um:mmgmmmmnmn'mmmnl L

3 12 3 01006 0808

 

 

 

 

 

 

 

JUN 14f??? W95

 

 

ABSTRACI‘

SOME SOCIAL ms 0N
ETING AND DRINKD‘G IN WHICE
PEROHYSCUS MIWMTW BAIRDII AND
P. H. GRACILIS

By
Jame Justin Cooper

Many experiments have shown that social facilitation of eating
occurs in dogs. chickens. and albino rats. These animals eat sore food
when fed in groups than when fed as isolated individuals. Ubreas, most
food and sucrose preference eXperinents have been conducted on individual
subjects. some investigators have used groups of subjects. In view of
the evidence for social facilitation of eating, it seems risky to
generalize from these group preference studies to the behavior of
isolated individuals .

Experiment I was done to see if social facilitation occured under
conditions like those used in sucrose preference studies. Forty-eight
pairs of deemioe were given ad libitum water, 8% sucrose solution, and
food. Half of the time the members of each pair were housed together
in one cage, half of the time they were housed apart in individual
036“.

The results indicated that given water. 2. g. agilis drank more
when housed apart than when housed together in pairs. and both subspecies
ate more when housed apart than when housed together. Given 8% sucrose
solution. both subspecies drank more when housed apart than when housed

‘0

U

:v\
Utcgctisr. although the social conditions did not reliably affect eating.

0 Jane Justin Cooper

Whether water or 8% sucrose solution was given. both subspecies consumed
more calories when housed apart than when housed together. Thus social
interference, the opposite of social facilitation occured.

Two mechanisms might have produced the social interference: (at) an
increase in arousal caused by the presence of other aninls. leading to
the enhancement of the most probable response. and to a decrement in all
other responses (Zajonc. 1965). and Q) social huddling. leading to
reduced heat losses. reduced energy needs, and thus. reduces caloric
consumption (Allee. 1938). hperinent II was an attempt to determine
which of these mechanism caused the social interference by mking
drinking the most probable response. Twelve pairs of deemioe fron
Experiment I were tested as before. except they received water or 8%
sucrose solution for only 1 hour per day. following 23 hours of water
deprivation. Food was still available ad libitun.

The results showed that even when drinking was the most probable
response , social interference of drinking occured. The social huddling
eXplanation of the social interference of caloric consumption is better
able to account for this data than the arousal eXplanation.

Aside from social interference. it was noted that g. g. bairdii
consumed more calories per gram-body-weight and drank less water than
3. g. mills when housed individually: g. g. g_r_a_cilis showed social
interference of water drinking. but 2. g. bairdii did not. 2. a.
bairdii drank less 8% sucrose solution than :1. 3. gracilis. but ate more
food given 8% sucrose solution. and obtained a higher proportion of their
calories from this food. It is possible that some of these differences
between the subspecies were produced by the difference in mean age

James Justin Cooper

between the 2- 2. Ellis and 2. 2. bairdii.

In addition. nice housed in small cages consumed more calories
given 8% sucrose solution than given water. then mixed pairs. con-
sisting of one :1. g. mills and one g. 5. bairdii were housed together.
their intakes did not differ from intakes predicted using the data of

hosogeneous pairs.

SOME 30cm. ms (1‘
EATER} AND DRINKING IN DER MICE
PEROMYBCUS MANICUIATUS BAIRDII AND
P. N. GMCILIB

By

James Justin Cooper

A TIEIS

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

MASTER OF ARTS

Department of Psychology

1 97h

mammalian-1

I an grateful to the members of w committee: Stanley Rather.
Janss Phillips. aid is chairman aalph Levine. without their patience
and understanding, this work could never have been completed. It would
also have been inpossible without the support of 11w long-suffering wife
Jayne. Thanks are also due to John King and his graduate students. who
kindly supplied the deermice used in these experiments.

11

TABLE or CWTMS

Page
LIST a? TABLES . . . . . . . . . . . . . . . . V
mama-um . . . . . . . . . . . . . . . . 1
W I . . . . . . . . . . . . . . . .
METHOD . . . . . . . . . . . . . . . . . .

Procedure . . . . . . . . . . . . . . . . 10
RESULTS . . . . . . . . . . . . . . . . . . 12
Infernal Observations . . . . . . . . . . . 12
Fluid Intake Results . . . . . . . . . . . . 12
Food Intake Results . . . . . . . . . . . . 15
Caloric Intake Results . . . . . . . . . . . 1 6
m Results . . . . . . . . . . . . . . . 19
Predicting T Intakes from A Intakes . . . . . . . 20
mm II . . . . . . . . . . . . . . . . 22
METHOD . . . . . . . . . . . . . . . . . . . 23

mm C O O ' O 0 O O 0 O O O O 0 O O O O 0 2"
DM$IW 0 C O 0 O O O O 0 O O O O I O O O 27

800131 Interfemnw o e s e e e e e s e e e e 27
Species Combination Differences . . . . . . . . . 32

Limitations ofthe Results . . . . . . . . . . 33

111

APPHDIX A:
AMI! Bi

ANALYSIS OF VARIANCE
PILOT mm .

memcm . 0's 0 0

iv

LISTG‘TABLES

The design of Experiment I

0% intake and food intake given 0%:

8% intake and food intake given 8%:

Experiment I
hperinsnt I

Caloric intake and proportion of calories from food .

Variance in T intake accounted for by three models

0%. 8%. and food intakes:

Experiment II

Complete fluid intake analysis of variance:

I O O O 0

(% analysis of variance:

8% analysis of variance:

Complete food intake analysis of variance: Experiment I
Food given 0% analysis of variance:

hperinent I

Ebcperiment I

Food given 8% analysis of variance:

Caloric intake analysis of variance:

Proportion of total calories from food given 8% analysis

of variance

Fluid intake analysis of variance:

Food intake analysis of variance:

Experinent II

Experiment II

Ebtperiment I
Experiment I

Experiment I

uperimnt

Liquid intakes for six pairs in three housing conditions:

Pilot studies

8% intakes for three pairs in four housing conditiais:

Pilot studies

Page
11
13
13
18
21

25

35
35

3?
37

39

1+1

43

45

MEWCl‘Im

Social facilitation is one of the most widely demonstrated phenomina
of social psychology. Scott (1968), following Crawford (1939). defined
social facilitation as ”any increment of performance resulting from
interaction between two or more individuals (p. 57)“ and social inter-
ference as "any decrement of performance resulting from social inter—
action (p. 57)." If. for example. two mice ate more food when they were
housed together. in one cage. than when they were housed apart. in two
cages. social facilitation of eating is said to have occured. but if the
mice ate more food when housed apart than when housed together. social
interference of eating is said to have occured. Social facilitation of
eating has been demonstrated in chickens (relish, 196a and 1965: Tolman
a wilsen. 1965). dogs (Ross 3: Ross, 19'+9a and 19191). James. 1953; James
8: Canon. 1955; James in Gilbert. 1955: James. 1960). and in albino rats
(Harlow, 1932).

The large amount of evidence for social facilitation of eating in
laboratory aninals leads one to question some of the methods which have
been used to test food preferences in the past. Whereas, in most food
preference studies the subjects are housed and tested in individual cages,
some sucrose preference studies have been conducted on animals housed in
groups with sucrose solutions available _a_d_ libitum (Jacobs .9: Scott. 1957;
Carpenter. 1958).

Similarly. in his food preference research. Young (1911;). :95. 19:6.
19+?) carried out a series of experiments in which rats were housed in

1

2

groups in cafeteria cages. In the first of these studies. Young listed
the advantages of this procedure and included the statement that ”The
average intake - intake perrat perday-canbe obtainedbydividing
the total intake for the group by the number of rats in the cage (Young.
19%. p. 372)." He went on to say that the major disadvantage of this
procedure is the loss of individual differences in intake in the group
results. Levine (1968) has shown tint this loss of individual differ-
ecnes in intake can lead to misinterpretation of results in sucrose
preference studies. Furthermore, the social facilitation literature
suggests that a second mjor disadvantage exists: It is unsafe to assume
tint the average per subject daily intake is the same for group-housed
and individually-housed subjects. Social facilitation may cause subjects
housed in groups to consume more food. sucrose solution. water, or other
substances than individually-housed subjects.

