HHHHM \H

    

H

_o
N

   

 

7* :3 Von-“ # q .
- , ‘7:—

m ‘8 MICHIGAN STATE UNIVERSITY LIBRARIES

2-1 - ll!“ UNI"!l1WINHIIWIHHHWWllllflllll

3 1293 01572 8870

 

 

 

 

 

 

 

This is to certify that the

thesis entitled

REPRODUCTIVE BIOLOGY AND BIOLOGICAL RHYTHMS IN
ARVICANTHIS NILOTICUS

 

presented by

Teresa L. McElhinny

has been accepted towards fulfillment
of the requirements for

_____M._S_._degree in immi—

5 g Major professor:f

Date 8"}3’fg

0-7539 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

LIBRARY
Michigan State
Unlvorsity

 

 

 

PLACE N RETURN BOX to man this chockou! from your mood.
TO AVOID FINES Mum on or bdoro duo duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

-L_i::l

 

REPRODUCTIVE BIOLOGY AND BIOLOGICAL RHYTHMS IN
Arvicanthis niloticus

By

Teresa L. McElhinny

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

MASTER OF SCIENCE

Department of Zoology

1996

ABSTRACT

REPRODUCTIVE BIOLOGY AND BIOLOGICAL RHYTHMS IN
Arvicanthis niloticus

By

Teresa L. McElhinny

Arvicanthis niloticus are small-bodied, murid rodents that inhabit sub-Saharan
Africa. The taxonomy of the genus Arvicanthis is the source of much debate: some
authors recognize just one species, whereas others describe up to six species. The
biology of A. niloticus is poorly understood, as the species has been little studied,
therefore the purpose of the work presented in this thesis was to elucidate aspects of the
reproductive biology and biological rhythms of captive A. niloticus. My data suggested
that multiparous female A. niloticus ovulate spontaneously, at intervals of 5-7 days. The
mean duration of the gestation period was 24.25 days. This species exhibits a copulatory
pattern similar to other murids. I found that A. niloticus are diurnal with respect to
general activity and core body temperature rhythms, although peaks in activity and
temperature occur at dawn and dusk. Copulation tends to occur during the morning peak

of activity, while parturition occurs outside the times of peak activity.

To Pat

iii

ACKNOWLEDGMENTS

I wish to thank my major professor Dr. Kay E. Holekamp, for her guidance,
support, and patience throughout my master’s program. I am very grateful to Dr. Laura
Smale for allowing me free reign to work in her laboratory.

I sincerely appreciate the suggestions, ideas and comments from the members of
my guidance committee, Dr. Kay E. Holekamp, Dr. Laura Smale, and Dr. Lynwood
Clemens.

There are two families that I wish to acknowledge. The McElhinny family, my
parents, brothers and sister, were very supportive of me, even though they probably
secretly worried that my higher education involved watching rats copulate. My second
family, the Holekamp lab, Micaela Szykman, Anne Engh, Carmen Salsbury, Erin
Boydston, Scott Nunes, and Amy Barber, were tremendously supportive, and I owe this
degree in part to them because of their infectious confidence in me.

Jacob Pegg, David Kent, and Justin Kiger provided long hours of assistance in the
laboratory. Bruce Bills translated an article from the original German. Dr. Evelyn Rivera
taught me the vaginal smear technique. Dr. Leslie Meek and Dr. Carmen Salsbury aided
in the development of the transmitter surgery protocol, and provided many useful

suggestions during the course of my research.

iv

Pat Bills provided invaluable logistical and emotional support, as well as 24 hour
access to his computer expertise.

Finally, I wish to thank Dr. James Atkinson and Dr. Martin Balaban for
encouraging me to attend graduate school, and for eliciting within me the courage

succeed.

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................ ix

LIST OF FIGURES ........................................................................................................ x

PREFACE ......................................................................................................................... 1
CHAPTER I

Arvicanthz’s niloticus NATURAL HISTORY .......................................................... 4

1.1 DISTRIBUTION AND HABITAT ......................................................................... 4

1.2 TAXONOMY AND SYSTEMATICS .................................................................... 9

1.3 PHYSICAL CHARACTERISTICS ...................................................................... 17

1.4 ECOLOGY ............................................................................................................ 18

I.4.a Diet ................................................................................................................ 18

I.4.b A. niloticus as a pest ..................................................................................... 20

I.4.c Predators ....................................................................................................... 23

I.4.d Activity ......................................................................................................... 23

I.4.e Demography and Social Organization .......................................................... 25

I.4.f Burrow Configurations .................................................................................. 27

1.5 REPRODUCTIVE BIOLOGY .............................................................................. 28

I.5.a Seasonality of Reproduction ................................. . ....................................... 29

Eastern Africa .................................................................................................. 33

Western Africa ................................................................................................. 37

I.5.b Estrous Cycle ................................................................................................ 4O

I.S.c Ovulation ...................................................................................................... 40

I.5.d Gestation ....................................................................................................... 41

I.5.e Litter Size ...................................................................................................... 43

I.5.f Lactation Interval .......................................................................................... 43

1.5. g Grth and Development ............................................................................. 44

I.5.h Ages for Males and Females at Sexual Maturity .......................................... 45

I.5.i Life span ........................................................................................................ 46

vi

CHAPTER II

THE NON-PREGNANT ESTROUS CYCLE AND OVULATION ............... 47
11.1 INTRODUCTION ................................................................................................ 47
11.2 CYTOLOGICAL ASSAYS OF VAGINAL SMEARS ....................................... 51

II.2.a General Methods .......................................................................................... 53
Subject Animals and Housing Conditions ....................................................... 53
Cytological Assay of Vaginal Smears ............................................................. 53

II.2.b Initial Evaluation of Cell Types .................................................................. 54
Purpose ............................................................................................................. 54
Methods ............................................................................................................ 54
Results .............................................................................................................. 54

II.2.c Females Housed in Small Cages ................................................................. 54
Purpose ............................................................................................................. 54
Methods ............................................................................................................ 55
Results .............................................................................................................. 56

II.2.d Females Housed in Large Enclosures ......................................................... 56
Purpose ............................................................................................................. 56
Methods ............................................................................................................ 56
Results .............................................................................................................. 59

1.2.6 Conclusions ................................................................................................... 59

11.3 BEHAVIORAL ASSAYS .................................................................................... 63

II.3.a Male-Female Interactions ............................................................................ 63
Purpose ............................................................................................................. 63
Methods ............................................................................................................ 63
Results .............................................................................................................. 66
Conclusions ...................................................................................................... 66

II.3.b Male Response to Female Urine ................................................................. 67
Purpose ............................................................................................................. 67
Methods ............................................................................................................ 67
Results .............................................................................................................. 68
Conclusions ...................................................................................................... 69

II.3.c Activity Rhythms ......................................................................................... 69
Purpose ............................................................................................................. 71
Methods ............................................................................................................ 71

Transmitter Implantation Surgery Protocol .............................................. 71
Results and Conclusions .................................................................................. 73
11.4 OVULATION ...................................................................................................... 73

II.4.a Purpose ........................................................................................................ 73

II.4.b Methods ....................................................................................................... 75

II.4.c Results ......................................................................................................... 75

II.4.d Conclusions ................................................................................................. 77

11.5 DISCUSSION ...................................................................................................... 77

vii

CHAPTER III

THE MORPHOLOGY OF COPULATORY BEHAVIOR ................................ 79
111.1 INTRODUCTION .............................................................................................. 79
111.2 METHODS ......................................................................................................... 81
111.3 RESULTS ........................................................................................................... 84
111.4 DISCUSSION ..................................................................................................... 87

CHAPTER IV

TEMPORAL PATTERNS OF ACTIVITY, TEIVIPERATURE, AND

REPRODUCTIVE BEHAVIOR ............................................................................... 89
IV.1 INTRODUCTION .............................................................................................. 89
1V.2 ACTIVITY AND TEMPERATURE RHYTHMS ............................................. 91

1V.2.a Purpose ....................................................................................................... 91
1V.2.b Methods ..................................................................................................... 92
1V.2.c Results ........................................................................................................ 94
IV.3 TIMING OF COPULATION AND PARTURITION ...................................... 104
IV.3.a Purpose ..................................................................................................... 104
IV.3.b Methods ................................................................................................... 104
IV.3.c Results ...................................................................................................... 105
IV.4 DISCUSSION ................................................................................................... 107

CONCLUSIONS AND RECOMIVIENDATIONS ............................................. 112

APPENDICES
Appendix A. Embedding Protocol for Ovarian Histology ........................................ 114
Appendix B. Staining Protocol for Ovarian Histology ............................................. 115

LIST OF REFERENCES .......................................................................................... 116

viii

 

Table 1.

Table 2.

Table 3.

Table 4.

Table 5.

Table 6.

Table 7.

 

 

LIST OF TABLES

Karyotypes of Arvicanthis from different geographical regions in Africa. ........ 12
Relationship between rainy season and reproductive season in nine African
Arvicanthis populations. ..................................................................................... 31
Arvicanthis litter size in captivity and in the field. ............................................. 42
Mean measurements (:tSE, [range]) of 45 female and 15 male of A. niloticus
from the M.S.U. colony. ..................................................................................... 45
Categorization of nonpregnant female reproductive cycles (afier Conaway,
1971). .................................................................................................................. 49
Number of multiparous and nulliparous female A. niloticus exhibiting signs of
ovulation. ............................................................................................................ 77
Percentage of one-hour bins during the dark phase in which activity levels fall

below the light phase trough. ............................................................................ 101

ix

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.
Figure 6.

Figure 7.

Figure 8.

Figure 9.

Figure 10.
Figure 11.
Figure 12.
Figure 13.

Figure 14.

Figure 15.

LIST OF FIGURES

Arvicanthis niloticus (Rosevear, 1969). .......................................................... 2

African countries in which Arvicanthis are known to occur (base map from

Afiica Today, an Atlas of Reproducible Pages. 1983) .................................... 6
Species range of Arvicanthis niloticus (Kingdon, 1974). ............................... 7
Distribution of vegetation types in Africa (Afiica Today, an Atlas of

Reproducible Pages. 1983). ............................................................................ 8
Distribution of described karyotypes of Arvicanthis across Africa. ............. 13
Activity rhythm of an individual Arvicanthis niloticus from Uganda. ......... 24

The relationship between rainfall, temperature, and breeding season in
Arvicanthis .................................................................................................... 33

Cyclic changes in vaginal cell types (afier Turner and Bagnara, 1971) ....... 52
Vaginal smear fi'om a nulliparous female A. niloticus. ................................. 55

Persistent estrous vaginal smears from a nulliparous female A. niloticus. ...57

Erratic vaginal smear series from a nulliparous female A. niloticus ............. 58
A five day estrous cycle from a nulliparous female A. niloticus. ................. 60
Sample check sheet for scoring male-female interactions ............................ 65

Male response to female urine sample represented by the number of seconds
spent sniffing the sample .............................................................................. 70

Number of standard deviations from mean activity measured over a period
of 31 days. ..................................................................................................... 74

Figure 16.
Figure 17.
Figure 18.
Figure 19.

Figure 20.

Figure 21.

Figure 22.

Figure 23.

Figure 24.

Figure 25.

Figure 26.

Figure 27.

Nulliparous and multiparous ovaries of female A. niloticus. ........................ 76
Patterns of mammalian copulation (Dewsbury, 1972) .................................. 80
Sample check sheet for scoring copulatory behavior .................................... 83
Number of mounts per ejaculation. ............................................................... 85
Postej aculatory interval per ejaculation. ....................................................... 86

Data collection scheme for the study of biological rhythms in
A. niloticus. ................................................................................................... 93

Biological rhythms in female A. niloticus ..................................................... 96
A comparison of females with running wheels (0) vs. females without
running wheels (0) with respect to core body temperature and gross motor

activity ........................................................................................................... 98

A comparison of patterns of gross motor activity and core body temperature
in female A. niloticus with running wheels (B) vs. females without running

wheels (A). .................................................................................................. 100
A comparison of the morning rise in gross motor activity and in

temperature. ................................................................................................ 103
Timing of copulation in captive A. niloticus ............................................... 106
Timing of parturition in captive A. niloticus ............................................... 108

xi

PREFACE

The most common mammals on earth are short lived, herbivorous rodents living
in tropical habitats. The Order Rodentia contains 40% of all living mammals, and four to
five times as many species of mammals live in the tropics as are found in the rest of the
world (Nowak, 1991). Despite the diversity and ecological importance of tropical
rodents, little is known about their reproductive biology. Most research in this area has
involved the use of domesticated laboratory species from temperate regions, and has not
reflected the importance or diversity of the group. In the interest of furthering the study
of reproductive biology in mammals, and to ensure that we account for the diversity that
exists, more consideration should be given to tropical mammals, and in particular, to
tropical rodents.

1 have studied several behavioral and physiological aspects of reproductive
biology, as well as circadian rhythms of temperature and activity, in the female of one
tropical rodent species, Arvicanthis niloticus (Figure l). A. niloticus are herbivorous
rodents which inhabit tropical Africa. With their blunt faces and stocky bodies, A.
niloticus are physically similar to voles (Kingdon, 1974; Carleton and Musser, 1984),
although they are larger, and have longer tails and more prominent pinnae. This species
is an excellent candidate for laboratory study because it is handled fairly easily, and it

breeds well in captivity. Relatively low levels of aggression allow animals of the same

1

3

5’
1
3

l

J

I

.1

.5

:f

g a
:
"1
..Ll

 

Figure 1. Arvicanthis niloticus (Rosevear 1969).

sex to be housed together, and males can remain housed with their mates without danger
of pups being cannibalized. Behavioral study of A. niloticus is facilitated by their
tendency to be active during hours of daylight.

This thesis is divided into chapters which address the specific topics of my
research. Chapter I is a review of the literature concerning Arvicanthis, including
taxonomy, systematics, ecology, and reproductive biology. In Chapter II, I discuss
experiments that I performed in order to elucidate the estrous cycle in A. niloticus. These
include vaginal smears, paired encounters between males and females, study of male
responses to female odors, and analysis of circadian rhythms in the female. Chapter 11
also contains discussion of an experiment in which I examined the ovaries of females of
different ages and levels of sexual experience in order to determine whether A. niloticus
ovulate spontaneously or reflexively. In Chapter 111, 1 present work in which I used video
analysis to describe the morphology of copulation in this species, to elucidate timing of
parturition, and to calculate a precise length of the gestation period. In Chapter IV, I

present work which describes several biological rhythms of A. niloticus. 1 explored the

3
circadian rhythrnicity of this species, describing rhythms of temperature, gross motor
activity, and wheel-running. I also looked at the timing of copulation and parturition in

relation to circadian activity and temperature rhythms.

CHAPTER I

ARI VCAN THIS NILOT 1C US NATURAL HISTORY

The purpose of this chapter is to present a review of existing information about
the genus Arvicanthis in general, and about Arvicanthis niloticus in particular. In
subsequent chapters 1 will attempt to add to this knowledge pool by describing my own

work involving the reproductive biology and biological rhythms of A. niloticus.

1.1 DISTRIBUTION AND HABITAT

Arvicanthis (Lesson, 1842) is a tropical genus, inhabiting the Nile valley, the
southwestern part of the Arabian peninsula (Harrison, 1972), and much of Africa from
Senegal to Somalia (Kingdon, 1974), south of the Sahara and north of the Zambezi river
(Neal, 1981). The common name, the Nile grass rat, refers to the fact that the first
specimen known to Europeans came from the Nile Valley (Rosevear, 1969). While the
range of Arvicanthis spans the continent, the members of this genus occur most
commonly in East Africa (Kingdon, 1974). Specifically, species belonging to this genus
occur to the west in Mauritania (Musser and Carleton, 1993), Senegal (Rousseau, 1983),
Mali (Rousseau, 1983), Burkina F aso (Rousseau, 1983; Karninski et al, 1984), Gambia
(Karninski et al, 1987), Guinea, Sierra Leone (Harrison, 1972), Ivory Coast (Musser and

Carleton, 1993), Ghana (Harrison, 1972), Togo, and Benin (Musser and Carleton, 1993).
4

5
In central Africa Arvicanthis have been found in Niger (Rousseau, 1983), Chad
(Rousseau, 1983), Nigeria (Harrison, 1972), and the Central Afiican Republic (Matthey,
1965; Rousseau, 1983). In northeastern Africa they occur in Egypt (Karninski et al,
1984), Sudan (Rousseau, 1983), Ethiopia (Corbet and Yalden, 1972; Dorst, 1972; Yalden
et al, 1976; Yalden, 1988), and the Somali Republic (Drake-Brockman, 1910). On the
Arabian peninsula they have been found in Yemen (Harrison, 1972). In eastern Africa
they occur in Kenya (Coe, 1973; Kingdon, 1974), Uganda (Bere, 1962; Delany and Neal,
1966; Delany, 1975), Tanzania (Allen and Loveridge, 1933; Kingdon, 1974; Packer,
1983; Senzota, 1983), Burundi, Zaire (Musser and Carleton, 1993), and south into
Zambia (Ansell, 1960; Ansell, 1978) (Figure 2).

The distribution of Arvicanthis within the countries mentioned above (Figure 3) is
dictated by the type of vegetation present. The preferred habitat of Arvicanthis is
described as dry savannas, woodlands and grasslands (Kingdon, 1974). Arvicanthis are
found in the northern and central Sudan along rivers and canals (Schmutterer, 1969), and
Delany (1986) observed A. niloticus living along river edges in Kenya. Arvicanthis are
also associated with water to the north where their range runs along the Nile in Egypt. In
areas where the species is commensal with man, they can be found in cultivated land,
native huts and grain stores (Delany and Neal, 1966). A. niloticus appear to be limited in
distribution by dense forest and desert (Poulet and Poupon, 1978). Comparing the
species distribution map in Figure 3 with the vegetation map in Figure 4, one can see that
Arvicanthis prefer open woodlands and savannas, but avoid the Sahara and Sudanese
deserts to the north, as well as the wet lowland forest of the Congo Basin in central

Africa.

 

     

"1

 

INDIAN

 
  

QCEAN
JTLANTIC
'- OCEAN .1
.
-' 0;-
fl .-
”ww-
can

mn—u—u-I—nnn-Ou-

“I‘— v-—-.O~l_~
' .-
1 l l 1

.—

”trauma-onus hm.
'
\ J | I

 

 

 

Figure 2. African countries in which Arvicanthis are known to occur (base map from
Africa Today, an Atlas of Reproducible Pages. 1983).

 

Figure 3. Species range of Arvicanthis niloticus (Kingdon, 1974).

 

 

3'
x,
I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

- Montana -“
Lowland forest _._.___

 

Moist woodlands. savannas,etc
Dry woodlands, steppe, etc.
D Desert and subdeeert

- Mocctsta ..

 

 

 

 

 

 

 

 

 

Figure 4. Simplified vegetation of Africa (Africa Today, an Atlas of Reproducible Pages.
1983).

1.2 TAXONOMY AND SYSTEMATICS

Rats of the genus Arvicanthis are ground dwelling rodents of the family Muridae,
subfamily Murinae. Murinae is the largest subfamily within the Muridae, and represents
a very heterogeneous group. Graur (1994) suggested that neither Muridae nor Murinae is
monophyletic, and that these taxonomic groups are in need of revision. Chromosomal
phylogenetics (V iegas-Pequignot et al, 1983; Chevret et al, 1993) and sperm morphology
(Baskevich and Lavrenchenko, 1995) have shown that, although Arvicanthis and Rattus
occur in the same subfamily, they are probably rather distant relatives. Viegas-Pequignot
et al. (1983) suggested a “dichotomic evolution” (p. 276) for the two groups from which
the genera Arvicanthis and Rattus arise. The electrophoretic studies of Bonhomme et al.
(1985) showed that the 16 African murid genera (including Arvicanthis) they analyzed
form a monophyletic subgroup separate fi'om Rattus and Mus. Within Africa, Arvicanthis
are generally allied with the other herbivorous murids, based on dental and cranial
characteristics (Misonne, 1969). These other herbivorous murids are: Pelomys,
Mylomys, Hybomys, Lemniscomys, Rhabdomys, and Dasymys (Kingdon, 1974).
Arvicanthis are distinguished from other African murid genera by their relatively dark,
coarse pelage and smooth incisors.

Taxonomic relationships within the genus Arvicanthis are poorly understood, with
17 to 37 forms recorded by early workers (Dollrnan, 1911; Allen, 1939; Ellerman, 1940).
Authors today suggest the existence of from one (Rosevear, 1969; Misonne, 1971;
Honacki et al, 1982) to five (Nowak, 1991; Musser and Carleton, 1993) distinct species.
The most recent treatment of the genus is by Musser and Carleton (1993). These authors

recognize 5 species: one (A. niloticus) ranges across Africa from Senegal to Ethiopia,

10
north to Egypt and South to Zambia, two species (A. abyssinicus and A. blicki) are
endemic to Ethiopia, and two Eastern African species (A. nairobae and A. somalicus)
occur within, and east of, the Rift Valley. There is some disagreement in the literature
regarding the scientific name of the most widespread species of Arvicanthis: some
investigators refer to it as A. niloticus (e. g. Musser and Carleton, 1993), while others
prefer A. abyssinicus (e.g. Nowak, 1991).

