Q ,
A

 

1."

 

A.

SERVALL MAMMAE. “ATCHES fit‘éDER "EW'S
EN‘EENSETEES GE'- BEZR ERGWSENG

Thesis £3: #56 Deg!» 0% M. S.
MiCHiGAN STATE COLLEGE
éack 9. Eeafiée
E3955

M11311 11111111111111

329100809 7267

b This .is. to certify til—at” the

thesis entitled

‘ —-.-. _ _-
—-—-— _..'...

SMALL Mabel. CAIthS UNDER “TWO Itdlhuolllhs CF

~-

, DEER Bacwsgcgl '_ ‘_ V

presented'b'g

Jack D. Beattiefl

V , - .- 42....

O
. 1.. ~
t'-

has been accepted towards fulfillment
of the requiremerfiS"Ior

M. s. , Fisheries and Wildlife
degree In“

L . W. Gysel

Major professor

 

Date Iiay 16. 1955

 

 

I,

u.wwww.pua-p. I

PLACE N RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or betore date due.

DATE DUE DATE DUE DATE DUE I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1"?

M80 In An Affirmative Action/Equal Opportunity Institution
cAdemM. 1

 

 

 

#x— _,__¥_

 

favorite.

SMALL MPJT'HL CATCHES UNDER THC) IN‘ECSNSITIES OF DEER BROT‘lfSING

by
Jack D. fleattie

ANABSTRACT

submitted to the School of Graduate Studies of Michigan
State College of Agriculture and.Applied Science
in partial fulfillment of the requirements
’ for the degree of

MASTER OF $IfllCE

Department of Fisheries and Wildlife

June 1955

THEbIS

 

 

 

ABSTRACT

A study was made in The Dead Stream Swamp area during the summer of
19514 in an attempt to determine the effects of deer browsing upon small
marshal populations. This was done by measuring compositions and densi-
ties of small manual populations in areas of comparatively unbrowsed and
overbrowsed deer yards in each of three cover types.

Vegetation data were gathered by use of nested quadrats located along
traplines. Vegetation was classed as trees, shrubs, or herbs and recorded
by stem counts and percent cover.

Two sizes of snap traps, placed on traplines, were used to obtain
manual data from each browse condition of each type.

Statistical analysis indicated highly significant differences in mn-
bers of animals between lightly browsed and heavily browsed areas and
significant differences among vegetation types when a combined error term
was used. More small manuals were found in lightly browsed than heavily
browsed areas. The few samles obtained did not provide a very powerful
test of differences between areas but the enmenditure of man power necess-
ary. to detect minor differences could not be justified.

Vegetation samples appeared to be too few to recognize any definite
relationship between amount of cover, food, and numbers of small mammals
captured. Differences in available browse were not due to previous deer
browsing alone but also to the age of the stand and self-pruning.

No correlation was suggested between mmbers of snail mammals captured

and differences in local climate or soil conditions.

SMALL WU'RVEAL CATCH‘ES UNDER THO INTENSITIES OF DEER BROWSING

By
Jack D. geattie

A THESIS

Submitted to the School of Graduate Studies of Michigan
State College of Agriculture and Applied Science
in partial fulfillment of the requirements
fer the degree of

MASTERiOF SCIENCE

Ebpartment of Fisheries and Wildlife

June 1955

TABL510F CONTENTS

IntrOduCt ion 0 O O O O O O O O O O O O O 0

Statement of the Problem

Review of Literature. . . . . . . . .

Description of the Area

Location . . . .

Physiography

Climate. . . . .

Soils. . . . . .

Methods and Procedures . .
Choice of Study Sites

Studied . . .

Cover Types . . . . . . . .

Measurements of vegetation.

Plot size and location

Tree measurements. . .

Shrub measurements

Herb measurements.

Mammal Census . . . .
ResultSe O O O O O O O O O

vegetation. . . . . .

Treatment of Data.

vegetation in Area

vegetation in Area

vegetation in Area

ii

354800

mmqmmn-wwuw

NMHHHHHHHH
HHOJNNNFWNN

Mammal Data
Discussion . . .
Conclusions . .
Appendix A . . .
Appendix B . . .
AppendixC . . .

Literature Cited

Selected Bibliography.

iii

2b
3?
ho
hi
he

in
So

ACKNOULEIIBENTS

Sincere thanks are expressed to the Game Division of the Michigan
Department of Conservation for furnishing equipment and facilities, which
made possible the completion of the problem. In particular, I wish to
thank Mr. R. A. Haciiillan, Biologist in Charge at the Houghton Lake wild-
life Experinent Station, for his fine co-operation and counsel in carrying
out the research, and who helped initiate the study.

The investigator wishes to sincerely thank Dr. Leslie U. Gysel, Asso-
ciate Professor of Fisheries and Wildlife who helped initiate the study,
and whose guidance and interest has helped immeasureably in completing the
problem.

Grateful acknowledgment is also due to Dr. Henry Ehrlington, Associate
Professor Emeritus, Botany and Plant Pathology; to Dr. Ivan P. Schneider,
Associate Ptofessor of Soil Science; to Dr. Charles I... Gilly, former
Assistant Professor of Botany and Plant Pathology; to it. George H. Parme-
lee, Curator of the Beal-Garfield Botanic Gardens; and to Dr. Carter 14.
Harrison, Professor of Farm Crops, for their helpful assistance and sugges-
tions concerning the problem.

Further acknowledgment is made to It. Don w. Hayne, Associate Pra-
fessor of Zoology; Dr. Peter I. Tack, Professor and Head of the Department
of Fisheries and Wildlife; and 11‘. John E. Cantlon, Associate Professor
of Botany and Plant Pathology, Michigan State College for their critical
reading of the manuscript. Dr. Hayne also gave invaluable time and aid
necessary for the completion of the statistical analysis of the small

manual data.
iv

Deep gratitude is due the Peoples of the United States who through
the 6.1. Bill of Rights, made it possible for the writer to pursue his
graduate studies.

Table I

Table II

Table III

Table IV

Table V

Table VI

Table VII

Table VIII

Table IX

Table X

LIST OF TABLES

Small Marmnal Catch by Lines According to Browse Condi-
tionsandCoverType..................

Analysis of Variance of logarithms of Embers (Plus one)
of Small tunnels Caught On 12 Traplines . . . . . . . .

Number of Tree Stems in "Good" and "Poor" Areas of Type
"A",OnIPeP'BCI‘B&SlS. e as e e 0 see a e e e e e

Nmber of Tree Stems in "Good" and "Poor" Areas of Cover
Me"3”0n&Per-icrcmls. e e as e e e e e e as e

timber of Tree Stems in "Good" and "Poor" Areas of Cover
Type"C",onaPer-acreBasis.............

Stem Counts of Shrubs - Based on Eight Plots in Each
cond!tion O O O O O O O O I O O O O O O O O O O O O O 0

Stem Counts of Herbs - Adjusted To 2h Plots . . . . . .

Comparisons of Bosses, Grasses, and Ferns in "Good" and
"Poor" Conditions of Type "A", using Eight Plots, .001
Acre inSize ForEachCondition . . . . . . . . . . . .

Comparisons of losses, Grasses, and Ferns in "Good" and
"Poor" Conditions of Type ”B", Densities Based on 18
Plots, .00025 Acre in Size For Each Condition . . . . .

Comparisons of Masses, Grasses, and Ferns in "Good” and

"Poor" Conditions of Type "C", Using 2h Plots, .00025 in
5128, h ken audition. . C . O . C C . O . O C . O O 0

vi

25

26

31

32
33

3h

35

Figure 1.
Figure 2.
Figure 3.
Figure h.
Figure 5.

Figure 6.

Figure 7.

LIST OF FIGURES

Diagram.of Nested Quadrats Used in Cover Type "A" . . .
Diagram of the Nested Quadrats Used in Cover Type "B" .
Diagram of Nested Quadrats Used in Cover Type "C" . . .