Not all studies, however, have demonstrated social facilitation of
eating. Shelley (1965) found that albino rats housed in groups ate less
and gained less weight under ad libitum food and water conditions than
rats housed individually. Thus he found social interference of eating.
not social facilitation.

This descrepanoy can be accounted for using a theory of social
facilitation suggested by Zajonc (1965 and 1968). The basic assumption
of the theory is that the presence of individuals of the same species
causes an increase in non-specific drive or arousal. Recently. Iatane’

a. Cappell (1W2) have demonstrated that in rats, heart rate increases
in the presence of other rats. supporting this assumption. This increase
in arousal causes the enhancement of the dominant or most probable

mapmse in the situation, and a decrement in all nondominant responses.

3

All of the above mentioned studies which denonstrated social facili-
tation of eating used feeding schedules or food deprivation schedules.
It seems likely tint at feeding time, the dominant response of an
animl on a feeding schedule is eating. If a second animl is present
in the feeding situation, one would, therefore, expect social facili-
tation of eating to occur. 01 the other tend, eating may never be the
most probable response for an animal receiving food _a_d_ libitum, as
Shelley's rats were (Shelley, 1965). The dominant response for such
animls might be moving around in their cages, for example, and the
enlnncement of this response caused by the presence of a second animal,
should, according to Zajonc's theory, result in a decrement in all other
responses including eating.

Based on data collected by Vetulani (1931) and Retzlaff (Personal
commication; both cited by Allee, 1938), Allee postulated another
mechanism which might produce social effects on eating behavior in some
circumstances. Retzlaff showed that given abundant food under high
temperature conditions (85° F.), young albino mice gained weight more
rapidly when isolated than when housed in groups. This result tends to
support Shelley's (1965) similar finding with young rats. l-I'owever, under
lower temperatures (61° F.) Retzlaff obtained the opposite result. He
found, as Vetulani had found earlier, that isolated mice grew more slowly
than amp-housed mice.

According to Allee (1938), the mice housed together grow more
rapidly under cool conditions, because they were able to huddle to
keep warm. Huddling animls have a lower body-surfaoe-area to weight
ratio than they would have if they were not huddling. Because heat loss
is proportional to body-surface-alea, the huddling animals lose heat

l;

more slowly than they would otherwise. Hence, by middling, the group-
housed mice conserved energ for growth which the isolated mice used to
maintain their betb' temperatures.

In support of this analysis, it has besn demonstrated that the
minimum rate of metabolism (rate of oxygen consumption) of four house
mice huddled together was only 2.2 tiles as great as that of a single
mouse, implying that less food per mouse was being oxidized by the hud-
dling mice than by the lone mouse (Pearson, 19W). Thus aninls
allowed to huddle should need to consume fewer calories to maintain
their body temperature than those prevented from huddling. This energy
savings might manifest itself as a faster rate of growth, or as a
decrease in food consumption.

Thus Zajenc and Allee proposed two different mechanisms which may
produce social effects on consumtory behavior: (a) arousal caused by
the presence of conspecifics, and (2) energy conservation produced by
huddling when groups of animals are housed together. Either or both of
these meclnnisms could have been at work in the previously cited food
and sucrose preference studies, and not generaliziable to isolated sub-
jects. Hence, the first experiment was done for two reasons: (_a_.) to
see if social facilitation of social interference effects occured in the
eating and drinking behavior of deemice (Peromcus mniculatus bairdii
and g. g. g_r_a_cilis) in conditions comparable to those used in sucrose
preference studies done in our laboratory and elsewhere, and (b) to see
if the mechanisms postualted by Zajonc and Allee could esplain such
effects, if any were found.

To make the present study similar to previously done sucrose
preference studies, food, water, and 8% sucrose solution were given to
the deernice ad libitum, as food was given to the animals in Shelley's
(1965) study and in the studies reported by Allee (1938). Specifically,
the question asked was: Do pairs of nice housed together drink and eat
more or less tien the same pair of mice housed individually?

The mice were tested with water (0%) and 8% sucrose solution (8%)
to see if social effects differed for these two liquids. Collier and
Bolles (1968) conceptualized the typical one-bottle sucrose preference
experiment as a situation in which the subjects select a diet from two
components, food and sucrose solution. Their results showed that rats
in one-bottle sucrose preference studies tend to consume a fixed number
of calories per day, and to get a fixed proportion of these calories
from the sucrose solution. Thus differences caused by changed in caloric
intake might be eXpected to appear in 8% intake, while such differences
should not appear in 0% intake.

Most studies of social facilitation have used only one species of
animal, but Ross 8: Ross (19+%) showed that social facilitation of eating
was greater for one breed of dogs than for another. Two subspecies of
deermice were used in the present study to see if similar social effects
on eating and drinking occured for the two subspecies. Mixed pairs,

made up of one subject from each subspecies, were also tested to

6

determine whether or not social effects for mixed pairs differed from

those for homogeneous pairs.

METHCD

Sub cts

The subjects were 96 male deermice bred in the Michigan State
University Zoology Department Colony: half of the mice were Peromcus
maniculatus bairdii and half were 2. 2. @1113. A sizable body of
psychological research has been done using members of the genus Perom-
g as subjects: see King (1968) for a summary of this work. At the
beginning of the experiment, the-mean ages of the mice were 315 and
576 days, with standard deviations of 58 and 414 days, and ranges of
120 - #10 and 126 - 1839 days for bairdii and asilis respectively.

Prior to the experiment , the mice were housed in groups of from two
to six animals in the small plastic rodent cages described below, with
the subspecies segregated. Both before and during testing, the subjects
were exposed to a 12 hours light - 12 hours dark schedule. They were
maintained on Purina Mouse Chow and tap water and had had no previous
experience with sucrose solutions. Four mice died during the course of
the experiment. All data from the pairs to which these mice belonged
was discarded and replacement pairs were tested.
' was

The mice were tested in the same room in which the mouse colony was
housed. Temperatures in the room generally varied about 5 or 6° F. over
each 24 hour period. The mean daily maximum temperature during the

experiment was 72° F. , the mean daily minimum temperature was 660 F.

8

Maximum and minimum temperatures ranged from 79 to 67° F. and 73 to 611.0
F. respectively.

Halfofthemicewere testedinlargecagesandhalfinsmallcages,
14 x 12 X 6.5 in. and 11 X '7 X 5 in. respectively. Both types of cages
were made of clear plastic and were equipped with tops nude from stainless
steal rods. These cage-tops were designed so that food could be placed
in a trough extending across the width of the cage. In all conditions,
food was spread out along the bottom of the trough, between the drinking
tubes, to preclude competition for food. The cage bottoms were covered
with a layer of wood-chips during testing; no additional nesting material
was given.

Solutions were presented to the mice in 50 ml. Pyrex graduated
cylinders (bottles) graduated in ml. The bottles were fitted with number-
three one-hole rubber stoppers into which Girton stainless steel drinking
tubes had been inserted. The bottles were placed on the cages so that
the drinking tubes protruded into the cages between the rods of the cage
tops. Under all conditions, two bottles were placed on each cage with
their drinking tubes about 4 in. apart. Care was taken to insure that
the tube placements were as similar as possible in the large and small
98883-

A Mettler P—6 electronic balance was used to weigh the mice and their
food, and assorted large beakers and bottles were used to mix and store
the fluids given to the mice.

Pram
The major independent variables of the experiment were:
1. Species combination: The % mice were randomly paired so that

16 paris consisted of two bairdii (BB), 16 pairs of one bairdii and one

9

gracilis (ac), and 16 pairs of two gains (60). These three types of
pairs constituted the levels of the species combination variables.

2. Social conditions hob pair was tested for 21+ days, divided
into four six-day periods. During two of .these periods, the pair was
housed apart, with one of its members in each of two cages (A). During
the other two periods, the pair was housed together in one cage ('1‘). The
social conditions were presented in two different orders: held of the
pairs received the sequence ATAT and half TATA. The third and fourth
six-day periods provided a replication of the first and second periods.