There has been a recent effort to revise the taxonomic status of Arvicanthis
through analysis of blood proteins and chromosome number. Kaminski et al. (1984) and
Kaminski and Petter (1984) studied the blood proteins albumin, transferrin, esterase and
6-phosphogluconate dehydrogenase in wild-caught animals from three Afiican countries:
Senegal, Burkina F aso and Egypt. Kaminski et al. (1984) found inter-population
differences and intra-population genetic polymorphism. Kaminski et al. (1984) concluded
that at least 3 forms could be recognized based on structural differences among the
proteins examined. The authors did not interpret this as evidence for speciation, but
rather as a step toward speciation. The polymorphisms in blood protein composition,
found between populations in east and west Africa, highlight the plasticity of
Arvicanthis, and may indicate that further speciation is in progress within this genus
(Karninski et al. 1984).

Kaminski et al. (1984) successfully bred a female Arvicanthis from Senegal with
a male from Cairo to produce four successful litters. Petter et al. (1969) were also
successful in breeding a series of mating couples crossed from Senegal and Ethiopia.
Production of offspring by pairing animals fi'om such distant localities strengthens the

argument of Musser and Carleton (1993) that one species inhabits the huge geographic

11
area shaded in Figure 3.

Kaminski et al. (1987) further analyzed the blood proteins of 152 Arvicanthis
from 11 localities in Senegal. They found polymorphisms among animals from the
northern and southern localities. Their northern collecting sites were along the Senegal
river and along Cape Vert. There were only two southern collecting sites, at Cape
Skirting in Gambia, and Kedougou in southeastern Senegal. These authors hypothesized
that the animals from the south were similar to the specimens they had previously studied
from Burkina Faso (Kaminski et al, 1984). Kaminski et al. (1987) determined that
together these animals from southern Senegal, Gambia, and Burkina Faso differed from
those from northern Senegal, constituting a species different from Arvicanthis niloticus.
Thus, two species may exist in western Africa, A. niloticus to the north in Senegal, and
another species to the south and east in Gambia and Burkina Faso. Kaminski et al.
(1987) found no major reproductive barrier between the two groups in the natural
environment, but stated that Petter unsuccessfully attempted to breed animals from
Senegal and Burkina F aso. The authors did not state whether they believed these two
forms to be sympatric in western Afiica.

The first studies of chromosome number in Arvicanthis were performed by
Matthey in 1959 and 1965 (see Table 1). Matthey (1959) described the first specimen as
A. abyssinicus, and reported a karyotype of 62, but did not report a collection site.
Matthey’s (1965) second specimen, described as A. niloticus, was collected in the Central
African Republic, and its karyotype was reported as 56 (Figure 5). Veigas-Pequignot et

al. (1983) reported the karyotype of an Arvicanthis from Cairo as 62 (n= 1 individual). In

12

Table 1. Karyotypes of Arvicanthis from different geographic regions in Africa.

 

Investigator Geographic Region Karyotype

Matthey (1959) ? 62

Matthey (1965) Bangui, Central African 56
Republic

Viegas-Paquignot et al. Cairo, Egypt 62

(1983)

Capanna et al. (1985) Afgoi, Somalia 44

Volobouev et al. (1987) Lakoumbala, Central 58
Afiican Republic

Volobouev et al. (1988) Bamako, Mali 62

Volobouev et al. (1988) Oursi, Burkina F aso 62

Volobouev et al. (1988) Brancon, Senegal 62

Orlov (1990) Gambella, Ethiopia 62
Ambo, Ethiopia 62
Awash River Valley, 56
Ethiopia

Gamo-Gofo, Ethiopia 60

 

13

 

r r V W
I‘ 0' fl"““

_,
9‘}.

 
  

‘temf

. I
PO‘ . V
ATLANTIC

P.‘
-.- J.-
- .
WWW”
“f."

mmmmmmamm... L ‘ W;
’ r r r *1

 

l

 

 

Figure 5. Distribution of described karyotypes of Arvicanthis across Africa. X ' and X"
karyotypes indicate the same species in different geographic areas (base map fromAfrica
Today, an Atlas of Reproducible Pages. 1983).

14

1987, Volobouev et al. compared this number with that of an individual from the Central
African Republic, which was 58 (n= 1). Based on differences in chromosome number
and chromosomal structure, the authors concluded that at least two distinct species
existed in Africa. Capanna and Civitelli (1988) found that a single specimen of
Arvicanthis from Somalia, which they called A. niloticus, displayed a karyotype of 44.
These authors concluded that each of the karyotypically differentiated populations
mentioned above should be considered a separate species. Capanna and Civitelli (1988)
hypothesized that the description of animals with 56 (Matthey, 1965) and 58 (V olobouev
et al, 1987) chromosomes in the Central African Republic may be the result of the
presence of an intermediate evolutionary stage of balanced chromosomal polymorphism.
As there is no further information available, 1 will consider the animals from the Central
African Republic to represent one species, different from A. niloticus.

Volobouev et al. (1988) examined animals from Egypt (n=1), Senegal (n= 1),
Mali (n= 1), and Burkina Faso (n= 1), and proposed that at least 3 species of Arvicanthis
exist across Africa. From their sample, these investigators found three groups of animals
with identical karyotypes from 1) Egypt and Senegal (karyotype= 62), 2) Burkina Faso
and Mali (karyotype= 62), and 3) the Central African Republic (karyotype=58). While
the animals from the first two groups described have the same number of chromosomes,
they are considered different karyotypes because the positions of the chromosomal
centromeres are different. In animals from Egypt and Senegal, all but one pair of
chromosomes were acrocentric, while in animals from Burkina Faso and Mali, only 22
pair were acrocentric, and the rest were metacentric and submetacentric. Volobouev et a1.

(1988) concluded that the karyotypes of these animals were adequately different to merit

15
classification as three distinct species. This would not conflict with results of the cross-
breeding experiments mentioned above (Kaminski et al, 1984, 1987), since Volobouev et
al. (1988) considered animals from Senegal and Egypt to represent the same species of
Arvicanthis, different from the species found in Burkina F aso. Volobouev et al. (1988)
proposed that animals from Ethiopia are probably of the same species as those from
Senegal and Egypt, since Petter et al. (1969) were able to successfully cross animals from
Senegal and Ethiopia. Volobouev et al. (1988) attempted to relate the three species that
they found to forms previously described which are reported to have similar geographic
distributions, but concluded that a more complete study is needed to establish the
distribution limits of each species. The taxonomy proposed by Volobouev et al. (1988) is
as follows: one species (A. niloticus) occurs in Senegal, Egypt and possibly Ethiopia, a
second (A. centralis) in the Central African Republic, and a third species (A. solatus) in
Burkina Faso and Mali.

The results reported by Volobouev et al. (1988) suggest that three species of
Arvicanthis exist in West Africa (A. niloticus, A. centralis and A. solatus), which would
increase the total number of species of Arvicanthis in Africa to seven, including species
based on morphological characteristics: the two Ethiopian endemics (A. abyssinicus and
A. blicki) and the two species present east of the Rift Valley (A. nairobae and A.
somalicus). Capanna and Civitelli (1988) maintain that their specimen from Somalia
(2n=44) is A. niloticus, not A. somalicus, which would further divide A. niloticus (2n=62,
as per Volobouev et al, 1988), and may bring the grand total of species within the genus
to eight.

The most recent treatment of chromosome number in Arvicanthis comes from

16
Orlov et al. (1990), in which the authors examined animals from four different regions in
Ethiopia. They found that all four areas yielded animals with karyotypes unique to one
another. Two groups possessed karyotypes of 62 chromosomes, yet the morphology of
the chromosomes was different enough that the authors classified them as different
species (A. abyssinicus and A. dembeensis). This classification is supported by analysis
of cranial morphology (Bekele et al, 1993). The other two groups described by Orlov et
al. (1990), with chromosome numbers of 56 and 60, were not classified. It is apparent
from the chromosome numbers however, that these two animals are species distinct from
one another, and from the aforementioned A. abyssinicus and A. dembeensis. Although
the taxonomy of Arvicanthis in Ethiopia remains puzzling, it is clear that the inclusion of
these animals in the taxonomy would put the total number of species in the genus
Arvicanthis above eight.

Kingdon (1974) indicated that a subfossil of Arvicanthis was found in Palestine,
and Capanna and Civitelli (1988) hypothesized that Arvicanthis colonized Africa through
a southern and southwesterly spread from Egypt. Arvicanthis are known from Algeria in
the Middle Pleistocene and the Holocene. Arvicanthis are thought to have migrated
through the Sahara and possibly along the Atlantic coast during the Middle Pleistocene
(Kowalski and Rzebik-Kowalska, 1991). Consequently, Capanna and Civitelli (1988)
suggested that the karyotype of the Egyptian animals (2n = 62; Viegas-Paquignot et al,
1983) is ancestral, and that the differences in chromosome number seen today are a result
of changes that occurred during the southerly range expansion across Africa.

The biological species concept describes a species as a group of interbreeding or

potentially interbreeding papulations of individuals incapable of breeding with

17
individuals from other such populations (Mayr, 1942). In the future, it would be prudent
to perform cross-breeding experiments, in addition to exploring the distribution patterns
of each of the Arvicanthis groups described here. This information would determine
whether description of these groups as species is warranted, according to the biological
species concept.

There is currently very little information available regarding blood proteins or
karyotypes from Arvicanthis captured in eastern Africa south of Egypt and Ethiopia.
Thus, the morphologically based taxonomy of Musser and Carleton (1993) of Arvicanthis
in eastern Afiica remains unmodified. The animals used in the experiments reported here
were captured at two locations west of the Rift Valley, and preliminary results from
sequencing of the Sry gene of these animals indicate that they are different from A.
nairobae (Barbara Lundigran, personal communication). Therefore, for the purpose of
this dissertation I will follow Musser and Carleton (1993) in assuming that the only
species of Arvicanthis present in Kenya west of the Rift Valley is Arvicanthis niloticus, a
species which apparently ranges across Africa, and exhibits a great amount of intra- and
interpopulation morphological variation. Some of the information reported in this
chapter may refer to other species of Arvicanthis according to recent reorganization of the
genus reviewed here. Therefore, I have indicated the collecting location or study site for

locations outside the east African countries of Kenya, Uganda and Tanzania,.

1.3 PHYSICAL CHARACTERISTICS
Arvicanthis have been compared to the microtine rodents with respect to body

conformation (Carleton and Musser, 1984), and they do generally resemble large voles,

18
but with longer, thicker tails and more prominent pinnae. As noted earlier, A. niloticus
have coarse, dark fur on their dorsal surfaces, and lighter fur on the belly. On both the
fore and hind feet, the second, third and fourth phalanges are considerably longer than the
first and fifth phalanges. The tail is fairly well haired. The cars are usually reddish, and
are fairly large and rounded. Females have six mammae. The incisors are ungrooved, in
contrast to those of the similar genus, Mylomys (Rosevear, 1969). The periosteum
covering the cranium is darkly pigmented with melanin, apparently serving to absorb
solanradiation (Carleton and Musser, 1984). Kingdon’s (1974) mean length
measurements for adult Arvicanthis in east Afiica are as follows: head and body: 106-204
mm; tail: 100-152 mm; hind foot: 23-32 mm; and weight 50-120 g. In an analysis of
morphological variation of Arvicanthis across Africa, Rousseau (1983) found that the
most noticeable differences among the animals sampled were in body size. There was a

west-to-east cline in body size, with the animals in the east being the smallest.

1.4 ECOLOGY

1.4.a Diet

Data from fecal analysis, stomach content analysis, and direct observation of
foraging have revealed that the diet of A. niloticus consists primarily of leaves and stems
of herbaceous vegetable matter (both monocotyledons and dicotyledons), supplemented
with insects, seeds and fruits (Delany, 1964; Delany and Neal, 1966; Coe, 1972; Nandwa,
1973; Neal, 1981; Rabiu and Fisher, 1989; Delany and Monro, 1986; K. E. Holekamp,

personal communication). Gautun et al. (1985) noted that during periods of high

19
population density, Arvicanthis in Senegal would resort to feeding on the bark of Acacia
trees, causing considerable damage. Rabiu and Fisher (1989) reported that the diets of
males and females were sirrrilar. Grasses predominated in the diet during the dry season
(Rabiu and Fisher, 1989; Neal, 1981), and this prevalence continued into the early rainy
season (Taylor and Green, 1976; Delany and Monro, 1986). Seeds were taken as they
began to appear during the rainy season (Taylor and Green, 1976; Rabin and Fisher,
1989), and continued into the beginning of the dry season (Delany and Monro, 1986;
Neal, 1 98 1).

Rabin and Fisher (1989) noted that in general Arvicanthis in Nigeria tended to
become more omnivorous with the commencement of the rains, presumably because it
was during the rains that more variation in food types became available. Seeds were
preferred over grass leaves when both foods were present in (Delany and Monro, 1986).
Delany and Kansiimeruhanga (1970) tested wild-caught Arvicanthis from Uganda for
food preferences using seeds, insects and a variety of vegetation types. The animals
preferred cassava and wheat kernels over grasses, insects or the leaves of coffee, sweet
potato and Eucalyptus. Senzota (1982) referred to A. niloticus as opportunistic feeders
because the extent to which a given species of grass was clipped by the rats was linearly
related to the abundance of the species. He also reported that the chance of a particular
grass being eaten increased with its proximity to an A. niloticus burrow system. All grass
species within the animals’ foraging area were eaten. The foraging radius was up to
seven meters around an A. niloticus colony in the Serengeti, with the radius depending on
the abundance of grasses (Senzota, 1983). Packer (1983) reported that the average

foraging distance from burrows in his Serengeti study site was six meters.

20

Senzota (1983) has suggested that there is resource partitioning between A.
niloticus and savanna ungulates such as wildebeest (Connochaetes taurinus) and
Thomson’s gazelles (Gazella thomsonii). He observed that in Tanzania, where the diets
of the two groups overlap somewhat, the rats eat species of grass that the ungulates tend
to avoid. Senzota (1983) further noted that A. niloticus prefer the stems of the grasses,
ripping them apart with their claws and teeth, discarding the leaves. Ruminant ungulates
prefer to take the leaves, as the stems are harder for them to digest. Senzota (1983)
concluded that by eating species and parts of the grasses that ungulates tend to avoid, A.

niloticus may have facilitated resource packing in this part of the Serengeti ecosystem.

1.4.b A. niloticus as a Pest

A. niloticus populations have been known to reach remarkably high densities, and
also to crash periodically (Delany and Monro, 1985b). Animal numbers in dry bush
savanna during a population explosion in Senegal went from O/ha to 100/ha in a matter of
months (Poulet, 1972). During an A. niloticus outbreak in Tanzania associated with
burning of the savanna, A. niloticus numbers were concentrated such that “one could
hardly avoid stepping on them” (p. 426, Hubbard, 1972). The dietary preference for
seeds shown by A. niloticus has resulted in their classification as a pest species in areas of
Afiica where cereal crops are grown. Dieterlen (1990) reported that A. niloticus is

6‘

Number One on Egypt’s most wanted” list, and that the species may have been
responsible for rat plagues dating back to the time of the Pharaohs. A massive explosion
of A. niloticus populations in Kenya and Tanzania in 1951 and 1962 resulted in damage

to corn, wheat and barley crops. Also in 1962, A. niloticus were responsible for damage

21

to cotton crops in Sudan, disturbing seedlings as well as mature bolls during the harvest
period (Ripper and George, 1965; Taylor, 1968). Schmutterer (1969) described A.
niloticus as minor pests in the Sudan, gaining major pest status in certain years. In the
Sudan they have attacked dura (a cereal), pennisetum (a grass), maize, wheat, millet,
groundnuts, sesame, haricot beans, tomatoes, stored cereal products of various kinds, and
guava and mango trees when fruiting (Ripper and George, 1965; Schmutterer, 1969). A.
niloticus were found eating coffee, cassava, and sweet potato on farms in Uganda (Delany
and Kansiimeruhanga, 1970). During a 1975-1976 outbreak in Senegal, A. niloticus
attacked rice, sorghum, millet, and cassava (Myllymaki, 1979), and stripped bark and
shoots from Acacia trees, causing damage to 80% of the trees in one km2 study area. A.
niloticus displayed adaptive behavior in response to these unusually high population
densities. At such times these animals not only invaded the arid habitats that they usually
avoid, but they also climbed trees in order to forage (Poulet and Poupon, 1978).

Measures to control A. niloticus numbers in Sudan and Kenya have included
baiting them with the poisons Warfarin, zinc phosphide (Ripper and George, 1965;
Schmutterer, 1969), and Endrin (Taylor, 1968).

A. niloticus have been identified as a carrier of plague (Davis, 1962; Hubbard,
1972), and Nairobi sheep disease, a tick-bome virus that affects sheep and goats (Cox,
1979). They are highly susceptible to Rifi Valley fever (Davis, 1962), a mosquito-bome
virus which causes serious disease in lambs, and occasionally infects humans (Cox,
1979). Therefore, the population explosions mentioned above could potentially raise
health concerns for sympatric human populations, in addition to concerns about their

documented effects on agriculture.

22

Arvicanthis outbreaks are reminiscent of population explosions of microtine
rodents in northern temperate zones, since they cause damage to cereal crops, and
Microtus aggrestis has been known to strip the bark from trees (Myllymaki, 1979).
However, abrupt increases in the numbers of Arvicanthis do not occur on an annual or
cyclical basis (Taylor, 1968) as they do among the microtines (for a review, see Southern,
1979). Population increases of Arvicanthis are believed to be caused by unusually high
rainfall, which increases food availability, and in turn lengthens the breeding season
(Taylor, 1968; Poulet and Poupon, 1978; Sicard et al, 1994). This simple cause-effect
relationship involving increased food availability has been rejected by those who study
microtines (Krebs, 1996). Fluctuations in numbers of voles and lemmings are believed
to be caused by an interaction of intrinsic effects of spacing behavior (affecting dispersal,
mortality and reproduction) and extrinsic effects, most likely predation and food
availability (Krebs, 1996). This explanation of microtine population cycles might also
apply to A. niloticus, but more detailed observations of Arvicanthis population outbreaks
are required to determine whether this is true. Most of the reports involving massive
increases in Arvicanthis numbers have focused on the resulting damage to crops rather
than on the dynamics of the rodent populations involved. One exception is Poulet’s
(1976) description of the 1975 population surge in Senegal. A severe drought in 1972,
followed by steadily increasing annual rainfall, led to dramatically increased population
numbers of several rodent species in 1975-1976. High population densities led to the
migration of A. niloticus from their normal habitat in nomad campsites and cultivated
areas to the less desirable dry savanna. This opportunistic dispersal may be an important

feature of spacing behavior in this species. Once established in the dry savanna,

23
immigrant rats reproduced quite rapidly, and population densities soon exceeded habitat
carrying capacities, resulting in starvation and high mortality. Mortality was increased by
predation due to abnormally high concentrations of diurnal Palearctic birds of prey.
Thus, this outbreak of animals was affected by the extrinsic factors of food availability
and predation, and spacing behavior in the form of dispersal to less populated areas,

increased reproduction after dispersal, and mortality due to famine.

1.4.c Predators

Indigenous predators in the natural habitat of A. niloticus include spitting cobras
(Naja nigricolis), black-backed jackals (Canis mesomelas), long crested hawk eagles
(Lophoerus occipitalis), black shouldered kites (Elanus caeruleus), black headed herons
(Ardea melanocephala) (Senzota, 1990), dwarf mongooses (Helogale parvula) (Packer,

1983), and native Africans (Homo sapiens) (V esey-Fitzgerald, 1966).

I.4.d Activity

In their natural habitat, A. niloticus have generally been observed to be diurnal
(Quilici et al, 1969; Delany and Kansiimeruhanga, 1970; Kingdon, 1974; Packer, 1983;
Rabiu and Fisher, 1989), or diurnal with crepuscular tendencies (Senzota, 1990; see also
Katona and Smale, in press). Figure 6 shows an actogram for A. niloticus from Delany
and Kansiimeruhanga (1970). These data are from direct observations of movement and
feeding of a caged, wild-caught animal kept in its natural photoperiod. The figure
clearly shows that the animal was more active during the day. However, some

investigators consider A. niloticus to be partially nocturnal (Ansell, 1960; Delany and

24

Neal, 1966;

MINUTES

 

Figure 6. Activity rhythm of an individual Arvicanthis niloticus from Uganda. Time
active is indicated by the open bars, and time spent feeding by the closed bars (Delany

and Kansiimeruhanga, 1970).

