QUANTITATION OF WATER DEPRIVATION
RESPONSES IN :MERIONES UNGUICULATUS '

Thesis for the Degree of M. S.
MICHTGAN STATE UNIVERSITY
DAVID L. NORTON
1970

     
   
   

University

mi,
I

Mia’s—i

;1 ? nah-Sine av
HIM & SIIII‘S'
BIIUK BINDER? INC.

 

 

 

 

 

ABSTRACT

QUANTITATION OF WATER DEPRIVATION RESPONSES
IN MERIONES UNGUICULATUS

 

BY

David L. Norton

Studies were undertaken to determine the effect
of prolonged water deprivation on the survival and metab-

olism of Meriones unguiculatus. Survival of this desert

 

species on a dry diet varying in protein content followed
the generalization that inability to conserve water when
urea must be excreted limits longevity. On a diet con-
taining 20% protein (Mouse Breeder Blox), mean survival
was only 5 weeks. On a 9% protein (barley) diet, mean
survival was in excess of 28 weeks yet the terminal whole
body water content (64% of body weight) was the same for
both groups. The conclusion was made that, although
water deprived animals were in negative water balance,
the proportion of water to protein remained essentially
the same as the animals lost weight and was unaffected by
dietary composition. The water content of ad libitum
control animals (59.4%) under constant illumination was

significantly different from that of "normal" controls.

David L. Norton

The reasons are as yet, uncertain, but a possible endo-
crine effect has been suggested.

The effect of water deprivation on oxygen consump-
tion has not heretofore been adequately investigated.
Fertig and Edmonds (1969) demonstrated a fall in oxygen
consumption for water deprived house mice, but these in-
vestigators failed to compare experimental and control
patterns. In the present study, it has been shown that,
aside from merely lowering oxygen consumption, water
deprivation causes a "shift" in the system relating
oxygen consumption (cc/day) to body weight (g). Quanti-
tation of the response resulted in two significantly
displaced regression lines representing control and water
deprived animals. Simple prediction equations for oxygen
consumption are, therefore, inadequate if the water status

of the animal is not considered.

.‘i .

 

 

 

QUANTITATION OF WATER DEPRIVATION RESPONSES

IN MERIONES UNGUICULATUS

 

BY

David L. Norton

A THESIS

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

MASTER OF SCIENCE

Department of Physiology

1970

Dedicated to my parents, Frank and Jessie
Norton, whose moral and financial
support made this research

possible.

ii

ACKNOWLEDGMENTS

The author wishes to express his sincere thanks
to Dr. Lester F. Wolterink for his encouragement and
guidance during the course of this study.

Appreciation is also extended to Dr. W. D. Collings
and Dr. W. L. Frantz for their advice and assistance in
this research effort.

Special thanks go to Mrs. Lucile Jolly for her

assistance in data tabulation.

iii

TABLE OF CONTENTS

Page
INTRODUCTION 0 O O O O O O O O O O O 0 O O O O O O O 1
REVIEW OF THE LITERATURE . . . . . . . . . . . . . . 6

Studies in Ad Libitum Water Intake . . . . . . . 8

Toleration of Water Deprivation in the
Order Rodentia . . . . . . . . . . . . . . . . 12

Dry Rodents . . . . . . . . . . . . . . . . . 12
Wet Rodents . . . . . . . . . . . . . . . . . 15

Studies in Body Water . . . . . . . . . . . . . . 17
Oxygen Consumption in Relation to Body Size . . . 18

Oxygen Consumption During Water Deprivation . . . l8
METHODS O O O O O I O O O O O O O O O O O O I O O O 22

Experimental Rationale . . . . . . . . . . . . . 22
Experimental Design . . . . . . . . . . . . . . . 22
Determination of Body Water . . . . . . . . . . . 24

Calculations and Assumptions in the
Determination of Oxygen Consumption . . . . . 24

Estimation of Body Weight Changes . . . . . . . . 25
Determination of Blood Osmolarity . . . . . . . . 28

Statistical Considerations . . . . . . . . . . . 29
RESULTS I O O O I O O O O O O O O I O I O O O O O O 30

Water and Food Consumption . . . . . . . . . . . 31

Quantitation of Ad Libitum Water Consumption:
The Heterogonic Equation . . . . . . . . . . . 34

Effects of a Deficient Water Supply on Body
Weight and Survival . . . . . . . . . . . . . 38

Quantitation of the Survival Response to
Water Deprivation . . . . . . . . . . . . . . 40

iv

Body Water in the Water Deprived Animal in
Light-Dark and Light-Light Environments . .

Effect of Water Deprivation on Mean Oxygen
Consumption. . . . . . . . . . . . . . . . .

Effect of Water Deprivation on the Quantitation
of the Oxygen Consumption Response . . . . .

Plasma Osmolarity Changes During Water
Deprivation in Animals on High and
Low Protein Diets O O O O O O O O O O O O 0
DISCUSSION 0 O O O O O O O O O O I O O O O O O O 0
SUMMARY AND CONCLUSIONS 0 O O O O O O O O I O O 0
LITERATURE CITED . . . . . . . . . . . . . . . . .

APPENDICES O O O O O O O O O O O O O O O O O O O O

Page

41

47

SO

53
56
65
67
71

LIST OF TABLES

Table
1. Ad Libitum Water Consumption in Some Rodents
2. Oxygen Consumption in the Order Rodentia . .
3. Oxygen Consumption when Different Foodstuffs
Are Oxidized . . . . . . . . . . . . . . .
4. Gerbil Survival without Water . . . . . . .
5. Gerbil Survival without Food . . . . . . . .
6. Body Water after Starvation . . . . . . . .
7. Plasma Osmolarity During Water Deprivation .

Vi

Page

10
20

25
39
39
46
54

LIST OF FIGURES

Figure Page

1. Food Consumption During Water Deprivation . . . 32

2. Quantitation of Ad Libitum Water Intake:
Relation of Water Intake to Mean Body
weight 0 O O O O I O O I I O O O I O O I O O O 35

3. Mortality Rates of LD and LL Animals on
Water Deprivation . . . . . . . . . . . . . . 41

4. Body Water in the Water Deprived Animal

on High and Low Protein Diets . . . . . . . . 43
5. Comparison of Mean Oxygen Consumption

During Water Deprivation . . . . . . . . . . . 48

6. Effect of Water Deprivation on the Quantita-
tion of Oxygen Consumption . . . . . . . . . . 51

vii

INTRODUCTION

The success of a species ability to survive in a
desert environment depends upon its capacity to solve the
problems of food, water, and body temperature maintenance.
The factors which determine the availability of the first
two commodities vary in different habitats and, as Chew
(1951) has so aptly phrased it, can "influence the dis-
tribution and abundance of mammals" on earth. It is not
a function of this dissertation to describe the ecological
patterns of the world's deserts but only to examine some
of the physiological mechanisms which arise from such an
existance and which are necessary for survival.

The problem of food is one for which mammals have
no physiological defense other than tolerance of starva-
tion or semi-starvation. Food must be supplied by the
environment. Survival depends upon it not only because
food itself is needed for survival but also because it is
a source of water; be it free or metabolic. The question
which now arises is whether an animal can live in the
absence of the second essential, water, if food is avail-
able. The answer will depend upon the animal's ability
to balance water loss with oxidation or metabolic water

and with that free water which may exist in the food.

Water is lost through the feces and urine, and by evapora-
tion at the respiratory surface, while gain, in this case,
can only be had via the food. Daily fluctuations in loss
to gain or in gain to loss may exist but, in the long run,
the animal must be in equilibrium if it is to survive.
This investigation is meant to examine the tolerance to
water deprivation of that group of desert mammals belong-
ing to the order Rodentia and, in particular, a species
of the genus Meriones, one of several in a large group of
desert rodents commonly referred to as the desert gerbil.
Schmidt-Nielsen §t_al. (1948) have demonstrated
the ability of the kangaroo rat (Dipodomys merriami) to
exist on a diet of dry grain with no exogenous water
available. Other investigators (Bartholemew and MacMillen,
1961; Kirmiz, 1962) have reported this independence of

exogenous water in the kangaroo mouse, Microdipodops

 

pallidus, and the jerboa, Jaculus. In contrast, Bartholemew
and Hudson (1959) have shown that the antelope ground

squirrel, Citellus leucurus, another desert rodent, has

 

need of succulent plants in its diet. This dissertation

attempts to classify the species Meriones unguiculatus,

 

as either a "wet" or a "dry" desert rodent.

The third stress of the desert is that of heat
and, as Schmidt-Nielsen (1964) points out, there are
basically three ways in which an animal can meet it;

evasion, tolerance, and thermoregulation. Evasion is

generally accomplished by burrowing and by being nocturnal
in nature. Misonne (1959) has reported that temperatures
at a one meter depth below the surface usually vary within
a range of 20-30 degrees centigrade; which is well under
the surface temperature during the day. Consequently,
burrowing animals do not meet with the severe stress of
the desert heat that one might expect. Furthermore, most
of the desert rodents are nocturnal, coming out at night
to forage for food and remaining in burrows during the
day. Hence they are further alleviated from the stresses
imposed upon them by the severe desert heat such as the
need to thermoregulate and the conservation of body water.
To passively put up with the heat would involve an ex-
tremely high tolerance to increases in body temperature
and, as yet, there are no indications that desert mammals
are able to surpass the limits of heat tolerance found in
most other mammals. The third mechanism is thermoregula-
tion whereby the animal combats the heat by evaporation.
This is the costliest method since it infringes upon the
animal's precarious water balance and, in fact, poses a
threat to survival. Desert species of the type studied
here are not particularly tolerant of high temperatures,
nor do they use water for heat regulation. It is the

fact that most desert rodents are both fossorial and noc-
turnal that contributes to their capacity to withstand the
aridity and temperature of the environment in which they

live.