3. Cage sizes Half of the pairs in each species combination. were
tesud in small cages and half in large cages. The cage size variable
was included to control for the possibility that the amount of cage-size
space per aninl might affect the dependent variables under study. If,
for example, only small cages were used, the cage-space did have an
effect on drinking, one might erroneously conclude that social condition
was producing the effect, since in the together condition, cage-space
per animal would be half as great as in the apart condition. Using large
and smll cages controlled for this possible confounding, since if cage-
space per animl produced a significant social condition effect, it should
also produce a significant cage size effect or Cage Size X Social Condition
interaction. because the large cages contained more than double the amount
of space contained in the smell cages.

4. Solution: Within each six-day social condition perdio, the
mice were given tap water (0%) for the first three days and 8% sucrose
solution (8%) for the last three days. Table 1 shows the basic design.

10

Procedure

The mice were run in four squads, each of which contained four pairs
fros each species combination. The pairs of mice were randomly assigned
to Squad 1 Cage Size 1 Order of Social Condition cells so that each cell
contained one pair of each species combination. _

The four squads were started on 5 February, 1 March, 26 March, and
23 April 1969. The replacement pairs were run as soon as possible after
it became apparent that a replacement would be required.

(h the first day for each squad, the mice were weighed and placed
in the proper cages. A weighed amount of Purina Mouse Chow was placed
in each cage-top. Cages containing one mouse were supplied with 7 pellets
of chow, those containing two mice with 11w pellets. Bottles containing
0%were placedonthecages, sndthsamount offluidgivenwasreadto
the nearest .2 ml. and recorded. Fresh bottles were supplied daily and
the amount left in the old bottles was recorded.

The mice received three days of 0% alternating with three days of
8% for 21! days. At the end of each three-day period, the food left in
each cage-top was weighed and fresh food given. At the end of each six-
day period , the mice were placed in clean cages and changed from one
social condition to the other. At the end of the 21+ days of testing,
the mice were given a final weighing, and a new squad was started.

The 8% solutions were mixed by dissolving 160 gm. of commercial
sugar in enough tap water to lake two liters of solution; thus the 8%
refers to weight of solute per unit volume of solution. Fresh solutions
were mixed about every other day, and both 0% and 8% solutions were
stored under refrigeration prior to use.

Table 1 .

11

The design of Experiment I.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Social Condition and Solution
Apart ‘— Together Apart Together
Cage Size
0% 0% 3% i 8%
Bairdii-Bairdii Pairs
‘ Pair 1 :'Pair 1— 2-Pa;ir 1 Pair 1
Le a... : : : + :
Pair 8 Pair _a_ Pair 8 Pair 8
Pair 9 Pair 9 Pair 9 ' Pair 9
terse Cage . ‘ . . .
o l o o 0
Pair 16 ' Pair 16 Pair 16 Pair 16
Bairdii-Gracilis Pairs
Pair 1"“? """“£r"1'7"““Pa m7“,*Fh 17'"
Small Cage : I I 3
Pair 2n Pair 21w Pair 24 Pair 2A
Pair 25 Pair 25 Pair 25 Pair 25
W (”3" : : : :
Pair Jz_ $1r 2 Pair 32 Pair g2
Gracilis-Gracilis Pairs
Pair—33-_—__—_3_—Pair 3 —""'""T"'Pair 3 T—ir 33‘
snail Case I I I 1
Pair 40 Pair two Pair two Pair #0
Pair #1 Pair #1 Pair #1 Pair 1+1
+£4.59 cage 0 o e 0
Pair 1&8 Pair 48 Pair 1:8 Pair as

 

 

REULTS

Informl Observations

The mice were generally inactive while the lights were on in the
laboratory. Typically they curled up in a corner of their cage: when
in the T condition, both mice usually shared the same corner. Obser-
vations wade during the 12 hour dark period, using red lights, showed
that the mice were very active in the dark and spent most of the time
running around their cages in stereotyped patterns: relatively little
time was spent eating and drinking.

Some pairs fought when they were first placed in the T condition.
The fighting seldom lasted more than 21+ hours, and most fights ended
after a few minutes. Bairdii seened more inclined to fight than mi:
ya, and often won fights with @1113, although they were usually
outweighed.
§‘_l_.w_i_i_d_ Mg Results

The individual fluid and food intakes for the two members of each
pair in the A condition were summed to obtain total intakes for the pair.
These intakes were analyzed with the corresponding intakes for each pair
in the T condition.

The mean daily 0% intakes for the six groups of pairs under the
four within-pair treatments are shown in Table 2, and the corresponding
8% intakes are shown in Table 3. Comparing these tables shows that all
groups drank more 8% than 0% in each condition. A complete analysis of

12

13

Table 2. 0% intake and food intake given 0%: Experiment I

 

Species Combination X Cage Size Group

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Social _ M :—
BB BB K} m GG GG
Condition
Snell Lame Small large Small Large
0% Intake in m1./day
:Kpart 11 .4 70.4 11.8 12.8 18.? 14. 5
Together 11.8 10.0 10.6 12.5 10.9 12.4
Food Intake given 0% in gnu/3 days
- Apart — 30.0 26.6 24.8 28.0 23.8 27.6
Together 27.1 24.2 21.9 24.4 21.0 24.7
Table 3. 8% intake and food intake given 851 Ehcperiment I
Social Species—Combination X Cage Size Group
BB BB K} BG GG GG
Condition Snell Large Smll large Stall large
8% Intake in ml./day
Apart 3‘h9 27.9 “5.6 37.3 47.8 48.9
Together 32.9 22.9 141.7 33.8 “2.1+ #2.1
Food Intake given 8% in gm. /3 days
Apart 23.4 20.4 16.2 18.9 15.1 16.4
Together 21.5 20.4 26.3 18.0 13.8 15.9

14

. variance (ADV) was performed on the daily as and 8% intakes. The
solution. social condition. Solution 1 Social Condition. species com-
bination. Solution 1! Species Combination, Social Condition X Species
Combination. Solution X Cage Size. pairs within groups, Solution X
Pairs. and replications effects were significant. p .05 (see Table
1A in Appendix A). Solution was by far the biggest effect, accounting
for 67% of the variance. Since there was no doubt that the nice drank
more 8% than 0%. and since so many other variables interacted with the
solution variables. separate AOVs were drawn on 0% and 8% intakes. To
simplify the calculation of these AOVs. means averaged over replications
and days were used in this analysis.

In the 0% ADV the Social Condition X Species Combination interaction
was significant (see Table 2A in Appendix A). Tests of the simple min
effects (Kirk. 1968) of social condition for the levels of species
combination showed that H: and 06 pairs drank more 0% in the A condition
than in the T condition, E_(1,’+2)=8.7 for 30. £(1,42)=70.6 for GO. 2 .01,
while BB pairs showed no social effect. 01‘ the 1&8 pairs tested. all but
13 drank less in the T condition than in the A condition; 8 of the 13
wereBBpairsandSwerempairs.

Tests of simple main effects of species combination at the two levels
of social condition indicated that differences between the species
combinations occured for the A intakes, §‘_(2,81&)=6.6, p .01. Tukey's
honestly significant difference (HSD) test (Kirk, 1968) revealed that
in the A condition the BB pairs drank less 0% than the G6 pairs, 10,84)
'5.1, 2 .01. and that the K; pairs did not differ significantly from

the BB or 66 pairs.

15

The 8% AOV showed that the mice drank more 8% in A than in T (see
Table 31 in Appendix A). Forty-two of the 48 pairs showed this social
interference effect. The six nonconformist pairs were evenly divided
among the species combinations.

The species combination effect was also significant for 8% intakes,
and Tukey's BSD test indicated that the BB pairs drank less 8% than the
00 pairs. g(3,llv2)-6.8. p .01. and the BC pairs, 3(3,42)=4.3, p .05,
but the latter two groups did not differ significantly.

The Solution x Cage Size interaction, which was significant in the
complete fluid intake AOV. was apparently caused by the large difference
between 8% intakes for the large cage and small cage groups. as compared
with a small difference in the opposite direction for 0% intake. However,
this difference in 8% intake for the cage size groups did not prove
significant in the AOV done on 8% intakes (see Table 3A in Appendix A).

