||HH|WHWIINIHIIWIll!WW\IIHIIWWIWI

   

LIBRARIES

" \llll‘llllll l lullll \llLllll

 

 

 

W518

 

LIBRARY
Michigan State
University

 

 

 

This is to certify that the

thesis entitled

Influence of Beaver on Three Watersheds
in the Houghton Lake Area of Michigan

presented by

Jennifer Dorset Derby

has been accepted towards fulfillment
of the requirements for

Master of Sciencedegree in Fish. & Wildl.

 

 

 

 

 

Major professor

Date March 30, 1995

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

___ __ we- -w-.__._.—.—§_.__ A 7 ‘ fl _ _ ’7

45—.— -v’————. ‘ .f

 

PLACE It RETURN BOXtonmavomI-chockoutfrom yourncord.
TO AVOID FINES return on or More data duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU I. An Afflmattvo Adlai/Equal Opportuntty Instltwon
Wm:

INFLUENCE OF BEAVER ON THREE WATERSHEDS
IN THE HOUGHTON LAKE AREA OF MICHIGAN

BY

Jennifer L. Dorset Derby

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTERS OF SCIENCE

Department of Fisheries and Wildlife

1 995

ABSTRACT
INFLUENCE OF BEAVER ON THREE WATERSHEDS
IN THE HOUGHTON LAKE AREA OF MICHIGAN
By

Jennifer L. Dorset Derby

The degree of impact on wetland habitat by beaver (Castor canadensis)
was evaluated in three watersheds, one with undeveloped land (public
ownership) and the other two with developed lands (private ownership). A
combination of National Wetland Inventory maps with Geographical
Information System and a 1993 field survey was used to evaluate landscape
and successional changes. Avian censusing was accomplished using fixed
radius circular sample plots on representative habitat types found in
beaver impacted wetlands representative of four successional stages.

More wetland area impacted by beavers was present on undeveloped
land with a four fold increase in wetland area impacted by beaver over a 12
year period. Successional changes in vegetative patterns resulted in
more diverse habitat types associated with beaver ponds on undeveloped
lands. Greater vegetative diversity and numbers of breeding birds and
more traditional wetland species corresponded with the diversity of

habitat types on older impoundments.

This master's thesis is dedicated to my husband, Ed, for his continued
support, encouragement and love.

ACKNOWLEDGMENTS

I would like to acknowledge some of the people who helped me in my
research. The five Michigan State University student workers who helped
at various times throughout the past two years provided invaluable
assistance. Cheryl Cunningham, David Morton, Carol Baumann, Mark Smith
and Daria Hyde assisted with the gathering of data in my field seasons and
also contributed priceless ideas and suggestions. I would especially
like to thank David Morton and Daria Hyde for their continued help after
the field season by lending their talents to lab work and data entry.

Dr. Harold H. Prince, my advisor, provided me with this research
opportunity and gave me the freedom to express myself while learning to
use the scientific skills I have studied. I would also like to thank my
committee members, Dr. Thomas Burton and Dr. Thomas Coon, for their
valuable input and encouragement at critical points in my studies.

The personnel at the Department of Natural Resources in Houghton
Lake, Michigan, helped me wherever possible. My appreciation goes out to
Rich Earle, John Nellist and Dan Moran for their interest, concern and
encouragement. Conservation officer Jim Espinoza along with Dan Moran
gain my unfailing gratitude for their efforts in locating us during an
eventful night in the Dead Stream Swamp. Thank you both also for opening
up your homes and lives to us lonely students!

The members of Dr. Prince's lab group also receive my thanks for

listening to my problems, giving advice and gathering supplies from

iv

v
campus. Cathy Flegel, Charlotte Lawrence, Mike Monfils, John Niewoonder
and Michael Whitt lent support at every turn. It would have been difficult
without your help. Thanks especially to two good friends, John and
Charlotte.

My family and friends were also very supportive along the sometimes
rocky road leading to this thesis. I appreciate all of the encouragement
from my parents, my husband Ed and my best friend Mikala.