Gerbils are rodents of the family Cricetidae, sub-

family Gerbillinae. They are referred to by a number of

 

common names such as sand rats, desert rats, antelOpe
rats, and jirds. Their natural diet consists of dry seeds
and roots. They are distributed throughout Africa, Egypt,
the Middle East, Southern Russia, Central and Eastern

Asia, and India. The habitat of Meriones unguiculatus is

 

described by Rich (1968) as extending from Mongolia to
the northern sections of the Sinkiang, Shensi, Ordos, and
Shansi provinces of China. Since there are twelve genera
presently known, it is inadequate in reports such as this
to ascribe only the word "gerbil" to the species under
investigation. The name Mongolian gerbil, when presented

in the literature, usually refers to Meriones unguiculatus

 

and it is to this species that this report makes reference.
This research focuses on the environmental aspects
of the animal's physiology with emphasis on water conser-
vation and oxygen consumption. The problem of heat regu-
1ation in Meriones is not considered because, in their
natural habitat, they exhibit the common behavioral
characteristics described above as being typical of most
desert rodents and, hence, are not subjected to the
stresses of the desert heat. It only remains then to ex-
amine the problems imposed upon the animal by the desert

dryness.

Of particular interest to this research was the
ability to quantitate, or to describe in mathematical
form, certain physiological responses as they relate to
some independent variable.

The questions of interest were as follows:

1. Can ad libitum water consumption be success-
fully correlated to body weight in an interaspecific
study such as this?

2. Is Meriones unguiculatus as adaptable to water

 

deprivation as other desert species?

3. Does dietary composition have a significant
effect on survival during water deprivation?

4. What effect does water deprivation have on
mean oxygen consumption?

5. Can oxygen consumption be successfully quan-
titated in an intraspecific study in terms of cc/(g)
(days)?

6. What effect does water deprivation have on
percent body water and plasma osmolarity?

7. Are these six factors affected by an artifi-
cial environment such as constant illumination?

The particular parameters studied include food
and water intake, body weight changes during water depri-
vation, oxygen consumption, body water, survival time as
a function of dietary intake, and plasma changes during

water deprivation.

REVIEW OF THE LITERATURE

The quantitation of a physiological process in-
volves the establishment of a relation between a response,
Y, and an independent variable, X, most notably surface
areas, body weights, or survival times. The equations
used for the expression of such relationships have been
termed heterogonic (Adolph, 1949) since they reveal the
degree of disproportionality existing between the two
variables investigated. If, for example, the logarithms
of a response Y are a linear function of the logarithms
of X, then obviously the response must be proportional to
a given power function of x. The equation is usually

written in the form

Y = a X (2.1)

in which Y is the dependent variable and X the independent
variable. Taking the logarithms of both sides, the

equation becomes

log Y = log a + k log X (2.2)

which has the same general form as the linear equation

Y = a + k X (2.3)

Linear equation (2.3) represents a constant absolute in-
crease in Y for a constant absolute increase in X, whereas
logarithmic equation (2.2) represents a constant percent-
age increase for the two variables. Both equations are
fitted to data by the method of least squares, but the
procedure for equation (2.2) involves a logarithmic trans-
formation of the measured data into terms of log Y and
log X.

Brody (1964) has concluded that it is often more
rational to assume that a given percentage deviation has
about the same significance for a large as for a small
animal than it is to assume that a given absolute devia-
tion has approximately the same significance for a large
as for a small animal. For example, if a 100 g animal
looses 5 g of body weight it has lost only 5% of its
initial weight, but the same absolute loss in a 10 g
animal constitutes a 50% reduction in weight. As a re-
sult, the logarithmic equation (2.2) has a greater signi-
ficance in such physiological quantitation than has the
linear equation (2.3).

Controversy has arisen over the value of the
exponent, "k," when body weight is the independent vari-
able. Adolph (1949) suggested that the two-third power
of body weight might be a common and reasonable coeffi—
cient of prOportionality, whereas Klieber (1961) showed

a greater proportionality between certain variables when

the three-fourth power of body weight was used. Thus,
studies on the interspecific comparison of metabolic rates
in 26 groups of animals resulted in a regression coeffi-
cient of 0.756. Intraspecific comparisons, however, have
usually resulted in higher exponential values. Lee (1939)
reported a "k" value of 0.82 after metabolic studies on
rabbits. Similarly, Benedict (1938), in his regression
line for metabolic rates in mice, gave a "k" value of
0.89. Since animals are strikingly dissimilar in size
and geometric shape, it seems illogical to equate "k"
with a constant in dealing with intraspecific studies.

It is more rational to take X k as the reference base,

the value of "k" being determined on the basis of actual

data (Brody, 1964).

Studies in Ad Libitum Water Intake

 

The use of ad libitum water intake as an index of
water requirements involves the assumption that it is an
accurate index of the water intake to which the species
is adapted. In support of this assumption, successful
correlation between ad libitum water intake and habitat
aridity has been shown for a number of rodents; mice
originating in areas with more mesophytic vegetation drank
more water than those from primarily xerophytic regions
(Lindeborg, 1952). Schmidt-Nielsen (1964) has shown that

those species of rodents most able to survive during

water deprivation consume the least amount of ad libitum
water and generally come from the most extreme desert
habitats.

Lee (1963) found no correlation of water intake
with habitat aridity in studies between coastal and

desert Neotoma lepida. MacMillen and Lee (1967) noted

 

that the Australian desert rodents, Notomys alexis and

 

Notomys cervinus, which are generally independent of

 

drinking water under moderate temperatures and a diet of
carbohydrate-rich seeds, will drink water "greedily" in

the laboratory. The jerboa, Jaculus jaculus, from the

 

Sahara, has been observed to drink 4.3% of its body weight
in water per day (Schmidt-Nielsen, 1964; see Table 1).

It is apparent that the variables involved in ad libitum
water consumption are numerous. Hudson (1962) found in-
creased water consumption with rising ambient temperatures

in studies on the desert ground squirrel, Citellus leucu-

 

rus. Schmidt-Nielsen (1964) has reported that kangaroo

rats, Dipodomys merriami, which ordinarily can survive

 

water deprivation indefinitely, could not do so in rela-
tive humidities below 15%. Adolph (1943) showed that a
decrease in water consumption occurs when food is restricted
in studies on the laboratory rat. Williams (1959) has
presented evidence that ad libitum water intake is af-
fected by dietary composition. Alteration of dietary

protein in studies on the deer mouse was directly related

10

 

 

vmma .cmmawflzlupflfinom macs II II “Emanumfi mSEOUomflo
vmma .cmmeHZIDpHE£om mcoc nu In msaaflnnmm msaaflnumw
vwma .cmmHmHZIupHEnom ones In In mammmno mmcoflnmz
vmma .cmmHmAZIuGHEEUm om.v In In msHsOMn mDHDUMb
vmma .ammHmHZIDCH830m mm.m I: In manommsn m>800¢
mmma .uouowmm v.ma In o.o> .mm msumOHHUOmmz
mmma .muonwccwq m.om HH n.vm msoflsm>ammccwm .m mDDOHOHz
vmma .ammawHZIupflenom mm.aa II II mscflufinmo mMEoom
mcsum mane o.HH ma m.¢m msumasoflsmcs mmcofiuwz
mama .cOmcsm H.~H OH m.Hm maafimm msaouomno
mmma .muonmpcwq m.m om H.Hm Oaaflcuou msmoosma msommfionmm
mmma .mmq H.ma m II mowmma mEouomz
mmmH .qu m.vN OH 5.0mH mwmwvmflm mfiouowz
NmmH .GOwUSE v.ma NH w.mm mDHDODwH mDHHmuHU
omma .Houommm n.5H II o.oma .mm COUOEmHm
Amo\.u3
mocmummmm cowMMWMmWOU z pnmflmbmwmmz mmflommm
umumz

 

 

.m#G®©O~m 050m CH GOHUQESWCOU HMHMZ EDfiHQHQ UAN

.H GHQMB

11

to the amount of water drunk. It is possible, therefore,
that, although a general correlation between habitat
aridity and water intake may sometimes exist, the measure-
ments of ad libitum water consumption in the laboratory
may not directly reflect the "normal" water consumption
in nature. The natural environment itself is not usually
"constant," unless behavioral adaptations carefully regu-
late the selection of the ambient environment.

The water requirements of Meriones unguiculatus
appear to be intermediate between those of Dipodomys

merriami and Citellus leucurus. Although both M. unguicu-

 

latus and g. leucurus require free water for maintenance,
the former consume less water ad libitum (Winkelmann and
Getz, 1962).

Adolph (1949) formulated an equation by which
interspecific comparisons of ad libitum water intake
could be made on the basis of body weight. His equation
was heterogonic and expressed water intake as being pro-
portional to the 0.88 power of body weight. Lindeborg
(1952) found no meaningful correlation between water in-
take and body weight in mice from xeric and mesic habitats.
Using a modification of Adolph's equation, Hudson (1962)
compared the water consumption of 14 species of rodents
with limited success. The trend relating water intake to
differences in the habitat of the various species studied

was also obscure. Dipodomys merriami, an animal from an

 

12

extremely arid region, had a consumption equal to the
predicted value, whereas the cotton rat, Siqmodon 52.,

an animal from a moist habitat, had a water consumption
higher than would be expected on the basis of body weight.

Microtus pennsylvanicus and Peromyscus leucopus tornillo

 

 

showed consumptions that correlated nicely with their
habitats (Lindeborg, 1952), that is, mesic versus xeric,
but did not follow the predicted values (Hudson, 1962).
Lee (1963) has attributed this discrepancy to the hetero-
geneous conditions under which the data were collected.
Hudson stated that comparisons of ad libitum water con-
sumption on a relative weight basis are complicated by
such variables as humidity, temperature, moisture content
of the food and activity. These factors are often not
explicitly acknowledged in the literature. Water intake
for a number of species has been tabulated by Spector
(1956).

Toleration of Water Deprivation in
the Order Rodentia

 

 

"Dry" Rodents

 

The ability of rodents, particularly the desert
species, to survive during periods of water deprivation
has been extensively studied. Body weight changes,

length of survival, and dietary intake are reviewed.

