Mammals

Mammals

Comparative Physiology of Thermoregulation

G. CAUSEY WHITTOW

DEPARTMENT OF PHYSIOLOGY, SCHOOL OF MEDICINE, UNIVERSITY OF HAWAII, HONOLULU, HAWAII

Table of Contents

Cover image

Title page

CONTRIBUTORS

Copyright

LIST OF CONTRIBUTORS

PREFACE

CONTENTS OF VOLUME I

Chapter 1: RODENTS

Publisher Summary

I Introduction

II Thermal Environment of Rodents

III Thermal Behavior

IV Precision of Homeothermy

V Temperature Regulation in Cold Environments

VI Temperature Regulation in Warm Environments

VII Activity and Temperature Regulation

VIII Altitude, Hypoxia, and Temperature Regulation

IX Sex Differences in Temperature Regulation

X Summary and Conclusions

Chapter 2: CARNIVORES

Publisher Summary

I Introduction

II General Thermal Defenses

III Protection against Hypothermia

IV Cold Acclimation

V Protection against Hyperthermia

VI Heat Acclimation

VII Thermoregulatory Importance of Hypothalamic Temperature

Chapter 3: UNGULATES

Publisher Summary

I Introduction

II Body Temperature

III Heat Production

IV Heat Loss

V Behavioral Thermoregulation

VI Food and Water Intake

VII Thermoregulation in the Young Animal

VIII Conclusions

Chapter 4: PRIMATES

Publisher Summary

I Introduction

II Normal Thermoregulation in the Monkey

III Behavioral Thermoregulation

IV Fever and Other Responses to Pyrogens

V Central Mechanisms of Thermoregulation

VI Neurochemical Control Mechanisms in the Hypothalamus

VII Conclusions

ACKNOWLEDGMENTS

Chapter 5: MAN

Publisher Summary

I Introduction

II Man’s Energy Exchanges Related to Temperature and Work

III Physiological Temperature Regulation

IV Behavioral Temperature Regulation: Thermal Comfort and Temperature Sensations

V Behavioral versus Physiological Temperature Regulation

VI A Mathematical Model of Man’s Thermoregulatory System

AUTHOR INDEX

SUBJECT INDEX

CONTRIBUTORS

THOMAS ADAMS

J.S. HART

A. PHARO GAGGE

R.D. MYERS

JAMES D. HARDY

J.A.J. STOLWIJK

G. CAUSEY WHITTOW

Copyright

COPYRIGHT © 1971, BY ACADEMIC PRESS, INC.

ALL RIGHTS RESERVED

NO PART OF THIS BOOK MAY BE REPRODUCED IN ANY FORM, BY PHOTOSTAT, MICROFILM, RETRIEVAL SYSTEM, OR ANY OTHER MEANS, WITHOUT WRITTEN PERMISSION FROM THE PUBLISHERS.

ACADEMIC PRESS, INC.

111 Fifth Avenue, New York, New York 10003

United Kingdom Edition published by

ACADEMIC PRESS, INC. (LONDON) LTD.

24/28 Oval Road, London NW1 7DD

LIBRARY OF CONGRESS CATALOG CARD NUMBER: 79-107580

PRINTED IN THE UNITED STATES OF AMERICA

LIST OF CONTRIBUTORS

Numbers in parentheses indicate the pages on which the authors’ contributions begin.

THOMAS ADAMS(151),     Department of Physiology, Michigan State University, East Lansing, Michigan

A. PHARO GAGGE(327),     John B. Pierce Foundation Laboratory, and Department of Epidemiology and Public Health, Yale University, New Haven, Connecticut

JAMES D. HARDY(327),     John B. Pierce Foundation Laboratory, and Department of Epidemiology and Public Health, Yale University, New Haven, Connecticut

J.S. HART(1),     Division of Biology, National Research Council of Canada, Ottawa, Canada

R.D. MYERS(283),     Laboratory of Neuropsychology, Purdue University, Lafayette, Indiana

J.A.J. STOLWIJK(327),     John B. Pierce Foundation Laboratory, and Department of Epidemiology and Public Health, Yale University, New Haven, Connecticut

G. CAUSEY WHITTOW(191),     Department of Physiology, School of Medicine, University of Hawaii, Honolulu, Hawaii

PREFACE

The purpose of this three-volume treatise is to present separate accounts of the means by which each of the major groups of animals regulates its body temperature, heat production, and heat loss. Readers who wish to obtain information on thermoregulatory processes in different species will, it is believed, welcome this as a convenient reference work. However, it is hoped that the treatise will have more than convenience value. The juxtaposition of chapters dealing with quite different animals might tempt the investigator to read about species other than the particular one in which he is immediately interested. In so doing he might well find that the specific problem with which he is concerned is better illuminated in another species. This, of course, is the essence of comparative physiology.

The composition of Volume I was determined partly by the amount of information available on invertebrates and on vertebrates other than mammals. Most experimental work on thermoregulation has been performed on mammals, and, accordingly, this volume deals exclusively with mammalian species. Thus, a rigid distinction between homeotherms and poikilotherms, ectotherms and endotherms, and regulators and nonregulators has been avoided.

