UNIVERSITY OF NEW SOUTH WALES Thesis/Project Report Sheet

SumameorFamilyname: ...... Jfl,.i;l].oueY, ...... First name: ...... 9.~~ ...... Othername/s: ...... J);~Y.ID......

Abbre vtation.or· · & d egreeasgtvenm· · theU ruvemtyca· · I e nd ar: ...... Ph.D...... Schoo); ...... f\i.gJgg;hg,~J.. ...?.s;;h~n9.~ ...... Faculty: .. J?.Jg,lgg:i,g.~.l .. liiDR. .. ~.. .a:v.iour.al .. Sr::iences...... TI&: .•...... •.• ~~.~~ .•.~~.

Abstract 350 words maximum: (PLEASE TYPE)

The emu, Ort?mBillS MI·'*Mllendiae, is a large ( 40 kg) flightless bird that inhabits areas as diverse in temperature regimes as the sno'w' country of the Great Dividing Range and the arid interior of the Australian continent. Several aspects of the emu's physiology suggest it is geared to life in the heat. The basal metabolic rate ( BMR) is lo'w'er than expected for a bird of its size. This may be a phylogenetic attribute, but it helps life in the heat by reducing energy needs and the endogenous heat load . The BMR and body temperature of resting males are lo'w'er than that of resting females . This may be a result of selection operating to reduce the energy needs of the male during his eight 'w'eek incubation fast in 'w'inter.

In hot conditions the emu maintains thermal balance by increasing evaporative 'w'ater loss, about 30% of'w'hich is cutaneous in origin. It avoids respiratory alkalosis 'w'hen panting by reducing tidal volume under mild heat stress. It is facilitated in its ability to remain active on hot, sunny days by the structural and optical properties of the plumage . less than 10% of solar radiation incident on the emu acts as a heat load at skin level. When 'w'i nd speed increases to 6 m/s the heat load at skin level is reduced to less than 1.5% of solar radiation. The heat load from radiation may be reduced further by the presence of a sub-cutaneous fat layer on the back in summer. When 'w'ater intake is restricted, osmoregulation is facilitated by a reduction i n the normal evaporative response to heat challenge. life in the co 1d is facilitated by the emu 's ability to i nc rease its oxygen extraction rate at 1O'w' ambient temperatures (T a), and by recuperative heat exchange in the nasal turbinates, 'w'hich allo'w's the emu to exhale air considerably cooler than body temperature at lo'w' Tas . These t'w'o factors acting in concert amount to savings of up to 20% of the energy consumption required to maintain body temperature at lo'w' ambient temperatures.·

Declaration relating to disposition of project report/thesis

I am fully aware of the policy of the University relating to the retention and use ofhigherdegreeprojectreports and theses. namely that the University retains the copies submitted for examination and is free to allow them to be consulted or borrowed. Subject to the provisions of the Copyright Act 1968, the University may issue a project report or thesis in whole orin part, in photostate ormicrofilrn orothercopyingmedium.

I~r~=T~,~~=~~~·t.•oct::~r:::•motioo&~::':~w1~:'~'. m m

The University recognises that there maybe exceptional circum tances requiring restrictions on copying or conditions on use. Requestsforrestriction for a period of up to 2 years must be made in writing to the Registrar. Reques oralongerperiod of restriction may be considered in exceptional circumstances ifa ccompanied by a Jetter of support from the Supervisor or Head of School. Such requests must be submitted with the thesis/project report.

FOR OFFICE USE ONLY DateofcompletionofrequirementsforAward: /'S. I I · (/'!, I I s ReKistrara!llDeputyPrincipal

TillS SHEET IS TO BE GLUED TO THE INSIDE FRONT COVER OF THE THESIS MBT 598.53041 1 Thermal Biology of the Emu

(Dromaius novaehollandiae)

by

Shane Kevin Maloney

Thesis submitted for the degree of Doctor of Philosophy to the University of New South Wales, Australia

July 1993 UNIVERSITY OF N.S.W. 2 o m::c 1993 LIBRARIES UNNERSITY OF NEW SOUTH WALES Thesis/Project Report Sheet

SumameorFamilyname: ...... ~.ialQUe.¥, ...... First name: ...... ~~~ ...... Othername/s: ...... J$§Y..;i;!.L ...... Abbreviationfordegreeas given in the University calendar: ...... ~!':.:.!?..~ ...... School: ...... l?.~g.:J.gg;hg§:f.. ... P.f1:1W.9.~ ...... Faculty: .. J2• .:i.glgg;;hg,?,.,l ... ?aP.Q, .. J?.OO~Y.iQW:.9..l .. SQieuc.es...... nt1e: ...... ~~.~~ ...~~S?.~9.9.¥. ...9.t ..t.h~ ... ~ ....mt.8U§i~~. noy~ehollancliaeL ......

Abstract 350 words maximum: (PLEASE TYPE)

The emu~ f)rQIMiU$ MI•'&?M/1Bndi8tl1 is a large ( 40 kg) flightless bird that inhabits areas as diverse in temperature regimes as the snow country of the Great Dividing Range and the arid interior of the Australian continent. Several aspects of the emu~s physiology suggest it is geared to life in the heat. The basal metabolic rate ( BMR) is lower than expected for a bird of its si2e. This may be a phylogenetic attribute, but it helps life in the heat by reducing energy needs and the endogenous heat load. The BMR and body temperature of resting males are lower than that of resting females. This may be a result of selection operating to reduce the energy needs of the male during his eight week i ncu bati on fast in winter. In hot conditions the emu maintains thermal balance by increasing evaporative water loss, about 30% of which is cutaneous in origin. It avoids respiratory alkalosis when panting by reducing tidal volume under mild heat stress. It is facilitated in its ability to remain active on hot, sunny days by the structural and optical properties of the plumage. Less than 1 0% of solar radiation incident on the emu acts as a heat load at skin level. When wind speed increases to 6 m/s the heat load at skin level is reduced to less than 1.5% of solar radiation. The heat load from radiation may be reduced further by the presence of a sub-cutaneous fat layer on the back in summer. When water intake is restricted, osmoregulation is facilitated by a reduction in the normal evaporative response to heat challenge. Life in the cold is facilitated by the emu's ability to increase its oxygen extraction rate·at low ambient temperatures (T a), and by recuperative heat exchange in the nasal turbinates, whtch allows the emu to exhale air considerably cooler than body temperature at low Tas. These two factors acting in concert amount to savings of up to 20% of the energy consumption required to maintain body temperature at low ambient temperatures.

Declaration relating to disposition of project report/thesis

I am fully aware of the policy of the University relating to the retention and use ofhigberdegree project reports and theses, namely that the University retains the copies submitted for examination andls free to allow them to be consulted or borrowed. Subject to the provisions ofthe Copyright Act 1968, the Universitymay issue a projectreportorthesis in whole orin part, in photostate ormicrofilm or other copying medium. ''""l~JJnl-icyMlaofilm•of o3S04nJ•""'"'inDh~"'lionA""""'"'"""tlonol(opp1Woblo "'""'J•••only), ...... : "...... : .. :;:: .:.... !...... 1 I 1...... w,;;;;; ...... !...... ?.!2/I:;.~:;; ......

The Universityrecognises that there may be exceptionalcircumftances requiring restrictions on copying or conditions on use, Requests for restriction for a period of up to 2 years must be made in writing to the Registrar. Reque;tJ;ralongerperiod ofrestriction may be considered in exceptional circumstances ifaccompanied by aletter ofsupport from the Supervisor or Head ofSchool. Such requests must be submitted with the thesis/project report.

FOR OFFlCE USE ONLY DateofcompletionofrequirementsforAward: /J. (I· li'S

Ige~straranaiJCuty JPri.l nCJpn

THIS SHBBTIS TO BEOLUBDTOTHBINSIDBFRONT COVBROFTHBTHBSIS 2

Table of Contents

TABLE OF CONTENTS 2 LIST OF FIGURES 7 LIST OF TABLES 12 DECLARATION OF ORIGINALITY 1 3

ACKNO~DGEMENTS 14 ABSTRACT 16

Chapter One: Thermoregulation in Birds: a Review 1 8 1.1 What is Thermoregulation? 1 8 1.2 Why Thermoregulate? 1 9 1.3 How is Thermoregulation Achieved? 21 1.4 Metabolism 22 1.5 Insulation 27 1.6 The Respiratory System 3 2 1.7 In the Cold 38 1.8 In the Heat 41 1.9 Control 4 6 1.10 Dehydration 4 8 1.11 Brain Temperature 48 1.12 The Case in Point, The Emu 50

Chapter 2: Basal Metabolism and Body Temperature 53 2.1 Introduction 53 2.2 Materials and Methods 55 2.2.1 Experimental· Animals 55 2.2.2 Metabolic Measurements and Procedure 55 2.2.3 Continuous Body Temperature Measurement 57 3

2.2.4 Statistical Analysis 57 2.3 Results 58 2.4 Discussion 60 2.4.1 Sexual Dimorphism in BMR 60 2.4.2 Body Temperature 61 2.4.3 BMR - Ratite Allometry 62

Chapter 3:Thermoregulation in the Emu 6 5 3.1 Introduction 6 5 3.2 Materials and Methods 6 5 3.2.1 Experimental Animals 6 5 3.2.2 Metabolic Meast.frements 6 6 3.2.3 Experimental Procedure 6 8 3.2.4 Continuous Body Temperature Measurement 6 9 3.2.5 Measurement of Cutaneous Evaporative Water Loss 69 3.2.6 Data Analysis 70 3.2.7 Surface Area 71 3.2.8 Statistical Analysis 72 3.3 Results 73 3.3.1 Body Temperature 73 3.3.2 Metabolism 74 3.3.3 Respiratory Quotient 78 3.3.4 Evaporative Water Loss 79 3.3.5 Cutaneous Evaporative Water Loss 80 3.3.6 Conductance 81 3.3.7 Surface Temperature 82 3.3.8 Heat Balance 84 3.4 Discussion 86 4

3.4.1 Body Temperature 86 3.4.2 Metabolism 88 3.4.3 Conductance 89 3.4.4 Evaporative Water Loss 94 3.4.5 More Female-Male Differences 97 3.4.6 Heat Balance 98 3.4.7 Legs as Radiators 99

Chapter 4: Ventilatory Accommodation of Oxygen Demand and Respiratory Water Loss 1 0 1 4.1 Introduction 1 01 4.1.1 In the Cold 102 4.1.2 In the Heat 1 03 4.2 Materials and Methods 1 0 3 4.2.1 Experimental Animals 1 03 4.2.2 Data Collection 1 0 3 4.2.3 The Barometric Method 1 04 4.2.4 Expired Air Temperature 1 0 6 4.2.5 Data Analysis 1 0 6 4.2.6 Statistical Analysis 1 0 7 4.3 Results 1 0 7 4.3.1 Ventilation 1 0 8 4.3.2 Gular Flutter 113 4.3.3 Sitting and Standing in the Cold 11 5 4.3.4 Expired Air Temperature 11 7 4.4 Discussion 11 7 4.4.1 Thermoneutral Ventilation 11 7 4.4.2 In the Cold 11 9 4.4.3 Respiratory Anatomy and Gas Exchange 1 22 5

4.4.4 Expired Air Temperature 1 24 4.4.5 In the Heat 125

Chapter 5: The Effect of Dehydration on Thermoregulation 1 28 5.1 Introduction 1 28 5.2 Materials and Methods 1 2 9 5.2.1 Preliminary Experiments 130 5.2.2 Experimental Procedure 131 5.2.3 Continuous Body Temperature Measurement 1 31 5.2.4 Blood Analysis 131 5.2.5 Statistical Analysis 1 3 2 5.3 Results 1 3 2 5.4 Discussion 138 5.4.1 The Effect on Body Mass and Plasma Osmolality 1 3 8 5.4.2 The Effect on Response to Heat Exposure 1 3 9 5.4.3 Protecting the Brain 1 4 0 5.4.4 Body Temperature Control 142

Chapter 6: The Heat Load from Radiation 143 6.1 Introduction 143 6.1.1 Animal Colour 143 6.1.2 The Heat Load from Radiation 144 6.2 Materials and Methods 1 4 7 6.2.1 Sample Collection 14 7 6.2.2 Measurement of Plumage Spectral Reflectivity 1 4 8 6.2.3 Measurement of Thermal Conductance 148 6.2.4 Measurement of Heat Load from Radiation 1 51 6

6.3 Results 153 6.3.1 Reflectivity 153 6.3.2 Thermal Conductance 154 6.3.3 Heat Load from Radiation 155 6.4 Discussion 157 6.4.1 Thermal Conductance 157 6.4.2 The Theory Relating Coats to the Heat Load from Radiation 157 6.4.3 What it Means to an Animal 158 6.4.4 A Simplified Model 159 6.4.5 The Effect of Convection 161 6.4.6 What's It All Mean to an emu? 162

Chapter 7: Conclusion 166

REFERENCES 170

APPENDICES 202 Appendix 1 List of Abbreviations 202 Appendix 2 Summary of the Model of Cena and Monteith (1975) relating Radiative Heat Load to Coat Structure and Colour 205 7

List of Figures

Page Figure 1.1 A plot of the typical endothermic response to different ambient temperatures, identifying the lower critical temperature (LCT) and the thermo neutral zone (TNZ) ...... 2 9 Figure 1.2 Diagram of the basic features of the avian lung depicting the air sacs (A=cervical, B=interclavicular, C=anterior thoracic, D=posterior thoracic, E=abdominal) and bronchi (1 ,2=primary bronchus, 3=ventrobronchi, 4=dorsobronchi, S=laterobronchus, 6=parabronchi). After Duncker (1972) ...... 3 3 Figure 1.3 Diagram showing direction of airflow in parabronchial lung during inspiration and expiration. During inspiration pressure drops in all air sacs, and increases in all air sacs during expiration (after Schmidt-Nielsen 1975) ...... 3 5 Figure 1.4 Density and distribution of emus in Australia (after Grice eta/. 1985) ...... 51 Figure 2.1 Basal metabolism (A), body temperature (B), and body mass (C) of female and male emus in summer and winter (Stars denote significant differences between the sexes in the same season, hashes denote significant differences between summer and winter in the same sex) .. 58 8

Figure 2.2 Twenty four hour body temperature record of female (dashed line) and male (solid line) emus. Shaded areas show SEM ...... 59 Figure 3.1 Flow rate of dry air into the metabolism chamber, and the resulting chamber vapour pressure of H20, at the ambient temperatures used in this study ...... 6 7 Figure 3.2 Areas of the emu where surface temperature measurements were made (after Eastman

1969) ...... 68 Figure 3.3 Body temperatures of female and male emus taken at the completion of metabolism experiments in winter and summer (different letters on the same sex denote significant differences between means; SNK P<0.05: Stars on the x-axis denote a significant difference

between the sexes; SNK P<0.05) ...... 7 4 Figure 3.4 Measurements of body temperature (B), evaporative water loss (C), and oxygen consumption (D), of an emu exposed to ambient temperature (A) increasing to 45°C ...... 7 5 Figure 3.5 Metabolic rates of female and male emus at a range of ambient temperatures in summer and winter ...... 76 Figure 3.6 Metabolic rates of female and male emus at a range of ambient temperatures in winter (different letters on the same sex denote significant differences between means; SNK

P<0.05) ...... 77 9

Figure 3.7 Respiratory quotient (VC02 I V02) versus

ambient temperature in winter (different letters denote significant differences between means; SNK P<0.05) ...... 7 9 Figure 3.8 Evaporative water loss versus ambient temperature for female and male emus in winter (different letters on the same sex denote significant differences between means; SNK P<0.05: Stars on the x-axis denote a significant difference between the sexes; SNK P<0.05) ...... 80 Figure 3.9 Emu conductance in winter expressed as watts/m2.oc (A) and watts/kg.°C (B). For explanation of symbols in Figure 3.9A, see text. (different letters denote significant differences between means; SNK P<0.05) ...... 81 Figure 3.10 Surface temperatures of various areas of emus in relation to ambient temperature (different letters denote significant differences between means; SNK P<0.05) ...... 8 3 Figure 3.11 Partitioning of heat loss by evaporative and non-evaporative means in winter and summer . 8 4 Figure 3.12 Leg radiative heat exchange as a percent of total radiative heat exchange (different letters denote significant differences between means; SNK P<0.05) ...... 8 5 Figure 4.1 Ventilation rate (A), tidal volume (B), and oxygen extraction rate (C), of female and male emus at a range of ambient temperatures ...... 1 0 9 1 0

Figure 4.2 Ventilation of female and male emus at a range of ambient temperatures (different letters on the same sex denote significant differences between means; SNK P<0.05: Stars on the x­ axis denote a significant difference between the sexes; SNK P<0.05) ...... 110 Figure 4.3 Oxygen consumption (A), ventilation frequency (8), tidal volume (C), and oxygen extraction (D), of emus at a range of ambient temperatures (different letters denote significant differences between means; SNK P<0.05) ...... 11 2 Figure 4.4 Emu head dissected to: expose the nasal turbinates ...... 11 4 Figure 4.5 Oxygen consumption (A), ventilation frequency (8), tidal volume (C), and oxygen extraction (D), of emus at 2soc and sitting and standing at - soc (different letters denote significant differences between means; SNK P<0.05) ...... 11 5 Figure 4.6 Relative contribution of ventilatory variables in the accommodation of increased oxygen consumption in sitting and standing emus at - 5°C, compared to 25°C ...... 11 6 Figure 4.7 Emu expired air temperature at a range of ambient temperatures ...... 11 7 Figure 5.1 Mass loss during dehydration versus the duration of dehydration ...... 1 3 3 Figure 5.2 Twenty four hour body temperature record of a dehydrated emu ...... 1 3 4 1 1

Figure 5.3 Measurements of body temperature (B), evaporative water loss (C), and metabolic rate (D), of a dehydrated emu as ambient temperature (A) was increased to 45°C 136 Figure 6.1 Diagram of the wind tunnel I radiation source experimental apparatus (a=sheet of glass painted except for a hole directly above the plumage sample, b=wind direction, C=hot wire anemometer, d=plumage sample, e=heat flow transducer, f=water filled hot plate connected to circulating water bath) ...... 1 50 Figure 6.2 Relative spectral distributions of sunlight and Arri Daylight lamp ...... 152 Figure 6.3 Spectral reflectance of winter and summer emu plumage samples ...... 1 53 Figure 6.4 Thermal conductance of plumage samples from emus in winter and·· summer (different letters denote significant differences between wind speeds; SNK P

List of Tables

Page Table 1.1 The relative importance of cutaneous evaporative water loss, as a percent of total evaporative water loss, in birds at moderate to

high ambient temperatures ( * = dehydrated) 43 Table 4.1 Ventilatory responses of individual emus at an ambient temperature of 35°C ...... 11 1 Table 4.2 Ventilation parameters measured by plethysmography at an ambient temperature when breath frequency is··a minimum ...... 120 Table 4.3 Maximum oxygen extraction reported for twenty species of birds ...... 1 2 3 Table 5.1 Characteristics of the blood and plasma of hydrated and dehydrated emus ...... 1 3 4 Table 5.2 Comparison of variables of hydrated and dehydrated emus exposed to ambient temperatures of 25oc and 45°C ...... 1 3 7 Table 6.1 Sex of emus from which plumage samples were obtained, and reflectivity of those samples ...... 154 1 3

Certificate of Originality

I hereby declare that this submission is my own work and that, to the best of my knowledge and belief, it contains no material previously published or written by another person nor material which to a substantial extent has been accepted for the award of any other degree or diploma of a university or other institute of higher learning, except where due acknowledgement is made in the text.

(Signed) ......

Shane Maloney July 1993 14

Acknowledgements

The emu being a rather large and at times unfriendly animal, life would have been somewhat more difficult without the technical help I received. My heartfelt thanks to Adam McLean and Steve McLeod who helped a lot and laughed along the way. Ray Williams, Geoff Vaughan and Jan Nedved were always willing to help with animal capture at Cowan. Dr. Jill Hallam and Liz May lent advice for four years and proof read this thesis without nodding off (so they say). Dr. David Croft helped significantly with matters statistical (please excuse the old joke).

Many people gave of their time and expertise to help with various aspects of the project, and all for free. Dr.Noel Bignow of the C.S.I.R.O. at Lindfield calibrated our flowmeters, the people in Dr. Peter Greenaway's lab showed me how to drive a spectrophotometer, Dr. Andrew Stevens of the School of Physiology and Pharmacology lent me his osmometer, and most notably Prof. Steve Dain of the School of Optometry spent many hours measuring plumage reflectivities. The staff of the faculty workshop, Gerry, Vince, Ed, Bob, John and Arvin, built the metabolism chamber and were always available with advice and help on other DIY projects. Ernie Gardner of Gardner Industries taught me how to tan skins.

My deepest gratitude is extended to Prof. Mark Chappell who I had the good fortune to help as a T.O. in 1988 before all this started. He helped set up the system with a heap of his gear which he subsequently left here, and was always just an e-mail away with an answer or advice when any of the many problems occurred. 1 5

This thesis wouldn't exist if not for my supervisor, mentor and drinking partner, Prof. Terry Dawson. He seems to have a lot more faith in me than I do myself, and I hope this work in some way repays that faith. He is not a standover type of a supervisor but I never seemed to have quite enough rope to hang myself. I owe you a few beers.

To my family who were always there and didn't laugh when I told them what I did with emus and thermometers.

Lastly to the missus to be, Cynthia, whose letters gave me something to look forward to every ·week and whose promise gives me something to look forward to for the rest of our lives.

During this work I was in receipt of an Australian Post Graduate Research Award. Financial support for the project came from an Australian Research Council grant to T.J. Dawson. 1 6

Abstract

The emu, Dromaius novaehollandiae, is a large (40 kg) flightless bird that inhabits areas as diverse in temperature regimes as the snow country of the Great Dividing Range and the arid interior of the Australian continent. Several aspects of the emu's physiology suggest it is geared to life in the heat. The basal metabolic rate (BMR) is lower than expected for a bird of its size. This may be a phylogenetic attribute, but it helps life in the heat by reducing energy needs and the endogenous heat load. The BMR and body temperature of resting males are lower than that of resting females. This may be a result of selection operating to reduce the energy needs of the male during his eight week incubation fast in winter.

In hot conditions the emu maintains thermal balance by increasing evaporative water loss, about 30% of which is cutaneous in origin. It avoids respiratory alkalosis when panting by reducing tidal volume under mild heat stress. It is facilitated in its ability to remain active on hot, sunny days by the structural and optical properties of the plumage. Less than 10% of solar radiation incident on the emu acts as a heat load at skin level. When wind speed increases to 6 m/s the heat load at skin level is reduced to less than 1.5% of solar radiation. The heat load from radiation may be reduced further by the presence of a sub-cutaneous fat layer on the back in summer. When water intake is restricted, osmoregulation is facilitated by a reduction in the normal evaporative response to heat challenge. 17

Life in the cold is facilitated by the emu's ability to increase its oxygen extraction rate at low ambient temperatures (T a), and by recuperative heat exchange in the nasal turbinates, which allows the emu to exhale air considerably cooler than body temperature at low T as. These two factors acting in concert amo(Jnt to savings of up to 20% of the energy consumption required to maintain body temperature at low ambient temperatures. 1 8

Chapter One

Thermoregulation in Birds: a Review

During the evolution of life on this planet at least two groups of animals have fine tuned metabolic processes to the extent that the maintenance of a reasonably constant body temperature (Tb) is essential for life. In doing so, mammals and birds have effectively emancipated themselves from fluctuations in environmental temperatures; fluctuations which render other animals inactive during cooler periods. The evolution of temperature maintenance by metabolic adjustments (endothermy:·. Greek, endo =inside; therme =heat) has not, however, occurred without cost. Associated with the development of endothermy in these groups is an increase in resting metabolic rate (tachymetabolism) which aids in maintaining body temperature but makes life more energetically expensive. Simply to exist an average endotherm must consume five to ten times as much energy as a similar sized reptile (Bennett and Ruben 1979, Yates 1981, Else and Hulbert 1987). The benefits of endothermy, however, must outweigh the costs if for no other reason than the a posteriori fact that endothermic animals are surviving and reproducing.

1.1 What is Thermoregulation?

Etymologically, thermoregulation derives from two Greek words: thermos meaning heat and regulare meaning to control. In essence this is precisely what a thermoregulating animal does; it controls the heat content of the body. The identity of the precisely 1 9

controlled variable remains enigmatic. Suggestions for the controlled variable include the temperature of the body core, hypothalamus, and spinal cord, or it may be a combination of these temperatures (Simon et a/. 1986). In practice body core temperature is usually measured, although the regulation of this temperature may be serendipitous. If this control is such that the temperature of the animal's body does not fluctuate by more than 4°C, then the animal is classified as a homeotherm (Greek: homoio =same, Simon 1987). The poikilothermic animal's Tb is not controlled as rigourously and usually falls when ambient temperature falls.

1.2 Why Thermoregulate?

For a trait as energetically expensive as tachymetabolism to persist it must confer some adaptive advantage. The evolution of thermoregulation in birds has received little attention, hence this section will be dominated by mammalian studies. Two main scenarios have been suggested for the evolution of homeothermy: the 'niche expansion' and the 'aerobic capacity' models. A major adaptive advantage of true homeothermy is liberation from fluctuations in environmental temperature, such that habitat selection can be made on other than thermal criteria (Hochachka and Somera 1984). The evolution of insulation (fur and feathers) and tachymetabolism allowed homeotherms to exploit niches previously unavailable, both temporally and spatially. Crompton et a/. (1978) proposed that mammalian homeothermy was acquired in two steps, the first enabling small mammals to invade a nocturnal niche without an increase in resting metabolism, the second when 20

mammals acquired higher body temperatures and resting metabolic levels. Their data suggest that some mammalian groups (e.g. the lnsectivora) have retained that first level of metabolism. Dawson and Grant (1980) criticize the 'normalisation' procedure used by Crompton et a/. and urge caution in the interpretation of data after such procedures. Previously the data from different mammalian orders had been used to suggest that homeothermic and metabolic capabilities evolved in a gradual manner (Dawson 1973).

Some fish can be classified as 'endothermic'. These fish maintain the central nervous system at a constant temperature despite fluctuations in external temperatures (they differ from true endotherms in that their heat of endothermy is myogenic, while in birds and mammals skeletal muscle contributes little to basal heat production, Bennett 1991 ). Tunas also maintain some muscle and viscera at a constant temperature. Block et a/. (1993) present evidence that 'endothermy' in fishes evolved at least three times, and suggest that the common factor in selective terms was always the invasion of new thermal niches.

Bennett and Ruben (1979) argue instead that the major selective force was for increased rates of aerobic metabolism; this is the 'aerobic capacity' model for the evolution of endothermy. This argument essentially hinges on a link between the basal metabolic rate (BMR) and maximum aerobic capacity. If a lineage could increase its aerobic capacity independently of BMR, then any increase in basal metabolism will be energetically wasteful and unlikely to be favoured by natural selection. There is some evidence of such a link (see Bennett 1991 for review, Bozinovic 1992), 21

though it remains essentially a generality (Bennett 1991 ). Bennett and Ruben (1979) also argued that an increase in aerobic capacity may have required an increase in the 'leakiness' of membranes, which would require the ubiquitous Na+;K+ pump to increase activity to maintain the normal high intracellular concentration of Na+, and low concentration of K+ (relative to extracellular fluid) seen in animals. Else and Hulbert (1987) have shown that the permeability of cell membranes to Na+ and K+ is several times higher in a representative mammal than in a representative reptile. This generality is supported by data from other species (Hulbert and Else 1990). This aerobic demand may also account for the increase in mitochondrial· surface area (Else and Hulbert 1985) and increased permeability of the inner mitochondrial membrane (Brand et a/. 1991) seen in endotherms. Additional evidence for this 'aerobic' scenario can be found in the action of the thyroid hormones. Thyroidectomy leads to lower levels, and thyroid treatment to higher levels, of resting metabolism (Sokoloff 1971 ). Ninety per cent of the increased heat production in rat liver and skeletal muscle slices caused by thyroid hormone can be attributed to increased activity of the Na+ pump (lsmaii-Beigi and Edelman 1970). There may be a link between leaky membranes, ion transport, thyroid hormones and the evolution of endothermy.

1.3 How is Thermoregulation Achieved?

That endotherms thermoregulate is beyond question. The most pertinent question to the research in this thesis is 'how do these animals thermoregulate?' Faced with a change in ambient conditions, how do these animals respond so that their body 22

temperature is maintained? Responses can be classified into two major categories, behavioural and physiological. Behavioural responses to thermal stress involve avenues such as changes in posture, orientation to the sun or wind, or ultimately avoidance of the stress by searching for a favourable microclimate (Cabanac 1975). These behavioural responses serve to narrow the thermal regime an animal must respond to physiologically. Physiological responses are numerous. The physiological thermoregulation of birds forms the basis for the rest of this review.

1.4 Metabolism

At the core of any discussion on temperature regulation in endotherms is metabolism. All animals possess the metabolic machinery needed to produce heat to maintain body temperature (Hochachka 1974) but only the mammals and the birds utilize this full potential and are homeothermic. Interestingly these are the tachymetabolic groups, having an elevated intrinsic 'idling' metabolism (the BMR). Even when the elevated metabolism of tachymetabolism is not required to maintain Tb at moderate to high ambient temperatures (in fact a high metabolic rate can be a disadvantage) metabolism cannot usually fall below the BMR. Basal metabolic rate is defined as the heat production or oxygen consumption of an organism in a rested, awake, fasting and thermoneutral state (Simon 1987). In this elevated resting metabolism lies the cost of homeothermy.

The BMR of all tachymetabolic animals is not equal. It is well known that the larger species have a higher BMR than the smaller 23

ones. But expressed in terms of body mass, small tachymetabolic species have higher BMRs than larger species. Late last century French scientists supposed that this was due to the fact that heat dissipation from endotherms must be proportional to their free surface. Consequently, they argued, small animals must have a high relative rate of heat production to counter their greater heat loss (Schmidt-Nielsen 1979). If an endotherm's BMR is determined by surface area, we might expect BMR to scale with the 2/3 power of body mass. Support for this hypothesis was provided by Rubner in 1883 (cited in Schmidt-Nielsen 1979) who studied dogs ranging in size from 3 to 31 kg. This relationship became known as the surface law. Prior to 1930 two independent groups used the same data set on humans to prove and disprove the surface law (Kleiber 1932).

Kleiber (1932) gathered metabolic data on thirteen groups of animals, ranging in size from a 150 g ring dove to a 680 kg steer, in an attempt to establish a function of body size to which metabolism might be more proportional than surface area. Plotting the log of fasting metabolic rate against log body mass he obtained a straight line with a slope close to 0.75. Brody and Proctor (1932) plotted data for birds and obtained a similar relationship, but with a slope of 0.64. Subsequent analysis on larger data sets has revealed exponents for birds ranging from 0.75 to 0.68 (King and Farner 1961, Lasiewski and Dawson 1967, Aschoff and Pohl 1970, Daan et a/. 1989). In 1938 Benedict calculated the relationship for mammals from 20 g mice to 4 ton elephants. He confirmed Kleiber's relationship but claimed it was an artificial relationship and of no physiological significance. Kleiber (1961) responded that "whether 24

such a relationship has physiological significance depends on the physiologist, just as it depends on the listener whether he hears a symphony or a series of noises".

The desire to describe the patterns underlying form are the reason for the search for a symphony. J.S. Huxley in 1932 commented that if the 'consequential' changes in proportion which automatically accompany changes in body size could be described, then strictly adaptive changes could be identified. Thus the allometric regression line acts as a 'criterion of subtraction' (Gould 1978).

The statistical method used to obtain allometric equations has been questioned. According to Zar (1968) use of the least squares regression method to fit a curve to log-transformed data presents problems, and he suggested the use of an iterative analysis on the untransformed data. However, the residuals of the untransformed data are heteroscedastic and therefore in violation of the assumptions of the least squares method. The log-transformed data meet the assumptions and describe the data better than Zar's iterative analysis (Lasiewski and Dawson 1969). Other alternatives to least-squares are major axis and reduced major axis methods but these require that the errors in the measurement of the X and Y variables be equal. This assumption is usually violated in body mass/BM R analyses and so the least squares method is an acceptable alternative (McArdle 1988).