13

Desert rodents of the family Heteromydae: Dipo-

 

domys gp,, Dipodomys merriami (kangaroo rats); and Perogna-

 

thus penicillatus pricei (pocket mice) were able to maintain

 

water balance on a diet of dry grain with no drinking water

(Schmidt-Nielsen, et a1., 1950). Dipodomys merriami sur-

 

vived for indefinite periods of time and were able to
maintain their body weights on air dried food without
access to drinking water when the diet consisted of
pearled barley or rolled oats. Survival was limited to
16 days when a high protein diet of soybeans was given
without water (Schmidt-Nielsen, 1964). Terminal weights
in these animals was 60% of initial. Bartholemew and
MacMillen (1961) have shown indefinite survival under

water deprivation for a third genus, Microdipodgps palli-

 

dus, the kangaroo mouse, when on a dry carbohydrate rich

diet. Pocket mice, Perognathus baileyi and Perognathus

 

 

penicillatus pricei, survived well without any moist food

 

and appeared to be even more independent of moisture than
the kangaroo rats (Schmidt-Nielsen, 1964).

A comparative study of the jerboa, Dipus aegyptius

 

(family Dipodidae) and white rat, Rattus E23! was reported

 

by Kirmiz (1962). On a dry grain diet (barley and wheat;
10% moisture) jerboas survived over a period of 1-3 years.
Weight loss followed 10 months of water deprivation. In
contrast, the white rats decreased food intake for the

first three days of water deprivation after which they

14

ceased to eat entirely. The finding for the rat was con-
sistant with that of Adolph (1943). Laboratory rats re-
fused to eat whenever some essential constituent of the
diet (such as water) was lacking. As a result, the body
weights of the rats diminished rapidly and survival was
only one week (Kirmiz, 1962). Schmidt-Nielsen, gt_§1.
(1948) reported a 21 day survival for water deprived rats,

Rattus norvegicus, accompanied by a 50% reduction in body

 

weight. The capacity of jerboas to live on a dry diet

was attributed to a reduced food consumption and metabo-
lism, whereas the white rats ceased to eat after the third
day on a dry diet. Total inanition as well as water fast-
ing were considered the lethal factors (Kirmiz, 1962).

Data for the subfamily Gerbillinae has been re-

 

ported by Burns (1956); Petter (1953); Schmidt-Nielsen

(1964). Gerbillus gerbillus, from Egypt, Meriones libycus

 

 

 

and Meriones crassus, both from the Sahara, lived well on

 

dry food and survived indefinitely, often with an increase
in weight. In a comparison of seven desert species,
Schmidt-Nielsen (1964) showed Gerbillus gerbillus to be
the most adaptive to water deprivation with Jaculus

jaculus and Acomys cahirinus the least adaptive. The

 

latter species suffered a 30-40% weight loss after three
weeks of water deprivation and are reported to consume

11.38% of their body weight daily in ad libitum water.

15

Final mention of the "dry" rodents must include a

recent study on the house mouse, Mus musculus (Fertig and

 

Edmonds, 1969). Mice kept on a dry grain diet maintained
themselves at full body weight for several months. When

a high protein diet was introduced, the mice consumed less
food, thereby subjecting themselves to a slow starvation.
The mice tolerated a temporary loss of body weight of
about 40%. High protein diets were lethal but urine con-

centrations often exceeded those of Dipodomys. The ability

 

of Mus musculus to survive on limited water intake ex-

 

ceeded that of Microtus (Chew and Hindegardner, 1957).
Independence of drinking water or succulent food is due
to an extreme ability to reduce urine water loss by form-
ing very hypertonic urine and by reducing evaporative
water loss by a decrease in oxygen consumption (Chew,

1961).

"Wet" Rodents

 

Studies in water deprivation have been done on
animals classified as "wet" rodents. Data on the North

American pack rat, Neotoma albigula, has been compiled

 

by the Schmidt-Nielsens, et a1. (1948) and compared to

the white rat, Rattus norvegicus. Like the kangaroo rat,

 

the pack rats needed no source of drinking water but could
not survive on air dried diets. Much of its moisture came

from succulent vegetation. When given only air dried

16

food, survival was only 4-9 days as compared to 15-21 days
in the white rat, also a "wet" rodent. The rate of weight
loss during water deprivation was similar for both animals
but the pack rats died after only a 30% reduction in
weight. White rats tolerated a 50% loss of weight.

Adolph (1943) found a 46% reduction in weight and a 6-15
day survival time for water deprived rats. Emphasis has
been placed on the seeming contradiction of a desert
species being less tolerant to water deprivation than a
relative of similar size with no special adaptation to a
desert existance.

Findings for the sand rat, Psammomys obesus

 

(Schmidt-Nielsen, 1964), and the carnivorous grasshopper

mouse, Onychomys torridus (Schmidt-Nielsen and Haines,

 

1962), were similar to those of the pack rat.

Related species, Neotoma lepida and Neotoma

 

fuscipes, have been studied by Lee (1963). Neither of

these species are able to maintain initial body weight or
to maintain a constant weight at a lower level when water
is withheld. Survival of water deprived wood rats ranges

from 2-16 days. Neotoma lepida experienced a 32.5% re-

 

duction in weight while Neotoma fuscipes lost a mean of

 

40.0% of their initial body weight. Animals which had
experienced partial dehydration, rehydrated, and were then
deprived of water, had survival times which were twice

those of unacclimated animals. Similar results were found

17

for Citellus leucurus (Hudson, 1962), but survival under

 

water deprivation reached a maximum level for all "wet"

rodents considered; 51 days.

Studies in Bodinater

 

Khalil and Abdil-Messeih (1954) reported a lower
water content in the tissues of desert animals than in
other animals. Sokolov (1966) refuted these findings and
found no such reduction in the tissues of desert rodents
when compared to the dog, rat, and man. Schmidt-Nielsen
gt_al. (1948) found the water content of kangaroo rats to
be 66.5% after 7 weeks of water deprivation on a diet of
pearled barley. Control animals averaged 67.2% after 54
days on fresh watermelon. Attempts at dehydration by
feeding the rats a diet of dry soybeans resulted in an
average body water content of 67.2% at the time of death.
Weight loss in these animals was 66% of initial. Nega-
tive water balance was achieved but the proportion of
water in the body remained the same as the animals lost
weight. This implied that the animals were not really
dessicated. Chew (1951) reported the same percent body
water (66%) for other small rodents on water deprivation.
Chew (1957) found a significant decrease in body water

in water deprived mice, Mus musculus.

 

18

Oxygen Consumption in Relation
to Body Size

 

 

The relation of metabolic rate to body size has
been reported by a number of investigators (Adolph, 1949;
Klieber, 1961; Brody, 1964). Interspecific comparisons
have shown that metabolic rate is most nearly proportional
to the three-fourth power of body weight or a regression
coefficient (in the heterogonic equation) of 0.756
(Klieber, 1961). Brody's analysis was shown in his
"mouse to elephant" curve.

Intraspecific comparisons of metabolic rate to
body size have yielded slightly higher coefficients; 0.89
for mice (Benedict, 1938), 0.82 for rabbits (Lee, 1939),

and 0.84 for dogs (Galvao, 1942). As a result, X R

has
been suggested as the reference base; the value of "k"
being determined on the basis of observed data (Brody,
‘1964).

Oxygen Consumption During
Water Deprivation

 

 

The ability of desert rodents to reduce metabolic
activity may be as important for water conservation as
for energy conservation (Bartholemew and MacMillen, 1961).
Schmidt-Nielsen (1964) has shown that a decrease in
evaporative water loss accompanies a decrease in oxygen
consumption. Since water conservation is essential to

the water deprived animal, a reduction in oxygen

19

consumption is therefore advantageous. Klieber (1961)
reported an inverse relationship between oxygen consump-
tion and days of starvation. Metabolism decreased with
increased time on starvation in laboratory rats. Fertig
and Edmonds (1969) have shown that water deprived house
mice on a lethal diet (high protein) entered a state of
torpor, as indicated by a reduced oxygen consumption, in
order to conserve energy. The reduction in oxygen con-
sumption was accompanied by a reduction in evaporative
water loss. Adolph (1943) found a similar reduction in
caloric output for water deprived rats.

A number of resting oxygen consumption values for

certain rodents has been compiled in Table 2.

20

IIIII. Aomaav
..Hm um .cmemAZIucflsaom

 

mm mm.m o.mm msHsomsE msz

Anomav .mmq mm mn.o In mmmHUmSM mEouomz

Ammmav .mmq am ms.o .. Aamummoov mended «souomz

Ammmac .mmq am mn.o .. Auummmuo mended msouomz

Ammmav .comesm om mo.a H.6m mausosma msaamuao

Ammmav .cOmzmo Hm om.a m.ma mausosma msaamuflo

Ammaav .comzmo mm om.a a.mm mscnucasmcmm mmsooomno

Asmmav .mmn a cmaaazomz mm Ha.a H.0OH mandamuommm masocomno

.aflMlmm .cmmHoHZInmwmmww mm ov.a H.ooa mnaanmuommm mssooomflo

Ammmflv .aOmgmo Hm o~.H n.4m nsmfiuums mssooomno

Insane .qu a cmaaflzomz mm as.m H.mm Aswannws masowomao

..AMIMM comeHz|u%meww mm vm.m H.mm HEMHMHmE mwaopomwa
U mmmummo Awmwwmwmw Amy

mocmummmm musumnmmEmB coaumESmcou pamflmz mmflommm

ucmflnfim mmmmxo com:

 

 

.mwpcmpom Hmpuo mnu ca coHumESmcoo cmmwxo

.N magma

21

 

AhmmHV .qu w cmaaflzomz
Ahmmav .mmq w cmHHHZOMz
Anomav .mmq a cmaaazomz
Anomav .aomcnnom
Ammmav .cOmcflnom

Aommav .20mnmmm

Acmmav

..Hm um .cmmamflzlupflfinom
Aommav

..Hm um .cmmHmHZIDwasnom
III 3mm:
..Hm pm .cmmamflzlupfiesom
Afloaav

.smaawzomz w 3mEmHonuumm
IIIII Aommav
..Hm um .cmmeHZIDUflfinom

IIIII AommHv
..Hm um .cmmamnz-ucasgom

mm
mm
mm
hm
mm
mm
mm
mm
mm
mm
mm

mm

mv.m
mm.H
«H.m
MH.H
mm.H
om.~
vH.m
mm.m
mm.H
om.H
wm.H

Hm.m

m.NH mflmcmmnsnmccmfiums mcwpmmmmq

>.vm mscfl>uwo mhfiouoz
m.mm mflxmam mmfiouoz
mealmn Sapfiamuwm msHHHQHmw
omlam mDuMHSOHsmcs mmcoflumz

o.m mfluonmmE m>EOMGOUs0H£uHmm

«.mm .mmm msg#mcmoumm
o.m~ msuflcano msommfionmm
H.mm msmnsm msumowno
«.ma mspflaamm mQOUomHGOHOAZ

H.moH Amuflnzv msowmm>uoc mspumm

«.mm Amuflczv msasomse ms:

METHODS

Experimental Rationale

 

Meriones' response to prolonged water deprivation
was observed over periods ranging from 20-40 days in
several experimental trials. The designs of all trials
were essentially the same although different parameters
were investigated after each experimental run. The abso-
lute values of the results obtained are of interest to
this research only to the extent that they have led to a
description or quantitation of a particular response.

