The influence of differential early social experience upon spatial distribution within populations of prairie deermice

THE INFLUENCE

OF DIFFERENTIAL EARLY SOCIAL EXPERIENCE

U P O N SPATIAL DISTRIBUTION WITHIN POPULATIONS OF PRAIRIE DEERMICE* BY C. R I C H A R D T E R M A N Biology Department, College of William and Mary, Williamsburg, Virginia Introduction

each study maintained under identical conditions of the physical environment indicates that factors intrinsic to each population were operating to bring about cessation of growth. The above studies and others have shown that when populations reached asymptote (stopped growing) the physiological alterations of the individuals of each were similar between populations regardless of the number of animals present (Brown, 1953; Christian, 1956, 1957; Clarke, 1955; Crowcroft & Rowe, 1957;

The importance of social factors to the dynamics of local populations of small mammals is becoming more and more apparent. Recent studies of laboratory and semi-natural populations of small mammals in which the physical environment was controlled have shown wide variations in the numbers of animals present at asymptote. Table I lists several studies and the range in maximum numbers among the several experimental populations of each study.

Table I. Variability of Population Asymptote Within Experiments.

Study

Animal

Enclosure size

Brown (1953)

Mus musctdus

6' × 4'

Christian (1956)

Mtts musctt[us

29" x 75" x 2 decks

Number of Populations

Number of Animals at Asymptote 3-- 19

4

21--130

Petrusewicz (1957)

Mus musculus

80 cm. × 80 cm.

x 15 cm.

'?

16-- 69

Southwick (1955a)

Mus musculus

6' x 25'

6

25--130

Clarke (1955)

Mierotus agrestis

67 ms

2

41-- 58

Louch (1956)

Mierotus pennsylvanieus

6' x 25'

3

28-- 67

In these populations food, water, and harbourage were supplied in excess of utilization. The data in the table cannot be compared between studies because criteria by which young were considered members of the total population varied, as did the environmental conditions of each study. The variation in maximum population size among the several experimental populations of

Christian & Davis, 1956; Louch, 1956; Southwick, 1955a; Strecker & Emlen, 1953). Thus, factors which control the growth of populations maintained under identical conditions of the physical environment may function at least partially independent of the number of animals m the populations. Social factors may operate in this way. If, as suggested above, social factors are important to the dynamics of populations, manipulation of these factors should be reflected in measurable attributes of the populations. This study was designed to combine laboratory and field techniques in order to examine the significance of social factors to the spatial distribution of prairie deermice

*This investigation was supported by the R. B. Jackson Laboratory, Bar Harbor, Maine, through Grant C.R.T. 5013, while the author was a Pre-Doctoral Research Fellow at that laboratory. Most of the data presented in this report was included in a dissertation submitted to the Zoology Department, Michigan State University, in partial fulfilment of the requirements for the Ph.D. degree. 246

TERMAN: SPATIAL DISTRIBUTION WITHIN POPULATIONS OF PRAIRIE DEERMICE within local populations. Further, an insight was sought into the part early experience may play in conditioning later social behaviour in populations. The objectives were: 1. To study the influence of social manipulation of deermice populations in the laboratory upon the subsequent spatial distribution of the same mice in a semi-natural environment. 2. To measure the individual awareness of and the preference for the home area as demonstrated by the homing performance of the mice. The influence of differential early experience upon these behaviours also was noted. The author wishes to express his sincere appreciation to Dr. John A. King, Roscoe B. Jackson Laboratory, Bar Harbor, Maine, who provided facilities, invaluable suggestions, and criticisms without which this study would have been impossible. Grateful acknowledgments also are due to Dr. James C. Braddock, Zoology Department, Michigan State University, and to Dr. Don W. Hayne, Patuxent Wildlife Research Centre, Laurel, Maryland, for stimulating suggestions and criticisms throughout this study. MATERIALS

Experimental Animals The prairie deermouse (Peromyscus man# culatus bairdii, Hay and Kennicott) inhabits prairies, open fields, beaches, and dense grass along fence rows in the mid-western United States. It was selected as the experimental animal because a grassland form was considered best suited to the study contemplated and quantities were available from a laboratory colony. No native grassland species of Peromyscus occur in Maine where this study was conducted. Food, nest sites, and nesting material in excess of utilization were provided in an attempt to correct for any ecological imbalance between the maritime environment and the adaptability of this subspecies. The biology of the experimental mice in terms of reproduction, weight, general body condition, and survival indicated that this attempt was successful, at least for the short periods under study.

Laboratory Facilities The mice were born in a laboratory colony of 60 bi-sexual pairs maintained at the Behaviour Division of the Roscoe B. Jackson Memorial Laboratory, Bar Harbor, Maine. All the animals were descendants of 12 original

247

pairs whose offspring had been in the laboratory for approximately 15 generations. The mice were kept in basement rooms and artificial light was provided from 7.30 a.m. to 4.30 p.m. each day. Each bi-sexual pair and its young were kept in 1 compartment of a standard 2-compartment wooden box. Each compartment measured 6 inches × 11 inches × 6 inches and a thin layer of wood shavings covered the floor. Purina laboratory pellets and water in excess of utilization were provided. Further, the mice were placed in other, clean boxes every 2 weeks.

Experimental Field The experimental area was a 0.9 acre field, 250 feet long by 165 feet wide, located at the Hamilton Station of the Jackson Laboratory. The vegetation was predominantly herbaceous, although forbs were found in quantity in parts of the field. The species composition and distribution, and the distribution of plant mass were recorded and are noted elsewhere (Terman, 1959). Enclosure. The experimental field was divided in two 0.44 acre plots, each 240 feet by 80 feet (Fig. 1). The plots were surrounded and separated by corrugated aluminium, 28 inches in width, buried on edge to a depth of 8 inches. (Since the 2 plots were adjacent, a single piece of aluminium served as a common inside boundary). The entire area was surrounded by a sevefoot high wood fence to which wire fencing was attached at the bottom and buried in the ground. Nest Boxes. Twenty-four subterranean boxes, represented by circles in Fig. 1, were placed in each plot along 3 files and 8 ranks. In Plot A these files were B, D, F, and in Plot B they were T, W, Y. The even-numbered ranks indicated nest boxes in both plots. Each nest box was 30 feet from its nearest neighbour in the same file or rank. Nest boxes in files B and F in Plot A, and T and Y in Plot B were 10 feet from the longitudinal aluminium fence nearest them. The end nest boxes in each plot were 15 feet from the end fence. The nest boxes were built of ½ inch lumber and had the following outside dimensions: 6½ inches wide by 5¼ inches deep by 7 inches high. Each box was completely buried in the ground and was covered by a piece of roofing material and a large piece of sod. The mice had access to the boxes via a rubber-hose tunnel of 1 inch inside diameter and 6 inches in length.

248

ANIMAL PLOT

!.