Lastly, I am grateful for the funding provided by the Michigan
Department of Natural Resources and the Michigan Agricultural

Experimental Station.

List of Tables

TABLE OF CONTENTS

 

List of Figures

 

Introduction

 

Methods

 

Results

 

Discussion

 

List of References

 

vi

vii

viii

18
22

LIST OF TABLES

Summary of individual beaver impacted wetland sites examined ..................

Watershed attributes compared on the watersheds of the Dead Stream,
the West Branch of the Muskegon River, and the Butterfield Creek in
the vicinity of Houghton Lake, Michigan

 

Distribution of MlRlS categories on 3 watersheds in the vicinity of
Houghton Lake, Michigan

Area (ha) of wetland and upland habitat in the surveyed portions of 3 .
watersheds in the vicinity of Houghton Lake, Michigan, based on 1981
NWI maps.

Number and area of active and inactive beaver impacted wetlands
within 3 watersheds in the vicinity of Houghton Lake, Michigan, based
on 1993 field survey

Proportion (%) of each Naiman et al. (1988) successional stage on
areas with active and inactive beaver sign in 3 watersheds in the

vicinity of Houghton Lake, Michigan, based on 1993 field survey ................

Frequency of occurrence (%) of bird species on plots in dominant
vegetation within specific beaver impoundments. Values of 20% and
greater for each species are reported

vii

..8

10

_12

13

13

14

17

LIST OF FIGURES

Schematic representation of the complex pattern of beaver pond
succession (from Naiman et al. 1988) 2

Location of study sites in the Houghton Lake Area within Missaukee
County, Michigan 6

 

Three watersheds examined in the vicinity of Houghton Lake,
Michigan - 7

Distribution of the number of individuals and species of birds
observed per visit on 1 ha survey plots on the individual wetlands
examined 1 6

viii

INTRODUCTION

Wetlands are formed by isolated geologic activities such as
depressions formed by glacial action (Mitsch and Gosselink 1986).
Wetlands also form along the shores of lakes, in protected shallow areas
(Prince 1985). Formation of riparian wetlands by the American beaver
(Castor canadensis) can be a dominant influence in first- through fourth-
order stream ecosystems. Most dams occur on these streams because dams on
larger streams are often removed by freshets (Naiman et al. 1988). Beaver
impoundments convert ecosystems, alter plant communities, and affect
pathways and rates of nutrient cycling. Beaver modify vegetative I
composition and succession by inundation from dam construction, tree
cutting and removal, and dewatering following abandonment (Broschart et
al. 1989). The creation of aquatic and semi-aquatic patches by beaver in
forests increase landscape heterogeneity (Johnston and Naiman 1990).
When beaver are left undisturbed by both managers and trappers, their
activities may influence a large proportion of streams in a drainage
network. Furthermore, these alterations may remain as part of the
landscape for centuries.

Naiman et al. (1988) predict a pattern of beaver induced alterations
in stream ecosystems that impact the ability of stream ecosystems to
resist and recover from disturbances. These impacts include altering the
stream hydrologic regimes, influencing invertebrate communities, and
changing the biogeochemistry of the stream (Naiman et al. 1988). Because

beaver ponds are created and maintained by living organisms, the ponds

1

2
themselves are dynamic, changing as they are colonized, flooded, and
abandoned by the beaver (Johnston and Naiman 1990). To better understand
beaver pond succession, Naiman et al. (1988) outlined a complex pattern
that may ultimately involve the formation of emergent marshes, bogs, and
forested wetlands (Figure 1). This cycle occurs naturally due to beaver
activities. Beaver activity impounds parts of stream ecosystems that
when left undisturbed may develop into a complex wetlands ecosystem. As
the beaver build dams the raised water levels and the use of available

trees for food ultimately eliminates the beaver food source. Abandoned

Emergent

Pond 0 P00 . Wetland
.Bog

Stream 0 oOId

 

 

 

Pond Forested
o ire, Disease . Wetland
Meadow
l 1-100years l50-200yrs l 200-2000years I
REVERSIBLE ’ IRREVERSIBLE
SUCCESSION SUCCESSION

Fig. 1. Schematic representation of the complex pattern of beaver pond
succession (from Naiman et al. 1988).