Ground Cover, Fbrest Fioor, and Composition of vegeta-
tion Found in "Heavily Browsedfl Conditions of Cover
me 3 '.A .' and "B" O O O O O O O O O O O O O O O O O O 0

Ground Cover, Forest Floor, and Composition of vegeta-
tion Found in "Lightly Browsed" Conditions of Cover
Mg: I'A“ “d "B" O O O O O O O O O O O O O O O O O O 0

Ground Cover, Ferest Floor, and Composition of vegeta-
tion Found in ”Heavily'Browsed" Conditions of Cover
Me "C". O O O O O O O O O O O O 0 O O O O O O O O O 0

Ground Cover, Forest Floor, and Comosition of vegeta-

tion Found in Part of the "Lightly Browsed" Condition
or cover me "C" O O O O O O O O O O O O O I O O O O 0

vii

10
11

19

20

22

23

INTRODQQTIQN
The Problem

The purpose of the problem was to compare small mammal populations
in comparatively unbrowsed and overbrowsed deer'yards by measuring com-
positions and densities of small mammal populations in these areas.

Bartlett (1950), reported that overbrowsing is not only detri-
mental to the deer but that it also deprives other game Species, such
as snossnoe hares and ruffed grouse of essential low-grcuing food and
cover. it might be assumed that overbrowsing also affects small mammal
populations such as shrews, voles, moles, and mice. It was hoped that
the samplino of these smaller mammal populations, correlated with sam-
ples of vegetation in the same areas, would show whether differences
in mammal populations did exist between "good" and "poor" areas of
browse conditions. If fairly'good quantities of food species were
available, the area was called "relatively good". If food species
were heavily browsed, the area was called "relatively poor". Evalua-

tions were based on ocular estimates.
Review of Literature

Studies of small mammal papulations are both numerous and extensive.

Likewise, sample studies of cover types are abundant. Methods and
technique in both cases vary greatly and there seems to be no particu-
lar answer as to which are the "best". While some studies have com-
pared the effects of large: :nimals on vegetation, few have attempted

to show relationShips between vegetation and small mammal populations.
1

Review of Literature

Dambach (l9hh) reports that his studies based on three nights trap-
ping showed results to be variable between grazed and ungrazed woodlots.
Where leaf litter was of similar depths, the grazed area had a slightly
higher population. However, where the amount of leaf litter was greater
in one area than the other, small mammals were found to be three and one-
third times more abundant in the ungrazed than grazed areas. An appar-
ent increase in abundance of mammals paralleled the development of new
'woody growth and better distribution of leaf litter in the ungrazed
woodldts.

According to Phillip (1936) who based his report on 5010 trap nights
in each area. deer mice were found to prefer overgrazed areas while cot-
ton rats preferred areas with more cover. Apparently, each species had
individual requirements as to cover. Studies by.Allen (1938) of various
small mammal species, (including prairie deer mice, while footed mice,
and shrews), found in various farm.land-use situations, data which
agreed with that of Phillip (1936) concerning. diferential use of cover
by various small mammal species. Blair (1938) found prairie deer mice
more abundant in blue grass situations than other types of cover in the
same area.

* Newbigin (1936), and Dice (1931) agree that the places in which

animals‘iiv! are determined by the nature of the vegetation and its use
as food and cover. Dice based his conclusions on some of his previous
work with snowshoe rabbits and small mammals and upon a review of work

by other investigators.

Description of the Area
9.263229.

The study was made on six plots located in the Dead Strem Swamp
area, which occupies an area about nine and one-half miles long and five
and one-half miles wide in Roscommon and Missaukee Counties in the I
north-central part of the Lower Peninsula of Michigan. Roughly, the
eastern boundary of the swamp lies between the western edges of
Houghton Lake on the south and Higgens Lake to the north. The swamp

then expands westward nine and one-half miles to highway H 7h.

Mam

Although most of this “swamp" area is low and wet, it is actually
of a very diversified nature; consisting of large areas of wet land, swamp,
and bog, interspersed with low sandy ridges and hills in the form of
islands and finger estending into and through the swamp. Veatch, Schoen-
mann, and Noon (1921:) state that the relief is a result of glaciation
and that the more level areas are part of outwash plains and that the till
plains or ridges are mainly land-laid moraines. Several creeks originate
in the swamp. Host of them empty into the Dead Stream, from which the

swamp gets its name.
Climate

Veatch (1921;) reports the chief climatic features for the area
studied to be: a mean annual temperature of about 111 degrees Farenheit;
a normal precipitation which includes melted snowfall of about 2? inches;

an average snowfall of 56.5 inches; low wind movement, low evaporation,
low percentage of possible sunshine; and moderately high humidity. Tem-
peratures below freezing occur from Nbvember through March and occasion;
ally in September and Hay. The frost-free season is about 120 days.

The average date of the last killing frost is June 10 and of the earliest
is September 10, but killing frosts have been recorded in every month

in the year.

Soils

 

Soils were field checked and identified by reference to the descrip-
tion of Veatch (1921;), (19141), and (1953).

Rifle peat is found in cedar-tamarack swamps; these soils have a
brown or dark brown, rather coarse and loose crumbly, moderately decayed
peat in.the surface two or three feet. This surface material usually
grades into yellowish brown, fibrous peat which is but slightly decayed.
Rifle peat is moderately acid and low in fertility. It occupies wet
flats and depressions with a.moderately highdwater table. The natural
water table is less than two feet below the surface for most of the summer.

Saugatuck sand occurs on moist, sandy plains. It occupies positions
but slightly higher than the wet lands or swamps which it commonly borders.
Subsurface-drainage is slow, due to a shallow water table or layer of clay
at or below five feet in depth which prevents free subsoil drainage.

The Newton loamy sands include these soils having dark gray loamy
sands rich in organic matter and acid in reaction, underlain by grayish
water-logged sands which continue to a depth of three or more feet. They
occupy the better drained edges of swamps and still better drained soils

occupy adjoining higher ground.

METHODS AND PROCEDJRES
Choice of Sites for Study

Study sites were limited to areas within the boundaries of the Dead
Stream Swamp. Conditions of browse within this swamp varied from loca-
tion to location; one part of the swamp was browsed quite heavily and an-
other part only a few miles away had considerable browse left, as shown
by field surveys.

To aid in selecting areas to be studied, aerial photos, Game Division
cover maps, and maps showing food conditions of deer yards were used. The
deer yard maps are prepared yearly, based on ocular evaluations made by
District Game Supervisors. Three forest cover types were chosen. Each
of these contained both "relatively good” and "relatively poor" available
browse according to the current naps. A preliminary survey of the cover
types, as recomended by Weaver, (1938), was made to check actual browse
conditions (available browse) against those described on the maps.

Three general cover types were used to compare small mammal popula-
tions. Within each, subjectively selected sites, as near homogeneous as
possible were chosen. In each cover type, areas were selected that were
comparable except in degree of browsing. No satisfactory quantitative
or quasi-quantitative means have yet been devised by which two people
can give independent identical estimates of available browse. Therefore,
the areas were selected by a general appraisal of food conditions. If

fairly good quantities of food species were available, it was called

"relatively good". If food species were heavily browsed, the area was
called "relatively poor". Later it was observed that this condition was
due to the stage of succession of the stand as well as to browsing. It

is granted that this is a loose classification, but deer browsing activi-
ties are so varied, that attempts to define precise limits have been un-
satisfactory, since no two people can. give independent identical estimates

based on ocular values.

Cover 3133

The area of cover type "A" was located in a swamp, with a good stock-
ing of cedar-spmce-tamarack, 0-9 inches d.b.h. in the upper story, and
with a swamp hardwood understory.

The area of cover type "B" was similar to "A". It was in swap,
with a good stocking of spruce-cedar-tamarack, 3-9 inches d.b.h. in the
upper story and also with a swamp hardwood understory.