The search for reasons why a given tachymetabolic species possesses a given BMR has spawned much literature. The empirically derived allometric relationship of body mass and BMR 25

(BMR=a·Mb) involves two variables: a, the mass-coefficient and b, the mass-exponent. Most of the discussion of the allometric equation has concerned the mass-exponent and its proximity in different groups of animals. Heusner (i 982) claimed that the inter­ specific relationship hides species differences in the mass­ coefficient, which he concluded was different for each of seven species he analyzed. He also obtained different values of the mass exponent for intra- (0.67) and inter- (0.776) specific comparisons. Feldman and McMahon (i 983) were quick to recognise that the mass-coefficient is correlated with species mass and that the 0. 75 mass-exponent has a component accounting for this. Subsequently, Bennett and Harvey (i 987) have reported also that the exponent differs depending on the taxonomic level of the data set. Their exponents varied, from 0.68 for an analysis of the average of orders within a subclass, to 0.82 for species within a genera. Later Dann et a/. (i 989) obtained an exponent greater than one for individuals within a species. Why the exponent should differ for intra- and inter-specific comparisons, and for different taxonomic levels, may be related to the similarity in shape within species, but that species of different sizes respond to the selective limitations of gravity and inertia in different ways (Feldman and McMahon "1983).

In spite of the shortcomings mentioned above, body mass remains the primary determinant of a species' BMR. This may be a result of the fact that the size of tissues that are metabolically active when an animal is at BMR (e.g. heart and kidneys) is correlated with body mass (Daan et a/. i 990). 26

Phylogenetic analysis of BMR has revealed some patterns. Within the mammals, when comparing similar sized animals, the marsupials exhibit lower BMRs than do eutherians (Dawson and Hulbert 1970, McNab 1986b) and the monotremes have lower levels again (Dawson and Grant 1980). Additionally, some orders within the eutherians have BMRs as low or lower than the marsupials and significantly lower than other eutherian orders (Hayssen and Lacy 1985). Such variation is also seen among the birds. Passerine birds have BMRs significantly higher than non-passerines (Lasiewski and Dawson 1967, Aschoff and Pohl 1970). The members of the order Falconiformes exhibit BMRs lower than other birds in general (Daan et a/. 1989), and the oldest group of extant birds, the suborder ratiti, may also possess BMRs below the non-passerine level (Calder and Dawson 1978, Withers 1983).

Also relevant are ecological factors, which appear to exert enough selective pressure to alter a species' BMR from that predicted by body mass. McNab (1980b, 1986a, b) developed an extensive analysis on ecological correlates of mammalian BMR and later did the same for birds (McNab 1988). His analysis of mammalian BMR has been criticised on the grounds that his analysis is equally resolved by phylogenetic differences (Elgar and Harvey 1987, Harvey and Elgar 1987). In answering this criticism, McNab (1987) says "the principal reason why most members of a higher taxonomic group are physiologically similar is that they are ecologically, behaviourally, and morphologically similar". This could alternatively be interpreted by suggesting the reason higher taxonomic groups are ecologically, behaviourally, and morphologically similar is that they are physiologically similar. 27

Latitude also exerts an influence on BMR, presumably via the action of ambient temperature regimes (Kendeigh et a!. 1977, Weathers 1979, Hails 1983). It would appear that within phylogenetic groups the intrinsic level of metabolism tends to be preserved, while selection, acting in the form of ambient temperature or restrictions on energy intake imposed by diet, may alter this intrinsic rate.

1.5 lnsu lation

The development of endothermy required two major engineering changes to animals. Firstly, the metabolic machinery had to be made to run at higher levels and secondly, animals needed to be able to regulate the flow of heat between the environment and the body. Birds largely achieved this second requirement with the development of an effective insulating layer, the feathers. The selective pressure leading to the evolution of feathers is unknown. It may have been thermoregulatory (Ostrom 1974), though this could not explain the very long wing and tail feathers on the oldest known bird, Archaeopteryx. These may have evolved as social or threat displays, insect 'nets', or aids to gliding or flight (Ostrum 1979, Molnar and Archer 1984). Whatever the selective force for feather origin, the practicality is that they adequately trap a layer of air against a bird's body and provide insulation.

Heat flow between an animal and its surroundings is proportional to the temperature difference between the animal and the environment and the heat transfer coefficient (commonly referred 28

to as the conductance of the animal) (viz. HF=(Ta- Tb)·C: where HF=heat flow, T a=ambient temperature, Tb=body temperature, C=conductance). Since Tb is maintained relatively constant in endotherms, and maximum conductance is finite, there is a minimum T a at which HF equals BMR (the lower critical temperature, LCT. See Figure 1 .1). At T as below this, heat loss exceeds BMR. The endotherms response is to increase heat production to balance the loss.

In birds this increase is usually achieved by shivering. Evidence for non-shivering thermogenesis (NST) has been reported for cold­ acclimated 7 and 11 month old chickens (EI Halawani 1970), ducklings (Barre et a/. 1989), and king penguin chicks (Duchamp et a/. 1991 ). Birds are not known to possess brown adipose tissue (BAT), which is the site of NST in mammals, however a 'brownish' adipose tissue with multilocular fat cells surrounded by numerous capillaries has been found. These bodies, however, lack sympathetic nerve endings typical of mammalian BAT. It is thought that this tissue may play a role in rapid fatty acid release for fuel during shivering (Hissa 1988). In birds, NST is thought to result from the uncoupling of oxidation and phosphorylation in muscle (Duchamp et a/. 1991 ). Cold acclimated fowl have been shown to have increased thyroid activity, which leads to increased metabolism without shivering (Hahn et a/. 1966).

The increase in metabolic rate (MR) below the LCT can be measured and plotted as in Figure 1.1. The better the animal's insulation (the lower its conductance) the lower will be the LCT and intensity of metabolic response at lower temperatures. If conductance is 29

constant below the LCT, the slope of the regression of MR on Ta can be taken as C (ml 0 2/g·hr·°C), otherwise C can be calculated from measurements of MR, Tb and Ta (Herreid and Kessel 1967).

(I) > .,_J (IJ -- Metabolism I.... 0 ...... Evaporative o..(J) (IJ (J) water loss > 0 LL.I.....I I.... I.... 0 (I) Eo+-J (J) (IJ •.-- 3: -0 ..Cl (IJ ~ ,, -- .,_J : ' ' I Q) ~~~- 1 ' I ~~ :I: ··------!------~~~I ' ' ' ' ' I I :~TNz_.,..: Tb Ambient Temperature

Figure 1.1 A plot of the typical endothermic response to different ambient temperatures, identifying the lower critical temperature (LCT) and the thermoneutral zone (TNZ)

The conductance of both mammals and birds might be expected, from physical considerations, to fall as body size increases. The surface area to volume ratio falls as size increases (Schmidt­ Nielsen 1979); larger body size permits the carrying of thicker coats which provide greater insulation (Scholander et a/. 1950) and the relationship of air boundary layer thickness to radius of curvature changes, such that large animals have thicker air boundary layers (Aschoff 1971, as cited in Aschoff 1981 ). Several 30

studies have shown this to be the case (Herreid and Kessel 1967, Drent and Stonehouse 1971, Aschoff 1981 ). In addition to showing that conductance was related to body mass, Aschoff (1981) analyzed data according to phase of daily cycle. Conductance measured during the active phase of the daily cycle of a bird or mammal was on average 50°/o higher than during the quiet phase. This mimics the dependence of BMR on phase (Aschoff and Pohl 1970). The difference in conductance between the phases is probably related to higher levels of evaporation in the active phase, since dry conductance shows no pattern with phase (data from Drent and Stonehouse reanalyzed by Aschoff 1981 ). On average the conductance of mammals is 30% greater than that of birds (Aschoff 1981 ), in spite of the many selective pressures on feather evolution other than thermoregulation.

A change in conductance is one of the major responses by birds and mammals to changes in the thermal environment. In the short term, conductance decreases in cold acclimated animals (Bech 1980, Rintamaki eta/. 1983, Whittow 1986). Conductance also shows a pattern with latitude, being lower as latitude increases (Scholander et a/. 1950, Blem 1974).

The thermoneutral zone (TNZ) is defined as that range of ambient temperatures in which temperature regulation is achieved by control of sensible heat loss, without changes in metabolism or evaporative water loss (Simon 1987). When an animal is at rest within its TNZ, heat production occurs mainly in organs deep in the body. The rate of conductance of this heat to the environment is a function of the thermal conductivity of whatever it passes through 31

{Conductivity in W/m 2 ·°C: fat=0.209, skin=0.337, muscle=0.50, air=0.0239) and the distance. The low thermal conductance of fat is exploited by some birds as subcutaneous insulation, especially by penguins and ducks, which have to deal with the higher rates of heat loss to water than to air {Whittow 1986). Alternatively, heat is convected by the blood, from the heat producing organs to the periphery. Depending on the circumstances blood flow can be manipulated to either hold heat within the core (by peripheral vasoconstriction) or transfer it to the periphery and thus the environment {by peripheral vasodilation). Vasomotor adjustments are facilitated by counter-current heat exchangers (rete) in the blood supply to peripheral areas of ·many species. The vessels in rete are arranged so that warm blood flowing from the core passes close to, and exchanges heat with, cold blood returning from the extremities. In this way heat is transferred back into the core rather than lost from the periphery. Rete occur in the legs of many wading birds (reviewed by Whittow 1986) and the flippers and feet of penguins (Frost et a/. 1975). Manipulation of blood flow is well illustrated in experiments on herons and gulls. Steen and Steen (1965) showed that at T a= 1 0°C less than 10% of metabolic heat is lost from the legs immersed in cold water, while at T a=35°C almost all metabolic heat is dissipated via this route. Rete are also important in regulation at high temperatures (see Section 1.11 ).

At temperatures below the LCT, when vasomotor and insulation adjustments cannot retard heat loss to BMR levels, the endotherm's heat production is increased. This means the body's demand for oxygen increases, a demand which must be met by the respiratory system. 32

1.6 The Respiratory System

Aspects of the avian respiratory system have been reviewed many times (Salt and Zeuthen 1960, Schmidt-Nielsen 1971, Lasiewski 1972, Scheid 1979, Fedde 1986). Despite the complexities, the anatomy and functional aspects of ventilation are now well established.

In contrast to the alveolar lung of mammals, which is basically a hose in a well folded bag, the avian respiratory system is somewhat complex, involving air sacs and a rigid parabronchial lung. Duncker (1972) examined the· lung-air sac system of 155 species of birds and described the general features common to all birds (Figure 1.2). The trachea divides at the syrinx into two primary bronchi (1 ,2), which pass through the lungs and into the abdominal air sacs (E). As the primary bronchus enters the lung, four ventrobronchi originate dorsa-medially (3). From the caudal end of the primary bronchus, before the abdominal air sac, the seven to ten dorsobronchi are directed cranially (4). Connecting the dorsa- to the ventre-bronchi are the parabronchi (6), which are small, rigid tubes. The walls of the parabronchi are sheathed in spiral muscular bands and small bundles of muscle surrounding the openings to atria (1 00-200 Jlm in diameter). These atria lead directly to air capillaries anastomosing with each other in all directions, where gas exchange occurs. The air capillaries are interlaced with a similarly structured network of blood capillaries. The rigidity of the parabronchial walls and air capillaries limits the danger of collapse and allows air capillary diameter to be very 33

E

D

Figure 1.2 Diagram of the basic features of the avian lung depicting the air sacs (A=cervical, B=interclavicular, C=anterior thoracic, D=posterior thoracic, E=abdominal) and bronchi (1 ,2=primary bronchus, 3=ventrobronchi, 4=dorsobronchi, 5=1aterobronchus, 6=parabronchi). After Duncker (1972) small compared to mammalian air capillaries, and consequently the exchange surface area per unit volume is much larger in birds (Duncker 1972). This arrangement of parabronchi has been termed the paleo-pulmo (Duncker 1972). Of 155 species studied only the penguins and the emu have solely this organization. All other species possess an additional parabronchial net, termed the neo­ pulmo, located laterally and ventrally to the posterior primary bronchus. In birds at rest, gas exchange in the paleo-pulmo exceeds that in the neo-pulmo (Scheid, Fedde and Piiper 1989). 34

In addition to the two pairs of large caudal air sacs (D,E), there usually exist three anterior air sacs (the paired cervical [A] and anterior-thoracic [C] and the single inter-clavicular [B]) which always connect to ventrobronchi.

During respiration the air sacs act as bellows to ventilate the rigid lung. It is well established that air flows through the parabronchi in one direction only, from dorsa- to ventra-bronchi, during both inspiration and expiration (see Figure 1.3; Bretz and Schmidt­ Nielsen 1971, Schmidt-Nielsen 1975, Scheid 1979, Fedde 1986). The simplest explanation for the observed flow pattern would be the presence of valves, but none have been observed (Bretz and Schmidt-Nielsen 1971 ). The shape and curvature of the primary bronchus, and the angle of departure of the secondary bronchi from it, were thought to establish flow resistances and shunt air in the directions observed (Scheid et a/. 1972) though this has been shown to be an insufficient explanation (Butler et a/. 1988). Banzett et a/. (1991) found that a constriction of the primary bronchus, just cranial to the first ventrobronchus, accelerates inspired air and can account for the observed shunting, at least in the goose.

Airway resistances within the system can also be modified by changes in the tone of bronchial smooth muscle. King and Cowie (1969) changed parabronchial resistance to airflow by electrically stimulating the peripheral stump of the vagus nerve, and with the sympathomimetic drugs acetylcholine, pilocarpine and histamine. Adrenaline and noradrenaline usually caused parabronchial dilation (King and Cowie 1969). Ray and Fedde (1969) and Molony eta/. 35

I nspi reti on lung

Ex pi ret ion ... ..__ ...... _ ...... __ ... .

Figure 1.3 Diagram showing direction of airflow in parabronchial lung during inspiration and expiration. During inspiration pressure drops in all air sacs, and increases in all air sacs during expiration (after Schmidt­ Nielsen 1975)

(1976) showed that changing the inhaled gas, especially the C02 content, could also affect bronchial smooth muscle and modify airway resistance.

When gas exchange occurs in an alveolar lung, the air and blood in close contact tend toward equilibrium. Oxygen (02) and carbon dioxide (C02) flow down their partial pressure gradients, 0 2 from air to the blood, and C02 from blood to air. Both these processes are 36

limited by diffusion rates, 02 more so than C02 (Berger et a/. 1979). The partial pressure of 02 in arterial blood leaving the alveolus

(Pao2) cannot exceed that in the air leaving the alveolus (PEo 2).

Similarly the partial pressure of C02 in alveolar air (PEco2) cannot exceed the arterial C02 pressure (Paco2) (Dejours 1981 ). In the parabronchial lung of birds, however, PEco2 has been observed to exceed Paco2 (Schmidt-Nielsen et a/. 1969, Piiper et a/. 1970) which is not consistent with alveolar diffusion equilibrium. Schmidt-Nielsen et a/. (1969) explained this phenomenon using a counter-current model of blood flow to airflow in the parabronchus. In this way blood about to exit the parabronchus contacts air with high Po 2 and low Pco2 just entering, and diffusion is enhanced. Zeuthen (1942, as cited in Schmidt-Nielsen 1975) had proposed a cross-current model for blood flow in the parabronchus which could also explain the observations. Scheid and Piiper (1972) realised that if the direction of air flow in the parabronchus could be reversed, the counter-current system would become con-current and thus blood and air approach diffusion equilibrium as in an alveolar lung. On the other hand the integrity of a cross-current system would be maintained. Reversal of the air stream in duck parabronchi supported the cross-current hypothesis (Scheid and Piiper 1972).

The utilization of unidirectional air flow and cross-current blood supply in the avian lung provides for very efficient gas exchange.

As a consequence, arterial Pco2 in birds tends to be lower than in mammals (birds=28 mm Hg, mammals=40 mm Hg; Calder and Schmidt-Nielsen 1968). While the cross-current model would predict Pao2 to exceed PEo 2, this is not observed. This could be 37

explained if the ratio of air (ventilation, V) to blood (perfusion, Q) that comes into contact in the parabronchi is altered. Ventilation I perfusion (V/Q) inhomogeneities have been shown to exist in birds

and these have a stronger effect on 0 2 than C02 exchange (Berger et a/. 1979).

The evidence available indicates that birds remove a larger

proportion of 0 2 in each breath than mammals. Lasiewski and Calder (1971) used data on seven bird species to determine the allometric relationships of respiratory variables and compared them to the mammalian allometric relationships obtained by Stahl (1967). These equations predict that for a given size, birds have a lower breath frequency (fR) and higher tidal volume (VT), and that total ventilation (VI), is smaller in birds. Since the metabolism of non-passerines and mammals is similar, the birds must remove more oxygen from a given volume of air, that is they must have a higher oxygen extraction (E02). The data used by Lasiewski and Calder (1971) to calculate their allometric equation have been questioned because the ventilatory system serves both metabolic and evaporative ends in birds. This results in respiratory variables changing markedly with T a, even within the TNZ (Bucher 1985). This leads to the problem of what defines 'resting' or 'basal' conditions for respiratory parameters. Nevertheless, the data available for birds to date agrees with the generalities stated above (see next section). 38

1.7 In the Cold

The increased heat production of birds in response to cold requires

an increase in the supply of oxygen (V02). Each of the three variables involved in this supply has limits. Tidal volume is limited

by the volume of the respiratory system; E02 , which is the proportion of inspired oxygen that is ultimately used by the animal, is limited by diffusion parameters; and fR is limited by the time required for contraction and relaxation of respiratory muscles. There theoretically exists an infinite number of patterns within these limits. The pattern chosen may represent a minimum rate of energy expenditure by respiratory muscles (Dejours 1981) or reflect physiological, phylogenetic and/or ecological factors (Chappell and Souza 1988).

Two patterns of response to increased oxygen demand have become evident from the limited data available on avian respiratory patterns. Some birds accommodate the demand by increasing ventilation (VI), while E02 remains constant. In some other species,

E02 increases with increasing demand. Early studies on ventilatory responses to cold exposure measured tidal volumes by pneumotachography (Brent et a/. 1983, 1984; Bech et a/. 1984) while later studies on different species have utilized plethysmography. Barnas and Rautenberg (1984) speculated that the two different patterns might arise from the two different methods. Bech et a/. (1985) found the two methods gave the same pattern in pigeons, although the absolute levels were different. There appears to be no relation between the maximum reported E02 for a species and its utilization of either strategy. The maximum extraction for 39

those species which increase E02 at low temperatures does not exceed the extraction of those that maintain E02. The question then remaining is, why do the former species have a low E02 at intermediate temperatures? At high temperatures the respiratory system has the alternate responsibility for increased levels of evaporative water loss, which leads to low E02 (see section 1.8). This factor does not seem to account for the lower levels at intermediate temperatures in the relevant species.

When inspired air is warmed and saturated with water vapour in the body and then exhaled, significant heat loss can occur. The amount of air which needs to be inhaled to satisfy oxygen demand is decreased with increasing E02. Increased E02 at low temperatures can be interpreted as a heat saving mechanism (Bech eta/. 1985). Chappell and Souza (1988) questioned this interpretation on the grounds that at low temperatures vo2 also increased in the species studied, so increased E02 may be an oxygen demand response. They calculated that the reduction in heat loss due to increased E02 was only 3-6o/o in several species studied (parrots, European coots, chukar partridges) and that a 1.5 fold increase in E02 in adelie penguins (from the observed 30-35% to 52%) would only reduce total heat loss by 1.9% at -20°C. On the other hand the increase in vo2 associated with flight at thermoneutral temperatures, and thus uncoupled from a minimum heat loss demand, is achieved without increasing E02 in the crow (Bernstein 1976), evening gosbeak, ring-billed gull, and black duck (Berger, Hart and Roy 1970). Such a comparison is clouded by the high heat production and the need for increased evaporation from the respiratory system in flying birds. While increased E02 at low 40

temperatures is advantageous in heat preservation, whether this is the sine qua non for its existence is unclear.

Perhaps the principal adaptation of the respiratory system for heat conservation at low T a is the manipulation of expired air temperature (T ex). Just as the circulatory system exploits spatial counter-current heat exchange for energy conservation, so birds utilize temporal counter-current exchange, also called recuperative or regenerative heat exchange (Mitchell et a/. 1987), in the nasal passages. The exchange occurs since inhalation of cool air results in the cooling of the nasal mucosa as heat flows into the cooler air. In addition, the warming of the inhaled air increases its saturation vapour pressure, causing evaporation from the moist mucosal surfaces which further cools them. When air saturated at body temperature is exhaled over these surfaces, heat flows from the air to the pre-cooled mucosa. Cooling of this air lowers the saturation vapour pressure and water condenses on the mucosa. In this way there is conservation of both heat and water. In the cactus wren the water savings can be as high as 74%, compared to expiring air saturated at body temperature (Schmidt-Nielsen et a/. 1970). The degree of conservation depends on the anatomy of the nasal passages. Wide, open tubes such as in humans allow for relatively little heat exchange, while the narrow passages of animals such as the cactus wren and kangaroo rat allow for greater exchange (Schmidt-Nielsen et a/. 1970). Decreased T ex with decreasing T a has been demonstrated in many species, including prairie falcons (Kaiser and Bucher 1985), European coots (Brent eta/. 1984) and penguins (Murrish 1973). 41

While this process results from the exploitation of physical principles, it is also under physiological control, at least in penguins (Murrish 1973). Heat loading at an ambient temperature of 5°C (by irradiation), or administration of an a-adrenergic blocking

agent, increased T ex from 10°C to near body temperature. A pre­ requisite for effective heat retention is constriction of blood supply to the nasal passages, otherwise they are re-warmed by the blood before air expiration. Heat loading and blocking a-adrenergic receptors leads to vasodilation of these vessels.

1.8 In the Heat

Plasticity of Tb regulation is a principal adaptation to heat exposure in many bird species. When heat stressed, Tb rises in a controlled hyperthermia. This is interpreted as a means of increasing the gradient for dry heat loss (or reducing the gradient for heat gain) rather than a failure of the controlling system (Schmidt-Nielsen 1964, Schmidt-Nielsen et a/. 1957, Weathers 1981, Dawson 1982). Eventually, however, the animal will need to rid the body of metabolic heat, plus any heat flowing in from the environment, or suffer heat death. Under the circumstances when T a exceeds Tb the evaporation of water is the only means an animal has to dispose of excess heat and maintain a stable Tb. Whenever a moist surface is in contact with unsaturated air, water evaporates from that surface. This process is endothermic (in this case meaning the process requires energy) with 2420 Joules being required to evaporate one gram of water. Air has a low thermal conductance and a low thermal capacity compared to animal tissue, so most of this heat is derived from the animal. Panting is a means 42

of increasing the contact of unsaturated air with moist surfaces and is an almost ubiquitous (there being one notable exception) response of birds to heat stress. However it is not the only means birds have to increase evaporative water loss (EWL). When heat stressed, storks and turkey vultures urinate onto the legs in a process called urihydrosis. The urine then evaporates from the legs and aids in thermoregulation (Kahl '1 963, Arad et a/. '1 989).

Until twenty years ago the lack of sweat glands in birds fostered the notion that the only other evaporation of any significance was respiratory (e.g. see Richards '1 970a). In the last twenty years it has been shown that cutaneous evaporative water loss (CEWL) is also an important avenue in many species (see Table '1. '1 ). The one notable exception to the panting response is pigeons acclimated since hatching to high temperatures, which do not pant at 50°C. These pigeons can maintain Tb by increases in CEWL (Marder and Gavrieli-Levin '1 987). The loose packing arrangement of cells in the stratum-corneum of the pigeon is thought to facilitate the movement of water through the skin (Elias et a/. '1 98'1 ). In fact, 60o/o of the increase in CEWL in pigeons in another study (Webster et a/. '1 985) could be explained by the increase in skin temperature. The remainder was explained by changes in epidermal lipids, or increased hydration of keratin layers due to increased cutaneous blood flow. Recently Marder and Raber ('1 989) have shown that cutaneous blood flow in the pigeon may be under ~-adrenergic control. Injections or oral administration of propranolol enhanced CEWL in resting birds at a Ta of 35°C, but not at T a of 50°C. They suggest that the cutaneous blood system in birds is normally kept 43

Table 1.1 The relative importance of cutaneous evaporative water loss, as a percent of total evaporative water loss, in birds at moderate to high ambient temperatures ( * = dehydrated).

Species name Common Name Mass CEWL Ref

Poephila castanotis Zebra Finch 12.5 30 63 a Poephila castanotis Zebra Finch 14 25 48 b 25* 27 Melopsittacus undulatus Budgerigar 31.6 30 59 a Excalfactoria chinensis Painted Quail 42.3 30 45 a Ploceus cucullatus Village Wea"¥er 42.6 30 51 a Phalaenoptilus nuttallii Poor-will 43.2 35 51 c Geophaps plumifera Spinifex pigeon 89 40 80 k Streptopelia senegalensis Palm Dove 112 20 62 f 45 79 Coturnix coturnix japonica Japanese quail 1 1 8 20 82 f 45 82 Streptopelia decaocto Turtle dove 168 20 78 f 45 75 Geococcyx californianus Roadrunner 274.2 35 51 c Columba Iivia Pigeon 220 30 45 30* 40 45 99 45* 35 Columba Iivia Pigeon 285 20 86 f 45 59 Columba Iivia Pigeon 473 20 60 30 60 44

40 53 Alectoris chukar Partridge 475 20 85 f 45 81 Falco mexicanus Prairie Falcon male 498 40 >50 h female 755 40 >50 Gallus domesticus Domestic Fowl 2040 20 51 e 30 33 40 25 Pekin Duck 2500 20 68 d 30 54 35 39 Struthio came/us Ostrich 95400 20- 40 g 33 Key to References in Table 1.1: a) Bernstein 1971, b) Lee and Scmidt-Nielsen 1971, c) Lasiewski et a/. 1971, d) Withers and Williams 1990, e) Richards 1976, f) Marder and Ben-Asher 1983, g) Withers 1983, h) Kaiser and Bucher 1985, i) Arad et a/. 1987, j) Webster and King 1987, k} Withers and Williams 1990 under a constricting f)-adrenergic tonus. Blockade of f)-receptors thus enhanced skin blood flow and increased CEWL.

The most documented response of birds to heat stress is panting. Panting has been described as the most efficient means for an insulated animal to dissipate body heat by the evaporation of water. Panting removes heat directly from the body core and the animal's insulating coat then retards the flow of external heat into the body. Alternatively, evaporation from external surfaces can increase the gradient for heat flow in from the environment (Dawson 1973). 45

The drawback with panting is the potential for washout of C02 from the blood, leading to hypocapnia and alkalosis (e.g. see Calder and Schmidt-Nielsen 1968, Marder and Arad 1989). However, several mechanisms have evolved which help to minimise or prevent C02 washout. During moderate levels of heat exposure, many species increase fR but reduce VT. This reduction in VT preferentially ventilates non-gas exchange areas of the respiratory system, such that the parabronchi are not hyperventilated and blood gases can be maintained. This response was originally reported in mammals and termed Phase I panting, as opposed to the deeper ventilation (increased VT) of Phase II panting during severe heat stress (see Hales and Webster 1967). Subsequently this type of response has been reported for many bird species and has been termed 'simple panting'. The species that have been shown to exhibit simple panting during heat stress include the chicken (Brackenbury et a/. 1982), flamingo (Bech et a/.1979), duck (Bouverot et at. 1974), Abdims stork (Marder and Arad 1975), mute swan (Bech and Johansen 1980), and rock partridge (Krausz et a/. 1977). Some other species exhibit 'flushout' panting, whereby VT is reduced to a level less than the dead space volume and intermittent deeper breaths provide parabronchial ventilation. A third pattern, 'compound' ventilation, has been reported for pigeons, involving high frequency, low VT panting superimposed on low frequency, high VT breathing (Ramirez and Bernstein 1976).

In addition, at least two studies suggest that other mechanisms can operate. In the ostrich, a prolonged exposure to over 50°C resulted in an initial drop in VT below the resting value of 1.2-1.5 litres but later it increased to up to 2.6 litres without a 46

significant change in Paco2 (Schmidt-Nielsen et al. 1969). The authors postulated that a shunt mechanism may operate whereby air passes directly to and from the posterior air sacs via the primary bronchus, without an increase in flow to the parabronchi via the dorsobronchi. Jones (1982b) reported that serial inhomogeneities in ventilation/perfusion ratios (V/Q) also exist in the panting ostrich, but that this effect alone was insufficient to explain the maintenance of Paco2 , and so concluded that a shunt also existed. In studies on another ratite, the emu, Jones et a/. (1983) recorded the development of only mild hypocapnia with prolonged panting, and paradoxically a slight hypoxia. A shunt would lead to deficiencies in both 02 and C02 exchange, and so hypercapnia and hypoxia. The emu situation can be explained if serial inhomogeneities exist in the V/Q ratio in the parabronchi, such that blood flows only to the proximal end of the parabronchi. In this case the blood and air would be in contact for only a short time, C02 diffusing quicker than 0 2 could still washout, while 02 exchange would become time limited and hypoxia thus develop.

1.9 Control

The possession of efficient effector mechanisms is not an end in itself. Proper regulation of body temperature requires appropriate control of those effectors. Models of the control of body temperature involve some sort of integrating controller receiving input from thermosensors and, according to that input, sending impulses to the relevant effectors. Thermoreceptive function has been proposed for nerve endings in the skin, spinal cord, deep body, and hypothalamus of birds (Hissa 1988). Whether the hypothalamus 47

of birds is a primary thermosensor or its thermosensitivity a product of temperature dependant signal transmission (Hammel 1968) is not known. Apart from the hypothalamus, manipulation of the temperature of these thermosensors leads to activation of appropriate effectors. However, there exists strong inhibitory cross-overs somewhere in the control system, since the degree of effector response produced by altering one thermosensor depends on the state of the other thermosensors (Simon et a/. 1986). Whereas hypothalamic temperature is thought to provide the main impetus for thermoregulatory effector response in mammals, the situation in birds is complicated by weak or inappropriate responses to hypothalamic temperature manipulation. Cooling the hypothalamus results in no effect, or in a decrease in heat production, in the pigeon (Rautenberg et a/. 1972), adelie penguin (Simon et a/. 1976), California quail (Snapp et a/. 1977), and Pekin duck (Simon-Opperman eta/. 1978), while appropriate responses have been recorded for the sparrow (Mills and Heath 1972) and emu (Jessen et a/. 1982). Heating the pigeon hypothalamus inhibits panting (Schmidt 1976), while penguins (Simon eta/. 1976) and Pekin ducks (Simon-Opperman et a/. 1978) show an appropriate, though very weak, responses to hypothalamic heating. Nevertheless, a model similar to mammalian control models is appropriate if the hypothalamic 01 o (which is the ratio of the values of any variable at two temperatures 10°C apart) for cold signal transmission is much greater than for warm signal transmission (Schmidt and Simon 1982). Thus the difference between mammals and birds is lower sensitivity to cold of the avian hypothalamus, and consequently in birds, extrahypothalamic thermosensors seem to play a more important role (Hissa 1988). Spinal cord heating and 48

cooling elicits appropriate effector responses in birds and this input is thought to exert the dominant influence in avian thermoregulatory control (Rautenberg and Necker 1975).

1.10 Dehydration

The thermoregulatory control system does not act in isolation from the rest of the central nervous system. Competing demands from separate regulated systems may require compromise. For example during exercise in the heat, increased blood flow is demanded by the muscles for oxygen delivery and by the periphery for heat dissipation, which results in sub-optimal delivery to each system (Rowell 1977). Negative feedback from the osmoregulatory system, as during dehydration, results in an increased Tb threshold for evaporative heat loss effector response (Arad 1982, 1983, Arad et a/. 1985, 1987, Kleinhaus et a/. 1985, Peltonen et a/. 1989, ltsaki­ Giucklich and Arad 1992). This increase in threshold is thought to come about due to inhibition from interneurons serving both the osmo- and thermoregulatory systems, such that increased input from osmoreceptors increases the threshold for thermoregulatory output (Hori et a/. 1988).