For example, the values for ad libitum water consumption
are not as critical as the system which relates water con-
sumption to body weight. The results formulate the system
but the system alone describes the response.

With the exception of oxygen consumption, all
parameters were observed directly. Oxygen consumption
was found by calculation from the food intake and adjusted

for body weight changes.

Experimental Design

 

Adult, male and female Mongolian gerbils, Meriones
unguiculatus, were placed in individual cages containing

a granulated corn cob litter and starved for 12 days in

22

23

order to reduce body fat. A reduction in body fat was
necessary to eliminate individual variations in percent
body water inherent between animals. During the starva-
tion period, water was available ad libitum. The animals
were then transferred to clean cages, with no litter, and
divided into two uniform groups of twenty animals each.
One group was housed in a constant light laboratory while
the other group was kept in the original laboratory under
light-dark conditions (12 hours of light and 12 hours of
darkness daily). The relative humidity in both labora-
tories ranged from 40-80%.

After the twelve days of starvation, each animal
received approximately 50 g of Wayne mouse breeder blox
(MBB) or barley (see Appendix I) and from then on food was
added as needed. Food remaining at the end of the trial
was measured and individual consumption was recorded as
mg/(g mean body weight)(day). Ten animals from each
group also received ad libitum water. The light-dark
watered animals (LDW) were considered the normal controls.
Body weights were measured every two days on a Mettler
balance to the nearest 0.1 9 beginning with the first day
of the starvation period. Water bottles were also weighed
at this time. Individual water consumption was measured
using the techniques described by Bartholemew and Hudson
(1959) using inverted water bottles fitted with L-shaped

drinking tubes to reduce spillage. No correction was made

24

for occasional spillage and evaporative water loss was
negligible. Thus, the values for mean water consumption,
and its error, may be slightly biased. Water consumed

was expressed as g/(g of mean body weight)(day). Measure-
ments made during the starvation period were separated

from those made during the experimental period.

Determination of Body Water

 

After each animal died, its carcass was placed
in a drying oven at 105 degrees centigrade and weighed
periodically. A constant weight was recorded after three
identical weighings. Body water was then recorded as a
percent of the terminal wet body weight. Control animals
were sacrificed after most of the water deprived animals
had died.

Calculations and Assumptions in the
Determination of Oxygen Consumption

 

Total food ingested was measured at the end of
the experimental period and corrected for digestibility.
Fat and starch were considered 90% digestible; protein
was considered 80% digestible. The percent composition
of the diets studied is given in Appendix I. Since these
compositions are at 40% relative humidity, there exists
a small percentage of water in each case. It was assumed
that the oxidation of digestible foodstuffs led to the

standard end products of carbon dioxide, water, and urea.

25

Table 3. Oxygen Consumption when Different Foodstuffs
Are Oxidized.

1 t

Liters of 0 Liters of O *

 

 

 

Gms. H 0 formed 2 %

Food Type consumed pe used per g .

per gm. fOOd gm. food water formed
Starch 0.556 0.828 1.489
Fat 1.071 2.019 1.885
Protein 0.396 0.967 2.441

 

*From Schmidt-Nielsen, Desert Animals, 1964.

 

Oxygen consumption values were found using the
values of Schmidt-Nielsen (Table 3) for calculating the
amount of oxygen needed to oxidize one gram of starch,
fat or protein. For example, the oxidation of one gram
of starch requires the consumption of 0.828 liters of
oxygen and from this gram of starch, 0.556 gms. of water
are produced metabolically. To find the total amount of
oxygen consumed, corrections had to be made for changes

in body weight.

Estimation of Body Weight Changes

 

It was assumed that the composition of weight
gain or of weight loss was probably a function of initial
body weight. It was therefore necessary to estimate the

initial and terminal body composition of each animal in

26

order to determine what percentage of the weight change
could be attributed to fat, protein, and water.

Terminal carcasses were ashed in a muffle furnace
at a temperature of not less than 675 degrees centigrade
for eight hours. The ash content of these carcasses was
then expressed as a percent of the terminal wet body
weight. The percent ash was considered constant during
either weight gain or weight loss.

The animal with the highest percent body water
was taken as the most "fat-free" animal. Its water con-
tent was associated with approximately 2.68% fat, a mean
percentage calculated from the values of Pitts and
Bullard (1968) on six small rodents trapped in the wild.
These species were considered relatively "fat-free." The
values for ash, body water, and estimated fat were sub-
tracted from 100% to give this animal a protein percentage
of 21.21% of its body weight.

J. T. Ried, et_al. (1968) have shown, by their
prediction equations for body composition in sheep, that
the protein/water ratio during weight change remains es-
sentially constant. It was assumed that this relationship
was also true of rodents and, hence, knowing the ratio
for one animal, it was possible to find the percentage of
protein in the terminal carcass of every animal using the

following identity:

27

69.57 = % body water (measured in every animal)
21.21 % proteinTYunknown)

 

The percent fat in the terminal carcass of each animal
was found by subtracting the values for water, protein,
and ash from 100%.

Regression analysis of percent body fat versus
terminal body weight showed that 40% of the variation in
body weight could be attributed to percent body fat
(Appendix II). Knowing the initial body weight, and
using the regression line as the best predictor of percent
body fat, the amount of fat present initially could then
be estimated. The percent ash present initially was con-
sidered unchanged. The remaining percentage consisted
of protein and water. The average protein/water ratio
was 0.306156. Multiplying this value by the remaining
percentage gave the percent protein present initially.
Body water was found by multiplying the remaining percent—
age by 0.693844 or by subtracting the other components
from 100%. The difference in composition between the in-
itial and terminal weights constituted the body weight
change. A detailed account of the procedure for estimat-
ing the composition of body weight change can be found in
Appendix II.

The amount of fat and protein lost by water de-
prived animals was added to the total digestible food and

the total amount of oxygen consumed was calculated. The

28

amount of fat and protein gained by water ad libitum
animals was subtracted from the food intake and again the

total oxygen consumed was calculated.
Determination of Blood Osmolarity

Animals deprived of water over periods ranging
from 20-46 days as well as ad libitum control animals
were ether anesthetized and a 0.6cc. sample of blood ob-
tained by direct heart puncture. Duplicate hematocrits
were taken immediately. Blood samples were covered dur-
ing the procedure to prevent evaporation and then centri-
fuged for 20 minutes. A 0.2cc. aliquot of plasma was
withdrawn using a 0.2cc. diSPO Prothrombin Pipet (accuracy
: 2%; Scientific Products) fitted with an airtight gasket
(Adams Suction Apparatus No. A-2473). In cases where
blood samples were small, a 20uL pipet was fitted to the
gasket. The 0.2cc. samples were diluted in 0.2cc. of
ammonia-free distilled water; the 30uL samples were
diluted in 0.3cc. of ammonia-free distilled water. All
samples were placed in a Precision Systems Osmette cali-
brated to i 3 mos. and the osmolarity measured in tripli-
cate after three successful runs. A mean value was
recorded for each sample. Comparisons were made between

LD and LL (constant light) animals on high and low pro-

tein diets.

29

Statistical Considerations

 

Statistical significance was determined using
either the Student's "t" test or the Analysis of Variance
F-test for one-way classification (Sokal and Rohlf, 1969).
Homogeneous within group variation was tested using the

critical values of Fm (Sokal and Rohlf, 1969). When

ax
significance was revealed in the analysis of variance,

the treatment sums of squares was partitioned into single-
degree of freedom orthogonal contrasts for determination

of the significant mean responses. Regression analysis

of plasma osmolarity changes with time and percent body

fat followed the procedures outlined by Sokal and Rohlf
(1969). Quantitating equations were found using the

method of least squares regression following a logarith-

mic transformation of the data points. Justification for
the log-log scale in the survival response is best explained

as "goodness of fit." A detailed account of the procedures

can be found in the Appendix.

RESULTS

The rates of many and diverse physiological pro-
cesses are proportional to some power function of body
weight. It is possible, therefore, to express certain
particular responses in such an equational form and thus
to interrelate those parameters whose values are dependent
upon the same variables. The purpose of this chapter is
to present the results of a number of studies involving
Meriones' response to water deprivation with significance
levels for the treatment combinations (i.e., water de-
prived vs. controls; light-light vs. light-dark). A more
detailed account of the statistical treatment is given in
Appendix III. Whenever the data for a particular para-
meter are amenable to mathematical interpretation, the
response is expressed in equational form. Predicted
values can be obtained directly from the graphs or through
a series of computations (see Appendix IV). (Although
several of the parameters already studied were quantified
in this manner, the interrelationships of other closely

allied responses were left for a future study.