The obvious omissions in Volume II will be made good in Volume III, which, in addition, will cover some special aspects of thermoregulation. Several of the chapters in Volume III will treat thermoregulation from the point of view of physiological systems. It is believed that this approach will complement the largely taxonomic organization of the first two volumes. It is also hoped that it will lend to the work as a whole a measure of integration which could not be achieved by considerations of species or systems alone.

Once again it is a pleasure to record my gratitude to Mrs. Jane Inouye, Mrs. Hilda Copeland, and to the staff of Academic Press for their willing help in the preparation of this volume.

G. CAUSEY WHITTOW

CONTENTS OF VOLUME I

Aquatic Invertebrates

F. JOHN VERNBERG AND WINONA B. VERNBERG

Terrestrial Invertebrates

J. L. CLOUDSLEY-THOMPSON

Fish

F. E. J. FRY AND P. W. HOCHACHKA

Amphibia

BAYARD H. BRATTSTROM

Reptiles

JAMES R. TEMPLETON

Birds

WILLIAM R. DAWSON AND JACK W. HUDSON

AUTHOR INDEX–SUBJECT INDEX

Chapter 1

RODENTS*

J.S. Hart

Publisher Summary

This chapter discusses the thermal environments of rodents. The rodents as a group consist of thermally sensitive animals of small body weight, which inhabit all regions of the earth from the tropics to polar region and from sea level to high-mountain climates. In arctic and subarctic climates, small animals are insulated from extreme conditions by their behavior in seeking warm microclimates, but these climates still appear to impose considerable metabolic burden particularly during foraging where subfreezing conditions are encountered. In cold climates, it is abundantly clear that the animals are exposed to much lower temperatures than in warm climates, in spite of behavioral tendencies tending to offset environmental influences. Studies on behavioral tendencies—such as huddling, nest building, and temperature selection—have shown the effectiveness of temperature compensation by this means and the seasonal variation in these qualities. The ability to tolerate elevations in body temperature in the heat and depression of body temperature in the cold is adaptive to variable conditions of environment, particularly in hot arid regions. Variability of body temperature is a mechanism for conserving energy and water. Examination of the effect of activity on temperature regulation shows that many rodents tend to increase activity with a fall in temperature and to reduce it greatly in extreme cold.

I. Introduction

II. Thermal Environment of Rodents

A. Arctic and Subarctic Regions

B. Cool Temperate Regions

C. Desert and Warm Climates

D. Special Environments

III. Thermal Behavior

A. Huddling and Nesting

B. Temperature Selection

C. Heat Reinforcement Behavior

IV. Precision of Homeothermy

A. Periodic Variations in Body Temperature

B. Body Temperature in Relation to Ambient Temperature

C. Body Temperature and Climate

V. Temperature Regulation in Cold Environments

A. Minimal Metabolism and Body Size

B. Newton’s Law, Critical Temperature, and Critical Gradient

C. Heat Production and Loss below Thermal Neutrality

D. Influence of Diet

E. Thermal Conductance

F. Thermal Limits

VI. Temperature Regulation in Warm Environments

A. Heat Production and Body Temperature

B. Partition of Sensible and Evaporative Heat Loss

C. Components of Pulmocutaneous Loss

D. Peripheral Heat Loss and Blood Flow

E. Acclimation to Heat

F. Thermal Limits

G. Heat Death and Tissue Injury

VII. Activity and Temperature Regulation

A. Amount of Activity at Different Temperatures

B. Effect of Activity on Body Temperature

C. Effect of Activity on Oxygen Consumption

D. Activity and Cold Resistance

E. Diurnal Metabolic Rhythms

VIII. Altitude, Hypoxia, and Temperature Regulation

A. Acute Effects

B. Chronic States

IX. Sex Differences in Temperature Regulation

X. Summary and Conclusions

References

I Introduction

The study of temperature regulation has been pursued more intensively in the rodents than in any other group during the last 25 years, particularly during the last decade. This interest has stemmed partly from the convenience in methodology for study of animals of small body size (10¹−10⁴ gm) but primarily from the great diversity in regulatory responses within the group. The latter aspect has attracted much attention from the comparative physiologists interested in hibernation, and in behavioral and physiological regulations during active states in relation to the great diversity of ecological and climatic conditions. These studies have shown that it has not been possible to understand the regulations in different species without knowledge of microclimatic conditions of their habitats and of the behavioral features of the species in question. While the characteristics of torpidity and hibernation will not be considered in this chapter, the microhabitat conditions and behavioral aspects of rodents will be analyzed, since these aspects are an important component of the total ability of a species to insulate itself from its environment.

While it may be concluded that the major interest of the zoological interpretation of temperature regulation has been in relation to the behavior of animals in nature, the rodents have been of primary interest to physiologists in the interpretation and understanding of the mechanisms of temperature regulation. Much of the work in this area has been on laboratory animals such as the white rat in which so much recent attention has been devoted to the study of responses and acclimation to cold. Much less study has been devoted to the acclimation of animals to heat.

In this chapter an attempt will be made to determine generalities common to different species and also to emphasize the species differences. The common factors include the general responses of heat production and loss in relation to body size, conductance, and fur insulation, as well as the general features of body temperature changes and thermal limits. To illustrate the theoretical aspects and definitions of some of these relationships, Fig. 1 shows some hypothetical curves of heat production at various air temperatures to illustrate tolerance and resistance limits, time changes, and acclimation.