3
ponds may develop into wetlands or revert to stream to continue the cycle.
The stages labeled I-V (Figure 1) in the successional sequence
encompassed the original idea of beaver pond succession showing the steps
from pond formation, through its aging process to meadow when eventually a
stream is reformed as a new channel is cut and the riparian vegetation
matures. Now it is recognized that a complex pattern of multisuccessional
pathways may be taken depending upon existing vegetation, hydrology,
topography, fire, disease, and herbivory (Naiman et al. 1988). Beaver
have been documented to cause problems for fish by increasing the water
temperature in the ponds, removing shade trees along the stream, and by
causing sediments trapped by the dam to cover spawning beds (Rue 1964).
Warming of downstream reaches by beaver ponds is commonly cited as a
detrimental effect of beaver on trout populations, but the literature has
established no clear relationship between different sizes or numbers of
impoundments and the degree of stream warming (McRae and Edwards 1994).
Air temperature is the single most important determinant of stream
temperature in the absence of other thermal inputs even with groundwater
discharge into the stream. McRae and Edwards (1994) found that local
differences in the degree of shading, groundwater inflow, and stream
volume make it difficult to generalize about the effect of beaver
ir"lixbundments on stream temperature. Their results suggest that large
Ponds (> 8 ha) act as thermal buffers, raising downstream water

tern peratures slightly in some cases, but also dampening the diel

fluetuation.

4

In more fragmented landscapes, such as private lands, the diversity
of human use discourages colonization by beaver with fewer wetlands being
a consequence. Stream ecosystems on undeveloped lands, with less habitat
fragmentation, should be impacted more by beaver than on developed lands
and may have more wetlands in stream ecosystems with no fisheries
conflict. Therefore, maximum stability and diversity of wildlife in
wetland habitats resulting from beaver activity should be achieved in
undeveloped contrasted with developed stream ecosystems. Vegetation and
respective avian use should also differ according to the wetland habitats
developed. We propose that aquatic habitats in watersheds with the least
amount of human activity on floodplain zones will have greater beaver
activity resulting in more wetlands as well as more diverse vegetation and
avian use.

A number of predictions can be made and evaluated based on the
hypothesis. First, we predict that watersheds on undeveloped lands
should have more wetland habitat due to beaver activity than watersheds on
developed lands. Next, wetland habitat impacted by beaver on undeveloped
lands should be in later succesSional stages with fewer new ponds (I, II)
and more old ponds (III, IV) and meadows (V) than the wetlands formed by
beaver on developed lands. Third, wetland habitat impacted by beaver
activity on undeveloped lands should have more diverse vegetative
patterns and correspondingly greater numbers and species of birds than in

wetlands impacted by beaver on developed lands.

METHODS

Three similarly sized riverine systems were examined in Missaukee
County of the northern Lower Peninsula of Michigan (Figure 2). The
watersheds of the Dead Stream, the West Branch of the Muskegon River, and
the Butterfield Creek were chosen based on recOmmendation of Department
of Natural Resources Wildlife Division personnel at Houghton Lake.

Individual watersheds were outlined by locating the highest elevation on
topographic maps between rivers and drawing the watershed boundary with
respect to highest elevations providing area of each watershed (ha).

Stream length (Km) and average gradient (lem) were measured from

. topographic maps. Soil types and percentage of hydric soils in each
watershed were determined using soil survey information (Frederick

1985).