30

31

Water and Food Consumption

At room temperatures between 25-27 degrees centi-
grade, and with the relative humidity between 50-80%, the
mean rates of ad libitum water consumption for light-dark
(LD) and light-light(LL) animals (n = 20 animals not
significantly different in body weight at time 0) were
0.113 g/(g mean body weight)(day) and 0.106 g/(g mean body
weight)(day) when food was available. The difference in
mean response was not significant (p > 0.05). When water
consumption for these control groups was compared with
water consumption during the initial starvation period,

a significant difference was found. Mean water consump-
tion during starvation was 0.046 g/(g mean body weight)

35) compared to 0.11 g/(g mean body weight)

(day) (n

(day) (n

19) when food was given ad libitum. The "t"
value of 9.13 was highly significant at p < 0.01.

Food consumption during water deprivation was
also measured (Figure 1). Ad libitum control animals
under LD and LL conditions had an average food consumption
of 80.59 mg/(g body weight)(day) and 77.69 mg/(g body
weight)(day). Water deprived animals under LD and LL
conditions had mean responses of 35.93mg/(g body weight)
(day) and 24.95 mg/(g body weight)(day). A one-way analy-
sis of variance was run to determine the significance of
the four groups as a whole. The analysis resulted in an

"F" value of 46.74 which was highly significant at p<:0.01.

Figure l.

32

Food consumption during water deprivation.
Vertical lines represent ranges; horizontal
lines represent the mean. Boxes imply : 2
standard deviations from the mean.

A: LD Controls
B: LL Controls
C: LL Water Deprived
D: LD Water Deprived

(Width of vertical bars in Figure l, and all similar
figures, has no significance.)

 

 

33

wmp\.u3 moon w :OHDQESchU poom

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure l

 

 

050505050505050
machommmmarmehnuauaui3
___74_.m_d_mJ_
_I
D
J
C
_ _
B
_ _
A-_ 4
_ _
_e__r_u__u__.___e__
0505050505050505050
m998877665544332211

mmo\m\mfi cofiugfismcou boom

34

An orthogonal breakdown of the treatment sums of squares

resulted in the following null comparisons:

Hol: LDW animals = LLW animals (controls)
H02: LDNW animals = LLNW animals (water deprived)
H03: LDW:LLW animals = LDNW:LLNW animals (controls

vs. water deprived)
The values of 02 for the first two nulls, testing the
differences in mean response under light-dark and light-
light conditions, were not significant (p > 0.05). It
follows that constant light does not lower food consump-
tion in animals under a similar water status.

The value of Q2 for the third null hypothesis
measured the significance between animals on a dissimilar
water status. It was highly significant at p < 0.001.

From the above facts, it is evident that food
deprivation causes a significant decrease in ad libitum
water consumption and, likewise, that water deprivation
significantly reduces food consumption. The absence of
one dietary factor will lower the animal's response to
the other. Light status apparently is not a significant

factor.

Quantitation of Ad Libitum Water
Consumption: The Heterogonic

Equation

 

The time rate of water intake (cc/day) in relation

to body weight (g) is shown in Figure 2. It is evident

35

Figure 2. Quantitation of ad libitum water intake;
relation of water intake to mean body
weight. log Y = 1.72 log X - 2.309
r = 0.868

Sy x = 0.0135
Dotted lines represent 95% confidence

on the line.

Ad Libitum Water Consumption cc/day

20--

 

36

 

l l I I l I A 1

 

10

20

1 U I I I T I I
30 40 50 60 70 80 90 100
mean body weight (g)

Figure 2

I
150

37

that the logarithms of water intake are directly propor-
tional to the logarithms of body weight. The data fit an

equation of the form
log Y = log a + k log X

where "k" represents the constant rate of change of the
dependent variable function, log Y, for unit increments
of log X, the independent variable ("a" is equal to
TEE—Y - k TEE—X). The values of "a" and "k" were found
and the line drawn using the method of least squares re-
gression following a logarithmic transformation of the
data points. The mean log of Y, for eighteen animals,
was 0.8524, and the mean log of X was 1.8371. The slope
of regression, "k," was 1.72. Expansion of the equation
yielded

log Y (cc/day) = Tog-Y - k TEE—R + k log X

0.8524 - (l.72)(l.837l) + 1.72 log x
1.72 log x - 2.309

The mean drinking response, 7, of eighteen animals averag-
ing 69.5 g to ad libitum water was 7.39 cc/day.

The general form of this equation, Y = a X k,
expresses the degree of disproportionality between the
physiological response, Y, and an independent variable, X.

The correlation coefficient, r, measuring the

degree of interdependence between water consumption and

body weight, was 0.868.

38

Effects of a Deficient Water Supply on
Body Weight and SurVIVal

 

Water deprived gerbils on a high protein diet of
Wayne mouse breeder blox were unable to maintain their
initial body weight or to maintain a constant weight at
a lower level. The average weight loss experienced by
constant light animals was 57.6% of their initial body
weight. Animals under a light-dark cycle lost 58.2% of
their initial weight. Upon the removal of drinking water,
animals on a low protein diet of pearled barley experience
an initial loss of weight followed by a plateau at about
66% of initial. At no time was a significant gain in
weight observed for water deprived animals on either
ration.

Survival of water deprived gerbils was found to
be inversely related to protein intake (Table 4). Mean
survival was slightly in excess of 22 and 35 days for
LD and LL animals on a high protein diet. Barley fed
animals had mean survival times of 92.3 days in LD and
199.9 days in LL. Statistical significance was found
between the mean survival response of animals on the two
diets as well as between animals on the same ration but
under different environmental treatments (LD vs. LL).
Animals under constant light survived for a significantly
longer period than did animals on the same diet but under

a light-dark cycle. Animals on a barley diet had

39

Table 4. Gerbil Survival Without Water.

 

mean -
survival protein diet light N
intake
(d)
22.0 ~6.4 MBB LD 11
35.0 6.1 MBB LL 10
92.3 2.8 BARLEY LD
104.3 2.3 SUNFLOWER LL
199.9 2.4 BARLEY LL

 

*mg protein/g mean body weight/day.

significantly longer survival times than those raised on
a mouse breeder block ration (p<:0.05).

Tolerance to food deprivation might be an important
physiological adaptation in desert species. In the ab-
sence of water, gerbil survival without food was only 21
days. A significantly longer survival period was observed
in LD animals given ad libitum water (Table 5). The in-
crease was also significant under LL but the mean re-
sponse was only 30.4 days making it significant from

water ad libitum animals under LD as well.

Table 5. Gerbil Survival Without Food.

 

In LD: In LL:
N days : S.E. N days : S.E.
No water 18 21.0 I.1°3 17 21.5 :_1.7
Water ad lib. 16 41.2 i 2.8 18 30.4 + 1.7

 

40

Quantitation of the Survival Response
To Water Deprivation

 

 

The effect of environmental treatment on the

mortality rate of water deprived animals is shown in Figure

3. Least squares analysis resulted in two statistically
different regression lines (Figure 3a; 3b). Figure 3a
represents the mortality rate of LD animals and Figure 3b
the mortality rate of LL animals. The logarithms of the
percents of the population surviving (Y) was inversely
related to the logarithms of survival days (X). The

slopes of regression were ka = 1.58 and k = 0.41. After

b
a logarithmic transformation of the data points, the mean
logs of Y for LD and LL animals were found to be 1.66 and
1.83 respectively. Likewise, the mean logs of X were 1.58

and 1.20. The heterogonic equations expressing the two

responses are as follows:
LD Water Deprived Animals

log Y

log Y + k log X - k log X
1.66 + l.58(l.32) 1.58 log X

= 3.75 - 1.58 log x

LL Water Deprived Animals

log Y log Y + k log X - k log X
1.83 + 0.41(l.20) 0.41 log X

2.32 - 0.41 log X

Under constant light, 50% of the population were alive

after 30 days of water deprivation whereas under a

41

cow
1

mmomo

ow om

ow

om

cowuo>wummo nmum3 mo ammo
on

x mom Ha.o . mm.~
x 00H mm.H I mn.m
.:0flum>flummp Hmumz so mHmEHcm Amv Ag pom Adv GA mo mmpmn wuwamuuoz

OH
.

m
.
.

m h m
p r p

» mod
w mod

m

P

D

“mm Guzman
"4m musmflm

m
q-N

.m musmflm

 

c

P
d

r
u

q

1

 

d

4

LI

OH

6N

bm

be
bm

do
65

dm
6m

 

doa

DUIAIAJHS uorqetndod go %

42

light-dark cycle 50% survived only 20 days. Since the
function describing the survival response of LD animals
appears sigmoid, the predicted values of the regression
line may be slightly biased. Nevertheless, significance
in survival was shown to exist between populations under
a different light status.

Body Water in the Water Deprived Animal

in Light-Dark and Light-Light
Environments

 

Since much of the variation in percent body water
between individual animals can be attributed to percent
body fat, animals were placed on a 12 day starvation per-
iod with ad libitum water. Afterwards, food was available
ad libitum and 10 animals from each environmental treat-
ment were water deprived. The starting weights of each
group did not vary significantly. After 40 days of ex-
perimental treatment, body water was measured.

The results in Figure 4 represent percent water
in the terminal carcass. LD and LL water deprived animals
had an average percent body water of 64.49 and 64.51 re-
spectively. Control animals averaged 64.06% under LD and
59.44% under LL. Water deprived animals on a diet of
pearled barley and under constant light had an average
percent body water of 65.8. A one-way analysis of vari-
ance determined the significance of environmental treat-

ment. An "F" value of 5.052 was significant at p< 0.01.

Figure 4.

43

Body water in the water deprived animal
on high and low protein diets. Vertical
lines represent ranges; horizontal lines
represent the mean. Boxes imply : 2
standard deviations.

: (n=10) LD Controls MBB
B: (n=10) LL Controls MBB
C: (n=10) LL Water Deprived MBB
D: (n=ll) LD Water Deprived MBB
E: (n= 6) LL Water Deprived Barley

 

Hmuoz hpom usmouwm
4. 2 0 8

 

--76
‘J74

“‘70

-472

8 6
6 6 6 6 6 5
_ _

_ _ _ _

 

E WIT
FIT

 

 

44

I——b
IhJ'

 

_———

 

 

 

 

-— 56
-~ 52
-—+-so
-— 48

fi54

Figure 4

 

 

 

B

 

 

 

 

<—_.