Vesey-Fitzgerald, 1966; Rosevear, 1969; Harrison, 1972), or primarily nocturnal
(Schmutterer, 1969; Ghobrail and Hodeib, 1982).

Laboratory evaluation of general activity rhythms of A. niloticus from a
population 13°N of the equator revealed a somewhat diurnal pattern, with activity rising
after the light phase began to a broad peak at midday, and falling again before the dark
phase began (Duplantier and Granjon, 1990). However, the absolute peak level of
activity occurred during the dark phase, due to another rise in activity that began shortly
before the dark phase began, and peaked two hours later. Totals of hours during which
activity occurred showed A. niloticus to be equally active in the dark and light phases.
The authors did not report data from individual animals (they only reported mean values
for several individuals), therefore it is unknown whether this was a general pattern, or

whether the outcome reported was skewed by the rhythms of a few individuals.

25
Since these earlier reports from the literature describing the activity patterns of A.
niloticus were conflicting, an analysis of their activity was undertaken in the present

study (see Chapter IV).

1.4.e Demography and Social Organization

A. niloticus live in social groups called colonies (Delany and Happold, 1979). All
colony members share a burrow system and a common foraging area (Packer, 1983). It is
currently unknown how animals distribute themselves when they are underground, inside
their burrows. Colonies contain approximately equal numbers of males and females
(Delany and Monro, 1985b; Senzota, 1990). In Packer’s two year study of a single
colony in Tanzania, colony composition averaged 2.6 adult females, 3.1 adult males, and
4.9 immature individuals old enough to emerge from burrows. Senzota (1990) observed
that animals of different age and sex categories in Tanzania seemed to associate with each
other at random. He further noted that adults did not seem to defend their offspring
against predators.

Packer (1983) found that male A. niloticus in Tanzania changed colonies with
significantly greater frequency than did females. Thus, the colony consisted of female
kin and their young, together with adult males originating from elsewhere. Females
apparently may disperse to form new colonies, but do not join existing ones (Packer,
1983). Senzota (1990) found that there was a greater tendency for males than females in
Tanzania to move from one colony to another, but that this tendency was not statistically
significant. He also found that the number of colonies founded by males and females did

not differ significantly.

26

The observations of Delany and Monro (1985b) suggested the occurrence of more
movements between colonies than was indicated by Packer’s (1983) data. Delany and
Monro (1985b) stated that animals of both sexes at their Kenyan study site explored new
habitats, and this included emigration to distant existing colonies. Such movements were
seen throughout most of the year (Delany and Monro 1985b). This finding was contrary
to what was observed at Packer’s (1983) site in Tanzania, where irnrrrigration appeared to
be sexually dimorphic and seasonal, peaking between July and September.

Delany and Monro (1985b) reported home ranges for A. niloticus living along
field edges in Kenya, measured as the length of the line over which they were trapped. In
October, home ranges were 86 m for males and 47 m for females. Home ranges
measured from February through April, when animal densities were high, were much
smaller (males: 37 m, females: 38 m). Muller (1977) measured home ranges of
Arvicanthis in Ethiopia using a grid of traps. Home ranges during the rainy season for
males were 2750 m2, and for females and juveniles were 950 m2. 1n the dry season,
during which there were higher densities of animals, home ranges were 1400 m2 for
males and 600 m2 for females and juveniles. From the estimations of Delany and Monro
(1985b) and Muller (1977), it appears that the home ranges of female and juvenile
Arvicanthis are smaller than those of males, and that home-range size decreases with
increased animal density.

One should note that the home-range sizes mentioned above are much larger than
the foraging radius values of 6-7 meters reported by Packer (1983) and Senzota (1983).
The observations made by these two investigators were concentrated around the burrow

systems. They did not use trap lines or grids in efforts to estimate range sizes.

27
1.4.f Burrow Configurations

Vesey-Fitzgerald (1966) reported that A. niloticus in the Rukwa Valley, on the
border between Zambia and Tanzania, dig their own burrows. Delany and Neal (1966)
also saw evidence of animals digging their own burrows in Uganda. Poulet and Poupon
(1978) noted that Arvicanthis modify preexisting cavities and crevices in the soil. Vesey-
Fitzgerald (1966) reported that the burrow holes appeared to occur in pairs: one entrance
with one exit. 1n taking a census of burrow systems, he found that larger groups of holes
yielded no more animals than did smaller ones, and proposed that large burrows were
indicative of longer use, not larger colonies. Delany and Neal (1966) reported that
burrows of animals in Uganda were usually 20-70 cm deep, and Poulet and Poupon
(1978) excavated a burrow that was 30 cm deep. The burrow had multiple entrances that
led to a central chamber. In addition to the burrows, Delany and Neal (1966) noted that
the animals also constructed surface nests in tussocks of grass, but did not report the
circumstances under which these nests were used.

Arvicanthis construct their burrow systems at the bases of bushes, trees, rock
piles, banks, trash heaps, and tennitaria (Delany and Neal, 1966; Packer, 1983; Senzota,
1982). Roots of trees and bushes loosen the soil and make it easier for the rats to
construct their burrows, while the canopy cover prevents rain from flooding the burrows
(Senzota, 1982), and sufficient sunlight filters through to allow grass to become
established (Coe, 1972).

The animals maintain runways through the grass, and use these for travel around
the home range (V esey-F itzgerald, 1966). The rats voluntarily cut plant material and

remove small objects in their paths. This behavior has been observed even on ground

28
lacking vegetation (Senzota, 1990). Delany and Neal (1966) found that, when the grass
was high, such as during the rainy season, these runways appeared as tunnels through the
vegetation. Runways were longer, more visible, and more numerous during the dry
season (Senzota, 1990). Senzota (1990) proposed that the evolution of runway
construction may have been driven partly by the need to escape predators. When the rats
sense danger, they run along a runway to the nearest burrow. Packer (1983) observed that
the only defense mechanism A. niloticus appeared to exhibit against predators was

immediate flight into a burrow.

1.5 REPRODUCTIVE BIOLOGY

We may be able to predict some aspects of the reproductive biology of A.
niloticus by looking to animals that are closely related taxonomically, or to animals that
occupy similar ecological niches. In general, murid rodents are spontaneous ovulators,
and are polyestrous. Although they may breed continuously in the lab, murids are usually
seasonal breeders in the wild (Asdell, 1964). The estrus cycle length can range from 4-11
days, and gestation from 17-47 days. Most species of murids exhibit a post-partum
estrus, although this is not universal, and delayed implantation and prolongation of
gestation are common during lactation. If a female does not become pregnant as a result
of post-partum copulations, cycling is usually interrupted nonetheless by lactational
anestrous (Asdell, 1964; Hayssen et al, 1993).

Ecologically, A. niloticus appear to resemble the rrricrotine rodents, in particular
those of the genus Microtus. Microtine rodents are members of the rodent Suborder

Myomorpha, but are usually assigned to Subfamily Microtinae, separate from the murids.

29
Members of both Arvicanthis and Microtus are small-bodied and herbivorous, and are
important prey species in their habitats. Members of the genus Microtus have been
described as diurnal, nocturnal and crepuscular, exhibiting seasonal shifts in the timing of
activity in some species (Madison, 1985). I have already made comparisons between the
two groups with regard to body conformation and population cyclicity. Among the
various species of Microtus, there is much variability with respect to reproduction. Most
species are induced ovulators, but at least one species is a spontaneous ovulator
(Seabloom, 1985). Length and timing of the breeding season are variable, especially
among voles found in areas of low snow cover during the winter. Environmental factors
which may influence the seasonality of breeding in voles are photoperiod, temperature,
water availability, and nutrition. Gestation duration among Microtus is generally 21 days

(Keller, 1985), and there is usually a post-partum estrus (Seabloom, 1985).

1.5.a Seasonality of Reproduction

Short-lived animals must reproduce as frequently and continuously as possible to
offset their brief life span. Whether an animal reproduces seasonally or continuously
depends upon the environment in which it lives (Bronson, 1989). Environmental factors
known to influence the seasonality of reproduction in mammals include food availability,
social cues, photoperiod, temperature, humidity, and rainfall (Bronson, 1987). Many
mammals living below 30° of latitude appear to breed opportunistically in relation to food
quality and availability, and thus often in relation to rainfall patterns (Bronson, 1989).

Delany (1986) described A. niloticus as highly adaptable, breeding both

seasonally and aseasonally, depending upon the location of the population and upon local

30
environmental conditions. Neal’s (1981) interpretation was that A. niloticus are both
physiologically and ecologically well adapted for continuous breeding. Where A.
niloticus breed aseasonally or continuously, litter sizes are small, but litters are larger in
areas where there is a distinct breeding season (Delany, 1986). In describing the
seasonality of reproduction in this species, Neal (1981) suggested that we should ask, not
what initiates breeding, but what prevents it from occurring continuously?

Reproductive activity in tropical African rodents has been related to rainfall by
some previous workers (Delany, 1972; Taylor and Green, 1976; Delany and Happold,
1979). Proposed mechanisms underlying a reproductive response to rainfall include:
increased water intake (Neal, 1981), increased quantity and quality of food (Delany and
Happold, 1979), and stimulation from chemical signals in the vegetation (Negus and
Berger, 1977). A secondary plant compound found in green vegetation, 6-
methoxybenzoxazolinone (6-MBOA) stimulates reproductive activity in the meadow
vole, Microtus montanus (Berger et al, 1981); and since 6-MBOA has a stimulatory
effect on the ovaries of another tropical murid rodent (Mastomys coucha) (Linn, 1991),
the chemical may affect Arvicanthis as well. Kingdon (1974) cites a study in which
Arvicanthis were induced to breed in the laboratory by supplementing the diet with fresh
greens, and providing nesting boxes (Weinbren and Mason, 1957). However, among A.
niloticus in the wild, the relationship between breeding and rainfall is variable (see Table
2 and Figure 7). In eastern Afiica (Figure 7: A, B, C, D, E) breeding commences (or
peaks, in the case of Packer’s (1983) work) in concert with, or shortly after, a rise in
rainfall. In Nigeria (F), however, breeding commences well before the onset of the rains.

Further west, in Burkina Faso and Mali (G, H), the breeding season occurs completely

31

Table 2. Relationship between rainy season and reproductive season in nine African
Arvicanthis populations. Note: Breeding seasons were adjusted for Taylor and Green
(1976) and Neal (1981), in which the authors reported the presence of pregnant females.
The authors’ breeding season estimations were adjusted for the gestation of the pregnant
females. All other authors were assumed to have made this adjustment in their estimation
of breeding season, as they indicated that breeding season was derived mainly by noting
the presence of pregnant or lactating females.

 

 

 

Investigator Latitude Study site Rainy season Reproductive
season
Packer (1983) 02 20 S Serengeti Natl. Park, March-May (peak) March-May
Tanzania (peak)
Ghobrail and ? Sudan July-September August-February
Hodeib (1982)
Delany (1964b) Equatorial Queen Elizabeth Park, April-May Only pregnant
Uganda September-November female was
found in
September
Delany and 00 16 S Nakuru, Kenya (Rift February-November Most intense in
Roberts (1976) Valley) March-April (peak) middle of the
wet season
Rabin and Fisher 12 03 N Kano, Nigeria May-September March-
(1989) November
(1984)
April-October
(1985)
Taylor and Green 00 16 S Kitale, Kenya April-mid-December May-December
(1976) April and August
(96316)
Delany and 00 16 S Nakuru, Kenya (Rift March-September April-February
Monro (1986) Valley)
Neal (1981) 00 11 S Mweya Peninsula, March-May, February-July,
Uganda
00 06 S Crater Track, Uganda September-November August-
(Queen Elizabeth Park) September
00 11 N Rojewero Plains, Kenya March-May, October- February-
December December
Sicard et a1. 14 00 N Burkina Faso June-September November-April,
(1992) 13 00 N Mali May-October September-June

 

32

Figure 7. The relationship between rainfall, temperature, and breeding season in
Arvicanthis.

Breeding season is represented by the cross-hatched boxes, rainfall by the open circles
(O), and temperature by the closed circles (e).

A. Kitale, Kenya (Taylor and Green, 1976), B. Nakuru, Kenya (Delany and Monro,
1985), C. Meru, Kenya (Neal, 1981), D. Queen Elizabeth Park, Uganda (Neal, 1981), E.
Serengeti National Park, Tanzania (Packer, 1983), F. Kano, Nigeria (Rabiu and Fisher,
1989), G. Burkina F aso (Sicard et al, 1992), H. Mali (Sicard et al, 1992)

For graphs A-F, temperature and rainfall data were collected from A groclimatological
Data for Afiica (1984). If a listing was not found for an exact location, the nearest
location with a similar elevation was used (graphs C-E). For graphs G and H,
temperature and rainfall data supplied in Sicard et al. (1992) were used.

 

Figure 7.

Rainfall (mm)

33

 

 
     

V////I///l//l/////l/L .
100 — 19

 

 

 

75

25

 

1 00
50

 

400 -
200 —

Temperature (°C)

 

225 —
150 .1
75 J

 

VI‘I/Illl/Illlh

300 - v _ 3(2)
150 — . ‘ l: 28
0 71/01/15“. 36

75 . _ 32

25 “v‘ T 28
fill/IILA'WIIA refill]. 24

 

 

‘.

JFMAMJJASOND

Month

34
opposite of the rainy season. Thus, although a correlation can be seen in eastern Afiica,
rainfall is not a good predictor of breeding season in the genus Arvicanthis across the
continent.
Eastern Africa
Reproduction in A. niloticus in eastern Africa appears to occur most commonly
during the rainy season (Delany and Roberts, 1976; Packer, 1983; Delany and Monro,
1986). Taylor and Green (1976) proposed that the breeding season of A. niloticus is
controlled by diet, with abundance of necessary foods mediated by rainfall. They found
that A. niloticus in Kitale, Kenya began breeding 2—3 months after the rains began, at the
time when weed seeds were first found in stomach contents. Breeding continued
throughout the rainy season and into the early part of the dry season, when it began to
decline. As breeding declined, weed seeds and cereals were still present in stomach
contents, but were less abundant. Later in the dry season, when breeding ceased, the diet
of A. niloticus switched first to the leaves and stems of dicotyledonous plants, and then
almost exclusively to grass. Fat deposits that were laid down while the more fatty seeds
and cereals were abundant were rapidly utilized at this time. Fat stores were often
completely depleted by the time breeding resumed. Taylor and Green (1976) proposed a
relationship between reproduction and diet, suggesting that breeding occurred when
seeds made up a major portion of the diet, and ceased when the diet switched to green
vegetation.
Taylor and Green (1976) tested their hypothesis that the breeding season of A.
niloticus is dependent upon food quality by provisioning animals with wheat for an entire

year. The provisioned animals bred continuously, while unprovisioned controls

35
continued to show seasonality of breeding. It is clear, therefore, that the role of diet in
regulating breeding of Arvicanthis in Kenya should not be discounted entirely. This
finding is important with regard to the population outbreak of Arvicanthis in Kenya
described by Taylor (1968). In years in which the rainy season is prolonged, the duration
of cereal production may also be lengthened (Poulet and Poupon, 1978; Sicard et al,
1994). Since A. niloticus in east Afiica appear to be more likely to breed when cereals
are present in the diet, prolonged production of cereals may lead to an extended breeding
season. An extended breeding season would result in higher than normal population
numbers, or population outbreaks.

Neal (1981) argued that the results from Taylor and Green’s (1976) study did not
indicate a breeding season defined by food quality. First, they noted that breeding
declined before nutritious food disappeared from the diet. Taylor and Green (1976)
explained this as “allowing resources from the last of the season’s weed seeds and cereals
to be diverted from reproduction to fat” (p. 373). This does not, however, indicate a
reliable cue for the inhibition of reproduction. Neal (1981) claimed that cessation of
breeding may actually have been due to increased density of rodents as a result of
decreased habitat due to harvesting in the study area. Second, in the second year of their
study, reproductive activity began before seeds or cereals were detected in the diet. The
commencement in activity was correlated with increased rainfall. Third, in Neal’s (1981)
observations in Kenya, there was no correlation between breeding and diet, and
reproductive activity began a few days after the rains began, long before seeds and cereals

were available.

36

Neal (1981) stressed the tie between dietary conditions and rainfall in governing
timing of reproduction in Arvicanthis. He pointed out that breeding in Taylor and
Green’s (1976) study declined before seeds became unavailable, and that therefore seeds
in the diet were not a reliable inhibitive cue. However, female mammals need to time
their reproductive activity such that optimal resources are available at the point in their
cycle when they need them most, usually during late lactation (Bronson, 1989). A female
Arvicanthis would need to ‘know’ how lo_ng seeds would be available, not just whether
they were available. So, while the presence or absence of seeds in the diet might not be
the proximate cue causing induction and cessation of reproduction, it might often be the
ultimate factor responsible. This hypothesis is consistent with the results of Neal’s
(1981) study. Perhaps Neal’s (1981) animals started breeding right after the rains began
because they used rain as a proximate cue for the ultimate factor of improved dietary
conditions.

Neal (1981) failed to find a relationship between diet and reproduction in Uganda
and Kenya, where breeding seemed to be controlled by temperature. He concluded that
breeding was most likely inhibited by water stress due to high temperatures, and that this
would explain the increase in breeding seen when the rains began. This is a point well
taken, for even a small increase in ambient temperature above a mammals’ thermoneutral
zone can increase demand for water, and decrease food intake and locomotor activity
(Bronson, 1989), creating suboptimal conditions for breeding.

In a comparison of animals from Uganda and Kenya, Neal (1981) found that
female reproductive activity and pregnancy were significantly higher during the wet

seasons in both places. Delany (1964b) examined 38 female Arvicanthis collected from

37
July to September in Queen Elizabeth Park, Uganda, for signs of reproductive activity.
Of these 38 animals, only one was pregnant and one lactating. The pregnant female was
found in September, 1963 while the rest were collected in July and August. According to
Delany (1964b) there is a lull in rainfall from June through August. It is possible that
Delany found such a low rate of reproductive activity due to the fact that he was
collecting animals during a time of low rainfall. In an earlier collection period in 1961
from July through December in Queen Elizabeth Park (Ankole and Karamoja), the author
found only one pregnant female (in August) out of 15 collected (Delany, 1964a). This is
in sharp contrast to Delany and Neal’s (1969) 1965-1966 studies in Queen Elizabeth
National Park, in which Arvicanthis bred continuously, and in which breeding peaked in
October, a month of high rainfall (these data appear to be the same as those Neal reported
in his 1981 paper). It is possible that, if Arvicanthis are using rainfall as a proximal one
of eventual food availability, they might be able to change their breeding season with
changing rainfall patterns.
Western Afiica
Rabiu and Fisher (1989) found that Arvicanthis in Nigeria began to breed 1-2
months before the rains, and stopped breeding at the end of the rainy season. The authors
describe 3 possible reasons or cues for this early commencement of reproduction: a rise in
ambient temperature in March and April, a fall in fat reserves towards the end of the dry
season, or a change in photoperiod, but these possible cues were not examined
experimentally. It is quite possible that photoperiod was the cue that triggered the
commencement of breeding. The effect of photoperiod on the breeding season of

Arvicanthis in western Africa is described below.

38

While Bronson (1989) stated that it is unlikely that animals living below 30° of
latitude respond to changes in photoperiod, he maintained that the capacity to respond
reproductively to photoperiodic cueing remains as a neural trait, depending upon the
animals’ evolutionary history. This assertion was mainly in response to a group of
investigators who studied Arvicanthis living further away from the equator than the
studies mentioned above (Sicard et al, 198 8). This group of investigators considered the
possibility that photoperiod may be governing the seasonality of reproduction of
Arvicanthis in western Africa (Sicard et al, 1992, 1993, 1994). In years of normal
rainfall, the breeding season of A. niloticus in Burkina F aso and Mali in west Africa does
not overlap with the rainy season. Other investigators in Ethiopia and Sudan have also
found that Arvicanthis are capable of breeding during the dry season (Milller, 1977;
Ghobrail and Hodeib, 1982). Sicard et al. (1988) tested the effect of photoperiod on
testicular function in sexually active male A. niloticus from Burkina Faso. Even though
the range of daylength in Burkina Faso is small (11 h 09 min-12 h 52 min), manipulation
of photoperiod had an effect on testosterone production in test animals, as compared to
controls maintained in a natural photoperiod. Interestingly, the effect was bimodal, there
was a photostimulatory phase between 8-11 h of daylight, and a photoinhibitory phase
when daylight exceeded 11 h 30 min.