The amount of wetland and upland habitat in the 3 watersheds was
estimated from the 1981 NWI maps and land use was estimated using the
Michigan Resource Information System (MIRIS). We surveyed a portion of
each watershed in 1993 along the main channel of the streams to compare
wetland and upland areas delineated on 1981 NWI maps (Figure 3). Wetland
and upland area (ha) in 1993 was determined by systematically searching
the areas along the rivers and comparing with 1981 NWI maps. Areas that
were discrepancies from the 1981 NWI maps were identified and classified.
Beaver impacted wetlands were classified as active or inactive in the
portions surveyed in 1993 based upon the presence or absence of recent dam

maintenance. We used the Naiman et al. (1988) successional stage

5

t1

 

 

» n

 

Missaukee
County

Fig. 2. Location of study sites in the Houghton Lake Area within
Missaukee County, Michigan.

LEGEND
m 1993 Survey Area
Dead Stream Walershed
West Branch Watershed

Butterfleld Creek Watershed '
1993 6 1994 Study Sites
1994 Only Study SIIes

 

Fig. 3.Three watersheds examined In the vicinity of Houghton Lake, Michigan.

8
continuum (Figure 1) to classify the stage of each beaver pond located on
the 1993 survey of each watershed.

Individual wetland sites impacted by beaver were chosen to be
representative of successional stages with their characteristic dominant
vegetative groups (Table 1). Dominant vegetative classes were noted at
each study site. Six beaver impacted wetlands corresponding to specific
successional stages were identified as follows: Addis Creek Dam (Stage
II) with 20% cover on the impoundment, Dead Stream Darn (Stage III)
supporting 50% cover, Intermediate Darn (Stage IVN) and Turnerville Road
site (Stage V) both with 95% cover, and WIllow Run Darn (Stage N") at the

lowest coverage, 5%. Avian species present and/or nesting at each study

Table 1. Summary of individual beaver impacted wetland sites examined.

 

 

Pond Successional Dominant Percent Cover
Site Area Stage Vegetation on lmpoundment
(ha)

Addis Creek Dam 11 II Alnus rugosa 20
Osmundia

Dead Stream Dam 15 III AIDUS rugosa/ 50
Populus spp.
Graminae

Intermediate Dam 9 IV Scirpus/ 95
Calamagmstis spp.
ljvpha Iatifolia

Turnerville Site 23 V Graminae! 95
Scirpus spp.
Typha Iatifolia

Willow Run Dam 10 MI Ainus rugosa 5

 

9
site were recorded between sunrise and 10:00 hours during the months of May,
June, and July in 1993 and 1994. All individuals and species of birds seen or
heard within fixed radius (18 m) circular sample plots (area = 0.1 ha) were
counted (Brown and Dinsmore 1986). Plots on each site were established and
observations were made during 6 minute periods on each circular plot with the
last 2 minutes of each count period devoted to replay of taped calls of secretive
birds (Brown and Dinsmore 1986). All responses were recorded for a period of 1
minute after the tape was stopped (Prince 1985). After the count period, the
area within 18 m of the same observation point (area = 0.05 ha) was searched
for nests (Brown and Dinsmore 1986). Nest searches were partitioned into 3
chronological periods: the first period from the last week of April until mid-May,
the second period from the last week in May through the first 2 weeks In June,
and the third period from the last 10 days of June through the first 10 days of
July (Prince 1985). Values were combined for both years and standardized per

number of visits made to each wetland site.

RESULTS
The Dead Stream at 13 Km in length had an average gradient of 0.7 lem
(Table 2). The West Branch and Butterfield Creek were over twice the distance
compared with the Dead Stream. The Dead Stream and the West Branch
streams were both Second Order streams while the Butterfield Creek was a First
Order stream (Morisawa 1968) and there were no human-made flow restrictions

along the rivers other than culverts beneath roads (Figure 3). Ownership of the

10

Table 2. Watershed attributes compared on the watersheds of the Dead Stream,

the West Branch of the Muskegon River, and the Butterfield Creek in the vicinity
of Houghton Lake, Michigan.