_
L.
_ _ _

 

76....
74—
72,—
7o—

64 " —-—L-—I—-

—

68r-

2 0 8
6 6 5

Hmumz moom unwoumm

 

52 P-

564—-
54*h-
50 —-
48-

45

Construction of orthogonal contrasts resulted in the fol-
lowing null hypotheses:

Ho : LDW = LLW

1
H02: LDNW = LLNW
H03: LDW:LLW = LDNW:LLNW

Comparison of barley animals with controls utilized the
Student's "t."

The first and third null hypotheses were rejected
at p> 0.05. No significance could be found between water
deprived animals under LD and LL. Barley animals did not
differ significantly from LD controls. The mean response
of LL ad libitum animals was low enough to produce signi-
ficance between water ad libitum and water deprived ani-
mals. When LD controls were compared to LD water deprived
animals, no significance was found (p:>0.05). Thus, it
was concluded that a significant decrease in percent body
water resulted from a constant light environment when
water was available. Since no significance was found
between water deprived animals, it was concluded that the
effect of constant light on body water is dependent upon
the "water--no water" regime. Within group variation
about the mean were homogeneous and hence not significant.

Body water after starvation was also measured
(Table 6). No significance could be shown between LD and
LL animals on a no water regime or between animals given

water ad libitum. The populations were combined in column 3.

 

46

«00.0 a «00.0 a 000.0 a

 

hN.o + mm.m® vm vv.o.H mm.mm wH Hm.o + hv.mm ma .QHH Um Hmflmg

mm.o H mm.©w mm mm.o H Nh.mw NH Nm.o H hH.hm ma HQHM3 OZ

.m.m H w : .m.m H w c .m.m H w c ucmEuwwue
"pmcflnaoo "qq cH "on cH

.coflpw>umum “ovum Hmumz hpom .m canoe

47

A significant difference in percent body water was observed
between water ad libitum and water deprived animals during
starvation. Environmental light status apparently has no

effect during starvation.

Effect of Water Deprivation on Mean
Oxygen ConsumptIOn

Comparisons of mean oxygen consumption values be-
tween water deprived animals and controls are shown in
Figure 5. The results are expressed in cc 02/( g mean
body weight)(day) on four treatment groups. LD controls
had a mean oxygen response of 56.72 cc/(g mean body weight)
(day) compared to 52.21 cc/(g mean body weight)(day) in LL
ad libitum animals. No significance could be shown
(p> 0.05) with the Student's "t." LD water deprived
animals had a mean oxygen consumption value of 41.27 cc/(g
mean body weight)(day) while LL water deprived animals
consumed an average of 41.62 cc/(g mean body weight)(day).
As with the control groups, no significance could be
shown.

A one-way analysis of variance was run to deter-
mine significance among the four groups as a whole. The
treatment sums of squares gave an "F" value of 5.3 which
was highly significant at p<<0.01. An orthogonal break-
down of the treatment sums of squares into single degree
(of freedom contrasts was used to determine the signifi-

czance of water availability (water status) since no

Figure 5.

48

Comparison of mean oxygen consumption during
water deprivation. Vertical lines represent
the ranges: horizontal lines represent the

means. Boxes imply :_2 standard deviations.

A: LD Controls
B: LL Controls
C: LL Water Deprived
D: LD Water Deprived

49

wmp\.u3 apon m coaumESmcou Gmmhxo

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 5

 

 

50 50505 05 05 05050505
xv LJ “8 8. 7 7 as «a :3 :3 .4 .w .3 .3 9“ 92 1h .L m
. q . _ _ _ _ . . . I. q . _ 4‘ a _ 1 _

_ .0

D

r _

. i.

j. __

a 1 1
pP.—._bL.bb_-_bb.__
505050505050505050
998877 66 5544332211

wmp\m\oo sOHumfidmcoo cmmwxo

50

significance was found between environmental treatments

(light status). The breakdown resulted in the following

null hypotheses:

H01: LLW = LLNW
H02: LDW = LDNW
H03: LDW:LDNW = LLW:LLNW

The values of Qi and 0: representing the first two nulls
were highly significant at p<<0.01 and p<:0.002 respec-
tively. Qi for the third null was not significant
(p>'0.25), as could be expected from the results of the
Student's "t." Water deprivation significantly lowered
mean oxygen consumption in those animals so treated as
compared with control animals on ad libitum water. No

significance could be attributed to the two conditions

of environmental light.

Effect of Water Deprivation on the
Quantitationof the Oxygen
Consumption Response

Graphic interpretation of oxygen consumption
(cc/day) and mean body weight (g) resulted in two statis—
tically significant regression lines. Figure 6a and
Figure 6b represent data from ad libitum and water de-
prived animals respectively. As evidenced by the graph,
the logarithms of oxygen consumption are prOportional to
the logarithms of mean body weight. The equations for

the regression lines are again heterogonic. Following a

51

zfday
a

l

I

N

O
L
I

 

Oxygen Consumption L O

1.0--

0.9"
0.8--L

0.7-b
0.64-

 

I g l I 11 44 I I I 4'4
10 2o 30 40 56 65 501090100 150
Mean Body Weight (g)

Figure 6. Effect of water deprivation on the quantitation of

oxygen consumption. Figure 6a - controls; Figure

6B - water deprived animals. 82 = standard error
of the line.

52

logarithmic transformation of the data, the slopes of re-
gression were found to be 0.736 for controls and 0.938 for
water deprived animals. The slopes of the two systems are
not significantly different. The degree of disproportion-
ality existing between oxygen consumption and mean body
weight is less than that found for water consumption and
mean body weight. The mean logs of Y were 3.58 and 3.24
for ad libitum and water deprived animals respectively;
the mean logs of X were 1.84 and 1.68. The prediction
equations for ad libitum and water deprived animals are

as follows:

Ad Libitum Controls (Y = cc/day; X = g)

Iog Y - k Iog X + k log X
3.58 - 0.736(l.84) + 0.736 log X
2.23 + 0.736 log X

log Y

Water Deprived (Y = cc/day; X = g)

Iog Y - k Iog X + k log X
3.24 - 0.938(1.68) + 0.938 log X
1.67 + 0.938 log X

log Y

It was concluded that water deprivation caused a signifi-
cant "shift" in the oxygen consumption to body weight
relationship.

The correlation coefficients, r, measuring the
degree of interdependence between oxygen consumption and
body weight were ra = 0.828 and rb = 0.888 for control

and water deprived animals respectively.

53

Plasma Osmolaritprhanges During Water
Deprivation in Animals on High and
Low Protein Diets
Blood osmolarity changes in animals on a water
deficient diet varying in protein content were measured
over a period of days (Table 7). Each animal was used
once and therefore variations between individual animals
were not accounted for. They appeared to be insignificant
(see standard errors). Regression analysis of osmolarity
(mOs/L) versus days yielded insignificant slopes (k = 0)
for all groups. Comparisons of mean responses utilized
the Student's "t" test. Animals under similar environ-
mental treatments (LD and LL) were compared to each other
to determine the effect of the diet on plasma osmolarity
changes. Likewise, comparisons were made between animals
on the same diet but under different environmental treat-
ments. No significance could be attributed to light
status (p> 0.10). Animals raised on a barley diet were
not significantly different from controls. The difference
between LL controls and LL animals raised on mouse breeder
blox bordered on significance. Only one animal survived
under LD and mouse breeder blox and hence, no meaningful
comparison could be made. Since the slope of regression
for LL animals on mouse breeder blox did not differ

significantly from zero after 20 days of water deprivation,

54

Houum pumpcoum H «

 

 

 

 

 

00.0“
00 000000 00.000 00.00 00-00 0000 2 0
00.0“
00 mm: 00.000 00.00 00-00 0:0: z 0
00.000
00 mm: 00.000 00.00 00-00 .n00 00 z 0
00.00“
00 000000 00.000 00.00 00-00 0000 2 0
00 mm: 00.000 00.00 00 mac: 2 0
00.mfl
00 mm: 00.000 00.00 00-00 .000 00 z 0
mx\EmOE m mwoo
x.
00000 0000 mn0um0osmo .060 00 00000 000000 umumz xmm xzv
.coflpm>00mmo Hmumz mcflusa muHHoHOEmO mEmmam .5 OHQMB

55

it was concluded that plasma osmolarity rose during the
initial phase of water deprivation (although probably not
to a significant level) and then leveled off during the

remaining days of the trial.

DISCUSSION

Among the species of rodents listed in Table 1,
the ad libitum water intake of Meriones unguiculatus (11%
of their body weights/day) is between that of Dipodomys
agilis and Acomys cahirinus. In contrast to the findings
of Lindeborg (1952) or of Lee (1963), successful correla-
tion of ad libitum water intake (cc/day) to body weight
(g) was found for this species. The correlation coeffi-
cient of 0.868 is well above those of Lindeborg (which
ranged from 0.032 in Noveboracensis bairdi to 0.536 in

Tornillo blandus). The reason for such a contrast might

 

be explained on the basis of individual methods.

In this study, prior to the measurement of ad
libitum water intake, all animals were starved in an at-
tempt to reduce individual variation in percent body water
due to the presence of fat. Fat deposition acts as a
"sink" in the physiological system and removes this
tissue from the lean body mass. Fat adds to the weight
of the animal but in no significant way does it reflect
the requirements of the active cell mass. As a result,
the highest correlation of a physiological process to
body weight would be expected to occur when the body

weight of an animal is a close approximation of the lean

56

57

body mass (i.e., an essentially "fat-free" animal). Evi-
dence for the assumption that ad libitum control animals
were relatively "fat-free" during the experimental period
is supplied by the fact that the percent body water in
these animals did not differ significantly from that of
water deprived animals. Water deprived animals (which
continue to loose weight during water deprivation) were
almost certainly low in body fat since the highest per-
cent body water recorded was for this group. It is as-
sumed, therefore, that the body weights of control animals
more accurately reflected the lean body masses of the
animals than did the body weights of the animals used by
Lindeborg and Lee, and are the reason for the high cor-
relation of water intake to body weight seen in this
study.