Sicard et al. (1992) repeated this experiment using animals from both Burkina
Faso and Mali. The range of daylength in Mali is similar to that in Burkina Faso (11 h 11
min-12 h 49 min). The investigators were able to replicate the findings of their 1988
paper, in that males from Burkina Faso displayed decreased testosterone production when

exposed to a daylength exceeding 11 h 30 min. The animals from Mali showed decreased

39

testosterone production in daylengths exceeding 12 h 15 min, although decreases of
testosterone in these animals did not reach the significance level of the data from the
Burkina Faso animals. This experimental evidence is supported by field observations of
the breeding seasons of animals in these two areas. Thus, having shown that a change in
daylength affects reproduction of Arvicanthis in some parts of Africa, Sicard et al. (1988,
1992) successfully responded to Neal’s (1981) appeal that the factors inhibiting
reproduction, rather than the factors facilitating reproduction in A. niloticus be
investigated.

While it may seem that a reproductive season governed by photoperiod would
leave little room for variation in length of breeding season, I have noted that it is a
lengthening of the breeding season of A. niloticus that is cited as facilitating the outbreak
of these rodents in Burkina Faso in 1986 (Sicard et al, 1994). In order to determine
whether photoperiodic regulation of the reproductive season could be modified by other
factors, Sicard et al. (1993) investigated the effects of temperature and water conditions
(relative humidity and availability of water rich foods) on testicular function in male A.
niloticus from Burkina Faso and Mali. Animals from Mali were tested in a short
photoperiod, while those from Burkina Faso were tested in both short and long
photoperiods. Testosterone levels increased in sexually inactive animals housed in long
daylengths but at low temperature and with abundant water. This increase was most
noticeable in animals housed in a short photoperiod.

To summarize, the onset and durations of breeding seasons of Arvicanthis near the
equator, as evidenced in this review by animals in eastern Africa, may be influenced by

food quality, water availability, or temperature. Arvicanthis living further from the

40
equator, like those in this review from Nigeria, Mali, and Burkina Faso, display breeding
seasons that appear to be mediated by photoperiod, and modified by food quality, water
availability, and temperature. The effects on breeding of social cues such as population

density have not yet been formally studied in Arvicanthis.

1.5.b Estrous Cycle

There is only one account of the estrous cycle of A. niloticus. Using the vaginal
smear technique, Compoint-Monmignaut (1968) described a 6 d estrous cycle for a single
female in the laboratory. This author did not report sample sizes or housing conditions,
but described all other females as being in permanent anestrus. While the author did not
offer an explanation for the incidence of permanent anestrus, it may have occurred due to
overcrowding, stress or poor husbandry. It is also possible that not all females cycle at all
times, or that there were confounding seasonal or ontogenetic variables. Ghobrail and
Hodeib (1982) took daily vaginal smears to elucidate the estrous cycle of this species.
However they merely stated in their results that the smears seemed to be comparable to
those of spontaneously ovulating rodents, and did not report a cycle length. As occurs in
most murids (Bronson, 1989), female A. niloticus exhibit a post-partum estrus (Ghobrail

and Hodeib, 1982; Delany and Monro, 1985a).

1.5.c Ovulation
Sicard et al. (1993) made reference to a thesis (Kyelem, 1993) that described A.
niloticus as spontaneous ovulators that do not exhibit a seasonal anestrus. The only

published study in which ovarian histology of A. niloticus has been examined to date was

41
that conducted by Ghobrail and Hodeib (1982), who collected animals over an 18 month
period in Sudan. They found that their wild-caught adult female A. niloticus were
continuous breeders, as indicated by the fact that their ovaries contained either large
follicles or corpora lutea of ovulation, and that there were no reproductively quiescent
animals. Corpora lutea were always present in the ovaries of pregnant and lactating
females. The ovaries of juveniles contained primary and secondary follicles, whereas
those of mature, non-pregnant adults contained follicles at all stages of development,
including fully mature follicles. Ghobrail and Hodeib (1982) reported that corpora lutea
were never seen in mature non-pregnant adults. This information, given in the text, is
contradicted by measurements of corpora lutea in recently mature and mature, non-
pregnant females presented in Ghobrail and Hodeib’s (1982) Table 3 (p. 326). The
authors did not report whether these females were nulliparous or multiparous. They
found that mature females kept solitarily, in pairs with males, or in heterosexual groups
showed corpora lutea of ovulation. The fact that the ovaries of solitarily housed females
contained corpora lutea suggest that A. niloticus is probably a spontaneous ovulator.
Since the exact conditions of the females in the above study were unknown, in the current
study I examined the ovaries of a variety of age groups in order to determine whether A.

niloticus are spontaneous or induced ovulators.

I.5.d Gestation
Previous reports of gestation length in A. niloticus have varied. Estimations for
gestation length among females not suckling a litter are as follows: 18 d (Delany and

Happold, 1979), 21-23 (I (Ghobrail and Hodieb, 1982), 22-24 (I (Delany and Monro,

42

Table 3. Arvicanthis litter size in captivity and in the field.

 

 

 

Investigator Study site Average litter size Range
Packer (1983) Tanzania above- 3.7 1-7
ground observations
Rabiu and Fisher Nigeria burrow 5.4 not given
(1989) excavations
Quilici et al. (1969) lab stock from not given 5-10
Senegal
Delany and Monro lab stock from 3.7 1-6
(1985a) Kenya
Ghobrail and Sudan lab stock: November-March: 4-12
Hodeib (1982) combined pup and 5.4
embryo counts April-August: 3.5
Taylor and Green Kenya capture data: 6.0 2-12
(1976) embryos
Coe (1972) Kenya capture data: 1968: 3.0 not given
embryos 1970: 6.5
Delany and Happold not given 4.0 3-10
(1979)
Rabiu and Fisher Nigeria capture 7.3 not given
(1989) data: embryos
K. E. Holekamp Kenya capture data: 1992-1993: 4 3-5
(personal pups born to
communication) females in captivity
that were pregnant

when captured

43
1985), or 24-26 d (Quilici et al. 1969). Ghobrail and Hodieb (1982) also reported a
gestation of 35-51 d for females suckling litters. Since these earlier estimations were

inconsistent, a study of gestation length in A. niloticus was undertaken in this thesis.

1.5.c Litter Size

Bronson (1989) stated that the reproductive strategies of murid rodents are driven
by the need to compensate for a very short life expectancy. One of the ways in which
murids make up for a short life span is to produce large litters at a high rate. Listed in
Table 3 are litter sizes for Arvicanthis from direct observation of litters in the wild, from
lab data, and from capture data. Please note that most of the capture data are reported in
counts of embryos, not of pups. Embryo counts of up to 10 (Delany and Happold, 1979)
and 12 (Taylor and Green, 1976) indicate that Arvicanthis has the capacity to produce
large litters. However, as the female has only six mammae, it is not likely that litters of
this size would be seen in the wild. This is also reflected by the fact that the average litter
sizes reported by Packer (1983) and Rabiu and Fisher (1989), were 3.7 and 5.4,
respectively. These averages are smaller than the reports of embryo counts mentioned
above. Packer’s (1983) range of 1-7 from observations of litters in Tanzania does not
approach the ranges of 5-10 (Quilici et al, 1969) and 2-12 (Taylor and Green, 1976) given

for the laboratory. We have had litters of 12 pups born in our M.S.U. lab colony as well.

I.5.f Lactation Interval
Delany and Monro (1985a) noted evidence of lactation in mothers for three weeks

after parturition. These authors proposed that since a second litter could be born at this

44

time, weaning probably took place during the third week post-partum.

1.5.g Growth and Development

Delany and Monro (1985a) charted growth of A. niloticus in the laboratory and in
the field. Neonatal mass in the lab (4.65 i 0.18 g) and in the field (4.25) were similar, but
laboratory animals exhibited considerably faster growth rates. Delany and Monro’s
(1985a) study of development in the lab revealed that A. niloticus were born with short,
dark fur, and that their coats took on the speckled coloration of adults within one week.
Incisor teeth in lab-reared pups erupted at 3 days. These pups opened their eyes at 7 days,
and had full motor ability by 14 days (Delany and Monro, 1985a). Quilici et al. (1969)
found that eyes were open and incisors apparent in lab-bred pups at five days, which is
closer to what was seen in our M.S.U. lab during this study. Delany and Monro ( 1985a)
found that weaning occurred in lab-born pups when they weighed 25 grams, which is
similar to Neal’s (1981) field estimate of 15-25 g.

Measurements of A. niloticus from our colony are listed in Table 4 (n=45 females,
15 males). In general, nulliparous females in our colony are smaller than males, but
pregnant or lactating females are often difficult to differentiate from males in terms of

body mass.

45

Table 4. Mean measurements (i SE, [range]) of 45 female and 15 male A. niloticus from

the M.S.U. colony.

 

 

Sex Weight Hind foot Ear Tail Head and Total Body
Body
Male 103.5 i 6.7 27.3 i 0.5 16.6 i 2.5 87.4 i 1.8 137.9 i 3.2 225.3 i 3.8
(55-103) (25-30) (1 1-20) (75-100) (111-156) (186-243)
Female 84.6 3: 3.1 27.3 i 0.5 15.5 i 0.3 87.3 i 1.1 132 i 2.7 219.2 i 2.9
(46-120) (21-35) (10-20) (68-104) (101-204) (187-280)

 

 

 

 

 

 

 

I.5.h Ages for Males and Females at Sexual Maturity

Age at sexual maturity probably depends upon environmental conditions. Neal
(1981) found that animals from Uganda and Kenya matured at different rates. Quilici et
al. (1969) found that A. niloticus in the lab were capable of breeding at 40-50 d, although
females matured slightly later than males. Delany and Monro (1985a) also noted scrotal
testes in lab-raised males at 40-50 days. First parturition in females can occur at 60 d
(Delany and Monro, 1985a). Given approximately 25 d of gestation, these data on age of
first parturition suggest that puberty in females occurs at approximately 45 (1.

Reproductive maturity in wild populations is usually related to body mass, since
exact ages are frequently unknown. Neal (1981) found that puberty occurred at very
different weights in animals in Uganda (males: 65-80 g; females: 55-65 g) and Kenya
(males: 33-40; females: 20-35), but this was probably due to a similar difference found in
the weights of adults in these two countries (Uganda males: 70-125, females: 60-110;
Kenya males: 60-90, females: 60-90). Taylor and Green (1976) reported that the weight
at which puberty occurred in free-living A. niloticus in Kenya was 45 g, based on the

lowest weight recorded for a pregnant female.

46

I.5.i Life span

Murid rodents usually live less than two years, and frequently live less than one
year (Walker, 1975). Females in Packer’s (1983) study in Tanzania lived an average of
10.2 months (maximum: 20 months), but the life span of males was more difficult to
quantify because of dispersal. Neal (1981) noted that few animals in Uganda and Kenya
lived as long as a year, and Delany and Monro (1986) found that their animals in Kenya
had a five percent chance of living to one year of age. Muller (1977) estimated that the
life expectancy for a one month old animal in his Ethiopia study was seven to eight
months. Thus in the field, the life span of Arvicanthis is less than one year, and it is
unknown at this point whether there is a sex difference in longevity.

Animals living in captivity, with the luxuries of ad Iibitum food and protection
from predators, often live longer than their counterparts in the wild. There is one record
of a captive Arvicanthis living for 6 years and 8 months (Jones, 1982, cited in Nowak,

1985).

CHAPTER II

THE NON-PREGNANT ESTROUS CYCLE AND OVULATION

11.1 INTRODUCTION

Most female mammals will accept copulation with males only at times when they
are physiologically ready to do so, and these short periods of sexual receptivity are
known as heat, or estrus. Periods of estrus are separated by longer periods of time in
which the female is not physiologically ready to mate, and in which she is unreceptive to
male advances. These periods of estrus and non-estrus may occur repeatedly and
sequentially, and they constitute the estrous cycle (Sadleir, 1973).

The estrous cycle has been studied in great detail in laboratory and domestic
species of rodents, such as Rattus norvegicus (Long and Evans, 1922), Mesocricetus
auratus (Kent and Smith, 1945), Cavia porcellus (Stockard and Papanicolaou, 1917), and
Mus musculus (Allen, 1922). While this work has been invaluable to the study of
reproduction, it does not reflect the diversity of the Order Rodentia (Hayssen et al, 1993).

The species mentioned above are highly inbred, specialized forms that are the result of
years of domestication and artificial selection, and the information gained from studying
these animals may have little bearing on what actually occurs in wild animals (Conaway,
1971). It has also been suggested that the domestication process for R. norvegicus has

selected for short ovarian cycles, as well as successful mating abilities under laboratory
47

48
conditions (McClintock and Adler, 1978).

Historically, reproductive cycles in female mammals have been classified into one
of two categories: those in which ovulation occurs spontaneously, and those in which
ovulation is induced by some form of stimulation (e. g. Eckstein, 1949). As our
knowledge of the field has grown, this dichotomy has proven not only inadequate, but
constraining. We have been eager to put a species or genus into one category or the
other, missing out on important subtleties essential to our knowledge of reproduction as a
whole. Conaway (1971) broadened the traditional dichotomy into a spectrum, noting that
spontaneous and induced ovulation are the extremes of a single continuum, and both
patterns of ovulation are capable of being modified by various factors. Intensive study of
domestic rodent species has led to an oversimplified conceptualization of mammalian
reproduction, and often to the belief that spontaneous ovulation is the rule, with induced
ovulation being the exception. In fact, as more information is obtained regarding the
reproductive cycles of various mammalian species, it seems more likely that induced
ovulation is the basic pattern, and spontaneous ovulation a specialization (Conaway,
1971).

In light of the common contemporary belief that spontaneous ovulation is the rule
among mammals, it is surprising that induced ovulation was thought to be the
predominant mammalian strategy until the beginning of the 19th century. It wasn't until
corpora lutea were found in the ovaries of virgin women that the idea of spontaneous
ovulation became popular (Ramirez and Soufi, 1994). The fact that human females are
spontaneous ovulators has perhaps led to an anthropocentric view of ovulation. However

we are learning that, as is often found in mammalian reproduction, many different means

49
exist to achieve the same end.
Conaway (1971) expanded the traditional spontaneous/induced categorization of
non-pregnant female reproductive cycles into three categories, with category assignments
made on the basis of spontaneous vs. induced forms of ovulation, and luteal function

(Table 5).

Table 5. Categorization of non-pregnant mammalian female reproductive cycles (after
Conaway, 1971).

 

 

 

Type Ovulation Luteal Function
Type I
Subtype A3 medium length spontaneous spontaneous
cycles.

Subtype B: long cycles

Type II
Subtype A3 medium 16118111 Induced spontaneous
cycles.
Subtype B: long cycles.
Type III spontaneous induced

 

In Type I, both ovulation and pseudopregnancy occur spontaneously. This
condition is typical of the hystricomorph rodents, ungulates, and primates. Cavia
porcellus, the guinea pig, fiequently used in laboratory studies, displays an estrous cycle
of 15-18 days (Subtype 1A). This is significantly longer than the cycle seen in the rat and
the hamster, because the corpora lutea in C. porcellus are functional in the absence of
copulation (F eder, 1981). The follicular phase is short, and the luteal phase lasts 12-13

days (van Tienhoven, 1983). Canids are classified in Subtype 1B, with non-pregnant

50
cycles that are over five weeks in length (Conaway, 1971).

Animals categorized as Type II are induced ovulators, with spontaneous corpus
luteum formation. Ovulation is induced by exogenous stimuli, most notably copulation.
This strategy is employed by several species of rodents in the genus Microtus. Microtus
are placed in Subtype 11A, with medium length cycles. In M. ochragaster, both
behavioral estrus and ovulation are induced, and if stimulation is continuous, behavioral
estrus can last for at least 30 days (Richmond and Conaway, 1969). Felids, which are in
Subtype 11B, come into behavioral estrus spontaneously, and the estrus period is more
fixed. Non-pregnant cycles in this subtype last 4-8 weeks (Conaway, 1971).

In animals that are classified as Type 111, ovulation is spontaneous, but luteal
function is contingent upon copulatory stimulation. Rattus and Mesocricetus fall into the
Type 111 category. The cycle is short, 4 days for hamsters and 4-5 days for rats
(Conaway, 1971).

Milligan (1975) suggested the addition of a fourth type, in which both ovulation
and luteal function are induced. In the field vole (Microtus agrestis), ovulation may be
induced by stimulation from a male, yet the stimulation may not be sufficient to induce
luteal function. That is, ovulation and luteal function appear to have different stimulation
thresholds for activation.

As I stated in Chapter 1, little is known about the reproductive cycle of A.
niloticus. Ovulation in wild-caught females was thought to be spontaneous by Ghobrail
and Hodeib (1982), and a 6 d estrous cycle was described for a single female housed in
captivity by Compoint-Monmignaut (1968). The purpose of the work presented in this

chapter was to describe the non-pregnant estrous cycle and ovulatory pattern of A.

51
niloticus. 1 used a variety of methods in my attempts to elucidate the estrous cycle,
including cytological assays of vaginal smears, behavioral assays, and long-term
monitoring of activity rhythms. Ovarian histology was performed at the conclusion of

most of these studies, in order to determine the pattern of ovulation in this species.

11.2 CYTOLOGICAL ASSAYS OF VAGINAL SMEARS

When Stockard and Papanicolaou reported the existence of a regular estrous cycle
in the guinea pig in 1917, they established the validity of the vaginal cytological assay.
The authors showed that changes in cell content of the vaginal lumen corresponded to
changes over the ovarian cycle, thus providing a relatively noninvasive marker of ovarian
activity. This assay has since been used to describe the estrous cycles of many
mammalian species (Rowlands and Weir, 1984).

By comparing the relative density of the 3 types of vaginal cells (cornified
epithelial cells, nucleated epithelial cells and leukocytes), one can assign each smear to
one of the 4 phases of the estrous cycle. These are estrus, metestrus, diestrus and
proestrus, and cytological changes in the vaginal lumen with the phase of the female
cycle are depicted in Figure 8. Estrus is indicated by a smear that consists solely of
cornified epithelial cells (Figure 8, C). A diestrus smear contains many leukocytes, and a
few cornified epithelial cells (Figure 8, D). Metestrus is indicated by a vaginal smear that
contains predominantly leukocytes (Figure 8, A). A proestrus smear contains mostly

nucleated epithelial cells, occurring singly or in sheets (Figure 8, B).

52

TYPES OF CELLS FEED
INTO VAGINAL LUMEN

PORT/0N or EP/THEUUM
VAGINAL 3 r ROMA I
mu.
E ‘ . ,‘fig, ‘ ,

 

   
     
 
    
  
 
 

":‘_" ’ 1.. Z ‘
all ‘..,fa§9§§t
. - 4 .

  

 

.8» ‘ “i
23', L- .>
1.3;" '-

Ho

3:3;1. )>'}’. _ ’1 'x 2"»
p" -.-' 1‘ y .3 . ‘92,!”
M255 If . - -

.-»’-é;1‘li""$ - '
ESQ):- ggrlg’gi

0

 

Figure 8. Cyclic changes in vaginal cell types.(after Turner andBagnara, 1971).

53
II.2.a General Methods
Subject Animals and Housing Conditions

The original stock of 29 A. niloticus from which the Michigan State University
breeding colony is derived was captured in July and August 1993, from two locations in
the Masai Mara National Reserve in southwest Kenya. A. niloticus in the M.S.U.
breeding colony were housed in plastic cages (3 8x14zl 6cm) with chipped aspen litter,
and cages were changed once a week. Animal rooms were kept at 23°C on a 12: 12
lightzdark cycle. Animals were provided ad libitum tap water and Harlan 8640 Teklad
22/5 rodent chow, supplemented with carrots and whole oats once a week. Litters born to
mating couples were weaned at 21 days, and the opposite-sex siblings were separated at
50 days, after which only littermates of the same sex were housed together.

Cytological Assay of Vaginal Smears.

Females were housed individually. In each experiment, vaginal smears were obtained
from each study animal at 24 h intervals at the same time every day, for a series of
consecutive days. Vaginal lavage was performed by releasing a few drops of 0.9% saline
solution into the rat's vagina with a glass eyedropper. The saline was then drawn back
into the eyedropper, and the sample was released and spread onto a microscope slide.
The smears were air-dried overnight. The slides were then stained in 0.25% aqueous
methylene blue, rinsed in distilled water, and stood on end to dry. The slides were
viewed under a light microscope at 10x and 40x magnification, and the cell compositions
of the slides were recorded. At the end of each experiment, the females were sacrificed,

and their blood and reproductive tracts were stored for further analysis. Blood samples

54
were collected via cardiac puncture, plasma was separated by centrifugation, and kept
frozen at -80°C. Ovaries and uteri were collected and stored individually in sealed bottles

of 10% buffered formalin.