 

 

Potential
Avg. Stream Hydric Wetland
Area Gradie Length Soils Area
Watershed ha m/Km Km % ha
Dead Stream 7920 0.7 13 60 4752
West Branch 9540 0.5 29 56 5342
Butterfield Creek 8580 0.8 31 36 3089

 

land In the Dead Stream watershed was undeveloped governmental lands (71%)
while ownership of land in the West Branch watershed was primarily private
(87%) following the amount of private land In the Butterfleld Creek watershed
(91%). Soils along the Dead Stream river belonged to the Lupton-Roscommon-
Tawas association. These soils are on nearly level, poorly drained and very
poorly drained mucky and sandy soil sites in upland depressions on outwash
plains, till plains and moraines. The surrounding upland soils were classified as
Rubicon-Montcalm-Graycalm association described as being nearly level to
steep, well drained and somewhat excessively drained sandy soils on outwash
plains, till plains and moraines. The Dead Stream watershed had approximately
60% hydric soils leading to a potential of 4752 ha of wetland area In the

watershed. Soils along the West Branch stream were classified as

11
Lupton-Roscommon-Tawas in association with upland soils classified as
Rubicon—Montcalm-Graycalm. Approximately 5342 ha of wetland habitat could
have been present in the West Branch watershed prior to development based on
an estimate of 56% hydric soils. Soils along Butterfield Creek were mostly
classified as AuGres-losco-Lupton association which is described as nearly level,
somewhat poorly drained and very poorly drained sandy and mucky soils on
outwash plains and till plains. The adjacent upland soils were generally Nester-
Kawkawlin-Manistee association defined as nearly level to gently rolling, well
drained and somewhat poorly drained loamy and sandy soils on till plains and
moraines. Butterfield Creek watershed contained about 36% hydric soils which
suggests there may have been as much as 3089 ha of wetland area. Potential
wetland area based on the amount of hydric soils was highest in the West
Branch and lowest for Butterfield Creek.

When comparing the amount of potential wetland area as indicated by the
1981 NWI maps to the actual amount calculated from MIRIS values (Table 3).
the Dead Stream had 93% of the potential wetland area. The Butterfield Creek
had 58% of the potential wetland area and the West Branch had only 45%.
Butterfield Creek watershed had the highest proportion of urban and agricultural
development, followed by the West Branch and the Dead Stream. The level of
urban and agricultural development was significantly different between the
watersheds (Chi Square, d.f.=2, p<0.001). Although the 3 watersheds were
similar in size, each contained significantly different amounts of wetland and
upland habitat area (Chi Square, d.f.=2, p<0.001). Dead Stream had the

greatest amount of wetland area as compared to the West Branch and the

12

Table 3. Distribution of MIRIS categories on 3 watersheds in the vicinity of
Houghton Lake, Michigan.

 

 

Development Dead Stream West Branch Butterfield Creek
ha % ha % ha °/o
Urban 237 3 76 1 34 1
Agricultural 127 2 2833 30 3543 41
Rangeland 713 9 735 7 412 5
Forest 2408 30 351 1 37 2789 32
Wetland a 4435 56 2385 25 1802 21

TOTAL 7920 9540 8580

 

a NWI estimates of wetland area: Dead Stream 4440 ha, West Branch 2340 ha, Butterfield Creek 1830 ha.

Butterfield Creek.

The 1993 status of wetland habitat was evaluated and contrasted with the
1981 NWI classification on survey areas ranging in size from 595 to 1334 ha
(Table 4). The amount of wetland area was not significantly different over time
for the 3 watersheds (Chi Square, d.f.=2, p<0.001). A

The Dead Stream survey area had the greatest number of beaver dams
present and greatest area impacted by beaver activity (Table 5). Six of the 14
dams were actively maintained by beaver in the Dead Stream watershed while 1
active dam'was maintained on the West Branch and the Butterfield Creek had no
active dams. The area impacted by beavers increased from 37 ha in 1981 to
172 he on the 1993 Dead Stream survey area. Beaver activity along the West
Branch and Butterfield Creek remained relatively constant from the 1981 NWI

13

Table 4. Area (ha) of wetland and upland habitat in the surveyed portions
of 3 watersheds in the vicinity of Houghton Lake, Michigan, based on 1981
NWI maps.