The area of cover type "C" was near the edge of the swamp, with a
medium stocking of poplar, 3-9 inches d.b.h., and a good stocking of
sprme-balsam-fir-cedar, 0-3 inches d.b.h. in the upper story while the
understory was mostly speckled alder.

In each of the cover types studied, the condition and composition of
the forest floor varied. Logs in various stages of decay, stumps, and
other characteristics varied considerably from place to place in the same
forest type.

The area of cover type "A" was low and level with a water table at
the surface in may places. Vegetation was found mostly on slight gromd
elevations little more than a few inches above the water line. Cover
type "B" was similar but had less area covered by standing water. Cover
type "C" was the dryest of those studied, with the water table slightly
lower than in the other cover types. However, it was still mainly a
«amp-bog condition with considerable amounts of standing water. The soils
here varied slightly from those of previous areas studied. A little

higher content of sand was found in the soil. This area was near the

swamp edge and was partially Newton Sandy loam while the soil associated

with previous cover types was found to be Rifle peat.

Measurements of Vegetation

Plot Size gn_<_i Location

 

Nested quadrats, located along each trapline (see section on small
manual census) to secure quantitative data for vegetation, although
altered in size, were similar to those described by Weaver and Clements
(1938), and recommended by Tansley (1920), Hardy (195), Elliott (1952),
Costing (1953), and Parmelee (1953). Bomdaries of quadrats were located
by means of a compass and tape. Edges of plots were marked with cotton
string. Rectangqu quadrats were used because they were considered to
be most efficient for sampling.

Different sizes of quadrats were used for trees, shrubs, and herbs.
The tree quadrats, first used, (type "11"), were 66 feet square. Inside
this, the shrub plot was 66 feet long and 6.6 feet wide. (he herb plot
six and six-tenths feet square was placed in the nearest end of the
shrub quadrat. Four tree quadrats were located on each trapline; each
quadrat was spaced 50 feet from the next, (see Fig. No. 1).

Because of the time limitations, quadrat numbers and sizes were
changed for cover types "B" and "C". Hauser (1953) reported the use of
varied numbers of quadrat plots in different areas of the same study where
the size of areas differed. Smaller tree and shrub plots, and smaller
but more mimerous herb plots were used. The four tree plots used on each

trap line were 16.7 feet square. One shrub plot 3.3 feet wide and 13.2

Figure No. l. - Diagram of Nested Quadrats used in Cover Type "A".

l 1:. Legend

 

.. _ 1:1 Trees
:8 ES Shrubs
E1 Herbs

Note: Four tree

 

 

 

 

plots were used

in Cover Type "A".

SO.‘

~1<

 

 

 

 

 

 

 

 

. 1
66’
F——— 663 ———>'-(—— SO,‘ ——->'-é———r 66.’

 

 

 

 

 

I /.'1’/////

I
f

10

Figure No. 2.- Diagram of the Nested Quadrats used in Cover Type "B".

 

 

 

 

 

 

 

 

1 [:1 Trees
1 - Shrubs
‘7
3 Herbs
1 Note: Two tree
plots were used in
Cover Type "B".
JL.
I
‘ 7:
‘3
H
1
I":
3
T.’
“i
oxen
ou-o
I‘M

 

 

 

 

 

 

 

7r-
‘5'
'1:

__]L_

11

Figure No. 3. - Diagrar: of Nested Quadrats used in Cover Type "C".

 

 

 

 

 

 

 

 

 

 

 

3'13.

h6.7'

;14
F

 

2H1

+———-— um-

‘13,

1+— 50..

>1

Legend

1:1 Trees
[\3 Shrubs
Herbs
Note: Four tree
plots were used

in Cover Type "C".

 

 

 

12

feet long was located in each tree plot. Herb plots were three and three-
tenths feet square. Three herb plots were located along the trap line in

each tree plot, one on each edge and one in the center, (See Figures 2
and 3).

Tree Measurements

All species one-inch d.b.h. and over were tallied in size classes.
Trees under one-inch d.b.h. were tallied with the shrubs. The density of
tree species was represented by a tally of numbers present in the quadrat.
An ocular estimate of crown class for each species was recorded as judged

to be codominant, dominant, or intermediate in type. Five vigor classes
based on ocular estimates were used to indicate condition of the species:
(1) Vigorous (fast growing); (2) slow growing; (3) stagnant; (h) Dying;
(5) Dead, but still standing.

The average height of the "browse line", (partially due to self -
pruning caused by shade and the age 01 the Stand), was based cn cooler

esrimatates checked occasionally by actual measurement.
Limb lessees” 5

All species were tallied by size class, and density. The frequency
of species used by animals for browse was noted. Size classes for shrubs

in all types were as follows:

£1.22 _____c1asses some 12 means
a O- 6 inches high
b 7 - 12 inches high
c 13 inches high and over but below

1 inch d.b.h.

11

 

 

13

Cover classes were used to indicate approximately how much of the area was
occupied by a species. Cover classes for shrubs were modified from Barick

(1950). For all plot sizes, density classes were:

My; gig—s; Percent 91 £195, covered
T 0 to 10 percent of the plot
1 10 to 33 percent of the plot
2 ' 33 to 66 percent of the plot
3 66 to 100 percent of the plot

Herb Measurements

All species were tallied by stem count; size class; and in some cases,
cover class values. Cover classes, in these plots, were used to indicate
the percentage of the area occupied by a species where stem counts would
be inaccurate and extremely difficult to obtain. Thus cover classes, in-
stead of stem counts were used for grasses, mosses, and ferns. Size clas-
ses for herbs were the same as those described for shrubs except that size
class c was 13 inches high and over, with no reference to diameter of the
stems. Cover classes for herbs in all plots were modified from Barick
(1950). Frequency of use by animals according to species, was noted al-
though use in this class was not necessarily considered to be due to deer
alone since other species such as rabbits and some of the other small mam-
mals known to utilize herbs were present in the areas studied.) Cover

classes for herbs were the same as those described for shrub plots.

 

 

._NI 3: F .

.\ Ufa? L. V a M -

 

 

11:

hmal Census

Procedures similar to those recommended in the North American Census
of Snail Manuals Standardized Sampling Procedure, (anonymous, no date),
were used.1 '

One set of two parallel lines was used in each location (browsed or
unbrowsed) in each cover type. Distance between lines was 1:00 feet. This
distance was supposed to prevent overlaping of trap areas. Each trap line
was started, (1st station), 200 feet from the center of the road (or edge
of the cover type when roads were not present), to exclude "edge effect“.

The line extended in a straight line 950 feet long toward the center of
the stand, or in such a manner as to pass through vegetation thought
representative of the type. In the "heavily-browsed" area of cover type
"C", parallel lines could not be used because of shape of the area, but
here an engineers compass was placed on an orientated map of the area to
determine an azimuth that would bisect the length of the cover type.
Then both trap lines were located along the azimuth. Precautions were
taken similar to those previously described to exclude "edge effect".
Lines in this "browse condition" were located between trails which crossed
the azimuth. Starting points for 'these trap lines were located 200 feet
from the center of the trail. It is recognized that the trap lines were
not random samples of the general area. There were twenty stations on

each trap line with intervals of 50 feet. At each station, three snap

1This is an anonymous, undated, mimeographed release of standardized
samling procedurea recomended by the North American Census of Small Mam-
mals organization“ s on file at the Michigan Department of Conservation,
Houghton Lake wildlife Experiment Station.

1 1

 

 

15

traps were placed within a five foot radius. In three nights, this plan
would total 360 trap-nights for each browse condition in each cover type.

Snap traps of two types were used: (1) Museum Special Traps, and
(2) Victor "No. 2" h-ways Traps (mouse size); both manufactured by the
Animal Trap Company, Leitz, Pennsylvania. At odd-numbered trapping sta-
tions, one Victor trap, (referred to as a small trap), and two Museum
Special traps, (referred to as large traps), were used. At even-numbered
stations, three Museum Special Traps were used. This arrangement of
traps was an attempt, with a limited number of "small" traps, to more
adequately sample the smaller species of small mammals. Ryel and Howe
(1953), suggested that Museum Special traps are somewhat large for sampl-
ing smaller species of Sorex.