A

C X

D

A

PLOT E X

0 X 0

FSI 0

F

HS Y X iX

0 X

X

E X X

0 X

0

W

Y

0 X

0

BEHAVIOUR,

FS5 0

X

Z

XI,

2-3

Additional information concerning the nest boxes, the vegetation in the field, and the design of the experimental area is given by Terman (1959, 1961, 1962). PROCEDURES

0

Laboratory Procedures

The young mice were separated from their X X X X X parents at 21 days of age and randomly placed either in a group or in isolation. Litters were 0 o ~?~ o o e~ 6 selected for social treatment by the following criteria: birth must have occurred in a 9-day X FS2 x x X FS6 X X 7 period centred around the 70th day prior to the e2d' ~04~ ~5g 8 scheduled date of release into the experimental ~35 -~6~ oh,~ field; each litter must have contained at least × X X X ) 9 9 one male and one female; and both parents must have survived until the date of weaning. IC -e.8,?, el 4,-6 5 "e~ 02~. I 0 Mice Raised in Isolation. Isolation, as used X FS3 X X II X FS7 X X II in this study, refers to a single mouse living in 1 compartment of a standard mouse box (Fig. 2) 0 0 12 d~3~ o o ,2 from weaning at 3 weeks until 10 weeks of age. Visual and tactile contact with other rnice during 13 X X X X X Xt3 the treatment period was prevented. In all cases, male and female sibs lived in different comO 0 0 " 14 0 0 O 14 partments of the same box. X ~5 FS4 X X FS8 X X 5 Mice Raised in Groups. The group social I0' treatment differed from the isolation treatment 0 0 6 i 6 I ~ x ~ e - - - s 0 " - - 4 o e--- 3 o ' - - ) o eL) in two ways. First, the partition separating the 17 x X X X 7 two compartments in each mouse box used in 0 - N E # T BOX the group treatment was made of ¼ inch hardX - T R A P ':STATION ware cloth instead of wood, as in the isolation FS' ~ F E E D I N G STATION boxes (Fig. 2). Second, a male and female were placed in each compartment of each box. All Fig. 1. Design of the experimental field and the pattern of introduction of the mice into the nest boxes. Individuals of four mice assembled in each mouse box were closely associated pairs were released into nest boxes char- strangers to one another and were young from acterized by identical symbols. four different litters. A male and female from one litter were cross-paired with a male and Box Traps. There were 4 files of trap stations female from another litter and the two cross(X in Fig. 1) alternating with the files of nest litter pairs thus formed were placed in two boxes in each plot. Due to a variation in the different mouse boxes. Since the partition beeastern outside boundary of Plot A, there were tween compartments was made of 2 pieces of only 35 trap stations in this plot, while Plot B hardware cloth, separated from each other by a contained 36. Two single catch, box-type live ½ inch space, each mouse in the group situation traps were placed at each trapping station within had tactile contact with only one other mouse, the area enclosed by the aluminium fence. At its pair-mate, and visual and olfactory contact each marginal trap station, one additional trap with the 3 mice in its box. Fig. 3 illustrates the was located outside the aluminium fence and procedure followed in placing mice in the social left set continually. treatment. Feeding Stations. Food, provided ad libitum, The group situation was devised to assure, was placed in a small hardware cloth hopper to as nearly as possible, similar social developprevent hoarding. One hopper was placed at each ment of successive populations of animals raised of 4 feeding stations (FS in Fig. 1) in each plot in groups at the time of their introduction into located at the intersection of files D and W with the field. The unpredictability and difficulty ranks 3, 7, 11, and 15. The feeding stations were of measuring social hierarchies, dominancecontinuously in the field and were distributed submission, etc., which exist in the larger equidistant from the nearest nest boxes.

TERMAN: SPATIAL DISTRIBUTION WITHIN POPULATIONS OF PRAIRIE DEERMICE

EXPERIMENTALPROCEDURE Aqe

Litter A

d?

Litter B

?d

d 9

11

Birth 3Weeks

IsolGtion Raised

roup Raised

Group RQised

IsolQtion Raised

10Weeks Plot B orA Experimental Field 13Weeks .

Remov01 end Dissection

Fig. 3. Experimental procedure. All litters did not contain 2 males and 2 females. Thus, the same litters were not represented in both treatment groups. The second pair in each group situation were from litters other than A or B. groups, were considered sufficient to necessitate as rigid control of social interaction as possible. During the treatment period food and water were supplied ad lib., and the animals were moved to other, clean mouse boxes every two weeks. Both treatment populations were kept in their respective social situations until release into the field at 10 weeks of age. At this age the mice were physically and reproductively mature (Clark, 1938; Dice & Bradley, 1942). Preparation of the Mice for Release. At 10 weeks of age, (Fig. 3), eight mice from each social treatment were released into different one-half acre plots of the experimental field. Animals to be released were selected according to two specifications. First, no pregnant females were acceptable. Pregnancy was ascertained by inspection and palpation. Second, populations released into the field were composed of a male and female from each of four litters. For the mice raised in isolation, this was accomplished by random selection of four mouse boxes each containing a male and female. As indicated previously, grouped males and females were

249

raised as cross-litter pairs. Thus, random selection of one pair necessitated the selection of the reciprocal mating involving the sibs of the first pair selected. Only one of the 2 pairs living in a single group mouse box was released into the field. Henceforth, in this discussion, those mice which were introduced into the field shall be referred to as experimentals and those remaining in the laboratory in their original social treatments as controls. On the day preceding scheduled introduction into the field, both the experimentals and the controls of each social treatment were etherized, weighed, and numbered by toe clipping. In addition, all experimental mice were tagged with a numbered fingerling tag which was used to attach a coloured celluloid disc to one ear. Immediately after weighing and tagging, experimentals raised in isolation were returned to their boxes where they remained in isolation until release into the field. Following weighing and tagging the individuals of each pair of experimentals raised in groups were separated and placed, one to each compartment, in a group-treatment mouse box. Field Procedure

Introduction Into the Field. The nest boxes in each plot were cleaned prior to the introduction of each experimental population into the field. Clean cotton nesting material and 5 pellets of Purina laboratory mouse food were then placed in each box. The experimental mice were released in the field at least one hour before sunset on the same day that the previous populations were removed. This was also true for the first experiment, since preliminary populations preceded it. Release of populations raised in each social treatment alternated between plots A and B for successive experiments. Since preliminary work indicated that releasing the populations at the centre of each plot resulted in high mortality, the mice were introduced into the nest boxes in a pattern which was identical for each plot and was repeated for the 8 experimental periods. The most closely associated pairs were the basic units by which the experimental mice were assigned to nest boxes (Fig. 1). The isolation treatment was represented by 4 sib-pairs, the individuals of each pair having been raised in separate compartments of the same box. The most closely associated pairs raised in groups were not sibs, as in the case of the isolates, but were mice living together since

T E R M A N : SPATIAL DISTRIBUTION WITHIN POPULATIONS O F PRAIRIE D E E R M I C E PLATE

I

Fig. 2. Mouse boxes used in the two social treatments, Left--Group Raised: box with hardward cloth partition; Right--Standard box used for Isolation Raised mice.