 

 

Linear Distance
Watershed Area Surveyed Sumed Wetlands Uplands
Dead Stream 595 3.2 428 167
West Branch 1334 14.4 932 434
Butterfield Creek . 720 9.4 286 402

 

Table 5. Number and area of active and inactive beaver impacted wetlands

within 3 watersheds in the vicinity of Houghton Lake, Michigan, based on 1993
field survey,

 

 

Wet'and Dead Stream West Branch Butterfield Creek
Status
. # dams 6 . 1 0
active
area (ha) 74 11 4
# dams 3 11 3
inactive

area (ha) 98 31 8

 

14
maps to the 1993 field survey. The amount of beaver impacted wetland area in
the Dead Stream sample area was signllficantly higher compared with West
Branch and Butterfield Creek (Chi Square, d.f.=2, p<0.001). whereas the area in
the West Branch was not significantly different than in the Butterfield Creek (Chi
Square, d.f.=2, p<0.001).

The greatest number of successional stages was found on the 14 sites of
beaver activity in the Dead Stream watershed as compared to the West Branch
and the Butterfield Creek for the area surveyed (Table 6). New pond area (II)
was recorded at one site on the West Branch while early successional stages
[new pond (II), pond (Ill), and old pond (IV)] were not present on the Butterfield
Creek. Most of the identifiable beaver activity in the West Branch and the

Table 6. Proportion (%) of each Naiman et al. (1988) successional stage
on areas with active and inactive beaver sign in 3 watersheds in the
vicinity of Houghton Lake, Michigan, based on 1993 field survey.

 

Successional Dead Stream West Branch Butterfield Creek

 

Stage (172 ha) (42 ha) (12 ha)
stream (I) 3 74 75
new pond (ll) 13 2 0
pond (Ill) 28 0 0
old pond (IV) 14 o o
meadow (v) 37 24 25

emergent wetland 5 0 0

 

15
Butterfield Creek watersheds were meadows (V) reverting back to wooded
floodplain on inactive sites with nonfunctional dams still visible .

Avian census data was combined for both years and standardized per visit
for each habitat type present on beaver impacted wetlands (Figure 4). The
greatest number of individuals per visit was recorded on survey plots on
impoundments with shrub/scrub and emergent vegetation habitat types. Species
numbers per visit remained fairly constant across habitat types.

Twenty species were recorded during 20% or more of the visits to the
various wetlands (Table 7). Seven species, wood duck (Aix sponsa), mallard
(Anas plafymynchos), blue-winged teal (A discors), blue jay (Cyanocitta
cristata), chipping sparrow (Spizella passen‘na), swamp sparrow (Melospiza
georgiana) and purple finch (Carpodacus purpumus), were counted at only 1 site
each. Four species, Eastern kingbird (Tyrannus Iyrannus), tree swallow
(Tachycineta bicolor), American robin (Turdus migraton'us) and red-winged
blackbird (Agelaius phoem'ceus), were recorded at 4 sites each. The greatest
number of species was recorded on the Intermediate Dam with 95% vegetative
cover, dominated by woolgrass, cattall and blue-joint grass.

Few nests were located on habitat types sampled on impounded areas.
One nest, tree swallow, was observed in 1993 located on the Intermediate Dam.
Two nests were observed in 1994, a tree swallow nest on the Intermediate Dam

and an Eastern Kingbird nest located at the Dead Stream site.

 

20

   
  

ividuals

-- ggecie visit
I li‘ews

 

Birds
8

 

 

 

Alnus rugosa Alnus rugosa Scirpus spp. Graminae Alnus rugosa
Graminae Typha laiifolia Scirpusspp.

Dominant Vegetation

Figure 4. Distribution of the number of individuals and species of birds observed
per visit on 1 ha survey plots on the individual wetlands examined.