1.11 Brain Temperature

Brain tissues are especially vulnerable to heat (Burger and Fuhrman 1964, Bowler and Tirri 1974) which limits the degree to which hyperthermia can safely develop. Many species are capable of regulating brain temperature below that of the body core, which increases the animal's tolerance to hyperthermia (Baker 1982, 49

Mitchell et a/. 1987, Elkhawad 1992). All birds thus far studied display a body-brain temperature differential of about 1°C in thermoneutral conditions. The species studied include the chicken (Richards 1970b, Arad et a/. 1984), rhea (Kilgore et a/. 1973), pigeon (Kilgore et a/. 1979, Bernstein et a/. 1979b, Pinshow et a/. 1982), American Kestrel (Bernstein et a/. 1979a), helmeted guinea fowl (Withers and Crowe 1980), zebra finch (Bech and M idtgard 1981 ), quail (Kilgore et a/. 1981 ), calliope hummingbird (Burgoon et a/. 1987), and Japanese quail (ltsaki-Giucklich and Arad 1992). This brain-body temperature difference is maintained when a bird is heat exposed but can be reduced if the bird is heat exposed when dehydrated (Arad 1983, ·· Arad et a/: 1984). The ability of birds to maintain this differential depends on evaporative cooling from the upper respiratory system, eyes, mouth, and nares. Cool venous blood draining from these regions enters the rete ophthalmicum, a vascular arrangement where counter-current heat exchange can occur between this cool venous blood and arterial blood destined for the brain (Bernstein et a/. 1979 a, 1979b, Pinshow et a/.1 9 82, Peltonen et a/. 1989). Brain cooling capacity is dependent on the surface area available for heat exchange between veins and arteries (Midtgard 1983, Midtgard et a/. 1983, Arad and Midtgard 1984, 1990). The increased surface area of the rete ophthalmicum in desert bedouin fowl partly explains their superior heat resistance over commercial breeds (Arad 1982, 1983). During dehydration evaporative cooling is reduced, at least in the fowl and Japanese quail, by a reduction in respiratory frequency. This reduction in evaporation from surfaces which supply cool blood to the rete results in a reduced body-brain temperature differential (Arad 1983, Arad et a/. 1984, ltsaki-Giucklich and Arad 1992). 50

1.12 The Case in Point, The Emu

The emu (Dromaius novaehollandiae) belongs to an ancient group of birds making up the suborder ratiti. It is thought that there were ratite ancestors widely dispersed on the Gondwanaland supercontinent and that after the breakup of that land mass, those ancestors evolved into the different ratites found on today's southern continents. These species are: in Africa the ostrich; in South America the rheas; in New Zealand the kiwis; and in Australia the cassowary and the emu (Cracraft 1974). Of the two Australian species, the cassowary is restricted to rainforests in the north of Australia, while the emu is dispersed widely over the continent. The habitat of the emu extends to areas as diverse in climate as the snow country of the great dividing range and the arid interior (Davies 1974, pers. obs.). The distribution and density of emus on the Australian continent, estimated from aerial surveys by Grice et a/. (1985), is shown in Figure 1.4. These authors estimate that at the time of the survey there were over 630,000 emus on the continent. The abrupt change from a high emu density in North-Western New South Wales, to a lower density in South­ Western Queensland and North-Eastern South Australia, is probably due to the exclusion of the predatory dingo from North-Western New South Wales (Caughley eta/. 1980).

The life history of the emu has been reported by Davies (1974). In late autumn to early winter (May-June) the female emu will lay up to 20 eggs which are incubated for approximately 56 days by the male. During incubation the male supposedly does not eat or drink. Aspects of emu incubation have been reported by Buttemer and 51

Dawson (1989) and Buttemer et a/. (1988). Approximately 56 days after the start of incubation, the striped chicks begin to hatch and all hatch within a few days of each other. The precocial chicks may remain with the male for up to 18 months.

Densi1y per squar• k1lometrt'

D

0.1-0.2

>0.2

NS Not Surveyed

Figure 1 .4 Density and distribution of emus in Australia (after Grice et a/. 1985) 52

The emu has low nitrogen and energy requirements (Dawson and Herd 1983). It possesses a simple digestive system, although it is capable of a high degree of fibre digestion (Herd and Dawson 1984). The kidneys are not capable of concentrating urine to the extent seen in some other desert adapted birds (Skadhauge 1974, Dawson eta/. 1985, 1991 ), however considerable amounts of solutes and water can be removed from ureteral urine refluxed into the rectum (Dawson et a/. 1985, Skadhauge et a/. 1991 ).

Most of the emu's daylight hours are spent foraging, with regular excursions to water. Groups of emus become active just prior to dawn and bed down just after sunset (Dawson et a/. 1984). During summer in the Australian arid zone, emus can be observed foraging in the open in ambient temperatures up to 42°C, when solar radiation levels can exceed 1000 watts/m2 (Dawson eta/. 1984, pers. obs.). The large temperature range covered by emu habitat, and their seeming disregard for high radiant heat loads, prompted this research. The aim of this research was to investigate the physiological. means by which the emu survives exposure to different ambient conditions, with the emphasis on temperature regulation. 53

Chapter 2

Basal Metabolism and Body Temperature

2.1 Introduction

The two variables under investigation in this chapter set the scene for thermoregulation in any endotherm. In endothermic animals the body temperature (Tb) is regulated within narrow limits, so a homeotherm's Tb determines the temperature gradient the animal will experience at different ambient temperatures (T a). The basal metabolic rate (BMR) is defined as .the energy consumption of an animal in a rested, awake, fasted, and thermoneutral state (Simon 1987). An animal's metabolism cannot usually fall below this level so this level of heat production sets the baseline for thermoregulatory responses to different T as.

The scaling of physiological variables such as BMR is of interest in comparative physiology because deviations from a relationship can indicate special features or adaptations (Gould 1978). The variable having the most bearing on a bird's BMR is body mass but BMR may also be correlated with other variables such as phylogeny (Lasiewski and Dawson 1967, Aschoff and Pohl 1970, Daan et a/. 1989), diet and habitat (McNab 1988) and latitude (Weathers 1979, Hails 1983).

The BMR-body mass relationship of birds has undergone many revisions since Kleiber (1932) included three birds in his initial examination of this relationship for homeotherms. Lasiewski and 54

Dawson (1967) calculated separate relationships for the passerines and non-passerines and later Aschoff and Pohl (1970) showed significant differences between the BMR measurements of birds in their normal quiet phase (p), as opposed to the active phase (a), of the daily activity cycle. Recently Dann et a/. (1989) published an equation based on p phase BMR measurements of 263 species. This equation predicts BMRs lower than the Aschoff and Pohl line for birds larger than 30 kg. They also report that a line representing 19 Falconiforme species is 28% lower than this line. Earlier, Withers (1983) found the BMR of the ostrich (Struthio came/us) was only 58% of that predicted from the Aschoff and Pohl non-passerine relationship, or 64% 'of that predicted using the Daan et a/. line for all birds. Calder and Dawson (1978) have also reported similar low BMR's for the three species of kiwi (Apteryx oweni 54%, A. australis 60%, and A. haasti 59%) and suggested that a low BMR may be a feature common to the ratite birds. They also obtained values lower than predicted (83%) for the emu but an earlier study by Crawford and Lasiewski (1968) had indicated that the BMR of the emu agreed well with predictions for non-passerine species. Both studies may have deviated from standardised conditions in that the emus in the 1968 study were hooded and restrained and the 1978 study was conducted in the early evening and used animals not familiar with the experimental procedure. have re-examined the BMR and resting Tb of the emu. I also examined the possibility of differences in BMR due to sex and time of year because of the long inactive period (8 weeks) associated with the incubation fast in winter of the male (Davies 1974). 55

2.2 Materials and Methods

2.2.1 Experimental Animals Five male and five female emus were used in this study. All had been raised in captivity. The sex of these birds was determined by cloacal examination. Experiments were carried out in winter (July­ September) and summer (November-February). None of the emus were laying or incubating during the experimental period.

Animals were transported to the University of New South Wales from their holding yards at least two weeks before data collection began. This two weeks was necessary to familiarize the birds with the experimental procedure; after this time they would stand or sit quietly in the experimental chamber. At the University, the emus were housed in pens (4.3 x 1.2 x 2.5 m) with food (pelleted feed, fruit, lettuce and bread) and water ad libitum. The animals were held for approximately six weeks, during which time body mass was maintained.

2.2.2 Metabolic Measurements and Procedure Experiments were carried out at night during the quiet phase of the emu's diurnal cycle between 18:00 and 03:30 local time. After 24 hours without food, an animal was placed in a large lexan chamber (1.8x1.0x0.6 m) inside a temperature controlled room where Ta was controlled at 25±0.5°C, which is thermoneutral for the emu (see Chapter 3). Ambient temperature was measured in the excurrent port of the chamber with a Jenco electronic thermometer. All temperature sensors used in the research reported in this thesis (thermocouples, thermistors) were calibrated at regular intervals 56

against a mercury in glass thermometer certified by the National Association of Testing Authorities, Australia.

Flow of dry air, at a rate of 149 + 6.3 SEM litres/minute, into the chamber was measured with a Hastings Mass Flowmeter (Model HFM-201 ). A sample (approx. 125 mUmin) of excurrent air was drawn from the chamber exit line by an Applied Electrochemistry R2 flow controller. This stream was dried, scrubbed of C02 with ascarite, redried and passed through an Applied Electrochemistry S3A-II oxygen analyzer. Five second averages from the sensors measuring Mass Flow and 0 2 concentration were logged on a personal computer via a 12 bit analog/digital converter (Sable Systems, USA) resulting in a maximum resolution of 0.001% 02. The system was calibrated several times in each season by the Iron­ burn method of Young eta/. (1984). Calibration factors varied between 0.97 and 1.02.

Before each experiment, an emu was weighed to the nearest 0.1 kg on a platform balance (Wedderburn Scales, Sydney) and then placed in the metabolism chamber. After 1-2 hours, data collection was initiated and lasted for an additional 2-3 hours. Upon completion of data collection, body temperature was measured 10 em into the cloaca (which would place the sensor in the rectum of the emu) with a calibrated Jenco electronic thermometer. Body temperature measurement was completed within five minutes of removing the bird from the chamber.

Oxygen consumption (V02) was calculated using equation 2 from Hill (1972) for a system measuring flow upstream of the chamber 57

and using dry, C02 free air as the reference :Vo2 = VsrPD · (Fio2 -

FEo2) I (1 - FEo2), where VsrPo=flow rate of dry air into the

chamber (STPD), and Flo2 and FEo2 are the fractional concentrations of 02 in air entering and leaving the chamber. BMR was defined as the lowest continuous ten minutes of vo2 in any collection period.

2.2.3 Continuous Body Temperature Measurement Body temperature of nine emus (5 male, 4 female) was measured continuously for 48 hours while they were housed at the University in summer. During this monitoring food and water were available in the pens. A thermistor attached to a temperature transmitter (J. Stuart Enterprises, USA) via wires inside a piece of flexible plastic tubing, was inserted 10 em into the cloaca. The transmitter was taped to the tail/tail feathers. Transmitter signals were received by a Telonics receiver/scanner. Pulse period of the signal (proportional to temperature of the probe) was measured with a Telonics TDP-2 processor. The voltage output from the processor was monitored by a personal computer via an analog/digital converter (ADC1, Remote Measurements, Washington, USA). Custom written software averaged each transmitter's pulse period for one minute every ten minutes. Before and after each use of the transmitters they were calibrated against a certified mercury in glass thermometer. In no case was there a change in the calibration of the transmitters during the time of use.

2.2.4 Statistical Analysis BMR, Tb at the end of the experiment, and mass, were analyzed using Student's t-tests for males versus females. Paired t-tests 58

were used to compare summer to winter data for each sex. Continuous Tb data were reduced to an average temperature for each animal for each 60 minute interval in 24 hours. A two-way repeated measures ANOVA was used to test for effects of sex and time on Tb. A Student-Newman-Keuls (SNK) multiple range test was applied to compare individual means. All values given in this thesis are mean+ SEM and all graphs show mean+ SEM.

2.3 Results

The BM R of the males was significantly lower than that of females in both summer (df=8, t=3.97, P=0.004) and winter (df=8, t=2.71, P=0.03, females summer 2.95+0.09, winter 3.08+0.29; males summer 2.42+0.29, winter 2.53+0.36 ml 0 2/kg.min; Figure 2.1 A).

B D Females 4 A 60 39 ~ Males

50 u 38 0

0" 4> -0" 40 .:..t: -I.... .:¥...... ,;:, N ID - 0__. 2 t 37 CL E E 4> -N 1- 0 ::7' ·> ""0 ~ 36 10

Summer Winter Summer Winter Summer Winter

Figure 2.1 Basal metabolism (A), body temperature (8), and body mass (C) of female and male emus in summer and winter (Stars denote significant differences between the sexes in the same season, hashes denote significant differences between summer and winter in the same sex). 59

39

...... u 0

Q) 5 38.5 +-1 (IJ +-1 L Q) c. E Q) ';, 38 "0 0 ID

37.5 0 4 8 12 16 20 24 Time of Dey

Figure 2.2 Twenty four hour body temperature record of female (dashed line) and male (solid line) emus. Shaded areas show SEM.

The data presented are from emus which were sitting in the chamber. These values for BMR are 77°/o (females) and 61.5% (males) of that predicted by the p-phase equation of Aschoff and Pohl (1970) for non-passerines, and similarly low compared to the equation of Daan et a/. (1989). The females were significantly heavier than the males in summer (df=8, t=2.38, P=0.04, females 45.4+2.8, males 40.7+3.4 kg, Figure 2.1 B). Both sexes were heavier in summer than winter (females df=4, paired t=3.63, P=0.02, males df=4, paired t=5.49, P=O.OOS, winter females 39.7+0.5, males 37 .0+1.2 kg).

Male Tb proved to be lower than female Tb in both seasons (summer df=8, t=3.81, P=O.OOS; winter df=8, t=2.7, P=0.028: females summer 38.3±0.2°C, winter 38.2±0.3°C, males summer 37.7±0.3°C, winter 60

37.7±0.2°C, Figure 2.1 C) and like BMR there was no difference between summer and winter. This difference in night time Tb prompted me to monitor Tb continuously to see whether this difference persisted for the entire day. Both sex (P=0.01) and time (P=0.05) proved to have a significant effect on Tb (Figure 2.2). Comparison of means with SNK showed time was significant for the males, but not for the females. Male Tb was significantly lower than female Tb at night, the males on average being about 1°C cooler (Figure 2.2). During daylight hours the males still tended to have a lower Tb than the females; however there was no significant difference for the 4 hours from 15:00 to 18:00. After 19:00, the males' Tb again became significantly lower.

2.4 Discussion

2.4.1 Sexual Dimorphism in BMR The notable result from this study was the 20% lower BMR of the male emu compared to that of the female. Sexual dimorphism in BMR has been reported for several avian species. However, this is usually associated with a dimorphism in size, with the larger of the sexes having the lower BMR (Rintamaki et a/. 1984, Kaiser and Bucher 1985, Daan et a/. 1989, Gorecki and Nowak 1990). The relationship of BMR with body mass scales with a power less than one (Aschoff and Pohl 1970, Bennett and Harvey 1987, Daan eta/. 1989) so theoretically it might be expected that larger animals will have a lower mass specific BMR. In the emu, there was no difference in mass between the sexes in winter, though the males' BMR was 20% lower than that of the females. In summer the males weighed less than the females and the lower BMR was maintained. 61

This value of BM R for male emus (2700 kJ/day) is close to the daily energy expenditure calculated for two male emus during their eight week incubation fast (2525 kJ/day, Buttemer and Dawson 1989). It is tempting to speculate that this low male BMR may result from selection operating during the males' incubation fast. Failure of incubation due to depletion of body reserves could exert strong selection pressure towards the reduction of male BMR in a population. The amount of fat carried by two male emus at the beginning and end of incubation can be calculated from total body water and mass data in Buttemer and Dawson (1989), assuming fat is 9.5%, and lean tissue 70%, water. The amount of fat decreased from 8.5 to 5.8 kg, and 6.8 to 3.7 kg, in the two emus. This suggests that fat reserves were not seriously depleted during incubation, but three factors bear on these data. Firstly the captive birds were larger, and so were probably carrying more fat, than any we have shot in the wild (see Skadhauge eta/. 1991 ). Secondly the 1989 study was conducted near Sydney, which has a mild climate, and so thermoregulatory costs during incubation would have been lower than in other areas. Thirdly, during starvation, protein depletion limits survival more than fat or carbohydrate depletion (Felig 1979). It is not known how severely protein stores were depleted during incubation but it is noteworthy that the emu also has a low nitrogen requirement (Dawson and Herd 1983). In another male­ incubating ratite, the kiwi, the lowest measured BMR was that of a male (Calder and Dawson 1978).

The Tb of ratites is lower than that of other non-passerine birds (Withers 1983); however, this did not account for the lower BMR of the ostrich (Withers 1983). For a 01 o of 2.5 to account for the 62

observed differences in BMR between the emu sexes, Tb would have to differ by 2.2°C. The male spends 8 inactive weeks incubating the nest and may need a lower working capacity (Drent and Daan 1980) than the female and so can afford a lower BM R.

2.4.2 Body Temperature The body temperatures measured in this study are similar to those reported previously for emus (Crawford and Lasiewski 1968, Calder and Dawson 1978, Buttemer and Dawson 1989). The Tb record of two incubating male emus (Buttemer and Dawson 1989) is comparable with our daily record for non-incubating males (Figure 2.2). The daily average Tb that they obtained for the two males (37.7 and 37.9°C) is similar to our measurement of 37.8°C (calculated from data in Figure 2.2).

The apparent lack of a diurnal cycle in female emus (Figure 2.2) is due to the averaging of data; one bird exhibited a drop in Tb during the day and another exhibiting constant Tb except for a rise at 20:30. Female Tb for most of the night is above the 95% upper confidence limit of the post metabolism experiment Tb (38.6°C, Figure 2.1 C). The emus had constant access to food during Tb monitoring. The females may have remained active and feeding during the night.

2.4.3 BMR - Ratite Allometry In a study on caged emus (Dawson and Herd 1983), the maintenance energy requirement was 70% of that predicted for non-passerine birds. Authors of studies on the BM R of several other ratite species, the kiwis (Calder and Dawson 1978) and the ostrich 63

(Withers 1983), have suggested that the metabolism of ratites may be below that of other non-passerine birds. These measurements on the emu support this hypothesis. If an average value for 'emus' from this study and single points for each species of kiwi and the ostrich are used to obtain a BM R-body mass relationship, the least squares regression line for the log-transformed data is; BMR (mL

0 2/h) =364·Mass0.763 (SEslope = 0.019, r2=0.998, P<<0.001). This line is 35% lower than the Aschoff and Pohl non-passerine line. The mass exponent of 0.763 is higher than expected for an analysis of families within a suborder (Bennett and Harvey 1987) and may be a reflection of the small data set. Deviations from 'normal' by any one of the five species will have a· large effect on the relationship. Verification of such a BMR-body mass relationship for all ratites will require knowledge of BMR in the cassowary and the rheas. These have been reported (cassowary, Benedict and Fox 1927, Rhea americana, Crawford and Lasiewski 1968) but these may not be true estimates of BMR. The cassowary measurement is lower than predicted for a non-passerine, but the bird may have been a juvenile. The rheas were hooded, restrained and measured during the day, so it is unlikely that these measurements were of BMR.

A species closely related to the ratites, the tinamou (Nothoprocta perdicaria), exhibits a BMR intermediate between this proposed ratite level and that of other non-passerines (Withers et a/. 1987). Ratites, along with the tinamou, are considered to be one of the earliest offshoots in bird evolution (Cracraft 1974), just as the monotremes represent the oldest mammals. Evolution seems to have favoured increasing levels of resting metabolism in mammals (Dawson and Hulbert 1970). A similar succession may have 64

occurred in bird metabolism from the ratites to carinate non­ passerines to passerines. This scenario would require either: 1) the tinamou to have increased BMR from the primitive 'ratite' level, independently from other carinates; or 2) for the ratites to have reduced metabolism from a primitive 'tinamou' level, since the tinamou are considered to be the earliest offshoot from the ratite­ tinamou lineage (Cracraft 1974). Ratites becoming flightless could have led to reductions in their required 'aerobic capacity'. If there is a relationship between BMR and maximum aerobic capacity (see Bennett 1991) this would have enabled the ratites to make energetic economies which resulted in a low BM R. 65

Chapter 3

Thermoregulation in the Emu

3.1 Introduction

The size of the emu limits the extent to which it can exploit behavioural thermoregulation by selecting appropriate micro­ climates, and so its range must reflect a generally successful physiological thermoregulatory capacity. Some of this success would undoubtedly be attributable to the degree of thermal inertia that large body size confers. Cracraft (1974) suggests that global cooling during the middle to late Cretaceous period may have provided the selective pressure leading to increases in size, and its associated advantages for temperature regulation, in the ratite lineage. However, the maintenance by an animal of a temperature gradient between itself and the environment requires regulation of heat production and loss. The experiments reported in this chapter were undertaken to investigate the emu's physiological responses to changes in ambient temperature.

3.2 Materials and Methods

3.2.1 Experimental Animals The ten emus used in these experiments were the same as those used in the experiments reported in Chapter 2. The transport and housing of the emus were the same. 66

3.2.2 Metabolic Measurements These experiments were also carried out during the quiet phase of the emu's diurnal cycle between 18:00 of one day and 03:30 of the next, except for one set of measurements at T a=25°C performed during the day. After 24 hours without food, an animal was placed in the metabolism chamber described in Chapter 2, which also had a mesh floor above a bath of vegetable oil to cover any excreta. The inside walls of the chamber were painted flat black to reduce radiation reflection in the chamber (Porter 1969). The chamber was situated in a temperature controlled room in which ambient temperature was controlled to ± 0.5°C of a set point. Measurements were made at five temperatures in summer (5, 15, 25, 35, 45°C), and at eight in winter (-5, 5, 15, 20, 25, 30, 35, 45°C). Ambient temperature was measured with a calibrated Jenco electronic thermometer, the probe being situated in the excurrent port of the chamber.

During these experiments flow rate of dry air into the chamber was manipulated so that water vapour pressure in the chamber remained low (Figure 3.1 ), and did not become limiting for the birds' physiological responses to ambient temperature (Lasiewski et a/. 1966). A sample (approx. 125 ml/min) of excurrent air was drawn from the chamber exit line by an Applied Electrochemistry R2 flow controller. The relative humidity of this sample was . me·asured with a CHK-Engineering capacitance type humidity sensor (±0.1% RH), which was calibrated at regular intervals with saturated solutions of lithium chloride and sodium chloride (Winston and Bates 1960). 67

260 1600 ~Flow (1/min) < Q) 230 - e - VP H20 1400 ~ .... 0 :::::J c: c: 1200 ~ E 200 .... I CD en I (/) -Q) (/) 1000 ....'­ I c:.... :=.. 170 I CD Q) I ca 800 N cc 0 ""0 3: 140 I - 0 I 600 m u.. 0 ~ 110 400 -

80 200 0 10 20 30 40 50 Ambient Temperature (°C)

Figure 3.1 Flow rate of dry air into the metabolism chamber, and the resulting

chamber vapour pressure of H20, at the ambient temperatures used In this study

The sample was then divided in two. One stream was dried and passed through an Anarad AR-50 Carbon Dioxide analyzer which was calibrated daily with ambient air and a gas of known higher C02 concentration (Commonwealth Industrial Gases Ltd ., Sydney) . The other stream was dried, scrubbed of C02 with ascarite, redried and passed through an Applied Electrochemistry S3A-II oxygen analyzer.

Five second averages from the four sensors measuring mass flow, relative humidity, and C02 and 02 concentrations were logged on a personal computer via a 12 bit analog/digital converter (Sable 68

Systems, USA). Maximum resolution for this combination of sensors and AID converter was 0.001% for 02 and 0.01% for C02. Three minutes of incurrent air was directed through all analyzers at the start and end of each data collection period to facilitate correction for drift in any sensor.

3.2.3 Experimental Procedure Before each experiment, an emu was weighed to the nearest 0.1 kg on a platform balance (Wedderburn Scales, Sydney) and then placed in the metabolism chamber. At temperatures below 20°C it was desired to obtain data from animals in both sitting and standing positions. Animals could be forced to stand by placing a large (0.65

Back

Belly

Toe

Figure 3.2 Areas of the emu where surface temperature measurements were made (after Eastman 1969) 69

x 0.4 x 0.4 m) plastic box in the chamber, which left insufficient room for sitting. Animals could be coaxed to sit, but not to remain sitting, which restricted the collection of data. After 1-2 hours at any T a. data collection was initiated. Data collection periods lasted from 1.5 to 4 hours, in most cases total exposure to any T a was more than 3 hours. At the end of an experiment plumage surface temperature of seven representative areas (Figure 3.2) were measured with an Everest infra-red thermometer. Cloacal temperature was then measured. This procedure normally took 4-5 minutes. Animals were sometimes measured at two temperatures on the one night. At least 48 hours elapsed between experimental sessions for any one animal.

3.2.4 Continuous Body Temperature Measurement Four emus were fitted with cloacal temperature transmitters (described in Chapter 2) and exposed to Ta=45°C. Custom written software provided a continuous display of probe temperature, which was recorded regularly by hand.

3.2.5 Measurement of Cutaneous Evaporative Water Loss Five emus exposed to T as of 25 and 45°C in summer were each fitted with two perspex capsules containing Drierite. The capsules consisted of a round sleeve (5.68 em diameter) and tight fitting caps which contained the Drierite below wire mesh. Two sleeves were glued (Superglue) to the emus, one to the back and one to the belly, to areas where the feathers had been trimmed. Then the animal was exposed to the relevant Ta for at least 30 minutes before a pre-weighed cap was fitted to each sleeve. At Ta=25°C the caps were left on the birds for 60 minutes, and at Ta=45°C for 30 70

minutes. After removal of the cap, it was immediately weighed and cutaneous evaporative water loss (CEWL) calculated as the weight gain of the cap. Ten estimates were made of the amount of water absorbed by the Drierite between weighing of the cap and placement on the emu, and removal and weighing. This gain averaged 7o/o of the measured average weight gain of the cap and each CEWL estimate was adjusted accordingly.

3.2.6 Data Analysis

Vo2 was calculated as in Chapter 2.

Vco2 was calculated as: Vco2 = VsrPD · (FEco2 - Flco2), where Flco2 and FEco2 were the fractional concentrations of C02 in air entering and leaving the chamber, respectively. Evaporative water loss (EWL) was calculated after relative humidity was converted to grams H20/Iitre in the excurrent air

(FEH 2o) using a predictive equation based on data in the CRC Handbook of Chemistry and Physics, and flow rate into the chamber converted to ambient temperature and pressure. Evaporative water loss was then calculated as: EWL = FEH2o · (V + Vco2 -Vo 2 ) I (1 -

FEH 2 o), since air entering the chamber was dry.

Metabolic Rate (MR) and EWL at each T a were defined as the lowest continuous five minutes of Vo2 and EWL, respectively, in any collection period, except at 35 and 45°C where EWL was an average after the animal had commenced panting. This period for MR estimation did not introduce a bias (see Discussion). Due to condensation of water in the chamber and excurrent lines at T a<0°C, EWL was not calculated for -5°C, but for calculation of conductance 71

was assumed the same as at 5° C. Heat equivalents of Vo2 and EWL were calculated assuming 20.08 kJ/Iitre 02, and 2427 J/g H20 respectively.

3.2. 7 Surface Area The total plumage surface area of each area in Figure 3.2 was measured in four emus. The neck, legs, and toes were assumed to be cylinders with a surface area equal to the average circumference times the length. All other areas (back, shoulder, hip, belly) were traced with tissue paper and the area calculated by trigonometry. Total surface area (SA) of the four emus was used to calculate a · · Meeh formula assuming a slope of 0.667: SA (cm2) = 7.5 (±0.16) · Mass (g)0.667 (Equation 3.1) This is a smaller mass coefficient than the 8.11 obtained by Walsberg and King (1978) as a general formula for external plumage surface of birds. This may reflect the relatively simple geometry of the emu's body and its lack of wings. However, the emu's SA includes the legs, which other SA analyses have not, because they form a large proportion (15%) of the standing emu's SA. Equation 3.1 was used to estimate the SA of each emu used in this study.

The average contribution of each surface area in Figure 3.2 to total surface area was used to estimate the area of each region for each emu. These estimates of the area of each region, and the surface temperature of each region, were used to calculate radiative heat loss for each region at each Ta, assuming chamber wall temperature was equal to Ta (this was confirmed several times), using the formula; 72

4 4 ci> = e 1 • e 2 · a · (T e - T a ) · SA (Equation 3.2) where: ci> = radiant heat exchange (watts)

e 1 = animal surface infra red emmissivity (assumed 0.98)

e 2 = chamber wall infra red emmissivity (assumed 0.98) a = Stefan-Boltzman constant (5.67x1 o- 8 wattsfm2.K4)

Te = relevant surface temperature (degrees Kelvin)

Ta = chamber wall temperature (degrees Kelvin)

SA = relevant surface area Total radiant heat loss from an animal at any T a was the sum of the radiant heat loss from each region.·

Investigators interested in the radiative loss of heat from animals have reduced total SA to an effective radiating surface area (Ar), since some areas of the body will be exchanging radiation with other surfaces on the animal, not the chamber walls. Taking the shape of the emu into consideration and comparing it with published values of Ar for other animals (Bond et a/. 1952, Blaxter et a/. 1959, Folkow and Mercer 1986), a value of 0.9 · SA was used as Ar for the emu. Thus radiative heat loss calculated from equation 3.2 was multiplied by 0.9 to obtain a final estimate of radiative heat loss for each animal.

3.2.8 Statistical Analysis Data were analyzed using Two-Way Repeated Measures Anovars for sex and temperature, and season and temperature for each sex. A Student-Newman-Keuls (SNK) multiple range test was applied when significant differences were indicated by the Anovar. If sex proved 73

significant, the sexes are presented separately, otherwise data for both sexes are grouped. Five out of the fifty individuai/T a surface temperature sets were not recorded due to problems with animals. Rather than reduce the data set to those individuals on which a full data set was obtained, thus allowing a paired analysis, sex and temperature differences were tested for with a factorial ANOVA and individual means compared with Fishers PLSD test. All graphs and tables show mean ± SEM. Means are considered significantly different if p<0.05.

3.3 Results

Different letters on means from the same sex denote significant difference between the means (SNK, P<0.05). Stars on the x-axis denote a significant difference between the sexes (SNK, P<0.05).

3.3.1 Body Temperature The temperature measured in the rectum of the emu will be referred to as the body temperature. At the completion of metabolism experiments (2-4 hours exposure), Tb had not changed from thermoneutral levels at any Ta, except at Ta=45°C where there was a slight but significant increase in Tb (Figure 3.3). There were no differences between summer and winter. The difference in Tb between female and male emus at thermoneutral Ta remained evident at all T as, although in winter at 15, 35, and 45°C, and in summer at 45°C, this difference was not significant. At 45°C there was a slight, but significant rise in Tb taken after several hours exposure to this temperature. The four emus fitted with temperature transmitters and exposed to Ta=45°C all began to pant 74

before Tb changed measurably. Two of these emus maintained Tb during the subsequent exposure, Tb rose in one bird, and was reduced by 0.2°C in one individual after an hour, when activity by the emu led to the termination of that experiment. The body temperature records from these experiments are shown in Figure 3.4, also noting the commencement of panting.

3.3.2 Metabolism There were no significant differences in MR between summer and winter at any ambient temperature for either sex (Figure 3.5A, B). Since the emus were measured at eight T as in winter and only five

39 Winter b Summer b

-u a.b . j .. z '8 37.S ftfl z fz o females ((') z z • males

37 -10 0 10 20 30 40 so 0 10 20 30 40 so .Ambient Temperature (°C) .Ambient Temperature (C)

Figure 3.3 Body temperatures of female and male emus taken at the completion of metabolism experiments in winter and summer (different letters on the same sex denote significant differences between means; SNK P

Figure 3.4 Continuous body temperature measurements and time when panting commenced of four emus (B-E) as ambient temperature (A) increased to 45°C

...... u 50 0 +J -(I) c !.... (I) 40 ...... :::3 .0 (C !.... E (I) <:t: 0. 30 E (I) 1- 20 ...... u 38.7 0 -(I) !.... ::» :::3 38.5 "C +J 0 (C £D !.... (I) experiment 0. 38.3 E terminated (I) 1- 38.1 ...... u 38.6 c 0 -(I) /panting !.... ::» :::3 "C +J 0 (C 38.5 £D !.... (I) 0. E (I) 1- 38.4 ...... u. 38.3 0 -(I) !.... ::» :::1 38.2 "C +J 0 (C £D !.... (I) 0. 38.1 E Q) 1- 38 ...... 37.9 u E 0 /panting -(I) !.... ::» :::1 "C +J 0 (C 37.8 £D !.... (I) 0. E (I) 1- 37.7 0 20 40 60 80 100 Time (minutes) 76

in summer, the analysis will focus on the winter results. All animals could be coaxed to sit and remain sitting at -5°C. At 5 and 15°C, attempts to gather full sitting data were not as successful. At T a=5°C sitting data were obtained from 2 females and 4 males and at Ta=15°C from 2 females and 3 males. At Tas above 20°C there was no difference in the MR of six emus between standing and sitting at 25°C (t=1.67, p=0.16) but the data presented are from sitting emus at these T as. Constant MR for the two sexes at Tas between 20 and 35°C, but at different levels (Figure 3.6), is further evidence that BMR differs between the sexes (Chapter 2) . The regression lines for female and male data from 20°C to 35°C both have a slope not significantly different from zero. The elevations of the two lines are significantly different, with the elevation of the male line being

Females Males

o Winter • Winter 1 o Summer • Summer Q l

t ~ j ~ ~ 0 ~ \ \

0.5 0 1 0 20 30 40 so 0 10 20 30 40 so Ambient Temperature (C)

Figure 3.5 Metabolic rates of female and male emus at a range of ambient temperatures in summer and winter 77

13% lower than the female line (females W/kg = 0.002 · T a + 1.005

(±0.046); males W/kg = 0.001 · T a + 0.878 (±0.059); t=4.39, df=37, P<<0.005). At T as below 15°C there was a significant increase in MR by both sexes when standing. Analysis of the means with a SNK multiple range test showed that at Ta=5 and 15°C there were no significant differences between the MR of the females and that of the males. At Ta=-5°C the MR of the females was significantly higher than the males. Since MR was the same for the sexes at two of the Tas

3 8. f 0 femeles 2.5 • meles b D femeles sit 0'> 2 z ~ c meles sit -(/) • -ro cd ?; 1.5

1

0.5

0 -10 0 10 20 30 40 so .Ambient Temperature (C)

Figure 3.6 Metabolic rates of female and male emus at a range of ambient temperatures in winter (different letters on the same sex denote significant differences between means; SNK P

below 20°C a single regression line relating MR to T a was calculated for data below 20°C. For standing emus the regression formula for data at Tas from -5 to 15°C was MR= -0.063 ·Ta + 2.2. For sitting emus the regression formula was MR= -0.028 · T a + 1.32. The slope of the standing line is significantly greater than the slope of the sitting line (t=2.95, df=47, p<0.005).