The small mammal populations in the three cover types were studied
separately. However, the animal populations in both "browse conditions"
in each cover type were sampled at the same time. Trap lines were set up
and traps placed but not set until all four lines were ready. Traps on
all four lines were set the same day. They were visited daily, (every
2h hours), in the same order each.tiae and as near to the same time of
day as possible. .After three days of trapping, the traps were moved to
a new cover type as recommended by Hardy (l9h5).

For bait, the following ingredients were mixed well together:

2 pounds melted beef fat

2 pounds peanut butter

2 pounds raisins ground with
2 pounds oatmeal

1 pound paraffin ( home canning type)

ill..- Janina-J”

. _

 

16

All specimens trapped, were labeled in the field as to trap location
and size of trap. Then the animals were placed in individual paper sacks
for later study. In the laboratory, the following measurements for .11
small manuals were made: total length, length of tail, length of hind
foot, and length of ear (from notch). Sex and age, as judged by external
characteristics were recorded. In general, the specimens were identified
from skin characteristics and measurements, but skulls of questionable
specimens were saved for later, more positive identification. Skin char-
acteristics did not always permit the separation of the young of some
species such as W leucoggg and W maniculatug gracilig,
but these species could be identified by skull examination.

Identification and measurements of other specimens trapped were

recorded on the same data sheet as those for small mammals.

17

RESULTS
Vegetation
Treatment 9; Data

In.most species of trees, yield of seed is related to stem size
and crown cover. In the present study many*mature species were present
that never reached the larger diameters and.yet.produced large quantitie
of seed. Thus it seemed that separation of stems on the basis of dia-
meter would not be justified. For this reason, total numbers of stems
were used and tree data were converted to a per-acre basis so that the
various cover types studied could.be compared as well as making com-
parisons'within each cover type.

A list of vegetation.used as food by deer was prepared.by Atwood
(l9hl), and supported in.part by Trippensee (l9h8) and Duvendeck
(l952).$ince all species of shrubs in the aUIhs studied were found on
the food list, all species were lumped together for comparison.by stem
counts.

Vegetation on.herb plots was seperated into two classes for study
of occurance. Vegetation which could be, was counted by stems. That
such as mosses, grasses, and ferns, which was not easily tallied by
numbers of stems and heights, was recorded only in cover classes. These
classes were converted to average per cent coverage for comparisons. The
number of plots in which each species occurred.was expressed as per cent

frequency of individual cover classes.

18

Vegetation in Area "A:

Chiniquare test for tree data from cover type "A" presented in
Table No. III showed "good" and "poor" areas to be similar in so far
as stem counts were considered by size classes. Overstory cover was
found to be very similar in both conditions for type "A". Shrub cover
was found to be greater in the "good" area than in the "poor" area.
Although more stems had been "used", in the "good" area more were
present there than in the "poor" area. Slrubs were more numerous in
size class "A" and less numerous in classes "b" and "c" in the "poor"
area where fewer animals were caught. The average cover of grasses,
ferns, and mosses (Table VIII) was greater in the "poor" area than
in the"good" area, but stem counts of the other herbs (Table VII)
showed the "good" area to contain more total stems than the "poor"
area.

A total view of the two conditions,not involving statistical
tests, showed overhead cover to be about the same but slightly more
cover of shrubs and herbs in the good area. Both areas were of types
with the "good" area slightly wetter than the "poor" one. Because
of the uneven tapography,the forest floor was not very homogeneous.
Some sample vegetation plots occurred on areas flooded with water.
For this reason vegetative values shown by the data were not con-

sid ered very reliable.

19

FIGURE h

GROUND COVER,FOREST FLOOR, AND COMPOSITION OF VEGETATION FOUND IN "HEAVILY
BROWSED" CONDITIONS OF COVER TYPES "A" AND "B".

Note the nearly bare forest floor and general lack of deer browse.

 

2. 7_/—\’~—\ VM‘g A\‘~\‘,—\x\“

20

FIGURE 5
GROUND COVER, FOREST FLOOR, AND COMPOSITION OF VEGETATION FOUND IN "LIGHTLY
BROUSED" CONDITIONS OF COVER TYPES "A” AND "B".

Note the abundant herbaceous growth on the forest floor and noticeable
amounts of Cedar and other available deer browse.

 

 

Vegetation _i_n Area :23:

Chi-square tests for tree data from cover type "B" presented in
Table IV showed more total trees present in the "poor" area than in the
"good" area. Thus if seed production were based on numbers of tree stems,
and small mammals related to that alone, it would seem that the "poor"
area could support more small mammals than the "good" area. Shrub cover
in all size classes was also found to be more numerous in the "poor" areas
than in the "good" area but greater numbers of those in the "poor" area
were browsed. The average cover of grasses, ferns, and mosses, and also
the ntnber of stems of other herbs were greater in the "poor" area than
in the "good" area.

This type was similar to type "A" but not quite as wet. Still, many
plots were flooded. The ”good" area was not as homogeneous as the "poor"
area. Part of one of its trap lines passed through a small area in which
cover was present but food species for the small mazmnals studied were
sparse. The water level in the "good" area was higher than that found in

the "poor" area.

Vegetation _i_r_1_ Area :91

Although stem counts for trees in this area, (Table V), showed more
trees over four inches d.b.h. present in the "poor" area than in the "good"
area, total stems were nearly the same in both areas. More shrubs of all
size classes were present in the "poor" area than in the good area, but
at the same time more of the shrubs were "used" in the "poor" area in pro-

portion to those in the "good" area. The average cover of grasses, ferns,

 

r ”3:43:01:
.

 

22

FIGURE 6

GROUND COVER, FOREST FLOOR, AND CQ’IPOSITION OF VEGETATION FOUND IN
"HEAVILY BROUSED" CONDITIONS OF COVER TYPE "C"

Note that although ground cover was fairly abundant,(from the summers
growth), shrub cover, used by deer during winter months, was quite sparse.

4" / I .4 ~

1”“ p,.\/-/l,J~J—_.f—\r-4,/~IAN—

W

 

r-v-I-sr Jar/fl

, _~~.\_my~/\K-H\ N,f\>MAA\—fl.~v\n\ . .M

23

FIGURE 7
GROUND COVER, FOREST FLOOR, AND COMPOSITION OF VEGETATION FOUND IN PART
OF THE “LIGHTLY BROUSED" CONDITION OF COVER TYPE "C"

Note that ground cover is fairly sparse in this area. Shrub cover
in adjacent areas was greater than that shown here.

Deflflk.

My.“ "1.: , ,
.,.

v
'1

 

A _ r... We , ._, A _
“ WV I _, /..’—~ r,“ / /_’ va -4 U.”— x4 J MVVIJVW «

E
<
3
f
5
fi
3
r
s
i
E
f

2b

and mosses, and the total number of other herb stems was greater in the
”good" area than the "poor" area.

This type differed from.those previously mentioned in that it was
considerably dryer and contained more hardwood species. In general,

ground cover was best in the "good" area. There was a difference in gen-

eral appearance between the "good" and the"Poor" areas of this type. The I-R
"good" area consisted of drier soils partly of the Newton Sandy loam type 3

and partly of Rifle Peat. The topography at the “poor" site was very un- % .
even and had many small ridges caused by up-rooted trees. The"good" site i”4

did not have as many of these ridges. Sampling indicated that the "poor"
site had a greater number of species of trees present although in smaller
total numbers than the good site. This variety in species might be thought
favorable to small mammals.

Mammal Data

O

The trap catch by lines shown in Table I is listed only as "small
mammals" but the species composition of the total catch is shown in
Appendix.A. An anabysis of variance, (Table II), was carried out for
the small mammal trapping data, using a logarithmic transformation,

(xv - log (x + 1), (Snedecor,1950). The results indicated highly sig-
nificant differences between good and poor areas and differences sig-
nificant at the five per cent level among types of areas, using a comb-
ined error term. .A combined error term was used here because the inter-
action.appeared negligible in the anabysis.

