefore, in active sites, the muskrats did significantly increase species richness but
this was only measurable in combination with the bank effect. Nyman
et al.
(1993) found
that plant species richness and muskrat activity were positively correlated in a brackish
Louisiana marsh.
Although muskrat effects on
Typha
biomass were seen in both active and abandoned
sites,
Typha
biomass was higher overall in the active sites. If conditions at the sites disturbed
by muskrats are not suitable for secondary species to colonize (as postulated above), it seems
logical that
Typha
biomass should be higher in abandoned sites that are no longer being
denuded by muskrats. The decaying detrital mat at the abandoned sites may inhibit re-
growth for a time. For example, site Z was located along an interior inlet which had a low
tidal flushing force compared to other areas in the marsh. This site also had the largest
detrital mat and the lowest amount of regrowth in the disturbed area. Another possible
contribution to decreased
Typha
growth in abandoned sites is that disturbance by cutting
60
143(1)
T
HE
A
MERICAN
M
IDLAND
N
ATURALIST
cattail shoots decreases its productivity in the following growing season. Jordan and Whigh-
am (1988) found that cutting or bending dead cattail shoots lowered aboveground pro-
duction and flowering in the following growing season because cutting and bending re-
sulted in lower oxygen concentrations in the rhizomes.
We found no significant results for muskrat or bank effect on species diversity (H9). It
should be noted, however, that without our conservative approach with the Bonferroni
correction there would have been a significant interaction between quadrat and age of site,
as explained above, for species richness. The interaction between quadrat and age for spe-
cies diversity is interesting given the significant interaction between quadrat and age for
species richness, and is worthy of further attention.
Although muskrats reduced the cattail biomass in Tivoli North Bay, we did not observe
a concomitant increase in other plant species stem count, richness or diversity. Perhaps the
density of muskrats, and hence muskrat disturbance, was not substantial enough to cause a
shift from the dominant cattail due to the short time lodges exist in this tidal marsh. A
follow-up study documenting revegetation of more lodge sites over a greater time period
would determine the long-term biotic response to muskrat disturbance in a freshwater tidal
marsh.
We expected muskrats to increase potential net nitrogen mineralization and nitrification,
yet there was no difference between transect one (muskrat) and two (reference). However,
age of lodges (which tested current versus old muskrat activity) significantly affected min-
eralization and nitrification rates suggesting muskrat impact. The two major differences
between sites were elevation and age (presence or past presence of muskrats). We then
considered whether elevation was confounding the age effect on soil mineralization and
nitrification rates. For example, lower sites would be less aerobic and would have lower rates
of potential net nitrogen mineralization and nitrification. Sites A, B and Z were located in
secondary stream channels and were lower in elevation than C, X and Y. Therefore, we
expected sites A, B and Z to have lower potential net nitrogen mineralization and nitrifi-
cation rates. However, we found no correlation between elevation and potential net nitro-
gen mineralization and nitrification rates in these sites. The active sites (A, B, C) had the
highest potential net nitrogen mineralization and nitrification rates. Scatterplots of miner-
alization vs. nitrification show a better correlation in active sites than in abandoned sites.
This correlation was seen in both transect one and two, which was why there was no sig-
nificant difference for the transect factor in the ANOVA.
Although we saw no visible sign of muskrat activity beyond 8 m, the lack of a significant
difference in the transect factor suggests to us that transect two was inadequate as a refer-
ence for soil nitrogen mineralization and nitrification rates. Errington (1961) found musk-
rats foraged within 45 to 92 m of their lodge or burrow. We suspect that the muskrats in
these Tivoli North Bay lodges were active beyond 8 m but that their activity was obscured
by tidal action.