Day time measurements of MR at Ta = 25°C averaged 5% higher than the night time measurements at 25°C (females night = 0.99±0.01, day = 1.1±0.02 W/kg; males night = 0.81±0.04, day = 0.84±0.04 W/kg). A two-way repeated measures ANOVA showed both sex and time to be significant (sex P=0.002·; time P=0.018).

3.3.3 Respiratory Quotient

Respiratory Quotient (RQ; \1co 2 I 'Vo 2) measurements for the winter emus are shown in Figure 3.7. At Tas from -5 to 35°C average RQs were between 0.7 and 0.8. At T a=45°C there was a significant increase in RQ to 0.85. Respiratory quotient estimates for summer emus were ignored when it was found that the stated C02 concentration of a Commonwealth Industrial Gases calibration gas was wrong. Respiratory quotients between 0.7 and 0.8 are consistent with postabsorptive metabolism of mostly fat. An energy equivalent of 20.08 kJ per litre of oxygen consumed was used to convert V02 into a heat equivalent (watts). The error induced by the use of this conversion factor at different RQ's will be minimal (Schmidt-Nielsen 1979). 79

0.9 b ...... c 0.85 Q) :;::::; f 0 a :::J 0.8 a 0 ,_>. a a a ...... 0 0.75 a I a f ,_ctS f ·a.. f en 0.7 I f Q) 1 a: 0.65

0.6 -10 0 10 20 30 40 50 Ambient Temperature (°C)

Figure 3.7 Respiratory quotient (VC02 / '\/02) versus ambient temperature in winter (different letters denote significant differences between means; SNK P

3.3.4 Evaporative Water Loss The total EWL of female and male emus in winter is shown in Figure 3.8. The differences between sexes in summer were not significantly different, compared to the winter values, at any T a· In winter at T as up to 25°C there was no significant difference in EWL between the sexes or between T as. At Ta=30°C the females increased EWL. At T a=35°C both sexes significantly increased EWL relative to the lower Tas. Female EWL was 43% higher than the males at this Ta (SNK, p=0.004). At T a=45°C both sexes EWL was significantly higher than at all other T as. Female EWL was 19% higher than the males EWL at 45°C (SNK, p<0.001 ). If EWL is converted to a heat loss equivalent (evaporative heat loss, EHL), 80

2 140 d 0- C\1120 I 0) ? -C/) 100 C/) +X 0 o females _J '- 80 males caQ) • ~ 60 c Q) > 9 40 b ~'- 0 c.. a 9. •y 20 a a ~ a eft o. z w Oe o. z z z z * * 0 0 10 20 30 40 50 . Ambient Temperature (C)

Figure 3.8 Evaporative water loss versus ambient temperature for female and male emus in winter (different letters on the same sex denote significant differences between means; SNK P<0.05: Stars on the x-axis denote a significant difference between the sexes; SNK P<0.05) and expressed as a percent of metabolic heat production, there were no significant differences between the sexes.

3.3.5 Cutaneous Evaporative Water Loss Average CEWL measured using drierite capsules at T a=25°C was 14.6±1.01 g H20/m2·h. There was no difference between the rates from the back and the belly. At T a=45°C data could not be obtained from two emus. CEWL for the other three emus was 21.1±1.24 g H20/m2·h. The rate of CEWL from the belly was higher in all three emus than from the back but this difference did not prove significant (paired t=1.97, n=3, P=0.19). For comparison, rates of CEWL estimated from the difference between EWL and respiratory 81

evaporative water loss (REWL, calculated from ventilatory data in Chapter 4) in summer were 7.4 g H20/m2·h at T a=25°C and 28.1 g H20/m2·h at Ta=45°C. These estimates were calculated using the formula for skin surface area of birds obtained by Walsberg and King (1978). The value obtained using the capsules at T a=45°C is within the 95°/o confidence intervals of the summer data. The value obtained at 25°C is above the 95% confidence interval of the summer data (upper 95% interval = 9.1) and may reflect the inherent low humidity in the capsules.

3.3.6 Conductance Dry conductance at each T a was calculated as C (WfJC) = (MR-EHL) I (Tb-T a) (Dawson and Schmidt-Nielsen 1966). Values of C were then

7 0.1S c A e B ...... u ...... u 0 6 0 N 0') .:::£ E S ...... b t ...... b f (/) 0.1 (/) o+J o+J o+J a;] o+J4a;] ~ ...... ~ (1) (1) 3 u J u J 8. a a. 8 c: t c: 8. + a;] 8. + a;] 8. o+J O.OS o+J 2 • u • u •t. ::l ::l • '0 • ' 8. ' "'0 8 • c: • § 1 0 u u 0 0 -10 0 10 20 30 40 so -10 0 10 20 30 40 so Ambient Temperature ( O()

Figure 3.9 Dry conductance in winter expressed as watts/m2·°C (A) and watts/kg·°C (B) . For explanation of symbols in Figure 3.9A, see text. (different letters denote significant differences between means; SNK P<0.05) 82

divided by either surface area or mass to obtain conductance estimates in units of Wfm2.°C (Figure 3.9A) and W/kg.°C (Figure 3.98). As might be expected from the fact that the heat production required to maintain Tb at low Tas was the same in summer and winter, there were no differences in the conductance of either sex between summer and winter. Conductance is minimum at 20°C, and conductance estimates at this Ta and below show no significant differences. Above 20°C, conductance increases to a maximum at 45°C. There was no significant difference between the conductance of females and males at any T a· Figure 3.9A also shows the conductance of sitting emus at T a=~5°C, calculated using total SA (open circle) and exposed SA (open square, see Discussion). A repeated measures ANOVA and SNK was used to compare these three measures of conductance at -5°C to the values at 20°C. All four estimates were significantly different from each other, in contrast to the analysis of the full range of standing data which indicated that the -5°C and 20°C measures were not different.

3.3.7 Surface Temperature All surface temperature measurements were made on standing emus. If an emu had been sitting at any T a. upon completion of metabolic measurements it was forced to stand, and 5-10 minutes later the surface temperature measurements were made. There were no differences between male and female surface temperatures of any region at any particular T a· Average surface temperatures of six of the seven areas identified in Figure 3.2 are shown in Figure 3.1 OA-F. Temperature of all surfaces were lower at lower Tas. The temperature gradient between emu surface and T a 83

Figure 3.1 0 Surface temperatures of various areas of emus in relation to ambient temperature (different letters denote significant differences between means; SNK P<0.05) 84

was lower for the well insulated areas of the body (shoulder, back, hip) than other areas not as well insulated.

3.3.8 Heat Balance Heat loss from an animal's body can take four forms: radiation,

200 Winter en en 150 0 _J 100 caQ) I 50 -0 c -Q) 0 (.) '- Q) a... -50 -100

0 REWL 200 Summer [J CEWL en en !ill] Radiation 0 150 _J II Cond + Conv ca 100 Q) I -0 50 c -Q) 0 ~ Q) a... -50

-100 -5 5 15 25 35 45 Ambient Temperature (°C)

Figure 3.11 Partitioning of heat loss by evaporative and non-evaporative means in winter and summer 85

evaporation, conduction, and convection. Evaporation was measured, radiative heat loss was calculated from surface temperature measurements, and so the contribution of conduction/convection could be calculated as the difference between total heat loss (which is metabolic heat production since Tb was constant) and EHL plus radiative heat loss. The contribution of each heat loss route to total heat loss in winter and summer is shown in Figures 3.11 A, B. There were no significant differences between the seasons in any of the parameters. As Ta increased EWL became progressively more important. At 35°C, radiative heat loss was close to zero, EWL accounting for 73% of total heat loss and conduction/convection for the remainder. At 45°C the Tb- T a temperature gradient was

40 b

Figure 3.12 Leg radiative heat exchange as a percent of total radiative heat exchange (different letters denote significant differences between means; SNK P

reversed and heat was entering the body via radiation and conduction/convection. At this T a, EWL increased considerably to enable the dumping of metabolic heat, plus the external heat load.

The percentage of total radiative heat exchange from the legs at each T a is shown in Figure 3.12. A factorial ANOVA and Fishers PLSD test were used to compare means at Ta=-5, 5, 15, 25, and 45°C. The T a=35°C data are negligible since total radiative heat exchange was less than 1% of total heat loss.

3.4 Discussion

3.4.1 Body Temperature There is some controversy in thermoregulatory studies regarding which temperature signifies the precisely controlled variable, be it deep body, hypothalamic, or spinal cord temperature (Simon et a/. 1986). Since deep body temperature is usually conveniently measured and is at least correlated with whatever is the controlled temperature, it is commonly measured. In this study the temperature in the rectum of the emu was measured. Continuous measurement of this temperature (Chapter 2) resulted in temperatures the same as those obtained from temperature transmitters in the peritoneum of two male emus by Buttemer and Dawson (1989). Thus temperature of the rectum can be regarded as a good estimate of deep body temperature. Body temperatures taken at the completion of metabolic measurements (Figure 3.3) indicate that the emus maintained Tb at all T as for the 2-4 hour period for which they were exposed. The difference in Tb between female and 87

male emus at thermoneutral temperatures (Chapter 2) was maintained at T as outside the TNZ.

The slight but significant increase in Tb following exposure to T a=45°C may be an artifact of experimental practice. Continuous monitoring of the Tb of four emus exposed to T a=45°C showed that they began to pant before Tb changed measurably and then they maintained Tb for at least two hours. These results suggest that there was no increase in Tb in undisturbed birds at T a=45°C. In the light of the continuous Tb monitoring it seems unlikely that the observed increase in Tb following exposure to 45°C is either a failure of the regulating system, or' a case of controlled hyperthermia.

The lack of, or inappropriate, response of many bird species to hypothalamic heating indicates that hypothalamic temperature is not a highly weighted variable in avian responses to heat (Simon et a/. 1986, Hissa 1988). The emu, however, does appear to show appropriate responses to hypothalamic temperature stimulation (Jessen eta/. 1982). Many bird species increase Tb before active EWL begins (Dawson 1982), indicating that a feedback temperature threshold from some internal receptor needs to be exceeded for active EWL to begin. There was no measurable change in Tb before the onset of panting in four emus exposed to Ta=45°C . Since cloacal temperature was used as a measure of Tb, it is not known whether hypothalamic or spinal cord temperature increased prior to panting initiation. If they did not, this indicates that peripheral temperature was an important variable in the initiation of active heat dissipation. 88

3.4.2 Metabolism At Tas from 20 to 35°C there were no significant differences in MR between Tas for either sex. Below Ta=15°C MR increased (Figure 3.6). This increase in metabolism facilitated the maintenance of Tb (Figure 3.3).

Hayes et a/. (1992) showed that the arbitrary choice of the length of time used to define MR can significantly affect MR estimates. The usual method used to obtain MR is to define a time period for MR measurement (e.g. 10 minutes) and then to select the lowest continuous 10 minutes of V02 recorded for each animal at each T a· A bias is introduced by this arbitrary choice of a time period, the probability being that the shorter the time period chosen, the lower the MR estimate will be. Hayes et a/. (1992) recognized that the bias will be greatest in active animals, where V02 varies considerably within short time periods. Such bias will especially affect conclusions drawn in a thermoregulatory study since the heat generated during active periods will help maintain Tb at low T as, but if the MR period excludes these active periods, the MR needed to maintain Tb will be underestimated.

The quiet response of the familiarised emus once in the metabolism chamber, and stable V02 measurements, meant that any bias introduced by the length of period chosen for MR estimation was minimal. Increasing the MR time period from five minutes to thirty minutes increased the average winter MR at T a=-5°C from 96.5 watts to 98.7 watts, an increase of 2.3%. At T a=25°C in winter the effect of the same increase in time period was to increase the average MR estimate from 36.1 watts to 37.7 watts, an increase of 89

4.6o/o. This type of analysis was not performed at each Ta· However, the behaviour of the birds, and vo2 consistency, were similar at each T a- Thus five minutes was used as the MR time period for these experiments.

3.4.3 Conductance It is a fortuitous accident of physics that the relationship between a homeotherm's MR and T a below the thermoneutral zone is a straight line (Burton and Edholm 1955). This relationship is often applied in what is known as the Scholander model of thermoregulation. The line relating heat production to T a below the

LCT can be used as an estimate of conductance (viz: Conductance = HP/[Tb-TaD and this line will extrapolate to Tb, when MR=O, in animals conforming to the model (Calder and King 1974, Schmidt­ Nielsen 1979). However, if physical regulation (e.g. posture, piloerection) changes as Ta falls, then conductance will continue to fall with decreasing T a and the conductance line will extrapolate to a temperature greater than Tb (Drent and Stonehouse 1971 ). This is the case for the sitting emus in this study; the conductance line extrapolates to 47°C (Figure 3.6). This is an indication that conductance must have decreased at temperatures below the LCT. Comparison of the sitting emu's conductance at -5°C with that at 20°C showed this was the case (Figure 3.9A). The birds could not be observed when in the chamber so the degree to which piloerection was utilized could not be determined. Emus incubating nests curl the neck into an S shape which contacts the shoulder area and thus reduces the exposed surface area. As soon as the temperature controlled room was opened, the emus looked up from inside the 90

metabolic chamber, and so it was not determined whether, or at what temperature, this posture was adopted.

The regression line relating Ta to M R for standing emus extrapolates to a temperature below Tb (35°C, Figure 3.6), indicating conductance must have increased with decreasing Ta. Comparison of the -5°C conductance data with the 20°C conductance data indicated that conductance did increase at the lower temperatures in the standing emus. This could result from more intense shivering as Ta decreases. This leads to increases in blood flow to the peripheral musculature and disturbs the air boundary layer which contributes td insulation both of which cause conductance to increase.

There were no significant differences in conductance between the sexes at any Ta (Figure 3.9A, B) nor between summer and winter values at any Ta· A change in conductance with a change in temperature regime (e.g. with changes in season) is a principal means of animal acclimation to environmental conditions (Bruck 1986). The emu does not show such a pattern, having the same conductance in both seasons. In winter a low conductance is advantageous, while in summer increased conductance would seem intuitively advantageous. However, the emu is active in summer on days when Ta is higher than Tb, and so a low conductance will reduce the environmental heat load. When radiant heat loads are considered (see Chapter 6) a low conductance can be especially advantageous in summer. 91

The transfer of heat from the body core to the cooler environment depends on a suite of factors. There is conduction through the tissues, convection by the blood, and radiant, convective, and conductive transfer through the plumage. Convective and conductive transfer contribute approximately equally to account for over 95% of transfer through bird plumage (Walsberg 1988a). From the external surfaces radiant, convective, and conductive transfer occurs, as well as evaporative heat loss from the skin and the ventilatory system. Because of the near impossibility of measuring the factors involved in each transfer process simultaneously (McNab 1980a), an estimate of total body conductance (Conductance

= Metabolic Heat Production I [Tb-T a·]) has been used by many workers. Such calculations include heat loss by EWL in the conductance estimate, and so are termed 'wet conductance'. The contribution of EWL to total heat loss at low T as is usually minor (5-15%, McNab 1980a). However Aschoff (1981) obtained significant differences between the wet conductance of birds measured during the active phase, as opposed to the quiet phase, of their daily cycle, a difference he attributed mainly to differences in EWL between the phases.

Contemporary thermoregulatory studies routinely measure EWL simultaneously with oxygen consumption. Subtraction of EHL in the conductance formula gives an estimate of the ease of overall dry heat flow from the animal to the environment (Conductance = [Metabolic Heat Production - Evaporative Heat Loss] I [Tb-TaD· This dry conductance is usually expressed in terms of body mass as watts/g.°C or some equivalent (King and Farner 1961, Herreid and Kessel 1967, Calder and King 1974). Because EHL becomes more 92

important as Ta rises, wet conductance becomes less meaningful at higher T as. At these higher temperatures dry conductance provides an indication of changes in the degree of insulation, a major part of which will be changes in peripheral blood flow. A change in dry conductance with increasing ambient temperature, as the system shifts demand from heat conservation to heat dissipation, is well illustrated in the emu (Figure 3.9A, B).

Given that the conductance of the sexes is the same, it may be expected that the relation of M R to T a for the sexes would become similar at Tas below the LCT, with the females having a lower LCT, due to their higher thermoneutral MR. Comparisons of female with male metabolism (using SNK) revealed no significant differences in MR at Ta=15 or 5°C. A factor which would be expected to affect the relationship of MR to Ta at these low temperatures is the higher Tb of the females (Figure 3.3), giving them a larger temperature gradient for dry heat loss. While the mean MR of the females was higher at these low Tas. the higher Tb would result in an increased heat loss of only 1.5% of SMR (e.g. at Ta=5°C the 0.5°C higher Tb of the females would result in a dry heat loss only 0.6 watts greater than that of the males).

Dry heat exchange between an animal and the environment occurs across the body surface, apart from the minor dry heat exchange associated with tidal air. Many authors express dry heat loss in terms of the body surface area (SA). The studies to date on birds are confused by the fact that most studies used a generalised Meeh formula for skin surface area, but heat exchange with the environment ultimately occurs across the plumage surface, or 93

exposed skin areas. Walsberg and King (1978) have shown that the commonly quoted Meeh formula overestimates SA of the plumage by 23% in species they measured. Studies that used this Meeh factor would have underestimated true area specific conductance by 23%. Posture changes also affect exposed surface area (for example if an emu sits down, exposed SA is reduced by 15%), so the question arises whether SA is the total SA or the exposed SA.

The data in Figure 3.9A were calculated using total SA for standing emus. For comparison, conductance was calculated for the sitting emus using both total SA and exposed SA. The surface specific conductance of sitting emus was significantly lower than that of standing emus, whichever definition of SA is used. This is a reflection of the effect of having only well insulated areas exposed when sitting, removing the legs as sources of heat loss had a substantial effect on heat balance. The conductance estimates of the sitting emus calculated using exposed SA were significantly higher than estimates made using total SA. This is the effect of calculating conductance using a smaller SA.

The dry conductance of the emu is higher than predicted by the allometric relationships of Herreid and Kessel (1967) and Aschoff (1981 ). It is also high if placed in the context of Figure 8 from Drent and Stonehouse (1971 ), even if the conductance values in that figure are adjusted by 1.23 (Walsberg and King 1978). Bech (1980) has argued that as body size increases, insulation may reach a maximum and so the relationships cited above cannot be applied to large birds (the largest bird in the analysis of Herreid and Kessel was 2.7 kg, in the study by Aschoff it was 2.4 kg, and in the data 94

set of Drent and Stonehouse the largest bird was 5 kg). Alternatively, flying birds may in general need lower conductances given that while flying they are exposed to high levels of convection. Flightless birds such as the emu may be able to survive in similar climates with lower insulation than flying birds. Most of the birds in the former analyses were flying birds and this may also affect comparisons of the emu to these relationships.

3.4.4 Evaporative Water Loss At low T as some loss of water via ventilation and across the general body surface is unavoidable. Loss of water when it is not essential for temperature regulation is disadvantageous to animals inhabiting arid areas, where pre-formed water supply can be unpredictable. The break even point for water loss I metabolic water production is 0.67 mg H20imL 02 consumed (Schmidt-Nielsen and Schmidt-Nielsen 1952). Some of the standing emus in this study were gainining more metabolic water from respiration than they were losing via EWL, the average ratios of mg water lost I mL

0 2 consumed being 0.67 at -S°C, 0.92 at 5°C, 1.6 at 15°C, and 2.6 at 25°C. None of the sitting emus were making a gain at low temperatures, although the ratios were low (1.06 at -S°C, 1.23 at S°C, and 1.99 at 15°C). These calculations, however, do not take account of water losses in the urine and feces. The ability to minimize water loss relative to water production will be important during times of limited water intake, for example during the male's incubation period. A factor contributing to the low ventilatory water loss is the ability to exhale air cooler than body temperature (see Chapter 4). 95

At higher Tas, the emus relied on increased levels of EWL to maintain thermal balance (Figure 3.8). Large animals such as the emu are forced to increase EWL at lower Tas than smaller animals, due to the relatively smaller SA available for dry heat loss (Weathers 1981 ). At very high T as (45°C) the gradient for dry heat transfer was reversed and heat was flowing into the body. At T a=45°C the heat lost by EWL in the emu was equal to 164% of metabolic heat production. Capacity for such high levels of EWL is common in birds, with rates at T a=45°C ranging from less than 100% to > 150% of metabolic heat production (Calder and Schmidt­ Nielsen 1967, Dawson and Bennett 1973, Hinds and Calder 1973, Marder 1973, Weathers 1981, Larochelle et at. 1982, Arad 1983, Dmi'el and Tel-Tzur 1985, Frumkin et a/. 1986, Roberts and Baudinette 1986, Arad et a/. 1987, Withers and Williams 1990, Hinsley 1992).

At Tas close to but lower than Tb, a high level of dry thermal conductance will allow dissipation of more heat by non­ evaporative means, and thus a saving of water. The emu's dry conductance does increase at Ta=35°C relative to lower Tas (Figures 3.9A, B). When Ta exceeds Tb, it is advantageous for the animal to decrease dry conductance. This means that less heat flows into the body, and the EWL needed to maintain Tb is lower than if conductance is high (Dawson and Schmidt-Nielsen 1966). This pattern of increasing dry conductance as Ta approaches Tb from below, but then decreasing dry conductance when T a exceeds Tb, has been observed in several bird and mammal species, such as the burrowing owl (Coulombe 1970), the pyrrhuloxia and Arizona cardinal (Hinds and Calder 1973), the monk parakeet (Weathers and 96

Caccamise 1975), the budgerigar (Weathers and Schoenbaechler 1976), the starling (Dmi'el and Tel-Tzur 1985), the chukar and sand partridge (Frumkin et a/. 1986), and the jack rabbit (Dawson and Schmidt-Nielsen 1966).

The emu does not conform to this pattern. Its dry thermal

conductance at T a=45°C is greater than at any other T 8 . At T a=45°C there was 2.4 (females) and 2.7 (males) times as much heat flowing into the body, resulting in EWL rates 24o/o (females) and 28% (males) higher, than if dry conductance were manipulated to the level observed at T a=25°C. This begs the question, why is the dry conductance so high at T a=45°C? High conductance at Ta=45°C may be the inevitable result of the contribution of CEWL to evaporative cooling. At T a=45°C approximately 30% of total EWL is CEWL in the emu. In other bird species, CEWL appears to be dependent on high levels of skin blood flow (Marder and Raber 1989). The maintenance of high levels of skin blood flow will also lead to high thermal conductance, thus further increasing heat flow from the environment into the body.

There would appear to be a trade off involved in using CEWL. The bird could vasoconstrict the periphery, absorb less heat, and so need a lower relative level of EWL, but the majority of that EWL would have to be achieved by REWL (see for example: Hinds and Calder 1973, Dmi'el and Tel-Tzur 1985). Alternatively it can rely more on CEWL, but consequently need a higher relative level of EWL to dissipate the extra environmental heat load (see for example Arad et a/. 1987, Withers and Williams 1990). Given that most bird species are threatened by pH imbalance when panting (Calder and 97

Schmidt-Nielsen "1 968, Marder and Arad "1 989), the trade off may be one of water economy versus pH balance. This trade off is undoubtedly affected by acclimatization, which is well illustrated by pigeons acclimated to high Ta, but with constant access to water (Marder et a/. 1989, Marder and Gavrieli-Levin "1 987). At high T as these pigeons do not pant, but dissipate all metabolic and the external heat load by CEWL, whereas non-acclimated pigeons pant and develop blood alkalosis (Calder and Schmidt-Nielsen "1 966, Bernstein and Samiengo "1 98"1 ).

The emu, however, presents an interesting case. It can pant without developing a severe pH imbalance ('Jones eta/. "1 983), so why should it choose the disadvantages associated with CEWL? Possibly diurnal foraging would be compromised by a need to pant on hot days.

3.4.5 More Female-Male Differences At the higher Tas used in this study (35 and 45°C) the EWL of females was significantly higher than that of the males (Figure 3.8). This is a result of the higher metabolism of the females (Figure 3.6) giving them a larger endogenous heat load to dissipate. The slightly higher Tb of the females, however, gives them a larger temperature gradient for dry heat loss at Ta=35°C. At Ta=35°C the Tbs of the sexes were significantly different in summer, but not in winter, so the calculations here will utilize the summer results. The conductance at Ta=35°C of the females was 3.4 watts/°C, and for the females and males respectively, metabolic rates were 45.5 and 33. "1 watts, EHL was 33.8 and 23.7 watts, and Tbs were 38.3 and 37.8°C. The females thus had a rate of dry heat loss "1 .8 watts 98

higher than the males. This helped to dissipate part of the 12.4 watts of additional endogenous heat they produced, but left a difference of 10.6 watts to be evaporated. The measured EHL of the females was 10.1 watts higher than the males, which compares favourably given that errors in conductance estimates become large at Tas close to Tb (Dawson and Schmidt-Nielsen 1966).

At Ta=45°C the 25% higher metabolism of the females again placed a larger burden on their EWL mechanisms relative to those of the males, resulting in the females having a 20% higher rate of EWL at this Ta.

3.4.6 Heat Balance The contribution of each form of heat transfer (conduction and convection, radiation, and respiratory and cutaneous evaporative water loss) to heat balance at each T a is shown in Figure 3.11. Estimates of radiative heat exchange are influenced by factors affecting the estimation of surface area (Section 3.4.3). Dry heat exchange follows mainly convective and radiative routes. The calculated rate of convective heat loss at low Tas accounts for about half of the dry heat loss, which might be expected given that both are proportional to surface temperature. However, convective exchange is influenced by wind speed while radiative exchange is not. Convective conditions in the chamber were not measured. With increasing T a the importance of evaporation increases as the Tb- T a gradient decreases and the emu is not able to dump all of the metabolically produced heat by non-evaporative means. At Ta=45°C the gradient for dry heat flow is into the emu and it is gaining heat by both radiative and convective routes. To maintain Tb the emu 99

dissipates this environmental heat load, plus the metabolically derived heat, by increasing the evaporation of water accordingly.

3.4. 7 Legs as Radiators The contribution of the legs to total radiative heat loss is shown in Figure 3.12. Since radiative and convective heat loss are proportional to surface temperature, the radiative loss from the legs will presumably reflect the contribution of the legs to total heat loss. At the low T as, the surface temperature of the legs fell considerably (Figure 3.1 OF). Because the legs are poorly insulated, such a low surface temperature is· an indication that the interior of the legs was not maintained at Tb. This does not mean that blood flow to the legs was retarded. Hyrtl (1863) has shown that the arteries and veins at the top of the emu's legs are arranged in a fashion that allows counter-current heat exchange (simple rete, sensu Midtgard 1981 ). Blood flow may be maintained with lower heat loss than if the counter-current exchanger were not present. If the surface temperature of the legs was equal to Tb at Ta=-5°C, radiant heat loss from the legs would be 29 watts, compared to the heat loss of 9 watts when surface temperature is the observed 1 0°C. Even with the rete, 20-30% of total radiant heat loss occurs from the legs at low T as. An indication of the effect of removing the legs as sources of heat loss can be gained by comparing the heat production of sitting and standing emus at low T as. Observations of emus sitting and standing at thermoneutral MR suggest that the cost of standing, per se, is not very large. Much of the reduction in MR in sitting emus will be due to reduced heat loss from the legs. 100

At Ta=25°C the controlling system would be shifting demand from heat conservation to heat dissipation. At this T a. radiant heat loss from the legs increases, indicating that leg surface temperature is relatively warmer (compared to the Tb- T a gradient) than at lower T as and that the legs may be acting as 'heat windows'. Poorly insulated areas of animals can be important areas for heat dissipation, increasing dry heat loss at moderate T as. Examples of such heat windows are elephant's ears (Phillips and Heath 1992), bare skin areas under the wings of ostriches (Louw et a/. 1969), and the legs and feet of birds (Steen and Steen 1965, Bernstein 1974, Kilgore and Schmidt-Nielsen· 1975, Baudinette et a/. 1976, Midtgard 1980, Arad et a/. 1989).

At T a=35°C, leg surface temperature was close to T a· In some individuals it was below T a. an indication that water was evaporating from the legs. Cutaneous evaporative water loss from the legs may be important to emus exposed to high T as and a radiant heat load. During periods of high insolation (i.e. around midday) the legs will be shaded by the body, and with CEWL could act as efficient heat dissipaters. This could also explain why emus are observed crouched in creek beds on very hot days in summer in the arid zone (T a>42°C, Dawson et a/. 1984). Crouching would expose the evaporating surfaces of the belly and legs to any ambient convection and increase evaporation. 1 01

Chapter 4

Ventilatory Accommodation of Oxygen Demand and Respiratory Water Loss

4.1 Introduction

The modelling of homeostatic systems in animals usually involves a central controller receiving input from sensors and, according to that input, sending a signal to an effector. In simple terms this describes homeothermic temperature regulation (Simon, Pierau and Taylor 1986). However, the homeothermic temperature regulation system has been described as an exception among homeostatic systems since, apart from brown fat and sweat glands in mammals, there are no effectors specific to the temperature regulation system (Mercer 1989). Effective control of body temperature thus relies on the controller usurping effectors from other functions. Circulation is manipulated according to the need to dissipate or retain heat. The ventilatory system of panting animals is used to increase evaporative water loss and in cold temperatures the metabolic machinery of muscle produces heat during shivering. An interesting aspect of this control is the involvement of the ventilatory system in the response to both cold and heat challenge. At high temperatures increased ventilation facilitates evaporative water loss, while at low temperatures the ventilatory system must accommodate the increase in oxygen demand associated with increased heat production. 102

4.1.1 In the Cold The supply of oxygen by the ventilatory system to satisfy metabolic demand involves three variables: the breath frequency (fR), the size of those breaths (tidal volume, VT), and the percentage of the oxygen in each breath which the animal ultimately uses (oxygen extraction, E02). Each of these variables has limits, although the number of combinations is theoretically infinite. The combination chosen by different species to satisfy V02 may reflect a minimum cost to respiratory muscles (Dejours 1981) or physiological, phylogenetic, or ecological differences (Chappell and Souza 1988). Despite the paucity of data, two general ventilatory responses of birds to lowered T a have been identified. In some species, E02 remains constant and the increased vo2 is met solely by increased ventilation, either through increases in fR or VT or a combination of the two. Other species increase E02 at low temperatures, and make up any shortfall in oxygen demand with changes in ventilation. Increasing E02 at low temperatures has often been interpreted as a mechanism to reduce respiratory heat loss (Johansen and Bech 1983). Chappell and Souza (1988) have questioned this interpretation since at low temperatures vo2 usually increases as well.

In this chapter I report studies on the ventilatory response of the emu to lowered T a. I also discuss the ventilatory response to low Ta at different levels of V02 (sitting down at low temperature and standing up at the same temperature). These data are used to distinguish between the effect of temperature and oxygen demand on ventilation. In these situations V02 varied by nearly 100%: 103

standing up at -5°C it is 2.6 x BMR, while sitting down it is 1.5 x BMR (Chapter 3).