Quantitation of ad libitum water consumption (cc/
day) versus mean body weight (g) resulted in a regression
equation of the form

y (cc/day) = 0.0045 x 1'72

On a weight relative basis (cc/g/day), the equation becomes

y (cc/g/day) = 0.0045 x 0'72

revealing that, relative to weight, water intake is pro-
portional, not to body weight, but to the 0.72 power of
body weight. The exponent of X, 0.72, is a close approxi-

mation of Klieber's (1961) use of body weight to the

58

three-fourth power as the reference base for physiological
quantitation. It must be said, then, that water intake is
relatively size dependent.

Water deprivation is an extreme procedure and
probably unnatural, but survival time without water may
reflect the extent to which a species is adapted to
limited amounts of water. In addition, this study has
shown, that dietary intake is an important factor to the
survival of water deprived animals in that survival time
was found to be inversely related to protein intake
(Table 4). Increasing the amount of dietary protein
significantly reduced the time at which 50% of the popu-
lation of water deprived animals survived. A significant
increase in the survival of constant light animals was
also observed but this was probably the result of a
slightly increased protein intake in these animals since
no significance could be shown in either the starting
weights or in the oxygen consumption of these animals as
compared to LD controls.

Aside from the small amount of metabolic water
formed from the oxidation of protein, the formation of
urea as a degredative product of protein metabolism
necessitates an increase in the obligatory water needed
for urine production. The combined effects of water
deprivation and a high protein diet are lethal to Meriones.

Animals fed a diet of pearled barley, however, had a

59

significantly longer survival time which is attributed to
a reduction in the need for urea processing.

It has been assumed (Howell and Gersh, 1935) that
certain desert species like the kangaroo rat, are able to
withstand a greater degree of dessication than other mam-
mals. The present study does not concur with this assump-
tion. Water deprived gerbils contined to lose weight on
a high protein diet but, after 40 days of water depriva-
tion, the body water percentage was not significantly
different from the normal LD controls given ad libitum
water (Figure 4). The difference between normal controls
and barley-fed animals was also insignificant. On a
starvation regime, however, water deprived animals had a
significant reduction in percent body water when compared
to animals given ad libitum water. Apparently, enough
free and metabolic water can be obtained from the food
intake to handle urine and evaporative water loss. The
conclusion must be made that, although water deprived
animals were in negative water balance, the proportion of
water in the body remained essentially the same as they
gradually lost weight. Even though a considerable amount
of water had been lost by these animals their bodies were
not really dessicated.

The average percent body water of 10 animals under
constant light was significantly lower than that for nor-

mal LD controls. The reasons for this finding are as yet

60

uncertain but a possible hormonal effect has been suggested.
Piacsek and Meites (1967), working with underfed rats,

found a decrease in reproductive function which they attri-
buted to a deficiency of follicle stimulating hormone (FSH)
and luteinizing hormone (LH) release from the pituitary.
Reactivation of gonadotropin release (FSH: LH) in the
starved rats was achieved under constant illumination.
Estrogen secretion was indicated by an increase in uterine
weight, enhanced mammary duct development, and by increased

pituitary weight. Although the animals in the current study

 

were randomly assigned to treatment groups, six of the 10
animals under constant light were females. It is suggested
that estrogen secretion in these water ad libitum animals
resulted in a synthesis and deposition of fat which
lowered the mean percent body water of the group as a
whole. Since no estrogen levels were measured, the prob-
lem, however interesting, needs further research.

Plasma osmolarity changes in the water deprived
gerbil appear to be insignificant under both environmental
treatments and type of diet. The plasma osmolarities of
animals on a high protein diet were higher but not signi-
ficantly higher than control or barley fed animals. Ap-
parently, plasma osmolarity rises slightly during the
initial phases of water deprivation after which the animal
retains relative homeostasis. The initial stages of water

deprivation are probably transient phases during which

61

time the kidney is adjusting to plasma osmolarity changes.
Homeostasis is regained but at a level slightly higher
than normal.

It is known that a reduction in food intake occurs
during periods of water deprivation in rodents (Klieber,
1961; Fertig and Edmonds, 1969). Moreover, during periods
of relative starvation, there is a decrease in thyroid
stimulating hormone (TSH) by the pituitary resulting in a
depression of thyroid function. This decrease in thyroid
activity causes an concommittant decrease in oxygen con-
sumption by the tissues. In the present study, water
deprived animals reduced their food intake and showed a
24% reduction in oxygen consumption below that of control
animals (Figure 5), supporting the work of these investi-
gators.

Robinson (1959) has reported a resting oxygen

consumption value of 2.14 cc/(g)(hr) for M. unguiculatus

 

at 25 degrees centigrade. In the current study, the mean
value for oxygen consumption in control animals was found
to be 2.27 cc/(g)(hr), which approximates Robinson's
figure. When gerbils are housed individually rather than
in groups, activity is very often at a minimum. It is
assumed, therefore, that the values obtained for control
animals approximated the normal resting condition of the

animal, being only 6% above Robinson's figure.

62

The physiological significance of a reduced meta-
bolic rate during water deprivation is two-fold. A re-
duction in energy expenditure, similar to that seen in
aestivating mammals, would prolong survival by extending
the period during which the energy reserves of the organ-
ism can last. Moreover, Fertig and Edmonds (1969), using
water deprived mice, have reported a decrease in evapora-
tive water loss following a reduction in oxygen consump-
tion. It is, therefore, apparent that a decrease in
oxygen consumption during water deprivation is as import-
ant for water economy as it is for energy conservation.

The system relating oxygen consumption (L/day) to
mean body weight (g) is shown in Figure 6. Regression
analysis of the data points resulted in two significantly
displaced regression lines representing control Figure
6a and water deprived Figure 6b animals respectively;
whereas the slopes of each system were not significantly
different. Three significant physiological features can
be elucidated from this regression analysis. First,
water deprived animals show a significant reduction in
oxygen consumption below control animals but the data
points for this group do not constitute the lower end of
regression line a, instead, the entire system correlating
oxygen consumption to body weight is shifted downward so
that animals of the same weight will have different oxygen

consumption values depending upon their water status. The

63

use of a prediction equation for oxygen consumption must,
therefore, acknowledge the water status of the animal
since two different systems relating oxygen consumption
to body weight can be shown to occur.

Secondly, there exists a high degree of correla-
tion between oxygen consumption (L/day) and body weight
(g). The correlation coefficient (rb) for water deprived
animals, 0.888, is slightly but not significantly higher
than that for controls (ra = 0.828). Since oxygen con-
sumption is a function of the active cell mass of an
animal, the high correlations suggest that the body
weights of the animals were a close approximation of the
lean body mass; the starvation period significantly re-
duced the variation inherent between individuals due to
the presence of fat.

The third significant feature of Figure 6 involves
the use of the heterogonic equations which describe the
response. The equations for control and water deprived
animals were as follows:

Control Animals:

Y (cc/day) = 54.95 x 0°736

Water Deprived Animals:
Y (cc/day) = 36.31 x 0°938

On a weight relative basis, the equations become

64

Control Animals:

Y (cc/g/day) = 54.95 X-o'264

Water Deprived Animals:

Y (cc/g/day) = 36.31 x‘°°162

the slopes of which are not significantly different from
zero. This implies that on a weight relative basis (i.e.,
cc/g/day) oxygen consumption is virtually size independ-
ent in intraspecific studies. It is possible, however,
that the range of body weights within this species is not
significant enough to show the inverse relationship of

O2 (cc/g/day) to body weight (g) that is seen between
species. It seems logical to assume that, given the con-
ditions of this study (i.e., relatively "fat-free" animals)
this relationship of oxygen consumption to body size

probably would not exist in intraspecific studies.

SUMMARY AND CONCLUS IONS

1. Physiological processes which reflect the
requirements of the active cell mass of an animal can be
successfully correlated to the body weight of the animal.
The highest correlations between the responses and the
independent variable (body weight) will occur when the
body weight is a close approximation of the lean body
mass.

2. Quantitation of ad libitum water intake showed
that water intake was proportional to the 0.72 power of
body weight.

3. Meriones unguiculatus appear to be less adapt-

 

able to water deprivation than the kangaroo rat, Dippdomys

 

merriami. On a dry barley diet, Meriones' survival is
indefinite but does not reach the limits seen for the
kangaroo rat.

4. Survival under water deprivation is inversely
related to protein intake. Increasing the amount of por-
tein in the diet necessitates an increase in the obliga—
tory water needed for urea excretion and infringes upon
the animal's water balance.

5. The percent body water of water deprived

animals does not vary significantly from that of control

65

66

animals and is not influenced by dietary composition.
6. A significant reduction in oxygen consumption
is observed during periods of water deprivation. The re-

duction is an important mechanism for energy conservation

and water economy.

7. Water deprivation causes a significant "shift"

in the system relating oxygen consumption (cc/day) to
body weight (g). Thus, animals of the same body weight

will have different oxygen consumption values depending

upon their water status.

LITERATURE CITED

LITERATURE CITED

Adolph, Edward E. 1943. Do rats thrive when drinking
sea water? Amer. J. Physiol. 123: 369-378.

Adolph, E. F. 1949. Quantitative relations in the
physiological constitutions of mammals. Science
109: 579-585.

Bartholemew, George A. and Jack W. Hudson. 1959. Effects
of sodium chloride on weight and drinking in the
antelope ground squirrel. J. Mamm. 40: 354-360.

Bartholemew, George A. and Richard E. MacMillen. 1961.
Oxygen consumption, aestivation, and hibernation
in the kangaroo mouse, Microdipodgps pallidus.
Physiol. Zoo. 34: 177-183.

 

Benedict, F. G. 1938. Vital Energetics: A Study in
Comparative Basal Metabolism. Carnegie Inst. of
Wash. Publ. 503: p. 65.

Brody, Samuel. 1964. Bioenergetics and Growth. Hafner
Publishing Company, Inc. New York. 1023 p.

Burns, T. W. 1956. Endocrine factors in the water meta-
bolism of the desert mammal, Gerbillus gerbillus.
Endocrinol. 58: 243-254.