The significant pairs within groups and Solution X Pairs effects
in the complete AOV suggest that some pairs in each group drank more
than others, and that these inter-pair differences were larger for 8%
intake than for 0% intake. The replications effect was significant
because of the small amount of variability between days within repli-
cations. '
filed _Ir_lt_ak__e_ Results

The man three-day food intakes for each group in each treatment,
given 0% and given 8% are shown in Tables 2 and 3 respectively. A
complete ADV done on the food intake data under both % and 8% conditi-
ions (see Table 4A in Appendix A) indicated that the solution, social

condition, Solution K Social Condition, species combination, Solution

16

X Species Combination, pairs within groups. and Solution X Pairs
effects were significant, p .05. Again. there was no doubt that

the mice ate more food under the 0% condition than under the 8% con-
dition, so separate AOVs were calculated for food intakes given 0% and
food intakes given 8%. These AOVs were done on the mean three-day
food intakes for each pair in the T and A conditions averaged over
replications.

The food intake given 0% of AOV indicated-that the mice ate more
food in the A condition than in the T condition (see Table 5A in
Appendix A). No other effects were significant. Again. a large major-
ity of the pairs (42) exhibited social interference to sow degree. Of
the six pairs which did not show interference. three were BB pairs, and
three were BC pairs.

The food intake given 8% AOV also had only one significant effect,
but it was species combination, not social condition (see Table 6A in
Appendix 11). Tukey's IBD test showed that BB pairs ate more given 8%
than cc pairs. g(3,42)=5.2, p .01, and that as pairs did not differ
on the average from either BB or GG pairs. Although the social inter-
ference effect was not significant, 30 pairs ate more in A than in T.

The significant pairs within groups effect. which accounted for 38;;
of the variance in the complete food AOV. suggests that differences in
food intake between pairs within groups were quite large.

Caloric Intake Results

Each pair's mean caloric intake per three days was calculated for
each within pair treatment. Each gm. of sucrose was counted as 3.85
calories . and each gm. of food as 0.147 calories. The resulting group

means are shown in Table 4. An AOV done on the caloric intake data

1?

indicated that the Solution X Social Condition and Solution X Cage
Size interactions were significant (see Table 7A in Appendix A), and
tests of simple main effects were carried out.

These tests indicated that caloric intake was higher in the A con-
dition than in the T condition whether the mice received 0% or 8%, E
(1.84)-48.4 for 0%, g(1.84)-15.8 for 8%. p .01. Given 8%, 34 of the
48 pairs consumed more calories in A than in T; five BB pairs, five K}
paris, and four CG pairs did not show this effect. The corresponding
figures for 0% rave already been given in the food intake results. since
in this case. the only source of calories was the food.

Tests for simple main effects also showed that the mice consumed
more calories given 8% than when given 0% only in the T condition, _F_
(1.810443, 2 .01. In the A condition, the mice consumed as many
calories given 0% as given 8%.

Tests of the simple main effects of solution at the levels of cage
size revealed that mice housed in small cages consumed more calories
when given 8% than when given 0%, _F_‘(1,l+2)=19.8, p .01, however, solution
differences in caloric intake for mice housed in large cages were not
siglificant. Moreover. differences in caloric intake between groups
housed in large and snall cages were not significant.

Relating the caloric intake results to the fluid and food results
shows that in the A condition, the mice beiaved as Collier and Bolles
(1%8) might have predicted: they mintained a constant caloric intake
whether or not sucrose was present. Pairs housed in large cages also
maintained a constant caloric intake in the T condition with or without
8%, however. pairs housed in shell cages in T consumd more calories when
8% sucrose was available than when it was not. 0f the 211' pairs in the

18

Table h. Caloric intake and proportion of calories from food.

 

Species Combination x Cage Size Group

 

 

 

 

 

 

 

Social
BB BB K} m CG CG
Condition
Snell large Small large Snell large
Caloric Intake given 0% in cal. /3 days
Apart 129 119 111 125 106 123
Together 121 108 % 111$ 91 111
Caloric Intake given 8% in cal./3 days
Apart 137 117 115 119 112 119
Together 126 112 111 112 101 116
Proportion of Calories from Food given 8%
Apart .76“ .772 .621 .702 .602 .617
Togetmr e 759 e 8% e 8‘2 e 709 o 607 e 658

small cage group, 19 pairs showed this effect.

when 0% was given, a mean social interference effect of 11.4
calories per three days was detected, a 9.6% decrease from A to T. When
8% was given. the corresponding effect was only 6. 5 calories. a 5. :6
decrease. Mean food intake given 8% decreased 2.5 calories (4.1%), but
mean calories from 8% decreased by 11,0 calories, a 10.7% reduction. There-
fore, the social interference of caloric intake in the 8% condition was
due mostly to social interference of 8% drinking. This suggests that the
average proportion of the total calories obtained from food, given 8%.

was greater in the T condition than in the A condition.

19

Table it shows the proportion of total calories obtained from food
in the A and T conditions for each group, given 8%. An AOV performed
on these proportions (see Table 8A in Appendix A) indicated that both
social condition and species coabination effected the proportion of
calories obtained from food. however, the effect due to social condition
was a small one: it accounted for only one percent of the varaince in
the proportions. and only 29 of the 118 pairs demonstrated it.

Tukey's 16D test revealed that the BB pairs obtained a higher pro-
portion of their calories from food than either the as pairs, 3(3,u2)-=
3.9. 2 .05. or 06 pairs, 9.(3.42)-5.7. P, .01, while no significant
difference occured between the Bg and 06 mean proportions. This was
to be expected, in ivew of the species combination differences in 8%
drinking and eating given 8%. discussed previously.

However, it was not expected that the total caloric intakes for
the three species combinations would be essentially equal. since the mean
body weights for pairs in the three conbinations varied considerably;
n.7, h7.5. and 56.9 em. for BB. m. and as pairs, respectively. An
AOV on caloric intake per gn. body weight indicated that differences
between the species combinations existed for this variable, £(2,l+2)-
30.2, p .01, and Tukey's HSD test revealed that BB pairs consumed more
calories per gm. body weight (2.9 calories/gm./3 days) than H} pairs (2.1+
cal./gm./3 days), 3(3.u2)-u.2, p .05, which in turn consumed more then
66 pairs (1.9 ca1./en./3 days). gem-3.5. p .05.

E Results

If social effects were different for K} pairs, consisting of indivi-

duals fron both subspecies, than for 06 or BB pairs. one would eXpect

the mean intakes for BC pairs in the T condition to differ from the

20

soon intakes for all other pairs in 'r. Schsffe's test (Kirk. 1968) was
used to test this contrast for 0% intake, 8% intake. food intake given
0%, food intake given 8%. and total calories given 8%. None of the diff-
erences between mean B} intake and mean honogenous pairs intake in T
proved significant. Hence. the effect of putting one member of each sub-
species together in one cage was substantially the same, as the affect of
putting two members of the same subspecies together in one cage. as far
as eating and drinking were concemed.

Three models of the relationship between intakes in the T and A
conditions were examined for 0%, 8%. food given 0%. food given 8%, and
total calories given 8% intakes. The models were:

1. The purelyadditiba model: T 'A +e whareT andA are

i i 1' i i
th

the T and A intakes of the 1 pair, and e1 is the error tern. This

model is called additive since it assumes that the sum of the A intakes
for the members of each pair is approximt'aly equal to the T intake for
the pair.

2. The additive model with a constant non-additive term: T =

i

A + K + e where K is a constant such that the expected value of the

i 1’
error term is zero.

3. The unrestricted regression model: T - C A + K + a where

i i i’
C and K are the usual regression coefficients. chosen to minimize the sum
of the squared error term.
‘8 might be eXpectad from the results presented previously, the
purely additive model accounted for a much lower proportion of the variance
in '1‘ than either of the other two models. except for food intake given

8% (see Table 5). The purely additive model seemed almost as

21

satisfactory as the other models for predicting this intake. For all
five intakes. the additive model with a constant non-additive term
accounted for only slightly less of the variance in T than the unre-
stricted regression model.

Table 5. Variance in T intake accounted for by three models.