4.1.2 In the Heat When birds are exposed to heat challenge, the flow of air over the moist respiratory surfaces is increased by panting. This increase in air flow is a potential problem since it can lead to C02 washout from the blood, and alkalosis. Jones et a/. (1983) showed that the emu can pant at 46°C for several hours and develop only mild hypocapnia (Paco2 going from 32 to 29 mm Hg). In that study qualitative estimates were made of VT during thermal stress. In this study VT was quantified during· mild (30, 35°C) and extreme (45°C) heat challenge. Ventilation frequency under these circumstances is also reported.

4.2 Materials and Methods

4.2.1 Experimental Animals Ventilation data were obtained concurrently with metabolism measurements (Chapter 3). The same animals and procedures were used in these experiments.

4.2.2 Data Collection The chamber used for metabolism measurements also acted as a whole body plethysmograph. Pressure changes caused by the warming or cooling and wetting of inspired air were measured with a pressure transducer (Omega PX164-01 0). The voltage output from the transducer was monitored every 0.1 or 0.2 seconds by a personal computer via an analog/digital converter. Files saved to 104

disk contained 64 or 128 seconds of ventilation data. The system was calibrated by injecting a known volume of air (630 ml) into the chamber after each experiment. The injection procedure was repeated ten times in a fashion such that deflection kinetics were similar to that during animal respiration. A mean value of the ten injections was used in calculations. To facilitate calculation of E02, a marker was placed on a metabolism file whenever ventilation data was saved to disk. This allowed for calculation of vo2 and vco2 for each ventilation file.

4.2.3 The Barometric Method The barometric method is the only means known of obtaining estimates of VT from undisturbed animals. The method has evolved somewhat since it was first applied by Chapin in 1954. The early closed systems involving a reference chamber have been replaced by open-flow systems incorporating a pressure transducer referenced to atmospheric pressure. Flow of dry air into an open­ circuit chamber alleviates the pressure build up inherent in the design of closed (Drorbaugh and Fenn 1955) or 'slow-leak' systems (Jacky 1978, 1980) due to the gradual warming and humidification of chamber air, and eliminates the need for base line corrections, such as those applied by Drorbaugh and Fenn (1955). In open-flow systems, chamber pressure, temperature, and humidity are held constant and so inspiratory pressure deflections are 'true' estimates of tidal volume (Bucher 1981 ), and corrections for the different pressure changes caused by inspiration and expiration, based on inspiration time and expired air temperature (Epstein and Epstein 1978, Epstein et a/. 1980, Jacky 1978, 1980), are not necessary. Using an open-flow system for measurements on the 105

little penguin, Stahel and Nicol (1988a) showed that uncorrected VT estimates (i.e. VT estimated directly from pressure deflections) were not significantly different from simultaneous estimates of VT obtained using pneumotachography. Correcting the barometric VT estimates for inspiration time and expired air temperature resulted in an estimate of VT which was 6.6% higher than the simultaneous pneumotachographic estimate of VT, although this increase did not prove to be statistically significant. In the experiments reported here, VT is estimated from pressure deflections on ventilation traces using equation 6 from Malan (1973), assuming lung temperature was equal to body temperature.

When ambient temperature approaches or exceeds an animal's body temperature the barometric system theoretically becomes unreliable, because the pressure change associated with temperature change of inspired air approaches zero. The low humidities used in these experiments resulted in the pressure change due to wetting of inspired air being quantitatively more important than temperature change at high T as. At T a=45°C the pressure increase due to wetting of an average tidal volume was twice the pressure decrease due to cooling of that air to body temperature. Repeatability of calibration injections is greater when simulating rapid inspiration kinetics than when inspiration is slower (pers.obs.). At the higher temperatures inspiration was very rapid, so errors in VT estimates will result mainly from errors in measurement of T a, Tb, or chamber water vapour pressure (which was calculated from chamber excurrent air temperature [Trhl and relative humidity [RH]). Overestimates of VT will result from underestimating Tb, or overestimating T a, Trh, or RH. Compounding 106

errors of 1°C in Ta, 0.3°C in Tb, 2°C in Trh, and 2% in RH, could result in a maximum 21% overestimate of VT.

4.2.4 Expired Air Temperature At ambient temperatures below 35°C expired air temperature (T ex) of eight emus was measured with a copper-constantan thermocouple held at the opening to the nares. I attempted to tape or glue the thermocouple in position and obtain undisturbed measurements of Tex. but all emus successfully dislodged the thermocouple. Eight of the emus would stand or sit quietly, seemingly undisturbed, while a thermocouple was held at the opening to the nares. Attempts to 'obtain measurements from the other two emus were abandoned since they were always agitated in the presence of a person. At ambient temperatures of 35 and 45°C the birds exhibited open mouthed panting. At these temperatures the thermocouple was held so that the tip was just above the trachea. Only three o.f the birds were co-operative for this procedure, and at these temperatures T ex is an average of these three emus. Temperature of the thermocouple was logged at 0.1 second intervals on a personal computer, via a thermocouple reference/ amplifier (AD595 chip and custom made circuit) and 12 bit analog/ digital converter (Sable Systems USA). The thermocouple was calibrated against a certified mercury in glass thermometer.

4.2.5 Data Analysis Between 5 and 25 files of ventilation data were collected for each animal at each temperature (Winter; -5, 5, 15, 20, 25, 30, 35, 45°C: Summer; 5, 15, 25, 35, 45°C). Ventilation data were rejected if V02 107

differed by more than 10% from MR for that animal at that temperature (Chapter 3). For statistical analysis a mean value for each variable was calculated for each animal at each temperature to avoid biasing in favour of animals with many samples. Tidal volume is presented as BTPS (body temperature, ambient pressure, saturated), but was converted to STPD (standard temperature and pressure, dry) for determination of E02, which was calculated as:

E02(%)=[V02/(FEo2 · VT · fR)] · 100, where FEo 2 is the partial concentration of oxygen in chamber excurrent air.

4.2.6 Statistical Analysis Data were analyzed using a Two-Wa.y Repeated Measures Anovar for sex and temperature. A Student-Newman-Keuls multiple range test was used when significant differences were indicated by the Anovar. All analyses were performed on a personal computer using Statistica/Mac software. If sex proved significant, the sexes are presented separately, otherwise data for both sexes are grouped. All graphs and tables show means ±SEM. Means are considered significantly different if p<0.05.

4.3 Results

There was no difference between summer and winter ventilation responses at any ambient temperature. Since measurements were made at eight temperatures in winter and five in summer, the analysis will concentrate on the winter results. 108

4.3.1 Ventilation While metabolism proved to differ significantly between the sexes (Chapters 2, 3), this difference was not accounted for by any single ventilatory variable. Females tended to have higher breathing frequencies (fR) at all temperatures (Figure 4.1 A) and some females tended to have larger VT (Figure 4.18), although no pattern was consistent between the sexes. This high intra-sexual variation meant that there were no significant differences in either fR or VT between the sexes. There was a difference in ventilation (VI) between the sexes (Figure 4.2) at ambient temperatures higher than 30°C.

The increased demand for oxygen at temperatures below 20°C (Figure 4.3A) was accommodated by an increase in VT (from 783 ml at 2S°C to 1173 ml at -S°C; Figure 4.3C), and an increase in E02 (from 21% at 2S°C to 28% at -S°C; Figure 4.30). These changes in VT and E02 accounted for 47% and 31%, respectively, of the change in V02 between 2S and -S°C. An increase in fR accounted for the other 22%.

At the higher temperatures the shift in demand on the ventilatory system from oxygen delivery to respiratory evaporative water loss (REWL) is indicated by the drop in E02 to 2.6% at 4S°C (Figure 4.30). The increase in VI at higher temperatures (Figure 4.2) is achieved by increasing fR (Figure 4.38). There was a significant drop in VT at 30 and 3S°C, to S81 and 607 ml respectively, compared to 783 ml at 2S°C. At 30°C all emus increased fR and dropped VT (Figure 4.38, C). At 3S°C individual ventilatory responses varied (Table 4.1 ). In summer two of the females increased fR to levels observed at 109

Figure 4.1 Ventilation frequency (A), tidal volume (B), and oxygen extraction (C), of female and male emus at a range of ambient temperatures 1 1 0

40 c

35 l ...... Q) -:::s c 30 1~ .E X

(J) 25 -Q) '- :!:::::: 20 b -c T 0 15 ! ·ca females a • ·.;::::; 10 2 c .,..a y 0 Q) a males a +a > 5 'o 'a z z z •o 0 -10 0 10 20 30 40 50 Ambient Temperature (°C)

Figure 4.2 Ventilation of female and male emus at a range of ambient temperatures (different letters on the same sex denote significant differences between means; SNK P

45°C, with VT being lower. The remainder of the birds regulated REWL by switching between a lower 'breathing' and a higher 'panting' frequency. Tidal volume was significantly lower when the animals were 'panting' than when 'breathing'(n=8, p=0.007). In winter at 35°C Ta. two females 'panted' the whole time, one male 'breathed' the whole time, and one male ventilated at a frequency intermediate between the average of the two levels. The remainder of the birds exhibited the switching response, with VT again being lower during 'panting' (n=7, p=0.013). Mean calculated REWL was 25% higher for females than for males in both seasons, although this result failed to reach significance (summer p=0.09, winter p=0.08). 1 1 1

Table 4.1 Ventilatory responses of individual emus at an ambient temperature of 35oc

Sex breath pant %time Ave- REWL

panting rage (gH 20/h) fR VT fR VT fR VT Summer Q 8.9 722 41.8 507 48 24.7 619 35.6 Q 10.5 1004 42 532 51 26.6 764 47.1 Q 6.6 1235 40.6 421 52 24.3 812 47.1 Q 42.3 456 100 42.3 456 44.3 Q 39,7 464 100 39.67 464 39.4 cJ 1 1 496 38 434 44 22.9 469 26.3 cJ 8.2 895 28.7 687 42 16.8 807 32.8 cJ 1 5 463 38.4 352 60 29.05 396 28.3 cJ 21.7 735 44.7 402 55 34.3 552 36 cJ 8.3 1122 40.1 493 43 22 851 28 Winter Q 38.5 484 100 38.5 484 46.1 Q 22.4 463 48.3 520 59 37.6 496 45.6 Q 9.2 879 33.3 553 26 15.47 794 31.13 Q 40.6 449 100 40.6 449 43.8 Q 10.6 774 34.6 546 42 20.6 680 33.7 cJ 24.5 384 100 24.5 384 24.5 cJ 16.7 794 31.6 424 45 23.4 628 35.14 cJ 10.9 801 31.7 457 34 1 8 684 29.8 cJ 13.24 851 37.8 616 30 20.6 780 40 cJ 9.9 700 0 9.9 700 17.4 112

At T a=45°C VT was 804 ml (Figure 4.3C). It was assumed in the tidal volume calculations that the emus inspired air which was at the average chamber humidity. The validity of this assumption depends on the rate of mixing of the chamber air and will become

so B

-c 40 'i§ .....4) C"' J .::J. c ...... 'i§ 30 c N ...... 0 0? ....I ..c. E -g 20 lo.. -N OJ 0 > 1 0 a. X 0

1300 ca. 35 D I i -30 1150 1 1 ~ ';;' 25 0 ~ 1000 :;::;: 4) g 20 E be ,;; :=!. 8 50 0 ~ ~ > ~ c 15 4) -<11 C"' i= 700 ::71 ~ 10

550 5 d X 400 0 - 10 0 1 0 2 0 3 0 40 50 - 1 0 0 1 0 2 0 3 0 40 50 Ambient Tempereture (° C)

Figure 4.3 Oxygen consumption (A), ventilation frequency (B), tidal volume (C), and oxygen extraction (D), of emus at a range of ambient temperatures (different letters denote significant differences between means; SNK P

more important at the higher ambient temperatures, where relatively larger saturation deficits exist in the chamber. If an emu inspired air with a water vapour pressure greater than the average in the chamber VT would have been underestimated. Tidal volume was used to estimate REWL and thus CEWL in Chapter 3. The difference between the measured rates of CEWL and the rate estimated using VT indicate that VT was not underestimated by more than 10%.

4.3.2 Gular Flutter When heat stressed, many species of birds use the gular area to facilitate REWL (Bartholomew et a'J. 1968). The process of gular flutter involves rapid oscillation of the gular area, which increases the movement of air over this highly vascularised region. In the emu, gular movements were synchronised with respiration, a pattern also exhibited by the ostrich (Schmidt-Nielsen et a/. 1969). The gular region was raised at the end of expiration and remained raised during most of the inspiration, falling just before the next expiration began. This process would result in low pressure at the back of the nasal turbinates during inspiration. This could be important in establishing flow of unsaturated air through the turbinates (Figure 4.4) during inspiration. The nasal turbinates in the emu are lined with water secreting serous cells, and the veins from it drain into an opthalmic rete (Smith and Dawson unpublished results). In other species this arrangement can be used to cool blood destined for the brain. However, no measurements of air flow through the turbinates were attempted. 114

Figure 4.4 Emu Head Dissected to Expose the Nasal Turbinates · 11S

4.3.3 Sitting and Standing in the Cold The difference in V02 between emus sitting and standing at -S°C is 1.S and 2.6 x BMR respectively (Figure 4.SA). Ventilatory parameters for standing and sitting emus at -S°C, and emus at

A B 350 ... 6 r- b T c 300 f.. - a ~ ·-E 5 1- ...... T ...... Q .... 250 1- ...... a ::;, :::n 4 1- c (..) T c ·-E f.. 200 ::;, ...... a T t:r N 3 0 ..._~ ...J 150 c E c 0 ...... , ...... 2 fD N 100 0 ...... > ....c 50 >

0 0

1500 35 D a c- r T a b 30 ...... T T ...J -...... ~ E a c 25 b ';: 1 000 1- 0 T :;::: T E (..) ::;, c fD 20 ~ 0 ...... X > a..J <0 c 15 "C j:: 500 :::n0'\ X 0 10 5

0 0 -5sit -5stand 25 -Ssit -5stand 25

Figure 4.5 Oxygen consumption (A), ventilation frequency (B), tidal volume (C), and oxygen extraction (D), of emus at 2soc and sitting and standing at -5°C (different letters denote significant differences between means; SNK P

25°C, are shown in Figure 4.58, C, D. There was an increase in VT (Figure 4.5C) and E02 (Figure 4.50) in both standing and sitting emus at -5°C compared to 25°C, while fA increased in the standing emus but was the same as at 25°C in the sitting emus (Figure 4.58). A major difference in ventilation between sitting and standing emus in the cold is the tendency for higher E02 in sitting emus. This can be illustrated by calculating the relative contribution of each variable to the increase in V02 (Figure 4.6). The predominance of E02 in the accommodation of increased V02 in the sitting emu becomes obvious.

1 so ..._

...... 0 •••• ~ • 0 •••• • 0 • • 0 • 0 • •• • • • 0 •• 0 • • •• ....._, • 0 •• 0 • ••• • 0 . .. • • • 0 0 0 ••• 0 • 0 •• • 0 • 0 • 0 • • 0 •• 0 • 0 • • ••• •• 0 ...... • 0 • • • • •••• 0 • • • • • • • • 0 0 ••••• 0 • 0 • 0 • 0 • • • • 0 •• c • • • • 0 0 0 • 0 ••••• • 0 •• • 0 •• • • 0 • • • • • 0 • • • 0 • • •••• 0 100 0 0 • 0 • 0 • • • • 0 •• • • • 0 0 •• 0 • •• 0 •• • • • • 0 ••• • •••• • • • 0 ••••• 0 •••• •• •• 0 0 • •• • • •• •• • •••• 0 • ••••• • 0 •• 0 0 • • • • • •• ':;) •• • 0 • • 0 0 • 0 • 0 0 • 0 •• 0 0 • • 0 • •• 0 • • • • 0 ••• 0 0 •• 0 0 0 0 0 • • • 0 • • ••• •• - • • • • • • • • • 0 • • •• .0 0 •• • • 0 • ••••• 0 • • • 0 • 0 •• • • • 0 • 0 II fR • ••••• •• •••• 0 0 • 0 0 0 •• 0 •• 0 • 0 • ~ 0 • •••• • 0 ••• • • • • 0 ...... 0 • •• 0 0 ...... 0 •••• • 0 0 • •••• • • • ••• • • • 0 •• ••• • • • • c • • • • • 0 • ••• •• 0 • ...... ~ ... - . ••• . . . ••. . • • . • 0 . 0 . • . • . • . lill1 Vr 0 so (_) ~ 0 E02 > -~ ~ 0 a:

-50 Sitting Standing

Figure 4.6 Relative contribution of ventilatory variables in the accommodation of increased oxygen consumption in sitting and standing emus at -5°C, compared to 25°C. 117

4.3.4 Expired Air Temperature Measurements of expired air temperature (Tex) at various T as are shown in Figure 4.7. These results indicate that the emu is capable of exhaling air which is considerably cooler than body temperature when ambient temperature is low. At ambient temperatures of 35°C and above (open mouthed panting) Tex was slightly lower than Tb

..--.. 40 0 n=3 0...__, n=3 X )( ,__

Figure 4.7 Emu expired air temperature at a range of ambient temperatures

4.4 Discussion

4.4.1 Thermoneutral Ventilation Resting ventilation frequencies have been reported twice previously for the emu. Jones et a/. (1983) measured a rate of 5.3 breaths/minute (b/min) at 21 °C, which is similar to the values of 4.5 (summer) and 4.8 (winter) b/min measured at 25°C in this study. Oxygen consumption was not measured during the 1983 study. Crawford and Lasiewski (1968) measured a resting 11 8

respiratory frequency of 7.1 b/min for their emus, which were hooded and restrained. This may have affected respiration (see Bucher 1985). Additionally, oxygen consumption was 17% higher in their study than that obtained for females in the present study, so ventilation was not measured under basal conditions.

Allometric relationships for resting ventilatory parameters in birds have been reported by Calder (1968), Lasiewski and Calder (1971) and Bucher (1985). The analysis of ventilatory parameters is confounded by the dual function of the ventilatory system in birds, it being used for oxygen delivery as well as for respiratory water loss (Bucher 1985). The type of problems identified by Bucher (1985) are well illustrated by results for the emu. Metabolism was consistently low at Tas from 20 to 35°C but ventilatory parameters changed markedly in this region. Respiratory frequency varies from 4.5 to 25.1 b/min (Figure 4.3B), tidal volume from 796 to 581 mL (Figure 4.3C, or 493 mL during 'panting' at 35°C), and oxygen extraction from 23% to 7% (Figure 4.30). The difficulty then is in defining 'resting' for ventilatory variables just as 'basal' is used to define metabolic measurements. Bucher (1985) proposed that a single specified T a. such as the lower critical temperature or the T a at which fR is minimal, be used when reporting standard values for comparative purposes. Additionally, she presented evidence that the method of measurement may influence the results obtained. Ventilation frequencies from studies where birds were restrained, masked, or anticipating exercise were consistently high relative to the allometric line for mass versus ventilation frequency that she obtained from unrestrained birds by plethysmography 119

(fR=i 0.3·Mass-0.32). She reasoned that the use of data from restrained, masked, or anticipatory birds in the Lasiewski and Calder (197"1) calculations could explain why that relationship (fR=i 7 .2·Mass-0.31) predicts fR values averaging 70% higher than the relationship she obtained. This contention is supported by ventilatory data from little penguins measured by plethysmography and by a Fleisch flow transducer inside a face mask (Stahel and Nicol 1988a). The use of a face mask caused a significant rise in fR, compared to unrestrained plethysmographic measurements.

Ventilation data from fifteen species of birds measured by whole body plethysmography and at ambient temperatures when fR is at a minimum have been published (Table 4.2). The least squares regression line of the log transformed data which best describes the data for fR versus body mass (kg) is; fR=i 3.8·Mass-0.308 (SEslope=0.04, r2=0.774; p

4.4.2 In the Cold Figure 4.30 shows that the emu increases E02 at low ambient temperatures. Part of the increased E02 would have been due to the 120

Table 4.2 Ventilation parameters measured by plethysmography at an ambient temperature when breath frequency is a minimum

Species Mass fR VT E02 Reference

Carpodacus mexicanus 20 36.5 0.4 44 Clemens 1988 Leucosticte arctoa 25 37.7 0.5 35.3 Clemens 1988 Oceanites oceanicus 34 74.8 0.6 14.5 Morgan et.al. 1992 Agapornis roseicollis 48 36.2 0.95 25.7 Bucher and Morgan 1989 Bolborhynchus linea/a 56 28.7 1.5 33.3 Bucher 1981 Geophaps plumifera 89 36.9 1.3 25 Withers and Williams 1990 Amazona viridigenalis 342 13.8 10.7 19.7 Bucher 1985 Alectoris chukar 475 23.3 6.4 25 Chappel and Bucher 1987 Falco mexican us Kaiser and Bucher 1985 Male 498 13.8 10.6 28 Female 755 13.6 12.1 23 Larus dominican us 980 18.1 29.1 27.8 Morgan et.al. 1992 Eudyptula minor 1082 6.9 14.9 55 Stahel and Nicol 1988b Catharacta 1250 11.0 46.4 30.9 Morgan et.a/. maccormicki 1992 121

Macronectes giganteus 3929 18.3 74.3 41.5 Morgan et.a/. 1992 Pygoscelis adeliae 4007 7.8 96.3 35 Chappell and Souza 1988 Dromaius 38350 4.5 783.1 21 This Thesis novaehollandiae increased VT (Figure 4.3C), resulting in the ventilatory dead space making up relatively less of each breath. The response of other species under similar conditions varies. Whether the increased E02 in those birds that increase E02 at low temperatures is a temperature or an oxygen demand driven response remains controversial. A major problem with resolving this controversy is the difficulty in uncoupling low T a from increased V02 in most species. In the adelie penguin, which has a lower critical temperature of -1 0°C, E02 at -20°C was less than at 10°C (Chappell and Souza 1988). Thus, in that species at least, low Ta did not stimulate an increase in E02. Data from four species of Antarctic sea-birds also suggests that low T a is not a driving force for increased E02 (Morgan eta/. 1992). In the emu the increase in V02 at -5°C in both sitting and standing emus was accompanied by an increase in E02 (Figure 4.5A, D). There was a tendency for E02 to be higher when sitting than when standing, despite a higher oxygen demand when standing. If perfusion of the parabronchi was similar under both conditions, the lower VI, and hence flow rate through the parabronchi would mean that each volume of air would be in contact with blood for a longer time in the sitting emu (i.e. it would have a lower V/ Q). A lower V/ a is favourable for oxygen 122

exchange, since 02 is limited by diffusion kinetics more than C02 (Berger eta/. 1979).

In contrast to findings in other species, it appears low Ta did stimulate increased E02 in the emu. The energy saving due to increased E02 was comparable to the energy savings calculated for other birds which increase E02 at low temperatures (standing emu 2%, sitting emu 4.4%, other birds 3-6%, Chappell and Bucher 1987, Chappell and Souza 1988). While this is a small fraction of an emu's daily energy expenditure the savings accrued while sitting at night in winter will reduce the time the emu needs to forage the following day by approximately 50· minutes.

4.4.3 Respiratory Anatomy and Gas Exchange Gas exchange by one-way air flow in parabronchi and cross-current blood supply in avian lungs is theoretically more efficient than in the alveolar lung. There are differences in the anatomy of the lung between bird species, with two general types being identified by Duncker (1972). The arrangement of parabronchi between dorsa­ and ventra-bronchi is thought to be the ancestral arrangement in · avian lungs and has been termed the paleopulmo (Duncker 1972). The penguins and the emu possess only this arrangement of parabronchi and consequently gas exchange is limited to the paleopulmo. All other birds (from 155 species in 47 families studied by Duncker 1972) have an additional network of parabronchi between the primary bronchus and the posterior air sacs, termed the neopulmo. The development of the neopulmo varies among species, being best developed in fowl-like birds and song birds. 123

Table 4.3 Maximum oxygen extraction reported for twenty species of birds

S ecific Name Max E02 Reference

Oceanites oceanicus 17.4 Morgan et a/. 1 9 9 2 Falco mexicanus 25 Kaiser and Bucher 1985 Rissa tridactyla 25.5 Brent et a/. 1983 Agapornis roseicollis 28.7 Bucher and Morgan 1989 Columba Iivia 28.8 Bech et a/. 1985 Amazona viridigenalis 29.5 Bucher 1985 Alectoris chukar 30 Chappell and Bucher 1987 Corvus ossifragus 30 Bernstein and Schmidt­ Nielsen 1974 Larus dominican us 30.9 Morgan eta/. 1992 Catharacta maccormicki 30.9 Morgan eta/. 1992 Cygnus o/or 3 3 Bech and Johansen 1980 Pygoscelis adeliae 3 5 Chappell and Souza 1988 Bolborhynchus linea/a 4 0 Bucher 1981 Anas platyrhynchos 41.4 Bech eta/. 1984 Geophaps p/umifera 42.9 Withers and Williams 1990 Macronectes giganteus 45.7 Morgan et a/. 1 9 9 2 Leucosticte arctoa 5 1 Clemens 1988 Carpodacus mexican us 52 Clemens 1988 Eudyptula minor 53.2 Stahel and Nicol 1988b Fulica atra 6 2 Brent et a/. 1 9 84

The levels of E02 for resting emus obtained in this study are not inferior to those reported for birds that possess a neopulmo (Table 4.3). The possession of only the paleopulmo lung does not appear to 124

be disadvantageous to gas exchange efficiency. One penguin species the little penguin, has the second highest E02 measured to date. It is worth noting that the neopulmo in the goose is responsible for only a small fraction of total gas exchange at rest (Scheid et a/. 1989). The ratio of air ventilation to blood perfusion in the paleopulmo (calculated to be 2 to 2.S times higher in the duck neopulmo than in the paleopulmo; data from Hastings and Powell 1986 reanalysed by Scheid et a/. 1989) is probably more important to gas exchange at rest than the presence or absence of a neopulmo.

4.4.4 Expired Air Temperature The major ventilatory adaptation for energy conservation in the cold is the manipulation of expired air temperature (Johansen and Bech 1983). Exhaling air at a temperature below body temperature results in considerable savings of both energy (heat) and water. The nasal turbinates in the emu are well developed (Figure 4.4) and permit the exhalation of air at a temperature cooler than body temperature (Figure 4.7). Exhaling air at 17°C, during exposure to -S°C ambient temperature, rather than at body temperature (38°C) results in a saving of 2.4 watts of heat, and 12.3 g H20/h, which has a heat equivalent of 8.3 watts. This saving of 10.7 watts is 11 °/o of the average heat production required to maintain Tb when standing during exposure to a T a of -S°C. When sitting the savings add up to 10.6% of the energy required to maintain Tb. The combination of increased E02 and lowered T ex at low ambient temperatures results in a saving of 16.7% (standing emu) and 19.7% (sitting emu) of the average heat production required to maintain Tb, compared to the situation if E02 remained at thermoneutral levels and air was expired at body temperature. 125

4.4.5 In the Heat The increase in fR at high temperatures to facilitate REWL is accompanied by a drop in VT at 30 and 35°C (Figure 4.3C). This is a common response to mild heat stress in many species of birds and mammals. An increase in fR accompanied by a drop in VT has been termed Phase I panting, to distinguish it from the increased fR and VT of Phase II panting during more severe heat stress (Hales and Webster 1967). The lower VT during Phase I panting is thought to ventilate only non-gas exchange areas of the respiratory system. In this way REWL can be increased without hyperventilating the parabronchi and blood gas homeostasis is not jeopardised. The regulation of REWL by switching between a 'breathing' and a 'panting' frequency, observed at 35°C in the emu, is similar to the diphasic or flushout ventilation in the black duck (Berger et a/. 1970), greater flamingo (Bech et a/. 1979), Chinese painted quail (Lasiewski and Seymour 1972), brown necked raven (Marder 1973), and the Pekin duck (Bouverot et a/. 1974). In these other species shallow panting is interrupted at regular intervals by deeper breaths, thought to provide increased parabronchial ventilation.

The regulation of REWL by switching between panting and breathing in emus, rather than panting at an intermediate frequency, suggests that the birds may be panting at the resonant frequency of the respiratory system. Panting at a resonant frequency has been found to be energetically less expensive than respiring at lower frequencies in dogs (Crawford 1962) and pigeons (Crawford and Kampe 1971). 126

The study of Jones et a/. (1983) showed that the emu could pant for several hours at T a=46°C and develop only a mild alkalosis. This was in contrast to many other birds which become markedly hypocapnic when panting (Calder and Schmidt-Nielsen 1968, Marder and Arad 1989). The ostrich is the only other bird which is known to be capable of maintaining blood gas integrity when panting during severe heat stress (Schmidt-Nielsen et a/. 1969). The easiest way to explain this phenomenon would be the presence of valves shunting air away from the parabronchi during panting, but no such valves have been found. Jones (1982a) suggested that aerodynamic valving could explain the shunting in large birds. Alternatively, the functional shuntiri'g may result from changes in airway resistance. Zeuthen (1942, as cited in Schmidt-Nielsen 1979) proposed that an increase in parabronchial smooth muscle tone could prevent parabronchial hyperventilation during panting. Molony et a/. (1976) demonstrated an increase in ventrobronchial airway resistance in duck lungs with decreasing levels of C02 in the ventilating gas, and King and Cowie (1969) have shown that application of cholinergic drugs to parabronchi, or stimulation of the vagus nerve, leads to increased parabronchial resistance in chickens.

Another possible means of impeding C02 washout from the blood during panting is to manipulate blood flow to the lung. Jones (1982a, b) concluded that there was a ventilation I perfusion (V/Q) inequality during panting in the ostrich, but that shunting was required in addition to account for blood gas maintenance. Later studies on the emu (Jones eta/. 1983) showed that V/Q inhomogeneities were responsible for blood gas maintenance. 127

Shunting would lead to diminution of both C02 and 02 exchange due to a reduction in parabronchial airflow resulting in hypoxia and hypercapnia. Emus panting for several hours became slightly hypoxic but paradoxically also slightly hypocapnic. The explanation, the authors reason, is a serial inhomogeneity in blood flow where blood perfuses only the proximal end of each parabronchus. This means that the time blood and air are in contact would be reduced. Such a reduction affects 02 exchange more than C02 exchange, because of the steeper C02 dissociation curve and higher solubility and effective diffusivity of C02 (Berger et a/. 1979). 128

Chapter 5

The Effect of Dehydration on Thermoregulation

5.1 Introduction

The emu's range in Australia extends into the arid interior, where rainfall is low and unpredictable (Stafford-Smith and Morton 1990). This results in marked fluctuations in surface water availability, and plant biomass and water content. During periods of drought animals may face prolonged periods of water scarcity and so increased tolerance to dehydration would be a useful survival strategy (Schmidt-Nielsen 1964). No matter what water conservation measures are engaged some loss of water is inevitable as a result of ventilation of the lungs and the need to excrete metabolic waste products.

When emus were dehydrated and offered dry feed by Dawson et a/. (1983), water efflux was reduced from 44.8 mllkg·day to 14.8 mL/kg·day, which resulted in a net water loss of 7.0 ml/kg·day. When incubating, the male emu can reduce efflux to less than one half of the dehydrated values (B~ttemer and Dawson 1989). Part of this reduction in efflux would have resulted from the reduced faecal output brought about by the emus not feeding after a few days without water and not eating at all when incubating (Dawson et a/. 1983, 1991 ). Additional economy can be afforded by a reduction in glomerular filtration rate and increased reabsorption of water from the filtrate by the kidneys, which results in an increased urine/plasma concentration ratio and a decrease in urine 129

flow rate (Dawson et a/. 1991 ). Additionally there may be further water removal from the urine when it is refluxed into the rectum (Skadhauge eta/. 1991 ).

However, regardless of the emu's ability to reduce water loss, if intake is severely restricted the animal will suffer some degree of osmotic stress. When emus are dehydrated, water is not lost equally from the compartments making up the total body water pool. The reduction in blood volume is disproportionately low compared to total body water loss, indicating that blood volume is preferentially maintained at the expense of other compartments (Dawson et a/. 1983).

During his osmoregulatory study on dehydrated emus, Skadhauge (197 4) noted that when emus were dehydrated they did not pant, even when ambient temperature rose to 38°C. Only when water was made available did they commence panting. Since panting is a major pathway for heat loss at high Ta in emus, this could affect thermoregulation and body temperature. The experiments reported in this chapter were undertaken to examine the effect of water restriction on thermoregulation in the emu.

5.2 Materials and Methods

Five emus were used in these experiments (2 female, 3 male). The emus were transported to the University two weeks before experimentation began, for the usual familiarisation procedure (Chapters 2, 3) during which time food and water were available in the pens ad libitum. Following the familiarisation period, 130

measurements were made of V02, VC02, EWL, fR, and Vr at ambient temperatures of 25 and 45°C, using the materials and methods described in Chapters 2, 3, and 4. Body temperature was measured with a calibrated Jenco electronic thermometer within five minutes of the end of metabolism measurements at each Ta. Calculations of REWL, CEWL, dry conductance, and E02 were done as described in Chapters 3 and 4. These measurements were used as the 'control' values for each animal. Several days after the last control measurements were obtained, 10 ml of blood was withdrawn from the jugular vein of each emu into heparinized syringes (lithium heparin, Sigma Chemicals). The next day the emus were weighed to the nearest 0.1 kg on a platform balance (Wedderburn Scales, Sydney) and water was removed from each emu's pen.

