ABSTRACT
INTERACTIONS BETWEEN SYMPATRIC MICROTUS AND SIGMODON
(RODEN'I‘IA: CRICETIDAE)
By

Max Robert Terman

Two species of grass eating, runway making myomorph rodents,

Microtus Ochrogaster and Sigmodon hispidus, were occupying a common

 

 

habitat (vegetationally dominated by bluestem grasses-—Andropogon spp.)

 

in central Kansas. Since these two rodents were living in the same

area, and since a decline in g, ochrogaster coincided.with a northward

 

spread of §,'hispidus, it was hypothesized that these two species were
interacting negatively with each other.

To test this hypothesis, a field and laboratory investigation was
carried out. It was hypothesized that the presence of Sigmodon in the
field would decrease mean body weights, change sex ratios, shorten sur-
vival rates, reduce mean residence times, increase range overlaps,
produce negative measures of interspecific association, decrease indi-
vidual and population movements, reduce home ranges, and reduce
trappability in MiCrOtus. In the laboratory, it was hypothesized that
Sigmodon would aggressively dominate and spatially exclude Microtus.

Six study plots were live trapped from April 16 through December 6,
1971 and then again on March 17 and April 14, 1972. In two plots all
MicrOtus were removed, in another all Sigmodon were removed, in the
fourth both species were removed and in two more plots neither species
was removed.‘ Sigmodon did not appear in the traps until late summer

which shortened the time of interspecific contact. The late appearance

Max Robert Terman

of Sigmodon reduced the between-the-plot comparisons but facilitated
the observation of single plots over time (before and after Sigmodon).

In the presence of Sigmodon, Microtus had a lower survival rate

 

and changed sex ratio in one of the two-species plots. Also, the two
species when left together were negatively associated. The population
movements of MiCrotus in the absence of Sigmodon were more widespread
than when Sigmodon was present. The number of captures per individual
Microtus dropped significantly after the appearance of Sigmodon in one
of the dual species plots.‘ Microtus also became less trappable in the
presence of Sigmodon. These results and the presence of wounds on
Microtus after the appearance of Sigmodon suggested a negative inter-
action between the two species in the field. The severity of the Kansas
winter may limit Sigmodon populations sufficiently for Microtus to
coexist.

Both genera were paired in a terrarium divided into two sections.
Sigmodon was clearly the dominant animal. The spatial distribution of
the two species was observed in a three-compartment laboratory arena.
Sigmodon excluded Microtus except under conditions of dense cover. In
a specially designed tunnel, the movements of Microtus were reduced by
a free ranging Sigmodon but not by a confined one.

The frequency of interspecific contact appeared to determine the
presence or absence of a negative interaction. A model utilizing the
contribution of various factors to the frequency of interspecific con—
tact was constructed. This offers a possible explanation of the rela-

tionship between Microtus Ochrogaster and Sigmodon hispidus. Also, a

 

 

model was proposed to explain the mechanism of competitive exclusion as

it might occur in myomorph rodents as a group.

INTERACTIONS BETWEEN SYMPATRIC MICROTUS AND SIGMODON

(RODENTIA: CRICETIDAE)

BY

Max Robert Terman

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Department of Zoology

1973

ax ACKNOWLEDGEMENTS

I wish to thank Dr. Rollin H. Baker, chairman of the guidance
committee, for his advice and encouragement in the completion of this
project. He not only provided much valuable knowledge of field work in
Kansas, but without his conscientious support while the author was off-
campus, this project could not have been completed. I also would like
to thank the other members of the guidance committee, Drs. James C.
Braddock, John A. King, and S. N. Stephenson. Dr. King provided valu-
able advice about rodent behavior and helped greatly in the analysis of
the data and the preparation of the manuscript.

Mr. Glen Wiebe and Dr. Lorin Neufeld of Tabor College, Hillsboro,
Kansas, aided greatly in the preparation and analysis of the data by
computer. I especially thank Dr. Neufeld for his help in writing com-
puter programs.

Dr. Webster Van Winkle of the Biomathematics Program, North
Carolina State University provided statistical advice. Dr. Bryan P.
Glass, Museum, Oklahoma State University, gave much valuable information
regarding rodent ecology. Dr. Richard Wiegert, Department of Zoology,
University of Georgia, graciously provided his data on rodent competi—
tion. Drs. S. D. Fretwell and C. S. Smith of the Division of Biology,
Kansas State University, and Dr. C. Richard Terman, Department of
Biology, College of William and Mary, set forth many valuable ideas

concerning the study. Dr. Dwight Platt, Department of Biology, Bethel

ii

College, Newton, Kansas, and Mr. Delbert Kilgore, Division of Cell
Biology and Physiology, University of Kansas contributed many valuable
comments about the particular rodents that were studied. Also, Dr. Walt
Conley and the graduate students in mammalogy, Department of Zoology,
Michigan State University offered many valuable comments. I wish to
thank Mrs. B. R. Henderson for her kind help in administrative affairs
throughout the duration of my doctoral program.

Many students in the Biology Department, Tabor College assisted in
the field work. I especially wish to thank Richard Wall for his invalu-
able help in securing a trapping area. Special thanks also go to Dr.

S. L. Loewen of Tabor College. Financial support was provided by a
grant from the Kansas Academy of Science and by awards from the Tabor
College Faculty Developement Program. Facilities for animal maintenance
and laboratory work were provided by the Department of Biology, Tabor
College, Hillsboro, Kansas. Computer facilities were provided by Tabor
College and the Associated Colleges of Central Kansas.

Finally, I wish to thank my wife, Janet, not only for her patience,
encouragement, and companionship throughout the study, but also for her

help in obtaining data and in all other aspects of the project.

iii

TABLE OF CONTENTS

INTRODUCTION
Objectives
Hypotheses
FIELD INVESTIGATION

Description of the Study Area
Plot Vegetation and Topography

Field Methods
Field Hypotheses
Analysis of Data

Field Results
Populations
Mean Body Weights and Sex Ratios
Survivorship
Mean Residence Times
Spatial Distribution
Microtus Range Overlap
Indices of Association
Movements
Migration Between Plots
Individual and Population Home Ranges

Trappability

iv

Page

10
13
15
15
26
27
27
30
30
37
39
44
44

49

Page

LABORATORY INVESTIGATION 51
Aggressive Behavior 52
Spatial Distribution 55
The Effect of Sigmodon on Microtus movements 63

DISCUSSION 71
Field Population 71
Mean Body Weights and Sex Ratios 73
Survivorship and Mean Residence Times 74
Spatial Distribution 75
Microtus Range Overlap 76
Indices of Association 76
Movements, Migration, Home Ranges, and Trappability 77
Aggressive Behavior 80
Spatial Distribution 80
The Effect of Sigmodon on Microtus Movements 81

INTERACTION BETWEEN SIGMODON AND MICROTUS 83

SUMMARY AND CONCLUSIONS 89

LITERATURE CITED 91

10.

ll.

12.

13.

14.

LIST OF TABLES

Weather data (1969-1972) taken from Marion Dam (8 km.
west of study area).

Percent coverage of common plant species found in each
plot.

Survival rates in Microtus and Sigmodon (per three weeks)
(1971).

Mean residence times (weeks) for Microtus Q: 1 SE).

Coefficients of interspecific association (C) for Microtus

and Sigmodon (all plots) (August-December 1971).

The mean number of individual Microtus per trap according
to the number of Sigmodon per trap (1971).

The median number of different traps and the median number
of captures per individual Microtus (1971).

The numbers of Microtus and Sigmodon invading or leaving
the plots (1971).

Mean home range statistics (A4) for Microtus individuals
in plots 1, 2, and 3 (1971).

Mean home range statistics (A4) for populations of

Microtus in plots 1, 2, and 3 (1971).

The median number of Microtus in each compartment per
observation under sparse cover (15 replications) and under
dense cover (13 replications).

Summary of nest location, water usage, and tracks in the
test chamber (all experiments combined).

The number of activity periods for individual Microtus
under no contact and contact conditions.

A tabular model of the Microtus-Sigmodon interaction:
trends to be expected.

Page

28

29

38

4O

45

46

47

48

6O

62

67

85

Figure

1.

10.

11.

12.

13.

LIST OF FIGURES

Distribution of Microtus ochrogaster and Sigmodon
hispidus in central United States.

 

Map of the study plots on the trapping area.

Minimum numbers of Microtus and Sigmodon on plot l (1971-
1972).

Minimum numbers of Microtus and Sigmodon on plot 2 (1971-
1972).

Minimum numbers of Microtus and the numbers of Sigmodon
removals on plot 3 (1971-1972).

Minimum numbers of Sigmodon and the numbers of Microtus
removals on plot 4 (1971-1972).

Minimum numbers of Sigmodon and the numbers of Microtus
removals on plot 5 (1971-1972).

The numbers of Microtus and Sigmodon removals on plot 6
(1971-1972).

Relative numbers of new and recaptured Microtus and
Sigmodon on plots 1 and 2 (1971).

Relative numbers of new and recaptured Microtus and the
numbers of Sigmodon removals on plot 3 (1971).

Relative numbers of new and recaptured Sigmodon and the
numbers of Microtus removals on plots 4 and 5 (1971).

Spatial distribution of centers of activity of individual
Microtus (Plot 1 - before appearance of Sigmodon).

Spatial distribution of centers of activity of individual

”§E£52£g§_and;§igmgdgn'(Plot l - after appearance of

14.

'Sigmodon).

Spatial distribution of centers of activity of individual
Microtus (Plot 2 - before appearance of Sigmodon).

vii

Page

11

16

17

18

19

20

21

23

24

25

31

32

33

Figure Page

15. Spatial distribution of centers of activity of individual
Microtus and Sigmodon (Plot 2 - after appearance of

Sigmodon). 34

16. Spatial distribution of centers of activity of individual
Microtus (Plot 3 - before appearance of and after removal

of Sigmodon). 35

17. The numbers of trapsites with captures of none, one, two,
etc. individual Microtus (before and after appearance of

Sigmodon - 1971). 36
18. Population movements of Microtus and Sigmodon on plot 1

(1971-1972). 41
19. Population movements of Microtus and Sigmodon on plot 2

(1971-1972). 42
20. Population movements of Microtus on plot 3 (1971-1972). 43

21. The experimental arena used to test for spatial
restriction. 57

22. The movement tunnel apparatus (A - no contact; B -
contact). 64

23. Activity pattern of Microtus alone, with Sigmodon con-
tact, and without'Sigmodon contact. 69

viii

INTRODUCTION

The prairie vole, Microtus ochrogaster (Wagner) and the hispid

 

cotton rat, Sigmodon hispidus Say and 0rd, are near the edges of their

 

respective ranges in central Kansas where they overlap. A field and
laboratory study of their relationship in a sympatric area was carried
out from April, 1971, through April, 1972. The purpose of this investi-
gation was to examine their relationship for a suspected negative
interaction.

The distribution of Microtus ochrogaster and Sigmodon hispidus is

 

 

shown in Figure l. Microtus ochrogaster has been a long time resident

 

of the Southern Great Plains, whereas Sigmodon hispidus is an invader

 

from the south, appearing in the central Kansas area about 40 years ago
(Cockrum, 1948). The northward spread of S, hispidus has coincided with

a decline in M. ochrogaster. For example, in Kansas Martin (1956, 1960)

 

and Frydendall (1969) found decreased Microtus population levels with

increasing Sigmodon populations. Microtus ochrogaster formerly was

 

trapped in north central Oklahoma but has not been taken there since
1957 and then only with a crash in the Sigmodon population (Bryan Glass,
personal communication). In South Carolina, Wiegert (1972) reported

that Microtus pennsylvaniCus lived and reproduced in field enclosures

 

when not in contact with Sigmodon hi3pidus. Odum (1955) noted a rare

 

occurrence of Microtus pennsylvanicus in Georgia following a low popu-

 

lation year for Sigmodon. Although the observed decline of M,

1

 

 

 

 

 

 

 

 

A - }
\.‘o ’ .{'-—...
\O
A - MICROTUS \_
Fb =- SIGMODON /"
C - ZONE OF OVERLAP '\ (- 0 430 L
’ \ ' Scale of Miles
.\)
l ,0 \
TUE 88

 

Figure 1. Distribution of Microtus ochrogaster and Sigmodon hispidus
in central United States (from Hall and Kelson, 1959; with
modifications from Easterla, 1968; Genoways and Schlitter,
1966; and Jones, 1960).

 

 

 

A
.o
‘ a
.
-
7
h
' ~
I 1 .‘ ‘
«
f
I. a
D
. " ‘ r
3‘; .
. ' o ‘ ".
‘ .
‘0
I‘-
< - '
‘$
.
I
u
' fl
..
“ ‘ a .
A ., ‘
s \.
‘ ~- ‘
- J‘
S
‘14
.. '2’.
V
f h .0...
Z
I
l .
n
1
V' I
r
‘.
' a
V .
v )
‘ i.
‘| ‘
sf 1 ’ ‘-. I
v
0 ‘ I
Q .
r . -.. ‘
I 3 . . ' Q
.' ‘fi _
‘ . \l
t’
‘\
."‘ p
A

. a .
I I . l"
: d . .'
‘ . I-— L o
I
I
I
+ ‘ ‘ 9 I! "
9
IL
|
l-
.
‘c‘
.
u
t
.
..
.
,-
v
.
'
n
_.

b
-4
.
Q
t
n
\- '
.. ~.-
'
..

ochrogaster might be the result of environmental factors having nothing

 

to do with the "newly arrived" Sigmodon in Kansas, these studies and
other observations (Baker, 1969, 1971; Dimmick, 1969; Fleharty and
Olson, 1969; Goertz, 1971; Hays, 1958; Whittaker and Zimmerman, 1968)
point to some kind of negative interaction between these two rodents.
These reports stimulated the questions which formed the basis for this

study.

Objectives

 

The objectives of the present study were (1) to investigate and

present data on the ecological relationship between Microtus ochrogaster

 

and Sigmodon hispidus in their common habitat; (2) to test hypotheses on

 

the presence or absence of a negative interaction between these two
species in (a) the field situation; and (b) in a controlled laboratory

situation.