FIG. 1 Metabolic response patterns and lethal limits in relation to time of exposure and acclimation. The zone of tolerance (TOL) is a temperature range in which no lethal effects take place, and the zones of resistance (RES) are ranges in which death results after varying periods of time depending on temperature of exposure. Response patterns also change with increasing length of exposure from t1 to t2. Acclimation to cold shifts the tolerance and resistance zones to lower temperature by change of slope of response curve from A to B or increasing the maximal response from W to C. Acclimation to heat may shift the zones of resistance and tolerance to the right by lowering the basal metabolism and changing the heat response from C to W. TB, body temperature; Lc, lower critical temperature; Uc, upper critical temperature.

In the left-hand section (Fig. 1) the thermal tolerance range (TOL) is defined as a zone within which no lethal effects of temperature occur after any period of time, while the zone of resistance (RES) eventually kills all the animals exposed to these temperatures. These definitions of thermal limits for heat and cold fellow those of Fry (1947) as used by Hart (1957, 1961). Within the zone of tolerance there is in almost all rodents a thermoneutral range in which heat production is constant, body temperature may or may not be constant, and which is bounded on its upper and lower limits by the upper (Uc) and lower (Lc) critical temperature limits. In this diagram TB is shown as a body temperature range which increases slightly with elevation in heat production above Uc. It is assumed that within the zone of resistance both TB and metabolism (M)* become time dependent, ending in the death of the animal with passage of time from t1 to t2.

Similarly, below Lc heat production in the zone of tolerance is assumed to increase in direct proportion to the temperature differences between TB and TA (air temperature) according to Fourier’s law of cooling (Scholander et al., 1950c). Below Lc thermal conductance [defined in the literature as Ms/(TB — TA)] is constant, while above Lc in the thermoneutral zone it increases with constant TB. In the zone of resistance for cold, oxygen consumption and body temperature are again time dependent. The lower the temperature, the faster TB and M fall; therefore determination of the maximal heat production (t1) can only be measured after a very short exposure, before M is reduced by hypothermia to t2 after a longer exposure period (Depocas et al., 1957).

In the right-hand section of Fig. 1 are shown some possible shifts with acclimation to heat or cold. As defined by Hart (1957), acclimation to cold and increase of cold resistance, as indicated by the left pointing arrow, could take place by increase of insulation (lowering of conductance) through a change in slope from A to B (insulative acclimation) without change in maximal heat production (W). Alternatively cold resistance could be increased without change in slope but with increase in capacity (C).

It will be shown that acclimation to high temperatures appears to take place in rodents by the lowering of the thermoneutral heat production. Consequently an increase in heat resistance, indicated by the right pointing arrow, could result from elevation of the Uc and lowering of the metabolism from C to W. It is also possible that acclimation to heat may produce greater resistance to hyperthermia, as suggested by the elevation of TB.

The theory and speculation raised by these diagrams will be assessed in the forthcoming sections on temperature regulation at high and low temperatures, particularly at the thermal limits. Both comparative aspects and regulatory mechanisms will be discussed. Additional sections dealing with the effect of activity and the effect of hypoxia on thermal responses are also included. This review is intended to stress aspects of temperature regulation not considered in depth in other reviews on the subject (Hart, 1957, 1963; Hensel, 1959; Carlson, 1962, 1966; Johansen, 1962a; Morrison, 1962a; R. E. Smith and Hoijer, 1962; Gelineo, 1964; Folk, 1966; Witke (1966); Himms-Hagen (1967); R. E. Smith and Horwitz (1969) and to concentrate on the more recent studies in this area. Taxonomy of rodents has followed Ellerman (1941), Ellerman and Morrison-Scott (1951), and Hall and Kelson (1959). Terminology for symbols follows in part that proposed by Gagge et al. (1967).

II

Thermal Environment of Rodents

The worldwide distribution of rodents from arctic to tropical regions of the earth and from sea level to altitudes exceeding 15,000 ft imposes a very diverse range of environmental conditions, and necessitates a great diversity of behavioral and physiological adaptations. Within these broad geographical life zones the habitat specializations of rodents, as exemplified by fossorial types which normally do not leave their underground burrows and by aquatic types such as beavers and muskrats, impose further modifications of environmental conditions with important implications for temperature regulation. In order to provide some background for an understanding of temperature regulation in rodents it is necessary to examine the actual environmental conditions of the habitat, and particularly the thermal conditions of the microclimate to which they are exposed. This includes consideration not only of the temperatures in these habitats but also other physical and ecological conditions such as snow cover, water and humidity, insolation, and protection by tunnels and vegetation. For these considerations it is convenient to discuss and contrast arctic and subarctic, temperate, and warm (mostly desert) regions separately. Although these are overlapping zones in the strict sense, the microclimatic differences are generally quite distinct.

A ARCTIC AND SUBARCTIC REGIONS

The biological, as contrasted with the physical, delineation of the arctic and subarctic has been designated by Kimble and Good (1955) in terms of the 10°C isotherm for July and in terms of the 10°C isotherm for the four warmest months, respectively. The lower boundary of the arctic combines the boreal forest and taiga or tundra forest to form the subarctic zone (Schwartz, 1963), which corresponds well with the extent of general permafrost (Morrison, 1966). It is also a region defined in terms of day length, with extremes in energy input from summer to winter, which have both direct and indirect consequences for the energetics of rodents. In regard to the seasonal variations in light and temperature given for Alaska (Irving, 1964; Morrison, 1966), which are typical of this general region, changes in mean lower temperature from about —25°C in winter (extremes to —50°C) to 10°C in summer could not be endured by any rodent without protection. However, these regions have been successfully colonized by nine species of microtine rodents, one sciurid, and three aquatic species (Schwartz, 1963) in Siberia, and ten microtines, two aquatic rodents, and eight sciurids in North America, in addition to Dicrostonyx and Zapus (Hall and Kelson, 1959).