Anita. Behav., I I , 2-3

250

ANIMAL

BEHAVIOUR,

weaning. Males and females of each closely associated pair were released into nest boxes separated from one another by a distance of 67.1 feet. In addition, the pattern of introduction of the animals of each population was such that the nearest neighbours of the same sex were separated by a distance of 42.4 feet, and those of the opposite sex by 30 feet. This procedure assured that: (a) Naive animals gained experience with at least one nest box. (b) Original distribution patterns were constant for each experimental population no matter into which plot it was placed. (c) Individuals of all closely associated pairs were initially equidistant from each other. (d) All mice were equidistant from the nearest neighbour of the same sex. (e) All mice were equidistant from the nearest neighbour of the opposite sex. Period of Population Establishment. After inspection of data collected in preliminary experiments, a period 17 days after introduction was considered sufficient for the mice to establish themselves in the field. During this "Period of Population Establishment", all the nest boxes in each plot were inspected daily, and the location of each mouse recorded. No time schedule was followed for examining the nest boxes in each plot, and since all mice wore celluloid discs of a different colour on one ear it was not necessary to handle them for identification. Two periods of live trapping were conducted during each experimental period. Both trapping periods were of four nights' duration and on each night traps remained set until two hours after sunset. At this time mice caught in the traps were recorded and released. The first trapping period began on the second night the mice were in the field and continued through the fifth night after introduction. The second trapping period was conducted on the 14th, 15th, 16th and 17th nights of each period of population establishment.

Combination of Laboratory and Field Procedures Homing. Homing is defined in this study as the ability of individual animals to return to previously occupied nest boxes in an experimental plot after being absent from the plot and nest boxes for approximately 36 hours. The homing test was used to measure the ability of the mice to " h o m e " ; measure the significance of previously established spatial distribution patterns; and provide basic data from which to make corn-

XI,

2-3

parisons of homing ability when the home nest boxes were empty and when aliens were in the home nest boxes. The influence of the differential early social treatments on the performance of the mice with respect to the above categories was noted. Establishment of the Homing Phenomenon. On the morning of the 17th day following the introduction of the populations into the field, all mice were removed from the experimental plots. Each was taken from a nest box at approximately 8.00 a.m., placed in a box trap, transported to the laboratory (a distance of about 500 yards), and placed in isolation in a standard mouse box. Each animal remained in the laboratory for approximately 36 hours with food and water supplied ad lib. Rarely all mice known to be in the field were not found on the morning they were scheduled to be removed and thus could not be taken to the laboratory for the total isolation period. In this event, the live traps in the plot were set that evening and the animals were taken to the laboratory when caught. Subsequent to the 36 hours of isolation, each experimental animal was reintroduced to its home plot at a point distant from the area of its most frequent previous occurrence. Following the technique of Hayne (1949) for calculating the centre of activity from trap data, a "Residence Centre" was calculated for each animal, using nest box records obtained during the last 10 days prior to removal from the field. The field was bisected transversely and mice whose residence centres occurred in one half of the field were reintroduced at the release point in the opposite half. In Plot A the release points were feeding stations 1 and 4, and in Plot B they were feeding stations 5 and 8 (Fig. 1). Reintroduction to the field was made at sunset of the day following removal. The mice were taken from the laboratory in box traps and at the appropriate release points the traps were opened and left on the ground. The investigator immediately left the area and the mice could leave the traps at any subsequent time.

Homing with Aliens Temporarily in the Field. A single alteration in the homing test was made in Experiments 6, 7 and 8. During previous tests all nest boxes in each plot were empty at the time of reintroduction of the experimental mice. Beginning with Experiment 6, however, the plots were transversely bisected, and a single young male deermouse, 20-30 days old, was

KAVANAU: STUDY OF SOCIAL INTERACTION BETWEEN SMALL ANIMALS

PLATE

II

Fig. 1 (b). View of the enclosure with ferromagnetic proximity sensor and passage-shutter switch in foreground.

Anita. Behav., 11, 2-3

TERMAN: SPATIAL DISTRIBUTION WITHIN POPULATIONS OF PRAIRIE DEERMICE placed in each of the 12 nest boxes located in half of each plot. The section of each plot which was selected to receive these aliens was the area in which the residence centres of most homing animals were located. The young aliens were placed in nest boxes approximately two hours before reintroduction of the residents and were retained there until the following morning. Retention of the aliens in the nest boxes was achieved by placing a collar on each animal and attaching a fine wire leader to each collar. The leader was then snapped on a wire loop in the cover of each nest box. All other techniques and procedures for this homing test were similar to those of the previously described test. Data Recorded. Following reintroduction of the resident populations, the location of each mouse was recorded for the next three days. For a mouse to be regarded as having homed successfully, it had to be recorded in a previously occupied nest box during this period. In those experiments in which aliens occupied the home nest boxes of residents, the presence or absence of the residents in the alien-occupied boxes and the condition of the aliens were recorded. As indicated previously, all aliens were removed from the plots on the morning following reintroduction of the residents. On the third morning of the homing phase of each experiment, the mice were taken from the nest boxes, killed with chloroform, weighed, and the adrenal gland removed and weighed. The nest boxes were then cleaned and readied for the introduction of the next experimental population. RESULTS This report is primarily concerned with those measurements which showed consistent differences between the populations raised in groups or in isolation. Discussions of general population phenomena and factors operating independently of the 2 social treatments have been presented elsewhere (Terman, 1961, 1962). A few descriptive data, however, will aid in an understanding of the adaptation of the populations to the experimental situation. Number of mice. There were eight experimental periods, each of three weeks' duration, between June 6th and November 21st, 1958. A population of 4 males and 4 females, raised either in isolation or in groups, was in each o f the two plots o f the field during every experimental period. Thus, the total number of mice

251

released into the plots was 128, of which 64 were reared in each of the two social situations. Selection and Use of Nest Boxes. During the 17-day period of population establishment, the mice were recorded in the nest boxes of the field 93.3 per cent. of the times possible. Not only did the mice use the nest boxes almost exclusively, but each mouse was recorded in a specific few of the 24 boxes available (Terman, 1961). During the total period of population establishment, the females and males raised in isolation were found in an average of 6.0 and 6.2 nest boxes respectively, while the females and males raised in groups were found in 5-7 and 6.3 nest boxes each. In the last ten days of each period of establishment, females raised in isolation were found in an average of 4.3 nest boxes while males were in 4-4. females raised in groups were in 4.2 boxes and males were in 4.3 during the same period of time. These data indicate no significant differences between sexes or social treatments in the average number of nest boxes occupied during each of the periods of measurement. The fact that so few of the 24 nest boxes in each plot were occupied by each mouse may be indicative of an apportioning of the area and nest boxes among the several individuals of each population.

Gregarious Behaviour. Table II indicates the total number of times mice were found in nest boxes alone, in pairs, or in groups of 3, subsequent to release in the field. It may be seen that Table II. Occurrence of Mice in Nest Boxes Alone, in Pairs, and in Groups of Three.

Isolation Raised

Group raised

Mice alone Males

379

351

Females

358

353

Pairs Male and female

84

103

Female and female

0

3

Male and male

4

2

Groups of three One male and two females Two males and one female

252

ANIMAL

BEHAVIOUR,

XI,

2-3

Table III. Analysis of Variance of the Proportion of Nest Box Records in which Mice were Alone.