Hypotheses

 

Baker (1969, 1971) has contended that when two species of grass
eating, runway making rodents of the genera Microtus and Sigmodon live
in the same area, MiCrOtus is generally replaced by the cotton rat or

there is spatial separation of the two species. Since M. ochrogaster

 

and_§. hispidus were resident in the mixed grass habitat of Marion
County, Kansas, this area offered an excellent opportunity to study the
alleged incompatibility of the two species.

Because of their apparent niche similarity (Calhoun, 1945; Svihla,

1929), it was postulated that the two rodent populations would negatively

affect each other. To test this concept the following two hypotheses

were developed:

(1)

(2)

It was hypothesized that Sigmodon would decrease mean body
weights, change sex ratios, shorten survival rates, reduce
mean residence times, increase range overlaps, produce nega—
tive measures of interspecific association, decrease indi-
vidual and population movements, reduce home ranges, and
reduce trappability in Microtus.

It was hypothesized that Sigmodon would aggressively dominate

and spatially exclude Microtus in laboratory experiments.

FIELD INVESTIGATION

Description of the Study Area

 

 

Trapping was conducted on a grassy pasture owned by Robert Navrat
located two miles north of the city of Marion in Marion County, Kansas.
Marion County is near the border between two biotic districts,the Mixed
Grass Plains and the Osage Savanna (Cockrum, 1952). The study area it-
self is dominated by big bluestem (Andropogon gerardi , little bluestem
'(Andr020gon scoparius), switchgrass (Panicum virgatum), and Indian grass
'(SorghaStrum nutans). The pasture in which the study area was located
was placed in the Soil Bank for ten years prior to the study, according
to owner Robert Navrat, and had not been disturbed by grazing, plowing
or burning during this time.

Cockrum (1952) lists Marion County as containing mammals of the
Great Plains Mammalian Distributional Area and mammals of the Central
Lowland Mammalian Distributional Area. Small mammal species observed or
captured on the study area were Sylvilagus floridanus, Lepus californi—

cus, PeromySCus'maniculatus, Reithrodontomys megalotis, Mus musculus,

 

MicrOtus ochrogaster, Sigmodon hispidus, and Cryptotis parva.

 

Peromyscus maniCulatus, R, megalotis, M. ochrogaster, S, hispidus,
and_§..pa£!a were the most common species observed.

Notable features of the Kansas climate are frequent and abrupt
changes. Summers are usually warm often.with periods of high tempera-
tures and low humidity. Winters are drier than the summers. The

5

average annual precipitation for the county of Marion is 782 mm. and
the average annual temperature is 13.5° C (56.6° F). Table 1 gives the

weather data for the study period.

Plot Vegetation and Topography

 

Moderately grazed pastures bordered the study area on all sides.
To the west and south there were fence rows with sparse to dense woody
vegetation. On the northern border there was a small creek with asso—
ciated woody vegetation. Grassland similar to that which composed the
study area was on the east border.

Six study plots, each 60 meters square, were chosen for this exper-
iment (see Figure 2). Each was placed at least 30 meters away from any
other plot or adjacent boundary. The percent coverage of the common
plant species (see Table 2) was determined for each plot during Septem-
ber by the line intercept technique described in Cox (1967). A Gossen
Luna-Pro light meter was used to measure relative cover by comparing the
amount of light penetration reaching the ground level beneath the vege-
tation in each plot (Mossman, 1955). The relative cover corresponded
directly with the percent coverage of bare ground in Table 2 (l2§°
plot 1 was most dense, then plot 3, and so on).

Plots 1 and 3 were located in a drainage area which was relatively
moist. This accounted for the relatively dense cover in these plots.
Plots 2, 4, and 5 were located on gentle lepes (Figure 2). Plot 6 was
on higher ground which gently sloped down into the drainage area con-
taining plots 1 and 3. The plots were not exact duplicates of each
other in terms of plant cover and species density. However, all plots

were part of a homogenous prairie area which occupied 80 acres, and the

momsmm .cmuumnamz .oouoaaoo mo usmEuumaon .m .D .moH>Hmm mung HmuaoacouH>cm Scum «

 

AHVHm ANva

ono ono
Asymma ono
s OH
0 a
m a
ma mm
m s
cm s
H s
m m
o a
.omn .>oz

ono
Aflvmm
ono

mH
0H
0H

HNH
ow
«OH

0H
NH
NH

ouoo

*.Ammum hvaum mo umo3 .ex wv 8mm aOHumz aoum amxmu ANNoH I momHv mumo Honummz

«N
MN
mN

Nm
omH
ow

HN
ON
ON

.ummm

5N
Nm
wN

MH
oN
NH

«N
wN
0N

.ws<

om
mN
mN

Ammmsmx .aomaHnousm um :mxmuv Avov

ooN
No
wnH

«N
0N
wN

sass

mN
«N
«N

«0N
NNH
wNH

mN
NN
HN

mash

Amvmom

Hovo

ono

ono

a¢vHOHAevamm AmHvoN

Amv~oa

ono

Amva

Nan
HNaH
ommH
NomH

wnbouw no mxmv mo Hones: was A.aov HHMmsoam Hmuoa

HN RH
«N mH
«N «H
mN NH
Hm «o
wHH ««
on ¢w
omH mNH
wH mH
oH «H
ON NH
wH NH
hm: .um<

NH
5

Nm
0
.NN

Q<f€h

.uwz

lan

.50

.nmm

Fir-IF)

Nan
Han
onaH
momH

N.OH um ousumuomaou HHom coax

N
RN
H

NNmH
Han
oan
momH

A.EEV aOHumuHmHomum sow:

WI
ml
ml.

.GMH

Nan
Han
oan
aomH

Avov musumuomaou cum:

.H mHam“.

Table 2. Percent coverage of common plant species found in each plot.

Species

Andropogon
gerardi

 

Panicum

virgatum

Andropogon
scoparius

 

Sorghastrum
nutans

 

Agrostis
alba

Agropyron
smithii

 

Sporobolus

8.8261“

Aristida
app.

 

woody plants

Bouteloua
curtipendula

 

herbs

Paspalum

circulare

bare ground

cover rank*

Plot 1

40.0

35.0

9.6

9.3

1.5

1.9

0.9

0.7

0.2

0.3

0.3

0.2

’ 0.1
100.0%

 

Plot 2

42.5

4.3

35.0

1.3

1.5

1.7

1.5

1.7

0.2

2.1

0.7

0.2

7.3

100.0%

 

 

 

Plot 3 Plot 4
54.0 40.1
25.5 0.5
10.3 44.6

2.0 4.5
0.8 0.1
1.5 0.2
0.8 1.5
3.3 0.5
0.3 0.3
0.2 0.9
0.5 0.7
0.5 ---
0 4 6.0

100 0 100.0%
4

 

Plot 5 Plot 6
59.6 35.0
15.3 10.5
19.8 43.5

2.0 1.2
0.3 0.2
0.7 0.4
0.6 0.3
0.1 0.5
0.2 0.2
0.1 0.5
0.5 0.5
0.2 -——
.6 7.0
100.0% 100.0%
3 5

 

* Determined by a light meter, ranked from most dense to least

dense.

Rank also related to amount of bare ground.

differences in cover, ground litter, soil type, and species composition

were not extensive.

Field Methods

 

On each 60-meter-square plot a live trap was placed at every 15
meter interval. Every plot thus contained 25 live traps of the type
described by Fitch (1950). Every three weeks traps were baited in late
afternoon, checked early the following morning, closed during the day
and opened again in late afternoon for four continuous days. Oats were
used for bait and the traps were partially wrapped in aluminum foil to
provide some overhead protection for the captured animals.

All plots were live-trapped from April 16 through December 6, 1971
and then again on March 17 and April 14, 1972. The catch of Microtus
and Sigmodon at each plot was manipulated according to the procedure
given in Figure 2. Other species of small mammals captured were identi-
fied and released.

The following information was recorded from each prairie vole and
cotton rat in each plot throughout the entire period of trapping (6300
trap nights): Species, sex, individual number (by toe-clipping),
weight, trap station, breeding condition, and presence or absence of
wounds.

Weights were taken to the nearest gram using a spring balance with
an alligator clip from which animals were suspended by their tails while
being weighed (see Krebs, 1969). Breeding condition was determined by
the position of the testes in males and by the condition of the vaginal
opening, lactation, and pregnancy in females. Juveniles, subadults, and

adults were categorized by weights. In Microtus juveniles were animals

10

.1; 22 grams, young adults were between 23 and 32 grams, and adults were
‘5: 33 grams (Krebs, 1969). In cotton rats juveniles were 1; 46 grams,
young adults were between 47 and 111 grams, and adults were it 112 grams
(Sealander and Walker, 1955). After information was gathered on each
rodent, the animal was marked (unless a recapture or one to be removed)

and released at its point of capture.

Field Hypotheses

 

The six field plots containing the two species were trapped utiliz-
ing selective release or removal procedures (see Figure 2). In two
plots all Microtus were removed, in another all Sigmodon were removed,
in the fourth both species were removed and in two more plots neither
species was removed. Once the field data were collected, comparisons
were made between the plots and within single plots. Between-the-plot
comparisons were made difficult by a late appearance of the cotton rats.
Sigmodon did not appear in the traps until late August which prevented a
year-around comparison as was originally intended. The late occurrence
of the cotton rats did, however, facilitate the single plot comparisons
since each Microtus population could be observed both before and after
the Sigmodon invasion. It must be stated that the latter comparisons
can introduce such variables as seasonal influences which can not be
controlled. For this reason, much field data can only be considered as
suggestive and not confirmative.

Where possible, the field data were subjected to statistical tests
in an effort to confirm or reject specific hypotheses. Much of the
data, for example population levels, were not testable by presently

known procedures. These types of data were not used to confirm or

11

small creek

    

  
    
 

10o slope

  

Plot 2
trap and

  

trap and release both
Microtus and Siggodon ’

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

    
 
 

 

 

 

“:* lane
70 80 I
/ o
3 10 slope
Plot 3
0 remove N
10 #sloPe/Ia ‘* Sigmodon A
60 T
Plot 4 '1 10° “'—
4‘3 remove 55 ‘slope
A Microtus I ’
Plot 6 I
remove
both
30 PIOt 5
remove
Microtus
100 85 Scale
I I f 1 cm. = 26 meters
I slope, I
‘140 fence row
4, gran o c: O ..--
I,
I
I
Figure 2. Map of the study plots on the trapping area

(distances in meters).

IA.-

 

12

reject hypotheses but were discussed because of their descriptive values
and the information they provided about the interspecific situation.

In the two field plots where both species were trapped, marked and
released (see Figure 2) it was expected that the Sigmodon populations
would cause certain changes to occur in the Microtus populations. When
compared with the plot from which Sigmodon was removed (plot 3) it was
hypothesized that the former two plots would have significantly lower
Microtus survival rates, lower mean residence times, more range overlap,
negative measures of interspecific association, decreased individual and
population movements and home ranges, and would be less trappable. I
hypothesized these things because of the suspected niche similarity and
dominance of Sigmodon (Martin, 1956). It should be noted now that the
late appearance of the cotton rat handicapped this design by reducing
the time of interspecific contact.

Because of the sudden appearance of Sigmodon, it was anticipated
that each plot would experience temporal changes in the respective vole
populations. In the plots containing both species, it was hypothesized
that the Microtus populations would change significantly in mean body
weights, sex ratios, mean residence times, range overlaps, trap capture
rates, individual movements, and individual and population home ranges.
The Sigmodon removal plot should experience changes of lesser magnitude.

The plots from which only Microtus were removed were expected to
provide data on Sigmodon survivorship in the absence of Microtus and on
interspecific association. It was hypothesized that the voles would not
affect cotton rat survival and that these plots would have neutral in-
dices of interspecific association. It was expected that plot 6 (both

species removed) would provide data on the total cotton rat and vole

13

populations available to the plots. If the animals of either species
were common and mobile on the grassland area as a whole, then there
should be a constant removal of that species from plot 6. Also, this
plot was hypothesized to have neutral indices of interspecific associa-
tion because of the removal program.

This trapping regime was also expected to produce useful data on
population levels, recruitment, immigration and emigration, and spatial

distribution in single and dual species conditions.

Analysis of Data

All field data were placed on paper and magnetic tape and were
analyzed on a PDP8 (Digital Equipment) computor through the time-sharing
facilities of the Associated Colleges of Central Kansas and Tabor
College.

Population estimates were made using the direct enumeration tech-
nique (Krebs, 1966), which reveals the minimum number of animals on each
area at time_£. The resulting figure is a summation of two counts:

(1) the actual number of individuals caught at time 5; and (2) the numr
ber of individuals marked previous to time_£'but caught after time_£,
and not at time E (Peterson, 1970).

Survival rates were calculated by computing the percentage of
animals surviving between successive trapping periods (per three weeks).
The mean residence times were computed by averaging the length of time
from first to last capture for all individual rodents in a given time
interval. Mortality here is assumed to include emigration or death.

Dispersion was determined by computing the center of activity for

each multi—captured individual (more than 2 captures) and for each

14

population in a given time interval. The center of activity is an
average x-y coordinate of all points of capture for an individual or
population. Centers of activity were then placed on grids representing
particular plots so that individual and population movements could be
observed in varying situations.

Another index of population dispersion was gained by counting the
number of trap sites which captured none, one, two, etc., individual
rodents. These data were then put on graphs to reveal population dis—
tributions in different situations.

Cole's (1949) coefficient of interspecific association was used to
ascertain the degree to which Microtus and Sigmodon associated with each
other in each plot. This involved the calculation of a coefficient of
association for data arranged in a 2 x 2 contingency table. This co—
efficient was then tested for statistical significance by using Cole's
specially designed Chi-square test formula. In addition, correlation
coefficients were calculated between the two species by using the number
of individuals of both species caught at every trapsite.