The insulative function of snow in providing warmer microclimates for the survival of small mammals in winter has been described mainly by Formazoff (1946) and Bashenina (1963) in Siberia, by H. M. Johnson (1951, 1957), Pruitt (1957), Weedfall (1963), and Morrison (1966) in Alaska, and also by Geiger (1950), Coulianos and Johnels (1963), and Dalenius (1963) in Europe. These authors indicate gradients through 20–30 in. of snow from —40°C or lower on the surface or ground snow interface on a cold winter day. The importance of the air space between ground and snow for small mammals is emphasized because it is in these lower layers that the smaller microtine rodents are mainly active. However, it is reported that some activity on the surface occurs at night in nocturnal Microtus and Clethrionomys at temperatures down to —20°C (H. M. Johnson, 1957). Temperatures in tunnel systems of lemmings in active subnivean feeding areas in a cold period were found to be as low as —23°C. The diurnal red squirrels were active on the surface at much lower temperatures in Alaska (Pruitt, 1958) and were commonly seen at temperatures down to —32°C (see Fig. 24).

FIG. 24 Winter activity in red squirrels in relation to ambient temperature. Closed circles indicate winter 1954–1955; closed triangles, winter 1955–1956; open circles represent winter 1956–1957. (Data from Pruitt, 1958.)

Seasonal variations in the microclimatic temperatures under snow cover indicate habitation by rodents in the 0°C to —20°C range in winter (Pruitt, 1953, 1957; Bashenina, 1963; Bashenina and Borovskaja, 1963; Coulianos and Johnels, 1963; Weedfall, 1963) and in the 0°C to 15°C range in summer (Table I). Considerably warmer temperatures than these could presumably be attained by nests and huddling (M. S. Johnson, 1926; O. P. Pearson, 1960b; Bashenina, 1963; Stark, 1963). However, any foraging would require exposures of individual mice to the temperatures indicated. This nocturnal activity has been noted in particular before the snow depth reaches 10–20 cm. Both Formazoff (1946) and Pruitt (1957) indicate that one of the most critical periods for small rodents during cold snaps is before arrival of an adequate snow cover, which may result in considerable mortality.

TABLE I

SELECTED THERMAL DATA IN RODENT MICROCLIMATES

aAir temperatures are taken in same microclimatic area as soil temperatures. Soil temperatures taken in unoccupied nest boxes or burrows at level occupied by rodents. Ranges give lowest recorded nocturnal and highest diurnal values.

The data in Fig. 2 show thermal conditions in air and in the soil at different times of the year in Alaska (Pruitt, 1957), south Michigan (Pruitt, 1959m) and Arizona (Vorhies, 1945). In Alaska [and also north Michigan (Pruitt, 1953)] soil temperatures (3 in. level) were largely higher than air temperatures during winter, but fell within this range of air temperature during winter in southern Michigan and Arizona (Fig. 2). These differences are related to the insulation of the ground provided by deeper snow in the northern locations. The damping of temperature variations by soil and snow is quite marked in the deeper soil layers of the taiga (Fig. 2). However, in spite of this protection, the small mammals living in these layers and in subnivean spaces (Fig. 2) are exposed to temperatures considerably lower than in more southerly environments.

FIG. 2 Seasonal variation in habitat temperatures in different climatic regions. Air temperature range given by shaded area. Soil temperature in rodent burrows or runways given by double lines bounding unshaded area (3 in. depth in Michigan and Alaskan taiga, 4 ft in Arizona desert in occupied nests of Spermophilus). -·-·, soil temperature at 9 in. depth; —, soil surface temperature. All temperature variations are grossly smoothed to show only general trends. Data: Arizona, Vorhies (1945); south Michigan woods, Pruitt (1959); Alaska, Pruitt (1957).

B COOL TEMPERATE REGIONS

The species from cool temperate regions are considered to lie south of the subarctic zones as defined above, but generally are within areas having winter precipitation in the form of snow, although this is light or absent for some of the Peromyscus areas in British Columbia (Hayward, 1965b). There is also generally an abundance of soil moisture (Pruitt, 1953, 1959; Daniel, 1964; Stark, 1963) and relatively high humidity in rodent burrows at all times of the year. The limited number of microclimatic records obtained throughout the year are for mostly the mice genera Peromyscus, Clethrionomys, and Microtus, and they show that there are marked seasonal changes in thermal conditions with a general level that is considerably warmer than those from subarctic areas (Table I). The temperatures obtained in mouse burrows (by allowing the mice to pull the temperature sensors into the nest area) by Hayward (1965b), or simply in soil in mouse areas, show that in the winter the lowest encountered temperatures range within narrow limits around the freezing point (Pruitt, 1959; Getz, 1961; Coulianos and Johnels, 1963; Hayward, 1965b). Minimal temperatures may be considerably warmer than this for some species, for example, Peromyscus sitkensis, and P. m. austerus in coastal British Columbia (Hayward, 1965b). As mentioned earlier mouse nests may have considerably warmer temperatures when occupied. Data obtained with various mice (M. S. Johnson, 1926; Bashenina, 1963; Stark, 1963; Daniel, 1964; Hayward, 1965b) have shown that temperatures of occupied nests are from 7° to 24°C higher than those of the immediate surroundings.