 

 

 

 

 

 

 

 

 

Type of Intake
Food Food Calories
0% 8% Given Given Given
Model
0% 8% 8%
Proportion of Variance Accounted for
T
T1 '3 ‘1 + 61 052 070 0M 089 073
T1 -A1+K +81 .68 .85 .70 .90 .82

 

T1 -CA1+K +°i .72 .88 .70 .90 .83

 

 

 

 

 

 

EXPERIMENT II

Experiment. I indicated that in the ad libitum situation frequently
employed in sucrose preference studies. social interference seemed to be
operating. Zajonc's (1965) theory states that the dominant response in
any situation should be socially facilitated, and nondominant responses
should undergo social interference. Thus if eating and drinking were men-
dominant in the situation used in Experiment I, which casual observations
indicated to be the case, then this theory could account for the social
interference.

Furthermore, Allee's (1938) huddling for energy conservation
mechanism could also have produced the observed interference in eating
given 0% and in 8% drinking. since the mice were observed huddling during
the 12 hour light period when in the T condition. Experiment II was per-
formd to discover whether or not drinking 0% and 8% would be socially
facilitated when drinking was made the dominant response by means of a
deprivation schedule. Under these conditions, Zajonc's theory would
predict social facilitation of 0% and 8% drinking. 0n the other hand,
the energy conservation through huddling explanation would lead one to
predict social interference with 8% drinking and eating and no effect on

0% drinking.

METHOD

The subjects used in Experimnt II were the 12 pairs of deermice
tested in small cages in squads two and three of the first experiment.
In Experimnt II. they were tested in the same pairs in which they were
run in Experiment I. The'apparatus used as the same as that used in
the previous experiment except no large cages were used. The design of
the present experiment was the same as that of the earlier experiment
except that cage size was not a factor, and only four pairs of each
species combination were run.

Each pair received the same order of presentation of social con-
ditions it had received in Experiment I, and the procedure was the
same as that used before except that 0% and 8% were available to the
mice only for the last hour of each 24 hour day; this hour began at
approximately 9: 00 A.M. Food was again available 251. libitum; food
deprivation was not attempted since deermice adapt poorly to food
deprivation schedules.

The mice were run in two squads; the pairs from squad two of the
Experiment I were started on 25 March 1969. and the pairs from squad

three were started on 22 April 1969.

23

RESULTS

Observation of the mice during the hour when fluid was available
indicated that they generally spent about five minutes drinking. immed-
iately after the bottles were placed on the cages. The rest of the hour
was spent eating. grooming. or sleeping, with some additional drinking.
A strong tendency to eat after the initial drinking bout was noticed.
There was no doubt that drinking was the dominant response during the
first five minutes of the hour.

As Table 6 shows. social interference, not social facilitation
of drinking occured. although much less fluid, especially 8%. was drunk
here than in Experiment I. An AOV performed on the complete fluid data
(see Table at in Appendix a) indicated that the solution and social
condition effects were significant. However. only six of the 12 pairs
drankmore0%inAthaninT.and9pairsdrankmore8%inAthaninT.
Thus the social interference here was less convincing than that found
in the first eXpariment. 0n the other hand. no evidence for social
facilitation of drinking was found. The solution effect indicated. of
course. that more 8% was drunk than 0%.

The pairs within species combination effect was also significant,
indicating differences between pairs in the species combination groups.
The significant replication effect seemed to be due to an increase in
fluid intake from the first to the second replication. probably caused
by adjustment to the deprivation schedule.

21+

25

Table 6. 0%, 8%, and food intakes: Experiment II

Solution X Species Combination Group

 

 

 

 

 

Social __ k
0% 0% 0% 8% 8% 8%
Condition
BB in ac BB Be as
Fluid Intake in ml./day
Apart— 6.1 . 5.8 at 6.9 6.5 7.5
Together 6.1 5.2 6.2 6.8 6.2 6.8
Food Intake in gm. /3 days
Apart 25.1 20.6 23.1» 22.8 18.5 29.8
Together 20.8 18.0 20.5 21.3 17.2 17.8

Mean three-day food intakes for the species combinations in each
treatment are also shown in Table 6. In conjunction with Tables 2 and
3 they indicate that less food was eaten in Experiment II than in
Experiment I, given 0%. An AOV performed on the food data revealed that,
just as in the first experiment, the mice ate more when given 0% than
when given 8%. and more in the A condition than in the T condition (see
Table 10A in Appendix A). Ten pairs ate more in A than in T when given
0%. and 11 pairs ate more in A than in T when given 8%. Again, differ-
ences between pairs within species combinations occured.

as pairs consumed less food and fluid than BB or 66 pairs in
Experiment II , although these differences were not significant. This
effect can be attributed to sampling error, since a similar effect

occured in biperiment I for these same pairs.

26

An AOV done on caloric intake per three days showed that the pairs
consumed more calories in A (100 ca1./3 days) than in T (89 cal./3 days),
E(1,9)-13.8, p .01. Eleven of 12 pairs showed this effect when given
8%. and . as untioned in the food results. 10 of the pairs showed the
effect when given 0%. As in the previous experiment. there were no
significant differences between mean caloric intakes for the species
combinations. An AOV carried out on three day caloric intakes per gm.
body weight shoved that differences between species combinations existed
for this variable, E(2,9)-8.3, 2 .01. Tukey's HSD test revealed that
BB pairs consumed more calories per gm. body weight (2.9 cal./gm/3 days)
than 96 pairs (2.0 ca1./gm./3 days), 1(319)'5.’+5. R .01. or 36 pairs
(2.2 ca1./gm./3 days), 1(3.9)=I+.a. B .05; the BG and ac pairs did not
differ significantly.

Another AOV indicated that the species combinations differed in the
proportion of calories obtained from food when 8% was given, {(2.9)-
15.7, p .05. Tukey's HSD test showed that BB pairs got a higher pro-
portion of their calories from food (.911) than cc pairs (.93). 1(3.9)=
Hull, 3 .05, but this difference seems ngeligible. Of course. the pro-
portion of total calories from food, given 8%. was much higher here than
in Experiment I, because of the restricted 8% intake in this study.

DISCUSS ID!

The major findings of Experiment I and II were (a) the observation
of social interference (not social facilitation) of eating and drinking.
and (b) the observation of differences in 0%, 8%, and food consumption
for the three species combinations.

M Interference

Social interference was observed in Eatperiment I for 8% drinking,
eating given 0%. total caloric intake given 0% and given 8%. and for 0%
drinking in BG and GG pairs. Only for food intake given 8% was the inter-
ference effect nonsignificant. and the purely additive model sufficient
to account for the ‘1‘ intakes of the pairs. These interference effects
cannot be attributed to variations in the amount of cage space per
animal. since differences between groups housed in large and snall cages
were not statistically reliable. Moreover. in Experiment II, social
interference was found for 8% drinking, eating given 0%, eating given 8%.
and total caloric consumption given 0% and given 8%.

Zajonc's (1965) idea that the dominant response in any situation
will be socially facilitated, while all other responses will show a social
decrement. can account for the social interference in Experiment I, if
one assures that eating and drinking were not dominant responses , and
therefore. showed a decrement when the dominant response was socially
facilitated. But it is difficult for this theory to explain why drink-
ing in Experiment II did not show social facilitation; there. drinking

was certainly the dominant response for the first few minutes after the

27

28

bottles were placed on the cages.

It is also hard for this theory to account for the lack of social
interference in eating. when 8% was given in Experiment I. Here. eating
was most likely less dominant when 8% was given than when 0% was given,
since less food was eaten given 8% than given 0%. Yet eating given 0%
showed social interference and eating given 8% did not. The absence of
social interference for 0% drinking in BB pairs presents a similar pro-
blem for Zajonc's theory.

0n the other hand, Allee's (1938) heat conservation mechanism can
explain the occurance of social interference in 8% drinking and food
consumption in both experiments. In both experiments. the subjects were
observed huddling during the light-on period of the day when in the T
condition. By this means. they could have reduced the amount of food
required to maintain stable body weights.

Similar effects have been noted in Peromyscus by other investiga-
tors; Kind (1968) stated that social huddling is one thermoregulatory
mechanism employed by this genus. Furthermore. in an earperiment with
Peromscus leucopus noveboracensis, Howard (1951) demonstrated the
importance of this nechanism for winter survival. The mice were housed
alone or in groups of two. three, or four and exposed to low temperatures
with limited food supplies. Those housed in groups of four survived
lonter than those housed alone or in snaller groups. According to
Howard. this result was due to reduced heat losses produced by huddling.
Howard (1951) also reported winter observations of aggregations of from
a few. up to a. dozen Mama maniculatus bairdii and g. leucopgg
under natural conditions. He attributed these agregations to the needs

to conserve food and to survive low temperatures.