To get a further insight into species association, traps were first
categorized according to the number of different cotton rat individuals
caught. Then the mean number of Microtus individuals was determined for
each trap of a particular category.

The last measure which gave an indication of species association
was the percentage capture rate for each trap station where both species
were taken. The vole capture rate at a trap should decrease when that
trap began to capture Sigmodon.

An index to population movements was gained by noting the monthly

movement of population centers of activity for Microtus and Sigmodon.

15

Other indices of movements included the mean number of different traps
and the mean number of captures per individual Microtus. Individual and
p0pulation home ranges were also indices to movements and were calcu-
lated by the computer using a method developed by Jennrich and Turner
(1969). This method is based on the determinant of the covariance
matrix of the capture points and can measure noncircular as well as
circular home ranges. The resulting figure is an A4 statistic which is

an index to the size of the home range.

Field Results

 

Populations

 

Preliminary trapping during October-December 1970, indicated that
both Microtus and Sigmodon were present on the study area and that there
seemed to be some degree of spatial separation between areas of similar
habitat.

When experimental trapping commenced on April 16, 1971, Microtus
was present in all plots except plot 2. In the next trapping period,
however, Microtus was trapped in plot 2 in numbers comparable to the
other plots. Sigmodon, on the other hand, was not captured until the
August trapping period in plots 1, 4, and 5 and not until the September
trapping period in plots 2, 3, and 6. This late arrival of Sigmodon
allowed the analysis of Microtus populations over time, 1.2. before
Sigmodon and after Sigmodon. Figures 3 through 8 show the population
levels for each species on each plot for the entire study (1971-1972).

The Microtus populations and removals showed a September—October
decline in all of the plots. This coincided with the first sign of

cotton rats. The voles appeared to recover in all plots by December

16

20 1" - - — — MICROTUS

" ———- SIGMODON

 

 

15 -
P
1
8’2 ’\
FIJ 10 - I ‘
E I \
.- \
’ \
’ \
.. ’ \
- I
I" I \‘ /
’1 \\ ’I \ o .
5 _ O/ \, \\ I’
\ I
\/

 

p
D
b
.-
b
L
h-
.-

 

A M M J J J A S 0
MONTHS

Figure 3. Minimum numbers of Microtus and Sigmodon on plot 1
(1971-1972).

 

17

 

20 - -—-~—- MICROTUS
"' -'—--—- SIGMODON
15 " o
H
I \
F' I \
b I ‘ ,
I \ ,I
I ‘ I ‘
I ‘\ I ‘
~ I ‘ I A ”.
a t l i ‘ I ” \
m 10 " I K I l .v"
I ‘ ' l ’
)- --—-o \ I I
\ I 1 I
_ ‘! \ I
| I
. \ ,’
' I
w 1 I
I /'
5 ~ \ ../ -
\
L l I
\ I
\
L \
_ \
L 1 t L L 1_ I 1 1 I l i 1 4 1
A M M J J J A S O 0 N D J F M
MONTHS
Figure 4. Minimum numbers of Microtus and Sigmodon on plot 2

(1971-1972) .

NUMBERS

20

15

10

18

 

 

 

 

- - — —— MICROTUS
* SIGMODON REMOVALS
I-
P
p
I- .
I\‘
I
L I ‘
I |
. I ‘
I \
I \
.- I \ .
I ‘ I"
I- - ‘ I \
I \ \
/ I ’
I \ I
I \ I
p-
rj J J I l 1 1
AMMJJJASOONDJFMA
MONTHS

Figure 5. Minimum numbers of Microtus and the numbers of Sigmodon
removals on plot 3 (1971-1972).

20

15

NUMBERS
|'-'
O

19

 

- -—-—MICROTUS REMOVALS
n SIGMODON
p
b
I
b
I-
I- l'\
. I \
/ \
_ . I ‘
n o " \ I \
/ \\I \. I
I- I ' \ I
I \ I
k .’ \’

 

I A 1 J 1 1 n\ l

A M M J J J A
MONTHS

 

Figure 6. Minimum numbers of Sigmodon and the numbers of
Microtus removals on plot 4 (1971-1972).

NUMBERS

 

20

 

 

 

20 .. --—- MICROTUS REMOVALS
I. -———--- SIGMODON
b
L
15 v
p
H
p
h-
10 P
'I
‘ O
\ ,I
L l ‘\
‘ \ 9..--.\
1 I | I ‘
t I I \ I ‘
| \ \
5 F' \ I | I, ‘
_ I ’ \ I ‘\
\ I, \ I ‘
p ‘ ’ I, ‘ .
‘I
. I I
\ I
\ I
F 0---o o
1 I I 1 IV 1 J
A M M J J J A
MONTHS
Figure 7. Minimum numbers of Sigmodon and the numbers of

Microtus removals on plot 5 (1971-1972).

-\

I"

20

15

NUMBERS

 

 

21

.. -- — — - MICROTUS REMOVALS

SIGMODON REMOVALS

 

plot 6 (1971-1972).

b
l
P I
I\
r- , \ I
’ \ ‘——--——.—~-
’ \x I’
b l 0‘ I
I \ I
-. ' \ I
\ K ‘
\ ’ \ I
h \ 's I, \ I
\ [I \ \ I
_ \I \’ \ I,
or. .
L A I n L 1 l 1 \‘1\1’/j\1 V j I V n J
A M M J J J A S O O N D J F M A
MONTHS
Figure 8. The numbers of Microtus and Sigmodon removals on

22

and increased thereafter, except in plot 1. Resident Sigmodon popula-
tions increased during September and October and then decreased in
December. Sigmodon removals from plot 3 decreased steadily, while those
from plot 6 were minimal, never more than one per period. Plots 2, 4,
and 6 showed a complete cessation of vole captures in October.

There were only two Sigmodon captured in the spring of 1972, both
in plot 1. Microtus populations were higher in the spring of 1972 than
in the spring of 1971 in all plots except plots 1 (where neither
Microtus nor Sigmodon was removed) and 5 (where Microtus, but not
Sigmodon was removed). Nine Microtus were known to overwinter, four
in plot 1, four in plot 2, and one in plot 3. One Sigmodon over-
wintered (plot 1). Plots 3 and 6 were the only ones that had captures
of Microtus juveniles in 1972.

Figures 9 through 11 reveal the numbers of new and recaptured
Microtus and Sigmodon on plots l-S. From these records, an indication
of recruitment and immigration can be gained. In plot 1, the recruit-
ment of voles and cotton rats increased steadily from September through
October. In November both rates decreased substantially. Microtus
recruitment increased again in December.

In plot 2, Microtus recruitment dropped to zero in early October.
In October, Sigmodon had no new captures and one recapture. From late
October, Microtus exhibited renewed recruitment and steady population
growth. From late October through December, Sigmodon had a constant
recruitment but little population growth.

In plot 3, Microtus recaptures dropped to zero in early October.
The population, however, remained constant until it declined in Decem-

ber. In December no Sigmodon were captured in plot 3.

NUMBERS

23

 

    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

22
F Plot 1
20 _
V
18 P NEW MICROTUS
V
16 L NEW SIGMODON g
14 . D ..
RECAPTURED MICROTUS .
12 _ .:
4 @ RECAPTURED SIGMODON r‘ '1 ‘
10 y. '3;
a - IE
" 2
6 D II
, I .
4 5 , % §
r 5 ‘ ~
2 a a a , § §
0. d 4 5 Ah
A M M J J J A s o D
16 P MONTHS
14 ,. _’
Plot 2
12 F
10 ' F"
b
r“ r

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8 . TV

6 b | ‘

4b 35'.

2 .

o - _ fl 3 _
A M O 0

M J J J A S D

2‘ llllflfl

Figure 9. Relative numbers of new and recaptured Microtus and
Sigmodon on plots 1 and 2 (1971).

NUMBERS

24

Plot 3

7
NEW MICROTUS

“
16 SIGMODON REMOVALS

14

1

[1I RECAPTURED MICROTUS
12

10 . ——

co
1

ON
I

 

 

 

 

 

 

 

 

 

 

MONTHS

Figure 10. Relative numbers of new and recaptured Microtus
and the numbers of Sigmodon removals on plot 3
(1971).

NUMBERS

12

10

16

14

12

10

 

 

 

 

, Plot 4

. MICROTUS REMOVALS

N
P . NEW SIGMODON
L- E RECAPTURED SIGMODON
F a

AM M J JJ ’A’s’oio

P Plot 5
F

 

‘\\7‘\t‘\\‘\\f‘§]

 

 

 

 

 

 

 

 

 

P

A

 

M J J J A
MONTHS

Figure 11. Relative numbers of new and recaptured Sigmodon

and the numbers of Microtus removals on plots
4 and 5 (1971).

26

The Sigmodon population in plot 4 showed sporadic recruitment.
Microtus removals declined in late October and recovered in November.
The Sigmodon population in plot 5, like that in plot 1, showed constant
recruitment from September through October when it also fell off to a
lower, more constant rate. Removals of Microtus declined steadily from

late October in this plot.

Mean Body Weights and Sex Ratios

 

The mean body weights and sex ratios (Joule and Jameson, 1972) of
Microtus were calculated for plots 1, 2, and 3 for the trapping periods
before and after the appearance of Sigmodon. For plot 1 (neither
Species removed) the Microtus mean body weights before and after
Sigmodon were respectively 35.2 grams and 35.8 grams. For plot 2
(neither species removed) these values were 36.7 grams and 32.4 grams.
The values for plot 3 (Sigmodon removed) were 34.0 grams and 38.3 grams.
Even though plot 3 differed from the two-species plots, none of the
differences were significant.

The sex ratio for Microtus in plot 2 (neither species removed)
changed significantly after the appearance of Sigmodon (p‘<.05,
x2 = 5.6). The pr0portion of females increased in the presence of
Sigmodon. Microtus sex ratios in plots 1 and 3 did not change.

Since Sigmodon did not appear until late August, it was not pos-
sible to get the response of the Microtus populations to the cotton
rats over what can be considered a reasonable amount of time. Also,
the cotton rats which did appear were mainly juveniles and subadults

which further reduced the likelihood that the voles came under the

necessary interSpecific pressure to reveal substantial declines. Other

27

than providing an indication of a negative interaction, the above data
on Microtus were experimentally inconclusive.

The late occurrence of the cotton rats indicated a low overwinter
survival for Sigmodon in Kansas. Preliminary trapping in 1969 and 1970
also revealed small or non-existent Sigmodon populations in the early
spring and summer. As will be discussed later, this is an important
point to consider since this might allow Microtus the time it needs to

complete its reproductive cycle.

Survivorship

 

Table 3 illustrates survivorship in Microtus and Sigmodon. The
Microtus survival rate was lower in the presence of Sigmodon (plot 2)
than in the plot without Sigmodon (plot 3) (p‘<.05, Mann-Whitney U
test). It was difficult to conclude, however, that the cotton rats were
the cause of this difference because of the low numbers of Sigmodon
which were caught in plot 2 (see Figure 4). Also the vole survival was
lowest on plot 2 for the year and for the interval before the appear-
ance of Sigmodon. This suggested that some other factors probably lim-
ited Microtus survival rates before Sigmodon appeared and that the in-
fluence of Sigmodon on Microtus was no more than slight. Microtus did
not affect the survival rates of the Sigmodon in any apparent way. The
plots from which Microtus were removed had the lowest Sigmodon survival

rates 0

Mean Residence Times

 

Table 4 gives the mean residence times for all Microtus in plots 1,
2, and 3. There were no significant differences when the dual species

plots were compared to the Sigmodon removal plot. The appearance of the

28

Table 3. Survival rates in Microtus and Sigmodon (per three weeks)

 

(1971).
Survival rate
Plot Species Time for all animals

1 Microtus year .59
2 Microtus year .36
3 Microtus year .57
1 Microtus before

Sigmodon .60
2 Microtus before

Sigmodon .49
3 Microtus before

Sigmodon .62
l Microtus after

Sigmodon .57
2 Microtus after

Sigmodon .18
3 Microtus after

Sigmodon .67
1 Sigmodon year .72
2 Sigmodon year .67
4 Sigmodon year .46
5 Sigmodon year .48

Table 4. Mean residence

Before Sigmodon

After Sigmodon

29

times (weeks) for Microtus Q: 1 SE).

Plot 1
6.6 _-_i-_ 2.7

6.3 + 1.7

Plot

6.9

3.6

l+

2.6

1.7

Plot 3

6.9 + 2.9

5.4 + 3.0

30

cotton rats did not significantly alter the residence times of the voles
when their use of the plots was compared over time. The values for
plot 2 were the only indication that the Sigmodon occupation had any
effect on the voles. Perhaps a longer exposure to a larger cotton rat

population would produce significant results.

Spatial Distribution

 

Figures 12 through 16 reveal the spatial distribution of centers
of activity of individual Microtus and Sigmodon in plots 1, 2, and 3.
It appeared that there was a change in the spatial distribution of the
centers of activity of the voles when the plots were compared over time.
It is debatable whether this change was a result of the Sigmodon inva-
sion or other factors related to the response of the Microtus popula-
tions to seasonal influences. This change in dispersion in plot 3
(Sigmodon removed) also could have been the result of undetermined
factors. Again, a longer period of exposure to cotton rats might have

produced more useful data to look at this alleged interaction.

Microtus Range Overlap

 

An index to the range overlap of a population can be gained by
counting the number of trapsites which have captured two or more indi-
viduals (Metzgar and Hill, 1971). Figure 17 illustrates these data for
plots 1, 2, and 3. When the dual species plots were compared to the
Sigmodon removal plot, there were no significant differences in the vole
range overlaps (X2 = 2.77, p‘<.30). When the findings from the plots
were compared over time, the Microtus in plot 2 were the only ones to
change significantly in range overlap after the appearance of Sigmodon

(l8 trapsites to 7 trapsites, X2 = 4.8, p‘(.05). In plot 3, the

31

 

 

 

 

 

~o———- 15 m-———. AA h :x
M
M
M
R4! M
M M
M

 

 

 

 

 

M - MICROTUS

Figure 12. Spatial distribution of centers of activity of
individual Microtus (Plot 1 - before appearance

of Sigmodon).