T Degrees of ._ freedom127

Source of variation

Sum of squares

Mean square

F value

1.56

61,978"49

Total Between experiments

7

5,511 '80

787.49

Between social treatments

1

2,269 "69

2,269.69

Between sexes

1

79 '69

79 "69

Social Treatment x experiment

7

3,914'12

559.16

Social treatment x sex

1

7'51

7.51

.01

Sex x experiment

7

962'62

137.52

.27

Sex x experiment x social treatment

7

647'8l

92.54

.18

96

48,585 '25

506.10

Error

4.48" .16 l.l

*Significant at 5 per cent. level. mice m o s t often were f o u n d alone and rarely combined in other than bisexual pairs. A three-way analysis o f variance of the gregarious behaviour is shown in Table III. The p r o p o r t i o n of the nest box records for each mouse in which the mouse was alone was used as the unit o f measurement o f gregarious behaviour. The above analysis assumes that the effects of the social treatments, experiments, and sex are fixed. It indicates that a significantly higher p r o p o r t i o n of the nest box records of mice raised in isolation than of those raised in groups was recorded while the mice were alone in nest boxes ( . 0 5 > P > - 0 1 ) . A n o t h e r way of saying this is that a higher p r o p o r t i o n of the nest box records obtained f r o m mice raised in groups than f r o m those raised in isolation was made while there were at least 2 individuals in a nest box.

W h e n a mixed model analysis o f variance was performed in which the effects o f the social treatments and sexes were considered fixed and those of experiments considered random, then the F value o f 4.06 obtained by testing the social treatments with the social treatments × experiments interaction was smaller than the value o f 5.59 required for significance at the 5 per cent. level. Bisexual Combinations. All instances in which at least 2 mice of opposite sex occurred together in a nest box are the basis for the following discussion. The important measurements of bisexual combination are given in Table IV. Fig. 4 shows the p r o p o r t i o n of each sex found with the opposite sex at least once and the average n u m b e r o f days spent in each different combination. There were no significant differences between social treatments with regard to

Table IV. Bisexual Combinations.*

Isolation raised

Group raised Females

Females

Males

Males

Number of mice in population (N)

30

30

32

30

Number of mice found with a mouse of the opposite sex (O)

22

18

27

24

Number of different bisexual combinations (~)

27

27

38

38

Number of days in bisexual combination ( # )

90

90

125

125

*Mice must have been present in the plots on the 9th day after introduction to be included in this table. (~, N O-Lsymbols used in Fig. 5),

T E R M A N : SPATIAL DISTRIBUTION WITHIN POPULATIONS OF PRAIRIE D E E R M I C E 10

253

and compared the two sexes and the social treatments on the basis of the number of animals which were found with the opposite sex. Ani alyses of the differences between social groups E'-are discussed below. z:3' S0 Different Bisexual Combinations. Comparisons of these data were made with t tests and although no differences were signicant at the 5 per cent. 2 level, there were several which showed a consistent low level of probability of occurrence by I 2a chance (Table V). Mice raised in groups occurred .2 in consistently, but not significantly higher 0 Duration of Each numbers of different bisexual combinations than With the Opposite Bisexual Combination Sex at Least Once those raised in isolation. Duration of Bisexual Combinations. DifferFig. 4. Proportion of mice found with the opposite sex ences between social treatments in the total and the mean duration of each bisexual combination. number of days mice were found in bisexual combinations (Fig. 5) were measured b,y t tests. these measurements, although the proportion Mice raised in groups were in combinations of mice combining with the opposite sex was (#/N) a longer period of time than isolates, larger for the mice raised in groups than for although not at the P=.05 level of significance those raised in isolation (. 1> P > .05). (Table V). No significant differences between Fig. 5 illustrates the number of bisexual the two sexes of each social treatment were combinations and the total number of days evident from these analyses. mice of each social treatment were found in a It should be mentioned that among those bisexual combination of any type. The number animals which combined (#/O), no differences of events (g~ in Table IV) in each category was approaching significance were apparent in any adjusted by dividing by (N), the total number of of the above categories. animals of the sex being measured or by (O), The influence of the social treatments upon the number of mice which combined in bisexual bisexual combinations may be summarized by combinations. This second method removed saying that a higher proportion of mice raised the effect of differential rates of combination in groups combined: they o" formed a greater number of J,6 0 Isolation Raised different combinations, and they o" 1.5 continued in combinations for I ~ Group Raised 1.4 a longer period of time than mice raised in isolation. Con( ~= 1.3 d sistent differences between an:~ "r.2 imals raised in isolation and in groups have not been demonO. strated to be statistically sigi.o. nificant by individual test comCr parisons. A significant difference ~.8 b. between the social treatments has been shown, however, by .7 the 3-way analysis of variance .6 shown in Table ]I. There the .5 @/N :14:/0 units of measurement were the #/0 ~/N Proportion ; Deys Mice Found Together proportion of nest box recordBisexual Combinations ings in which mice were either = Frequency of Event alone or combined with other N = Number of Animals of Each Sex in Population 0 = Number of Animals of Each Sex Found in Bisexual Combinotion~ mice. Assuming a type I model, mice raised in groups were in Fig. 5. The number of different bisexual combinations and days in combinations combinations for a significantly Iso!ation Raised

(.B4)

il

'-?'

INN

~

Group Raised

254

ANIMAL

BEHAVIOUR,

XI,

2-3

Table V. Comparing Social Treatments as to the Number and Duration of Bisexual Combinations. r

[ i

Larger number of events

Comparison

: i

#/N

Significance level

~¢/O

Number of Combinations All mice

G>I

(.2>P> "1)

(.5>P>.4)

Females

G>I

('3>P>.2)

(.4>P> .3)

Males

G>I

('3>P>'2)

(p>.5)

Duration of Combinations

i

J

All mice

G>I

(.2>P> .1)

(.4>P> .3)

Females

G>I

(.1>P>.05)

(.3>P> .2)

Males

G>I

(.2>P> "1)

(P> .5)

~

@ = Frequency of event. N=Number of animals in populations. O =Number of animals combining with the opposite sex. I=Isolation raised mice. G=Group raised mice. larger proportion of nest box records than

isolates (.05 > P > .02). Latency of Combining with the Opposite Sex. Data were assembled indicating the number of days that elapsed following introduction to the field before each mouse was found with a mouse of the opposite sex. The percentage of mice of each social treatment found for the first time with an animal of the oppsite sex for each day is indicated in Fig. 6, as art accumulative percentage curve. The distributions for the social treatments were compared by the KalmogorovSmirnov test (Siegel, 1956) and the mice raised in groups combined with the opposite sex significantly sooner (.001 > P ) after introduction to the field than those raised in isolation. No comparison was made of the latency of combining with the same sex because of the low frequency of such events.

Spatial Patterns, Spatial distribution was measured as the distance to the nearest neighbour (Clark & Evans, 1954, 1955). The following categories of nearest neighbour measurements were taken, the means of which are listed in Table VI. Distance to : Nearest neighbour of the same sex. Nearest neighbour of the opposite sex. Nearest neighbour of either sex. In all but one distance measurement taken, mice raised in isolation were more widely separated than those raised in groups. The only exception was that males grouped during early life were dispersed a greater average distance than those kept in isolation for the same length of time. Statistical comparisons between social treatments for mice of the same sex were made for

Table VI. Average Daily Distance to Nearest Neighbour.*

Group raised

Isolation raised Males

Females

Males

Females

Same sex

56"9

65.18

58 "4

62.3

Opposite sex

38'49

38"18

34 "82

33.73

Either sex

30 '64

34"48

30'32

28.98

To mice of:

*Distance measured in fee t,

TERMAN: SPATIAL DI8TRIBUTION WITHIN POPULATIONS OF PRAIRIE DEERMICE e¢-.

~ IOO, E

8 80"

-

-

-

-

"

"

-

~

g 60'

o'" . . . . .