Actual small mammal populations of each area could not be determined

by the sampling procedures. According to Dice (1938), no single statist-

25

ical sample, unless it be very large could give more than a rough app-
roximation of the size of the population from which it were drawn. Leraas
(1938) and Stickel (l9h6) found that population figures based on a three-
day snap trap catch were exaggerated, but Dice (1938) and Jameson (l9hh)

' stated that such trap lines gave reliable indications of relative abund-

ance of species in different areas. Hayne, (l9h9), discussed the estimat- $1?'
ion of population size from the decrease in rate of capture as animals are ;
removed. This method was not appropriate in the present study because of E
smallness of catch on.most lines. El

TABLE I

SHALL HAHHAL CATCH BY LINES.ACCORDING TO BROWSE CONDITIONS AND COVER
TYPE (Species Composition is shown in Appendisz)

Browse Conditions

Poor Good
Cover Type Line Line Total Line Line Total
1 2 3 h
A 1 o 1 h h 8
B 0 1 1 2 6 8
c 3 6 9 12 h 16
Total for lines h 7 18 1h

Total for Conditions 11 32

 

“34.3 «A nip-15 "a _.

 

26

TABLE II

ANALYSIS OF VARIANCE CF LCIEARITHHS OF NUMBERS (PLUS ONE) OF SIAM. MANUALS
CAUGHT ON 12 TRAPLINES.

(Using (Using a

 

 

fli‘ifii #7—- "'"T'
'

ELLE—1' :1
’4
._1

Degrees Sum Interaction Combined
of of Mean Error), Error Term)
£01122 £13212: miles 32222. F F
Total 11 1.333
Type 2 .hm .220 5.7 h.9 as
Good vs Poor 1 .529 .529 13,8 11.7 as
Interaction
[Type x (Good vs Poor)] 2 .077 .0383 .8
Error 6 .287 .0118
(Combined Error) (3) ~ _ .363 .oh5

 

Note:- Computations in the above table were carried out to four decimal
places and then rounded off to three decimals under Sum of Squares
and.lean Square, and to one decimal under "F".

1* indicates significant differences

as indicates highly significant differences (see following discussion)

In reference to Table 11 showing analysis of variance of the small
mammal data, the combined error term.yields the best estimate of error

of measurement, the mean square for combined error being 0.0h5 with a

standard deviation of 0.21. The approximate 95 per cent confidance limits

of average catch in "good" areas are 2.9 to 7.6 animals per line and in

the "poor" areas are 0.5 to 2.3 animals. Following the reasoning of Hin-
sor and Clarke (l9h0),(See also Ball and Kayne, 1952 and Snedecor,l9h6),

27

with a standard deviation in logarithms of 0.21, the antilogarithm 1.63
would indicate that the catch has been multiplied (or divided) by 1.63.
Thus one standard deviation in‘ the logarithm corresponded to a percent-
age standard deviation or coefficient of variation of 63 per cent in the
catch.

With the information on standard deviation, it is possible to esti-
mate the number of samples (trap lines) needed for confidance limits of
a stated size assuming the variability of random samples to be the same
as that of the present non-random samples. It would be possible to view
the information in still another m. If we desire to know the size of
sample necessary to allow us to discriminate between two areas that dif-
fer in catch by a stated amount, we must consider the two kinds of errors
inherent in any such test for a difference, as discussed by Walker and
Lev (1953). Rejecting a hypothesis when it is true is called error of the
first kind. The statistical level of significance is the probability of
making an error of this kind, alpha error. The probability of accepting
a hypothesis when some alternative to the hypothesis is true is- called
error of the second kind or beta error.

In the present study, the ivpothesis tested could be stated thus,
"The two kinds of areas are the same with regard to average catch",
recalling that we assume catch to be related to, but not identical with,
Population level. Error of the first kind would be deciding that a dif-
ference exists where there is really none, while error of the second kind

would be failing to detect real difference in catch.

 

‘

2
_=!

 

 

28

with error of the first kind set at 0.05 and error of the second kind
at 0.20 then computation has been made of the number of traplines necess-
any to detect population differences at the two levels. If one population
' were twice the size of the other, then eight trap lines in each sample
should be adequate. However, if one were only fifty per cent greater than
the other, 23 such lines would be needed in each sample. The greater
sample number here reflects the better information.

It is apparent from the data in this study that the few trap lines
used did not provide a very powerful test for differences between areas.

There is a real question in such a study as this as to whether one
should be satisfied with detecting probable major differences, or whether
one should expend.many more hours of labor to detect lesser differences.
For example from the experience gained in this study, it is judged that
to compare two kinds of areas in similar swamp types, using 23 trap lines
in each condition, would involve about five and one-half weeks of labor
for one individual.

It does seem probable that the small mammal populations studied were
greater under the "good" browse conditions than under the "poor". There
is less definite evidence that the catches reflect real differences among

the cover types chosen for study.

’5' '1

AT?

1.!

u.
a

29

 

 

.12....
A? New 83
- .. - m - - - Na- .. mam - mama .......... mm- -ams- -amam .. - - - mumaym whom
............................... N... .. .. .. .. .. .amummusmammw
N N mN 3N m 2 om :88 325
m N :25 manners.
N a song
8 Nmo 83¢ suing—m
N N :9: .833.
N a beam 833
m a o 35 use
ma ms: N8 R m: a: «68 £33 Suntan
a NN mm a S 2 83am x88
N m a as :2 :88
2 Nm 4% N 3 a: be 3:8
n «B 5285.
Ids...” INA.“ 10.3 4.4 a [NAN Iona INN.
.adé - team 5 “3.55 :36 - «£05 a nuts-B
.38 8.2 as .8 8.328 8mm“ 8.89m

 

 

 

 

 

mg amassed a 8 .5. mg no £5 .88.? 88?. E amen mmfi mo «was:

33%

3O

 

836:8 .3 H38.

2

8H
2
2

man

was INN? 6.. 4 4|.”
.53. - mass 5 39.55

900m

omfi

4N a: 32

m 3 an

m m m

a 2. m2
3 m3
m 2

m 8 93
m m

3“" NTH 034 “la
£56 1 mucus—H 5 Banana
0006

.84 as .3 833:8 8.55

 

:83 3:5
5&4 2:28...
x333.

834 uoaxuoom

human outflow

3%: com

.38 3:5 Esauoz
832x88

a: 525

 

mouuoaw

mama 28.5: a 8 ..m.. mat. HEB mo mama :62. 92 .589. E mam was: ac Ems.

>H Bah

2w .
“I’ll. l” t
ls
. . 1 <. . lowlyglhir

 

 

31

 

3% o 2 8.29.8 3 ~38.
2 3 2.“ Sp 3 :3 4% on.» an :38.
- - -n ..................................... a.“ “a?“
m mm mm m 3 :83 8:5
m on m. mmH mm m: t. 5&2 @5358.
2. 8m 83¢ 828%
m 8: .5
m m 3 mm m o 3%: 3m
8 .506 a:
m m4 m2 .380 32: 823qu
m m =8 cam 8233:
2 x82»: £38m
2 m duo 3258
m R m3 83am x38
m cm 3 :3 :38
m mm mm mm mm a; 5.28
a an. 50232
afld. .3 fl 3 wan M3 WA 3
£56 - «ANN 5 uuuoaaa £56 .. uuwwmw 5 .8358 832w

00.3 23 mo £33280 338m
gig gm 4 B :0: ME $00 «8 mg .308: 024. 580.. E” 3.8 E m0 mug

 

 

> an.

 

32

m8
«mm

owe

unnasz
H38.