 

32

 

 

 

 

 

 

 

._ 15 m _Lt‘ M
5 IN
IM
4% ~M———«
s
(Er M S ‘5
M
M
E
M
s

 

 

 

S L
T s
M = MICROTUS
5 = SIGMODON
Figure 13. Spatial distribution of centers of activity of

individual Microtus and Sigmodon (Plot 1 - after
appearance of Sigmodon).

33

 

 

 

M
M N
M M
M
r M
M

7
.3

 

:3

 

 

 

 

 

 

M . MICROTUS

Figure 14. Spatial distribution of centers of activity of
individual Microtus (Plot 2 - before appearance

of Sigmodon).

34

 

1"“
;+——-15 m «———h‘

 

3:

4

 

 

00
U I

 

 

 

 

 

 

M - MICROTUS.
S = SIGMODON

Figure 15. Spatial distribution of centers of activity of

individual MicrOtus and Sigmodon (Plot 2 - after
appearance of Sigmodon).

35

 

 

 

 

1 f !A

x a a .
SIGMODON
REMOVAL
AREA

0 X‘
CDC
0
X

 

 

 

 

 

 

 

C) = Microtus before Sigmodon
X = Microtus after Siflodon

Figure 16. Spatial distribution of centers of activity of
individual Microtus (Plot 3 — before appearance of
and after removal of Sigmodon).

NUMBERS OF TRAPSITES

36

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

BEFORE SIGMODON AFTER SIGMODON
20 _ 20
Plot 1 P Plot 1
15 - l5 .
10 ." 10 r-
5 r ___7 ‘ 5
o l o
O l 2 3 4 5 6 O,— l 2 3 4 5 26*
20 F 20 _
15 _ Plot 2 15 _ Plot 2
10 P 10 _
5 F—f’fl_ fl 5"
O . - O
O 1 2 3 4 5 6 O 1 2 3 4 5 6
20 r_ 20 F
15 . Plot 3 15 _ Plot 3
___1‘
10,- 10 b
5 L- l-
o 7 t
O 1 2 3 4 5 6 O l 2 3 4 5 6

NUMBERS OF INDIVIDUALS AT EACH TRAPSITE
IN EACH PLOT

Figure 17. The numbers of trapsites with captures of none, one, two,
etc. individual Microtus (before and after appearance of

Sigmodon-‘l97l)-.

37

difference was fairly large (X2 = 1.60, p‘<.2) but the range overlaps
did not differ enough to be significant. This index is probably af-
fected by radical changes in population density (Metzgar and Hill, 1971)
and since plot 2 experienced a population decline in Microtus in

October, the value was suspect.

Indices of Association

 

Coefficients - Table 5 gives the coefficients of association (C)

 

for Microtus and Sigmodon. Plot 1 had the only significant negative
association as revealed by the number of traps which captured none,

one, or both species (see Cole, 1949). The Sigmodon were most dense in
this plot which probably resulted in more frequent interspecific con-
tact. Plot 2 did not have many Sigmodon at any time which probably
resulted in the neutral index revealed in this plot. Plots 3, 4, 5, and
6 were expected to have neutral coefficients due to the removal of one
or both species. This reduced the opportunity for interspecific contact
on a prolonged basis. In contrast to the sign negative association
index (Cole, 1949), no plot had a significantly negative correlation
coefficient. The lack of conformity between the two tests may be due

to a difference in their ability to reveal differences in trap utiliza-
tion by the two species.

Trap and Capture Indices of Association - When the mean number of

 

 

Microtus individuals per trap was compared to the number of Sigmodon
individuals per trap, it was noticed that there were no significant
differences in the numbers of voles at a trap due to the presence of
two or more cotton rats. The reduced numbers at highly frequented

Sigmodon traps and the near significant values for these comparisons

38

Table 5. Coefficients of interspecific association (C) for MiCrotus
and Sigmodon (all plots) (August-December 1971).

m _c_
1 -1.0 *
2 0.10
3 0.08
4 0.09
5 0.03
6 -0.07

 

C is computed from the number of traps catching none, one, or both
species. Values run from -l.0 to +1.0 (negative association
to positive association).

* Significant negative association.

39

(p<.1, Mann—Whitney U test) did, however, suggest that a negative
interaction could be suspected. Table 6 gives the values for the
selected plots.

When the capture rates of the stations taking both Microtus and
Sigmodon were compared over time, plots 1 and 2 had significantly
different percentage capture rates of Microtus after the cotton rats
appeared (p'<.05, Mann-Whitney U test). The traps in plot 3 (where
Sigmodon was removed) were not different in their vole capture rates.
It is reasonable to at least partly attribute this change to the cotton
rats since any seasonal factors would have caused the capture rates in
plot 3 to change also. Ecological differences in the plots may have

produced these effects but such seems unlikely.

Movements

Figures 18 through 20 show the monthly movements of the Microtus
and Sigmodon pOpulations. These data were analyzed by measuring the
distances moved by each vole population in each plot. Before the ap-
pearance of Sigmodon, there were not significant differences between
the dual species plots and the Sigmodon removal plot (the monthly move-
ments were between 2 - 16 meters). After the appearance of Sigmodon,
the population of Microtus in plot 3 (Sigmodon removed) moved signif-
icantly greater distances (6 - 52 meters) than the voles in plots 1 and
2 (p<(.05, Mann-Whitney U test). This indicated that the Microtus
population in plot 3 was more wide-ranging. This difference might have
been due to the absence of'Sigmodon or to intraspecific factors oper-

ating in the individual Microtus populations.

40

Table 6. The mean number of individual Microtus per trap according
to the number of Sigmodon per trap (1971).

Mean number of individual Microtus per trap

.ElEE _:5 1 Sigmodon _§: 2 Sigmodon
1 3.50 1.73
2 1.11 0.60
4 0.72 0.75

5 1.0 0.75

 

41

 

 

 

 

 

I1 9". 3"“ '0.

 

 

 

 

4 June
5 July
6 July
7 Aug
8 Sep
9 Oct

 

 

 

 

10 Oct
11 Nov
12 Dec
13 Mar
14 Apr

(1972)

 

Microtus population movements

 

-..-- Sigmodon p0pulation movements

Figure 18. Population movements of Microtus and Sigmodon on
plot 1 (1971-1972). (Locations numbered according

to legend.)

 

42

 

 

 

 

 

 

 

 

 

»
B
I
I e '
I
I
I 7
I 7. .13 ,
I, g/ flN.If A"
I fl \ f
I
I
' » No
l
u: ". A05
\
. \!'i
4 4 June 10 Oct
5 July 11 Nov
6 July 12 Dec
7 Aug ~13 Mar (1972)
8 Sep l4.Apr
9 Oct (no data)

 

 

 

Microtus population movements
___”_ Sigmodon population movements

 

Figure 19. Population movements of Microtus and Sigmodon on
plot 2 (1971-1972). (Locations numbered according
to legend.)

43

 

,§___ 15'm ___)

  
 

 

 

 

 

//:IGMODON

REMOVAL
AREA

 

 

K 18

 

 

 

 

V
4 June 9 Oct (no datab
5 July 10 Oct
6 July 11 Nov
7 Aug 12 Dec
8 Sep 13 Mar (1972)
14 Apr

 

 

 

 

 

 

Microtus population movements

Figure 20. Population movements of Microtus on plot 3 (1971-
1972). (Locations numbered according to legend.)

44

An index to individual movements is given in Table 7. The median
number of different traps and the median number of captures per
individual Microtus were not significantly different between the plots
when Sigmodon was present. The only significant change occurred when
the plots were compared over time (each plot before and after Sigmodon).
The number of captures per individual Microtus dropped significantly
after the occurrence of Sigmodon in plot 1 (p«<.05, Mann-Whitney U
test). The number of traps per individual Microtus dropped also but the
difference was not significant (p'(.07). If the Sigmodon had been more
dense in plot 2 perhaps this plot would have shown a difference compara-

ble to plot 1.

Migration Between Plots

 

The plots were invaded or vacated by Microtus and Sigmodon as shown
in Table 8. Migration between plots characterized 13 percent of the
recaptured voles and 16 percent of the Sigmodon individuals. Twenty-six
percent of the Microtus migrants were active in the April-July period;
74 percent were active in the August-December period. These migrants
were released or removed according to the experimental treatment of the
plot that they invaded. The increased migration of the voles during the
autumn periods may have been due to seasonal influences rather than the

occurrence of the cotton rats.

Individual and Population Home Ranges

 

Tables 9 and 10 give the individual and population home range
statistics (A4) for Microtus. When compared over time the major changes
in the individual home ranges occurred in plots 1 and 2. These changes

were not significant however, and the home ranges for the individuals

45

Table 7. The median number of different traps and the median number
of captures per individual Microtus (1971). (Numbers of
individuals are in parentheses.)

Median number of different Median number of captures
traps per individual per individual
Before After Before After
Plot Sigmodon Sigmodon Sigmodon Sigmodon
l 2 (19) l (27) 2 (l9) 1 (27)
2 1 (26) l (15) l (26) 1 (15)
3 l (17) l (11) l (17) l (11)

 

The only significant difference was in captures per individual
in plot 1 when compared before and after Sigmodon.

46

Table 8. The numbers of Microtus and Sigmodon invading or leaving the
plots (1971).

 

Invaders Departures
_Plg£ Microtus Sigmodon Microtus Sigmodon
l 6 6 6 3
2 7 4 2 5
3 O 1 9 --
4 3 3 -- l
5 O 0 -- 4

 

47

Table 9. Mean home range statistics (A4) for Microtus individuals
in plots 1, 2, and 3 (1971).

 

 

.Plpp .N Before Sigmodon N_ After Sigmodon
1 4 17.7 4 11.5
2 6 16.3 1 0.0
3 2 8.3 2 6.7
All 12 15.4 7 7.5 *

 

N equals the number of individuals captured 4 times.
* Significantly different after Sigmodon when compared to
before Sigmodon (Mann—Whitney U test, see Siegel, 1956).

Table 10.

Plot
1

2

48

Mean home range statistics (A4) for populations of Microtus

in plots 1, 2, and 3 (1971).

Before Sigmodon

 

34.12

35.80

33.80

After Sigmodon

 

35.22

40.30

44.10

 

49

had to be pooled for all three plots before the ranges were signifi-
cantly reduced over time (p4<.05, Mann-Whitney U test). The numbers of
individuals were too few to allow valid comparisons between the plots.
Table 10 reveals that there likewise were no significant differences
between the plots and over time using the population home ranges. This
failure to find significant changes makes it difficult to derive con-
clusions regarding the effect of Sigmodon on the home ranges of
Microtus. The pooled comparison using individual home ranges was comr
plicated by many uncontrolled variables. The population home range
increase in plot 3 was nearly twice that in the other plots but the

difference was not significant.

Trappability

 

Trappability (Krebs, 1969) was computed for the Microtus and
Sigmodon populations. For the year Microtus had a trappability of .80;
Sigmodon had a trappability of .85. In plots 1 and 2, the vole trap—
pability was, respectively, .82 and .86 before Sigmodon and .76 and .62
after Sigmodon. In plot 3, the trappability was .81 before Sigmodon
and .90 after Sigmodon. The Microtus in plot 2 were the only ones
which differed significantly from those in plot 3 in trappability during
the cotton rat occupation (p'<.05, Mann-Whitney U test). If Microtus is
less trappable in the presence of Sigmodon then they should have become
less trappable in plot 1 also since it had the highest cotton rat den-
sities. Unless the plot vegetation (dense in plot 1, sparce in plot 2,

see Table 2) affected the Microtus-Sigmodon frequency of contact“ it is

 

difficult to explain this discrepancy in the results.

50

All captured Microtus were examined for the presence of wounds,
especially the large slashes indicative of a Sigmodon encounter (Terman
and Johnson, 1971). During autumn, plots 1, 2, and 4 had voles which
showed large slashes on the back and hindquarters. Such individuals,
however, were rare in the samples. No wounds were noticed before the
cotton rats were on the plots.

Field results of experimental studies on interspecific interaction
are rarely conclusive because of the many environmental factors which
remain uncontrolled. These field results were complicated by seasonal
factors, plot differences, and changes in population densities. If a
negative interaction occurred, it probably was not well detected be—
cause techniques for monitoring the individual interactions were not
available. Such individual interactions appear to be the mechanism of
competition (Grant, 1970). In order to investigate these interactions

a laboratory study was undertaken.

LABORATORY INVESTIGATIONS

Many of the variables which complicate the field experiment can be

controlled in the laboratory. For this reason, the alleged Microtus-

Sigmodon interaction was examined in the laboratory. The hypotheses

tested were suggested by the field results of the present study and by

the observations and comments of other investigators.

(l)

(2)

(3)

If interspecific aggression between Microtus and Sigmodon is
the mechanism favoring cotton rats over voles (Baker, 1971;

Martin, 1956), then Sigmodon should be aggressive toward

Microtus and dominant in the interspecific encounters. I

hypothesized that Sigmodon would actively attack Microtus and
would be aggressively dominant in interspecific pairings.

If Sigmodon limits the spatial distribution of Microtus
(Baker, 1969; Fleharty and Olson, 1969; Terman and Johnson,
1971), this phenomenon should be observable in a laboratory
arena. I hypothesized that voles would reduce their use of
areas occupied by Sigmodon. Further, if the amount of habitat
cover (complexity) affects this relationship (Crombie, 1946;
Krebs pp a1., 1971; MacArthur, 1972), then I hypothesized that
dense cover would enable Microtus and Sigmodon to occupy the
same area.

If the Microtus-Sigmodon interaction lacks interspecific

 

recognition and avoidance (Baker, 1971; Brown, 1966; Calhoun,

51

52

1963; Jameson, 1947), cotton rats should affect the movement
of voles only when Sigmodon comes into actual physical contact
with Microtus. I hypothesized that vole movements would not
be affected by a confined cotton rat but would be affected by
a free ranging cotton rat.