II.2.b Initial Evaluation Of Cell Types
Purpose
The purpose of this pilot study was to determine whether vaginal smears of A.
niloticus contain the distinct cell types described in other rodents.
Methods
Vaginal smears were taken from two nulliparous females for 11 days and one
multiparous female for 9 days.
Results
The three types of cells expected: leukocytes, cornified epithelial cells, and
nucleated epithelial cells were readily visible in the vaginal smears of A. niloticus (Figure
9). Cell types changed in relative abundances in the smears of the virgin females, but
estrous smears were never seen. The multiparous female gave a typical estrous smear

every day.

II.2.c Females Housed 1n Small Cages
Purpose
The objective of this study was to systematically examine the vaginal smears of a
large number of females in an effort to elucidate the estrous cycle of A. niloticus, and to

compare multiparous females to nulliparous females with respect to estrous cyclicity.

 

Figure 9. Vaginal smear from a nulliparous female A. niloticus.

Methods

Daily vaginal smears were taken from 5 multiparous females and 10 nulliparous
females, between 1000 h and 1100 h for 29 days. Two of the multiparous females were
of unknown age, and the other three were between 200-216 doa. Each of the multiparous
females had delivered at least 2 litters. The nulliparous females were all at least 104 days
old (range: 104-116 days).

Ovaries were fixed in paraffin (Appendix A), sliced on a rotary microtome, fixed
onto microscope slides, and stained with hematoxylin and eosin (Appendix B). Ovarian
sections were viewed under alight microscope at 10x and 40x magnification, and were
scored for the presence of corpora lutea. Each section was scanned for luteal tissue, and a

yes/no classification was recorded.

56
Results
All five multiparous and three of the nulliparous females gave estrous smears
every day (Figure 10). Leukocytes would occasionally appear in an animal’s smear, but
in very low abundances. The remaining nulliparous females gave smears containing all
three types of cells; although these cell types changed in relative abundances, they did not
show the expected cyclic pattern, and estrous smears were never seen (Figure 11).
Ovarian histology revealed that the ovaries of all five of the multiparous females
contained corpora lutea, suggesting that all five females had ovulated. Luteal tissue was
absent from the ovaries of all nulliparous females. A Student’s t-test revealed that uterine
weights, corrected for body weight, did not differ between multiparous and nulliparous

females (t= -1.15, at: 13, p= 0.272).

II.2.d Females Housed In Large Enclosures
Purpose
I thought that perhaps the results from the first study did not follow the expected
pattern of smear types due to the stress of captivity and of daily handling. Therefore the
objective of this study was to modify the conditions of the previous experiment, in order
to alleviate stresses on the animals.
Methods
I attempted to ease the stresses of captivity and handling by housing females
individually 3.5 ft x 3.5 ft enclosures, and by anesthetizing the animals with Metofane
prior to taking the samples. I took daily vaginal smears from four mature nulliparous

females for 15 consecutive days. Having learned in a previous experiment (described in

 

57

 

Figure 10. Persistent estrous vaginal smears from a nulliparous female A.niloticus.

58

 

day 5 day 6

 

day 7

 

Figure 11. Erratic vaginal smear series from a nulliparous female A. niloticus.

59
Chapter IV) that COpulations involving nulliparous female A. niloticus occurred around
the time of the beginning of the light phase, 1 took the samples at 0700 h, just after the
lights came on in the animal room.
Results
One female gave an estrous smear on the first day of the experiment, then

proceeded to give smears that indicated a five-day cycle, and gave another estrous smear
on the sixth day of the experiment (Figure 12). After this initial cycle, the smears from
this female showed no evidence of cyclicity for the remainder of the experiment. The
smears of the remaining three females did not give any indication of cyclicity. The
ovaries from the one female that showed a vaginal estrous cycle contained no corpora
lutea. Two of the remaining females similarly showed no evidence of ovulation, but the

last female had one regressing corpus luteum in one of her ovaries.

II.2.c Conclusions

A single vaginal estrous cycle of five days was observed in one female. However,
as the female that appeared to show a cycle did not show evidence of ovulation
histologically, I conclude that this apparent vaginal cycle was not a reflection of an
ovarian estrous cycle.

The pattern of continuous estrous smears shown by all multiparous females and
three nulliparous females is puzzling. It is possible that this smear pattern may be the
result an insufficient diet. Greenwald (1956) mentioned that a prolonged estrous smear in
an isolated female rodent could conceivably be due to a vitamin A deficiency. In R.

norvegicus, continuous estrous smears are indicative of aging, and of the cessation of

day 1: estrus

e9.

       

Figure 12. A five day estrous cycle from anulliparous female A. niloticus.

61
estrous cyclicity. In one experiment, the ovaries of rats in constant estrus did not contain
corpora lutea (Huang and Meites, 1975), indicating that ovulation had not occurred. The
presence of corpora lutea in the ovaries of multiparous females indicates that constant
estrous smears did not simply reflect aging in A. niloticus in this experiment.

Breed (1967) reported persistent estrous smears in the reflexively ovulating short-
tailed vole (Microtus agrestis). yet corpora lutea were not visible in the ovaries of these
females. Peterson (1986) found in the gray-tailed vole (M. canicaudatus) that the
presence of persistent vaginal cornification was age-related, and that two-thirds of
females 150-200 days of age showed persistent vaginal cornification. This is fairly early
in the reproductive lives of these voles, since they will reproduce in the laboratory for 12-
16 months (Peterson, 1986). Therefore, persistent vaginal cornification in the reflexively
ovulating Microtus species is probably not analogous to the persistent estrous state
associated with aging in the spontaneously ovulating R. norvegicus. Since corpora lutea
were found in the ovaries of female A. niloticus that had exhibited persistent vaginal
estrus, it is not likely that this smear pattern is analogous to constant estrus in the
microtines.

Continuous estrus periods have also been observed in smears obtained from
Rhabdomys pumilio (Dewsbury et al, 1984), a species closely related to A. niloticus.
These periods were very similar to those seen in A. niloticus, with the occasional
appearance of leukocytes on one day within the period of constant estrus. Dewsbury et
al. (1984) found no obvious explanation for constant vaginal estrus, but mentioned that it
may be advantageous for a female to be in a prolonged state of receptivity. Dewsbury et

al. (1984) did not, however, comment specifically on the receptivity of the females

62
showing constant estrous smears. The authors merely stated that copulation was more
likely to occur during proestrus, but that it may occur during estrus as well.

It is possible that continuous estrous smears in A. niloticus are due to high levels
of circulating estrogens. Perhaps high hormone levels are ‘masking’ the vaginal
representation of the estrous cycle. It is also possible that estrogen levels are normal for a
murid rodent, but that this species has a lower threshold to estrogen stimulation for
eliciting an estrous smear. Another possibility is that the stress of captivity and daily
handling renders the vaginal cytological assay ineffectual in A. niloticus. Whatever the
ultimate reason, it is clear that the cytological assay of vaginal smears was not an
effective indicator of estrous cyclicity in our captive M.S.U. colony of A. niloticus.

The presence of corpora lutea in the ovaries of females isolated from all
conspecifics for 30 days, suggests that A. niloticus can ovulate spontaneously. There is a
remote possibility that this species is an induced ovulator, and that the females in this
study were induced to ovulate by the stimulation of the glass eyedropper in the vagina
during the smearing procedure. This seems unlikely, however, since luteal tissue is
absent from the ovaries of virgin females. This finding may indicate that isolated virgin
females do not ovulate. There are many possible reasons for this. Multiparous females
may ovulate spontaneously, while nulliparous females may be induced to ovulate. That
is, the female's first ovulation may be induced by the presence of a male, as occurs in the
dwarf hamster (Erb et a1, 1993). Multiparous females may actually be induced ovulators
, but may have a lower threshold for ovulation to be induced, and may have been
stimulated by the sampling process. It is also a possibility that the stress of handling

suppresses ovulation in nulliparous females, but not multiparous females.

63

11.3 BEHAVIORAL ASSAYS

II.3.a Male-Female Interactions

The manner in which male and female mammals behave toward one another
changes over the course of the female’s estrous cycle. Sexual behavior of female
mammals can be divided into three components: attractivity, proceptivity, receptivity
(Beach, 1976). Attractivity is the female’s stimulus value to male conspecifics,
proceptivity is the extent to which the female initiates or solicits copulation, and
receptivity reflects the stimulus value of a female for educing an intravaginal ejaculation
from a male. These three behavioral parameters vary with circulating levels of steroid
hormones produced by the ovary during the estrous cycle, and in many mammals
estrogen enhances female sexual behaviors and progesterone inhibits them.

Purpose

The objective of this study was to elucidate the estrous cycle by observing the
copulatory behavior of male-female pairs over a series of days. I hoped to observe
cyclical changes in male-female interactions, reflecting cyclic changes in the female’s
reproductive physiology. My specific goal was to calculate lordosis quotients (LQs) in
female A. niloticus, and to monitor changes in LQs overtime. The LQ is the ratio of the
number of lordosis postures shown in response to a fixed number of mounts (usually 10)
x 100 (Beach, 1976).

Methods
1 paired two couples, consisting of mature, sexually naive animals, in separate 15

gallon aquaria for 10 minutes a day over 10 consecutive days. Testing began between

64

0900 h-1000 h. Except during testing, the females were housed alone, and the males

were housed together. Animals were placed in the aquarium on either side of a masonite

divider for an acclimatization period of five minutes. At the end of this time, I raised the

door in the middle of the divider to allow the animals access to one another. On days 3-

10 of testing, the testing room was illuminated by only a 60 watt bulb to decrease the

impact of the presence of the investigator. On days 7-10, in an attempt to make to testing

situation more like the female’s home cage, the female’s bedding was put into the

aquarium on the side in which she was acclirnatized. Behaviors were recorded on a data

sheet (Figure 13). Recorded behaviors were:

female approach: female approached the male (subsequent behaviors imply
male approach)

male roll: male rolled onto his back underneath the female

male sniff: male sniffed the ano-genital region of the female

mount: male attempted to mount the female

I: intromission

E: ejaculation

groom: both autogrooming and allogrooming were noted

These behaviors were recorded numerically in order of occurrence in 30 second time slots

on the check sheet.

65

 

 

Behavioral Check Sheet: Paired Encounters

Date:

Time:

Observer:

Animal id: male: female:

Time Female Male Male Mount 1 E Comments
Roll Sniff

:00
:30
1:00
1:30
2:00
2:30
3:00
3:30
4:00
4:30
5:00
5:30
6:00
6:30
7:00
7:30
8:00
8:30
9:00
9:30

 

 

Figure 13. Sample check sheet for scoring male-female interactions.

 

66

Female response was recorded following the numerical notation of each event
using the following abbreviations:

o R: run away

0 K: kick

o B: bite

o KT: kick threat, female raised leg but did not strike

0 BT: bite threat, female bared teeth but did not strike

0 P: posture, a specific response to male sniffing category in which the female

raised her hindquarters
Results

Males showed initial interest by sniffing the females’ ano-genital regions, but
females usually responded by kicking the male or running away. Aggression in the form
of biting was not seen. Interest shown by the male attenuated with time during the course
of each trial. Mounting was never seen, and therefore female LQs could not be
calculated. Lowering the lights decreased the animal’s apparent awareness of the
investigator. Adding the female’s litter to the testing arena did not appear to afiect the
behavior of the female, however the male spent most of his time sifting through the litter,
apparently foraging for oats.

Conclusions
Two possible interpretations of the results of this experiment are: l) the females

were not cycling, or 2) the animals were not exposed to each other at a time of day

67
conducive to mating. I did not examine the ovaries of these females, so 1 don’t know
whether they were cycling. If they were indeed cycling, it is possible that I did not see
any mounting because 1 was not pairing the animals at the right time of day. In a
subsequent experiment (described in Chapter IV), I examined the timing of copulation of
eight mating couples, and found that copulations involving nulliparous females began
from 0319 h-0600 h. In the present experiment, testing began after 0900 h, therefore the

females may not have been physiologically ready to mate.

II.3.b Male Response To Female Urine
Female mammals advertise their readiness to mate via olfactory signals (Johnston,
1983), and the most likely source of such signals is urine, which contains the metabolic
products of many hormones, as well as vaginal secretions. Hormone secretion fiom the
gonads varies over the course of the reproductive cycle, and chemical contents of urine
may vary with gonadal output (Johnston, 1983). If a male is receiving information about
a female’s reproductive state from her urine, his interest in the urine should fluctuate over
the course of the female’s estrous cycle.
Purpose
The objective of this study was to elucidate the estrous cycles of female A.
niloticus by observing changes in male behavior in response to the daily presentation of
female urine samples.
Methods
I collected daily samples of urine from eight mature female A. niloticus. Seven of

these females were nulliparous (162-167 doa), and one had delivered one litter (174 doa).

68

Females were housed in metabolism cages, which had wire mesh bottoms that allowed
all waste to fall through. The waste was collected in a test tube via a firnnel that was
suspended beneath the cage. Feces were kept separate from the urine sample by a plastic
mesh screen placed in the bottom of the funnel, such that only liquid waste could pass
through to the test tube. Females were fed powdered food from a cup that prevented the
mixing of food with the urine sample. Once a week, the funnels and screens were washed
and dried. At the termination of the experiment, the females were sacrificed, and their
ovaries were removed and stored in 10% buffered formalin for future histological
examination.

Urine samples were collected daily at 1700 h, and presented to eight males, along
with a distilled water control, during a 20 min test every day for 21 (1. Testing was done
in a 15 gallon aquaritun that was reduced to 1/3 floor space by a masonite divider, in
order to retain the male in the area of the samples. 1 pipetted 200 pl of urine or water
onto one inch squares of absorbent paper that were taped to the bottom of the test arena,
about five inches apart. The placement of the samples was rotated, with water on the
right on even numbered dates, and on the left on odd numbered dates. Testing was done
in two runs of four males each, and all tests were videotaped for later analysis. There was
no investigator in the testing room during testing. Videotapes were scored individually
for each male, and the male’s level of interest, as reflected in the amount of time the male

spent sniffing and licking the urine sample, was recorded.

69
Results

The urine samples differed in quality between females. One female gave about 2
ml of light yellow colored urine every day, three females gave over 9 ml of light yellow
colored urine, and three females gave 1-2 mls of urine which ranged in color from light
brown to dark brown. The quality of urine did not appear to affect the male’s level of
interest.

Fluctuations of levels of interest were evident, but there did not appear to be any
cyclic pattern (Figure 14). I thought that perhaps the females were not cycling, and upon
histological examination of the ovaries, 1 found that four of them contained no corpora
lutea, but that the other four (including the multiparous female) did display evidence of
ovulation. Post-hoe analysis then revealed that the four males that had received urine
from females who had ovulated showed greater fluctuations of interest than did the other
males (Figure 14, A,C,G,H). Males exposed to urine from ovulating females showed
peaks of interest with a range of 5-7 days between peaks.

Conclusions

These results were not significant, and although there is a trend that suggests that
the cycles of the females involved in this study may have been somewhere between 5-7
days in length, there is no way to correlate the peaks in interest with an estrous cycle of

with ovulation.

II.3.c Activity Rhythms
Changes in activity rhythms, such as an increase in activity, or phase advance of

activity, have been associated with the estrous cycle in Rattus norvegicus (Wang, 1923),

70

 

 

 

 

 

 

 

A C
80.00 80.00
70.00 70.00
60.00 60.00
50.00 50.00
E 40.00 E 40.00
30.00 30.00
20.00 20.00
10.00 \A" M 1000 f M
0.00 ~ - 0.00 , - - - -
B D
80.00%; 80.00
70.00 -- 70.00
60.00 ~- 60.00
50.00 1. 50.00
3.....- g...»
30.00 «- 30.00
20.00 ‘r 20.00
10.00 -» M 10.00
0.00 “A” - 0.00
E G
80.00 v 80.00
70.00 .. 70.00
60.00 ~» 60.00
50.00 ~~ 50.00
40.00 -~ E 40.00
30.00 ~~ 30.00
20.00 ~~ 20.00
10.00 -~ 5 10.00 A
0.00! 0.00 - - -
F H
80.00 80.00 7
70.00 70.00 --
60.00 60.00 ~-
50.00 50.00 «-
3.... 3.....-
30.00 30.00 ~l
20.00 20.00 ~~
10.00 10.00 c
0.00 0.005
eaaaasgeesss aaaeaeggeesa
Day Day

Figure 14. Male response to female urine sample represented by the number of seconds
spent sniffing the sample.

71
Mesocricetus auratus (Zucker, 1980), and Octodon degus (Labyak and Lee, 1995).
Activity peaks on the day of estrus in R. norvegicus (Wang, 1923). In M. auratus,
activity is phase advanced, or begins earlier, by about 15 minutes on the day of estrus. 0.
Degus display both of the aforementioned patterns: on the day of estrus, activity onset is
advanced two h, and shows an overall increase (Labyak and Lee, 1995).
Purpose

The purpose of this experiment was to elucidate the estrous cycle of A. niloticus

by following the activity rhythms of several females over time.
Methods

Seven mature nulliparous female A. niloticus were implanted intraperitoneally
with Mini-Mitter transmitters (Mini-Mitter, 1nc., Sunriver, OR) to measure gross motor
activity. Counts of activity were recorded as the transmitter moved over a wire grid in a
receiver plate underneath the cage, and this information was relayed to a computer in an
adjacent room. Data were collected in five minute intervals, or bins, using DataQuest 111
software (Mini-Mitter, 1nc.).
Transmitter Implantation Surgery Protocol. Surgical instruments, autoclips, and cotton
swabs were autoclaved before use. During the surgery, instruments were kept immersed
in Nolvasan. Instruments were sterilized with 260°C dry heat between sequential
surgeries with an Inotech dry sterilizer (model Steri 350). Animals were anesthetized
with methoxyflurane (Metofane) in a glass chamber at the beginning of the procedure,
and kept unconscious during the surgery using a nose cone filled with Metofane-soaked

cotton. Surgeries were performed under a hood, and a heating pad was used to keep the

72

animal warm during the procedure. The abdominal area was shaved and wiped with
Nolvasan, and an incision was made along the midline of the abdomen. Transmitters
were kept immersed in Nolvasan for at least 1/2 h before surgery, and rinsed with sterile
saline before insertion into the intraperitoneal cavity. Plain gut absorbable suture (5-0)
with attached needle (PS-2) was used, and the muscle wall and the skin of the abdomen
were sutured separately. Autoclips were used over the external sutures for reinforcement,
and to prevent the animals from removing the sutures. The closed incision was covered
with Nolvasan topical antibiotic cream (1% chlorhexidine acetate). The animal was
placed in a normal tub cage with bedding and a ‘house’. A heating pad was placed under
half of the cage until the animal was completely recovered from the anesthetic.
Following surgery, animals were treated with 1 cc of Lactated Ringer’s solution and .01
cc buprenorphine hydrochloride (Buprenex) as an analgesic. For three days following
surgery, animals received 12 mg of sulfamethoxazole and trimethoprim (SMZ-TMP), an
oral antibiotic, in their drinking water. On the day of surgery, animals received softened
rodent chow placed on the floor of the cage. For one week following surgery, rodent
chow and cats were placed on the floor of the cage to prevent the animal from standing
up to reach for food from the overhead bin, and putting strain on the incision. Autoclips
were removed one week after surgery, and Nolvasan topical antibiotic cream was applied
to the incision, which was checked for signs of infection. Incisions that were not
completely healed, or that appeared off-color were flushed with Betadine surgical scrub
(ammonium nonoxynol-4-sulfate), and Nolvasan topical antibiotic cream was applied.

Data from the same 31 day period were used for each of six animals. The daily

records were scanned for evidence of activity onset. The mean daily level of activity

73
(measured as the number of grid wires crossed per time bin) was calculated for each
individual, and these values were used to calculate overall mean values for the test period.
Daily values were then plotted as standard deviations from the mean for each animal.
The ovaries of the subject animals were prepared histologically as described above, and
scanned for the presence of corpora lutea.
Results and Conclusions

There was no evidence of advancement of activity onset. There was no pattern of
cyclic increases in activity, as illustrated by activity values plotted as standard deviations
in Figure 15. Corpora lutea were present in the ovaries of females T1, T5, and T6,
indicating that these animals had ovulated, but these ovulation events were not evident in
the activity records. The results from this experiment showed that activity rhythms were

not good indicators of the estrous cycle in female A. niloticus.

11.4 OVULATION
11.4.a Purpose
The results from the experiments described earlier in this chapter prompted me to
look more closely at ovarian histology, in order to determine whether nulliparous females
were indeed cycling. Therefore I conducted a study in which the objective was to
determine whether the incidence of ovulation differs between nulliparous and multiparous

female A. niloticus.