5.2.1 Preliminary Experiments In an initial set of experiments water was withheld for 14 days and then the emus were exposed to T as of 25 and 45°C. During these experiments the emu's Tb was monitored continuously using the temperature transmitter probes described in Chapter 2. In these experiments three of the emus showed no deviation from the control pattern of thermoregulation when exposed to a T a of 45°C. The other two emus postponed the onset of panting until Tb had risen by at least 1°C, and showed a reduction in the EWL needed to maintain this elevated Tb. Analysis of the blood samples from the three emus exhibiting the control response at 45°C following dehydration revealed no increase in plasma osmolality (average increase = 2.3%, n=3, paired t=3.6, P=0.07). The osmolality of the plasma of the other two emus had increased by more than 7%. 131

Consequently it was decided to repeat the experiment using an increase in plasma osmolality of at least 7% as the criterion for dehydration.

5.2.2 Experimental Procedure Five emus were used in the second set of experiments which were carried out ten months later. Control measurements were carried out as described above. After 16 days without water, 2 ml of blood was obtained from each animal and plasma osmolality measured. This procedure was repeated every second day until the desired increase in plasma osmolality was observed. When it had been determined that plasma osmolality !lad increased by more than 7%, 10 ml of blood was obtained from each emu. The birds were then weighed and exposed to Tas of 25 and 45°C in the metabolism chamber, and a 'dehydrated' set of measurements made. During these experiments Tb was monitored continuously using the temperature transmitter probes described in Chapter 2. Expired air temperatures at T a=45°C were also measured in three co-operative dehydrated emus by the method outlined in Chapter 4.

5.2.3 Continuous Body Temperature Measurement On day 15 of dehydration, three of the emus were fitted with temperature transmitters and Tb was recorded for 24 hours by the methods outlined in Chapter 2.

5.2.4 Blood Analysis Within 30 minutes of blood being obtained from an emu, five aliquots were taken and plasma hematocrit determined using a micro hematocrit centrifuge and reader (Hawksley and Sons 132

Limited, England). The remainder of the blood was centrifuged for ten minutes in a Rota Uni II centrifuge and the plasma subsequently placed in sealed tubes and refrigerated for later analysis. Plasma osmolality was measured using a Fiske One-Ten freezing point depression osmometer calibrated with distilled water and a 400 mosm/kg standard solution. Plasma sodium and calcium were determined on a GBC 906AA atomic absorption spectrophotometer after dilution in cesium chloride (sodium) or lanthanum chloride (calcium).

5.2.5 Statistical Analysis Control values for each variable were compared to the dehydrated values by paired t-tests using Statview software on a personal computer. Each emu acted as its own con'trol and differences between the sexes (Chapters 2,3) did not affect analysis. Results were considered significantly different if p<0.05. Tables show means ± the standard error of the mean.

5.3 Results

All of the emus had ceased eating by day four of dehydration which is consistent with past observations on water restricted emus (Dawson et a/. 1983, 1991 ). After 16 days without water, the plasma osmolality of two of the emus had increased by more than 7% (increases of 7.7 and 11.9%). The third emu was dehydrated for 18 days (9.1% increase in plasma osmolality). After 20 days without water, plasma osmolality of the other two emus had not risen by 7% and it was decided not to prolong the dehydration. Dehydrated measurements were made on these emus on days 21 and 133

22 of dehydration. An initial attempt to perform measurements on one female was abandoned when an attempt to escape from the harness used to transport the emus resulted in her Tb rising to 39.8°C before she was placed in the chamber. Ambient temperature was 24°C and she was not panting. When her Tb rose to 40°C, she began to pant. Measurements on this animal were postponed until the next night. The next night, when exposed to a T a of 45°C, she began to pant when her Tb rose to 38.8°C

6.5 6 -C) ~ • 5.5 en -en 0 5 ....! • • en 4.5 en ctS 4 • ::E 3.5 • 3 15 17.5 20 22.5 25 Days Dehydrated

Figure 5.1 Mass loss during dehydration versus the duration of dehydration

Mass lost during the dehydration period varied from 3.6 to 6.0 kg. Mean mass loss was 4.8 kg which was 11% of average mass before dehydration. The emus dehydrated for the longest time tended to have a lower mass loss (Figure 5.1 ). The regression of mass loss on days dehydrated was not statistically significant (n=5, r2=0.5, P=0.2), however it is an indication that the activity, and hence daily energy expenditure, of the emus in their pens varied considerably. The change in blood osmolality also tended to be less tor the birds that were dehydrated for longer but this correlation 134

Table 5.1 Characteristics of the blood and plasma of hydrated and dehydrated emus

Osmolality Hematocrit Na Ca Na/Ca

CONT 295.1 ± 1.0 44.0 ± 0.6 138.6 ± 0.8 3.09 ± 0.2 45.4 ± 2.4 DEHYD 318.6 ± 3.5 48.5 ± 0.9 149.1 ± 2.5 3.3 ± 0.3 46.5 ± 3.9 0.002 0.004 0.01 0.33 0.69 was not significant either (n=5, r2=0.6, P=0.12). Characteristics of the blood and plasma of the emus are given in Table 5.1. The percent increase in plasma osmolality following dehydration varied from 5.0 to 11.9%, with an average increase of 8%. Plasma hematocrit increased from 44.0 to 48.5% during the dehydration. Plasma sodium concentration increased from 138.6 to 149.1 mmoi/L, while plasma calcium levels did not change. The ratio of sodium to calcium in the plasma was not affected by the dehydration.

40 -0 0 39.5 -Q) '- :::::::1 39 ai '- Q) c. 38.5 E Q) 1- 38 >. "'C 37.5 co0 37 1600 2000 2400 0400 0800 1200 1600 Time of Day

Figure 5.2 Twenty four hour body temperature record of a dehydrated emu 135

One of the emus whose Tb was monitored continuously on day 15 of dehydration subsequently was shown to have experienced a 7.7% increase in plasma osmolality up to this time. His Tb for the 24 hours from the afternoon of day 15 of dehydration is shown in Figure 5.2.

When emus were dehydrated and exposed to Ta=45°C, the Tb in all of the emus rose before they began panting. The degree to which Tb rose varied among animals. Data from one bird is shown in Figure 5.3A-D. There was a tendency for the initiation of panting to occur after a greater change in Tb (from its value at the start of the 45°C exposure) in the emus which exhibited the greatest change in plasma osmolality but this was not significant (n=5, r2=0.4, p=0.2).

Means for all variables measured in the control and dehydrated emus are given in Table 5.2. Body temperature showed a significant increase at Ta=25°C when emus were dehydrated, as did the calculated levels of CEWL. The observed increase in Tb when dehydrated and exposed to T a=25°C may be an artifact of experimental practice (see discussion). At T a=45°C, apart from the increase in Tb, there was a significant reduction in EWL and dry conductance in the dehydrated emus.

Expired air temperature (Tex) measured in the three dehydrated emus averaged 0.6°C below Tb (Tb =38.9±0.17°C, Tex=38.3±0.09°C). A T ex 0.6°C below Tb was used to calculate respiratory evaporative water loss. 136

- 50 p- 45 c: -Q) ~ 40 ..a ~ E 05 35 <( a.

EQ) 30 I- 25 39 -() 0 -Q) 38.5 I... >.. ::::J "'0 0 ~ I... 38 co Q) a. E Q) 37.5 I- 37 50

40 panting ~en- ~ 1a 30 W3:- 20

10 80 E .en 60 o-00 ..a~ $;:: 40 m- ::E 20 0 20 40 60 80 100 120 Time (minutes)

Figure 5.3 Measurements of body temperature (B), evaporative water loss (C), and metabolic rate (D), of a dehydrated emu as ambient temperature (A) was increased to 45oc 137

Table 5.2 Comparison of variables of hydrated and dehydrated emus exposed to ambient temperatures of 2soc and 45oc.

VO;. EWL fR VT E02 mL02/kg·m gH20/h b/min mL %

25°C Control 2.36 ± 0.09 15.7±1.3 4.1 ± 0.14 917 ± 38 17.4±1.1 Dehyd 2.81 ± 0.21 22.7 ± 2.2 4.2 ± 0.37 779 ± 47 22.9 ± 2.1 P= 0.13 0.07 0.92 0.09 0.08 45°C Control 3.00 ± 0.19 98.1 ± 4.9 41.3 ± 1.0 805 ± 36 2.6 ± 0.2 Dehyd 3.10 ± 0.1 78.2 ± 6.4 37.3 ± 1.9 761 ± 65 2.9 ± 0.3 P= 0.90 0.048 0.08 0.62 0.44

Tb Cond CEWL REWL% oc H20/m2·h Total EWL 25°C Control 37.6 ± 0.2 1.9±0.2 10.4 ± 1.5 39.7 ± 3.7 Dehyd 38.2 ± 0.1 1.6±0.3 20.2 ± 2.1 22.9 ± 2.4 P= 0.002 0.46 0.03 0.013 45°C Control 38.5 ± 0.2 3.5 ± 0.2 25.1 ± 5.1 76.7 ± 4.6 Dehyd 38.8 ± 0.2 2.0 ± 0.4 11.7 ± 7.4 89.3 ± 8.1 P= 0.03 0.05 0.09 0.09 138

5.4 Discussion

5.4.1 The Effect on Body Mass and Plasma Osmolality The water restrictions imposed on the emus during this study caused a loss of body mass and an increase in plasma osmolality in all emus. However, the loss of body mass and the increase in plasma osmolality showed wide variation both within this study, and between this and past studies. The emus in the study by Skadhauge (1974) were dehydrated for 3 days, lost 11 °/o of original body mass, and experienced a 13% increase in plasma osmolality. In the study by Dawson et a/. (1991 ), 10 days of dehydration resulted in a 19.3% loss of body mass and a· 10.4% increase in plasma osmolality. During the present study, after an average 18.6 days dehydration, the emus lost an average 11% body mass and plasma osmolality rose by an average 8%. A major influence on the rate of water loss would have been the environmental conditions. The previous two studies were done in summer which could have led to increased rates of EWL for thermoregulation.

In the present study the three emus losing the most mass, and experiencing the largest increase in plasma osmolality, were males. It is possible that hormone levels, influenced by the proximity of the breeding season, may have led to increased activity of the males in their pens, compared to the females, during this study. There was no difference between thermoneutral resting metabolism measured at 25°C during these experiments and measurements at other times of the year (control vs winter [Chapter 2];t-test, t=1.144, df=6, p=0.296). 139

5.4.2 The Effect on Response to Heat Exposure The inhibition of the normal thermoregulatory pattern of EWL when dehydrated emus were heat exposed is a similar phenomenon to that reported for other species of birds (Crawford and Schmidt­ Nielsen 1967, Arad 1982, 1983, Arad et a/. 1985, 1987, ltsaki­ Giucklich and Arad 1992) and mammals (Schmidt-Nielsen et a/. 1957, Taylor 1970a, 1970b, Doris and Baker 1981 ). The suppression of EWL responses has been reported for panting and sweating species and involves the threshold for EWL being shifted to a higher Tb. There is a high degree of convergence of thermosensitive and osmosensitive neurons in the pre-optic anterior hypothalamus (Silva and Boulant 1984, Nakashima et a/. 1985, Hori et a/. 1988). The inhibition of EWL during dehydration is thought to result from increased thermoregulatory thresholds set up in these interneurons due to increased osmoregulatory input. Increased osmoregulatory input results from increased plasma and cerebral fluid osmolality. It has been shown that dehydration or intravascular, intracarotid, or intracerebral ventricular infusions of hypertonic saline or sucrose leads to reductions in EWL in heat exposed mammals (Turlejska-Stelmasiak 1974, Baker and Doris 1982a, 1982b, Baker and Dawson 1985, Turlejska and Baker 1986). Additionally, cerebral ventricular infusions of water into dehydrated, heat exposed cats led to an increase in EWL to levels observed in hydrated cats (Doris 1983).

The bird hypothalamus is not a major organ of thermosensitivity although temperature changes in the hypothalamus do affect thermoregulation, presumably due to the effects of temperature on neuronal transmission (Simon et a/. 1986). The scenario of local 140

effects of temperature and osmolality observed in mammals could be similar in birds. Simon and Nolte (1990) have demonstrated the interdependence of the thermo- and osmoregulatory systems in the duck. Cooling the duck hypothalamus results in a reduced osmoregulatory response to increases in plasma osmolality.

This inhibition of EWL in the emu can be interpreted as an adaptive response since non-lethal increases in Tb store a heat load which would usually be evaporated, and the resultant higher Tb is advantageous because the thermal gradient for dry heat gain is made more favourable, which saves water. Once Tb had risen and panting was initiated in the dehydrated emus they were able to maintain Tb for the duration of the 2 hour exposure to a Ta of 45°C. During this time Tb maintenance was achieved with a significantly lower level of EWL than that seen during the heat exposure of hydrated emus. The increased Tb of the dehydrated emus at Ta=45°C would give them a lowered dry heat gain. However the increased Tb when dehydrated can only account for 2.4 watts of the observed 13.3 watts reduction in EHL. The major contributing factor to the reduced EWL of the dehydrated emus was the reduction of dry conductance to 57% of the hydrated value (Table 5.2) resulting in a lower environmental heat load.

5.4.3 Protecting the Brain An interesting response to dehydration is observed in the eland, an animal that usually both pants and sweats when heat exposed. When dehydrated and heat exposed, the eland reduces sweating EWL but increases ventilation frequency (Finch and Robertshaw 1979). This increases evaporation from surfaces in contact with venous blood 141

which drains into the carotid rete, where counter current cooling of blood destined for the brain can occur. The eland can allow Tb to rise, but protects the thermally sensitive brain. Birds usually maintain the brain at a temperature about 1°C cooler than the body (see Section 1.11) but when dehydrated and heat exposed the brain­ body temperature difference is reduced (Arad 1983, Kleinhaus et a/. 1985). Presumably this is due to the reduction in ventilation frequency observed in many dehydrated, heat exposed birds (Crawford and Schmidt-Nielsen 1967, Arad 1983, Arad et a/. 1984, Kleinhaus et a/. 1985, ltsaki-Giucklich and Arad 1992). This would result in less cooling of the venous blood draining into the opthalmic rete, which in birds is where counter current cooling of blood destined for the brain occurs (Bernstein et a/. 1979a, 1979b, Pinshow et a/. 1982, Pelton en et a/. 1989).

The emu has an opthalmic rete (Smith and Dawson, unpublished results) and so it is likely that the brain is maintained cooler than the body. There was a tendency for ventilation frequency to be lower in the dehydrated state (Table 5.2) but REWL was the same in both states, due mainly to the increased Tb and expired air temperature when dehydrated. It is possible that in the emu, like other bird species, the degree to which hyperthermia can develop during heat exposure is facilitated by preferential cooling of the brain.

Studies on other species have suggested that changes in the plasma Na+/Ca+ ratio can cause a shift in the 'set point' for temperature regulation (see Senay 1979, Arad and Skadhauge 1986, for review). The emu's plasma Na+/Ca+ ratio was not affected by dehydration 142

(Table 5.1) and the night time Tb of one dehydrated male emu was similar to hydrated levels (Figure 5.2). The large increase in Tb of this emu during the day compared to hydrated male emus (Figure 2.2) suggests that Tb was not defended as strongly when the dehydrated animal became active. If Tb was elevated before the dehydrated emus were placed in the metabolism chamber at 25°C, T b not being as rigourously defended in the dehydrated state could have led to the observed higher Tbs at T a=25°C.

5.4.4 Body Temperature Control The pattern of thermoregulation when T a was raised to 45°C suggests that there was inhibition ·of heat loss responses until some critical threshold was exceeded. The reference for this critical threshold is more complex than simply an increased deep body temperature (and consequently probably neither hypothalamic nor spinal cord temperature) because the 'critical' Tb in one emu was apparently raised by lowering the T a· If the observations on this one emu are indicative of emus in general, then it seems there is peripheral input into the initiation of active heat dissipation when dehydrated. This is not surprising in view of the fact that peripheral temperature is probably the major impetus for active heat dissipation in normally hydrated heat exposed emus (Chapter 3). 143

Chapter 6

The Heat Load from Radiation

6.1 Introduction

Electromagnetic radiation from the sun reaches the earth's atmosphere at an average rate of 1370 wattsfm2 (Hickey et a/. 1982). Upon entering the atmosphere, part of this radiation is reflected back to space, part is transmitted or scattered forward towards the earth, and part is absorbed by gases or dust in the atmosphere. Variations in the composition of the atmosphere above the earth result in variations in the amount of radiation incident at ground level. In places where the air is clean and dry, ground radiation levels can exceed 1 000 wattsfm2 (Walsberg 1992). For an animal the size of the emu, which presents an area of about 0.4 m2 perpendicular to the sun at midday, this is a potential heat load of over ten times the resting metabolic rate. The emu's dark colour means that it absorbs most of that incoming radiation, however, the emu actively forages in the open on clear, hot days in the Australian arid zone (Dawson et a/. 1984).

6.1.1 Animal Colour The perception of animal colour results from the reflection from an animal's surface of radiation in the visible part of the spectrum (wavelengths from 380 to 760 nm). Animal coloration is thought to be influenced by a number of selective pressures, including: 144

i) the need for conspicuousness, either for intra-specific social interactions, or aposematicism, the advertisement of harmful or distasteful qualities ii) the need for crypsis, to avoid predation (Cioudsley-Thompson 1979) iii) reduced abrasiveness of feathers containing melanic pigments (Burtt 1981) iv) thermal influences. Since different colours result from the differential reflection of visible radiation, and approximately 40% of sunlight is in the visible region (Gates 1966), different coloured animals will absorb different amounts of solar radiation. The need to increase· or lower heat loads from radiation is thus a potential selective force on animal coloration.

6.1.2 The Heat Load from Radiation That solar radiation can supplement endogenous heat production has been demonstrated in experiments which showed that, at a low ambient temperature, the energy consumption (measured either directly as food consumption or indirectly as oxygen consumption) of irradiated animals was less than that of non-irradiated animals (Hamilton and Heppner 1967a, Morton 1967, Lustick 1969, 1971, Ohm art and Lasiewski 1971, Neal and Lustick 1974, de Jong 1976, Wunder 1979, Lustick et a/. 1980). Generally, these experiments confirmed the intuitive feeling that a dark coloured animal absorbs more radiation and experiences a higher heat load from radiation than a lighter coloured animal (Hamilton and Heppner 1967a, Lustick 1969, 1971, Lustick et a/. 1980). 145

If these results were indicative of a general rule, then dark coloured desert inhabitants and light coloured polar inhabitants would appear to be maladapted in thermal terms. This has led some authors to dismiss the significance of thermal selection, and conclude that other selective pressures are more relevant (e.g. Cowles 1967, Cloudsley-Thompson 1979), while others urged that thermal influences should not be considered in isolation from other possible selective pressures (Hamilton and Heppner 1967b, Wunder 1979). Of course, depending on the behaviour of a species, the heat load from solar radiation will vary in importance to an animal. For example the desert dwelling hill kangaroo, or euro, is dark in colour. However, it lies up in caves during summer days so it is unlikely that the thermal effects would have been important in the selection of its colour (Dawson and Brown 1970). A reminder that these factors should be viewed in an overall sense is provided by studies on desert goats. The dark colour of desert goats appears intuitively maladaptive and in fact black goats need to evaporate more water on hot, sunny days than do white goats (Finch et a/. 1980). However, it can be argued that desert goats are black because the energy savings associated with increased radiation absorption in winter outweigh the requirement for increased water loss in summer (Dmi'el et a/. 1980). Whether this is in fact the case or other factors preside in the selection of black goats, it is a reminder that all the advantages need to be weighed against all the disadvantages of a particular colour.

Evidence has been found that at least in some animals the heat load from solar radiation is not simply a by-product of a coat colour selected for other reasons, but that thermal effects can provide 146

selection pressure. Summer and winter rock squirrels ( Spermophi/us variegatus) differ in their heat loads from radiation by about 20% (lower in summer) but coat reflectivity, and hence coat colour, is identical during the two seasons (Walsberg 1988b, Walsberg and Schmidt 1989). This suggests that the coat is seasonally altered to increase solar heat gain in winter, but reduce heat gain in summer, independently of the animals colour. Two other species of ground squirrel ( S. latera/is and S. saturatus) differ markedly in coat colour (reflectivities; S. latera/is 0.29, S. saturatis 0.19) but have the same heat loads from radiation (Walsberg 1990).

A theoretical analysis of the effect of solar radiation on the thermal balance of animals by Kovarik (1964, 1973) indicated that the story was not as simple as black and white. The situation is complicated by the penetration of radiation into the fur or feathers (rather than all absorption occurring at the coat surface as earlier workers had assumed, e.g. Burton and Edholm 1955, Priestley 1957, Joyce et a/. 1966). The resultant change in thermal gradients due to penetration meant that the greatest heat load at skin level was experienced by a coat of intermediate colour. Studies on the effect of penetration on the heat load under coats of different colours (Hutchinson and Brown 1969) confirmed the theoretical predictions of Kovarik. They also showed that coat density was an important factor: the denser a coat the more likely penetrance was to be low.

Theoretical modelling of the fate of radiation incident on an animal's coat revealed that physical factors such as wind speed can influence the heat load resulting from solar radiation (Gena and 147

Monteith 1975; see Appendix 2). Walsberg et a/. (1978) used plumage samples from black and white pigeons to confirm empirically the Gena and Monteith predictions. While at low wind speeds the black plumages experienced a higher heat load than the white ones, as wind speed increased the two converged, and at wind speeds greater than 3 m/s the white plumages experienced a higher heat load than the black. This phenomenon results from a higher proportion of forward scattering in the white plumage, which results in the radiation that isn't reflected being absorbed deeper in the coat. The resultant heat is then prevented from flowing to the environment by the insulation of the coat. Conversely, radiation is ·absorbed n·ear the surface of the black plumage and, with increasing wind speed, more heat is convected away.

In this context the dark coloration of the emu may not necessarily be a disadvantage in terms of solar heat gain on hot summer days. The experiments reported in this chapter were undertaken to quantify the effect of radiation on thermoregulation in the emu.

6.2 Materials and Methods

6.2.1 Sample Collection Pelts from eight emus (four summer, four winter) were collected at the University of New South Wales Arid Zone Research Station at Fowlers Gap in North Western New South Wales, in July-August 1991 (winter) and February-March 1992 (summer). Skins were salted, frozen and transported to the laboratory in Sydney. Skins were thawed and a section, approximately 35 x 35 em, was cut 148

from the mid-dorsal region. These sections were stripped of subcutaneous fat and tanned in a salt/alum solution (Dimpel 1971) with pH adjusted to 4 during soaking and returned to 7 prior to removal (E. Gardner, Gardner Industries, Botany, NSW. pers.comm.). During drying each section was pegged so that dimensions of the dried skin were the same before and after tanning.

6.2.2 Measurement of Plumage Spectral Reflectivity For measurement of reflectances an Optronics model 746-IRD spectroradiometer (Optronics Inc, Orlando, FL, USA ) was used. An Optronics Model 740-20 lamp housing was attached to the monochromator entrance port and an Optronics Model 740-70 reflectance attachment was attached to the exit port. The measurements were made with reference to a compressed barium sulphate disc using standard published values for its reflectance. The reflectance attachment included an integrating sphere coated with barium sulphate paint. Measurements were made with a 10 nm half band width between 350 and 750 nm. For measurements between 800 and 21 00 nm the same spectroradiometer was used but measurements were made with a 50 nm half band width.

Total solar reflectances of the plumage samples were calculated using the relative spectral power of sunlight given by the American Society for Testing & Materials ASTM E424-71. Reflectances of the plumage samples under the Arri Daylight lamp used in these experiments were calculated using the relative spectral distribution of the lamp measured by the spectrophotometric facilities of the Australian Broadcasting Corporation at Gore Hill, New South Wales. 149

6.2.3 Measurement of Thermal Conductance Prepared samples were mounted on the upper surface of a heat flux transducer (HFT) I temperature controlled plate apparatus. The water filled plate (25 em diameter x 6 em deep) was maintained at approximately 39°C via connection to a circulating water bath (Torithermo MD1, Torika Co. Japan). Three 2x3 em HFTs (Thermonetics Corporation, USA, model HA 13-18-1 OP) were embedded in the plates upper surface. Voltage output from the three HFT's was logged on a personal computer via a Datataker analog/digital converter (Data Electronics Australia P/L, model 1OOF). The plate/transducer apparatus was calibrated using two pieces of insulating polystyrene of· known thermal conductance (Boral Industries, Sydney).

Temperature of the skin surface was measured by two 0.07 em diameter copper/constantan thermocouples fed through oblique holes from beneath the skin. These and similar thermocouples which measured plate and air temperatures were referenced to a Datataker Isothermal Block (Data Electronics Australia P/L) and logged on a personal computer via the AID converter mentioned previously. Plumage surface temperature was measured with two thermocouples attached to flexible wires, which were positioned so that the thermocouple tips rested on the plumage surface. All thermocouples were calibrated against a certified mercury in glass thermometer.

Conductance of each sample was measured as a function of wind speed inside a glass wind tunnel (80 x 45 x 8.5 em; Figure 6.1 ). Wind speed was varied by adjusting the power supplied to a fan 150

ARRI Dey light -

b ec

Figure 6.1 Diagram of the wind tunnel I radiation source experimental apparatus (a=sheet of glass painted except for a hole directly above the plumage sample, b=wind direction, C=hot wire anemometer, d=plumage sample, e=heat flow transducer, f=water filled hot plate connected to circulating water bath)

forcing air through the tunnel and was measured 2 em above the sample with a Datametrics 81 OL thermoanemometer calibrated by the method of Walsberg (1988c). Turbulence was minimised by placing matching flutes at the tunnel entrance and exit. Variation in wind speed across the tunnel was less than 5%. Each sample was measured at six wind speeds (1, 2, 4, 6, 8, and 10 m/s). Air temperature was controlled to ±1 °C by placing the wind tunnel inside a temperature controlled room. Air temperature cycled every seven minutes as heaters cut in and out in the room . Hence, after i 51

stabilization at any wind speed, heat flow was taken as the average of the three transducers for seven minutes. Similarly, all temperatures were averaged for the same seven minutes. While plate (T p) and air (T a) temperature varied between experiments, both were measured for each experiment and used in subsequent calculations. Thermal conductance was calculated as: C (Wim2·°C) = b. I (Ts-T a) where O=heat flow through the sample and Ts=skin temperature. The contribution of the air boundary layer to total insulation was obtained from calculations of plumage conductance (Cp = Q I (Ts-T e) and air boundary layer conductance (Ca = Q I (Te- T a), where T e was the plumage surface temperature.

6.2.4 Measurement of Heat Load from Radiation Following measurements of conductance, the experiments were repeated with 590 Wlm2 of short wave radiation incident on the sample. Radiation was supplied by an ARRI spotlight (ARRI Daylight 575W) containing a 575 watt metal halide lamp (ILC Technology, DMI575). The spectral distribution of these lamps contains some spikes not present in daylight, however such lamps provide the best available source of simulated solar radiation (Beeson i 978). The relative spectral distribution of the lamp is shown compared to sunlight in Figure 6.2. To minimize heating of the wind tunnel surrounding the sample and transmission of infra-red radiation to the sample from the hot lamp body, a piece of glass, painted black except for a hole allowing light to penetrate in the area of the sample, was placed between the lamp and the wind tunnel. The lamp and second sheet of glass were cooled by forced convection. Radiation at the level of the top of the coat was measured with a CSI RO SRI4 radiometer after the beam passed through two sheets 152

1

r:: 0 ".+=: 0.8 ..c:::J ·c...... ,o en 0 ca 0.6 ...... '- (.) 0 Q) 0.. CJ) 0.4 Q) 0 0 ·ca> Q) 0.2 a: --sunlight · · · · · Arri Daylight 0 300 400 500 600 700 800 900 1000 1100 Wavelength (nm)

Figure 6.2 Relative spectral distributions of sunlight and Arri Daylight lamp of glass the same as in the wind tunnel set-up. Radiation was maintained at 590 Wfm2.

The heat load from radiation (HLR) was calculated as: HLR = (heat flow without radiation) - (heat flow with radiation) and expressed as a percent of radiation incident on the coat surface. It was difficult to match exactly Tp and T a conditions of the radiation measurements with those during no radiation measurements. A value for conductance of the entire sample (skin, feathers and air boundary layer) during the no radiation measurements, calculated as: Ctot = Q I (T p-Ta), was used to adjust heat flow without radiation to the exact conditions of T p and T a during the radiation experiment. 153

6.3 Results

6.3.1 Reflectivity Spectral reflectances of the summer and winter plumage samples are shown in Figure 6.3. Total solar reflectances were: summer 0.17±0.02 and winter 0.23±0.01. This difference in reflectances was significant (unpaired t=2.62, df=6, P=0.03). Total reflectances under the spectrum emitted by the Arri Daylight lamp in the wind tunnel experiments were: summer 0.09 ± 0.02 and winter 0.14 ± 0.01. This difference was also significant (unpaired t=2.9, df=6, P=0.027). The reflectivity was lower under the spectrum emitted by the Arri Daylight than in sunlight by 45% for the summer pelts (paired t=16.36, df=3, P<0.001) and 37% for the winter pelts (paired t=26.8, df=3, P<0.001 ). The sex of the emus from which plumage samples were obtained and the solar reflectivity of the plumage samples are given in Table 6.1. Comparison of the two

0.7

0.6 0.5 ·:;:c :;::::; (.) 0.4 Q) ;:;:= .. Q) 0.3 ~ -· 0: . 0.2 I ' --Winter 0.1 -----Summer 0 300 700 1100 1500 1900 2300 Wavelength (nm)

Figure 6.3 Spectral reflectance of winter and summer emu plumage samples 154

Table 6.1 Sex of emus from which plumage samples were obtained, and reflectivity of those samples

Pelt # Sex Reflectivity Winter 1 cJ 0.231 2 Q 0.197 3 cJ 0.245 4 Q 0.238

Summer 1 cJ 0.132 2 cJ 0.221 3 Q 0.136 4 Q 0.178 females with the two males in winter showed there was no significant difference between the total reflectivity of the sexes (t=0.97, df=3, P=0.43), though this may be a reflection of the small sample size.

6.3.2 Thermal Conductance Thermal conductances of the plumage samples at various wind speeds are shown in Figure 6.4. There was no significant difference between the conductances of summer and winter samples at any wind speed. Wind speed had no significant effect on conductance up to 4 m/s. At wind speeds greater than 4 m/s conductance increased significantly for each increase in wind speed. At wind speeds above 2 m/s the plumage samples were depressed, so the surface temperature thermocouples were measuring temperature above the plumage surface. Thus at wind speeds above 2 m/s the contribution of the air boundary layer to total insulation was not calculated. At 1 m/s wind speed, the air boundary layer contributed 8.8 ± 0.5 155

9 d

c (]) (.) -() 6 c 0 ...... rn (.) C\1 b :::::s E H "'0 c 0 -$: 3 () - a h a a !x x Summer F r • Winter 0 0 2 4 6 8 10 12 Wind Speed (metres I second)

Figure 6.4 Thermal conductance of plumage samples from emus in winter and summer (different letters denote significant differences between wind speeds; SNK P

(summer) and 8.8 ± 0.4 (winter) percent of total insulation. At 2 m/s wind speed the air boundary layer contributed 5.6 ± 1.1 (summer) and 7.1 ± 0.3 (winter) percent of total insulation.