Should most of these questions be answered, a model of the

Microtus—Sigmodon interaction could then be constructed which might

 

help explain the numerical and spatial observations on these two species
throughout their geographical ranges.

Rodents used for the laboratory experiments were removed from the
field during 1970-1972. These animals were maintained in the live
animal facility of the Department of Biology, Tabor College, Hillsboro,
Kansas, and were caged under conditions described by Colvin and Colvin
(1970). The facility at Tabor College is located in a heated basement
which remains relatively cool in the summer and warm in the winter.
Relative humidity was always rather high and light was supplied through
overhead windows and by a 100 watt incandescent bulb. Light intensity
in the room during the day was always near 10 fc. The animals were kept
on a light cycle of 12 hours daylight and 12 hours darkness. Water and

food (oats and grass) were given ad lib.

Aggressive Behavior

 

Interspecific aggression has been proposed as a mechanism by which
Sigmodon may exclude Microtus (Baker, 1971; Martin, 1956). Terman and

Johnson (1971) noted aggressive behavior between Sigmodon hispidus and

 

Microtus pennsylvanicus when the two were together in a confined

laboratory arena. Sigmodon appeared to be dominant and actually killed

53

and ate some juvenile Microtus. Martin(1956) reported that_§. hispidus
killed and ate Microtus when the two were captured in the same live
trap. Eugene Fleharty (personal communication) observed S, hispidus

to be dominant over Microtus ochrogaster when the two were placed to-

 

gether in a terrarium.
On the basis of this information, I hypothesized that Sigmodon
hispidus would be aggressive towards and dominant over Microtus

ochrogaster in interspecific encounters.

 

Subjects

All animals tested were wild males recently taken from the field.
Before being used, each rodent was caged for one month to allow it to
acclimate to the laboratory environment. Only adult animals (as defined

under Field Methods) were used and each animal was tested only once.

Treatments

 

The purpose of this experiment was to determine if Sigmodon was
aggressive towards and dominant over Microtus. Individuals of each
species were tested in (l) a neutral arena; and (2) in an arena formerly
inhabited by the other species. If one species was aggressive towards
and dominant over the other species in both of these treatments, it was
concluded that this species would be victorious in most of the en—

counters that might take place in the field.

Apparatus and Procedures

 

Each test arena consisted of a terrarium (91 x 30 x 30 cm.) with a
solid, central, removable partition and a floor covering of soil and

grass. For each test a cotton rat was placed in one section and a vole

54

in the other. The test animals were allowed to acclimate, and then the
partition which separated them was removed. The behavior patterns ex-
hibited were observed and recorded for 30 minutes (under red light,
following Finley, 1959), and the dominant animal was determined by the
aggressiveness (attack) or avoidance that was exhibited. Thirty inter-
specific pairings were observed, 15 in each of the treatments (neutral

and resident arenas).

Analysis and Results

 

A noticeable interaction occurred in all but 8 of the 30 pairings.
Sigmodon dominated Microtus in each encounter. Sigmodon was aggressive
towards Microtus even in those trials where Microtus had been the resi-
dent species in the arena. Ten Microtus eventually died as a result of
injuries which were received in the pairings. Three of eight trials
without interactions were in the neutral arena, one occurred when
Sigmodon was the resident species, and four occurred when Microtus was
resident.

Sigmodon exhibited either attack (Krebs, 1970) or did not react to
Microtus. Microtus exhibited agonistic behavior only after it came
into physical contact with the cotton rat. Teeth chattering (Getz,
1962) was a common reaction as was submission, avoidance, and threat
(Krebs, 1970).

During 10 preliminary interspecific pairings, it was noticed that
Sigmodon subadults (j; 60 grams) were non-aggressive toward Microtus.
This point deserves further study for it could be important to the

interaction in the field.

55

 

Spatial Distribution

 

When Sigmodon becomes abundant in an area, it is often noted that
Microtus becomes scarce or nonexistent (Baker, 1969; Fleharty and Olson,
1969; Frydendall, 1969; Martin, 1956, 1960). Terman and Johnson (1971)

reported that S. hispidus limited the movements of M. pennsylvanicus

 

when the two were in a confined area. The purpose of the present ex-
periment was to determine if Microtus would use areas occupied by
Sigmodon under (1) sparse cover conditions, and (2) under dense cover
conditions. I hypothesized that Microtus would reduce their use of
those areas occupied by Sigmodon when cover was sparse but would not Lm;
reduce their use of those areas when cover was dense. The data from

Terman and Johnson (1971) indicated that co-utilization would not occur

under sparse cover conditions. The results of Crombie (1946), Morris

(1969), and the theories of MacArthur (1972) suggested that dense cover

would allow co-utilization of an area.

subjects

All Microtus and Sigmodon used in these tests were adults recently
captured from the field. For the tests using sparse cover conditions,
15 groups of three male adult Microtus and 15 bisexual pairs of adult
Sigmodon were used. For the tests using dense cover conditions, 13
different groups of Microtus and Sigmodon were tested. Each group was

used only once so all animals were naive to the experimental conditions.

Apparatus
A test chamber (2.4 x 1.2 x 1.2 m.) similar to that described by

Terman and Johnson (1971) was constructed (see Figure 21). Observations

56

could be made over the top of the 1.2 m. sides of the chamber without
disturbing the animals. Also, a red light was used to make the observa-
tions. Water and food (oats) were supplied in excess in each compart-
ment. The compartments were connected by holes measuring 2.5 cm. and
1.3 cm. in the partitions at floor level. These openings could be
closed at will. Sigmodon was able to go through the large holes but not
through the small holes. Between observations the arena was covered
with a canvas which caused the floor area to be similar to natural
habitat in light intensity.

Two environmental conditions were used: sparse and dense cover.
Sparse cover consisted of a mixture of grass and soil placed on the
floor of the arena allowing the animals to be easily seen from overhead.
Dense cover consisted of the grass-soil substratum with shredded paper
added in excess so that the animals could not be seen from overhead
without moving the cover material. A stiff wire probe was used to
separate the cover at places where there was activity in order for the

observer to see the animals.

Treatments and Procedures

 

To test the hypothesis that voles would reduce their use of those
parts of a confined area occupied by cotton rats (restriction of spatial
distribution), fifteen groups of three male adult voles and fifteen bi-
sexual pairs of adult cotton rats were subjected to the following three
trials under sparse cover conditions. After the voles had acclimated to
the entire arena for three days, the first experimental trial consisted
of introducing the pair of Sigmodon into compartment A. Subsequently,

locations of the individuals by compartment were recorded at 18

57

 

CI” C)
water water water

1|
0

A. \\ B ;'\\ C
1.3 cm. 2.5 cm.
holes holes

I

Figure 21. The experimental arena used to
test for spatial restriction.

 

 

 

 

58

irregular intervals over a three-day period. The voles could freely
move throughout all sections but the cotton rats were confined to sec—
tion A.

The second experimental trial consisted of placing the pair of
cotton rats in compartments B and C. The voles were left in the com-
partments in which they were found at the end of trial one. They could
freely move throughout all the sections, but the cotton rats were now
confined to sections B and C. Observations were recorded as in trial Q
one.

To exclude the possibility of position effect, a control test
(trial 3) in which only voles were used was run. During these observa—
tions the voles were able to move freely from one compartment to an-
other. During the trials, if an animal died it was replaced by another.

To determine the effect of dense cover on this interaction, thir-
teen different groups of Microtus and Sigmodon were tested with com-
partments B and C containing dense cover. After the voles had accli-
mated for three days, trial one in this experiment consisted of placing
the Sigmodon in B and C. Observations were carefully taken at 18
intervals over a three-day period as in the first experiment.

For the second trial, the cover material was carefully removed,
leaving the animals in sparse cover as in the first experiment. After
one day, observations were recorded for the three-day period. This
trial was used to verify the effect of the cover in trial one.

Trial three was a control test in which only voles were used.

Before these observations, the cover was replaced in the compartments B

and C.

59

Two different groups of three male adult Microtus were examined in
a third experiment. Two pairs of Sigmodon were placed in dense cover
along with the voles. At this high Sigmodon density, it was expected

that the voles would not coexist with the cotton rats.

Analysis and Results

 

The results are presented in Table 11. The number of Microtus in
each compartment was compared between treatments using the Friedman two-
way analysis of variance for the data from the sparse cover conditions.
The number of Microtus in each compartment varied significantly between
the treatments (p‘(.001) under sparse cover. This indicated that the
Sigmodon significantly excluded Microtus from compartments it occupied
under sparse cover conditions.

Under dense cover conditions, the number of Microtus in each come
partment was examined when Sigmodon was present in compartments B and C
and when Sigmodon was absent from these compartments. The Wilcoxon
matched-pairs signed-ranks test was used to make this comparison. The
number of Microtus in each compartment did not vary significantly between
these two treatments. This indicated that dense cover allowed the
Microtus to use those areas occupied by Sigmodon. When the cover
material was removed (but not the Sigmodon) the distribution per com-
partment differed significantly from when the cover was present with the
Sigmodon (p‘<.001). This again indicated that cover was essential for
the co-utilization of compartments.

Although not enough groups were tested to be statistically eval-
uated, the experiment using four Sigmodon instead of two suggested that

a high density of Sigmodon might override the effect of the cover. Both

60

Table 11. The median number of MicrOtus in each compartment per
observation under sparse cover (15 replications) and under

dense cover (13 replications).

 

 

 

 

SPARSE COVER
Compartment Compartment Compartment
Treatment A B C
Sigmodon in A 0 (0) l (12) 2 (l7)
Sigmodon in B, C 3 (l7) 0 (l) 0 (l)
Sigmodon not present 1 (15) l (11) l (12)
DENSE COVER
(in B, C)
Sigmodon in B, C 0 (6) 1 (12) l (11)
Sigmodon in B, C -
cover removed 3 (l7) 0 (3) 0 (3)
Sigmodon not present -
cover in B, C O (6) l (15) l (14)
Numbers in parentheses are the mean numbers of observations

(rounded off) per three-day observation period that Microtus

was observed in each compartment.

61

of the observed groups of Microtus declined to enter the compartments
containing the four Sigmodon. They were noticed to remain in compart-
ment A much more than did the Microtus in the other dense cover experi-
ments.

I also observed that MicrOtus would spend more time in compartment
A (not accessible to Sigmodon) when Sigmodon was especially active, even
in situations of dense cover. A rustling of the cover material usually
signaled increased movement by Sigmodon and this observation was often
coupled with the presence of Microtus in compartment A.

Other observations were taken on the location of nests, water
usage, and tracks on smoked cards. Table 12 gives these data. A nest
was recorded as being in a compartment if a nest was noticed in 70 per-
cent of the groups. Water usage was termed "considerable" for a com-
partment when an average of ;Z_30 percent was used during the trials and
"slight" when an average of :g_30 percent was used. In each compartment
smoked cards were placed in a small cage accessible only to Microtus.

At the end of every trial, these cards were checked for vole tracks.
The percentage of groups making tracks in a particular compartment was
then computed. These observations supported the distribution revealed
by the animal counts.

Upon removal from the arena, all animals were checked for the
presence of wounds. Only Microtus exhibited visible lacerations. In
experiment one (sparse cover), 12 percent of the animals were wounded
and 12 percent were found dead. In experiment two (dense cover), 35
percent of the animals were wounded and 7 percent died. In the two
trials using four cotton rats, four of the voles had wounds and three

of the voles died. One cotton rat died in the high density trials,

62

Table 12. Summary of nest location, water usage, and tracks in the test
chamber (all experiments combined).

 

Nest Water Percent of groups
Test Location Usage * making tracks
Sparse Cover A ' E .9 A 13. 9. A E E.
Sigmodon in A 0 X X 0 f + O 100 100
Sigmodon in B, C X 0 0 + 0 0 100 O 0
Sigmodon not
present X X X + + + 100 100 100
Dense Cover
Sigmodon in B, C X X 0 + + + 31 69 61
Sigmodon in B, C
(cover removed) X 0 0 + O 0 38 0 0
Sigmodon not
present (cover
replaced) 0 X X + + + 15 54 46
Four Sigmodon in
B, C (dense
cover) X 0 0 + + 0 50 50 0

 

* + Considerable water usage (£330 percent of water used)
0 Slight water usage (SL30 percent of water used)

63

presumably from intraspecific strife.

The Effect pf_Sigmodon‘pp_Microtus Movements

 

Baker (1971) has contended that since Microtus and Sigmodon are
both runway prone grass eaters, they may have many head-on meetings in
the narrow trails that are used by both species. Jameson (1947)
reported capturing both species in the same runways. It has also been
suggested that neither Microtus nor Sigpodon seem to have the complex
social organization required to avoid frequent inter—individual contacts
(Brown, 1966; Calhoun, 1963; Getz, 1972). The question arises as to
what mechanism may then allow Sigmodon to restrict the movements of
Microtus. I hypothesized that‘Sigpodon would only affect the movements
of individual Microtus when there were frequent contacts (tactile)

between the two.

Subjects
A total of 16 male adult Microtus and 16 adult Sigmodon were used

for this experiment. All animals tested were recently removed from the
field and thus new to the experimental situation. The animals were kept
in the laboratory for at least a.week before they were used for experi-

mental purposes.

Apparatus

To test the hypothesis that Microtus movements would only be
affected by frequent contact with a cotton rat, a tunnel was constructed
(1.8 x 0.5 x 0.5 m.) with an adjoining cage (see Figure 22). For non-
contact encounters a wild—caught, healthy, active cotton rat was placed

in the cage. The cage was then closed and placed midway across the

issw'

 

64

age

Food and Water

 

 

 

AREA holes holes

 

 

 

 

 

l Treadle ‘
.. 1.3}o"““’ ’3 ’ _1.3 NEST AREA
RECORDING " tin: ~. .. " ‘ "cm.