Ground temperatures in appropriate habitats, which appear to give good indications of the seasonal microclimatic temperatures to be expected in rodent burrows, show annual variations of less than 10°C in some areas (Hayward, 1965b, Table I). In wooded areas in northern Michigan (Pruitt, 1959) and in various habitats of Microtus and Synaptomys in southern Michigan (Getz, 1961) the annual range of soil temperature is about 20°C with summer temperatures of 15°–20°C. In less protected areas summer soil temperatures reach much higher levels and winter temperatures fluctuate more widely (Pruitt, 1953).

In all these areas the air temperatures in the microhabitats show much greater diurnal and annual variations than the relatively stable soil temperatures, particularly under a snow cover (Table I). However, these air temperatures are of some significance to foraging mice, since it is known that Peromyscus and other mice are active at night above the soil or snow surface (M. S. Johnson, 1926; Swan, 1952; Gentry and Odum, 1957; Beer, 1961). Under these conditions it would be expected that the mice would be exposed to temperatures down to —15°C or below in winter, and up to 30°C or higher in summer (Table I), with even greater thermal extremes being encountered occasionally.

In an important review of the Russian literature on ecology of the common vole, Microtus arvalis, Bashenina (1963) describes the thermal conditions in habitat in the summer and in winter under the snow. The temperature profiles taken through snow and soil show gradients from —25°C on the snow surface in March to —3°–5°C at the snow–soil interface, where small mammals exist in large aggregates of up to 10–29 mice. The temperatures in the aggregates are 10°C or more above that of the surroundings depending on the number of animals present. This paper should be consulted for further details of microclimates in nests and burrows in different seasons and geographic areas. The comparison of smoothed seasonal changes in microhabitat temperatures in a typical cool temperate forest area (Pruitt, 1959) with similar changes by the same author (Pruitt, 1957) in a subarctic area (Fig. 2) shows a generally similar annual variation at a considerably higher level. The general impression gained from these comparisons is that while considerable protection is afforded by the semifossorial habits of small mammals, especially under snow cover (M. S. Johnson, 1926; Coulianos and Johnels, 1963) and huddling in nests (Sealander, 1952; O. P. Pearson, 1960b; Muul, 1968), the thermal conditions encountered in active foraging animals would be well below the thermoneutral range of most small rodents at almost all times of the year in arctic and subarctic regions, as well as in most temperate microclimates during certain seasons of the year. While the winter microclimate seems somewhat less severe in cool temperate regions than in the subarctic areas studied it would appear that in both regions the microclimates would entail considerable expenditure of metabolic energy for temperature regulation in proportion to the time spent outside the nest in foraging or in other activities.

C DESERT AND WARM CLIMATES

While the areas discussed in this section lie generally south of the cooler regions discussed above, they overlap and lie geographically within the north temperate zone and are distinguished mainly on the basis of extremes of heat and aridity during summer months, in contrast to the cold and generally mesic microclimates previously considered. Precipitation in these areas is very low and there is an absence of snow in the winter. While winter temperatures may be well below freezing in certain localities, particularly in high altitude deserts, e.g., in habitats of the Sonoran deer mouse, Peromyscus m. sonoriensis (Hayward, 1965b), the winter climate is generally milder than that in the north temperate field and forest areas (Table I).

Seasonal changes in desert microclimatic temperatures of rodents measured in Arizona (Vorhies, 1945; Brown, 1968), in Iran (Misonne, 1959), the Sahara (Pierre, 1958), and California (A. K. Lee, 1963; Carpenter, 1966) and Utah (Brown, 1968) have shown pronounced diurnal fluctuations, which are greater and at a higher level than those in nondesert areas. One of the most extreme microclimates is that shown by Vorhies (1945) for the ground squirrel, Spermophilus tereticaudus, in Arizona (Fig. 2). While the annual range in burrow temperature was similar to that in Michigan, but at a higher level on the temperature scale, the diurnal fluctuations in air temperatures were greater and this imposed thermal conditions not encountered in the more mesic northern climates. Spermophilus and other desert ground squirrels are active during the hottest periods of the day (Vorhies, 1945; Hudson, 1964a) and are exposed to intense insolation and to soil surface temperatures reaching a maximum of 70°C in summer and 32°C during winter. When retiring to an extensive burrow system to dissipate the heat accumulated outside (Hudson, 1964a) the ground squirrel would be exposed to temperatures slightly below 30°C, which is well below the range requiring active defenses against heat.