 

 

E S 3 «S N a H 0

8n o2 S :8 N a H m

2. m moo «H N a H 4
Jul In: It... smug

096.5 9: :93:

nuuuaflo swam 130g» uaoau ocugaiyh
unnusn

museum Ho acouuuucoo uopm

am
:3
pm

 

.3855
«308

 

no mm 0 mm a a m 0

03 mm 9 we a a m m

40m 0mm new mm” a a m 4
llmaflmmml

lwl Im—l lwl 8393 as .3852

3330 33 039.3 2.3» 0538“.
3&5:

85m no 3333600 U000

 

 

onhHono

gagmflmfimaoommdmlgnamgfihm

H> mans.

33

TABLE VII

STEH{COUNTS.0F'HERBS -.ADJUSTED’TO 2h PLOTS

Trapline

Number

A 1 a 2
A 3 & h
B 1 a 2
B 3 a h
c 1 a 2
c 3 1 h

1h91
1656
1022
328
375
656

Total Ntmber of Stuns
Diameter in inches

.1;;g_ 13" - 1n dbh 19521
75 18 ‘ 1581.
1&1 15 1812
110 20 1152
ho 73 hhl
22 9 ho6
25 1 682

Median percentages fbr cover classes were as fbllaws:

Cover
Class

T

1
2

3

Percent
Cover

0-10
10-33
33-66
66-100

median %
Cover

5;o%
21.5%
h9.5%
83.0%

Average per cent cover in the fbllowing_three tables was determined.by

use of the fblloving formula:

Average % Cover - Sum of Median Percent

ludber of quadrats

”1h
:4

3b

TABLE VIII

COMPARISQIS 0F IDSSES, GRASSES, AND FERNS IN "Goon" AID "POOR"
CONDITIONS OF TYPE "A", USING BIG-IT PLOTS, .001 ACRE IN SIZE
FOR EACH CONDIT 1w

 

 

 

Species Frequency m
Cover Average
Class Cover Class Total Average #
T l 2 - Value Cover Class
113.119.29.111 ens A2.(§«;r_censi£ien272 _. .. .. .. _ .. - .. .. .. _. 4. _. Z - -
percent percent percent
Calamgrostis canadensis 13 1. 1.
mar“ feline-mas 1'3 3.
Onoclea sensibilis . 13 6. 9.
Hnium mtatmn 13 13 11.
Ptilun crista castrensis 13 10.
gleam spp. 38 .31- 52-
Total average per cent cover for “poor“ conditions: 62.

Osmunda cinnamomea 13 l.
Osmunda regalis - 13 6. 7.
him mtatm 13 3.,
Pleurozium schema ' 13 . 3.
m spp. 38 13 29. 35.

Ed‘s-.nuga.’ '1 1.. - A _ .
. q
‘

35

TABLE IX

COMPARISWS OF mosses, GRASSES, AND FERNS IN "GOOD" AND "POOR" CONDI-
TIQIS OF TYPE "B", DENSITIES BASED ON 18 PLOTS, .00025 ACRE IN SIZE
FOR EACH CWDITION.

 

 

 

 

2m
Species Frequency Averaqe Total Average ‘3:
Cover Class* Cover Class Cover Class 53*

-1L—-l———§_y;l. ._JEE£EE ___ :

Lisamessfil ans §2_Teosr_csnéit~.iens 3. ___________ 5
percent percent percent ;

Calamagrostis canadensis 17 08 08 17 21. 21. ;
Dryopteris felix-mas 25 08 3. g ., 4'
Onoclea sensibilis 17 1. ii
M cinnamomea 08 0.11 1M4
ygcgggdium spp. 17 8.
Mnium rostratum 08 h.
M 22.21.2222 ~ 17 17 58 60- 72.
Zola; 93281151: £02 Spoor ".6an 111.2%: ___________ 9 1-11
922.11%; §3_a2d_BE 199.09. 2°2d1t10281= ____________
Calamaggostis canadensis 17 22 06 17 22. 22.
giggpteris felixrmas 11 O6 O6 6.
Onoclea sensibilis 17 ' l.
Osmmda regalis 11 06 11 16. '
Pteridium gggilinum 17 06 O6 10. 33.
M91112 app. 11 1 .
3111-1321214: castaway. .. .02 .02 .11 .1411 - .. £15.. ._ .. - .112...
102:1 $118.31: 101'. 1'. oos'_<=2n21212n2: ____________ 09. ._

*Per cent frequency cover classes; and.Average cover class values
in the "poor" areas above were based on 12 plots instead of 18.

36

TABLE X

COHPARISCNS 0F PDSSES, GRASSES, AND FMS IN 'GOOD" AND "POOR" C(llDI-
TIQJS OF TYPE "C", USING 2h PLOTS, .00025 ACRE IN SIZE, IN EACH
CONDITION.

 

 

Spec ies Frequency m
Cover Average
Class Cover Class Total Average ‘3-
T 1 g 3_ Value Cover ClasL
LIB.W£3_EJ; £9.81. _C2_lgogr_cgn_c_i_ .. 2“ 3 ................
percent percent percent
Calanagrostis canadensis 25 21 13 0h 15. 15.
Mteris {elm-93g 08 Oh l.
Pteridium milinun 08 h. S.
Hetemnim Holdanianum 011 l.
momdium spp. 08 08 08 13.
Hnium mtatum 08 1 o.h
Hniun rostratun Oh 0h 11.
Pleuros'ium Schreberi Oh 01: h.
Ptilm crista Castrensig oh oh oh oh 7.
Total density for "poor" condition: 119.1:
Ens: Eli-lit? EBIaEdICE IéOE EOEdItIOh'sIC - I I I I I I I I I - I :
Calamagrostis canadensis 13 13 21 25 3h. .
maria felix-m_a_s_ 13 oh 2. '
Onoclea sensibilis Oh 0h 1. 3.
gtemhyllim Holdanianua 011 3.
Lyggmdim spp. 08 0.14
Mniun Matt!!! 08 is.
him rostratun 0h 1.
Pleurozium Schreberi Oh 0h 6.
Ptilum crista castrensis O8 21 21.
Tetrgphis pellucidum 0h 2. 37.11

Total density for "mg” conditions: 7&4;

 

 

 

 

37

DI$USSION

The first thing that should be discussed is the terminology used by
the investigator. The title of the thesis is "Small Mammal Catches Under
Two Intensities of Deer Browsing". Those Species of small mammals used in

this report included: Clethrionoms e i, Blarig brevicauda, 5.9.38

. -'.-Th-‘O_—J

cinereus, Sorex alustris, W leucgpgs, and Peromscus maniculatug
939111 . Other "larger" small mammals such as Glaucoma sabrinus, _.

 

.. .1!

Tamiascurius hudsonicus, and 13%; striatus, although caught, were not J
used for comparisons because of the lack of adequate sampling with the
size traps that were used.

Careful field examination of the "browse conditions" of various areas,
and examination of data from vegetative quadrats seemed to indicate that
the lack of understory browse material was actually due. to a combination
of the past use of browse material by deer and also the effect of the age
of the stand and self-pruning. These observations agree with those of
Duvendeck, (1952), who found that self-pruning significantly reduced the
amount of available browse on white cedar. This added factor may have had
some effect upon the lack of conclusive differences in vegetation of “good"
and 'poor" browse areas.

In all areas, small maml captures were greater in "good" areas than
in "poor" areas. However, small mamal captures in the cover types ob-
served could not be related directly to the density of any one "layer" of

cover. Over-all estimates of cover also showed lack of correlation with

iii.

L.

.

 

 

38

small mammal captures. Tw0 areas studied had more over-all cover in
"good" areas while one area had more cover in "poor" areas.