/

• Group

Raised N=62

Isolation

f.,...~ ......... ""

Raised

N. 60

2 g

o <

i

oN" 0 2

.

4

.

.

6

.

.

8 I0 DAYS

12

14

16

18

Fig. 6. Latency of combining with the opposite sex. Mice must have been present in the plots on the 9th day after introduction to be included (N). the daily distance measurements in the 3 categories outlined above. The results of these analyses are shown in Table VII. There were no significant differences between females raised in isolation and those raised in groups and between males of the two social treatments in the distance to the nearest neighbour of the same sex when compared for the total period of establishment. Since it was not possible to catch and remove all the animals of each population on the final day of each experimental period, only 5 populations of each social treatment had 17 day periods of population establishment. Thus, a radical difference in the distance to the nearest neighbour on day 17 for one experiment could greatly affect the average for all experiments. This happened in

255

the measurement of distance to the nearest neighbour of the same sex. Therefore, the social treatments were compared using data for only the first 16 days. Females raised in isolation were significantly further from other females than were females raised in groups ( . 0 5 > P>.02). Conversely, males raised in groups were further from other males in the populations than were males that experienced the isolation treatment ( . 1 > P > . 0 5 ) . The average daily distance to animals of the opposite sex was significantly greater for females ( . 0 2 > P > . 0 1 ) and males ( . 0 5 > P > . 0 2 ) raised in isolation than for their counterparts raised in groups. Comparison of the distance to nearest neighbours of either sex showed that females raised in isolation were significantly more dispersed than those raised in groups (.001 > P ) while males of the 2 social treatments did not differ significantly (P>.5). The above analyses indicate that females raised in isolation maintained a greater distance to their nearest neighbours than did those experiencing the group treatment. Males were not so consistent as females. The only significant difference found when comparing males showed that those raised in isolation were further away from females than were grouped males. Males raised in groups occurred further away from other males than those that experienced isolation, although not significantly so. These differences in male and female behaviour may be indicative of a differential effect of the social treatments upon the two sexes.

Establishment of the Homing Phenomenon. Homing was established as a phenomenon in a series of 9 experiments involving 4 populations of mice raised in isolation and 5 populations

Table VH. Comparisons of the Average Daily Distance to the Nearest Neighbour During the Period of Population Establishment.

Significance of greater distance to: Same sex

Opposite sex

Either sex

Females

I>G ('05 >P > '02)

I>G (.02>P>.01)

I>G (.001>P)

Males

G>I (.l>P>.05)

I>G (.05 > P > .02)

I>G (P > .5)

Comparison Isolation vs. group

I=Isolation raised mice. G=Group raised mic,,

256

ANIMAL

BEHAVIOUR,

raised in groups. Following the 36 hour period in the laboratory, a total of 30 mice raised in isolation and 35 raised in groups were reintroduced into the field. Fig. 7 illustrates the homing performance of these mice. Of the 30 mice raised in isolation which were reintroduced into the field, 25 homed by day and 1 and 29 by day 3. Twenty of the 35 mice raised in groups homed by day 1 and 23 by day 3. iO0

i

30

XI,

2-3

the first night following reintroduction of the residents. Table VIII. Establishment of the Homing Phenomenon.

Larger frequency and significance

Comparison

Day 1

Day 3

Observed homing vs. that expected by chance:

35

Isolation raised

90

B>E (.001 >P)

8O

Group raised

2..L3

70

6o

20

5o

222

..~ 40

5572

z

Isolation raised v. Group raised

2C-."

222

2(> I0. .~ Isolation Raised

i

Group

Reised

!

•

RELEASE

[ ] o.Y,

[]

I

I>G ! .05 > P >.02) I

I>G (.01 >P>.001)

B =Homing observed. E=Homing expected by chance. I=Isolation raised mice. G=Group raised mice.

30

0

B>E i (.Ol>P>.OO1

DAY 3

Fig. 7, Percentage of mice homing during experiments prior to those during which aliens were temporarily in the plots. Significance tests were made comparing the observed homing behaviour of the mice with that which would be expected if there were no homing. The levels of significance are indicated in Table VIII, and it may be seen that homing in all categories was greater than expected by chance. A comparison of the homing performance of the mice of the two social treatments (Table VIII) showed that a significantly larger proportion of mice raised in isolation than of those raised in groups homed by day 1 ( . 0 5 > P > . 0 2 ) and by day 3 ( . 0 1 > P > . 0 0 1 ) .

Fig. 8 shows the homing performance for all mice released during these experiments. A total of 19 mice raised in isolation were released while aliens were in the plots, of which 5 homed by day 1 and 13 by day 3. Of the 23 mice raised in groups which were reintroduced into the plots, 6 homed by day 1 and 13 by day 3. Table IX shows the results of tests comparing the homing performance before and after aliens were temporporarily in the plots. Exact probabilities were calculated in those comparisons in which expected values less than 5 occurred. Table IX. Comparisons of Homing Performance Before and After Alien Introduction.

Comparison Homing before aliens vs. that after aliens Isolation raised

Homing with Aliens Temporarily in the Field. As mentioned in the procedures, the homing phase of experiments 6, 7, and 8 differed from those of the preceding experiments in which homing was established as a phenomenon. During the later experiments, young male aliens were retained in one-half the nest boxes during

Larger~equency and significance ! Day 1 Day 3

!

B>A ('001 >P)

B>A (-02>P>-01)

i

Group raised

B>A (.05>P> .02)

ND

(p>.5)

B=Homing before aliens. A=Homing after aliens ternporarily in field. ND =Difference in frequencies sma!l or none,

TERMAN: SPATIAL DISTRIBUTION WITHIN POPULATIONS OF PRAIRIE DEERMICE

257

homing performance to each type of nest box is also compared. Mice raised in isolation homed significantly less after aliens than before aliens no matter whether they were returning to empty or to occupied nest boxes. Of the 8 mice returning to empty nest boxes (Fig. 9), only 3 homed by day 1 (.02>P>.01) and 6 day day 3 (.05>P>.02). Only 2 of the 11 mice homing to alien occnpied nest boxes did so by day 1 (-001 >P) and 7 by day 3 (.02>P>-01). Thus, the presence of aliens in the field during the first night after reintroduction of the residents disrupted the homing performance of mice raised in isolation which were returning to empty, as well as to occupied nest boxes. This was true on both day 1 and day 3.

IsolationRaised

'l

Group Raised

I

~

RELEASE ~

DAY I

N

DAY3

Fig. 8. Percentage of mice homing during the experiments in which aliens were in the plots for one night.