.29

Social huddling. rather than it; libitun feeding might be one
reason Shelley (1965) found social interference of eating in young rats.
His group-housed subjects could huddle. but his isolated subjects could
not. In constrast. studies which found social facilitation of eating
not only used feeding schedules. as noted before; they prevented huddling
for all subjects (see Barlow, 1932. for sample) or allowed huddling
for all subjects (see Ross at Rose, 194%., for example), regardless of
whether the subjects were fed singly or in groups. Thus huddling could
not lave been a factor in these studies.

To test the lurpothesis that huddling could account for the social
interference of caloric intake found in the present experiments and in
Shelley's (1965) study. the present study could be repeated under high
temperature conditions. as in Betzlaff's study (cited by Allee. 1938).
Because the high temperature would minimize the advantage in heat con-
servation which subjects in T have compared to subjects in A. differences
in caloric intake should be minimized under these conditions. By.the
same reasoning. lowering the temperatures should increase social inter-
ference with caloric consumption to the extent that huddling produces
this effect.

One result that the huddling hypothesis does not explain is the
social interference of 0% drinking found for Bg and GG pairs in the first
experiment; 0% drinking does not contirbute to caloric intake, wry when.
should it decrease when huddling is possible? Perhaps reduced eating
leads to reduced drinking. The fact that for 00 pairs the amount of
social interference of 0% drinking in Experiment I correlated .61 with
the amount of social interference of eating, given 0%. supports this

idea. Moreover. Bartoshuk. (1971) cited a number of rat studies which

30

indicated that water deprivation produced reduced food intake and 11.20;
m. It might be that g. g. mills drink less when they eat loss.
while 2. 5. bairdii do not.

One might ask how other variables mnipulated in this study might
have influenced the huddling belavior of the subjects. In this regard.
the following variables should be enmined: (a) cage size. (2) subspecies.
(2) solution. and (g) pair age.

It seems that a smaller cage might increase the probability of
huddling by reducing the opportunity for other competing activity. If
this were the case. one would expect to find greater social inhibition
of caloric intake for snall-cage subjects than for large-cage subjects.
assuming that huddling produced such inhibition. However. no significant
cage size min effect or Cage Size K Social Condition interaction was
found for caloric intake . so no support for the above reasoning can be
found.

Similarly. one might expect that the larger mice in the experiments
would show greater savings in caloric intake than the smaller mice. when
the A and T conditions were compared. This follows from the facts that
the surface area reduction for two huddling nice . as compared with the
same mice not huddling. is proportionate to the total surface area of
the mice. and heat loss is proportionate to surface area. Again. no
support was found for this idea. since the larger Ellis showed no more
social inhibition of caloric intake than the smaller bairdii. In add-
ition. within groups. weight of the pair usually correlated negatively
with the amount of social inhibition. For BB. K}. and 00 the correlations
of weight with the mgnitude of inhibition of caloric intake were -.12.

-.26. and -.17 for 8%. and 546. .10. and null for 0%.

31

On the other land. it nkes no intuitive sense to assume that the
nice would vary their huddling behavior when given 8% instead of water to
drink. Thus it is not surprising to find that social inhibition of
caloric intake occured for both solutions. But it is surprising to find
that in the T condition the subjects consumed more calories given 8%
than given 0%. If huddling is to account for this affect. one must assume
that the nice huddled less when given 8% than when given 0%. It is even
less reasonable to attribute this effect to a liking for auger solution.
since in the A condition no similar increase in caloric intake when given
8% was detected. Further research is required to clarify this issue.

Finally. age did not seen to have a clear-cut effect on huddling.

As stated earlier. the species combinations did not differ in the amount
of social inhibition of caloric intake observed. although the genie
were older on the average than the bairdii. Within pair types. age did
appear related to the amount of social inhibition of caloric intake
observed. For BB pairs. caloric-intake social inhibition was negatively
correlated with pair age given 0% (-.30) and 8% (-.3+). but for 00 pairs.
age correlated positively (.37 and .62 for 0% and 8%) with these varia-
bles. Thus the younger bairdii and the older gracilis pairs showed more
social inhibition than the older bairdii and the younger gracilis pairs.
This suggests that younger bairdii and older milis pairs say have
huddled more frequently than other pairs. The results for as pairs are
difficult to interpret. since when given 8%. age and inhibition correlated
as with the BB pairs (506). while when given 0% their pair ages correl-
ated like those of the as pairs with social inhibition of caloric con-

smtion (.31).

32

It is obvious from the above analysis that some effects which the
huddling explanation of social inhibition of caloric consumption might
lead one to expect were not detected. while other effects were detected
which are not easily understood in term of huddling. In addition.
huddling alone cannot account for the observed social facilitation of
caloric intake for a few pairs of deermice. or for the greater sensi-
tivity of 8% intake to social condition changes contrasted with that of
food intake given 8%. Thus it is probable that factors other than
huddling. such as those discussed by Zajonc were also influencing caloric
consumption in these studies.

Species Combination Differences

In Experiment I. differences among the species combinations were
found for 0% and 8% drinking. and for eating when 8% was available. For
0% drinking. mean intakes for the species combinations did not differ
significantly in the T condition. For the A condition. and for 8%
drinking. BB pairs drank less than 60 pairs and H} pairs fell in between.

0n the other hand. BB pairs ate more food given 8% than 06 pairs
with m pairs again in the middle. Collier and Bolles' (1968) idea of
treating sucrose preference experiments as diet selection experiments
seems useful in summrizing these 8% eating and drinking results: 2. 1?.-
bairdii selected a diet with more food and less sucrose than did 3. g.
ggcilis. This result tends to support Drickamer's (1972) finding that
2. g. bairdii were conservative in sampling unfamiliar foods. when given
a choice between new and old foods.

Species combination differences were also found for caloric intake
per gm. body weight in both experiments; BB pairs consumed more calories

per gm. body weight than 60 pairs with as pairs again in between.

33

Although these differences between g. 3. bairdii and g. g. milis
were observed. they were almost all differences in the amount of son
behavior exhibited by both subspecies. The sole exception was the social
interference of 0% drinking which occured for ac and 06 pairs but not for
BB pairs. It seems reasonable to assume that the mine showed this
effect in mixed and homogeneous pairs. but the bairdii did not.

Finally. the beinvior of the E: pairs can be more economically
accounted for in terms of a sort of average of the behaviors of the BB
and 00 pairs. Mixing the subspecies did not mm to add any new dimen-
sions to the variables stud.ied here.

Limitations 3f the Results

These experiments have demonstrated that social interference effects
occur for eating and drinking in deermice in an experimental sitmtion
often used for sucrose preference testing. The lagnitude of the inter-
ference effects does not seem large enough to preclude group preference
testing using these subjects and a similar procedure under similar tem-
perature conditions. However. caution should be exercised in drawing
conclusions from such group data about the behavior of isolated animals.
and no: 2258.2- '

No specific mechanism or mechanisms have been shown to account for
the social interference effects that were observed. Social huddling my
be one factor that is envolved. These experiments present only a crude
picture cf the social effects occuring in the test situation. For instance.
no information about the behavior of the individual minis in the
together condition was obtained. and such informtion is certainly a

necessary prerequisite for understanding what as going on.

APPHDICEE

APMDDK A: ANALYSE 0F VARIANCE

3“

 

 

Table 1A. Complete fluid intake analysis of variance: Experiment I.
Source 9: LE r:
Solution (A) 1 199920.09 11118.8“
Social Condition (B) 1 2077.17 67.0”
A x B 1 877.63 “1.0"
Species Combination (c) 2 7509.62 22.5”
A x c 2 4740.19 10.6%
B x c 2 149.56 n.82*'
A x B x c 2 7.03 1
Cage Sise (D) 1 1705.01; 2.5
A x D 1 2505.tm 5.6*
B x D 1 26.92 1
c x n 2 881.69 1.3
A x B x n 1 32.91 1.5
A x c x n 2 453.88 1.0
nxcxn 2 4m” 13
A x a x c x D 2 5.00 1
Pairs within c x D (E) #2 667.87 18.3H
A x a 42 “5.49 12.2"
B x E 42 31.02 1
A x B x E #2 21.39 1
Replications within
A X B X E 1% 36.16 3.4-!”
Days within Replications 768 10.59
Total 1151

35

Table 2A. 0% analysis of variance: Experiment I.