 

 

TO RECORDER A . -__...J u

 

 

_—-—-'--b
:"--‘

I V9

Cage (opens up into center)

 

 

 

 

 

 

 

 

 

i__..Jf

Figure 22. The movement tunnel apparatus (A - no contact; B -
contact). ‘

65

tunnel (Figure 22-A). The approaching Microtus could see, hear, and
smell the cotton rat but could not enter into combat. The vole was free
to pass over the cage and enter the recording area which had a treadle
connected to a recorder (Heath-Kit) on a 24 hour cycle. When the vole
stepped on the treadle, an electrical circuit was closed which caused
the recorder to make a mark at the corresponding time of the cycle.

The lights in the laboratory were set to turn on and off at sunrise

and sunset.

To allow contact, the cage was simply moved to one side so that it
opened up into the central part of the tunnel (Figure 22—B). The tunnel
had sliding panels with 1.3 cm. sized holes which prevented the Sigpodon
from entering the recording area and the nest area of Microtus. The
nest area was darkened and contained cotton batting. The recording area
was painted white and contained approximately 75 percent of the food and
water required by an individual rodent in 24 hours. The nest area con—
tained the rest of the food and water which prevented starvation but did
not remove the motivation to go into the recording area.

To determine the activity of Sigpodon, the partitions were removed.
The cotton rat could be tested alone or with a Microtus confined in the

cage.

Treatments‘and‘Procedures

 

Eight adult Microtus were subjected to noncontact encounters with
Sigpodon and eight other adult voles were subjected to contact en—
counters. All animals tested were unfamiliar with the tunnel and in
most cases were recently captured from the field. Ten adult Sigmodon

were also tested with a MicrOtus confined in the cage.

66

Each animal was acclimated to the tunnel for one day. It was then
tested alone, with the other species, and then alone again. The tests
were run in sequence. The last test should reveal any lasting effects
of the interspecific encounter.

Activity periods were defined as 30-minute intervals with at least
two marks. The number of activity periods was tallied for each animal
and then recorded for each test. Table 13 gives the results of this
experiment.

The proportion of total marks falling into three-hour intervals was
computed. Figure 23 illustrates these data. The peaks of activity for
both species occurred during 1800 - 0300 hours. This agrees favorably

with Calhoun's data (1945) for these same animals.

Analysis and Results

 

That Microtus lived in the nest area and had to pass the Sigmodon
to get to the recording area was indicated by the presence of nests and
the abundant feces in the nest end of the tunnel. Also, when the ani-
mals were checked periodically during the experiment, Microtus was
invariably in or near the nest area. Evidently the presence of cotton
batting, the small amount of food and water (one-fourth of daily re-
quirement), and the closed, darkened area were more desirable to the
Microtus than the open recording area with its more abundant food and
water but whitened background and no nest material. Microtus evidently
moved to the recording area for the desired food and water but quickly
returned to the seclusion of the nest area.

In both tests (contact and no contact) the number of MiCrotus

activity periods was less when'Siggodon was in the tunnel. However, the

67

Table 13. The number of activity periods for individual MicrOtus under
no contact and contact conditions.

No Contact (Sigmodon confined)

 

 

 

Sigmodon
Animal Alone "ip'Cage Difference
l 13 6 + 7
2 10 6 + 4
3 13 1 +12
4 5 7 - 2
5 17 13 + 4
6 24 20 + 4
7 9 3 + 6
8 19 11 + 8
Contact (Sigmodon free)
'Animal Alone .Sigmodon free Difference
l 19 2 +17
2 l8 0 +18
3 l4 0 +14
4 8 0 + 8
5 20 3 +17
6 16 1 '+15
7 20 0 +20
8 l7 9 + 8

 

Differences were ranked and compared using the Mann-Whitney U test.

68

free ranging Sigmodon caused these differences to be much greater than
the confined Sigmodon (p‘(.01, Mann-Whitney U test). This indicated
that Microtus movements were more affected by'Sigmodon when the two
animals could physically contact each other. Inspection of the raw data
indicated that this contact was sufficient to keep some Microtus from
ever crossing the center area of the tunnel. Four of the voles remained
in the nest area despite a shortage of food and water. After each test
with the Sigmodon, the Microtus were again tested when alone. There was
also a reduction in the number of activity periods in this last test.
The differences between the treatments were not significant. This
activity reduction may have been due to an increased familiarity with
the tunnel.

Cotton rats did not significantly reduce their activity in the pres-
ence of voles although the number of activity periods was less. Cotton
rats did not show the reduced activity in the last test as did Microtus.

The Microtus activity pattern during Sigmodon contact did not
appear to be different from the other activity patterns (see Figure 23).
It appeared that the amount of activity changed but not the scheduling
of the activity. The peaks of activity for both species were in the
same time interval (1800 - 0300 hours). This supported Calhoun's (1945)
contention that these two species may be active at the same time in the
field.

These laboratory results suggested that Sigmodon would be dominant
in an interspecific encounter should it occur in the field. Sigmodon
also seemed to have the ability to exclude MicrOtus spatially. Dense
cover affects this relationship by allowing Microtus to enter areas

occupied by Sigmodon. At higher Sigmodon densities, however, this

69

Aiwas mos mua>auom a coasts sa waau mo
«wasp m muoooaaoo mooaa awoauum>v .uomusoo
Gloom-awflm. usosuas cow .uomusoo a

:33 .mooam 3 mo ~53qu mua>auo<

Amazonv MSHH
comm ccmc cog ocam coca coma coma oomc coco

.mN musmam

 

‘

HU<HZOU FHHZ mDHOMOHZ
M7812 mPHOMUHS

< l

d -

 

HU<HZOU oz EH3 mPHOMUHZ Illl

d

4 d

. .Uo|.

 

1\\.

’-
/ ’/
\.L.Va....
/.
/ \

 

.\.
IMAM/z . 3

.II
\ .11 m

[’2]

I
ma

 

LLIAILOV 'IVLOJ. JO NOILHOJOHJ

7O

effect seemed to be lessened. The freedom of Sigmodon to contact
Microtus seemed to be necessary before it could significantly affect

the movements of Microtus.

DISCUSSION

 

Field‘Popplations

Ayala (1972), Hutchinson (1953), Klopfer (1969), Fretwell (1972),
and MacArthur (1972) point out the significance of environmental factors
to the outcome of interspecific competition. Since Sigmodon is near the
northern edge of its range (Cockrum, 1948), it was expected that over-
winter survival in this species would be low in Kansas thus releasing
Microtus from any major competitive pressures in early spring.

The late occurrence of Sigpodon in 1971 and the low population
levels in the spring of 1969, 1970, and 1972 did strongly suggest that
overwinter survival for this southern species may be low in Kansas.

This phenomenon would appear to allow MicrOtus to be free from heavy
Sigmodon contact for a relatively long time (for my data, late December
until August). The severity of the Kansas winter is prdbably the prin-
cipal factor. The climatic data in Table 1 reveal that the winters of
1970 and 1971 were severe in terms of snowfall. The spring trapping
periods of these years yielded not a single cotton rat. The winter of
1972 was mild in comparison (little snowfall) and the April trapping
period of that year was the only one that yielded any Sigmodon at all.
The trapping records reported by Gier (1967) indicated a similar situa-
tion throughout the state of Kansas. There were only isolated instances

of cotton rats being abundant in the spring of the year and it was a

71

72

common observation to see them disappear after a severe winter. The
states south of Kansas experience a milder winter which presumably
allows more Sigpodon to overwinter. This permanent Sigmodon residence
may contribute to the absence of the prairie vole from states such as
Oklahoma.

The late summer emergence of Sigmodon and the shortened time of
interspecific contact affected the experimental design of the field in-
vestigation. It was originally intended to observe the year—around
response of one species population to the presence or absence of the
other. As it happened, this situation was only partially realized.
Between-the-plot comparisons were made over a relatively short period of
time and with only a small population of Sigmodon being present. It is
doubtful whether such conditions allowed the time and frequency of con-
tact necessary to produce reliable results. The single plot comparisons
(one plot observed over time) was affected by the presence of seasonal
factors which may obscure any effects of interspecific interaction. The
field results, consequently, merely suggested a negative interaction
between the two species. The laboratory experiments examined this rela-
tionship between the two species in order to learn if their interactions
were negative.

The October appearance of Sigmodon coincided with a sharp drop off
in Microtus captures and recruitment. Data from Carley gp_§l. (1970),
Cooksey (1971), Fitch (1957), Keller and Krebs (1970), Krebs pp 31.
(1969), Martin (1956), Meserve (1970), and Yang g£_§1. (1970) did not
indicate that a normal MicrOtus population cycle should crash this
abruptly in October. At the same time, it was not possible to attribute

the vole decline solely to the presence of the cotton rats because of

73

the decline in plot 3, a plot which continually had its Sigmodon pop-
ulation removed. Since the plots were all part of a contiguous grass—
land area, it can be argued that all the plots, regardless of treatment,
were equally affected by the Sigmodon invasion. Such things as immi-
gration might be affected when field plots are contiguous. Such state-
ments speak to the designs of future experiments. It may not be
possible to look for population changes in plots which are in the same
area and contiguous.

The plots containing both species did indicate that some coexist-
ence was possible for the observed time interval. It may also be stated
that the presence of good cover and food supply facilitated the co-
existence. Plot 1 (abundant vegetation and cover) had a more constant
Microtus population growth than plot 2 (lesser vegetation and cover).

If these two plots had been observed over a longer period, this rela-
tionship might have become more evident.

The plots with low Sigpodon densities (plots 2, 3, 4, and 6) in
1971 had the highest Microtus densities in 1972. Although it can not be
verified that the cotton rats caused this, such data do suggest a
possible negative interaction. The presence of wounds on Microtus, the
absence of Microtus juveniles from all plots except the Siggodon removal
plots, the winter increases in Microtus population levels, and plot
differences in immigration also suggest that the interaction between the

two species is negative.
Mean Body Weights and Sex Ratios

It was hypothesized that the mean body weights of Microtus would

decrease over time in those plots where neither species was removed.

74

Joule and Jameson (1972) noticed a weight increase in Sigmodon hispidus

 

females when Reithrodontomys and Ogyzogys were removed from the habitat

 

being used in common. Although the mean body weights decreased in the
dual species plots and not in the Sigmodon removal plot, the differences
were not significant. A longer period of interspecific contact may
have produced significant results.

It was also hypothesized that the sex ratios of Microtus would
change in the presence of Sigpodon. Joule and Jameson (1972) found this
to occur in Siggodon when its competitors were removed. This hypothesis
was confirmed for plot 2 (p (.05) but not for plot 1 (p (.1). Again,
the period of interspecific contact may not have been long enough for
both plots to have significant values. These results, although not
conclusive, suggest that Sigmodon might affect Microtus negatively in

situations providing sufficient interspecific contact.

Survivorship and Mean Residence Times

If Sigpodon affects Microtus adversely, this should show up in
reduced survival rates for Microtus in the dual species plots. It was
hypothesized that Microtus survivorship would be lower in the presence
than in the absence of‘Sigmodon. This hypothesis for lower vole survi-
vorship was confirmed for plot 2 but not for plot 1. The reason for
this was not clear. It seemed more probable that Microtus survival
rates in plot 1 would drop since it had more Sigmodon. Unless the
differences in cover between the two plots caused this discrepancy, it
seems more likely that Microtus survival in plot 2 was more affected by
environmental factors than by the few cotton rats which were captured

on the plot.

75

It was also hypothesized that the mean residence times of Microtus
would be lower in the presence of Sigmodon than in their absence. This
hypothesis was not confirmed for either plot. All that can be said is
that there might have been an inhibitory influence. The time of inter-
specific contact was probably too brief to cause significant differ-
ences. Also, the numbers of Sigppdon on the study area as a whole were
probably low as was shown by the removal rates from plot 6 (both species
removed). While Microtus were taken with regularity from this plot,
Sigmodon were taken only sporadically with never more than one indi-

vidual being taken in a trapping period.

Spatial Distribution

 

In the absence of procedures for analyzing the data on spatial
distribution, examination of the raw data did not suggest any plot dif-
ferences between the dual species plots and the Sigmodon removal plot
during the'Sigmodon occupation. There did seem to be major differences
which appeared when the plots were compared individually over time.
When Sigmodon appeared in plot 1, MiCrotus seemed to leave former areas
of concentration. Also, there were few Microtus around areas of con—
centrated Sigpodon activity. This was also the case in plot 2. Since
the Siggodon in plot 3 were removed, it was expected that this plot
would not show obvious changes in Microtus spatial distribution over
time. However, Microtus in this plot also appeared to change which
leads to the conclusion that factors other than the presence of Sigmodon
were operative. It was not possible to ascertain the role of inter-
specific strife in these distributional changes which again means that

the data were only suggestive.

76

Microtus Range Overlap

 

At the onset of the field experiment, it was expected that the
Sigmodon would cause the individual ranges of MicrOtus to overlap due
to the movement limitations that the cotton rats might impose on the
voles. This hypothesis was rejected when the plots were compared with
each other. When the individual plots were compared over time, plot 2
was the only plot which changed significantly in range overlaps.
However, this change was not in the expected direction. The ranges
became more exclusive which indicated that something other than the
Siggodon (which were few in number) caused this change. Since both
plots 2 and 3 showed large changes, it was concluded that changes in
population density were the cause of these changes. Both plots exper-
ienced sudden declines which probably reduced the chances for any given
trap to capture more than two voles. This measure is probably not very
useful in long term studies where population densities may change

radically.

Indices pf Association

 

Coefficients - Cole's (1949) coefficient of association indicated

 

a negative interaction between the two species in plot 1 (see Table 5).
Plot 2 presumably did not show this because of lower densities of
Sigmodon which resulted in reduced opporunities for interspecific con-
tact. Plots 3, 4, 5, and 6 showed neutral values primarily because of
the reduced contact due to the removal program. Since plot 2 did not
have an adequate density of cotton rats, plot 1 was the only one that

could be used to verify the hypothesis that the two species were

77

negatively associated. Even though the correlation coefficient for this
plot was not significantly negative, the very significant value for
Cole's index remains as convincing evidence for a negative interaction.
In summary, this index revealed a negative interaction where the
opportunity for interspecific contact was high. Seasonal influences,
plot differences, and fluctuating rodent populations all acted to reduce
the statistical validity of the index. Even considering these factors,
there does appear to be evidence that the Sigmodon and Microtus reacted
negatively to each other.