Other desert rodents living under similar environmental conditions occur both in the desert regions of the southern United States and in the Sahara. In the United States the various pocket mice (Perognathus) species and species of kangaroo mice, Dipodomys (Vorhies, 1945; Carpenter, 1966), and in the Sahara the jerboa, Dipus aegiptius* (Kirmiz, 1962), and the jird, Meriones unguiculatus (P. F. Robinson, 1959), may evade exposure to the most extreme heat by remaining in cooler burrows during the day (see Table I) and by nocturnal activity at lower temperatures. Some of these species live in areas of sparse vegetation and can thrive on dry food, requiring no added water in the diet (B. Schmidt-Nielsen and Schmidt-Nielsen, 1951; K. Schmidt-Nielsen, 1964a; Kirmiz, 1962). This ability to survive with a very limited water supply is aided by the relatively high humidity in the burrows compared to that in the outside air and by a remarkably efficient kidney as shown for Dipodomys (K. Schmidt-Nielsen et al., 1948).

Avoidance of high temperatures by the use of burrows and by nocturnal habits is also an adaptation of Peromyscus m. sonoriensis to air temperatures exceeding 40°C in summer. This species has a relatively stable burrow temperature of 25°–27°C in summer and 3°–5°C in winter, but it may be exposed to air at 0°C and —10°C while foraging at night (Hayward, 1965b). The wood rats, Neotoma lepida and N. fuscipes, are also noctural but have a very loosely constructed nest without a true burrow system (A. K. Lee, 1963). Both the humidity of these nests and the temperatures follow the environmental conditions much more closely than in species with extensive burrows (Vorhies, 1945; A. K. Lee, 1963). These well-ventilated nests result in exposure of these rodents to quite high temperatures and low humidities in spite of their nocturnal habits. In the white-throated wood rat, Neotoma albigula (Brown, 1968) a rather extensive tunnel system may be constructed. The temperature in the areas occupied by desert wood rats are not as high as those to which the ground squirrels are exposed owing to the presence of considerable cover of sagebush, creosote bush, cactus, Joshua trees, and other vegetation (Vorhies, 1945; Lee, 1963) on which it depends for its water requirements.

D SPECIAL ENVIRONMENTS

Most rodents are semifossorial, making use of burrows and nests to a considerable extent and foraging outside of these burrows during some portion of the day or night. However, two other groups, the aquatic rodents and truly fossorial rodents, occupy unique environments. In the former, as exemplified by the wide-ranging muskrat, Ondatra zibethica, and the more restricted beaver, Castor canadensis, the aquatic habit poses special problems of temperature regulation resulting from exposure to cold water rather than air. In the latter, the rodents virtually never leave the burrow systems, which are environments of almost constant temperature and low heat-dissipating power, owing to the high humidity and low convection and relatively high temperatures—up to 29°–31°C in Heterocephalus burrows according to McNab (1966).

Just as the protection of burrows under a snow cover has insulated most rodents from the marked winter cold in northern latitudes and permitted their extension into unfavorable regions of the earth, the aquatic environment has provided the thermal stability for beavers and muskrats necessary for a wide distribution in the most northern fresh waterways of the earth (Hall and Kelson, 1959). Thus, the winter temperatures experienced by these animals, being limited by the freezing point of water, must be relatively independent of latitude over much of their range. In beaver (Novakowski, 1967), there is winter foraging, especially in young animals, with evidence of reduced activity in this season in older animals. Winter foraging under the ice is evidently quite extensive in the muskrat even in northern areas such as the Mackenzie and Slave River delta (Law, 1949; W. A. Fuller, 1951; Stevens, 1955). The occupation of houses which are protected by a layer of snow apparently provides a warm refuge with temperatures in the range of 10°–20°C (Hart, unpublished observations), which are within the thermoneutral range of the muskrat in air. In water the thermoneutral range, however, is in the 25°–30°C range (Hart, 1962), which is higher than that of the foraging environment, particularly during winter. Under foraging conditions there is probably a considerable expenditure of metabolic energy for temperature regulation with some development of mild hypothermia (Hart, 1962; Shcheglova, 1963). However, any heat debt incurred could be repaid by return to the thermoneutral environment of the house. An important factor in the ability of aquatic mammals to cope with the greatly increased cooling power of water is a special property of their fur which, in the muskrat, holds a volume of air next to the skin equivalent to about 25% of the volume of the animal (Johansen, 1962c).

It may be concluded that rodents, because of their small size and thermal sensitivity, have made maximum use of burrows, nests, and other means to avoid exposure to environmental extremes that would not be tolerated without protection and this has allowed extension of their distribution ranges to almost all areas of the earth. Nevertheless, the protection by means of subnivean tunnels in arctic and subarctic and burrows and nocturnal habits in desert environments, as well as by various other behavioral processes, is obviously insufficient to insulate these rodents completely from temporal and spatial variations in their environments. The evidence for this is found in the geographic and seasonal adaptations in the temperature regulation of rodents to their environment which will be considered subsequently.

III Thermal Behavior

It may be appreciated from the foregoing discussion that small mammals can survive the rigorous variations in their microclimates only by appropriate behavioral responses which greatly reduce the physiological effects of heat and cold. A survey of some of the literature on huddling and nest-building, temperature selection, and behavioral thermal regulation will illustrate the importance and extent of some of these responses.