 

 

 

 

30“” 9-1 2!: .13
Social Condition (A) 1 21.28 #23"
Species Combination (B) 2 26.71 2.8
A x B 2 9.07 18.3”
Case Size (0) 1 6.10 1
A x C 1 .03 1
n x C 2 21.61 2.3
A X B X C 2 1.39 2.8
Pairs within B x c 42 9.57
A 1 Pairs 02 .50
Total 95
mi ‘
Table 3A . 8% analysis of variance: Experiment I.

Source .4: 8 LB. .1:
Social Condition (A) ;— 1 455.88 51.2%»
Species Combination (B) 2 2061.96 12.0%»
A x B 2 14.56 1.6
Cage sine (c) 1 660.16 3.8
A x c 1 7.82 1
B x C 2 220.54 1.3
A1310 2 an 1
Pairs within a x C ' 02 172.52
A 1 Pairs 42 8.90
Total fi

 

H2 .01

%

Table 4A. Complete food intake analysis of variances Experiment I.

._____.___ A_‘

 

 

“E“ 2: E; E
Solution (A) A 1 1 5150.21 530.9“
Social Condition (B) 1 263.51 30.8**
A x B 1 76.59 10.9**
Species Combination (C) 2 609.89 4.2*
A x c 2 110.36 11.4**
B x C 2 1.13 1
Axnxc 2 an 1
“@SHeO) 1 1%58 1
A10 1 1&W L9
3x0 1 LW 1
C x D 2 276.51 1.9
Axexn 1 L% 1
Bxcxn 2 231 1
A x C x.n 2 13.49 1.4
Axexcxn 2 &W Li
Pairs within C x D (a) 42 143.41 22.5**
A x r 42 9.70 1.5*
B x B 42 8.5“ 1.3
AXBXE M mm L1
Replications within
AXBXE 1% 6%

Tan $3

37

Table 5A. Food given 0% analysis of variance: Experiment I.

in

 

 

 

 

 

“ Tm ‘— if 715. I:
Stein). Condition (A) 1 156.06 36.0H
Species Combination (B) 2 50.42 1.3
A x a 2 1.26 1
Cage Size (0) 1 53.70 1.4
Axc 1 .m 1
B x C 2 101.56 2.6
A x B x C 2 .46 1
Pairs within B x c 42 38.81
A 1: Pairs 42 4,311
Total __ 9.1
as 2 .01
Table 6A . Food given 8% analysis of variance 1 Experiment I.

"source E: 24.8; 1:
8:61; Condition (A) —_ 1 7.40 3.1
Species Combination (B) 2 283. 51 7.21”r
A x B 2 .95 1
Case Size (0) 1 16.13 1
A x C 1 6.03 2.5
BIC 2 5mm 13
A x B X c 2 6.24 2.6
Pairs within B x C 42 39.55
A 21 Pairs 42 2.40
Total 95

HR .01

38

 

 

 

Table 7A . Caloric intake analysis of variance 1 Experiment I.

Source if. .113; E
Solution (A) 1 449.60 7.6“
Social Condition (B) 1 3848.09 44.5H
A X B 1 286.90 6.8*
Species Combination (C) 2 2130.17 1.3
A x C 2 7.67 1
B X C 2 7.16 1
A X B X C 2 28.83 1
C886 3129 (D) 1 363.49 1
A x D 1 743.06 12.6H
B X D 1 31.116 1
c x D 2 3639.5? 2.2
A X B X D 1 39:11 1
A X C X D 2 74.11 1.2
B X C X D 2 25.82 1
AXBXCXD 2 72.44 1.?
Pairs within 0 x D (E) 42 1648.71
A x a 42 59.20
B x r 42 86.50
A X B X E ’42 142.22
Total 191
we 2 .01

39

Table 8A. Proportion of total calories from food given 8% analysis
of variance: lxperinnt I.

 

30“” if. 14.8.. 1".
Social Conditions (A) 1 .006952 9. 6"
Species Combination (n) 2 .200573 15.7"
A x n 2 .000245 1
Case sue (a) 1 .048875 3.8
A x C 1 .002465 3.4
a x c 2 .005190 1
A x a x C 2 .001766 2.4
Pairs within B x C 42 .012753
A x Pairs 42 .000723
Total 95

 

40

 

 

 

Table S. Fluid intake analysis of variance: Experiment II.

Some A «1.: in: .1:
Solution (A) 1 £14.97 32.3"
Social Condition (B) 1 7.03 6.8*
A X B 1 .39 1
Species Combination (C) 2 15.79 1.3
A X c 2 .15 1
B X c 2 1.32 1.3
A X B x c 2 .65 1
Paire within c (D) 9 12.15 9.6H
A X D 9 1.39 1.1
B X D 9 1.0” 1
A X B X D 9 1.01 1
leplicatione within
A X B X D ’48 1.27 2.0“
Days within Replicationa 192 .63
Total 28?
*«l 2 .01

Table 1m. Food intake analysis of variances

Experimnt II.

 

Source if. is 1:
Solution (A) 1 79.21 22.1H
Social Condition (B) 1 140.65 13.4**
A x B 1 16.17 2.5
Species Combination (C) 2 124.61 1.6
A X C 2 11.84 3.2
B X C 2 1.7“ 1
A X B X C 2 1.83 1
Pairs within C (D) 9 78.4? 10.1?”
A x n 9 3.58 1
an 9 1mw 1A
Axnxn 9 aw 1
Replicat ions within
A X B X D 118 7.52
Total 95

 

I"! 2 .01

APPENDIX B: PM HPERDIENI‘S

1+2

APPUDH Bi PILOT mums

A pilot study was done to investigate the possibility of social
facilitation of drinking in Peromcus miculatus (deer-ice). Two
subspecies, g. g. bairdii and g. g. Escilis were used. The six subjects
were pamd off so that one pair consisted of two bairdii (BB). one of
two gggcilis (CG). and one of a gaggilis and.a bairdii (BC). The subjects
were tested in three different housing conditions: (a) apart in snail
cages (As). (1:) together in large cages (TL). and (3) together in large
cages with a hardware-cloth barrier down the Biddle of the cage to sep-
arate the nice (Barrier). In the together conditions. only the two
nenbers of a pair were housed.in one cage. not all six mice. In each
housing condition. the nice were first given water for three days. than
given an 8% sucrose solution for three days. Food was available ad
libitun in all conditions.

In a second study. three other pairs of nice. one pair of each
species combination. were tested in the sane three housing conditions.
They were given 8% sucrose solution in two other conditions as well: (a)
apart in large cages (AL). and (2) together in suall cages (TS).

In both studies. intakes in the apart conditions were summed and
compared with intakes in the together conditions for’each pair of mice.
Mean intakes for the different species combinations in the A8. TL. and
Barrier conditions are presented.in.Table 13 for the nice in both pilot
studies.

Comparing 8% sucrose intakes for the apart and together conditions
makes it clear that interference of drinking occured.in the together

condition. It is not clear whether this was due to the presence of two

“3

Table 1B. Liquid intakes for six pairs in three housing cmditionea
Pilot studies.

Species Combination
Housing Condition

 

BB K: GG

M Sucrose Solution in m1./day

Together Large Cage 13.6 30.1} “3.6
Apart Snell Cage 18.“ 38.8 49.0

Together Large with Barrier 16.4 39.6 57.6

water in m1./day

 

Together Lei-‘88 Case 8.6 12.6 12.0
Apart Snail Cage 8.6 8.6 12.8
Together Large with Barrier 9.0 12.0 1114+

nice in the same cage. or to the difference in cage size. however. The
Barrier condition produced facilitation of 8% sucrose drinking for 66
pairs but not for BB or m pairs. An analysis of variance done on the TL
and A8 conditions for the 8% intakes showed that the effects of housing
conditions were significant. £(1,3)-d6.8. p .05. as were the effects
of species combination. E(2.3)-18.5. p .05. The interaction of these
factors was not significant. This indicated that the housing condition
effect was the same for each species combination.