Various foods are used by small mammals. Shull (1907), found that
the short-tailed shrew used snails, moles, beetles, and earthworms for
food. Hamilton (19111), reported the food of the long-tailed shrew to be
mostly insects, earthworms, caterpillars, millipedes, sowbugs, snails,
spiders, some vegetable matter, small salamanders, and mice. Townsend
(1935), Hatfield (1938) and Hamilton (19in) reported Peromyscus to be
mainly vegetarian in habits, although large numbers of insects were used
too. Foods used by Peromyscus, based on stomach analysis, cheek pouches,
and catches, according to Hamilton are listed in Appendix C. Hamilton fur-
ther stated that the food of Clethrims gm; consists of seeds of
M, Amelanchier, Vaccinium, plus insects, beetles, adult flies, lepi-
dopterous larva, centipedes, spiders, and fungi. From observations in
the present study, plentiful amounts of food species of small mammals
were present in most areas studied, (see Tables VII-X and Appendix C).

Small manual populations were judged on total numbers of species.
This technique was used because of the very small numbers of individual
species captured. However, it would be reasonable to expect species
differences amng small mammals. For example, total clearing eliminates
some and allows others to multiply. Also small ml catch does not
necessarily equal populations present. If, for some reason, areas in
the two browse conditions were not comparable except for the amount of

available browse, trapping results could be biased.

39

There may be reasons other than those measured in this study affect-
ing numbers of small mammals found in certain areas. However, small mam-
mals may simply prefer certain areas which the deer avoid, assuming that
"good" and "poor" mean "little used" and "much used" by deer. whatever

the reasons, there were differences in small mammal nmbers caught between
"good" and "poor" browse areas. This would indicate that both the investi-
gator and the small mammals are reacting to certain differences although
these differences may not be the same in both cases.

9

‘h .wv-a. ‘L— __ .. .;
.3; _

7

b0

CONCLUSIONS

1. According to an ocular estimate, a difference in the amount of
available browse in the areas was present. This difference was due to _‘
previous browsing by deer and also to the degree of selfepruning. :13?
2. Catches of small mammals differed significantly between "good"
and "poor" browse areas according to results of statistical analysis using - i
a combined error term. More small mammals were caught consistantly in if It
areas classified as containing "good" browse.
3. Features of the vegetation, as analysed, failed to show an obvious
correlation with catch of small mammals. This could well have been due
to very inadequate sampling of both vegetation and snail mammal papulations.
h. The variability in catch on various trap lines was large. A large
number of trap lines would be necessary to determine anything less than

major population differences in various browse conditions.

 

APPENDIX A

ANIMALS CAUGIT IN TRAPS

Note: Numbers in parenthesis behind common names
indicate numbers of each species caught.

Note: Common and scientific names of mammals are
listed according to Burt (19116); those of
birds are according to Peterson (1934?); and
those of amphibians according to Pope (19117).

Comon Name

Red-backed vole (6)
Deer mouse (7)
white-footed mouse ()4)
Snort-tailed shrew (12)
Hater shrew (h)
hsked shrew (10)
Northern flying squirrel (5)
Red squirrel (1)

01 ive-backed thrush (1)
Hood frog

Bull frog

Green frog

Leopard frog

Pickerel frog

111

Scientific Name

Clethrionoms 19925;;
Peromscug maniculatus

Peggyscus leucoggs

Blarina brevicauda
_S_g_r_9_t palustris
_S_gr_e_g_c cinereus
Glaucoms sabrinus

Tamiasc iurus hudsonicus

filocichla ustulata

Rana gylvatica
Ema catesbeiana

Rana clamitans
Rana pipiens
Rana glustris

APPENDIX B

VEGETATION PRESEUCE LIST

Note: Comon and scientific names of all vegetation are
listed according to Robinson and Femald (1938)

unless otherwise noted.

Tree Plots

eel-m .1192
Ash, black

Alder, speckled
Aspen, trembling
Birch, white

Birch, yellow
Cherry, pin

Cedar, Northern white
Dogwood

Elm, American

Fir, balsam

Hemlock, Eastern
hple, red

Maple, silver

QR, Northern Red
Pine, red

Pine, white

112

Scientific N23

Fraxinus nigra

11.92.; means.
Loggia; tremuloideg

21219. are.

2212.212. 122:;

P_t_'\_n;_us__ zmglvanica

111312. occidentalis

92121.2 app.

y_1_1_1_u_s_ americana

_A_b_i_eg balsamea

1:335 canadensis

.422: m

595; saccharinum

Quercus rubra var. borealis
£11315 resinosa

Pinus strobus

Comon Name
Serviceberry
Spruce, black

Tamarack

Alder, speckeled
Balsam fir
Black spruce
Black ash

Gray dogwood
Leatherleaf
Mountain holly
Northern red oak
Nine bark

Red maple

Red dogwood
Ribes
Serviceberry
Tamarack
Vacciniun

Wild raisin
mite birch

willow

APPENDIX B (continued)

Sarub Plots

143

Scientific Name
Anelanchier canadensis
Picea nariana

Larix laric ina

:b

lnu incana

 

 

Abies balsamea

 

Picea marina

Fminus ni._g__ra

_C_gg_nu_s_ miculata
Chamem cal ata
lemmthug mucmnata
Qgrcus m var. borealis
flysocgEus omlifolius
Ase; m

C__9__rn_u_s stolonifera
5.1.1293 spp.

Amelsnchier canadensis
_L§r__i__15 laricina
Vaccinium spp.

Viburnum cassinoides
m 9.1.119.

Salix spp.

lib

APPENDIX B (continued)

Herb Plotg

Scientific Names

9.22.9 app-

gggg balsama
W M
£2; gem

M udicoides
Calanagrostis canadensig
Clieggggdim album

 

992333 trifolia

CM canadengis
Clintoni‘a borealis
Chamaeda re calfiulata

Corning miculata
”Calm 221.2%

Chiogenes higpidula
Mteris was;
eron spp.
Eghorbia spp.
Egmtorium naculatg
Ell—‘9; trifidm
Hetemnim l'ioldaniamm (b) (Grev.)
Lycogodium spp.

5‘

dun menlandicua
iua rostratm Schrad.

§ .

 

l'him 29222.29 L-

 

Hitchella m

rhiantm canadeng
w nucronata
Onoclea flsibilis

Q_n__l_i_g Acetosglla

Osmmda regglis

9mg; cinnamnea

BELLE! M castrensis L.
Pteridium milinun

Pleurozium Schreberi (Brid.) Mitt.

lus tremuloides
Polyggnun spp. ?

Mole. minor
Rubus pubescens

Ribes lacustre

Ribes spp. 7

Solidago rugosa
Mboricgms alba

15

APPENDIX B (continued)

Scientific Names

§pirea spp. ? Thalictrium mlmamum

Sparting Michggxiang

Spiranthes lucida

 

§2hagnum spp.

Smilacina trifolia Mg spp.
Trientalis mnericana Viburnum spp.
Tetggphis pellucida Hedw. Viburnum cassinoides

Thu 19. occidental i 5

Note: Three species were not
identified.

APPENDIX C

FOOD LIST FOR PEROMYSGJS

(Based on stomach analysis, cheek pouches, and catches; compiled by

Hamilton, 19141. )

Mg canadensis

Cl intonia borealis
hianthemum canadense

Polygonum spp.
Coglus anericana

£5893 gr_a_.n_difolia
Ribs; anericanua
Fragaria viminiana
m app-

m virginiana
.1453; verticillata
99:33; miculata
Fraxinus americana
Hitchelia m
Sambucus spp.

116

Tang! canadens i s

Smilicina racemosa

Hedeola virginica
C333 ovata
Ostm virginiana

Caulgphzllun thalictrium
Amelanchier spp.

m app.

m mmlvanica
m gratina
mtiens spp.
Vaccinium canadenae
Solaniul spp.
Viburnum spp.

5.82% am

LITERATURE CITED

Allen, Durward L. 1938. "Ecological Studies on the Vertebrate Fauna of
a 500 Acre Farm in Kalamazoo County, Michigan" Ecological Monographs

8: 31:74:36.

Atwood, Earl L. 19111 "White-tailed Deer Foods of the United States."
Journal 9; Wildlife Management. 5 (3): 3114-332.