The presence of aliens significantly reduced the homing preference during the first night for mice raised in both social treatments. Homing performance by day 3 was significantly poorer than before aliens for mice raised in isolation only. Differential Homing with Aliens in the Field. As indicated in the procedure, the plots were divided transversely and aliens were placed in the 12 nest boxes in one-half of each plot. Thus, only half the mice of each reintroduced population were homing to nest boxes into which aliens had been placed, Fig. 9 illustrates the homing performance of the mice under these differential conditions. Table X shows the results of statistical comparisons of homing to each type of nest situation with the performance prior to alien introduction. The

Subsequent to alien introduction, mice raised in groups did not significantly differ in their homing success to empty boxes from the rate observed prior to introduction of aliens. Of the 12 mice returning to empty nest boxes, 6 did so by day 1 and 8 by day 3. There were 11 mice raised in groups which were homing to occupied nest boxes. Of these, none homed by day 1 and 5 did so by day 3. Comparison of the proportion of homing success with that observed in establishing homing as a phenomenon indicated a significant difference on day 1 only ('01>P>.001). Further, the deleterious effect on homing performance of introducing aliens was evident only for those mice homing to alien occupied nest boxes. This was true only for day 1, the period during which the young aliens were in the boxes,

ALLEN

"~ ~,oo. ~ g 8o,., 6o. ~ ~ 4o. w zo ~o

ALLENED OCCUPI

EMPTY

OCCUPIED

II

~3

EMPTY "

12

~"

6

Isolation Raised

[]RELEASE

"

~

Group Roised

DAYI

[]

DAY 3

Fig. 9. Percentage of mice homing to empty nest boxes and to those occupied by aliens for one night.

258

ANIMAL

BEHAVIOUR,

Table X. Homing to Alien Occupied or Empty Nest Boxes.

Larger frequency and significance

Comparison

Day 1

Day 3

B>A (.02>P>.01)

B>A .05>P> .02)

-I

Homing before aliens vs. that after aliens i To empty nest boxes i Isolation raised mice

ND

Group raised mice

ND

To occupied nest boxes B>A .02>P>.01)

B>A (.001 > P)

Isolation raised mice

B>A •5 > P > .3)

B>A (.01 > P > .001)

Group raised mice Homing success to I each type of nest t box I Isolation raised mice !

i [ M>O ND

I !

M>O ND

i

Group raised mice

I M> O l i(-01 > P > .001) !

ND

'l

B-~Homing before aliens. A=Homing after aliens temporarily in field. M~Homing to empty nest boxes. O=Homing to alien-occupied nest boxes. ND=Difference in frequencies small or none. and did not have the longer lasting effects observed for the mice raised in isolation. Comparisons were made between the homing performance to nest boxes in which aliens had been temporarily retained and the performance to nest boxes which had not contained alien mice. Table X indicates that mice raised in isolation exhibited no significant difference in homing to the two nest box conditions by day 1 or day 3. Mice raised in groups, however, homed significantly more frequently to alienfree (empty) nest boxes than to occupied nest boxes by day 1 (.01 > P > . 0 0 1 ) . By day 3, the homing performance to empty boxes was not significantly different from that to boxes previously occupied by aliens. Differences in homing behaviour following introduction of aliens were evident for mice of both social treatments when compared with the rate of homing established before aliens. Generally, the homing performance of mice

XI,

2-3

raised in isolation was more adversely affected by the presence of aliens than was that of the mice raised in groups. Comparison of Social Treatments. Following alien introduction, there were no significant differences between social treatments in homing per se, to empty nest boxes, or to alien-occupied nest boxes. DISCUSSION This study was designed to manipulate two social variables and measure their differential effects upon the spatial distribution of mouse populations in a semi-natural environment. While some measurements revealed behavioural patterns that were unaffected by the social manipulations and were, therefore, considered characteristic phenomena of the populations (Terman, 1961, 1962), other measurements showed consistent differences between the mice raised in groups and in isolation. The measurements of gregarious behaviour indicated that mice raised in isolation were recorded alone in nest boxes proportionately more often than those raised in groups. The reciprocal of this is that the proportion of nest box records in which one mouse was combined with at least one other mouse was larger for mice raised in groups than for those experiencing the isolation treatment. N o significant differences were found between social treatments when comparing various data having to do with bisexual combinations, although consistent trends were noted. Mice raised in groups were consistently different from those raised in isolation in that a larger proportion combined with the opposite sex, they were found in a larger number of different combinations, and they were in combinations for a longer period of time. Mice raised in groups combined with the opposite sex significantly sooner after introduction into the plots than did those raised in isolation. Daily comparisons of the average distance between the nearest neighbours of the same, opposite, or either sex showed that females raised in isolation were consistently and significantly further away from their neighbours than were females raised in groups. Male mice raised in isolation were signicantly further from females than were males raised in groups. Males raised in groups, however, were further from other males than were males raised in isolation, but not signicantly so (.1 > P > .05). It is possible

Tt~RMAN: SPATIAL DISTRIBUTION WITHIN POPULATIONS OF PRAIRIE DEERMICE that there was a differential sex effect of the social treatments which an eventual analysis of the data will reveal. The homing performance after aliens were introduced into the nest boxes showed no differences between social treatments in homingper se, to nest boxes in which Miens had never been placed, or to alien-occupied nest boxes. Mice of both social treatments homed to alien occupied nest boxes significantly less frequently than they did when homing was established as a phenomenon. Mice raised in groups showed no significant difference between their established rate of homing and their homing performance to alien-free nest boxes after the aliens were in the field. During and following the time aliens were in the field, however, mice raised in isolation homed significantly less often to boxes which never received aliens than was the case in the experiments establishing homing as a phenomenon. The significant difference between pre- and post-alien homing performances to empty nest boxes for mice experiencing the isolation treatment, was due to their better homing performance as compared to that of the mice raised in groups during the experiments before aliens were in the field. Thus, the introduction of aliens to half the nest boxes had a more adverse effect upon mice raised in isolation than upon those raised in groups whether they were homing to empty or to alien occupied nest boxes. Mice raised in groups homed to empty nest boxes significantly more often than to alien occupied on day 1 only. Mice raised in isolation did not home with a significantly different frequency to either type of nest box. The comparison of social treatments during the experiments when homing was established as a phenomenon showed that although mice of both social treatments homed by day 1 significantly more often than expected by chance, the isolates homed significantly better than those grouped early in life. Not only did mice raised in isolation home more often than those raised in groups but the difference in homing performance between the two social treatments was greater and was significant at a higher level of probability on day 3 ( . 0 1 > P > . 0 0 1 ) than on day 1 (.05 > P > .02). Thus, it appears that animals raised in isolation which did not succeed in homing by day 1 sought to return to a previously occupied nest box by day 3 more often than did mice raised in groups.

259

A summary of the behavioural characteristics of mice raised in isolation as opposed to those raised in groups is as follows: Mice isolated during an early part of their life combined with others less often than those raised in groups; were slower in combining; generally, with few exceptions, maintained a greater distance from their fellows; and homed significantly more often. Differential homing behaviour after aliens were in the field showed mice raised in isolation to be more adversely affected by the introduction of aliens than were mice raised in groups. Mice raised in isolation thus appeared to be less sociable and more spatially orientated than those raised in groups. The effect of the specific early experience of physical isolation upon the subsequent social behaviour of mammals has not been clearly determined (Beach & Jaynes, 1954; Fredericson, 1951; King, 1956b, 1958; Scott & Fredericson, 1951; Scott & Marston, 1950). King & Gurney (1954) and King (1957) reported that male C57B1/10 mice, raised singly from weaning, were less aggressive than their controls raised in groups. Kahn (1954) found that male Mus raised in isolation from weaning were more aggressive than were males raised with their mothers. King & Eleftheriou (1957) raised Peromyscus in isolation and in groups and then released them into the wild in an effort to ascertain the effects of social experience upon adaptation to the natural environment. They were greatly hampered by a precipitous decline in the populations by the end of the first week, but observed that mice raised in isolation were found together in nest boxes less frequently and moved about the field more and to greater distances than mice raised in groups. Evidence previously reported for Peromyscus (Stickel, 1946; Terman, 1961) and other mice and small mammals (Calhoun & Webb, 1953; Orr, 1955) suggested that the spatial arrangements within local populations circumscribed by either natural or artificial barriers, or even in neighbourhoods of a larger more continuous population, exist as a dynamic equilibrium recognized and maintained by members of the population. With the above facts in mind, the greater homing performance of mice raised in isolation than of those raised in groups may be hypothetically explained in the following manner: Spatial and social patterns of distribution were of greater significance to the mice raised in isolation. Mice raised in isolation, after once