Trap and Capture Indices pf Association - It was expected that

 

 

traps which captured two or more Sigmodon would capture lower numbers
of Microtus than traps which had fewer Sigmodon (Table 6). Although the
number of Microtus at these traps was lower, the difference was not
significant. It was also expected that the trap capture rates of
MicrOtus in the dual species plots would change more than the trap
capture rates in the removal plots. The rates in the dual species plot
did change significantly while those in the removal plots did not
change. This indicated that Sigmodon caused Microtus to use different
traps or to become less trappable (see discussion of trappability).
This suggested a negative association between the two species.. The
spatial distribution of Microtus and Sigmodon in plot 1 (see Figures 12
and 13) also suggested that the trap capture rates of Microtus might

change with the appearance of Sigmodon.

Movements, Migration, Home Ranges, and Trappability

 

 

The hypothesis for significantly different population movements

between the dual species plots and the Sigpodon removal was confirmed.

78

These results suggested that Sigppdon may limit the normal population
movements of Microtus.

It was hypothesized that there would be significant differences
between the plots in the median numbers of different traps and captures
per individual Microtus. This hypothesis was not confirmed. It was
also hypothesized that the plots containing both species would change
significantly in the number of traps and captures per Microtus when com-
pared over time. This hypothesis was confirmed for plot 1 but not for
plot 2. The interplot differences were not significant but the over
time comparison (before and after Sigppdon) for plot 1 was significant.
This makes it difficult to conclude that Sigppdon negatively influenced
Microtus. Were seasonal factors the cause of the change over time in
plot 1? Also, why were there no interplot differences between the
removal plots and the plots containing both species?

The increased amount of inter-plot movement by Microtus in autumn
probably illustrated a response to seasonal factors rather than to the
presence of cotton rats. That this increased movement takes place casts
further doubt on the validity of making certain comparisons on the voles
over time. Future experiments should take this into consideration.

It was hypothesized that individual and population home ranges of
MicrOtus in the dual species plots would be smaller than those in the
Sigpodon removal plot. However, there were too few individuals to make
any comparisons between the plots using the individual home ranges.

The individual home ranges were then pooled for all three plots and the
difference was significant when the three plots were compared over time.
Although seasonal factors complicated the comparison, it is noteworthy

that the individual ranges decreased in size when it was expected that

79

they would increase (as the autumn inter-plot movements would indicate).
The hypothesis for decreased Microtus population home ranges was also
rejected although the voles on the dual species plots had smaller home
ranges than on the Sigpodon removal plot. When the population home
ranges on each plot were compared over time they showed an increase.
Plot 3 (Sigmodon removed) showed the largest increase but this was not
significant.

It was hypothesized that Microtus trappability would be lower in
those plots containing Sigpodon than in the Sigmodon removal plot. The
Microtus in plot 2 were the only ones that were significantly less
trappable than the ones in the Sigmodon removal plot. This result also
should have been observed in plot 1 since it had the highest cotton rat
densities. This is difficult to explain unless the differences in cover
between the plots allowed for differing degrees of interspecific
contact.

In summary, the field results of this study were at most suggestive
of a negative interaction between Microtus and Sigmodon. The results
were inconclusive mainly because of the late appearance and low numbers
of'Siggodon. Also, the weaknesses of live-trapping as a technique for
monitoring small mammal interactions were apparent. Results of the
present study suggest that in Kansas, this negative interaction may
only occasionally become severe enough to result in the exclusion of
MicrOtus. If Sigmodon can adapt to the severe winters so that it can
maintain a more stable population level, or if winters become less

severe Siggodon may eventually exclude MicrOtus.

80

Aggressive Behavior

 

It was hypothesized that Sigmodon would be aggressive towards and
dominant over MiCrOtus in interspecific encounters. This hypothesis was
confirmed. If interspecific encounters between individuals occur in
the field and if the laboratory behaviors are real, Sigmodon should have
a negative survival influence on Microtus. How many encounters are
necessary to affect noticeably a papulation of Microtus is debatable.
Also, how long might it take? With what frequency must these inter-
actions occur? These questions must be answered if the mechanism of

interference is to be understood.

Spatial Distribution

 

It was hypothesized that Microtus would not use those areas of a
laboratory arena occupied by Sigmodon. This hypothesis was confirmed
under sparse cover conditions. It was also hypothesized that dense
cover would allow co-utilization of the compartments. This was also
confirmed for the experimental conditions. At higher Sigpodon densities
the degree of co—utilization of a compartment by the two species seemed
to be lessened. It appeared that the cover conditions affected the
frequency at which the two species came into contact. If this frequency
was high (as under sparse cover conditions which allowed the animals to
see and smell each other) the Sigpodon would exclude Microtus. If the
frequency of contact was low (as under dense cover conditions which
permit concealment) coexistence resulted. High densities (and greater
activity) increase the frequency of contact which leads to spatial ex-

clusion. It appeared that the densities used in the laboratory were

81

sufficient to reveal interference in space utilization. What densities
are required in the field? Does there have to be a certain number of
adults?' The tentative observation that subadult‘Sigpodon.were non-
aggressive toward Microtus suggests that there might be a certain number
required. Are such sensory modalities as sight and smell operative in
the exclusion? MicrOtus did not stay out of areas which had been
previously occupied by Sigpodon, whose remaining odors did not seem to
affect Microtus. The frequency of interspecific contact seems to be

a crucial point.

The Effect_p£ Sigmodon pp Microtus Movements

 

It was hypothesized that Sigmodon would only affect the movements
of individual Microtus when the two species made frequent contact.
This hypothesis was confirmed when it was observed that a free ranging
’Sigpodon affected Microtus movements more than a confined Sigmodon.
Here again a situation, which provided contact, resulted in spatial
exclusion. Why did the contact and not the mere presence of the cotton
rat result in restriction of movements? Is the learning ability of
Microtus involved? An aggressive encounter is a stimulus which may
not be forgotten. If enough of these encounters occur, would the
learning of avoidance by all members of a population then result in
complete exclusion? The animals are active at the same times (see
Figure 23), they are both runway oriented, and they seem to strive for
the same food resources (Fleharty and Olson, 1969). The possibilities
for contact appear to be present. The question arises as to what

factors may control the number of interspecific contacts.

82

The results from the laboratory investigations suggested that the
Microtus-Sigmodon interaction might be one based on the frequency of
interspecific contact. If this is the case, it might be possible to
explain many of the observations on these two species in terms of this

concept. Since this concept will be referred to later, it is abbre-

viated to FIG.

INTERACTION BETWEEN SIGMODON AND MICROTUS

From the results of the present study, it was difficult to conclude

with confidence that Microtus ochrogaster and Sigmodon hispidus were in

 

competition. The most that can be said is that they interact negatively
and frequent interactions are necessary before the effect of Sigpodon
on Microtus can be measured in the field.

Other workers have suggested that these two genera may be in
competition as it is described by Birch (1957) and Miller (1967).

Fleharty and Olson (1967) noted that Microtus ochrogaster and Sigmodon

 

hispidus in Kansas consumed the same species of plants, lived in the
same area, and appeared to be in competition for space. Since both
species seem to under-utilize food resources (Golley, 1960; Fleharty
and Olson, 1969), competition for space is a tenable hypothesis.
Wiegert (1972) implicated competition in South Carolina when he
observed that Microtus pennsylvanicus, which is rare or nonexistent

in the state, lived and reproduced in enclosures when not in contact
with Sigmodon. Odum (1955) indicated that there might be an inter-
specific interaction between these same two species in Georgia. Bryan
Glass (personal communication) pointed to interspecific competition as
a probable reason for the prairie vole's general absence from Oklahoma.
Data from other Oklahoma studies (Calhoun, 1950; Goertz, 1964, 1971;

Hays, 1958; Phillips, 1936) support this contention. Howell (1954) did

83

84

not report any M.'ochroga3ter with the Sigmodon that he caught in

 

Tennessee. Baker (1969) suspected competition between Microtus and
Sigmodon in Mexico.

Occasionally small populations of Microtus have appeared sporad-
ically in areas where Sigmodon was the dominant grass eating rodent.
This occurred in Oklahoma in 1957 following a Sigmodon "crash" in
population (Bryan Glass, personal communication). The same phenomenon
appeared to happen in Georgia (Odum, 1955; Ramsey and Briese, 1971).
Whittaker and Zimmerman (1968) noted an unusual appearance of M.
ochrogaster in Alabama. In Tennessee, there was also an unexpected
appearance of this species (Dimmick, 1969; Whittaker and Zimmerman,
1968). Baker (1969) observed presumed differential exclusion of

Microtus mexicanus by Sigmodon in Mexico.

 

These observations indicated that if there was any competitive
exclusion (Hardin, 1960) of the voles by the cotton rats, it was only
partially realized. The question arises as to what mechanisms and
factors could result in the above mentioned observations. It is pro-
posed that a model of the interaction based upon the frequency of
interspecific contact (FIC) could explain these observations. Such a
model is given in Table 14.

The frequency of Microtus-Sigmodon contact (FIC) facilitated by

 

each of the factors is the important point. In Kansas, Microtus experi—
ences low FIC much of the time because severe winters seem to decimate
Sigpodon populations (Gier, 1967). It follows that Microtus only
occasionally comes into frequent contact with Sigmodon; this infre-
quency seems to permit voles to coexist with the sporadic cotton rats.

This factor is probably important throughout the zone of overlap

85

Table 14. A tabular model of the Microtus-Sigmodon interaction:
trends to be expected.

 

 

Microtus
FaCtor Factor value .ELEHXElEE *
A. Sigmodon overwinter survival poor 1
moderate 2
good 3
B. Present Sigmodon population low 1
density
moderate 2
high 3
C. Adult proportion of Sigmodon always low 1
population
periodically high 2
constantly high 3
D. Habitat cover dense 1
moderate 2
sparse 3
E. Extent of habitat area ;Z_10 hectares l
2 hectares 2
:g_1 hectare 3
F. Habitat diversity high 1
moderate 2
low 3
PrediCtion formula If: equals expect

 

—A+B+C+D+E+F=l3-18 exclusion

8 - 12 noticeable
inhibition

6 - 7 coexistence

 

* Frequency of Interspecific Contact (a value of 1 indicates low FIC).

86

between the two genera (Dunaway and Kaye, 1961; Mohlenrich, 1961).
Periodic Sigmodon crashes (Goertz, 1964; Haines, 1971; McCulloch, 1962)
reduce FIC in southern states. The Sigmodon population, however, is
stable enough in the southern states (Chipman, 1966) that this alone
would not allow Microtus to exist there. The necessity for high FIC
allows MiCrotus to inhabit refuges which, due to a variety of micro-
climatic factors, maintain only small or occasional populations of
Sigmodon. The low FIC in these areas provides for the presence of
vole "reservoirs" which expand when the cotton rats experience a wide-
spread population decline. Other factors which reduce the Microtus
FIC are a low proportion of Sigmodon adults, dense habitat cover,
extensive habitat area, and high habitat diversity (for discussions

of these concepts see Caldwell and Gentry, 1965; K10pfer, 1962; and
MacArthur, 1972). Martin (1956) reported that during periods of high
Sigmodon densities, the voles in his Kansas study retreated to refuge
areas which were rarely used by Sigmodon. Barbehenn and Strecker
(1962) and Ecke (1954) reported this phenomenon as being important in
the interaction between Old World rats (Muridae).

In the field experiment of the present study there are many factors
which could result in low FIC. First, the winter before the period of
trapping was severe and Sigmodon was absent during spring and early
summer. Second, when the cotton rats did appear, most of the indi—
viduals were juveniles and subadults. Third, the densities of Sigmodon
were relatively low. Fourth, the plots were dense in habitat cover,
structurally complex (refuges), and part of an extensive grassland area.
These factors might have been reasons for the general lack of conclusive

results from the field experiment. The high FIC facilitated by the

87

laboratory experiments likewise may have exaggerated a negative inter—
action. Further studies using FIC as a hypothesis are needed to
verify the meanings of these results.

FIC may characterize other competitive situations involving
myomorph rodents (Findley, 1954; Delong, 1966; Getz, 1962; Koplin and
Hoffman, 1968; Murie, 1971; Wendlend, 1970). Many other such cases are
in the literature. Getz (1972) found no evidence of a formalized

social structure in Microtus pennsylvanicus. This observation provides

 

added support for the hypothesis that these animals need contact for
their spatial organization. Batzli (1968), Goertz (1971), Grant (1971),
and Van Vleck (1968) all listed high densities as being central to the
competitive interactions documented. Raun and Wilks (1964) reported

that dense cover was necessary for coexistence between BaiOmys taylori

 

and S, hispidus in Texas. Morris (1969) reported that snow cover

facilitated coexistence between Microtus and Clethrionomy§_in Canada.

 

Explanations of the mechanism of competitive exclusion need to be
generated. In addition to Table 14, I propose the following explanation
concerning competitive exclusion as it may take place in myomorph
rodents. When two ecologically similar species meet, the individuals
of each species enter into aggressive encounters. When the frequency
of these interspecific encounters (contacts) becomes high enough,
individuals of the subordinate species develop an avoidance behavior.
The acquisition of this behavior causes an individual rodent to leave
an area should it continue to receive enough negative stimuli (contact,
sight, smell) from the members of the dominant species. When enough

subordinate individuals acquire the avoidance behavior and if the

88

dominant species continues to provide negative reinforcements, exclusion
results. The time required for exclusion to occur depends upon the
frequency and the duration of interspecific contacts (FIC) and on the
ability of the subordinate species to learn and exercise avoidance
behavior.

If situations of species coexistence and exclusion are to be
explained, testable hypotheses need to be generated. Models such as
the one described provide a source for these hypotheses. Whether this
one, or others like it, stand or fall depends on further experimenta-

tion.

2.