Hater intake was effected much less by the changes in housing
conditions. No consistent facilitation or interference effect can be
found. An analysis of variance done on the TL and AS data showed that
the housing condition effect was significant. _F_'_(1,3)-10.6. p .05. but

in.

so was the Housing Condition X Species Combination interaction. {(2.3)-
20.15. p .05. Furthermore. the interaction accounted for 18% of the
total variance. while the housing effect accounted for only 5%. It can
be seen from Table 1B that the m pairs were causing the lame interaction
by drinking more in TL than in AS. Possibly this effect was linked to
the large amount of fighting observed when the m pairs were first put
together in the same cage.

Table 2Bgivesthe mean8%intakes foreachpairofnice inthe
second study for four housing conditions. It is evident that all three
pairs drank less in the TL condition than in the AL calditicn. Similarly.
boththe BBandmpairsdranklessintheTSconditionthanintheAS
condition. Only the 06 pair drank more in the TS condition than in the
AS condition. hence a social interference effect was obtained in five
of six possible comparisons.

Moreover. comparing the means of the TL and AL cmditions to those
of the TS and AS conditions indicates that the nice drank less in large
cages than in small cages in four out of six cases when the social
condition did not change. Thus it seem likely that the interferince
effect in the Table 1B data. was due to changes in both cage size and
social condition.

The data in Table 1B Show a large difference in mean 8% sucrose
intake between the BB and 06 pairs; their water intakes can be seen to
be much more closely clustered. Species combination accounted for about
80%ofthevariance forthe 8%data. andonlyabout 36%of the variance
for the water data. In both cases. pairs within species combinations
was the next most important source of variance. accounting for 7% and

27% of the variance for 8% sucrose and water respectively. Thus there

“5

Table 2B. 8% intakes for three pairs in four housing conditions:
Pilot studies.

Species Combination
Housing Condition

 

 

BB E} 66
Together in Large Cage 12.“ 28.0 50.8
Apart large Cage 16.6 “1.8 56.6
Together Smll Cage 16.6 36.8 61.“
Apart Snell Cage 20.2 38.8 53.8

 

Note. - The above intakes are in ml./day.

appears to be an intersubspecific difference in drinking behavior for
these mice. as well as differences between particular pairs.

LIST (1“ MOE

LISTCFWCE

Allee. H. C. The Social life of animals. London: William Heinemn.
1938.

Bartoshuk. A. x. Motivation. In J. v. Kling a. L. A. Riggs (Eds.).
woodworth and Schlosberg's experimental chol (3rd ed. ).
New York: Holt. Rinehart and Finston. 1E1.

Carpenter. J. A. A comparison of stimulus presentation procedures in
taste-preference experiments. Journal of Co tive and
2113301051081 Fathom. 1958. 21. 351-335.

Collier. G.. A: Belles. R. Some determinants of intake of sucrose

solutions. Journal of Comparative and Pgsiological Psycholog.
1%80 fit 37g:%§e

Crawford. M. P. The social psychology of the vertebrates. Psychological
Bulletin. 1939. 2. “07.1%.

Drickamer. L. C. Experience and selection behavior in the food habits
of peronvscus: Use of olfaction. Behaviour. 1972 . fl. 269-287.

 

 

Harlow. H. F. Social facilitation of feeding in the albino rat.
Journal of Genetig_Psxcholog. 1932. fl. 211-221.

Howard. H. E. Relation between low temperatures and available food to
survival of small rodents. Journal of Mammollggy. 1951. Z, 300-
312.

Jacobs. H. L.. 3: Scott. M. L. Factors mediating food and liquid intake
in chickens: I. Studies on preference for sucrose or sacchrine
solutions. PfllLrLScience. 1957. $0 8-15.

James. W. T. Social facilitation of eating behavior in puppies after
satiation. Journal of Comparative and Pagiological PsycholggzI
1953. 26;. #277125—

James. H. T.. a Cannon. D. J. Variation in social facilitation of
eating behavior in puppies. Journal of Genetic Psycholofl. 1955.
81! 225.2280

James. H. T.. 8: Gilbert. T. F. The effect of social facilitation on
food intake of puppies fed separately and together for the first
90 days of life. British Journal of Animl Behaviour. 1955. i_1_1_,
131-133.

“6

4?

Jams. H. T. The developnent of social facilitation of eating in
puppies. Journal of Genetic szgholog. 1960. %. 123-127.

King. J. A. Psychology. In J. A. King (BIL). Biology of perouscus
(rodentia). Amrican Society of Mammologists. 1968.

Kirk. R. E. Eagerinental design Procedures for the behavioral
sciences. Belmont. California: Brooks7Cole. 1968.

I
Latane. B.. a: Cappel. H. The effects of togetherness on heart rate
in rats. Psychononic Science. 1972. _2_9. 177-179.

Levine, R. L. Stability of sucrose drinking curves in genetically
heterogeneous nice. Psychononic Science. 1968. 13. 11-12.

Pearson. O. P. The rate of metabolism of some small mammals. Ecolgg.
19"?! 2&0 127-1“.50

Retzlaff. E. Personal commieation. Cited by H. C. Allee. [he
social life of animls. London: Villiam Heinemn. 1938. p.

8'3".

Ross. 8.. J: Ross. J. G. Social facilitation of feeding behavior in
dogs: I. Group and solitary feeding. Journal of Genetic

Pereholgg. 19w. 23. 97-108. (a)

Ross. 8.. at Ross. J.G. Social facilitation of feeding behavior in
dogs: II. Feeding after satiation. Journal of Genetic

Pszgholw. 19+9. 23. 293-3ou. (b)

Scott. J. P. Social facilitation and alleloinimetic behavior. In
E. c. Simmol, a. A. Hoppe. a: c. A. Milton (Eds.). Social
facilitation and imitative behavior. Boston: Allyn and Bacon.
1968.

Shelley. H. P. Eating behavior: Social facilitation or social
inhibition? lisychonomic Science. 1965. 2, 521-522.

 

Tolmn. C. H. Social Facilitation of feeding behavior in domestic
chicks. Animl Behavior. 1965. 1.2. 134-1h2.

Tolmn. C. it. Emotional behavior and social facilitation of feeding
lbegt‘iavior in domestic chicks. 131ml Behaviour. 1965. 12. 1493-

Tolnn. C. 3.. 8: wilson. G. F. Social feeding in domestic chicks.
gig-:1 Behaviour. 1965. 12. 131+-1lb2.

#8

0: fl
Vetulani. T. Untersuchungen uber das Hachstun der Saugetiere in
Abhlngigkeit von dar Anzahl Zusannengehaltener Tiers. I. Teil:
Beobachtungen an Hansen. Biological Genetics. 1931. z. 71-%.
- Cited by H. C. Allee. The social life of annals. London: Hillian

Heinenn. 1938. p. 31.

Young. P. T. Studies of food preference . appetite and dietary habit:
II. Group self-selection nintenance as a method in the study of
food preferences. Journal of Cogparative Psychom. 19M. 22.

371-391 .

Young. P. T. Studies of food preference. appetite and dietary habit:
IV. The baJance between hunger and thirst. Journal 9: gm-

tive Psycholgg. 1945. 28_. 135-174.

Young. P. T. Studies of food preference . appetite and dietary habit:
VI. ihbit. palstability and the diet as factors regulating the
selection of food by the rat. Journal or; Comtive Psycholgg.
1%. 22' 139-1760

Young. P. T. Studies of food preference. appetite and dietary habit:
VII. Palatability in relation to learning and performance. Journal

of Comrative g; Pusiolggical Pszcholgg. 19W. 22. 37-72.
Zajonc. R. B. Social facilitation in cockroaches. In E. C. Sinnel.

B. A. Hoppe. a: G. A. Milton (£33.). Social facilitation and
imitative behavior. Boston: Allyn and Bacon. .

Zajonc. R. B. Social facilitation. Science. 1965. m2. 269-274.

nICHIan STATE UNIV. LraRnaIEs
I“WIN‘IIIHINIWIHIWWWWWNW!1|
31293010

060808