Several types of muskrat disturbance activities could increase the nitrogen mineraliza-
tion and nitrification rates of soil such as reduced plant uptake due to herbivory, lodge
decomposition, fecal accumulation and aeration of the soil. The first two activities listed
are not likely to cause the significant effect seen with mineralization and nitrification
rates because the effect is only seen in active lodge sites. For example, reduced plant
uptake is unlikely because
Typha
biomass was higher in active lodge sites. We expected
areas where
Typha
was denuded to have higher mineralization and nitrification rates
because the nitrogen was not being taken up by the plants, but abandoned lodge sites
had lower mineralization and nitrification rates. Likewise, lodge decomposition is not the
source of the difference in mineralization and nitrification rates between sites because
2000
61C
ONNORS ET AL
.: M
USKRAT
D
ISTURBANCE
abandoned lodges had lower rates. Fecal accumulation is also unlikely because the sites
are inundated and drained twice a day. This tidal action would flush away or at least move
around any fecal pellets, thereby eliminating fecal pellets as a source of increased min-
eralization and nitrification rates.
We hypothesized that aeration of the soil may increase potential net nitrogen minerali-
zation and nitrification rates of soil. Muskrats aerate the soil through burrowing and tread-
ing activities. It is possible that the muskrat burrow system is extensive enough to influence
aeration of soils at 8 m. Tivoli North Bay muskrat burrows can persist for many years after
abandonment of a lodge site (E. Kiviat, pers. obs.). However, if these burrows were the
source of aeration and increased mineralization and nitrification rates, then abandoned
sites would have had rates as high as active sites. Thus burrows are not important in mod-
ification of the mineralization and nitrification rates of soils.
It is important to remember that plants are important in the sediment aeration in tidal
freshwater wetlands because live plant roots leak oxygen and other oxidized substances
(Howes
et al.,
1981). The more aerobic soil conditions stimulate nitrogen mineralization
and nitrification (Bowden, 1987). The lower rates of nitrogen mineralization and nitrifi-
cation in the abandoned lodge sites could be associated with the lower plant biomass on
these sites. However, the ANOVA results did not support this conclusion. In both active and
abandoned lodge sites biomass was significantly higher in the reference transect than in
the muskrat transect, yet there was no difference in nitrogen mineralization and nitrification
rates between the two transects.
Muskrat treading activities may create small aerobic zones that stimulate mineralization
and allow nitrification. The muskrat is a lightweight animal (about 1 kg) whose tracks do
not compact the saturated wetland soils found in Tivoli North Bay. The soil probably re-
distributes with each step, and tracks allow air and water to infiltrate. Since treading is the
only disturbance activity that solely affects active sites in this study we interpret the increased
potential net nitrogen mineralization and nitrification rates of soil in the active sites as a
temporary effect related to active muskrat disturbance.
This study showed that muskrat disturbance effects in a freshwater tidal wetland increase
potential net nitrogen mineralization and nitrification rates of soil, most likely through soil
aeration. Most wetland sediments have low nitrite and nitrate concentrations due to low
oxygen tensions that favor denitrification (Bowden, 1987). Denitrification is important as a
potential sink for excess nitrate, which can cause eutrophication in marine ecosystems
(Ryther and Dunstan, 1971). Our data indicate that muskrats can increase nitrate levels, at
least in small areas of the marsh. This increase may foster the occurrence of certain plant
species that favor nitrate and may increase nitrate losses from the marsh.
A number of rodent species are known to have dramatic effects on the availability of
resources to other species. For example English and Bowers (1994) found that woodchuck
(
Marmota monax
) activity increases plant species richness at intermediate distances from
burrows and decreases plant species richness both near and distant from burrows. Pocket
gopher activity increases the heterogeneity of the soil surface by creating patches of high
and low soil fertility ( Johnston, 1995). Schiffman (1994) found giant kangaroo rat (
Dipo-
domys ingens
) activity increases the diversity of exotic grass species while decreasing the
diversity of native species in a California grassland and the North American beaver (
Castor
canadensis
) is notable for its extensive alteration of resources (Naiman, 1988). Our study
establishes the muskrat as modifier of potential net nitrogen mineralization and nitrification
rates of soil and previous studies (Danell, 1977; Kiviat, 1978; Weller, 1987; Nyman
et al.,
1993) suggest that the muskrat affects diversity of both vegetation and animals. Therefore,
the impacts of muskrats are likely to influence wetland restoration and management.