6.3.3 Heat Load from Radiation The proportion of radiation incident on the plumage surface which acted as a heat load at skin level at each wind speed is shown in Figure 6.5. There was no significant difference between the heat load under summer and winter samples at any wind speed. Heat load from radiation decreased significantly as wind speed increased up to 6 m/s. At wind speeds faster than 6 m/s, the heat loads at skin level were very low, averaging 1.35 to 0.65 percent of incident radiation and were not significantly different. 156

Figure 6.5 Heat load from radiation at skin level as a percent of incident radiation on winter and summer plumage samples from emus (different letters denote significant differences between wind speeds; SNK P

Plumage conductance and plumage reflectivity were used as predictors of HLR by multiple regression analysis. Conductance accounted for more of the variation in HLR than did reflectivity at all wind speeds; however the regression proved significant only at 1 m/s (P<0.01 ). At 1 m/s wind speed the partial correlation coefficients of conductance and reflectivity were 0.905 and -0.77 respectively. The partial correlation coefficient represents the unique contribution of the respective predictor to the prediction of heat load when the other predictor variable is held constant (Sakal and Rohlf 1981 ). 157

6.4 Discussion

6.4.1 Thermal Conductance That the conductance of the plumage samples was not different between summer and winter is in accord with the observation that whole body conductance showed no change between summer and winter (Chapter 3). Many birds and mammals show a seasonal change in conductance, which is interpreted as an adaptive response to changes in the ambient temperature regime (BrOck 1986). It is reasoned that plumage with a low conductance will be advantageous in winter, whereas the maintenance of low conductance in summer ·will be a disadvantage. The latter may not be the case if ambient temperature regularly exceeds body temperature, when a low conductance will actually reduce the environmental heat load. Additionally a good layer of insulation helps to reduce the heat load of animals exposed to sunlight. A shorn camel needs to evaporate 50% more water than a furred camel on hot sunny days (Schmidt-Nielsen et a/. 1957) and shorn sheep standing in the sun at a high T a exhibit respiratory rates twice those of unshorn sheep (Parer 1963). So, if significant radiant heat loads are encountered a low conductance in summer may also be advantageous.

6.4.2 The Theory Relating Coats to the Heat Load from Radiation The heat load from radiation experienced by the emu at the skin is lower than might be expected given the dark colour of the emu plumage (Figure 6.3). The factors contributing to the low heat load from radiation are the insulation provided by the plumage, and the 158

plumage colour and plumage structure. Importantly for the emu, the plumage colour and structure will result in little penetration of radiation into the plumage. The phenomenon of radiation entering a constant density medium has been treated many times but the equations derived by Kubelka and Munk in 1931 are the most widely used (van de Hulst 1980). The Kubelka~Munk equations were applied theoretically to the special case of radiation incident normally on an animal coat by Cena and Monteith (1975) (See Appendix 2).

The validity of such a model has been verified empirically (Walsberg et al. 1978, Walsberg 1988b, 1988c, 1990, Walsberg and Schmidt 1989). Differential penetration allows a reinterpretation of the seemingly maladaptive dark desert and light polar animal coats. All other things being equal radiation will penetrate further into a light coloured coat than into a dark coat and the coat insulation between this absorbed radiation and the environment will differ. Different levels of absorption means that not only is the total coat insulation important in determining heat load from radiation, but that convection affects the heat load from radiation differently for coats of different colours. This is a result of the influence of convection on air boundary layer insulation. Coats with a high amount of radiant penetration have the heat produced by radiation absorption protected from loss to the environment by the coat and air boundary layer insulation.

6.4.3 What it Means to an Animal If considered in terms of maximization of resource utilization by animals, then polar animals should be attempting to gain heat from radiation when T a is low, while desert animals should reduce water 159

losses if they can minimize heat gain from radiation when Ta is high. In this respect, the white polar bear is thermally well adapted. The hairs making up the coat of the polar bear are hollow and transparent. These 'optical fibres' act as light guides to the dark skin. The result is that of the radiation which is not reflected, most is transmitted forward and absorbed by the dark skin (0ritsland and Ronald 1978, Grojean eta/. 1980).

At the opposite end of the 'optical' scale are coats with low penetrance. All other things being equal (e.g. total coat insulation) this type of coat would be advantageous to desert inhabitants like the emu. The application of the Ceria and Monteith (1975) model requires that p (the probability of radiation striking a feather element within a given layer) and a (the fraction of intercepted radiation that is absorbed within a given layer; see Appendix 2) be constant, or vary in a predictable manner, with coat depth. The structure and colour of the emu plumage means that p.a will vary considerably with depth. The feathers are longer than coat depth and the feathers overlay each other at the outer layer of the coat. This results in the feathers being more dense at the coat's outer layer and p being greater in this region. Colour also varies along the length of each feather, it being darker at the tips. This means that a is greater in the black tips. The combination of these two factors indicates that the absorption of radiation is greater at the coat's outer layer.

6.4.4 A Simplified Model A simple model of the heat gain from radiation can be developed if all radiation is assumed to penetrate to a single layer, z, such that 160

the heat load resulting from this average penetration equals the heat load experienced with nonlocalized absorption (Hutchinson and Brown 1969). In this case the heat load from radiation (HLR) is equal to:

HLR = R-A·(rz+re) I (rc+re) (Equation 6.1) where R=the strength of radiation incident on the coat, A=the coats absorptivity, which equals 1-reflectivity, rc=the insulation of the coat, rz=the insulation of the coat between the point of absorption and the coats outer, and re=the insulation of the air boundary layer. If HLR is expressed as a percent of incident radiation then:

HLRIR = A·(rz+re) I (rc+re) (Equation 6.2)

For the special condition where all radiation is absorbed on the coats surface, rz=O and:

HLR = R·A·(re) I (rc+re) (Equation 6.3)

Equation 6.3 is the formula used by early workers who assumed radiation was absorbed on the coat's surface (e.g. Burton and Edholm 1955, Priestley 1957, Joyce et a/. 1966). Assuming rz=O led to underestimations of HLR of 300% in cattle (Hutchinson and Brown 1969) and 600% in pigeons (Walsberg et a/. 1978). Assuming that rz=O for the emu leads to HLR values at 1 mls wind speed of 0.080 (summer) and 0.075 (winter) which are less than the observed values of 0.090 (summer) and 0.081 (winter). This means that there was some penetration of radiation into the plumage. 161

Solving equation 6.2 for rz leads to an estimate of the degree of radiation penetration in the emu. The calculated values of rz are 1.21% of rc in summer and 1.11% of rc in winter. The average plumage depths were 4.50±0.37 em in summer and 4.45±0.29 em in winter. From these results the average penetration depths are calculated to be 0.05 em in both seasons, if rc is assumed to be constant with coat depth. It was suggested earlier that this assumption is not valid, and so the actual average penetration depth would be less than this calculated value. The low value for average penetration depth is consistent with the prediction of low penetration from considerations of the optical and structural characteristics of the plumage.

6.4.5 The Effect of Convection When wind speed increased above 1 m/s re will be decreased and accordingly in the emu HLR fell, being less than 2°/o of incoming radiation at wind speeds greater than 4 m/s. There was a tendency for the darker summer coats to have higher HLR at low wind speeds and lower HLR at the higher wind speeds but this was not significant. The emu's range extends into the alpine snow country, where a higher HLR would be an advantage, especially to males incubating nests in winter. It would be interesting to compare the HLR of emus from those areas to that of these desert emus.

In general, the structure of birds' plumage is similar to that of the emus, with feathers overlaying each other at the outside of the coat. This means that it is likely that p will be high at the outer layer of the plumage for birds. The use of small birds and minimal convective conditions in the early experiments into heat load from 162

radiation (Hamilton and Heppner 1967a, Lustick 1969, 1971, Lustick et a/. 1980) may have contributed to the conclusion that dark colour meant a greater HLR.

6.4.6 What Does It All Mean for an Emu? On an average summer day at the University of New South Wales Fowlers Gap Research Station in arid Australia, the sun is up for 14-15 hours. At dawn and dusk, radiation levels are low and during the middle 12 hours of the day the average radiation load at ground level is 570 wattsfm2 (unpublished data). A 40 kg emu presents a cross sectional area of approximately 0.4 m2 perpendicular to solar radiation at midday. This area will' be slightly different when the sun is lower in the sky, but for the purpose of these calculations it will be assumed 0.4 m2 is intercepting radiation for twelve hours. In this time the emu will absorb 886 kJ at skin level if wind speed is low (1 m/s). At higher wind speeds the total heat absorbed will be less (Figure 6.5). When Ta is less than Tb, this heat could be transferred by the blood and lost to the environment from other surfaces of the body. Taking a worst case scenario, assuming Ta is greater than Tb and wind speed is low all day, the emu will need to evaporate 886 kJ to maintain Tb. in addition to that required to evaporate the heat load by convection and conduction from the warmer air. Accomplishing this would require the evaporation of 365 grams of water.

In the above calculations it was assumed that the HLR on the head and neck was equal to that on the rest of the body, which is probably not the case. The head is covered by dark feathers but these will provide less insulation than the feathers on the back. 163

The top of the neck is bare, but light in colour which will mean much of the incident radiation is reflected. The cross sectional area of the head and neck is approximately 0.03 m2 and, even assuming HLR on the head and neck is as high as 40% of incoming radiation, the water needed to evaporate the entire heat load from radiation under the worst case scenario is 459 grams. This is less than 2o/o of the body water pool of a 40 kg emu (assuming TBW=60%·mass, Dawson eta/. 1984). Under field conditions the total radiant heat load will be less than this. Ambient temperature is never higher than 38°C all day and natural convection in addition to the convection provided by movement during foraging will reduce the heat load. Dawson et a/. (1984 )' measured no difference in the water turnover of emus at Fowlers Gap in summer compared to winter, which suggests a very low need for water for thermoregulation in summer. However, ambient conditions were mild during their summer measurements (T.J. Dawson pers. comm.).

Under the simplified model (Equation 6.2) it can be seen that HLR is proportional to the ratio of insulation outside the average penetration layer to insulation beneath this layer. During the wind tunnel experiments the heat load at skin level was measured. The emus from which plumage samples were obtained in summer had a noticeable layer of subcutaneous fat on the back (Figure 6.6) which was not observed in winter (unfortunately the winter samples were collected and prepared first, so a quantitative comparison was not possible). The extra insulation provided by the layer of fat (assuming it acts as insulation, i.e. blood flow to the skin on the back is retarded) will reduce the flow of heat from the skin into the body. 164

I 2 em I

Figure 6. 6 Fat Layer Beneath Back Skin in a Summer Emu 165

The observation that emus tend to water twice a day during summer (Dawson et a/. 1984) would indicate that there is nothing special about the emu's ability to remain active on hot days during summer. The structural and optical properties of the plumage provide for good protection from a high radiant heat load and the heat load that is experienced can be dissipated by evaporating a minor amount of water which, it appears, is replaced at regular intervals. 166

Chapter 7

Conclusion

The diversity of temperature regimes in areas inhabited by the emu suggests the emu possesses generally successful thermoregulatory abilities. The emus large size limits the extent to which it can exploit thermally advantageous micro-climates and so its range implies well developed physiological thermoregulatory abilities.

In the cold the emu remains homeothermic, at least in the conditions of this research. Whole· body thermal conductance is higher than predicted for a bird of this size, but even at -S°C the heat production required to maintain body temperature (Tb) is only 2.S times the basal metabolism rate (BMR). Heat conservation in the cold is aided by the presence of vascular rete at the top of the legs (Hyrtl 1863). If the emu sits down, the cost of homeothermy is reduced to 1.S times BMR at -S°C, principally due to the removal of the naked legs as sources of heat loss.

Body temperature regulation at low ambient temperatures (T a) is aided by the emu's ability to increase the oxygen extraction rate (E02) in response to cold challenge, which results in lower ventilation and thus reduced ventilatory heat loss. Further energy economies are afforded by the ability to exhale air that has been cooled considerably below body temperature. The combination of increased E02 and reduced expired air temperature results in savings of 1S-20% of the energy expenditure required to maintain Tb at T a=-S°C. 167

At thermoneutral T as the emu's resting metabolism is lower than predicted from allometric relationships for birds. Other ratites that have been studied also have lower than predicted BM Rs. Whether a low metabolism is an adaptation by the emu or a phylogenetic attribute, it will be advantageous for life in the arid zone. Arid zones are areas of low productivity and a low BMR reduces the birds energy needs. Low BMR also means a lower endogenous heat production, and so in hot conditions the heat load to be dissipated is reduced. It is also usually associated with low water needs, and the emu has been shown to have low rates of water turnover (Dawson et a/. 1983).

The BM R and Tb of the male emu are lower than the females. The lower BMR of the males is possibly related to their lower Tb, but the very large temperature effect (01 o of greater than 25) suggests that there is some difference in the regulated system, not simply in the regulation of the system. The male's lower BMR may be the result of selection operating to reduce the energy cost of the eight week incubation which it undertakes in winter. During incubation the male does not eat and relies on body reserves for survival.

The large size of the emu reduces the surface area (relative to that of smaller birds) available for dry heat dissipation. There is some evidence that dry heat dissipation is enhanced at moderate T as by the legs acting as heat windows. At T as above 25°C T b maintenance is achieved by increases in evaporative water loss (EWL). At T a=45°C 30o/o of total EWL is cutaneous in origin. This evaporation occurs from all surfaces, including the legs. The remainder of EWL is accomplished by panting. 168

During mild heat exposure (T a=35°C) tidal volume (VT) is reduced, which could minimize parabronchial hyperventilation and so reduce C02 washout. At this Ta the emu switches regularly between this reduced VT panting and a normal breathing mode of ventilation. During severe heat stress VT is higher, which should result in parabronchial hyperventilation. It has been shown previously that respiratory alkalosis is avoided under these conditions by changes in blood perfusion of the lungs (Jones et a/. 1983). During severe heat stress only the proximal end of each parabronchus is perfused, reducing the contact time between blood and air, and reducing C02 washout.

When the emu's water intake is restricted it can reduce water efflux considerably (Dawson et a/. 1983). Water economy is facilitated by the inhibition of the normal evaporative response to heat challenge. When exposed to heat the dehydrated emu's Tb rises before panting commences. This elevated Tb is then defended with an EWL rate 20% lower than that of hydrated emus. This is achieved partly because the elevated Tb reduces the gradient for environmental heat gain, but mainly by a reduction in dry thermal conductance when dehydrated. Associated with this reduction in dry conductance is a tendency for a lower reliance on cutaneous evaporative water loss.

Possibly the most impressive aspect of emu thermoregulation is their ability to remain active on hot days in areas where ground radiation levels can exceed 1 000 wattsfm2. The structural and optical properties of emu feathers results in little penetration of 169

radiation into the plumage. The insulation provided by the plumage between the skin and the absorbed radiation retards heat flow to the skin and results in less than I 0% of an incident radiation load acting as a heat load at skin level when convection at the coat/environment barrier is minimal. With increasing convection the heat load at the skin level diminishes dramatically. At wind speeds greater than 6 m/s less than 1.5% of incoming radiation acts as a heat load at skin level. The heat load that the emu experiences in sunny conditions will be attenuated further by the presence of a layer of sub-cutaneous fat on the back in summer. A reduction in the fat layer in winter would increase the heat load from radiation, which would be advantageous in cold conditions.

The emus thermoregulatory abilities in concert with its mobility allow it to exploit areas on most of the Australian continent and to move to more favourable areas when conditions change. This ability will be especially important when the adults are augmenting body reserves in preparation for the winter breeding season. 170

References

Arad, Z. (1982). Effect of water deprivation and heat exposure on body weight loss and drinking capacity in four breeds of fowl (Gallus domesticus). Comp. Biochem. Physiol. 73A,297- 299 Arad, Z. (1983). Thermoregulation and acid-base status in the panting dehydrated fowl. J. Appl. Physiol. Respir. Environ. Exercise Physiol. 54,234-243 Arad, Z. and U. Midtgard (1984). Differences in the structure of the rete opthalmicum among three breeds of domestic fowls, Gallus gallus f. domesticus (Aves). Zoomorphology (Bert.) 104,184-187 Arad, Z., U. Midtgard and E. Skadhauge (1984). Effect of dehydration on body-to-brain temperature difference in heat-stressed fowl (Gallus domesticus). J. Comp. Physiol. B. 154,295-300 Arad, Z., S.S. Sighvatur, A. Chadwick and E. Skadhauge (1985). Osmotic and hormonal responses to heat and dehydration in the fowl. J. Comp. Physiol. B. 155,227-234 Arad, Z. and E. Skadhauge (1986). Body temperature and plasma arginine vasotocin in fowls adapted to high- and low- NaCI diets. Br. Poult. Sci. 27,11 5-1 21 Arad, Z., I. Gavrieli-Levin, U. Eylath and J. Marder (1987). Effect of dehydration on cutaneous water evaporation in heat-exposed pigeons (Columba Iivia). Physiol. Zoo/. 60,623-630 Arad, Z., U. Midtgard and M.H. Bernstein (1989). Thermoregulation in turkey vultures: vascular anatomy, arteriovenous heat exchange, and behaviour. Condor 91,505-514 171

Arad, Z. and U. Midtgard (1990). Ontogenetic development of the opthalmic rete in relation to brain cooling in chickens and pigeons. American J. Anat. 187,98-103 Aschoff, J. (1981 ). Thermal conductance in mammals and birds: its dependence on body size and circadian phase. Comp. Biochem. Physiol. 69A,611619 Aschoff, J., and H. Pohl. 1970. Der Ruheumsatz von vogeln als

funktion der tageszeit und der korpergro~e. Journal Fur Ornithologie. 111,38-47. Baker, M.A. (1982). Brain cooling in endotherms in heat and exercise. Annu. Rev. Physiol. 44,85-96 Baker, M.A. and P.A. Doris (1982aJ. Control of evaporative heat loss during changes in plasma osmolality in the cat. J. Physio/. (Lond.) 328,535-545 Baker, M.A. and P .A. Doris (1982b). Effect of dehydration on hypothalamic control of evaporation in the cat. J. Physiol. (Lond.) 322,457-468 Baker, M.A. and D.O. Dawson (1985). Inhibition of thermal panting by intracarotid infusion of hypertonic saline in dogs. Am. J. Physiol. 249,R787-R791 Banzett, R.B., C.S. Nations, N. Wang, J.J. Fredberg and J.P. Butler (1991 ). Pressure profiles show features essential to aerodynamic valving in geese. Respir. Physiol. 84,295-309 Barnas, G.M. and W. Rautenberg (1984). Respiratory responses to shivering produced by external and central cooling in the pigeon. Pfluegers Arch. Eur. J. Physiol. 401 ,228-232 Barre, H., F. Cohen-Adad, C. Duchamp and J.L. Rouanet (1989). Muscular non-shivering thermogenesis in birds. In 'Thermal Physiology'. (Ed. J.B. Mercer), Proceedings of the 172

International Symposium on Thermal Physiology, Tromso, Norway, 16-21 July 1989. (Excerpta Medica: Amsterdam). Bartholomew, G.A., R.C. Lasiewski and E.C. Crawford (1968). Patterns of panting and gular flutter in cormorants, pelicans, owls, and doves. Condor 70,31-34 Baudinette, R.V., J.P. Loveridge, K.J. Wilson and K. Schmidt-Nielsen (1976). Heat loss from feet of herring gulls at rest and during flight. Am. J. Physiol. 230,920-924 Bech, C. (1980). Body temperature, metabolic rate, and insulation in winter and summer acclimatized mute swans (Cygnus olor). J. Camp. Physiol. 136,61-66 Bech, C., K. Johansen and G.M.O. Maloiy (1979). Ventilation and expired gas composition in the flamingo (Phoenicopterus ruber) during normal respiration and panting. Physiol. Zoo/. 52,313-328 Bech, C. and K. Johansen (1980). Ventilation and gas exchange in the mute swan, Cygnus olor. Respir. Physiol. 39,285-295 Bech, C. and U. Midtgard (1981 ). Brain temperature and the rete mirabile opthalmicum in the zebra finch (Poephila guttata). J. Comp. Physiol. 145,89-93 Bech, C., K. Johansen, R. Brent and S. Nocol (1984). Ventilatory and circulatory changes during cold exposure in the pekin duck, Anas platyrhynchos. Respir. Physiol. 57,1 03-112 Bech, C., W. Rautenberg and B. May (1985). Ventilatory oxygen extraction during cold exposure in the pigeon (Columba Iivia). J. Exp. Bioi. 116,499-502 Beeson, E.J.G. (1978). The CSI lamp as a source of radiation for solar simulation. Lighting Research and Technology 10,164- 166 173

Benedict, F.G., and E.L. Fox. (1927). The gaseous metabolism of large wild birds under aviary life. Proc. Am. Phi/as. Soc. 66,511- 534. Bennett, P.M (1991 ). The evolution of activity capacity. J. Exp. Bioi. 160:1-23. Bennett, A.F. and J.A. Ruben (1979). Endothermy and activity in vertebrates. Science 206,649-654 Bennett, P.M., and P .H. Harvey (1987). Active and resting metabolism in birds: allometry, phylogeny and ecology. J. Zoo/. (Land.) 213:327-363. Berger, M., J.S. Hart and O.Z. Roy (1970). Respiration, oxygen consumption and· heart rate ·in some birds during rest and flight. Z. verg/. Physio/ogie 66,201-214 Berger, R.E., M. Meyer, W. Graf and P. Scheid (1979). Gas exchange in the parabronchial lung of birds: experiments in unidirectionally ventilated ducks. Respir. Physiol. 36,19-37 Bernstein, M.H. (1971 ). Cutaneous water loss in small birds. Condor 73,468-469 Bernstein, M.H. (197 4). Vascular responses and foot temperature in pigeons. Am. J. Physio/. 226,1350-1355 Bernstein, M.H. (1976). Ventilation and respiratory evaporation in the flying crow, Corvus ossifragus. Respir. Physiol. 26,371- 382 Bernstein, M.H. and K. Schmidt-Nielsen (1974). Ventilation and oxygen extraction in the crow. Respir. Physiol. 21,393-401 Bernstein, M.H., M.B. Curtis and D.M. Hudson (1979a). Independence of brain and body temperatures in flying american kestrels, Falco sparverius. Am. J. Physiol. 237,R58-R62 174

Bernstein, M.H., I. Sandoval, M.S. Curtis and D.M. Hudson (1979b) Brain temperature in pigeons: effects of anterior respiratory bypass. J. Camp. Physiol. 129, 11 5-11 8 Bernstein, M.H. and F.C. Samiengo (1981 ). Ventilation and acid-base status during thermal panting in pigeons (Columba Iivia). Physiol. Zoo/. 54,303-315 Blaxter, K.L., N. McGraham, F.W. Wainman and D.G. Armstrong (1959). Environmental temperature, energy metabolism and the heat regulation in sheep. II. The partition of heat losses in closely clipped sheep. J. Agric. Sci. 52,25-40 Blem, C. (1974). Geographic variation of thermal conductance in the house sparrow Passer domesticus. Camp. Biochem. Physiol. 47A,101-108 Block, B.A., J.R. Finnerty, A.F.R. Stewart and J. Kidd (1993). Evolution of endothermy in fish: mapping physiological traits on a molecular phylogeny. Science 260,210-213 Bond, T.E., C.F. Kelly and H. Heitman (1952). Heat and moisture loss from swine. Agric. Eng. 33,148-154 Bouverot, P., G. Hildwein and D. LeGoff (1974). Evaporative water loss, respiratory pattern, gas exchange and acid-base balance during thermal panting in pekin ducks exposed to moderate heat. Respir. Physiol. 21,255-269 Bowler, K. and R. Tirri (1974). The temperature characteristics of synaptic membrane ATPases from immature and adult rat brain. J. Neurochem. 23,611-613 Bozinovic, F. (1992). Scaling of basal and maximum metabolic rate in rodents and the aerobic capacity model for the evolution of endothermy. Physiol. Zoo/. 65,921-932 175

Brackenbury, J.H., P. Avery and M. Gleeson (1982). Effects of temperature on the ventilatory response to inspired C02 in unanaesthetized domestic fowl. Respir. Physiol. 49,235-250 Brand, M.D., P. Couture, P.L. Else, K.W. Withers and A.J. Hulbert (1991 ). Evolution of energy metabolism. Biochem. J. 275,81- 86 Brent, R., J.G Rasmussen, C. Bech and S. Martini (1983). Temperature dependence of ventilation and oxygen extraction in the kittiwake, Rissa trydactyla. Experientia 39,1092-1093 Brent, R., P.F. Pederson, C. Bech and K. Johansen (1984). Lung ventilation and temperature regulation in the European coot, Fulica atra. Physiol. Zoo/. 57,19-25 Bretz, W.L. and K. Schmidt-Nielsen (1971 ). Bird respiration: flow patterns in the duck lung. J. Exp. Bioi. 54,103-118 Brody, S. and R.C. Proctor (1932). Relation between basal metabolism and mature body weight in different species of mammals and birds. Missouri Ag. Exp. Stat. Res. Bull. 166,89-101 Bruck, K. (1986). Basic mechanisms in thermal long-term and short-term adaptation. J. Therm. Bioi. 11,73-77 Bucher, T.L. (1981 ). Oxygen consumption, ventilation and respiratory heat loss in a parrot, Bolborhynchus linea/a, in relation to ambient temperature. J. Camp. Physiol. 142,4 79- 488 Bucher, T.L. (1985). Ventilation and oxygen consumption in Amazona viridigenalis: a reappraisal of 'resting' respiratory parameters in birds. J. Camp. Physiol. B. 155,269-276 Bucher, T.L., and K. Morgan (1989). The effect of ambient temperature on the relationship between ventilation and 176

metabolism in a small parrot (Agapornis roseicollis). J. Camp. Physiol. B. 159,561-567 Burger, F.J. and F.A. Fuhrman (1 964). Evidence of injury by heat in mammalian tissues. Am. J. Physiol. 206,1057-1061 Burgoon, D.A., D.L. Kilgore and P .J. Motta (1987). Brain temperature in the calliope hummingbird (Stellu/a calliope): a species lacking a rete mirabile opthalmicum. J. Camp. Physioi.B. 157' 583-587 Burton, A.C. and O.G. Edholm (1955). 'Man in a cold environment.' (Edward Arnold Ltd: London.) Burtt, E.H. (1981). The adaptive ness of animal coulors. Bioscience 31,723-729 Butler, J.P., R.B. Banzett and J.J. Fredberg (1988). Inspiratory valving in avian bronchi: aerodynamic considerations. Respir. Physiol. 72,241-256 Buttemer, W.A., LB. Astheimer and T.J. Dawson (1988). Thermal and water relations of emu eggs during natural incubation. Physiol. Zoo/. 61,483-494 Buttemer, W.A., and T.J. Dawson. (1989). Body temperature, water flux and estimated energy expenditure of incubating emus. Camp. Biochem. Physiol. A. Camp. Physiol. 94,21-24. Cabanac, M. (1975). Temperature regulation. Annu. Rev. Physiol. 37,415-439 Calder, W.A. (1968). Respiratory and heart rates of birds at rest. Condor 70,358-365 Calder, W.A. and K. Schmidt-Nielsen (1966). Evaporative cooling and respiratory alkalosis in the pigeon. Proc. Nat. Acad. Sci. 55,750-756 177

Calder, W.A. and K. Schmidt-Nielsen (1967). Temperature regulation and evaporation in the pigeon and the roadrunner. Am. J. Physiol. 213,883-889 Calder, W.A. and K. Schmidt-Nielsen (1968). Panting and blood carbon dioxide in birds. Am. J. Physiol. 215,477-482 Calder, W.A. and J.R. King (1974). Thermal and caloric relations of birds. In 'Avian Biology.' (Eds D.S. Farner and J.R King) (Academic Press: New York.) Calder, W.A., and T.J. Dawson. (1978). Resting metabolic rates of ratite birds: the kiwis and the emu. Camp. Biochem. Physiol. A. Camp. Physiol. 60,479-481. Caughley, G., G.C. Grigg,· J. Caughley and G.J.E. Hill (1980). Does dingo predation control the densities of kangaroos and emus. Aust. Wild/. Res. 7,1-12 Cena, K. and J.L. Monteith (1975). Transfer processes in animal coats. I. Radiative transfer. Proc. Roy. Soc. Lond. B. 188,377- 393 Chapin, J.L. (1954). Ventilatory response of the unrestrained and

unanaesthetized hamster to C02 . Am. J. Physiol. 179,146- 148 Chappell, M.A. and T.L. Bucher (1987). Effects of temperature and altitude on ventilation and gas exchange in chukars (Aiectoris chukar). J. Comp. Physiol. B. 157,129-136 Chappell, M.A. and S.L. Souza (1988). Thermoregulation, gas exchange and ventilation in adelie penguins (Pygoscelis adeliae). J. Comp. Physiol. B. 157,783-790 Clemens, D.T. (1988). Ventilation and oxygen consumption in rosy finches and house finches at sea level and high altitude. J. Camp. Physiol. B. 158,57-66 178

Cloudsley-Thompson, J.L. (1979). Adaptive functions of the colours of desert animals. J. Arid. Environments. 2,95-104 Coulombe, N.H. (1970). Physiological and physical aspects of temperature regulation in the burrowing owl Speotyto cunicularia. Comp. Biochem. Physiol. 35,307-337 Cowles, R.B. (1967). Black pigmentation: adaptation for concealment or heat conservation? Science 158,1340-1341 Cracraft, J. (1974). Phylogeny and evolution of the ratite birds. Ibis 116,494-521 Crawford, E.C. (1962). Mechanical aspects of panting in dogs. J. Appl. Physiol. 17,249-251 Crawford, E.C. and K. Schmidt-Nielsen (1967). Temperature regulation and evaporative cooling in the ostrich. Am. J. Physiol. 212,347-353 Crawford, E.C., and R.C. Lasiewski. (1968}. Oxygen consumption and respiratory evaporation of the emu and rhea. Condor 70,333- 339. Crawford, E. C. and G. Kampe (1971 ). Resonant panting in pigeons. Comp. Biochem. Physiol. 40A,549-552 Crompton, A.W., C.R. Taylor and J.A. Jagger (1978). Evolution of homeothermy in mammals. Nature 272,333-336 Daan, S., D. Masman, A. Strijkstra, and S. Verhulst (1989). Intraspecific allometry of basal metabolic rate: relations with body size, temperature, composition, and circadian phase in the kestrel, Falco tinnunculus. J. Bioi. Rhythms 4:267-283. Daan, S., D. Masman and A. Groenewald (1990). Avian basal metabolic rates: their association with body composition 179

and energy expenditure in nature. Am. J. Physia/. 259 (Reg. Int. Camp. Physial. 28), R333-R340 Davies, S.J.J.F. (1974). The natural history of the emu in comparison with that of other ratites. Proceedings of the 16th International Ornithological Congress, Canberra, Australia. 1974. pp:1 09-120. Dawson, T.J. (1973). 'Primitive' mammals. In 'Comparative Physiology of Thermoregulation Vol. Ill.' (Ed. G.C. Whittow) (Academic Press: New York.) Dawson T.J. and K. Schmidt-Nielsen (1966). Effect of thermal conductance on water economy in the antelope jack rabbit, Lepus alieni. J. Cell. Physidl. 67,463-4 72 Dawson T.J. and G.D. Brown (1970). A comparison of the insulative and reflective properties of the fur of desert kangaroos. Camp. Biachem. Physial. 37,23-38 Dawson T.J. and A.J. Hulbert (1970). Standard metabolism, body temperature, and surface areas of Australian marsupials. Am. J. Physia/.218,1233-1238 Dawson T.J. and T.R. Grant (1980). Metabolic capabilities of monotremes and the evolution of homeothermy. In 'Comparative physiology: primitive mammals.' (Eds K. Schmidt-Nielsen, K. Balis and C.R. Taylor) (Cambridge University Press: Cambridge.) Dawson, T.J. and R.M. Herd(1983). Digestion in the emu: low energy and nitrogen requirements of this large ratite bird. Camp. Biachem. Physial. 75A,41-45 Dawson, T.J., R.M. Herd and E. Skadhauge (1983). Water turnover and body water distribution during dehydration in a large arid 180

zone bird, the emu, Dromaius novaehollandiae. J. Camp. Physiol. 153,235-240 Dawson, T.J., D. Read, E.M. Russell and R.M. Herd (1984). Seasonal variation in daily activity patterns, water relations and diet of emus. Emu 84,93-102 Dawson T.J., R.M. Herd and E. Skadhauge, (1985). Osmotic and ionic regulation during dehydration in a large bird, the emu (Dromaius novaehol/andiae): an important role for the cloaca-rectum. Quart. J. Exp. Phys. 70,423-436 Dawson T.J., S.K. Maloney and E. Skadhauge, (1991 ). The role of the kidney in electrolyte and nitrogen excretion in a large flightless bird, the emu, du'ring different osmotic regimes, including dehydration and nesting. J. Camp. Physiol. B. 161,165-171 Dawson, W.R. (1982). Evaporative losses of water by birds. Camp. Biochem. Physiol. 71 A,495-509 Dawson, W.R., and A.F. Bennett (1973). Roles of metabolic level and temperature regulation in the adjustments of western plumed pigeons (Lophophaps ferruginea) to desert conditions. Camp. Biochem. Physiol. 44A,249-266 de Jong, A.A. (1976). The influence of simulated solar radiation on the metabolic rate of white-crowned sparrows. Condor 78,174-179 Dejours, P. (1981 ). 'Principles of Comparative Respiratory Physiology, 2Ed.' (Elsevier: Amsterdam.) Dimpel, H. (1971 ). 'Tanning large skins for museum collections.' (C.S.I.R.O.: Melbourne.) 181