A HUDDLING AND NESTING

It has been noted by many authorities that Peromyscus and other small mammals tend to form aggregations and to build protective nests during winter and to disperse during warmer periods of the year (Nicholson, 1941; Thomsen, 1945; Burt, 1946; Stark, 1963; Hayward, 1965a; Muul, 1968). Nest-building and huddling are considered to be important behavioral adjustments to cold and to have major ecological significance (M. S. Johnson, 1926; Thomsen, 1945; O. P. Pearson, 1947, 1960b; Howard, 1951; Sealander, 1952). It has frequently been observed that rodents build much more elaborate nests in cold environments (Kinder, 1927; Sealander, 1952; Thorne, 1958) by increasing shredding behavior with paper and other materials. Barnett (1956, 1965) has described the progressively enhanced nest-building capability in mice living many generations at —3°C.

The view has sometimes been expressed that mice huddling in a nest may elevate their own ambient temperature to such an extent that a metabolic response to cold may not be necessary. However, measurement of the extent of the elevation in temperature has little quantitative significance since it depends on the position of the thermocouple in relation to the animals in question. A more meaningful approach has been to measure the effect of the nest and of the aggregation on heat production or on cold resistance.

The physiological effect of nest building was compared in the white-footed mouse, Peromyscus leucopus (Sealander, 1952), caught during summer. When provided with nest materials and placed at —32°C to —35°C without food or water summer animals survived 1.5–7.8 hours compared to 7–50 hours for winter mice. While part of the difference was associated with physiological superiority of the winter mice, part was also associated with an important behavioral difference in the ability to construct nests, which apparently varied seasonally. In summer mice the nest, however poor, was equivalent to an 18°C elevation of air temperature. Another approach taken by O. P. Pearson (1960b) with the harvest mouse, Reithrodontomys megalotis, was to determine the extent to which the nests lowered metabolism in the cold. In this species the lowering was about 17% at 12°C and 24% at 1°C. The effect was much greater with several mice than with one. In the round-tailed ground squirrel, Spermophilus tereticaudus, allowed to build a nest of cotton in a respiration chamber at 9°C, the oxygen consumption was reduced to an amount equivalent to raising the air temperature by about 15°C (i.e. to 24°C) which was close to the thermoneutral range for this species (Hudson, 1964a,b).

King and Connon (1955) have shown that the survival of C57BL/10 mice, 2 weeks old, depended significantly on proper thermal conditions which may be supplied by huddling. The importance of huddling on survival of mice in the cold was also observed by Sealander (1952) who showed that the survival time at —23°C was increased from about 1–3 hours to 3–10 hours by huddling, and that the increase was proportional to the number of mice in the group (varying from 1 to 4). The effect of huddling on the metabolic response to cold has been extensively studied by Ponugaeva (1960), who also examined the effect of distance between animals not actually in contact. O. P. Pearson (1947, 1960b), Prychodko (1958), Gorecki (1968), and others have determined the reduction in heat production caused by huddling. One example (Fig. 3) shown by Prychodko illustrates the reduction in food consumption and energy expenditure resulting from increasing the number of mice (MAf/SP males) in the group at different temperatures. Thus, with 5 mice, the food consumption at —3°C was 0.25 gm, per gram of mouse per day compared to 0.35 gm for 1 mouse, a reduction of nearly 30%, which is equivalent to an elevation in ambient temperature of about 11°C. In the example given by O. P. Pearson (1960b), huddling by 3 mice at 1°C reduced the oxygen consumption by 29% below that for 1 mouse. In the flying squirrel, Glaucomys volans (Muul, 1968), 3 squirrels together with nesting material reduced thermoregulatory energy by 33% compared to that for one exposed animal. The energetic studies indicated that the insulative value of wood and nest in trees inhabited by this species was sufficient to largely offset the increased heat loss in winter, so that winter and summer energy expenditures were about equal. The seasonal behavioral changes were photoperiodically determined.

FIG. 3 , 5 per cage. (From Prychodka, 1958.)

B TEMPERATURE SELECTION

A new physiological tool was developed when it was demonstrated by Herter (1936) that mice, when placed in a horizontal gradient consisting of a copper Bar heated at one end and cooled at the other, would select a range of temperature in which they apparently prefer to rest. Herter placed thermometers at regular intervals and showed that with a uniform gradient of about 5°C per inch, house mice selected a temperature averaging 34.6°C. The selected temperatures of rodents were found to be specific to the species (Herter, 1940) but declined with age. Later Bodenheimer (1941), with the same apparatus, obtained a selected temperature averaging 31.4°C for the house mouse, 3°C lower than that obtained by Herter.

Subsequent studies have shown that selected temperatures are not a fixed feature of the species but vary with age, previous climatic history, and various other factors. Baccino (1935), Bashenina and Boroskaja (1963), and Ogilvie and Stinson (1966a) have shown that the selected temperature of rodents declined sharply with age during the first weeks of life.

The influence of climatic factors on selected temperatures of rodents in horizontal gradients has been examined under changing constant temperatures, seasons and geographic location. Stinson and Fisher (1953) acclimated the same four individuals of deer mice, Peromyscus manicultatus bardii, to 22°, 33°, and 15°C and observed selected temperatures averaging 26.5°, 22.5°, and 25.5°C, respectively. In other words, the highest acclimation temperature gave the lowest selected temperatures. In the same study, it was concluded that the mice were selecting for floor temperature because there was no change in position when an air current was blown through the tube, destroying the gradient in air temperatures. A similar conclusion was reached by D. A. Smith and Fisher (1956) for the lemming, Dicrostonyx groenlandicus, which selected temperature in a gradient only when the fur was removed from the foot pads. Under these conditions the animals selected temperatures between 28° and 35°C, apparently as a result of stimulation of receptors in the feet.