74

 

 

O
A

M
A

‘
A

 

 

e
d
A A

It:

Mmbatolmnoudawtetlonurommnacttvtty
O

 

 

 

 

 

 

 

 

a
I
l
l
I
I
l
I
I
I
i
I
I
i
l
I
I
l
I
l
l
l
i
i
l
l
l
l
l
I
I
I

H
A

‘
A

 
  

{all

7 I 1213“

 

 

A

-
‘
A

 

lb

Woietendarddavtemtrommenactlvtty
O

b

p.-...—_-——_—---_——_—————_———_—___———

 

 

 

 

 

 

 

 

U
.

N
A

‘
A

 

 

I
.-
A

Nunberoletanderddevletlomtrommoenacttvtty
O

 

 

 

 

 

 

 

 

 

Figure 15. Number of standard deviations from mean activity measured over a period of
31 days.

75

11.4.b Methods

1 processed a total of 19 ovaries from multiparous females, and 59 ovaries from
nulliparous females using the methods previously described. The multiparous females
ranged in age from 155-580 (1, and the nulliparous females ranged in age from 47-253 (1.
All but three of the multiparous females had been separated from their male partners for
at least 50 days, and the three mating couples that remained together had stopped
producing young. A chi-square test was performed to determine whether nulliparous

females differed from multiparous females with regard to ovulation.

11.4.c Results

All of the ovaries from multiparous females contained corpora lutea, indicating
that ovulation had occurred (Table 6, Figure 16B). Of the 59 nulliparous ovaries
processed, only 10 exhibited evidence of ovulation, the remaining ovaries contained
growing follicles, but no corpora lutea (Table 6, Figure 16A). Ovaries of multiparous
females were significantly more likely to contain corpora lutea than were those of
nulliparous females (X2=42.442, df=1, p<0.001). Older (>100 doa) nulliparous females
did not differ significantly from younger (<100 doa) females (two-tailed Fisher’s exact

test, p=0.476).

76

 

Figure 16. Nulliparous and multiparous ovaries of female A. niloticus. A nulliparous
ovary (A), and a multiparous ovary (B), with 2 corpora lutea (CL) and one regressive
corpus luteum (RCL). Note the size difference. These photos were shot at the same
magnification (10X).

77

Table 6. Number of multiparous and nulliparous female A. niloticus exhibiting signs of
ovulation.

 

Corpora lutea Corpora lutea 11

present absent
Multiparous l9 0 l9
Nulliparous 10 49 59
n 29 49 78

 

11.4.d Conclusions

The multiparous females in this study were spontaneous ovulators. Since they
were not giving daily vaginal smears, there is no possibility that they were being induced
to ovulate by the actions of the investigator. These results also indicate that most virgins
that are isolated from males do not ovulate, as is also the case in the dwarf hamster,
Phodopus campbelli (Erb et al, 1993). P. campbelli exhibits social control of puberty in
that most females do not begin ovulatory cycling until they are exposed to a male.
Female A. niloticus that did show evidence of ovulation in this experiment may have been

induced to ovulate by olfactory cues emanating from males housed in the same room.

11.5 DISCUSSION
The estrous cycle of A. niloticus could not be elucidated by means of the vaginal
cytological assay, behavioral assays, or monitoring of activity rhythms. Spontaneous
ovulation occurred in all multiparous females that had been isolated from all conspecifics
for 30 days. There is some indication that nulliparous females in this species may not
begin ovulatory cycles when isolated from male conspecifics. This sort of social

mediation of puberty is seen in the dwarf hamster (Erb et al, 1993). Delay of sexual

78
maturation in the female has also been seen in other rodents (reviewed by Levin and
Johnston, 1986): Mus musculus, Microtus californicus, M. ochragaster, Peromyscus
maniculatus, Notomys alexis, and Meriones unguiculatus. Levin and Johnston (1986)
found a correlation between the lability of puberty onset and the degree of gregariousness
of the species. Individuals of a gregarious species might be more likely to be influenced
by the presence of conspecifics than would individuals of a species that live solitarily
(Levin and Johnston, 1986).

Conaway (1971) questioned the importance of the non-pregnant estrous cycle in
small rodents in natural populations. He referred to a cycle in which pregnancy does not
occur “a pathological luxury which cannot be tolerated” (Conaway, 1971, p. 239). A
non-pregnant estrous cycle is physiologically expensive, therefore it would be
energetically efficient to delay cycling until such time as there is a high chance of
becoming pregnant. While the results of the preceding experiments may be due to the

stress of captivity, the hypothesis of delayed puberty merits further testing

CHAPTER III

THE MORPHOLOGY OF COPULATORY BEHAVIOR

III.1 INTRODUCTION

The pattern of behaviors that mammals display during copulation, or the
morphology of copulation, varies widely among species, but is quite stereotyped among
individuals within a species. Information on the morphology of copulation in a variety of
species give us insight into the processes underlying the evolution of behavior and
reproductive systems. However in this branch of the study of reproduction, as in the
study of estrous cycles and ovulation, domesticated laboratory rodents are among the
best-studied species; and domesticated animals are the least likely candidates to provide
information on the relationship between behavior and ecological variables
(Dewsbury,1972).

Dewsbury (1972) established criteria for the description of copulatory behavior in
mammals, and outlined a process of study which involves three steps. The first step,
classification, requires assigning a species to a category on the basis of the presence or
absence of four behaviors: copulatory lock, intravaginal thrusting, multiple intromissions
per ejaculation, and multiple ejaculations per copulatory series. A dichotomous tree of
these four criteria results in 16 possible pattern types (Figure 17). The second step,

elaboration, involves the quantitative description of the frequency and patterning of

79

80

Multiple Multiple ' Pattern .

 

 

 

 

 

 

 

 

 

 

 

Lock? Thrusting? . intromission? ejaculation? number ‘
Ye5\ .
Yes Y 3
NO \ es
Yes
YCS\T—_Yes .5
No "0 ‘5
No #:Yes 7
N0 '—j—.-- 8
' e-‘Yes--.—- 9
. YCS\
< ...._.....
Yes
NO \Yes_——ll
No ---—12
No ,

 

 

’ Yes\
No -—-—14
No< ~ _ 15

Figure 17. Patterns of mammalian copulation (from Dewsbury, 1972).

81
copulatory events, as well as the description of the conditions required for copulation,
courtship behavior, and postures assumed during copulation. The third step involves
experimental analyses of variables influencing copulatory behavior, such as the effects of
genetics, experience, age and hormone levels. In this study of the morphology of
copulation in A. niloticus, I will describe the classification of the species, and provide a

quantitative description of the frequency and patterning of mounting and ejaculation.

111.2 METHODS

A total of fifteen mating couples of mature, sexually naive A niloticus were paired
for use in this study, but only eight were actually used in data collection. Of the seven
mating couples that were eliminated from the study, four never produced a litter, and
three were excluded because of the death of one of the animals. Mating couples were
housed in 15 gallon aquaria and fihned around the clock, using low-intensity red light for
nighttime filming. In order to facilitate filming, animals were not given a house box, or
refuge. Aquaria floors were covered with bedding, and food and water were available ad
libitum. Filming was time-lapsed, recorded at 0.2 second intervals, such that 24 hours of
fihn was recorded on two hours of tape. F our mating couples were filmed
simultaneously, and the 15 mating couples were rotated through the video room over a 10
month period. Mating couples were removed from the video room at the cage change
following the birth of their second litter. This filming schedule allowed the recording of
the frrst (nulliparous) copulation, two births, and the post-partum copulations associated

with these births.

82

When a litter was found, the videotape for the corresponding day was scanned,
and the time of birth, as indicated by the appearance of the first pup, was recorded.
Nulliparous copulations were found by subtracting 26 days (the longest estimation for
gestation [Quilici et al, 1969]) from the birth date, and scanning videotapes in a reverse
chronological order from that date. Post-partum copulations were found by scanning the
videotape following the birth. In some cases, mating couples copulated a second time, as
early as 15 (1 following post-partum copulations. These secondary copulations, in
addition to the nulliparous and post-partum copulations, were scored using behavior
check sheets (Figure 18), on which the frequency and patterning of copulatory events
were recorded.

A total of 28 copulations were recorded: 8 nulliparous, 15 post-partum, and 5
secondary or multiparous. There was one case in which copulation did not occur post-
partum, and one case in which a multiparous copulatory bout was not scored due to
camera malfunction. 1 will describe only the nulliparous, multiparous and post-partum
copulation bouts that resulted in fertile pregnancy, and the infertile post-partum
copulatory events will be excluded.

The morphology of copulation in A. niloticus was described in accordance with
Dewsbury’s (1972) criteria. The number of mounts per ejaculation, and the duration of
the postecjaculatory interval (PEI), defined as the interval between ejaculation and the
resumption of copulation, were recorded for each ejaculation. This study also allowed
me to establish the timing of several reproductive events. I calculated a definitive
gestation for this species, and studied the timing of copulation and parturition (described

in Chapter IV).

RAT SEX DATA SHEET

 

GRASS RAT COUPLE: DATE: TAPE STARTING TIME

MALE FEMALE OBSERVER

time male female male groom male female comments
chases approaches mounts (m/f) intromits lordoses
female male female

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 18. Sample check sheet for scoring sexual behavior.

 

84
111.3 RESULTS

Copulation in this experiment consisted of the male chasing the female, who
would lordose in response to a mount by the male. Copulation was easily recognized
when scanning tapes due to its noticeable pattern, a circular chase interrupted by the short
stops for mounting and ejaculation. The circular pattern of locomotion was probably due
to the small size of the enclosures. There was no struggle or delay as males and females
separated after intromission or ejaculation, so 1 concluded that A. niloticus do not display
copulatory lock. From the time-lapse video, it did not appear that males were displaying
intravaginal thrusting, and this was supported by direct observation of copulatory
behavior in animals outside of this study. Male A. niloticus exhibited a series of shallow
pelvic thrusts, a pattern which has been shown in the rat to be used to locate the opening
of the vagina (Bennant, 1965). A. niloticus exhibited multiple intromissions per
ejaculation, and multiple ejaculations per copulatory bout. This pattern of no lock, no
thrust, multiple intromissions and multiple ejaculations can thus be assigned to pattern
number 13 on Dewsbury’s (1972) dichotomous tree (Figure 17).

It was impossible to determine whether an intromission was actually achieved
during a mount, so only mounts and ejaculations were scored. In all but two cases, the
number of mounts per ejaculation first decreased and then increased across a series of
ejaculations within a copulatory bout(Figure 19). Length of the PEI increased with
successive ejaculations within a copulatory bout (Figure 20).

Gestation ranged from 23.3-25.46 days (average 24.25 d). It appeared initially
that some post-partum gestations were longer than nulliparous gestations, with a range of

30-52 d. However, by scanning the tapes 23-25 (1 before the date of the birth, I found

85

 

70- 1'

60-

50—

 

40—

30—

Number of Mounts

20—

10-

 

 

lllililillll
1234567891011n-1n

Ejaculation

Figure 19. Number of mounts per ejaculation.

 

86

 

(A)
01

(A)
O
l

M
(II
I

 

N
O
I

 

.3
01
l

 

Postejaculatory interval (PEI)

(It
i

 

 

0 l l i l l I l I T

l l l
12345678910n-1n
Ejaculation

Figure 20. Postejaculatory interval per ejaculation.

 

87
that copulation was occurring a second time after post-partum copulation (secondary or
multiparous copulations). Gestations resulting from these copulations were of durations
comparable to the nulliparous gestations, and are included in the gestation range and

average listed above.

111.4 DISCUSSION

The description of the copulatory morphology of A. niloticus on the basis of
Dewsbury’s four basic criteria allies this species with rodents in the genera Rattus,
Mesocricetus, Peromyscus (the deer mouse), Sigmodon (the cotton rat), Meriones (jirds),
and Oryzomys (the rice rat) (Dewsbury, 1975), which are murid and cricetid rodents. The
decreasing, then increasing, pattern of mounting frequency over the course of a
copulatory bout has been found in Meriones unguiculatus, Oryzomys palustris, and
Rattus norvegicus. The pattern of increasing PEI length over the copulatory bout is seen
in most rodents (Dewsbury, 1975).

This experiment resulted in the first accurate description of gestation for A.
niloticus. The discovery of the secondary or multiparous copulations is significant.
Ghobrail and Hodieb (1982) reported gestations of 35-51d for females suckling litters.
The authors offered no hypothesis to explain the lengthened gestations (they reported
gestations of 21-23 (1 for females not suckling litters). Without any further information,
one might hypothesize that these lengthened gestations were the result of delayed
implantation. However, in this experiment, I found that in apparent cases of prolonged
gestation resulting from post-partum copulation, the animals actually copulated a second

time. These secondary or multiparous copulations occurred from 15-28 d following the

88
post-partum copulations (in one instance in which the pups were eaten by the parents,
copulation occurred five days after the post-partum copulation). If copulation and
ovulation can occur 15 d after the birth of a litter of pups, lactation may be waning at this
point. In Rattus norvegicus, females do not enter estrus until the young are weaned, at
about 25-30 days (Nelson, 1995). Thus, perhaps weaning is taking place sooner in A.
niloticus than the three week approximation reported by Delany and Monro from

observations in the lab (1985a).

CHAPTER IV
TEMPORAL PATTERNS OF ACTIVITY, TEMPERATURE, AND

REPRODUCTION

IV.1 INTRODUCTION

Biological rhythms synchronize the physiological, biochemical, and behavioral
processes of an individual with environmental conditions that vary in a predictable
rhythmic fashion. The physiological system that measures time and synchronizes these
processes with daily events is known as the circadian timing system. The word circadian
is used to describe the approximately 24-hour cycles that an organism endogenously
generates (Moore-Ede et al, 1982)). Most research on the mechanisms controlling
circadian rhythms in mammals has focused on a restricted number of animal models.
Specifically, nocturnal rodents such as Rattus norvegicus and Mesocricetus auratus have
been investigated extensively. Data on circadian rhythms and their neural substrates in
diurnal mammals are relatively sparse, and we know nothing about how the substrates
controlling circadian rhythms differ in nocturnal and diurnal species. One reason for this
has been the lack of suitable diurnal rodent models with which to conduct experimental

research.

89

90

As noted in Chapter 1, reports from the literature regarding activity patterns in A.
niloticus are conflicting, with the species described as diurnal (Delany and
Kansiimeruhanga, 1970; Kingdon, 1974; Packer, 1983; Rabiu and Fisher, 1989), diurnal
with crepuscular tendencies (Senzota, 1990), partially nocturnal (Ansell, 1960; Vesey-
Fitzgerald, 1966; Harrison, 1972) and completely nocturnal (Schmutterer, 1969; Ghobrail
and Hodeib, 1982). Data on the activity rhythms of A. niloticus in the field have been
collected via direct observations (Harrison, 1972), trapping (Senzota, 1990; Rosevear,
1969) and shooting (Rosevear, 1969) records.

Evaluation of general activity rhythms of A. niloticus taken into captivity provides
little clarification. Delany and Kansiimeruhanga (1970) reported the rhythm of only one
animal. Duplantier and Granjon (1990) studied the rhythms of 11 individuals, but
presented mean data for these animals, and did not describe rhythms from individuals.
These two studies, the first of which describes only one subject, and the second of which
describes the population rhythm, yield only an incomplete picture of activity in A.
niloticus.

A. niloticus in our laboratory have displayed a range of diurnal and crepuscular
wheel-running behavior (Katona and Smale, submitted); but the significance of wheel-
running with respect to the behavior of animals in their natural habitat is unknown
(Mather, 1981). Information about the general activity of a species aids in the study of its
ecology. The adoption of a temporal niche of activity is as ecologically important to a
species as its spatial niche (Moore-Ede et al, 1982), and this is particularly evident with

respect to reproduction.

91

Reproductive events can be related to endogenous rhythms of activity. For
example, receptivity in female R. norvegicus occurs predictably during the course of the
24 hour period (Keuhn and Beach, 1963; Hardy, 1972), and parturition in many
mammalian species tends to occur at a specific time during the light-dark cycle (Jolly,
1973; Rowland et al, 1984). The timing of both receptivity and parturition are especially
important in small-bodied, short-lived mammals. A female who is receptive during an
active period is more likely to come across potential mates, and a female who gives birth
during an inactive time is less likely to be disturbed by predators or conspecifics that may
cannibalize her pups (Jolly, 1973).

The objective of this study was to determine whether the biological rhythms of
wheel-running, gross motor activity, and core body temperature in A. niloticus, together
with the timing of copulation and parturition, form a coordinated pattern, and if so,

whether the pattern is nocturnal, diurnal, or crepuscular.

1V.2 ACTIVITY AND TEMPERATURE RHYTHMS

1V.2.a Purpose
The purpose of this experiment was to characterize the body temperature rhythm
in this species, and to examine the relationship between rhythms in body temperature and

rhythms in general activity and wheel running.

92
1V.2.b Methods

1 used a combination of radio-telemetry transmitters and running-wheels to
monitor activity and core body temperature. Seven mature virgin female A. niloticus
were implanted intraperitoneally with Mini-Mitter telemetry transmitters to measure core
body temperature and gross motor, or general, activity (for surgery protocol, see Chapter
11). Three of the animals were placed in cages containing running wheels (26 cm
diameter, 8 cm width), and four were left in standard cages without wheels. Data from
the transmitters and the running wheels were recorded by a computer in an adjacent room
using the DataQuest 111 data collection system (Mini-Mitter, Inc., Sunriver, OR) that
stored information in five-minute bins. Each wheel revolution was recorded when a
magnetic bar attached to the running-wheel strut closed a switch mounted on the wheel
housing. A receiver plate was placed beneath each cage to relay the individual
temperature and activity signals from each female's transmitter to the computer (Figure
21).

Here, I present data for each animal from a 10 day period beginning 3 weeks after
surgery (with the exception of animal T3, for which we began data collection 9 weeks
after surgery, due to technical difficulties). Histograms representing the average activity
for each hour in a 24 h period bins were used to graphically analyze data for the
description of daily rhythms in gross motor activity and core body temperature. Morning
and evening peaks were evident in wheel-running, temperature, and gross motor activity
patterns of each animal, and several features of these peaks were measured from 24 h
histograms divided into 20-minute bins (see Methods: Activity and Temperature

Rhythms). To evaluate the relationship between gross motor activity and core body

93

 

   

      
   

Wheel
Revolutions

     
 

General

Temperature Activity

 

 

 

Receiver Plate Magnetic Switch

 

 

 

 

 

 

 

 

 

 

 

Matrix Box

T
'3,

 

 

 

 

 

 

 

 

 

 

 

 

DataQuest 111

Figure 21. Data collection scheme for study of biological rhythms in A. niloticus.

94
temperature, the morning rises in activity and temperature were compared using
histograms divided into 10-minute bins. In order to determine whether A. niloticus are
diurnal with respect to general activity and body temperature, I compared the hourly
values of these variables in the dark phase with the trough from the light phase, and I

calculated the percent of one h bins which fell below the level of the light-phase trough.

1V.2.c Results

Daily patterns of wheel-running, gross motor activity, and body temperature of A.
niloticus were extremely predictable and precise (Figure 22). The three animals with
running-wheels were decidedly crepuscular in both general activity and wheel-running,
with peaks immediately before the beginning, and after the end, of the light phase
(Figures 22, 23). Individuals housed in cages without running wheels also displayed
crepuscular peaks in general activity, but activity levels remained higher relative to the
peaks during the light phase, than did activity in animals housed without running wheels.

Individuals housed in cages without wheels displayed absolute peaks in general
activity from 0611 h-0729 h and at 1975 h, and individuals with running wheels
displayed absolute general activity peaks from 0552 h-0650 h and 1926 h-1945 h. While
there was variability among individuals with respect to level of activity, general activity
peaked at higher levels in animals outfitted with running wheels than in animals housed
in standard cages (Figure 23). Animals in standard cages displayed higher light phase
trough levels than did animals with running-wheels, and dark phase trough levels were
similar between the two groups (Figure 24).

Peaks in core body temperature were associated with peaks in general activity. As

95

Figure 22. Biological rhythms in female A. niloticus. Actograms of general activity, core
body temperature and wheel-running activity of animals T1 and T6, divided into 10
minute bins, throughout ten 24 h periods.

96

 

T1 | | T6

 

general activity

'.889 12.88 16.88 213.86
True (Hours)

 

temperature

M
.' 2 88 16.88 28.80
inc (Hours)

1"

g .
4. 8 8.8?0 12.90 16.66 25.88
1&9

(Hours)

 

Figure 22.

97

Figure 23. A comparison of female A. niloticus with running-wheels ( 0), vs. females
without running-wheels ( O) with respect to core body temperature and gross motor
activity. Patterns of core body temperature (A) and gross motor activity (B), are divided
into 1 h bins and averaged over 10 days.