62
143(1)
T
HE
A
MERICAN
M
IDLAND
N
ATURALIST
Acknowledgments.
—Plant species verification was kindly provided by Jerry Jenkins. We thank Pablo
Inchausti for reading a draft of this paper, Jane Wolfson for help with data analysis, Gay Hanson for
assistance with soil analysis and Christopher Connors for technical assistance. The Shannon diversity
analysis was kindly provided by Felicia Keesing. We thank J. Pastor and three anonymous reviewers
whose comments substantially improved the clarity of the paper. This paper is part of the first author’s
Master’s Thesis from Bard College and is Bard College Field Station—Hudsonia contribution 71.
L
ITERATURE
C
ITED
A
LEXANDER
, M. M. 1956. The muskrat in New York State. State University of New York College of
Forestry, Syracuse, N.Y. 15 p.
A
LPKEM
C
ORPORATION
. 1986. Operator’s manual and methodologies for the RFA-300. Alpkem Corp.,
Clackamas, Oreg.
B
ENNINGTON
,C.C.
AND
W. V. T
HAYNE
. 1994. Use and misuse of mixed model analysis of variance in
ecological studies.
Ecology,
75:717–722.
B
ERG
,K.M.
AND
P. C. K
ANGAS
. 1989. Effects of muskrat mounds on decomposition in a wetland eco-
system, p. 145–151.
In
: Sharitz, R. R. and J. W. Gibbons (eds.). Freshwater Wetlands and Wild-
life, U.S. Department of Energy Symposium Series 61.
B
OWDEN
, W. B. 1987. The biogeochemistry of nitrogen in freshwater wetlands.
Biogeochemistry,
4:313–
348.
B
UTLER
, D. R. 1995. Zoogeomorphology: animals as geomorphic agents. Cambridge University Press,
New York, N.Y. 231 p.
C
ONNELL
, J. H. 1978. Diversity in the tropical rain forests and coral reefs.
Science,
199:1302–1310.
D
ANELL
, K. 1977. Short-term plant successions following the colonization of a northern Swedish lake
by the muskrat,
Ondatra zibethica. J. Appl. Ecol.,
14:933–947.
DE LA
C
RUZ
, A. A. 1978. Primary production processes: summary and recommendations. p. 79–86.
In
:
Good, R. E., D. F. Whigham and R. L. Simpson (eds.). Freshwater wetlands: ecological pro-
cesses and management potential. Academic Press, New York, N.Y.
D
UNCAN
,C.P.
AND
P. M. G
ROFFMAN
. 1994. Comparing microbial parameters in natural and constructed
wetlands.
J. Envir. Qual.,
23:298–305.
E
NGLISH
,E.I.
AND
M. A. B
OWERS
. 1994. Vegetational gradients and proximity to woodchuck
Marmota
monax
burrows in an old field.
J. Mammal.,
75:775–780.
E
RRINGTON
, P. L. 1961. Muskrats and marsh management. University of Nebraska Press, Lincoln, Nebr.
183 p.
G
AUCH
, H. G., J
R
. 1982. Multivariate analysis in community ecology. Cambridge University Press, Cam-
bridge, England. 298 p.
G
LEASON
,H.A.
AND
A. C
RONQUIST
. 1991. Manual of vascular plants of northeastern United States and
adjacent Canada. Second edition. New York Botanical Garden, Bronx, N.Y. 910 p.
H
ANSELL
, M. H. 1993. The ecological impact of animal nests and burrows.
Functional Ecology,
2:5–12.
H
OWES
, B. L., R. W. H
OWARTH
,J.M.T
EAL AND
I. V
ALIELA
. 1981. Oxidation-reduction potentials in a
salt marsh: Spatial patterns and interactions with primary production.
Limnol. Oceanogr.,
26:
350–360.
H
UNTLY
,N.
AND
R. I
NOUYE
. 1988. Pocket gophers in ecosystems: patterns and mechanisms.
BioScience,
38:786–793.
J
OHNSTON
, C. A. 1995. Effects of animals on landscape pattern, p. 57–80.
In
: Hansson, L., L. Fahrig
and G. Merriam (eds.). Mosaic landscapes and ecological processes. Chapman and Hall, Lon-
don.