260

ANIMAL

BEHAVIOUR,

establishing themselves in a few nest boxes, sought to return to this social and spatial equilibrium to a greater extent than did the more sociable animals raised in groups. This undoubtedly was not an active seeking but rather that the balance of social and spatial stimuli was not similar to earlier adapted levels until the mice were back in their home areas. Since mice raised in groups were socially better orientated, adjustments to different social and related spatial stimuli were more easily made. The fact that there was no difference in homing performance between mice raised in isolation and in groups after aliens were in the field, while there was a significant difference before aliens were introduced, suggests that the introduction of aliens had a more adverse effect upon the homing of the former. The lack of homing difference between social treatments after aliens appeared indicates that the avoidance of aliens or Mien occupied nest boxes may be a natural population phenomenon, and that the basic equilibrium pattern existent in the study area before introduction of the aliens had been disrupted (Terman, 1961). As was mentioned earlier, homing with aliens in the field took place in a series of experiments immediately following those establishing homing as a phenomenon. The difference in the time when each series of experiments was performed should not be considered influential in the poorer homing performance while aliens were in the field, since the mice raised in groups showed no reliable differences in homing to empty nest boxes before or after aliens. In a previous paper it was shown that differences in the social behaviour of individuals are important in determining the spatial distribution within local populations of prairie deermice (Terman, 1961). Calhoun (1949, 1950) with rats, King (1956a) with guinea pigs, and Mykytowycz (1958, 1959, 1960) with rabbits have similarly demonstrated the importance of the behaviour of individual animals in determining the spatial distribution and dynamics of populations. Evidence was presented earlier indicating that population growth may be controlled by factors operating at least partially independent of the number of individuals in a population. That the damping effects on popnlation growth may be a result of intraspecific competition or "social pressure" indicating some qualitative measure of social interaction has been suggested by several workers (Brown, 1953; Christian, 1950, 1956, 1957, 1959, 1960; Clarke, 1955; Nicholson,

XI,

2-3

i933, 1954, 1957; Petrusewicz, 1957; Southwick, 1955a, 1955b). Nicholson (1933, 1957) has suggested that increasing intraspecific competition among animals of a population for requisites of the environment as a result of increasing density is effective in limiting population growth. Christian (1957) postulated that the "social structure of a population in terms of aggressiveness of individual members, their equality or lack of equality and other less well known behaviour factors would determine the maximum population density. . . . . " Thus, in different populations varying numbers of animals may comprise each unit of social pressure as a result of individual differences in behaviour (Brown, 1953; Southwick, 1955a, 1955b; Christian, I957). The total units of social pressure for asymptotic populations under identical conditions of the physical environment would be the same although numbers of animals in each population might differ markedly (Christian, 1957). It is not unlikely that the operation of such social factors may be related to patterns of spatial distribution. Indeed, such a relationship between behaviour and spatial distribution has been suggested by Frank (1957) for Microtus arvalis and M. agrestis and termed the "Condensation Potential". The condensation potential is "based on all intraspecific and especially social behaviour that favours the increase of density" and it "is normally limited by intrinsic behaviour, especially territoriality, to a saturation point which is approximately adapted to the carrying capacity of the environment" (ibid). Davis (1958) referred to the same relationship in a reverse way as the "individual distance tolerance limit" which may be "so low that individuals can never associate together". This is apparently the "individual distance" concept suggested by Marler (1956). Within the matrix provided by the physical environment in any given area, differences in the behaviour of the individual animals of a species population may operate to determine the spatial arrangements existent within that population. The increase in social pressure reflecting a rise in intraspecific competition in combination with increased population density may be a result of exceeding the individual distance tolerance limit determined by the characteristic requirements of the species, behavioural differences of individuals, and the carrying capacity or availability of requisites of the environment. Thus,

]'ERMAN: SPATIAL DISTRIBUTION WITHIN ]?OPULATIONS Ol) PRAIRIE DEERMICE the spatial patterns of distribution may reflect or precipitate the pressures which operate via physiological pathways to curtail population growth. Social behaviour has been shown to be an influential force shaping the patterns of spatial distributions within populations. It is not inconceivable that wild, free living mammals may experience behavioural manipulations no less rigorous than the techniques employed in these experiments. The resultant behavioural variations produced by such manipulations would have significant implications with regard to the genetics, evolution, and dynamics of populations. Summary This study measured the influence of different early social experience upon the spatial distribution of populations of prairie deermice in a semi-natural environment. Successive populations of deermice were raised in the laboratory from weaning age (21 days) in isolation or in groups. At 10 weeks of age, 4 bisexual pairs from each social treatment were systematically released into different .44 acre "mouse p r o o f " plots. There were 8 experimental periods between June 6th and November 21st, 1958, during which 8 successive populations of 4 bisexual pairs were living in each of the 2 plots. Following release into the plots, the daily occurrence of the mice in nest boxes was recorded. Seventeen days after release all mice were taken to the laboratory and isolated for 36 hours. Each was then reintroduced into its home plot at a point distant from its previously established home area. The location of each mouse for the next 3 days was recorded. In the last 3 experiments young Mien mice were retained in onehalf of the nest boxes during the first night after reintroduction of the residents. At the conclusion of each three-week experimental period, all field experimentals and their laboratory controls were killed, weighed, and adrenals removed and also weighed. The differential effects of the social treatments were as follows: (a) The mice raised in isolation combined with others less often than the mice raised in groups; were slower in combining; and generally maintained a greater distance from their fellows. (b) The mice raised in isolation homed significantly more often than those raised in groups

261

during the experiments when homing was established as a phenomenon. The introduction of Miens into one-half of the nest boxes had a more adverse effect upon the homing performance of mice raised in isolation than of those raised in groups. (c) The mice raised in isolation appeared to be less sociable and more spatially orientated than those raised in groups. (d) The data suggest that spatial patterns of distribution existent in the plots were largely determined by social interaction and were of greater significance to mice raised in isolation than to those experiencing the group treatment. Mice raised in isolation adapted less easily to changes in the social and related spatial stimuli than the more sociable group raised mice and, thus, more frequently returned to the earlier established spatial patterns. The introduction of aliens disrupted the social-spatial equilibrium existent in the plots. This disruption had a more severe and longer lasting effect upon mice raised in isolation than upon group raised, due perhaps to the inability of the former to adapt quickly to the environmental changes. Differences in social behaviour have been shown to be important factors determining spatial patterns of distribution within local populations of prairie deermice. The hypothetical interrelationships between social behaviour, spatial distribution, and availability of requisites of the physical environment as factors operating via physiological mechanisms to cause cessation of population growth were discussed. REFERENCES Beach, F. A. & Jaynes, J. (1954). Effects of early experience upon the behaviour of animals. Psychol. Bull., 51,239-263. Brown, R. Z. (1953). Social behavior, reproduction, and population changes in the house mouse (Mus musculus L.). Ecol. Mono., 23, 217-240. Calhoun, J. B. (1949). A method of self-control of population growth among mammals living in the wild. Science, 109, 333-335. Calhoun, J. B. (1950). The study of wild animals under controlled conditions. Ann. N.Y. Acad. Sci., 51, 1113-1122. Calhoun, J. B. & Webb, W. L. 0953). Induced emigrations among small mammals. Science, 117, 358-360. Christian, J. J. (1950). The adreno-pituitary system and population cycles in mammals. J. Mamm., 31, 247-259. Christian, J. J. (1956). Adrenal and reproductive responses to population size in mice from freely growing populations. Ecology, 37, 258-273.