98

 

 

 

 

 

 

 

 

 

4
_ O
. ..e .2
O
1111. 2
l1.
:0
C
- .. -m
9. .
ao
.
1 .6. 1471.
I I8
0 4
e .
A .0
O
_ _ A _ _ _ d _ _ 0
0 9 8 7 6 0 0 0. 0 O
4 3 3 3 3 0 0 0 0
4 3 2 1

Aoev EEEmQEB.
zoom emu—2

38286526...
.562 $20 :85.

Figure 23.

99

Figure 24. A comparison of patterns of gross motor activity and core body temperature in
female A. niloticus with running wheels (B) vs. females without running wheels (A).
Patterns of gross motor activity ( O) and core body temperature (0), divided into 1 h bins
and averaged over 10 days.

100

Gov mSfiEaEm... boom cams.

 

 

 

 

 

 

  

 

 

 

0 9 8 7 6 9 8 7 6
4 3 3 3 3 3 3 3 3
_ _ _ _ _ r _ _
O- O!
.A. o
.-m e
‘.1\1 litIl-II‘ .
e.
.lv 0
ti o
. I
e
e
e
01
l.’
e
e
o
A B ..
_fi______J _ — _ 1 4
0000000000 0 0 0 0 0
9876.54.37.41 0 0 0 0
4 3 2 1

EEC-3853204. .222 $20 :82

8 12 16 20 24
Time (h)

4

Figure 24.

101

with general activity, peaks in temperature reached higher levels in individuals with
running wheels (morning peak range: 38.45°C-39.69°C; evening peak range: 39.29°C-
39.57°C) than in animals without running wheels (morning range: 38.09°C-38.9°C;
evening range: 38.09°C—39.09°C) (Figure 23). Morning rises in activity occurred almost
simultaneously with rises in core body temperature (Figure 25). This is especially clear
in animals outfitted with running wheels (Figure 25: D, E, F).

In all seven animals, the majority of values for the 12 one-hour bins of activity in
the dark phase were lower than the lowest value in the light phase (range: 58.33-75%)
(Table 7). This trend was also observed for values of body temperature, although there

was one female in which 50% of the values for the dark phase were below the light phase

trough.

Table 7. Percentage of one-hour bins in the dark phase with mean values below those of

 

the light phase trough.
Animal id Activity Temperature
T1 66.67% 66.67%
T2 58.33% 50.0%
T3 58.33% 66.67%
T5 66.67% 75.0%
T6 75.0% 75.0%
T7 75.0% 75.0%
T8 66.67% 66.67%

 

102

Figure 25. A comparison of the morning rise in gross motor activity and in temperature.
The pattern of gross motor activity ( O) and core body temperature ( C) from 0300b-
0600h, are divided into 10 minute bins and averaged over 10 days, in females without
running-wheels (A-D) and females with running wheels (E-G).

104

IV.3 TIMING OF COPULATION AND PARTURITION

IV.3.a Purpose

The purpose of this experiment was to characterize A. niloticus with respect to
temporal organization of copulation and parturition.
IV.3.b Methods

Data from the study on the morphology of copulation (Chapter III) were used in
this study. Mature, sexually naive males and females were paired in 15 gallon aquaria
and filmed around the clock. Data were collected from 8 mating couples that produced
litters during a ten month study period. Couples were filmed from the time of pairing
through the birth of their second litter. Thus, initial sexual encounters, plus two births
and two post-partum copulatory series were fihned for each couple. Pups were removed
at 21 days. When a new litter was found, the entire videotape for that day was scanned,
and the time of parturition (the sighting of the first pup) was noted. The time of
initiation of copulation was recorded for three types of copulatory series: the first series
that occurred for a given mating couple (nulliparous, n=8), post-partum copulation
(n=15), and secondary copulation, which occurred if the post-partum copulation did not
result in a successful pregnancy (multiparous, n=9).

Using data from the experiment on activity and temperature rhythms, peak periods
of general activity were calculated for animals housed without wheels, and the timing of
copulation and parturition of animals in this experiment were compared with these peaks,
to determine whether these events were more likely to occur during times of activity or

rest. The rise of each peak was defined as beginning at the lowest 20-minute bin before a

 

105
steady increase to the peak, and the end of the peak was defined as the lowest 20-minute
bin following the peak. The starting and ending points of the morning and evening peaks
of general activity were averaged to obtain times of peak activity. Using these periods of
peak activity, the 24 hour day was split into active (total 5.09 h) and inactive (total 18.91
h) periods. The binomial test (Snedecor and Cochran, 1967) was used to test whether the

distribution of timing reproductive events over these two periods was random.

IV.3.c Results

The eight copulations involving nulliparous females all began between 0319 h-
0600 h (Figure 26). Six of the eight copulations were initiated during the morning period
of peak activity. Postpartum copulations were initiated between 0511 h and 1937 h. A
cluster of 12 copulations occurred between 0511 h and 0640 h (Figure 26). Of the
remaining three instances of mating, one occurred at 0956 h, one at 1140 h, and the last at
1937 h. There was only one instance in which a mating couple did not copulate post-
partum. The parturition to post-partum copulation interval ranged from 0216 h-2024 h
(n=15). Interestingly, the majority of post-partum copulation (n=9) events were not
fertile. Fertile post-partum events (n=6) occurred with a parturition to post-partum
copulation interval range of 0216 h-1124 h.

When the first post-partum mating event was infertile, second bout of mating
occurred between 15 to 28 days after the first post-partum copulation. In one case, there
was no nighttime video due to camera malfunction, so information on initiation time was
not available for that bout of copulation. However these animals were found mating

when room lights came on. In the remaining five cases, copulation occurred between

 

106

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Post-Partum — 7;? 7/
Nulliparous — O 0 Z %//

Time (h)

Figure 26. Timing of copulation in captive A. niloticus. The onset of 28 copulations in A.
niloticus ( O), and the beginning of heat in R. norvegicus ( C), after Hardy, (1972)
plotted over the peak periods of general activity (indicated by the hatched bars).

107
0411 h and 0633 h (Figure 26).

Of the total of 28 copulations observed, 23 began during times of peak activity as
defined with the general activity data. The probability of this distribution occurring by
chance alone (P=0.50) is p=.0024.

The 16 recorded parturitions occurred in two clusters, one during the light phase
from 1025 h-1516 h (five cases), and the other during the dark phase from 1945 h-0410 h
(11 cases) (Figure 27). Sixty-nine percent of the births occurred during the dark period,
and 87.5% occurred outside the periods of peak activity. The probability of this

distribution occurring by chance alone (P=0.50) is p=0.016.

IV.4 DISCUSSION

A. niloticus has previously been shown to be diurnal and crepuscular with respect
to wheel-running activity (Katona and Smale, in press). Further exploration of biological
rhythms in this species has revealed that A. niloticus is a diurnal mammal, and displays a
suite of adaptations to this general pattern, involving general activity, sexual activity, and
temperature. In this experiment, wheel-running was crepuscular. Previous wheel running
experiments with this species showed some individuals to be diurnal, and others
crepuscular (Katona and Smale, in press). Patterns of general activity and temperature
were both diurnal, although peaks occurred just before the start of the light phase, and
just after the beginning of the dark phase. Levels of general activity and temperature
remained higher during the light phase than in the dark phase, giving both patterns a
strong diurnal component. The general pattern of activity seen in A. niloticus of two

peaks separated by a trough is the most common among wild animals, and is different

 

108

 

 

 

AA AA A AA A AAA

 

.\\\\\\‘i§

 

 

u\\\\\\\’

 

 

 

 

 

 

 

_ . .

l
0 4 8 12 16 20 24
Time (h)

I

Figure 27. Timing of parturition in captive A. niloticus. Time of beginning of parturition
(A) plotted over the peak activity periods represented by the hatched bars.

109

from the one-peak pattern seen in most laboratory animals (Aschoff, 1966). Thus the
peaks seen in A. niloticus do not warrant the species’ classification as crepuscular, but
rather as a diurnal animal with crepuscular tendencies.

It is widely believed that, although high levels of activity can cause slight
elevations in body temperature, the circadian rhythm of body temperature is not simply a
secondary effect of the circadian rhythm of activity. One reason that investigators have
come to this conclusion is that in htunans and many animals, core body temperature starts
rising before the individual awakens and becomes active (Refinetti and Menaker, 1992).
One of the most interesting findings of this study is that in A. niloticus, temperature rises
in concert with gross motor activity in the morning.

In R. norvegicus, heat, signified by the display of lordosis and low incidence of
rejection of the male, begins around dusk, and females are maximally receptive for six to
nine hours (Keuhn and Beach, 1963; Hardy, 1972) thus female R. norvegicus are
receptive during their periods of peak activity. Similarly, of the 28 instances of mating
recorded in A. niloticus, all but six (78.6%) began during the narrow window of the
morning peak of activity, 0440h-0729h . Of the six mating bouts that did not occur
during this time, four were closely associated with one of the two daily peaks of activity.
Three of these copulation bouts began within 1.5 h before the morning rise in activity at
0440h, and one began during the evening activity peak. Thus, in both R. norvegicus and
A. niloticus, mating occurs at the beginning of the active phase, but these species are
opposite one another in phase. In R. norvegicus, copulation occurs at dusk, whereas it

occurs just before dawn in A. niloticus.

 

110

Diurnal New and Old World monkeys give birth at night, while nocturnal
prosirnians deliver during the day (Jolly, 1973). The nocturnal species R. norvegicus and
M. auratus tend to deliver their young during the light period (Mitchell and Yochim,
1970; Lincoln and Porter, 1976; Connor and Davis, 1980; Viswanathan and Davis, 1992),
which represents their inactive phase. In R. norvegicus, it is the biological rhythm of the
mother that determines the time of birth (Lincoln and Porter, 1976). Female A. niloticus
delivered 68.75% of their litters during the dark phase, and 87.5% of all litters were
delivered outside of periods of peak activity. Thus, the timing of parturition in A.
niloticus appears to exhibit a pattern that might best be described as anti-crepuscular.

In R. norvegicus, timing of post-partum copulation depends upon the timing of
parturition, and is indirectly dependent on the time of day in that parturition usually
occurs during the light phase (Connor and Davis, 1980). In R. norvegicus, females
display lordosis in response to manual stimulation 4-36 h after parturition (Blandau and
Soderwall, 1941), and display peak receptivity to an active male an average of 12 h after
parturition (Connor and Davis, 1980). A. niloticus are different from R. norvegicus with
respect to the timing of post-partum copulation in two ways. First, A. niloticus are
capable of copulating much sooner after parturition than are R. norvegicus. The shortest
parturition to copulation interval that resulted in pregnancy recorded here for A. niloticus
was 0216 h, and all fertile post-parturn copulations began earlier than the average peak
receptivity seen in R. norvegicus at 12 h post-partum. A. niloticus may also differ from
R. norvegicus with respect to the timing mechanism controlling post-partum copulation.
Twelve of the 15 instances of post-partum copulation in A. niloticus began during the

morning peak. However, in contrast to the findings of Connor and Davis (1980), this

 

111
clumping of post-partum copulation times in A. niloticus was not a simply the result of
clumping of parturition times. Female A. niloticus appeared to be 'waiting' until their
period of peak activity in order to copulate. Recall, however, that most of these A.
niloticus copulation events (9/15) did not result in fertile pregnancy. Although we
initially ascribed this low post-partum success rate to a problem with pacing (the female
had no place to take refuge from the male), it may be that the experimental conditions
caused some anomaly in timing. Perhaps the reason that the copulations did not result in
pregnancy was that the females waited until the next morning to copulate.

In conclusion, the rhythms of Arvicanthis niloticus are predictable, clear, and

 

most certainly diurnal. This species is an excellent candidate for a diurnal model with
which to study circadian rhythms. The bimodal activity pattern displayed by A.

niloticus are more typical of wild animals than of other lab rodents.

CONCLUSIONS AND RECOMMENDATIONS

Conventional means for elucidating the estrous cycle in rodents have failed with
Arvicanthis niloticus. Although two lines of evidence suggested the existence of a 5-7
day cycle length, the search for an estrous cycle in this species revealed no conclusive
evidence. It is possible that the non-pregnant estrous cycle may be unimportant to female
A. niloticus in the wild. In a rodent that exhibits a post-partum estrus, the non-pregnant
estrous cycle may be obsolete, with conception occurring at each post-partum estrus
during the breeding season. Captive A. niloticus females are easily brought into
behavioral estrus with the priming protocol of estradiol and progesterone injections
(Carmen Salsbury, personal communication) used for R. norvegicus. Therefore, future
studies that require an estrous female could use primed subjects, and the need for
predicting the time of natural estrus is reduced.

Regardless of whether the non-pregnant estrous cycle occurs in the wild, the
question of ovulation in captive A. niloticus requires further study. One area of research
that deserves deeper investigation is that of the nature of the corpus luteum. Knowing the
lifespan of the corpus luteum in A. niloticus, and its cycle of degeneration, would allow
us to determine when a female ovulated, and to speculate how the female ovulated:
spontaneously, or in response to some form of stimulation. Knowing more about the

nature of the corpus luteum would also be helpful in other studies regarding the

112

 

113
reproductive biology of female A. niloticus. I suggested in Chapter II that puberty might
be delayed in females that are not exposed to males. A simple experiment would
determine whether this is true. Comparison of the ovaries of females housed alone at
weaning with those of females housed with male conspecifics would address this
question.

A. niloticus display a diurnal activity pattern which is quite different from the
nocturnal pattern of their well-studied cousins in the genus Rattus. The other six murid
rodents within Kingdon’s (1974) Arvicanthis division (see Chapter I), Dasymys,
Mylomys, Hybomys, Rhabdomys, Lemniscomys, and Pelomys, all show at least some
component of diurnality in their activity rhythms (Kingdon, 1974). Perhaps diurnality
evolved within this group as it developed separately from that of other murid taxa.
Comparative study using A. niloticus and R. norvegicus as diurnal and nocturnal models
would greatly benefit the study of circadian rhythms, and possibly the study of the
evolution of rhythms.

The utilization of Arvicanthis niloticus as a laboratory model opens new doors in
the study of behavior. A. niloticus offer three unique characteristics to laboratory study.
The first is that they have been recently brought into captivity. When studying
domesticated animals like Rattus norvegicus and Mesocricetus auratus, one has to
wonder whether the results are applicable to natural rodent populations. The second
exceptional quality of A. niloticus is that they are diurnal, and a reliable diurnal rodent
model is lacking. The third attribute that makes A. niloticus stand out is the
gregariousness of the species, providing opportunity for the study of social behavior in

the lab.

APPENDICES

 

50% EtOH
70% EtOH
95% EtOH
95% EtOH
95% EtOH
100% EtOH
100% EtOH
100% EtOH
Xylene
Xylene
Paraffin

Paraffin

30 min

30 min

30 min

30 min

30 min

30 min

30 min

30 min

30 min

30 min

30 min

30 min

114

APPENDIX A

Embedding Protocol for Paraffin Sections

 

115

APPENDIX B

Staining Protocol for Paraffin Sections

Hemo De 15 min
Hemo De 3 min
100% EtOH 2 min
100% EtOH 2 min
95% EtOH 2 min
70% EtOH 2 min
deO 2 min
hematoxylin 5 min
deO dip
eosin l min
70% EtOH dip
95% EtOH dip
100% EtOH dip
100% EtOH dip
Hemo De 5 min
Hemo De 5 min

All alcohol was diluted with double distilled water.

LIST OF REFERENCES

 

116

LIST OF REFERENCES

Afiica Today, an Atlas of Reproducible Pages. 1983. World Eagle, Inc., Wellesley, Mass.

Agroclimatological Data for Africa, Volumes I and 2. 1984. Food and Agriculture
Organization of the United Nations. Rome.

Allen, B. 1922. The oestrous cycle in the mouse. Am. J. Anat. 3: 297-371.

 

Allen, G. M. 1939. A Checklist of Afi-ican Mammals. Bull. Mus. Comp. 200]. (Harvard).
83. E

Allen, G. M., and A. Loveridge. 1934. Reports on the scientific results of an expedition to
the southwestern highlands of Tanganyika Territory. Bull. Mus. Comp. Zool.
(Harvard). 75: 47-140.

Ansell, W. F. H. 1960. Mammals of Northern Rhodesia. The Government Printer Lusaka,
Rhodesia.

Ansell, W. F. H. 1978. The Mammals of Zambia. National Parks and Wildlife Service,
Zambia.

Asdell, S. A. 1964. Patterns of Mammalian Reproduction. Cornell University Press, New
York.

Aschoff, J. 1970. Circadian activity pattern with two peaks. Ecologi. 47: 657-662.
Baskevich, M. I., and L. A. Lavrenchenko. 1995. On the morphology of spermatozoa in
some African murines (Rodentia, Muridae): the taxonomic and phylogenetic

aspects. J Zool Syst Evol Res. 33: 9-16.

Beach, F. A. 1976. Sexual attractivity, proceptivity, and receptivity in female mammals.
Harm. Beh. 7: 105—138.

117

Bekele, A., E. Capanna, M. Corti, L. F. Marcus, and D. A. Schlitter. 1993. Systematics
and geographic variation of Ethiopian Arvicanthis (Rodentia, Muridae). J. Zool.
Lond. 230: 117-134.

Bere, R. M. 1962. The Wild Mammals of Uganda and Neighboring Regions of East
Africa. Longmans, London.

Berger, P. J., N. C. Negus, E. H. Sanders, and P. D. Gardner. 1981. 6-
methoxybenzoxazolinone: a plant derivative that stimulates reproduction in
Microtus montanus. Science. 214: 67-70.

Berrnant, G. 1965. Rat sexual behavior: photographic analysis of the intromission I
response. Psychon. Sci. 2: 55-56.

Blandau, R. J .; Soderwall, A. L. 1941. Post-parturitional heat and the time of ovulation in
the albino rat. Data on parturition. Anat. Rec. 81: 419-429.

 

Bonhomme, F., D. Iskandar, L. Thaler, and F. Petter. 1985. Electromorphs and phylogeny
in murid rodents. Pp. 671-683 in Evolutionary Relationships Among Rodents: a n
Multidisciplinary Analysis (W. P. Luckettand J-L. Hartenberger, eds.). Plenum
Press, New York.

Breed, W. G. 1967. Ovulation in the genus Microtus. Nature. 214: 826.

Bronson, F. H. 1989. Mammalian Reproductive Biology. University of Chicago Press,
Chicago.

Capanna, E., and M. V. Civitelli. 1988.A cytotaxonomic approach of the systematics of
Arvicanthis niloticus (Desmarest 1822) (Marnmalia Rodentia). T rop. Zoo]. 1: 29-
37.

Carleton, M. D. and G. C. Musser. 1984. Muroid rodents. Pp. 289-380 '_rr_1_ Orders and
Families of Recent Mammals of the World (S. Anderson and J. K. Jones, eds).
John Wiley and Sons, New York.

Chevret, P., C. Denys, J. J. Jaeger, J. Michaux and F. Catzeflis. 1993. Molecular and
Paleontological aspects of the tempo and mode of evolution in Otomys
(Otomyidae: Muridae: Marnmalia). Biochem. Syst. Ecol. 21: 123-131.

Coe, M. J. 1972. The South Turkana Expedition. IX. Ecological studies on the small
mammals of South Turkana. Geo. J. 138: 316-338.

Compoint-Monmignaut, C. 1968. Cycle oestral d’un gerbellide et de quelques murides
d’Afrique central. Mammalia. 32: 524-525.

118

Conaway, C. H. 1971. Ecological adaptation and mammalian reproduction. Bio. of Rep.
4: 239-247.

Connor, J. R.; Davis, H. N. 1980. Postpartum estrus in Norway rats. 1. Behavior. Biol.
Rep. 23: 994-999;

Corbet, G. B., and J. E. Hill. 1991. A World List of Mammalian Species. Oxford
University Press, Oxford.

Corbet, G. B., and D. W. Yalden. 1972. Recent records of mammals (other than bats)
from Ethiopia. Bull. Brit. Mus. Nat. Hist. 22: 213-252.

Cox, F. E. G. 1979. Ecological importance of small mammals as reservoirs of disease. Pp.
213-238 in Ecology of Small Mammals (D. M. Stoddart, ed.). Chapman and Hall,
London.

Davis, D. H. S. 1963. Wild rodents as laboratory animals and their contribution to
medical research in South Africa. S. Afi'. J. Med Sci. 28: 53-69.

Delany, M. J. 1964a. A study of the ecology and breeding of small mammals in Uganda.
Proc. Zaal. Soc. Land. 142: 347-370.

Delany, M. J. 1964b. An ecological study of the small mammals in Queen Elizabeth Park,
Uganda. Rev. 2001. Bot. Afr. 73: 339-344.