Seasonal changes in preferred temperatures have been studied in the common mouse, Mus musculus (Ladygina, 1952), the common vole, Microtus arvalis (Mokrievich, 1966), the Eversman hamster, Cricetulus eversmanni, the gray hamster, Cricetulus migritorius (Mokrievich, 1965), and in the wood mouse, Apodemus sylvaticus and A. flavicollis, the yellow-throated mouse (Kalabukhov, 1939, 1953; Kalabukhov and Ladygina, 1953). The changes, amounting to 3°–7°C, are quite pronounced in some species, particularly in Microtus arvalis and Mus musculus (Fig. 4). The direction of the seasonal trends appears to parallel in a general way the seasonal changes in temperature, or even to anticipate these changes by several months. However, the variability in some of these tests was large and the significance of some of the differences is questionable.

FIG. 4 Temperatures selected in a gradient by mice at different seasons. Data: Mus musculus, Ladygina (1952); Microtus arvalia, Mokrievich (1966); Cricetulus migratorius, Mokrievich (1965); Apodemus flavicollis, Kalabukhov (1953).

Seasonal changes in preferred temperatures related to hibernation have been shown in hamsters by Gumma et al. (1967) using the method of choice of different temperature compartments by the animals. It was found that they chose a relatively cool (8°C) compartment prior to hibernation and a relatively warm one (24°C) following arousal.

Not only specific, but intraspecies and geographic differences were demonstrated by Kalabukhov and Ladygina (1953), who found that Apodemus flavicollis from the Crimea selected temperatures nearly 10°C lower than did those from Kharkov. When Crimean mice were raised for several generations at Kharkov the racial differences still remained distinct. Since the Kharkov area, from which the mice were obtained, was far to the north of the habitat for the Crimean mice, the southern (warmer) race apparently preferred lower temperatures than did the northern (colder) race. On the other hand, Microtus arvalis from more northern regions tended to select lower temperatures than did those from more southerly regions (Bashenina, 1963).

Considering the species, seasonal, and geographic differences in preferred temperatures, it is important to conduct comparisons of different species at the same season and with the same thermal history. When previous history was equalized by acclimation at 22°C (Ogilvie and Stinson, 1966b), the temperatures selected by mice in a horizontal gradient (5°–45°C) averaged as follows:

The standard deviations of these species increased with decrease in selected temperature, so that differences between Peromyscus m. bairdii and P. m. gracilis and between P. m. noveboracensis and Mus were not significant. The most interesting difference was found between P. m. gracilis and P. leucopus, which is in accord with the geographic difference in the range of these species, P. leucopus having a more southerly distribution. However, the correlation does not hold for P. m. bairdii. There are also considerable differences in cold resistance between P. leucopus and P. m. gracilis (Riedel, 1967). Klein (1960) had already suggested that surface soil temperature may be a significant factor in their differential habitat selection in central New York State.

The selected temperature appears to be one of the factors in the choice of a suitable microclimate by small mammals and to be an important ecological and behavioral characteristic of the species. However, the usefulness of temperature selection as a tool in precise physiological taxonomy is limited by its variability and the marked influence of the environment as noted for other physiological characteristics. In general, the relatively high levels of temperatures selected by most small mammals, close to the range of thermoneutrality, undoubtedly indicate an important behavioral tendency to select warm microclimates and to avoid conditions leading to high rates of energy expenditure.

C HEAT REINFORCEMENT BEHAVIOR

The ability of laboratory animals to press a bar activating a mechanism supplying heat with or without a delay period has been used to measure the effect of various treatments on this behavioral response at different temperatures. This has provided considerable information on the physiological factors involved in thermal behavior.

Work on heat reinforcement behavior began quite recently (Carleton and Marks, 1958) with the observation that cold exposed rats learned to press a switch that turned on a hot air blower. Subsequently various workers have shown that the frequency of heat reinforcement by rats pressing a bar for heat in the cold was increased prior to feeding (Hamilton and Sheriff, 1959), when food was deprived (B. Weiss, 1957; Hamilton, 1959), or when the thyroids were removed (Laties and Weiss, 1959). Administration of triiodothyronine produced a significant decrease in heat reinforcement rate in hypothyroid rats.

B. Weiss and Laties (1961) have described the behavior of clipped rats when pressing a bar for heat at different temperatures. Rats previously living at room temperature waited over 5 hours, with an 8°C fall in subcutaneous temperature, at 2°C, before learning to rewarm by pressing a bar. But those previously placed for 5 hours in the cold, started pressing it after only 2 hours, presumably because the subcutaneous temperature was initially low. The same authors (Laties and Weiss, 1960) found that rats placed at 2°C for 16 hours each day, for 31 days, pressed less for heat than did controls at 25°C, thus showing an effect of cold acclimation. The subcutaneous temperature fell to 34°C when first placed at 2°C for 16 hours but was maintained at 38°C to 39°C in the last 2 hours by bar pressing, which fell off as the subcutaneous temperature increased.

Another effect of cold acclimation was shown by Revusky (1966), who exposed hairless mice to about 10°C for 6+ hours each day for a 30-day period with opportunity for heat reinforcement. During the course of this training, the fall in skin temperature declined, and the frequency of heat rewards also declined. In contrast, Baldwin and