262

ANIMAL

BEHAViOUR,

Christian, J. J. (1957). A review of the endocrine responses in rats and mice to increasing population size including delayed effects on offspring. Nay. med. Res. Inst. Res. Repts., 57, 445-462. Christian, J. J. (1959). The roles of endocrine and behavioral factors in the growth of mammalian populations. Proc. Columbia Univ. Symp. on Comparative Endocrinology. 71-97. John Wiley & Sons, New York. Christian, J. J. (1960). Chapter in Physiological Mammalogy. Acad. Press, New York. ed. by W. Mayer & R. Van Gelder. Currently available in preliminary form as "Endocrine adaptive mechanisms and the physiologic regulation of population growth." Nay. med. Res. Inst. Repts., 60: 2, 49-150. Christian, J. J. & Davis, D. E. (1956). The relationship between adrenal weight and population status of urban Norway rats. J. Mamm., 37, 475-486. Clark, F. (1938). Age of sexual maturity in mice of the genus Peromyscus. Y. Mamm., 19, 230-234. Clark, P. F. & Evans, F. C. (1954). Distance to nearest neighbor as a measure of spatial relationships in populations. Ecology, 35, 445-453. Clark, P. J. & Evans, F. C. (1955). On some aspects of spatial pattern in biological populations. Science, 121, 397-398. Clarke, J. R. (1955). Influence of numbers on reproduction and survival in two experimental vole populations. Proe. roy. Soc., B., 144, 68-85. Crowcroft, P. & Rowe, F. P. (1957). The growth of confined colonies of the wild house-mouse (Mus musculus L.). Proc. zool. Soc. Lond., 129, 359-370. Davis, D. E. (1958). The role of density in aggressive behaviour in housemice. Anim. Behav., 6, 207- 211. Dice, L. R. & Bradley, R. M. (1942). Growth in the deermouse, Peromyscus maniculatus. J. Mamm., 23, 416-427. Frank, F. (1957). The causality of Microtine cycles in Germany. J. Wildl. Mgt., 21, 113-121. Fredericson, E. (1951). Competition: the effects of infantile experience upon adult behaviour. J. abnorm, soc. PsychoI., 46, 406-409. Hayne, D. W. (1949). Calculation of size of home range. J. Mature., 30, 1-18. Kahn, M. V. (1954). Infantile experience and the mature aggressive behavior of mice, some maternal influences. J. Genet. Psychol., 84, 65-75. King, J. A. (1956a). Social relations of the domestic guinea pig living under semi-natural conditions. Ecology, 37, 221-228. King J. A. (1956b). Sexual behavior of C57B1/10 mice and its relation to early social experience. J. genet. PsychoI., 88, 223-229. King, J. A. (1957). Relationships between early social experience and adult aggressive behaviour in inbred mice. J. genet. Psychol., 90. 151-166. King, J. A. (1958). Parameters relevant to determining the effect of early experience upon the adult behaviour of animals. Psychol. Bull., 55, 46-58. King, J . A . & Eleftheriou, B. E. (1957). Effects of social experience in the laboratory upon adaptation of Peromyscus to a natural environment. Anat. Rec., 128, 576. King, J. A. & Gurney, N. L. (1954). Effect of early social experience on adult aggressive behaviour in C57B1/10 mice. J. eomp. physiol. PsyeboL, 47, 326-330,

XI,

2-3

Louch, C. D. (1956). Adrenocortical activity in relation to the density and dynamics of three confined populations of Microtus pennsylvanicus. Ecology, 37, 701-713. Marler, P. (1956). Studies of fighting Chaffinches. (3) Proximity as a cause of aggression. Brit. J. anita. Behav., 4, 23-30. Mykytowycz, R. (1958). Social behaviour of an experimental colony of wild rabbits, Oryctolagus euniculus (L.). 1, Establishment of the colony. C.S.LR.O. Wildl. Res., 3, 7-25. Mykytowycz, R. (1959). Social behaviour of an experimental colony of wild rabbits, Oryctolagus cuniculus (L). II. First breeding season. C.S.LR.O. Wildl. Res., 4, 1-13. Mykytowycz, R. (1960). Social behaviour of an experimental colony of wild rabbits, Oryctolagus cuniculus (L). III. Second breeding season. C.S.LR.O. Wildl. Res., 5, 1-20. Nicholson, A. J. (1933). The balance of animal populations. J. anita. Ecol., 2, 132-178. Nicholson, A. J. (1954). An outline of the dynamics of animal populations. Aust. J. Zool., 2, 9-65. Nicholson, A. J. (1957). The self-adjustment of populations to change. Cold Spring Harbor Symposia Quant. BioL, 22, 153-173. Orr, H. D. (1955). Ranging activity of the northern White-footed mouse, Peromyscus leucopus noveboraeensis Fischer. Ph.D. Thesis, U. Pittsburgh. Petrusewicz, K. (1957). Investigation of experimentally induced population growth. Ekologia Polska, Set. A., Polska Akad. Nauk., 5, 281-309. Scott, J. P. & Fredericson, E. (1951). The causes of fighting in mice and rats. Psysiol. Zool., 24, 273-309. Scott, J. P. & Marston, M. V. (1950). Critical periods affecting the development of normal and maladjustive social behaviour of puppies. J. genet. Psychol., 77, 25-60. Siegel, S. (1956). Non-parametric statistics for the behavioral sciences. New York: McGraw-Hill. Southwick, C. H. (1955a). The population dynamics of confined house mice supplied with unlimited food. Ecology, 36, 212-225. Southwick, C. H. (1955b). Regulatory mechanisms of house mouse populations; social behaviour affecting litter survival. Ecology, 36, 627-634. Stickel, L. F. (1946). The source of animals moving into a depopulated area. J. Mamm., 27, 301-307. Strecker, R. L. & Emlen, J. T. (1953). Regulatory mechisms in house mouse populations--the effect of limited food supply on a confined population. Ecology, 34, 375-385. Terman, C. R. (1959). Social factors influencing spatial distribution in populations of prairie deermice. Ph.D. Thesis, Mich. State Univ. Tenaaan, C. R. (1961). Some dynamics of spatial distribution within semi-natural populations of prairie deermice. Ecology, 42, 288-302. Terman, C. R. (1962). Spatial and homing consequences of the introduction of aliens into semi-natural populations of prairie deermice. Ecology, 43, 216-223.

(Accepted for publication 14th August, 1962. Ms. number: 257)