Delany, M. J. 1975. The Rodents of Uganda. Trustees of the British Museum (Nat Hist),
London.

Delany, M. J. 1986. Ecology of small rodents in Africa. Mammal Rev. 16: 1-41.

Delany, M. J. and D. C. D. Happold. 1979. Ecology of African Mammals. Longman, New
York.

Delany, M. J ., and W. D. K. Kansiimeruhanga. 1970. Observations of the ecology of
rodents from a small arable plot near Kampala, Uganda. Rev 2001 Bat Afi'. 81 :417-
425.

Delany, M. J ., and R. H. Monro. 1985a. Growth and development of wild and captive
Nile rats, Arvicanthis niloticus (Rodentia: Muridae). Afi. J. Ecol. 23: 121-131.

Delany, M. J ., and R. H. Monro. 1985b. Movement and spatial distribution of the Nile rat
(Arvicanthis niloticus) in Kenya. J. Trap. Ecol. 1:111-130.

Delany, M. J ., and R. H. Monro. 1986. Population dynamics of Arvicanthis niloticus
(Rodentia: Muridae) in Kenya. J. Zool, Land. 209: 85-103.

 

119

Delany, M. J ., and B. R. Neal. 1966. A Review of the Mmidae (Order Rodentia) of
Uganda. Brit. Mus. (Nat. Hist.) 2001. Bull. 13: 295-355.

Delany, M. J ., and B. R. Neal. 1969. Breeding seasons in rodents in Uganda J. Reprod.
Fert, Suppl. 6: 229-235.

Delany, M. J ., and C. J. Roberts. 1978. Seasonal population changes in rodents in the
Kenya Rift Valley. Bull. Carnegie Mus. Nat. Hist. 6: 97-108.

Demeter, A. 1983. Taxonomical notes on Arvicanthis (Mammalia: Muridae). from the
Ethiopian rifi valley. Proc. 3rd int. coll. ecol. & taxon. African small mammals.
Ann. Mus. Ray. Afi. Centr., Sc. 2001. 237: 129-136.

Dewsbury, D. A. 1972. Patterns of copulatory behavior in male mammals. Q. Rev. Bio.
47 : 1-33.

Dewsbury, D. A. 1975. Diversity and adaptation in rodent copulatory behavior. Science.
190: 947-954.

 

Dewsbury, D. A., B. Ferguson, and D. G. Webster. 1984. Aspects of reproduction,
ovulation, and the estrous cycle in African four-striped grass mice (Rhabdamys
pumilia). Mammalia. 48: 417-424.

Dieterlen, F. 1990. True or long-tailed mice. Pp. 152-205 i_n Grzimek’s encyclopedia of
mammals. Volume II: Rodentia and Carnivara (S. P. Parker, ed). McGraw-Hill
Publishing Company, New York.

Dollman, G. 1911. New and interesting mammals from East Africa. Ann. Mag. Nat. Hist.
8(43): 124-131.

Dollman, G. 1911. On Arvicanthis abyssinicus and allied East-African species, with
descriptions of four new forms. Ann. Mag. Nat. Hist. 8: 334-353

Dorst, J. 1972. Notes sur quelques rongeurs observes en Ethiopie. Mammalia. 36: 182-
192.

Drake-Brockman, R. E. 1910. The Mammals of Samaliland. Hurst and Blackett, Ltd,
London.

Duplantier, J.-M., and L. Granjon. 1990. Rhythms d’activité chez six especes de Muridés
du Sénégal appartenant aux genres Mastamys, Arvicanthis, Myamys et Dasymys.
Mammalia. 54: 173-182.

120

Ellerman, J. R. 1940. The Families and Genera of Living Rodents. II. Muridae. Trustees
of the British Museum (Natural History), London.

Eckstein, P. 1949. Patterns of the mammalian sexual cycle. Acta Anat. 7: 389-410.

Erb, D. E., H. E. Edwards, K. L. Jenkins, L. C. Mucklow, and K. E. Wynne-Edwards.
1993. Induced components in the spontaneous ovulatory cycle of the Djungarian
hamster (Phadapus campbelli). Phys. Beh. 54: 955-959.

Feder, H. H. 1981. Estrous cyclicity in mammals. Pp.79-348 i_n Neuraendocrinalagy of
Reproduction: Physiology and Behavior (N. T. Adler, ed.). Plenum Press, New
York.

Fisher, M. 1991. A reappraisal of the reproductive ecology of Arvicanthis in Afiica. Afi.
J. Ecol. 29: 17-27.

Gautun, J .-C., M. Trainer, and B. Sicard. 1985. Liste préliminaire des ronguers du
Burkina Faso (ex Haute-Volta). Mammalia. 49: 537-542.

Ghobrail, L. I., and A. S. K. Hodieb. 1982. Seasonal variation in the breeding of the Nile
rat (Arvicanthis niloticus). Mammalia. 46: 319-333.

Graur, D. 1994. Molecular evidence concerning the phylogenetic integrity of the
Murinae. Israel. J. Zool. 40: 255-264.

Greenwald, G. S. 1956. The reproductive cycle of the field mouse, Microtus califarnicus.
J. Mamm. 37: 213-222.

Hardy, D. F. 1972. Sexual behavior in continuously cycling rats. Behaviour. 41:228-297.
Harrison, D. L. 1972. The Mammals of Arabia. VIII. Ernest Benn Ltd, London.

Hayssen, V., A. van Tienhoven, and A. van Tienhoven. 1993. Asdell ’s Patterns of
Mammalian Reproduction. Cornell University Press, New York.

Honacki, J. H., K. E. Kinman, and J. W. Koeppl. 1982. Mammal Species of the World: a
T axanomic and Geographic Reference. Allen Press, Inc., Lawrence, Kansas.

Huang, H. H., and J. Meites. 1975. Reproductive capacity of aging female rats.
Neuroendacrinalagy. 17: 289-295.

Hubbard, C. A. 1972. Observations on the life histories and behaviour of some small
rodents from Tanzania. Zaal. Afr. 7: 419-449.

 

121

Johnston, R. E. 1983. Chemical signals and reproductive behavior. Pp. 3-37 i_n
Pheromones and Reproduction in Mammals (J. G. Vandenbergh, ed.). Academic
Press, New York.

Jolly, A. 1973. Primate birth hour. Int. Zoo Yb. 13:391-397.
Jones, M. L. 1982. Longevity of captive mammals. Zaal. Garten. 52: 113-128.

Kaminski, M. and F. Petter. 1984. Variabilité electrophorétique des protéines et enzymes
sanguins comparée chez quelques muridés Afiicans (Arvicanthis niloticus,
Mastomys erythraleucus et Mastomys huberti). Bull. Soc. Zaal. de France. 109:
399-409. .

Kaminski, M., M. Rousseau and F. Petter. 1984. electrophoretic studies on blood
proteins of Arvicanthis niloticus. Biochem. Syst. Ecol. 12: 215-224.

Kaminski, M., M. Sykiotis, J. M. Duplantier and A. Poulet. 1987. Electrophoretic
variability of blood proteins among populations of the genera of African rodents:

Arvicanthis and Mastomys from Senegal. Genetic polymorphism and geographic
differences. Biochem. Syst. Ecol. 15: 149-165.

 

Katona, C. 1., and L. Smale. 1996.

Keller, B. L. 1985. Reproductive patterns. Pp. 725-778 i_n Biology of New World
Microtus (R. H. Tamarin, ed.). The American Society of Mammalogists.

Kent, G. C., and R. A. Smith. 1945. A study of the estrous cycle in the golden hamster,
Cricetus (Mesocricetus) auratus Waterhouse. Anat. Rec. 92: 263-271.

Keuhn, R. E.; Beach, F. A. 1963. Quantitative measurement of sexual receptivity in
female rats. Behaviour. 21 :282-294.

Kingdon, J. 1974. East African Mammals: an Atlas of Evolution in A fiica, Volume IIB.
The University of Chicago Press, Chicago.

Kowalski, K., and B. Rzebik-Kowlaska. 1991. Mammals of Algeria. Polish Academy of
Sciences, Warsaw.

Krebs, C. J. 1996. Population cycles revisited. J. Mamm. 77: 8-24.
Kyelem, M. 1993. Indicatuers cytologiques et stréoidiens du cycle oestrien des Muridés

et Gebillidés Sahéliens du Burkina F aso: variations saisonniéres et
expérimentales. These 3 cycle, Université de Ougadougou, Ouagadougou.

122

Labyak, SE, and T. M. Lee. 1995. Estrus- and steroid-induced changes in circadian
rhythms in a diurnal rodent, Octodon degus. Phys. Beh. 58: 573-585.

Levin, R. N., and R. E. Johnston. 1986. Social mediation of puberty: an adaptive female
strategy? Behav. Neur. Bio. 46: 308-324.

Lincoln, D. W.; Porter, D. G. 1976. Timing of the photoperiod and the hour of birth in
rats. Nature. 260:780-78.

Linn, I. J. 1991. Influence of 6-methoxybenzoxazolinone and green vegetation of
reproduction of the mulitmarnmate rat Mastomys caucha. S. Afi. J. Wildl. Res. 21:
33-37.

Long, J. A., and H. M. Evans. 1922. The oestrous cycle in the rat and its associated
phenomena. Mem. Univ. Calif 6: 1-148.

Madison, D. M. 1985. Activity rhythms and spacing. pp. 373-419 i_n Biology of New
World Microtus (R. H. Tamarin, ed.). The American Society of Mammalogists.

Mather, J. G. 1981. Wheel-running activity: a new interpretation. Mammal Rev. 11:41 -5 l.

Matthey, R. 1959. Forrnules chromosomiques de Muridae et Spalacidae: 1a question du
polymorphisme chromosomique chez les mammifieres. Revues Suisse de
Zaolagie. 66: 175-209.

Matthey, R. 1965. Etudes cytogenetique sur des Murinae Africains apparterant aux genres
Arvicanthis, Praamys, Acomys et Mastomys (Rodentia). Mammalia. 29: 228-249.

Mayr, E. 1942. Systematics and the Origin of Species. Columbia University Press, New
York.

McClintock, M. K., and N. T. Adler. 1978. The role of the female during copulation in
wild and domestic Norway rats (Rattus norvegicus). Behaviour. 67: 67-96.

Milligan, S. R. 1975. Mating, ovulation and corpus luteum function in the vole, Microtus
agrestis. .1. Rep. Fert. 42: 35-44.

Misonne, X. 1969. Afiican and Indo-Australian Muridae. Evolutionary Trends. Ann. Mus.
Ray. Afi'. Centr.

Misonne, X. 1971. Order Rodentia, Part 6 i_n The mammals of Afi'ica: an identification
manual (J. Meester and H. W. Setzer, eds.). Smithsonian Institution Press,
Washington, DC.

 

123

Mitchell, J. A.; Yochim, J. M. 1970. Influence of environmental lighting on duration of
pregnancy in the rat. Endacrin. 87:472-480.

Moore-Ede, M. C., F. M. Sulzman, and C. A. Fuller. 1982. The Clocks that Time Us.
Harvard University Press, London.

Muller, J. P. 1977. Populatiosokologie von Arvicanthis abyssinicus in der grassteppe des
Semien Mountains National Park (Athiopien). Z Saugetierkd. 42:145-172.

Musser, G. C., and M. D. Carleton. 1993. Family Muridae. Pp. 401-755 _i_r_r Mammal
species of the world: a taxonomic and geographic reference (D. E. Wilson and D.
M. Reeder, eds.). Smithsonian Institution Press, Washington.

Myllymaki, A. 1979. Importance of small mammals as pests in agriculture and stored
products. Pp. 239-280 i_n Ecology of Small Mammals (D. M. Stoddart, ed.).
Chapman and Hall, London.

Nandwa, S. M. 1973. Feeding ecology of rodents in Kenya. E. Afr. Wildl. J. 11: 407-408.

Neal, B. R. 1981. Reproductive biology of the unstriped grass rat, Arvicanthis, in East
Afi'ica. Z. Saugetierk. 46:174-189.

Negus, N. C., and P. J. Berger. 1977. Experimental triggering of reproduction in a natural
population of Microtus montanus. Science. 196: 1230-123 1.

Nelson, R. 1995. An Introduction to Behavioral Ecology. Sinauer Associates, Inc.
Massachusetts.

Nowak, R. M. 1991. Walker ’s Mammals of the World, Volume 2. The Johns Hopkins
University Press, Balitrnore.

Orlov, V. N., M. I. Baskevich, and N. Sh. Bulatova. 1992. Chromosomal sets of rats in
the genus Arvicanthis (Rodentia, Muridae) from Ethiopia. Zaal. Zh. 71: 103-112.

Packer, C. 1983. Demographic changes in a colony of Nile grass rats (Arvicanthis
niloticus) in Tanzania. J. Mamm. 64:159-161.

Peterson, S. L. 1986. Age- and horrnone-related changes in vaginal smear patterns in the
gray-tailed vole, Microtus canicaudatus. J. Reprod. Fert. 78: 49-56.

Petter, F., M. Quilici, Ph. Ranque, P. Camerlynck. 1969. Croisement d’Arvicanthis
niloticus (Rongeurs, Muridés) du Sénégal et d’Ethiopie. Mammalia. 33: 540-541.

Poulet, A. R. 1976. Evolution of the rodent population of a dry bush savanna in the
Senegalese Sahel from 1969 to 1977. Bull. Mus. Nat. Hist. 6: 113-117.

 

124

Poulet, A. R., and H. Poupon. 1978. L’invasion d’Arvicanthis niloticus dans le Sahel

Séne'galais en 1975-1976 et ses consequences pour la strate ligneuse. Terre et Vie.
32: 161-194.

Quilici, M., Ph. Ranque and P. Camerlynck. 1969. Elevage au laboratiore d’Arvicanthis
niloticus. Mammalia. 33: 245-247.

Rabiu, S. and M. Fisher, 1989. The breeding season and diet of Arvicanthis in northern
Nigeria. J. Trap. Ecol. 5: 375-386.

Ramirez, V. D., and W. L. Soufi. 1994. The neuroendocrine control of the rabbit ovarian
cycle. pp. 585-611 i_n The Physiology afReproductian (E. Knobil and J .D. Neill,
eds.). Raven Press, New York.

Refmetti, R., and M. Menaker. 1992. The circadian rhythm of body temperature. Phys.
Beh. 51: 613-637.

Richmond, M., and C. H. Conaway. 1969. Induced ovulation and oestrus in Microtus
ochragaster. J. Reprod. Fert., Suppl. 6: 357-376.

Ripper, W. E., and L. George. 1969. Cotton Pests of the Sudan. Blackwell Scientific
Publications, Oxford.

Rosevear, D. R. 1969. The Rodents of West Afi'ica. Trustees of the British Museum
(Natural History), London.

Rousseau, M. 1983. Etude des Arvicanthis du Muséum de Paris par analyses factorielles
(Rongeurs, Muridés). Mammalia. 47:525-542.

Rowland, D. L.; Helgeson, V. 8.; Cox, C. C. III. 1984. Temporal patterns of parturition in
mammals in captivity. Chranabialagica. 43: 31-39.

Rowlands, I. W., and B. J. Weir. 1984. Mammals: non-primate eutherians. Pp. 455-658 i_r_r_
Marshall’s Physiology of Reproduction: Reproductive Cycles of Vertebrates (G.
E. Lamming, ed.). Churchill Livingstone, New York.

Sadleir, R. M. F. S. 1973. The Reproduction of Vertebrates. Academic Press, New York.

Seabloom, R. W. 1985. Endocrinology. Pp. 685-724 in Biology of New World Microtus
(R. H. Tamarin, ed.). The American Society of Mammalogists.

Schmutterer, H. 1969. Pests of Craps in Northeast and Central Africa. Gustav Fischer-
Verlag, Portland.

 

125

Senzota, R. B. M. 1982. The habitat and food habits of the grass rat (Arvicanthis
niloticus) in the Serengeti National Park, Tanzania. Afr. J. Ecol. 20: 241-252.

Senzota, R. B. M. 1983. A case of rodent-ungulate resource partitioning. J. Mamm. 64:
326-329.

Senzota, R. B. M. 1990. Activity patterns and social behavior of the grass rats
[Arvicanthis niloticus (Desmarest)] in the Serengeti National Park, Tanzania.
Trap. Ecol. 31: 35-40.

Sicard, B., D. Maurel, J. Gautun and J. Boissin. 1988. Activation ou inhibition testiculaire
par la photopériode chez plusiers especes de rongeurs sahéliens: premiere rrrise en .

évidence d’une variation circadienne de la photogonadosensiblité. C. R. Acad. Sci.
Paris. 307: 11-17.

Sicard, B., D. Maurel, F. Fuminier and J. Boissin. 1992. Circadian rhythm and
photosensitivity and the adaptation of reproductive function to the environment in

two populations of Arvicanthis niloticus from Mali and Burkina Faso. J. Reprod.
Fert. 95:159-165.

 

E‘E

Sicard, B., F. Fuminier, D. Maurel and J. Boissin. 1993. Temperature and water
conditions mediate the effects of day length on the breeding cycle of a Sahelian
rodent, Arvicanthis niloticus. Biol. Reprod. 49: 716-722.

Sicard, B., D. Maurel, F. Fuminier and J. Biossin, 1994. Climate, trophic factors, and
breeding patterns of the Nile grass rat (Arvicanthis niloticus solatus): a 5-year
study in the Sahelian region of Burkina Faso (formerly Upper Volta). Can. J.
Zool. 72: 201-214.

Snedecor, G. W., and W. G. Cochran. 1967. Statistical Methods. Iowa State University,
Iowa.

Southern, H. N. 1979. population processes in small mammals. Pp. 63-102 i_n Ecology of
small mammals (D. M. Stoddart, ed.). Chapman and Hall, London.

Stockard, C. R., and G. N. Papanicolaou. 1917. The existence of a typical oestrous cycle
in the guinea-pig, with a study of its histological changes. Am. J. Anat. 22: 225-
283.

Taylor, K. D. 1968. An outbreak of rats in agricultural areas of Kenya in 1962. E. Afi.
Agric. For. J. 34:66-77.

Taylor, K. D., and M. G. Green. 1976. The influence of rainfall on diet and reproduction
on four African rodent species. J. Zaal., Land. 180: 367-389.

126

Turner, C. D., and J. T. Bagnara. 1971. General Endocrinology. W. B. Saunders,
Philadelphia.

van Teinhoven, A.. 1983. Reproductive cycles. Pp. 196-249 i_n Reproductive Physiology
of the Vertebrates. Cornell University Press, London.

Vesey-Fitzgerald, D. F. 1966. The habits and habitats of small rodents in the Congo river
catchment region of Zambia and Tanzania. Zoo. Afr. 2: 111-122.

Viegas-Pequignot, E., B. Dutrillaux, M. Prod’Homme and F. Petter. 1983. Chromosomal
phylogeny of Muridae: a study of ten genera. Cytogenet. Cell Genet. 35: 269-278.

Viswanathan, N.; Davis, F. C. 1992. Timing of birth in Syrian hamsters. Biol. Rep. 47: 6-
10.

Volobouev, V. T., E. Viegas-Pequignot, F. Petter and B. Dutrillaux. 1987. Karyotypic
diversity and taxonomic problems in the genus Arvicanthis (Rodentia, Muridae).
Genetica. 72: 147-150.

Volobouev, V. T., E. Viegas-Pequignot, M. Lombard, F. Petter, J. M. Duplantier and B.
Dutrillaux. 1988. Chromosomal evidence for an polytypic structure of Arvicanthis
niloticus (Rodentia, Muridae). Z. 2001. Syst. Evalut-farsch. 26: 276-285.

Walker, E. P. 1975. Mammals of the World. Volume 11. Johns Hopkins University Press,
Baltimore.

Wang, G. H. 1923. The relation between ‘spontaneous’ activity and the oestrous cycle in
the rat. Camp. Psych. Mon. 2: 1-27.

Weinbren, M. P., and P. J. Mason. 1957. Rift Valley fever in a wild field rat Arvicanthis
abyssinicus. A possible natural host. S. Afi. Med. J. 31.

Yalden, D. W. 1988. Small mammals of the Bale Mountains, Ethiopia. Afi'. J. Ecol. 26:
281-294.

Yalden, D. W., M. J. Largen, and D. Kock. 1976. Catalogue of the mammals of Ethiopia,
II: Insectivora and Rodentia. Ital. J. 2001. 1: 1-118.

Zucker, I. 1980. Light, Behavior, and Rhythms. Pp. 93-101. i_n Neuroendacrinolagy (D.
T. Krieger and J. C. Hughes, eds.). Sinauer Associates, Inc, Massachusetts.

 

MICHIGAN smTE UNIV. LIBRt-IRIES
IIHIIWIW llll HIWW Ill W WIN W I‘ll!” W W
31293015728870