Chew, Robert M. 1951. The water exchanges of some small
mammals. Ecol. Monogr. 21: 215-225.

Chew, Robert M. 1961. Water metabolism of desert inhabit-
ing vertebrates. Biol. Rev. 36: 1-31.

Chew, Robert M. and Ralph T. Hindegardner. 1957. Effects
of chronic insufficiency of drinking water in
white mice. J. Mamm. 38: 361-374.

Dawson, William R. 1955. The relation of oxygen consump-
tion to temperature in desert rodents. J. Mamm.
36: 543-553.

67

68

Fertig, Daniel S. and Vaughan W. Edmonds. 1969. The
physiology of the house mouse. Sci. Amer. 221:
103-110.

Howell, A. Brazier and I. Gersh. 1935. Conservation of
water by the rodent Dipodomys. J. Mamm. 16:
1-9.

 

Hudson, Jack W. 1962. The role of water in the biology
of the antelope ground squirrel, Citellus leucurus.
Univ. of Cal. Publ. Zoo. 64: 1-56.

Khalil, F. and G. Abdil-Messeih. 1954. Water content of
tissues of some desert reptiles and mammals. J.
Exp. Zoo. 125: 407-413.

Kirmiz, John P. 1962. Adaptation to Desert Environment,
A Study on the Jerboa, Rat, and Man. Butterworths,
London. 154 p.

Kleiber, Max. 1961. The Fire of Life. John Wiley and
Sons, Inc. New York. 454 p.

Lee, Anthony Kingston. 1963. The adaptations or arid
environments in wood rats of the genus Neotoma.
Univ. of Cal. Publ. Zoo, 64: 57-96.

Lee, R. C. 1939. Size and basal metabolism of adult
rabbit. J. Nutr. 18: 489-500.

Lindeborg, R. G. 1952. Water requirements of certain
rodents from xeric and mesic habitats. Contr.
Lab. Vertebr. Biol. Univ. Mich. 58: 1-32.

MacMillen, Richard E. and Anthony K. Lee. 1967. Australian
desert mice: Independence of exogenous water.
Science. 158: 383-385.

Misonne, Xavier. 1959. Analyze zoogeographique des
mammiferes de 1'Iran. Bruxelles, Inst. Royal des
Sci. Nat. de Belgique, Memoires. 59: 157 p.

Pearson, Oliver P. 1960. The oxygen consumption and
bioenergenics of harvest mice. Physiol. Zoo.
33: 152-160.

Petter, F. 1953. Note preliminaire sur l'ethologie et
l'ecologie de Meriones libycus (Rongeurs, Gerbil-
lides). Mammalia. 17: 281-294.

 

69

Piacsek, Bela E. and Joseph Meites. 1967. Reinitiation
of gonadotropin release in underfed rats by con-
stant light or epinephrine. Endocrinol. 81:
535-541.

Pitts, Grover C. and T. Robert Bullard. 1968. Some
interspecific aspects of body composition in
mammals. £3 Body Composition in Animals and
Man. National Academy of Sciences. Washington,
D. C., 45-70.

Reid, J. T., et a1. 1968. Some peculiarities in the
body composition of animals. £2 Body Composition
in Animals and Man. National Academy of Sciences.
Washington, D. C., 19-44.

 

Rich, Sigmund T. 1968. The Mongolian gerbil (Meriones
unguiculatus) in research. Lab. Anim. Care.
18: (2) 235-241.

 

Robinson, Paul F. 1959. Metabolism of the gerbil,
Meriones unguiculatus. Science. 130: 502-503.

 

Robinson, Paul F. and Roy V. Henrickson. 1961. Metabolism
of Gerbillus pyramidum. Nature. 180: 637-638.

 

Schmidt-Nielsen, B. and K. Schmidt-Nielsen, A. Brokaw and
H. Schneiderman. 1948. Water conservation in
desert rodents. J. Cell. Comp. Physiol. 32:
331-336.

Schmidt-Nielsen, B. and K. Schmidt-Nielsen. 1950. Eva-
porative water loss in desert rodents in their
natural habitat. Ecol. 31: 75-85.

Schmidt-Nielsen, Knut. 1964. Desert Animals: Physio-
logical Problems of Heat and Water. Oxford
University Press. New York and Oxford. 277 p.

Schmidt-Nielsen, K. and H. B. Haines. 1962. Water balance
in a carnivourous desert rodent, the grasshopper
mouse. J. Cell Comp. Physiol. 52: 259-265.

Sokal, Robert R. and F. James Rohlf. 1969. Biometry:
The Principles and Practices of Statistics in
Biological Research. W. H. Freeman. San Fran-
cisco. 776 p.

Sokal, Robert R. and F. James Rohlf. 1969. Statistical
Tables. W. H. Freeman. San Francisco. 253 p.

7O

Sokolov, W. 1966. Water content in the tissues of
desert animals. Nature. 211: 545.

Spector, William S. (compiled by). 1956. Handbook of
Biological Data. W. B. Saunders Company.
Philadelphia and London. 584 p.

Williams, Olwen. 1959. Water intake in the deer mouse.
J. Mamm. 40: 602-606.

Winkelmann, John R. and Getz, Lowell L. 1962. Water
balance in the Mongolian gerbil. J. Mamm. 43:
(2) 150-154.

APPENDICES

APPENDIX I

Percent Composition of Air-Dry Diets

 

Mouse Sunflower Pearled

Breeder Blox Seeds Barley
Water* 13.3 4.8 11.1
Ash 4.7 4.0 0.9
Starch 55.1 19.9 78.8
Fat 9.1 47.3 1.0
Protein 17.8 24.0 8.2
kcal/lOOgm 356 560 349

*water content variable with humidity;
values are for equilibration with 40% R.H.

Metabolic Parameters for (a) Diets as Fed
(40% R.H.) and (b) Fat Free Weight Loss

Mouse Sunflower Pearled Weight
Breeder Blox Seeds Barley Loss
Caloric Density
(kcal/lOOgm) 356 560 349 130
Protein/Energy
Ratio (g/kcal) 0.0500 0.0429 0.0235 0.2419
Urea/Oxygen
(mg/m1) 0.0685 0.0555 0.0341 0.3548
Urea/Total Water*
(mosmols/liter) 1527 1644 721 2222
TOtal water1"/°xygen0.748 0.562 0.788 2.664

(mg/ml)

*Total Water = free water in diet plus water
formed by oxidation.

71

APPENDIX II

Regression Analysis of Body Fat Least Squares
Prediction from Body Weight

Regression ANOVA

Source of Variation d.f. SS M88 F
Regression 1 481.274 481.274 25.21
Residual 39 744.662 19.094

Total 40 1225.935

P(F = 25.21) < 0.001

Least Squares Regression Line
Y = a + bX

b = 0.17 a = Y}- bXI= 2.01

Y = estimated mean % fat X mean body weight
Y' = 2.01 + 0.17X
39.25% of variation in body weight is due to fat content

r = 0.625

72

73

xha.¢ + Ho.m u .M .mm0.o u ucmwoflmmmoo GOHumamuuoo
.unm003 upon 800m sofluowomum mmumowm ummma “you Soon mo mammamcm aoflmmmummm

Amfionmv Davao: hoom Hmcfifiuma

 

moa mm mm mh mm mm mv mm mm
— _ . . F . p _
0 0 . 0 1 0 _ 0
o
o
0
o o
o o
0 O 0 0 O O
O
o o o O. Q
o o o. o. o, o
O O
o o
0 o
o 6
O
O
0

OH

om

fin

190 Kpos % penemtqss

APPENDIX III

Statistical Treatment of Parameters
Analysis of Variance and Orthogonal

Linear Contrasts

N = number of observations

n = sample size

a = number of treatment groups
Ey.= grand sum

y = sample mean
Ey2 = sum of squared observations
Eyi = sum of sample ni

CT

grand total squared and divided by N

The Sum of Squares:
SS (Total) = By2 - CT
SS (Treatments) = (Eyi)2/ni - CT

SS (Error) = SST - SSt

The anova table is constructed as follows:

Source of variation d.f. SS M88

Among groups a-l SSt SSt/a-l
Within groups N-a SSe SSe/N-a
Total N-l SST

MSt/MSe

When significance is found with the F test, the among group
sums of squares can be partitioned into single-degree of
freedom contracts which are orthogonal and hence ask in-
dependent questions about the treatment combinations.

74

75

The Orthogonal Breakdown

There are as many orthogonal contrasts in a statis-
tical analysis as there are degrees of freedom in the
treatment sums of squares. An orthogonal set consists of
scalars which sum to zero when any combination of "vectors"
is cross multiplied. The test statistic is Q .

Z Z

02 = (EMT)2/NEM2
M = scalar value = l,0,-l
T = treatment total
Q2 = 88 + sum of squares

The anova table is identical to that of the analysis of
variance F test. The mean sums of squares are equal to
the sums of squares since each contrast has but one
degree of freedom.

Variance Homogeneity

2

F(max) = S max/82min

p = 0.05

Method of Least Squares Regression
(log transformation)

X = log X Y = log Y
Ex = sum of independent variable
Ey = sum of dependent variable
Ex ,Ey = sum of squared variable
Exy = sum of cross products (SP)
b = slope of the line = SP/SSx
a = ordinate intercept = TEE—Y - b TEE—2'

Y = a + bX

Coefficient of Correlation

r2 = SSregression/SStotal

r

 

76

Student's "t"

t = Y1 - YZ/SE
t = test statistic
Y1,Yé = mean response
SE = standard error = (SE1)2 + (SE2)2

"t" is significant when p is less than or equal to 0.05.

APPENDIX IV
Log Computation of the Heterogonic Equation
Y = an
log Y = log a + k log X
a = 1739—? - k TEE—7

log Y = log Y - k Iog X + k log X

The value of any one Y can be found by taking the
antilog. The least squares regression line uses only
the logarithmic values of the data points.

77

 

MICHIGAN STATE UNIVERSITY LIBRARIES

IIIIIIIGIIIIIIIIII

904

I
5! III
I