17

Table 7. Frequency of occurrence (%) of bird species on plots in dominant
vegetation within specific beaver impoundments. Values of 20% and greater for
each species are reported.

 

% Presence

 

Bird Avian census 2 Alnus rugosa 2 ScI‘rpus spp. 2 Graminae
SpeCleS plots 3Afnus rugosa 1 Graminae 1 Typha latibh’a 1 Typha latifolia 3Alnus rugosa
Wood Duck . — 25 — — -
Mallard — 31 — — —
Blue-winged Teal — - — 22 —
Northern Flicker 20 - 20 22 —
Great Crested Flycatcher — — 20 55 —
Eastern Kingbird 27 50 27 - 27
Tree SWallow 47 44 80 — 67
Blue Jay 33 — — - —
American Crow — - — 22 —
Black-capped Chickadee - — — 22 20
American Robin 47 25 40 44 —
Cedar Waxwing ‘ — — 40 . 22 ‘ 20
Common Yellowthroat — -.- 27 33 -
Chipping Sparrow — — 20 — -
Song Sparrow 33 — 33 - 33
Stamp Sparrow — — 33 - —
Red-winged Blackbird 33 — 80 55 80
Common Crackle — 63 40 — 67
Baltimore Oriole — 38 — — -
Purple Finch 20 — — - -
Total 6 7 12 - 9 7

 

 

DISCUSSION

Beaver activity impacted the floodplain habitat the most along the Dead
Stream followed by the West Branch and lastly by the Butterfield Creek. The
Dead Stream was located on almost entirely state-owned land with little human
interference and there was continuous spatial impact by beaver along the
floodplain. The West Branch had more human activity on the floodplain with the
exception of one 7 Km section, observed during the 1993 field survey, where
there appeared to be little to no attempt made by property owners to control
beaver activity. Limited beaver activity occurred along the remaining 22 Km in
the agricultural areas where beavers are most likely controlled. Signs of past
beaver activity was visible along Butterfield Creek while most of the recent
activity was curtailed. This watershed, with the highest level of human activity,
supported very little beaver activity with no active dams.

Land use has a controlling effect on beaver activity. We predicted that
watersheds on undeveloped lands should have a greater proportion of wetland
habitat associated with increased and uninterrupted beaver activity compared
with watersheds on developed lands. According to the 1981 NWI maps, the
Dead Stream had the greatest area of wetlands compared with the other 2
watersheds. The West Branch watershed had more wetland area than
Butterfield Creek watershed as reflected by land use activities on the West
Branch being primarily restricted to camping, fishing, and hunting.

The proportion of the different types of beaver impacted wetlands present in
the landscape is dynamic and will vary due to the multidirectional and nonlinear

successional stage replacement related to beaver activity. This pattern of

18

19
successional change is driven in part by beaver colonization, abandonment, and
recolonization and tends to perpetuate a dynamic state of vegetative community
structure and composition, increasing edge habitat, and promoting habitat
diversity (Broschart et al. 1989). The Dead Stream showed the greatest number
of Naiman et al. (1988) successional stages. Although we predicted the
occurrence of varied successional stage beaver ponds on undeveloped lands,
such as the Dead Stream, the scarcity of early successional stages on the
developed watersheds suggests that beaver activity is being actively controlled
as indicated by the beaver nuisance records from the DNR. Undeveloped lands,
with less human disturbance, were expected to have beaver ponds at various
successional stages from new ponds to emergent and possibly forested
wetlands, supporting many different vegetative patterns.

Beaver pond succession should be in earlier successional stages on
developed lands because the human disturbances will not allow the beaver
ponds to proceed further along successional patterns which also maintains
certain vegetative patterns due to the pond successional stage. The beaver
activity on the West Branch and Butterfield Creek is not only being controlled,
but is declining as shown by the wetland area in the meadow stage and little
area in earlier successional stages. A greater variety of successional stages on
beaver ponds were found on undeveloped watersheds again due to the lack of
human activities. The successional cycle of beaver ponds on undeveloped
watersheds tend not to be interrupted by human pressures and therefore have a
chance to proceed naturally through the successional pattern.