SUMMARY AND CONCLUSIONS

The hispid cotton rat, Sigmodon hispidus, and the prairie vole,

 

Microtus ochrogaster, are sympatric in central Kansas.

 

Sigmodon did not reach a substantial population density until late
summer and early autumn. This allowed Microtus to be free from
heavy Sigpodon contact during winter, spring, and summer. The
severity of the Kansas winter was probably responsible for this
relationship.

The results of the field experiment were only suggestive and not
confirmative of a negative interaction between Microtus and
Sigmodon. A late appearance of the cotton rats affected the
experimental design of the field experiment.

Sigmodon (adults but not subadults) was aggressive towards Microtus
and was dominant in interspecific encounters.

Prairie voles did not use the areas of a confined space occupied by
cotton rats unless the cover was dense. However, high Sigmodon
densities excluded the voles even in dense cover.

Microtus movements were not affected by a confined cotton rat but
were affected by a free-ranging one.

A high frequency of interspecific contact between the two species

was prOposed as the necessary condition for a negative interaction.

89

90

8. A.model utilizing the contribution of various factors to the
frequency of interspecific contact was constructed to explain
observations on these two species.

9. A model attempting to explain the mechanism of competitive

exclusion in myomorph rodents was proposed.

LITERATURE CITED

Ayala, F. J.
1972. Competition between species. Am. Scient., 60:348-357.

Baker , R. H.
1969. Cotton rats of the Sigmodon fulviventer group. Misc. Publ.
Mus. Nat. Hist., Univ. Kansas, 51:177-232.

 

1971. Nutritional strategies of myomorph rodents in North American
grasslands. J. Mammal., 52:800-805.

 

Barbehenn, K. R., and R. L. Strecker
1962. Intra-and interspecific behavior. A. Field data. Pages
166-173 i2 T. I. Storer, ed. Pacific island rat ecology.
Bishop Mus. Bu11., 225:1—274.

Batzli, G. O.
1968. Dispersion patterns of mice in California annual grassland.
Jo Mamalo, 49:239-250.

Birch, L. C.
1957. The meanings of competition. Amer. Nat., 91(no. 856).

Brown, L. E.
1966. Home range and movement of small mammals. Pages 111-142 in
P. A. Jewell and C. Loizos, eds. Play, exploration and
territory in mammals. Academic Press, Inc., London.

Caldwell, L. D., and J. B. Gentry
1965. Interactions of Peromyscus and Mus in a one-acre field
enclosure. Ecology, 46:189-192.

 

Calhoun, J. B.
1945. Diel activity rhythms of the rodents, Microtus ochrogaster
and'Sigpodon hispidus. Ecology, 26:251—273.

 

 

1950. North American census of small mammals. Release no. 4.
Roscoe B. Jackson Memorial Laboratory, Bar Harbor, Maine.

 

91

92

1963. The social use of space. Pages 1-187 13 W. V. Meyer and
R. G. Van Gelder, eds. Physiological mammalogy. Vol. I.
Academic Press, Inc., London.

 

Carley, C. J., E. D. Fleharty, and M. A. Mares
1970. Occurrence and activity of Reithrodontomys megalotis,
Microtus ochrogaster, and Peromyscus maniculatus as recorded
by a photographic device. Southwestern Nat., 15:209-216.

 

 

 

Chipman, R. K.
1966. Cotton rat age classes during a population decline.
J. Mammal., 47:138-141.

Cockrum, E. L.
1948. The distribution of the hispid cotton rat in Kansas.
Trans. Kansas Acad. Sci., 51:306-312.

 

1952. Mammals of Kansas. Univ. Kansas Publ., Mus. Nat. Hist.,

Cole, L. C.
1949. The measurement of interspecific association. Ecology,
30:411-429.

Colvin, M. A., and D. V. Colvin
1970. Breeding and fecundity of six species of voles (Microtus).
J. Mammal., 51:417-419.

Cooksey, B. F.
1971. Conditions required for reproduction of Microtus ochrogaster,
in the field and in the laboratory. Diss. Abst. Int. 3.,
32 (l):234.

 

Cox, G. W.
1967. Laboratory manual of general ecology. W. C. Brown Co.,
Dubuque, Iowa. 195 pp.

Crombie, A. C.
1964. Further experiments on insect competition. Proc. Roy. Soc.
(London), Ser. B., 133:76-109.

Delong, K. T.
1966. Population ecology of feral house mice: interference by
Microtus. Ecology, 47:481-484.

Dimmick, R. W.
1969. The distribution of Microtus ochrogaster in Tennessee.
J. Mammal., 50:126.

 

93

Dunaway, P. B., and S. U. Kaye
1961. Cotton rat mortality during severe winter. J. Mammal.,

42:265-268.

Easterla, D. A.
1968. Hispid cotton rat north of the Missouri River. Southwestern

Nat., 13:364-365.

Ecke, D. H.
1954. An invasion of Norway rats in southwest Georgia. J. Mammal.,
35:521-525.

Findley, J. S.
1954. Competition as a possible limiting factor in the distribution
of Microtus. Ecology, 35:418-420.

Finley, R. B., Jr.
1959. Observations of nocturnal animals by red light. J. Mammal.,

40:591-594.

Fitch, H. S.
1950. A new style live-trap for small mammals. J. Mammal., 31:

 

1957. Aspects of reproduction and development in the prairie vole
(Microtus ochrogster). Univ. Kansas Publ., Mus. Nat. Hist.,
10:129-161.

 

Fleharty, E. D., and L. E. Olson
1969. Summer food habits of Microtus ochrpgaster and Sigmodon
hispidus. J. Mammal., 50:475-486.

Fretwell, S. D.
1972. Populations in a seasonal environment. Princeton Univ. Press,
Princeton, N. J. 217 pp.

Frydendall, M. J.
1969. Rodent populations on four habitats in central Kansas.
Trans. Kansas Acad. Sci., 72:213-222.

Genoways, H. H., and D. A. Schlitter
1966. Northward dispersal of the hispid cotton rat in Nebraska and
Missouri. Trans. Kansas Acad. Sci., 69:356-357.

Getz, L. L.
1962. Aggressive behavior of the meadow and prairie vole.
J. Mammal., 43:351-358.

1972. Social structure and aggressive behavior in a population of
‘Microtus pennsylvanicus. J. Mammal., 53:310-317.

 

94

Gier, H. T.
1967. The Kansas small mammal census: terminal report. Trans.
Kansas Acad. Sci., 70:505-518.

Goertz, J. W.
1964. The influence of habitat quality upon density of cotton rat
populations. Ecol. Monographs, 34:359-381.

 

1971. An ecological study of Microtus pinetorum in Oklahoma.
Am. Midlo Nat., 86:1-120

 

Golley, F. B.
1960. Energy dynamics of a food chain of an old field community.
Ecol. Monographs, 30:187-206.

(kmu,P.R.
1970. Experimental studies of competitive interaction in a two-
species system. II. The behaviour of Microtus, Peromyscus,
and Clethrionomys species. Anim. Behav., 18:411-426.

 

 

1971. Experimental studies of competitive interaction in a two-
species system. III. Microtus and Peromyscus species in
enclosures. J. Animal Ecology, 40:323-350.

 

 

Haines, H.
1971. Characteristics of a cotton rat (Sigmodon hispidus) popula-
tion cycle. Texas J. Sci., 23:3-27.

 

Hall, E. R., and K. R. Kelson
1959. The mammals of North America. Vol. II. Ronald Press,
New York.

Hardin, G. _
1960. The competitive exclusion principle. Science, 131:1292-1297.

Hays, H. A.
1958. The effect of microclimate on the distribution of small

mammals in a tall-grass prairie plot. Trans. Kansas Acad.
Sci., 61:40-63.

Howell, J. C.
1954. Populations and home ranges of small mammals on an over-
grown field. J. Mamma1., 35:177-186.

Hutchinson, G. E.

1953. The concept of pattern in ecology. Proc. Acad. Nat. Sci.,
105:1-12.

95

Jameson, E. W., Jr.
1947. Natural history of the prairie vole. Univ. Kansas Publ.,
Mus. Nat. Hist., 1:125-151.

Jennrich, R. T., and F. B. Turner
1969. Measurement of non-circular home range. J. Theoret. Biol.,
22:227-237.

Jones, J. R., Jr.
1960. The hispid cotton rat in Nebraska. J. Mammal., 41:132.

Joule, J. J., and D. L. Jameson
1972. Experimental manipulation of population density in three
sympatric rodents. Ecology, 53:653-660.

Keller, B. L., and C. J. Krebs
1970. Microtus papulation biology; III. Reproductive changes in
fluctuating populations of M. ochrggaster and M. pennsyl-

vanicus in southern Indiana, 1965-1967. Ecol. _Monographs,
40: 263-294.

 

Klopfer, P. H.
1962. Behavioral aspects of ecology. Prentice-Hall, Inc.,
Englewood Cliffs, N. J. 166 pp.

 

1969. Habitats and territories: a study of the social use of space
by animals. Basic Books, Inc., New York. 117 pp.

Koplin, J. R., and R. S. Hoffman
1968. Habitat overlap and competitive exclusion in voles (Microtus).
Am. Midlo Nat., 80:494-507.

Krebs, C. J.
1966. Demographic changes in fluctuating populations of Microtus
californicus. Ecol. Monographs, 36:239-273.

 

., B. L. Keller, and R. Tamarin

1969. MiCrotus population biology: Demographic changes in fluctu-
ating populations 0f.§r ochrogaster and M, pennsylvanicus
in southern Indiana. Ecology, 50:587-607.

 

 

 

 

1970. Microtus population biology: Behavioral changes associated
with the population cycle in M. ochrogaster and.M§ pennsyl-
vanicus. Ecology, 51. 34-52.

 

., B. L. Keller, and J. H. Myers
1971. Microtus population densities and soil nutrients in southern
Indiana grasslands. Ecology, 52:660-663.

 

96

MacArthur, R. H.
1972. Geographical ecology: patterns in the distribution of
species. Harper and Row, Publ., New York. 269 pp.

Martin, E. P.
1956. A population study of the prairie vole (Microtus ochrogaster)

in northeastern Kansas. Univ. Kansas Publ., Mus. Nat. Hist.,

 

1960. Distribution of native mammals among the communities of the
mixed prairies. Fort Hays Studies (n.s.), Sci. Ser., 1:1—26.

 

McCulloch, C. Y.
1962. Populations and range effects of rodents on the sand sage-
brush grasslands of western Oklahoma. Oklahoma State Univ.
Publ., 59:1-112.

Meserve, P. L.
1970. Population ecology of the prairie vole, Microtus ochrogaster,
in the western mixed prairie of Nebraska. Am. Midl. Nat.,
86:417-433.

 

Metzgar, L. H., and R. Hill
1971. The measurement of dispersion in small mammal populations.
J. Mammal., 52:12-20.

Miller, R. S.
1967. Pattern and process in competition. Pages 1'74.$E.J° B.
Cragg, ed. Advances in ecological research. Vol. 4.
Academic Press, London.

Mohlenrich, J. S.
1961. Distribution and ecology of the hispid and least cotton rats
in New Mexico. J. Mammal., 42:13-24.

Morris, R. D.
1969. Competitive exclusion between Microtus and Clethrionomys in
the aspen parkland of Saskatchewan. J. Mammal., 50:291-301.

 

Mossman, A. S. .
1955. Light penetration in relation to small mammal abundance.
Jo Manual. , 36:564-566.

Murie, J. 0.
1971. Behavioral relationships between two sympatric voles
(MicrOtus): relevance to habitat segregation. J. Mammal.,
52:181-186.

Odum, E. P.
1955. An eleven year history of a Sigmodon population. J. Mammal.,
36:368-378.

97

Petersen, M. K.
1970. Competition between the cotton rats, Sigmodon fulviventer
and_§.”hi$pidus. Unpublished Ph. D. thesis, Michigan State
Univ., 110 pp.

 

Phillips, P.
1936. The distribution of rodents in overgrazed and normal grass-
lands of central Oklahoma. Ecology, 17:673-679.

Ramsey, P. R., and L. A. Briese
1971. Effects of immigration on the spatial structure of a small
mammal community. Acta Theriologica, 16:191-202.

Raun, G. G., and B. J. Wilks
1964. Natural history of Baiomysrtaylori in southern Texas, and
competition with Sigmodon hispidus in a mixed population.
Texas J. Sci., 16:28-49.

 

 

Sealander, J. A., and B. Z. Walker
1955. A study of the cotton rat in northwestern Arkansas. Proc.
Arkansas Acad. Sci., 8:153-162.

Siegel, S.
1956. Nonparametric statistics for the behavioral sciences.
McGraw-Hill Book Co., New York. 312 pp.

Svihla, A.
1929. Life history notes on Sigmodon hispidus hispidus. Ecology,
10:352-353.

 

Terman, M. R., and D. I. Johnson
1971. The effects of Sigmodon hispidus on Microtus pennsylvanicus

 

 

in a confined area. Trans. Kansas Acad. Sci., 74:217-220.

Van Vleck, D. B.
1968. Movements of Microtus pennsylvanicus in relation to depop-
ulated areas. J. Mammal., 49:92-103.

Wendlend, V.
1970. Vikarianz bei der Nordischen Wilhmaus (Microtus oeconomus)
und der Erdmaus, Microtus agrestis, im Westberliner Raum.
Z. Saugetierk, 35:51-56.

 

 

Whittaker, J. 0., and E. G. Zimmerman

1968. Microtus ochrogaster in Alabama and Tennessee. J. Mammal.,
49:328..

 

Wiegert, R. G.
1972. Population dynamics of cotton rats (SigModon M. hispidus)
and meadow voles (Microtus p. pennsylvanicus) in field
enclosures in South Carolina. Bull. Georgia Acad. Sci.,

_i_p press .

 

98

Yang, Kee-Chum, C. J. Krebs, and B. L. Keller
1970. Sequential live-trapping and snap—trapping studies of
MicrOtus populations. J. Mammal., 51:517-526.

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

um"WWIIulmrummlungumwuw