J
ONES
, C. G., J. H. L
AWTON AND
M. S
HACHAK
. 1994. Organisms as ecosystem engineers.
Oikos,
69:373–
386.
J
ORDAN
,T.E.
AND
D. W
HIGHAM
. 1988. The importance of standing dead shoots of the narrow leaved
cattail,
Typha angustifolia
L.
Aquatic Botany,
29:319–328.
K
ANGAS
, P. C. 1985. Vegetation on muskrat mounds in a Michigan marsh.
Am. Midl. Nat.,
13:392–396.
K
IVIAT
, E. 1978. Vertebrate use of muskrat lodges and burrows.
Estuaries,
1:196–200.
M
EADOWS
,P.S.
AND
A. M
EADOWS
. 1991. The environmental impact of burrowing animals and animal
2000
63C
ONNORS ET AL
.: M
USKRAT
D
ISTURBANCE
burrows: the proceedings of a symposium held at the Zoological Society of London on 3rd
and 4th May 1990. Clarendon Press, Oxford, London. 349 p.
M
ITSCH
,W.J.
AND
J. G. G
OSSELINK
. 1993. Wetlands, 2nd ed. Van Nostrand Reinhold, New York, N.Y.
722 p.
N
AIMAN
, R. J. 1988. Animal influences on ecosystem dynamics.
BioScience,
38(11):750–752.
N
YMAN
, J. A., R. H. C
HABRECK AND
N. W. K
INLER
. 1993. Some effects of herbivory and 30 years of weir
management on emergent vegetation in brackish marsh.
Wetlands,
13:165–175.
O
DUM
, W. E., M. L. D
UNN AND
T. J. S
MITH
III. 1978. Habitat value of tidal freshwater wetlands, p. 248–
255.
In
: P. E. Greeson, J. R. Clark and J. E. Clark (eds.). Wetland functions and values: the
state of our understanding. American Water Resources Association. Minneapolis, Minn.
P
ICKETT
,S.T.A.
AND
P. S. W
HITE
, eds. 1985. The ecology of natural disturbance and patch dynamics.
Academic Press, New York, N.Y. 472 p.
R
EICE
, S. R., R. C. W
ISSMAR AND
R. J. N
AIMAN
. 1990. Disturbance regimes, resilience, and recovery of
animal communities and habitats in lotic ecosystems.
Environ. Manage.,
14:647–659.
R
EZENDES
, P. 1992. Tracking and the art of seeing: how to read animal tracks and sign. Camden House
Publishing, Charlotte, Vt. 320 p.
R
YTHER
,J.H.
AND
W. M. D
UNSTAN
. 1971. Nitrogen, phosphorous and eutrophication in the coastal
marine environment.
Science,
171:1008–1113.
S
AMUELS
, M. L. 1989. Statistics for the life sciences. Dellen Publishing Company, San Francisco, Calif.
597 p.
S
CHIFFMAN
, P. M. 1994. Promotion of exotic weed establishment by endangered giant kangaroo rats
(
Dipodomys ingens
) in a California grassland.
Biodiversity and Conservation,
3:524–537.
S
MITH
, F. R. 1938. Muskrat investigations in Dorchester County, MD 1930–1934. Washington, D.C.: U.S.
Department of Agriculture Circular 474, 24 p.
S
WIHART
, R. K. 1991. Influence of
Marmota monax
on vegetation in hayfields.
J. Mammal.,
72:791–795.
S
WIHART
,R.K.
AND
P. M. P
ICONE
. 1991. Effects of woodchuck activity on woody plants near burrows.
J. Mammal.,
72:607–611.
V
ELLEMAN
, P. F. 1988. Data Desk handbook, version 4. Data Description (Inc.) Ithaca, N.Y.
W
ELLER
, H. W. 1987. Freshwater marshes: ecology and wildlife management, 2nd ed. University of
Minnesota Press, Minneapolis, Minn. 150 p.
S
UBMITTED
9N
OVEMBER
1998 A
CCEPTED
2J
ULY
1999