A beaver impounded landscape is a mosaic of different vegetative types

20
due to the dynamic hydrology of beaver ponds, the diversity of pre-impoundment
vegetation, and the changes caused by beaver foraging in the riparian zone
(Naiman et al. 1988). The successional diagram from Naiman et al. (1988)
(Figure 1) was used to compare physical and vegetative differences between
wetlands in vaIying successional stages. Naiman et al. (1988) predict a
successional pattern of riparian zone changes due to beaver impacts. The
beaver concentrate on clear-cutting the preferred dominant deciduous species
such as trembling aspen (Populus tremuloides). Initially, the riparian Zone
becomes more open as shrubs become the dominant growth form. Eventually
the nonbrowsed species such as black spruce (Picea mariana) and balsam fir
(Abies balsamea) may overtop the shrubs. The impact of the beaver activities
initially reduces vegetation height and then alters biomass partitioning (Naiman
et al. 1988). In our individual beaver impacted wetland sites we observed that
vegetative diversity was tied to beaver pond succession on both developed and
undeveloped lands.

While reasoning that beaver impacted wetlands on undeveloped lands will
support more diverse vegetative patterns, we predicted that these same
wetlands would support greater numbers and species of birds. Vegetative
communities influence the avian composition of an area since avian use Is
directly related to the type, structure, and composition of the vegetation present
in the wetland. There were low densities of birds In all of the beaver impacted
wetlands with the greatest numbers being recorded on mid successional staged
wetlands with greater vegetative diversity. Early and late successional stages

supported more avian species associated with upland ecosystems, less diverse

21
vegetative communities and correspondingly fewer numbers of birds. Vegetative
diversity and avian use are directly related to beaver pond successional stage.
When beaver ponds are allowed to progress successionally, the vegetative and

avian species naturally respond.

LIST OF REFERENCES

LIST OF REFERENCES

Broschart, M.R., C.A. Johnston, and R.J. Naiman. 1989. Predicting beaver
colony density in boreal landscapes. J. Wildl. Manage. 53(4):929-

934. '

Brown, M. and J.J. Dinsmore. 1986. Implications of marsh size and
isolation for marsh bird management. J. WIldl. Manage. 50:392-397.

Frederick, W.E. 1985. Soil survey of Missaukee County, Michigan. U. 8.
Dept. Agr. Soil Conserv. Serv. 138 pp.

Johnston, CA. and R.J. Naiman. 1990. The use of a geographic information
system to analyze long-term landscape alteration by beaver.

Landscape Ecology 4(1):5-19.

McRae, G. and OJ. Edwards. 1994. Thermal characteristics of Wisconsin
headwater streams occupied by beaver: implications for brook trout
habitat. Transactions of the American Fisheries Society 123:641-656.

Michigan Resource lnforrnation System (MIRIS). 1975. Michigan Land Use
Classification and Referencing Committee. Division of Land Resource

Program, Department of Natural Resources.

Mitsch, W. and J. Gosselink. 1986. Wetlands. Van Nostrand Reinhold, N.Y.
539 pp.

Morisawa, M. 1968. Streams: their dynamics and morphology. McGraw-Hill,
New York.

Naiman, R.J., C.A. Johnston, and J.C. Kelley. 1988. Alteration of North
American streams by beaver. Bioscience 38(11):753-762.

Prince, H.H. 1985. Avian communities in controlled and uncontrolled Great
Lakes Wetlands. Pages 99-119 in H.H. Prince and FM. D'ltri, eds.
Coastal wetlands. Lewis Publishers, Inc., Chelsea, MI. 286 pp.

Rue. L.L. III. 1964. The world of the beaver. J.P. Lippincott Co.,
Philadelphia, PA.

22