Barick, F.B. 1950. "The Edge Effect of the Lesser Vegetation of Certain
Adirondack Forest Types with Particular Reference to Deer and
Grouse." Roosevelt Wildlife axlletin 9 (1) Roosevelt Wildlife
Forest Experiment Station. New York College of Forestry at Syracuse
University, Syracuse, New York.

Bartlett, LR. 1950. Michigan Deer, Game Division, Michigan Department
of Conservation Bulletin. Lansing, Michigan.

Blair, U.F. 1938. "A Study of Prairie Deer-Mouse Populations in Southern
Michigan" American Midlend Naturalist 214: 273.305.

*___, 191:1. “Techniques for the Study of Small Maximal Popula-
tions." Journal 2f Mammalogy. 22 (2): 1148-157.

Innbach, C.A. 19111;. "A Ten-year Ecological Study of Adjoining Grazed
and Ungrazed woodlands in Northeastern Ohio. " Ecological Mono-
g_r_ap_h3 114: 255-270

Dice, L.R. 1931. "The Elation of Mammalian Distribution to Vegetation
Types" Scientific lbnthly 33: 312-317.

, 1938. ”Some Census Methods for hauls” Journal 2;:
Wildlife Manamnt 2 (3) : 119-130.

Duvendeck, J.P. 1952. "Some Effects of Deer Browsing on Iorthern
Michigan Forest Plants." Thesis for M.S. Degree, Michigan State
College, East Lansing, Michigan.

Elliot, J.C. 1952. "The Hwtosociology of the Upland Hardwoods of
Missaukee County, Michigan" Thesis for PHD. Degree, Michigan
State College, East Lansing, Michigan.

Hamilton, N.J., Jr. 19111. "The Food of Snll Forest Manmls in the
Eastern United States." Journal 9; hamlogy 22 (3) 250-263.

Hardy. Ross. 1915. "The Influence of Types of Soil Upon the Local
Distribution of Son: hmls in Southwestern Utah" W
Monogmhs 15 (1): 71-108. h

7

 

h8

Hatfield, D.M. 1938. "Studies on Rodent Populations in a Forested Area."
Journal 2; Mammalggy 19 (2): 207-211.

Hauser, 12.3. 1953. "An Ecological Analysis of the Isolated Prairies of
Newaygo County, Michigan. PhD. Thesis, Michigan State College, East
Lansing, Michigan.

Mayne, D.U. 19b9. "Two Methods for Estimating Populations from Trapping
Records". Journal gf'Mammalggy 30 (h): 399-h11.

Jameson, E.U., Jr. 19h9. "Some Factors Influencing the Local Distribution
and Abundance of Woodland Small Mammals in Central New York." Journal

52$ amalggz 3o (3) 221-235.

Leraas, R.T. 1938. "Ecological Distribution of Mammals in the Cranbrook
Area." Cranbrook Institute of Science, Bulletin No. 13. Bloomfield
Hills, Michigan.

Newbigin, M.I. 1936. Plant and Animal Geography Dutton, etc. Company
3rd Edition, New Yorfi.

Costing, H.J. 1953. 123 Study g£_Plant Communities. U.H. Freeman and Company.
SanFrancisco, California.

 

Parmelee, G.N. "The Oak Upland Continuum in Southern Michigan " PhD Thesis,
Michigan State College, East Lansing, Michigan.

Phillip, Paul. 1936. "The Distribution of Rodents in Overgrazed and
Nermal Grasslands of Centaal Oklahoma." Ecology 17 (h): 673-679.

Ryel, L.A. and D. Howe. 1953. "Small Mammal Studies". Game Division,
Heughton Lake Wildlife Experiment Station Report. Michigan Depart-
ment of Conservation. July.

Shull, A.F. 1907. "Habits of Short-tailed Shrew, Blarina brevicauda."
Sociegy of American Naturalists bl: h95-522.

Stickel, L.F. 19h6. "Experimental Analysis of Methods for Measuring Small
Mammal Populations". Journal g£_wildlife Management 10 (2): 150-159.

Tansley, A.G. and T.F. Chipp. 1926. "Aims and Methods in the Study of
Vegetation." The British Empire Vegetation Committee and the Crown
Agents for the Colonies.

Townsend, M.T. 1923. "Studies on Some of the Small Mamals of Central
New'York.” Roosevelt Wildlife Annals h (1): 1- 120.

 

Trippensee, R.E. 19b8. Wildlife Man ement Vol. I. Mcgraw-Rill Book
Company Inc. New‘York, (P. 9 .

119

Veatch, J.0. et. al. 19234. United States Department of Agriculture, Soil
Survey of Roscommon County, Michigan, Bureau of Oiemistry and Soils,
UOSODOA.

Veatch, J.0. 19m. "Agricultural Land Classification and Land Types of
Michigan,” with map. Michigan State College Agricultural Experiment
Station, Section of Soils, Special Bulletin 231. East Lansing, Mich.

, J.0. 1953. "Soils and Land of Michigan. The Michigan State College
Fess, East Lansing, Micfiig .

Weaver, J.E. and Clements. 1938. Plant Ecolo 601 pages, Mc Gray-Hill
Book Company Inc. New York, Toronto, Lon on.

 

SELECTED BIBLIOGRAPHY

Billington, Cecil, 19h9. Sirubs g Michigan, Cranbrook Institute of Science.
Bulletin No. 20, 2nd Edition.

. , 1952. Ferns 9_i_‘__ Michigan, Cranbrook Institute of Science.
Bulletin No. 32.

Burt, R.E., 19h6. The Manual: oLMichigan, University of Michigan press,
Ann Arbor, Mich-{Earn

Darlington, H.T., E.A. Bessey, C.R. Megee, and B.H. Grigsby, 1951. Some
rtant Michi an Weeds Special Bulletin 30h. Michigan State
0 ege grimgur eriment Station, East Lansing, Michigan.

Fassett, N.C. 19h0. A Manual 9_i_‘_ Agggtic Plants, McGraw-Hill Book Coapany,
Inc. New York, tondon.

Gleason, H.A., 1952. lie! Britton and Brown Illustrated Flora of the Eggh-
eastern United States _a__nd Id 1acent Canada. lew York Botanical-Edens,
EECC VOIWCSQ

Harlow and Harrar. 1937. Textbook 9_i_'_ Dendrologz , Mc Graw-Hill Book Company
Inc. New York, London, lst Edition.

Muenscher. U.C. 1936. K 3 3:2 MPlants, Ithaca, New York. Constock
Publishing Company c.

Peterson, Roger T. 191:7. A Field Guide £2 53 Birds. 2nd Edition
Houghton - Mifflin Conpany, Boston.

Pope, C.H. 19M. 19117. Ajphibians 59g ngtiles of _t_._h_e Chicago Area.
Chicago Natural History Museum. Chicago, Illinois.

Robinson, B.L. and M.L. Fernald, 1908. 9.1312 y_e_w Manual ifl‘
American Book Company, New York, Chicago.

Schneider, LF. 1951. "Preliminary Land Type Legend. Missaukee County,
Michigan,“ with nap, Michigan State College Agricultural Experiment
Station. (Field work in 19h3).

Snedecor, GJI. 1950. Statistical Methods. 14th Edition, Iowa State
College Press, Ames Iowa.

50

 

51

University of Illinois Agricultural Experiment Station. Weeds g£_£h£_North
Central States. Publication No. 36, Circular no. 715.

U.S. Department of Commerce. 195k. Climatological data For Michigan.

Weather Bureau, F.U. Reichelderfer, Chief. Volume LXIX, Numbers
6,7,8, and 9.

walker, H.M. and J. Lev. 1953. Statistical Inferences, Henry Holt and
Company, New York.

winsor, C.P. and G. L. Clarke. 19h0. “A statistical Study of Variation
in the Catch of Plankton Nets." Sears Foundation. Journal 22
Marine Research 3: l-3h.

 

 

“M”! Mr": Xi.
gawk?!” i‘Jj‘L E Y
WRW (ql

 

 

"Ililk‘i‘fliiiflii“iii“