Dmi•el, R., A. Prevulotzky and A. Shkolnik (1980). Is a black coat in the desert a means of saving metabolic energy? Nature 283,761-762 Dmi•el, R. and D. Tel-Tzur (1985). Heat balance of two starling species (Sturnus vulgaris and Onychognathus tristram1) from temperate and desert habitats. J. Comp. Physiol. B. 155,395- 402 Doris, P.A. (1983). Osmotic regulation of evaporative water loss and body temperature by intracranial receptors in the heat stressed cat. Pfluegers Arch. Eur. J. Physiol. 398,337-340 Doris, P.A. and M.A. Baker (1981 ). Effect of dehydration on thermoregulation in cats exposed to high ambient temperatures. J. Appl. Physiol. 51,46-54 Drent, R.H. and B. Stonehouse (1971 ). Thermoregulatory responses of the Peruvian penguin, Spheniscus humboldti. Comp. Biochem. Physiol. 40A,689-71 0 Drent, R.H. and S. Daan (1980). The prudent parent: energetic adjustments in avian breeding. Ardea 68:225-252. Drorbaugh, J.E. and W.O. Fenn (1955). A barometric method for measuring ventilation in newborn infants. Pediatrics 16,81- 87 Duchamp, C., H. Barre, J.L. Rouanet, A. Lanni, F. Cohen-Adad, G. Berne and P. Brebion (1991 ). Nonshivering thermogenesis in king penguin chicks. I. Role of skeletal muscle. Am. J. Physiol. 261 ,R1438-1445 Duncker, H.R. (1972). Structure of avian lungs. Respir. Physiol. 14,44-63 Eastman, M. (1969). 'The Life of the Emu.' (Angus and Robertson: Sydney.) 182

Elgar, M.A. and P .H. Harvey (1987). Basal metabolic rates in mammals: allometry, phylogeny and ecology. Funct. Ecol. 1 ,25-36 EI-Halawani, M.E, W.O. Wilson and R.E Burger (1970). Cold acclimation and the role of catecholamines in body temperature regulation in male Leghorns. Poult. Sci. 49,621- 632 Elias, P.M., E.R. Cooper, A. Kore and B. E. Brown (1981 ). Percutaneous transport in relation to stratum corneum structure and lipid composition. J. Invest. Dermatol. 76,297-301 Elkhawad, A.O. (1992). Selective brain cooloing in desert animals: the camel (Came/us dromedarius). Comp. Biochem. Physio/. 101 A, 195-201 Else, P.L. and A.J. Hulbert (1985). An allometric comparison of the mitochondria of mammalian and reptilian tissues: the implications for the evolution of endothermy. J. Comp. Physiol. B. 156,3-11 Else, P.L. and A.J. Hulbert (1987). Evolution of mammalian endothermic metabolism: 'leaky' membranes as a source of heat. Am. J. Physiol. 253,RI-R7 Epstein, M.A.F. and R.A. Epstein (1978). A theoretical analysis of the barometric method for measurement of tidal volume. Respir. Physiol. 32, I 05-120 Epstein, R.A., M.A.F. Epstein, G.G Haddad and R.B. Mellins (1980). Practical implementation of the barometric method for measurement of tidal volume. J. Appl. Physiol. Respir. Environ. Exercise Physiol. 49,1107-1115 Fedde, M.R. (1986). Respiration. In 'Avian Physiology 4Ed.' (Ed. P.O. Sturkie) (Springer-Verlag: New York.) 183

Feldman, H.A. and T.A. McMahon (1983). The 3/4 mass exponent for energy metabolism is not a statistical artifact. Respir. Physiol. 52,149-163 Felig, P. (1979). Starvation. In 'Endocrinology.' (Eds L.J DeGroot, G.F. Cahill, W.O. Odell, L. Martini, J.T. Potts, D.H. Nelson, E. Steinberger and A.l. Winegrad) (Grune and Stratton: New York.) Finch, V.A. and D. Robertshaw (1979). Effect of dehydration on thermoregulation in eland ( Taurotragus oryx) and hartebeest (Aice/aphus buselaaphus cocki1). Am. J. Physio/. 237, R 192- R196 Finch, V.A., R. Dmi'el, R. Boxman, A: Shkolnik and C.R. Taylor (1980). Why black goats in hot deserts? Effects of coat colour on heat exchanges of wild and domestic goats. Physiol. Zoo/. 53,19-25 Folkow, L.P. and J.B. Mercer (1986). Partition of heat loss in resting and exercising winter- and summer- insulated reindeer, Am. J. Physiol. 251 ,R32-R40 Frost, P.G.H., W.R. Seigfried and P.J. Greenwood (1975). Arterio­ venous heat exchange systems in the Jackass penguin Spheniscus demersus. J. Zoo/. (Lond.) 175,231-241 Frumkin, R., B. Pinshow and Y. Weinstein (1986). Metabolic heat production and evaporative heat loss in desert phasianids: chukar and sand partridge. Physiol. Zoo/. 59,592-605 Gates, D.M. (1966). Spectral distribution of solar radiation at the earths surface. Science 151,523-529 Gorecki, V.A., and M. Nowak. (1990). Durch die jahreszeiten bedingte thermoregulationsveranderungen beim fasan (Phasianus colchicus. L). Z. Jagdwiss. 36,236-243. 184

Gould, S.J. (1978). Generalities and uniqueness in the history of life: an exploration with random models. Bioscience 28,277- 281 Grice, D., G. Caughley and J. Short (1985). Density and distribution of emus. Aust. Wild/. Res. 12,69-73 Grojean, R.E., J.A. Souza and M.G. Henry (1980). Utilization of solar radiation by polar animals: an optical model for pelts. Appl. Optics 19,339-346 Hahn, D.W., T. Ishibashi and C.W. Turner (1966). Alteration of thyroid hormone secretion rate in fowls changed from a cold to a warm environment. Poult Sci. 45,31-33 Hails, C.J. (1983). The metabolic rate of tropical birds. Condor 85,61-65 Hales, J.R.S. and M.E.D. Webster {1967). Respiratory function during thermal tachypnoea in sheep. J. Physiol. 190,241-260 Hamilton, W.J. and F. Heppner (1967a). Radiant solar energy and the function of black homeotherm pigmentation: an hypothesis. Science 155,196-197 Hamilton, W.J. and F. Heppner (1967b). Reply to Cowles, R.B. "Black pigmentation: adaptation for concealment or heat conservation?" Science 158,1341 Hammel, H.T. (1968). Regulation of internal body temperature. Annu. Rev. Physiol. 30,641-71 0 Harvey, P.H. and M.A. Elgar (1987). In defence of the comparative method. Funct. Ecol. 1,160-161 Hayes, J.P., J.R. Speakman and P.A. Racey (1992). Sampling bias in respirometry. Physiol. Zoo/. 65,604-619 185

Hayssen, V. and R.C. Lacy (1985). Basal metabolic rates in mammals: taxonomic differences in the allometry of BMR and body mass. Comp. Biochem. Physiol. 81A,741-754 Herd, R.M. and T.J. Dawson (1984). Fiber digestion in the emu, Dromaius novaehollandiae, a large bird with a simple gut and high rates of passage. Physiol. Zoo!. 57,70-84 Herreid, C.F. and B. Kessel (1967). Thermal conductance in birds and mammals. Comp. Biochem. Physiol. 21 ,405-414 Heusner, A.A. (1982). Energy metabolism and body size. I. Is the 0.75 mass exponent of Kleiber's equation a statistical artifact? Respir. Physiol. 48,1-12 Hickey, J.R., B.M. Alton, F.J. Griffin, H. Jacobowitz, P. Pellegrino, R.H. Maschhoff, E.A. Smith and T.H. Vander Har (1982). Extraterrestrial solar irradiance variability: two and one half years of measurements from Nimbus 7. Solar Energy 29,125-127 Hill, R.W. (1972). Determination of oxygen consumption by use of the paramagnetic oxygen analyzer. J. Appl. Physiol. 33,261- 263. Hinds, D.S. and W.A. Calder (1973). Temperature regulation of the pyrrhuloxia and the Arizona cardinal. Physiol. Zoo/. 46,55-71 Hinsley, S.A. (1992). Bioenergetics of a desert specialist, the double-banded sandgrouse, and the problem of stress induced by experimental conditions. Comp. Biochem. Physiol. 1 02A,433-439 Hissa, R. (1988). Controlling mechanisms in avian temperature regulation: a review. Acta. Physiol. Scand. 132, Supplementum 567 186

Hochachka, P. (1974). Regulation of heat production at the cellular level. Fed. Proc. 33,2'162-2'169 Hochachka, P.W. and G.N. Somera ('1984). 'Biochemical Adaptation.' (Princeton University Press: Princeton, New Jersey.) Hori, T., T. Nakashima, H. Koga, T. Kiyohara, and T. Inoue ('1988). Convergence of Thermal, Osmotic and Cardiovascular Signals on Preoptic and Anterior Hypothalamic Neurons in the Rat. Brain Research Bulletin 20,879-885 Hulbert, A.J. and P.L. Else ('1990). The cellular basis of endothermic metabolism: a role for 'leaky' membranes. News in Physiological Sciences 5, 25-28 Hutchinson, J.C.D. and G.D. Brown '('1969). Penetrance of cattle coats by radiation. J. Appl. Physio/. 26,454-464 Hyrtl, J. (1863). Neue wundernetze und geflechte bei vogeln und saugethieren. Denkschriften der mathem. naturw. C/. 22,'1'13-'152 lsmaii-Beigi, F. and I.S. Edelman ('1970). Mechanisms of thyroid calorigenesis: role of active sodium transport. Proc. Nat. Acad. Sci. 67, I 071-'1 078 ltsaki-Giucklich, S. and Z. Arad (1992). The effect of dehydration on brain temperature regulation in Japanese Quail ( Coturnix coturnix japonica). Comp. Biochem. Physiol. 101 A,583-588 Jacky, J.P. (1978). A plethysmograph for long-term measurements of ventilation in unrestrained animals. J. Appl. Physiol. 45,644-647 Jacky, J.P. (1980). Barometric measurement of tidal volume: effects of pattern and nasal temperature. J. Appl. Physiol. 49,319-325 187

Jessen, C., J.R.S. Hales and G.S. Molyneux (1982). Hypothalamic sensitivity in an emu, Dromiceius novae-hol/andiae. Pf/uegers Arch. Eur. J. Physiol. 393,278-280 Johansen, K. and C. Bech (1983), Heat conservation during cold exposure in birds (vasomotor and respiratory implications). Polar Res. 1,259-268 Jones, J.H. (1982a). Pulmonary blood flow distribution in panting ostriches. J. Appl. Physiol.: Respirat. Environ. Exercise Physiol. 53, 1 411 -1 41 7 Jones, J.H. (1982b) .. Hot ostriches: intraparabronchial blood flow changes during panting. Fed. Proc. 41,1096 Jones, J.H., B. Grubb and K. Schmidt-Nielsen (1983). Panting in the emu causes arterial hypoxemia. Respir. Physio/. 54,189-195 Joyce, J.P., K.L. Blaxter and C. Park (1966). The effect of natural outdoor environments on the energy requirements of sheep. Res. Vet. Sci. 7,342-359 Kahl, M.P. (1963). Thermoregulation in the wood stork, with special reference to the role of the legs. Physiol. Zoo/. 3 6, 1 41-1 51 Kaiser, T.J. and T.L. Bucher (1985). The consequences of reverse sexual size dimorphism for oxygen consumption, ventilation and water loss in relation to ambient temperature in the prairie falcon Falco mexicanus. Physiol. Zoo/. 58,748-758 Kendeigh, S.C., V.R. Dol'nik and V.M Gavrilov (1977). Avian energetics. In 'Granivorous Birds in Ecosystems.' (Eds J. Pinowski and S.C. Kendeigh) (Cambridge University Press: Cambridge.) Kilgore, D.L., M.H. Bernstein and K. Schmidt-Nielsen (1973). Brain temperature in a large bird, the rhea. Am. J. Physiol. 225,739-742 188

Kilgore, D.L. and K. Schmidt-Nielsen (1975). Heat loss from duck's feet immersed in cold water. Condor77,475-478 Kilgore, D.L., D.F. Boggs and G.F. Birchard (1979). Role of the rete mirabile opthalmicum in maintaining the body-to-brain temperature difference in pigeons. J. Camp. Physiol. 129,119-122 Kilgore, D.L., G.F. Birchard and D.F. Boggs (1981 ). Brain temperature in running quail. J. Appl. Physio/. 50,1277-1281 King, J.R. and D.S. Farner (1961 ). Energy metabolism, thermoregulation and body temperature. In 'Biology and comparative physiology of birds.' (Ed. A.J. Marshall) (Academic Press: New York.) King, A.S. and A.F. Cowie (1969). The functional anatomy of the bronchial muscle of the bird. J. Anat. 105,323-336 Kleiber, M. (1932). Body size and metabolism. Hilgardia 6,315-353 Kleiber, M. (1961 ). 'The fire of life.' (John Wiley and Sons: New York.) Kleinhaus, S., B. Pinshow, M.H. Bernstein and A.A. Degen (1985). Brain temperature in heat stressed, water deprived desert Phasianids: sand partridge (Ammoperdix heyt) and chukar (Aiectoris chukar sinaica). Physiol. Zoo/. 58,105-116 Kovarik, M. (1964). Flow of heat in an irradiated protective cover. Nature 201,1085-1087 Kovarik, M. (1973). Radiation penetration of protective covers. J. Appl. Physio/. 35,562-563 Krausz, S., R. Bernstein and J. Marder (1977). The acid-base balance of the rock partridge (Aiectoris chukar) exposed to high ambient temperatures. Camp. Biochem. Physiol. 57 A,245- 247 189

Larochelle, J., J. Delson and K. Schmidt-Nielsen (1982). Temperature regulation in the black vulture. Can. J. Zoo/. 60,491-494 Lasiewski, R.C. (1972). Respiratory function in birds. In 'Avian Biology.' (Eds D.S. Farner and J.R King) (Academic Press: New York.) Lasiewski, R.C., A.L. Acosta and M.H. Bernstein (1966). Evaporative water loss in birds I. Characteristics of the open flow method of determination, and their relation to estimates of thermoregulatory ability. Comp. Biochem. Physiol. 19,445- 457 Lasiewski, R.C. and W.R. Dawson (1967). A re-examination of the relation between standard metabolic rate and body weight in birds. Condor 69,13-23 Lasiewski, R.C. and W.R. Dawson (1969). Calculation and miscalculation of the equations relating avian standard metabolism to body weight. Condor 71,335-336 Lasiewski, R.C. and W.A. Calder (1971 ). A preliminary allometric analysis of repiratory variables in resting birds. Respir. Physiol. 11,152-166 Lasiewski, R.C., M.H. Bernstein and R.D. Ohmart (1971). Cutaneous water loss in the roadrunner and poor-will. Condor 73,4 70- 472 Lasiewski, R.C. and R.S. Seymour (1972). Thermoregulatory responses to heat stress in four species of birds weighing approximately 40 grams. Physiol. Zoo/. 45,106-118 Lee, P. and K. Schmidt-Nielsen (1971). Respiratory and cutaneous evaporation in the zebra finch: effect on water balance. Am. J. Physiol. 220,1598-1605 190

Louw, G.N., P .C. Belonje and H.J. Coetzee (1969). Renal function, respiration, heart rate, and thermoregulation in the ostrich ( Struthia came/us). Scient. pap. Namib Desert Res. Stn. 42,43-54 Lustick, S.J. (1969). Bird energetics: effects of artificial radiation. Science 163,387-390 Lustick, S.l. (1971 ). Plumage colour and energetics. Condor 73,121- 122 Lustick, S.l., M. Adam and A. Hinko (1980). Interaction between posture, colour, and the radiative heat load in birds. Science 208,1052-1053 Malan, A. (1973). Ventilation measured by body plethysmography in hibernating mammals and in poikilotherms. Respir. Physial. 17,32-44 Marder, J. (1973). Body temperature regulation in the brown-necked raven (Corvus carax rufical/is). I. Metabolic rate, evaporative water loss and body temperature of the raven exposed to heat stress. Camp. Biachem. Physiol. 45A,421- 430 Marder, J. and Z. Arad (1975). The acid base balance of abdim's stork (Sphenarhynchus abdimil) during thermal panting. Camp. Biachem. Physio/. 51 A,887-889 Marder, J. and J. Ben-Asher (1983). Cutaneous water evaporation. I. Its significance in heat-stressed birds. Camp. Biachem. Physial. 75A ,425-431 Marder, J. and I. Gavrieli-Levin (1987). The heat-acclimated pigeon: an ideal physiological model for a desert bird. J. Appl. Physiol. 62,952-958 1 91

Marder, J. and Z. Arad (1989). Panting and acid-base regulation in heat stressed birds. Camp. Biochem. Physiol. 94A,395-400 Marder, J., Y. Arieli and J. Ben-Asher (1989). Defense strategies against environmental heat stress in birds. lsr. J. Zoo/. 36,61-75 Marder, J. and P. Raber (1989). Beta-adrenergic control of trans­ cutaneous evaporative cooling mechanisms in birds. J. Camp. Physiol. B. 159,97-1 03 McArdle, B.H. (1988). The structural relationship: regression in biology. Can. J. Zoo/. 66,2329-2339 McNab, B.K. (1980a). On estimating thermal conductance in endotherms. Physiol. Zoo/. 53,145-156 McNab, B.K. (1980b). Food habits, energetics and the population biology of mammals. Am. Nat. 116,106-124 McNab, B.K. (1986a). The influence of food habits on the enrgetics of eutherian mammals. Ecol. Monogr. 56,1-19 McNab, B.K. (1986b). Food habits, energetics, and the reproduction of marsupials. J. Zoo/. (Lond.) 208,595-614 McNab, B.K. (1987). Basal rate and phylogeny. Fun ct. Eco/. 1,159-167 McNab, B.K. (1988). Food habits and the basal rate of metabolism in birds. Oecologia (Ber/.) 71,343-349 Mercer, J.B. (1989)(Ed.) 'Thermal Physiology.' Proceedings of the International Symposium on Thermal Physiology, Tromso, Norway, 16-21 July 1989. (Excerpta Medica: Amsterdam.) Midtgard, U. (1980). Heat loss from the feet of mallards Anas platyrhynchos and arterio-venous heat exchange in the rete tibiotarsale. Ibis 122,354-359 Midtgard, U. (1981 ). The rete tibiotarsale and arteriovenous association in the hind limb of birds: a comparative "192

morphological study on counter-current heat exchange systems. Acta Zoo/. (Stockh.) 62,67-87 Midtgard, U. ("1983). Scaling of the brain and the eye cooling system in birds: a morphometric analysis of the rete opthalmicum. J. Exp. Bioi. 225, "197-207 Midtgard, U., Z. Arad and E. Skadhauge (1983). The rete mirabile opthalmicum and the relation of its size to the body-to­ brain temperature difference in the fowl (Gallus domesticus). J. Camp. Physiol. 1538,241-246 Mills, S.H. and J.E. Heath ("1972). Responses to thermal stimulation of the preoptic area in the house sparrow (Passer domesticus). Am. J. Physio'J. 222,914-9"19 Mitchell, D., H.P. Laburn, M.J.M. Nijland, Y. Zurovsky and G. Mitchell (1987). Selective brain cooling and survival. South African Journal of Science 83,598-604 Molnar, R. and M. Archer (1984). Feeble and not so feeble flapping fliers: a consideration of early birds and bird-like reptiles. In 'Vertebrate zoogeography and evolution in Australasia.' (Eds M. Archer and G. Clayton) (Hesperian Press: Carlisle, Western Australia.)

Molony, V., W. Graf and P. Scheid (1976). Effects of C02 on pulmonary air flow resistance in the duck. Respir. Physiol. 26,333-349 Morgan, K.R., M.A. Chappell and T.L. Bucher (1992). Ventilatory oxygen extraction in relation to ambient temperature in four Antarctic seabirds. Physiol. Zoo/. 65,1 092-1113 Morton, M.L. (1967). The effects of insolation on the diurnal feeding pattern of white-crowned sparrows (Zonotrichia leucophrys gambelit}. Ecology 48,690-694 193

Murrish, D.E. (1973). Respiratory heat and water ~xchange in penguins. Respir. Physiol. 19,262-270 Nakashima, T., T. Hori, T. Kiyohara and M. Shibata (1985). Osmosensitivity of preoptic thermosensitive neurons in hypothalamic slices in vitro. Pfluegers Arch. Eur. J. Physiol. 405,112-117 Neal, C.M. and S.l. Lustick (1974). The effects of artificial radiation on energetics of the Eastern chipmunk, Tamias striatus. Camp. Biochem. Physiol. 47 A,277-281 Ohmart,R.D. and R.C. Lasiewski (1971 ). Roadrunners: energy conservation by hypothermia and absorption of sunlight. Science 172,67-69 0ritsland, N.A. and K. Ronald (1978). Solar heating of mammals: observations of hair transmittance. Int. J. Biometeor. 22,197-201 Ostrum, J.H. (1974). Archaeopteryx and the origin of flight. Q. Rev. Bioi. 49,27-47 Ostrum, J.H. (1979). Bird flight: how did it begin? Am. Sci. 67,46- 56 Parer, J.T. (1963). Wool length and radiant heating effects in sheep. J. Agric. Sci. 60,141-144 Peltonen, L.M., Y. Arieli and J. Marder (1989). Brain temperature regulation of panting and non panting pigeons exposed to extreme thermal conditions. Camp. Biochem. Physiol. 92A,91-96 Phillips, P.K. and J.E. Heath (1992). Heat exchange by the pinna of the African elephant (Loxodonta africana). Comp. Biochem Physiol. 101 A,693-699 194

Piiper, J., F. Drees and P. Scheid (1970). Gas exchange in the domestic fowl during spontaneous breathing and artificial ventilation. Respir. Physiol. 9,234-245 Pinshow, B., M.H. Bernstein, G.E. Lopez and S. Kleinhaus (1982). Regulation of brain temperature in pigeons: effects of corneal convection. Am. J. Physiol. 242,R577-R581 Porter, W.P. (1969). Thermal radiation in metabolic chambers. Science 166,115-117 Priestley, C.H.B. (1957). The heat balance of sheep standing in the sun. Aust. J. Agric. Res. 8,271-280 Ramirez, J.M. and M.H. Bernstein (1976). Compound ventilation during thermal panting in pigeons: a possible mechanism for minimizing hypocapnic alkalosis. Fed. Proc. 35,2562-2565 Rautenberg, W., R. Necker and B. May (1972). Thermoregulatory responses of the pigeon to changes of the brain and the spinal cord temperatures. Pfluegers Arch. Eur. J. Physiol. 338,31-42 Rautenberg, W. and R. Necker (1975). A comparison between peripheral and centrally generated cold sensitive signals in regulating heat production in the pigeon. In 'Depressed metabolism and Cold Thermogenesis.' (Ed. L. Jansky) (Charles University Press: Prague.) Ray, P.J. and M.R. Fedde (1969). Responses to alterations in

respiratory Po2 and Pco2 in the chicken. Respir. Physiol. 6,135-143 Richards, S.A. (1970a). The biology and comparative physiology of thermal panting. Bioi. Rev. 45,223-264 Richards, S.A. (1970b). Brain temperature and the cerebral circulation of the chicken. Brain Res. 23,265-268 195

Richards, S.A. (1976). Evaporative water loss in domestic fowls and its partition in relation to ambient temperature. J. Agric. Sci. 87,527-532 Rintamaki, H., S. Saarela, A. Marjakangas and R. Hissa (1983). Summer and winter temperature regulation in the black grouse, Lyrurus tetrix. Physio/. Zoo/. 56,152-159 Rintamaki, H., L. Karplund, H. Linden and R. Hissa (1984). Sexual differences in temperature regulation and energetics in the capercaillie, Tetrao urogallus. Ornis Fenn. 61 ,69-7 4 Roberts, J.R. and R.V. Baudinette (1986). Thermoregulation, oxygen consumption and water turnover in stubble quail, Coturnix pectoralis, and king quail, Cbturnix chinensis. Aust. J. Zoo/. 34,25-33 Rowell, L.B. (1977). Competition between skin and muscle for blood flow during exercise. In 'Problems with Temperature Regulation during Exercise.' (Ed. E.R. Nadel) (Academic Press: New York.) Salt, G.W. and E. Zuethen (1960). The respiratory system. In 'Biology and Comparative Physiology of Birds.' (Ed. A.J. Marshall) (Academic Press: New York.) Schied, P. (1979). Mechanisms of gas exchange in bird lungs. Rev. Physio/. Biochem. Pharmacal. 86,137-186 Schied, P. and J. Piiper (1972). Cross current gas exchange in avian lungs: effects of reversed parabronchial air flow in ducks. Respir. Physiol. 16,304-312 Schied, P., H. Slama and J. Piiper (1972). Mechanisms of unidirectional flow in parabronchi of avian lungs: measurements in duck lung preparations. Respir. Physiol. 14,83-95 196

Schied, P., M.R. Fedde and J. Piiper (1989). Gas exchange and air-sac composition in the unanaesthetized, spontaneously breathing goose. J. Exp. Bioi. 142,373-385 Schmidt, I. (1976). Paradoxical changes of respiratory rate elicited by altering rostral brain stem temperature in the pigeon. Pfluegers Arch. Eur. J. Physiol. 367,111-113 Schmidt, I. and E. Simon (1982). Negative and positive feedback of central nervous system temperature in thermoregulation in pigeons. Am. J. Physiol. 243 (Reg. Int. Comp. Physiol.

12) I R363-R372 Schmidt-Nielsen, K. (1964). 'Desert Animals: Physiological Problems of Heat and Water.' (Oxford University Press: New York.) Schmidt-Nielsen, K. (1971). How birds breathe. Sci. Am. 225,72-79 Schmidt-Nielsen, K. (1975). Recent advances in avian respiration. Symp. Zoo/. Soc. Lond. 35,33-47 Schmidt-Nielsen, K. (1979). 'Animal Physiology: adaptation and environment 2Ed.' (Cambridge University Press: Cambridge.) Schmidt-Nielsen, K. and B. Schmidt-Nielsen (1952). Water metabolism of desert mammals. Physiol. Rev. 32,135-166 Schmidt-Nielsen, K., B. Schmidt-Nielsen, S.A. Jarnum and T.R. Houpt (1957). Body temperature of the camel and its relation to water economy. Am. J. Physiol. 188,103-112 Schmidt-Nielsen, K., J. Kanwisher, R.C. Lasiewski, J.E. Cohn and W.L. Bretz (1969). Temperature regulation and respiration in the ostrich. Condor 71,341-352 Schmidt-Nielsen, K., F.R. Hainsworth and D.E. Murrish (1970). Counter-current heat exchange in the respiratory passages: effect on water and heat balance. Respir. Physiol. 9,263-276 197

Scholander, P.F., V. Walters, R. Hock and L. Irving (1950). Body insulation of some arctic and tropical mammals and birds. Bioi. Bull. 99,225-236 Senay, L.C. (1979). Temperature regulation and hypohydration: a singular view. J. Appl. Physiol.: Respirat. Environ. Exercise Physiol. 47,1-7 Silva, N.L. and J.A. Boulant (1984). Effects of osmotic pressure, glucose, and temperature on neurons in preoptic tissue slices. Am. J. Physiol. 247, R335-R345 Simon, E. (1987). Glossary of terms for thermal physiology. Pfluegers Arch. Eur. J. Physiol. 410,567-587 Simon, E., C.H. Simon-Opperman, H:T Hammell, R. Kaul and J. Maggart (1976). Effects of altering rostral brain stem temperature on temperature regulation in the adelie penguin, Pygoscelis adeliae. Pfluegers Arch. Eur. J. Physiol. 362,7-13 Simon, E., F.K. Pierau and D.C.M. Taylor (1986). Central and peripheral thermal control of effectors in homeothermic temperature regulation. Physiol. Rev. 66,235-300 Simon, E. and P. Nolte (1990). Temperature dependence of thermal and nonthermal regulation: hypothalamic thermo- and osmoregulation in the duck. In 'Thermoreception and temperature regulation.' (Eds J. Bligh and K. Voigt) (Springer Verlag, Berlin.) Simon-Opperman, C., E. Simon, C. Jessen and H.T. Hammel (1978). Hypothalamic thermosensitivity in conscious Pekin ducks. Am. J. Physiol. 235 (Regulatory Integrative Comp. Physiol. 4),R130-R140 Skadhauge, E. (1974). Renal concentrating ability in selected West Australian birds. J. Exp. Bioi. 61,269-276 198

Skadhauge, E., S.K. Maloney and T.J. Dawson (1991 ). Osmotic adaptation of the emu (Dromaius novaehollandiae). J. Camp. Physiol. B. 161,173-178 Snapp, B.P., H.C. Heller and S.M. Gasper (1977). Hypothalamic thermosensitivity in California quail (Lophortyx californicus). J. Camp. Physiol. B. 117,345-357 Sakal, R.R. and F.J. Rohlf (1981 ). 'Biometry 2Ed.' (W.H. Freeman and Company: New York.) Sokoloff, L. (1971 ). The action of thyroid hormones. In 'Handbook of Neurochemistry.' (Ed. A. Lajtha) (Plenum Press: New York.) Stafford-Smith, D.M. and S.R. Morton (1990). A framework for the ecology of arid Australia. J:· Arid Environments 18,255-278 Stahel, C.D. and S.C. Nicol (1988a). Comparison of barometric and pneumotachographic measurements of resting ventilation in the little penguin (Eudyptula minor). Camp. Biochem. Physiol. 89A,387-390 Stahel, C.D. and S.C. Nicol (1988b). Ventilation and oxygen extraction in the little penguin (Eudyptula minor), at different temperatures in air and water. Respir. Physiol. 71,387-398 Stahl, W.R. (1967). Scaling of respiratory variables in mammals. J. Appl. Physiol. 22,453-460 Steen, I. and J.B. Steen (1965). The importance of the legs in the thermoregulation of wading birds. Acta. Physiol. Scand. 63,285-291 Taylor, C.R. (1970a). Strategies of temperature regulation: effect on evaporation in East African ungulates. Am. J. Physiol. 219,1131-1135 199

Taylor, C.R. (197Gb). Dehydration and heat: effects on temperature regulation of East African ungulates. Am. J. Physiol. 219,1136-1139 Turlejska-Stelmasiak, E. (1974). The influence of dehydration on heat dissipation mechanisms in the rabbit. J. Physiol. (Paris) 68,5-15 Turlejska, E. and M.A. Baker (1986). Elevated CSF osmolality inhibits thermoregulatory heat loss responses. Am. J. Physiol. 251 ,R749-R754 van de Hulst H.C. (1980). 'Multiple light scattering. Tables, formulas, and applications.' (Academic Press: New York.) Walsberg, G.A. (1980). The glossy appearance of a black bird: thermal effects. J. Therm. Bioi. 5,185-187 Walsberg, G.A. (1983). Coat color and solar heat gain in animals. Bioscience 33,88-91 Walsberg, G.A. (1988a). Heat flow through avian plumages: the relative importance of conduction, convection, and radiation. J. Therm. Bioi. 13,89-92 Walsberg, G.A. (1988b). The significance of fur structure for solar heat gain in the rock squirrel, Spermophilus variegatus. J. Exp. Bioi. 138,243-257 Walsberg, G.A. (1988c). Consequences of skin color and fur properties for solar heat gain and ultravioloet irradiance in two mammals. J. Comp. Physiol. B. 158,213-221 Walsberg, G.A. (1990). Conergence of solar heat gain in two squirrel species with contrasting coat colors. Physio/. Zoo!. 63,1025-1042 Walsberg, G.A. (1992). Quantifying radiative heat gain in animals. Am. Zoo/. 32,217-224 200

Walsberg, G.A. and J.R. King (1978). The relationship between external surface area of birds to skin surface area and body mass. J. Exp. Bioi. 76,185-189 Walsberg, G.A., G.S. Campbell and J.R. King (1978). Animal colour and radiative heat gain: a re-evaluation. J. Comp. Physio/. 126,211-222 Walsberg, G.A. and C.A. Schmidt (1989). Seasonal adjustment of solar heat gain in a desert mammal by altering coat properties independently of surface coloration. J. Exp. Bioi. 142,387-400 Weathers, W .W. (1979). Climatic adaptation in avian standard metabolic rate. Oeco/ogia (Berl.) 42,81-89 Weathers, W.W. (1981 ). Physiological thermoregulation in heat­ stressed birds: consequences of body size. Physio/. Zoo/. 54,345-361 Weathers, W.W. and D.F. Caccamise (1975). Temperature regulation and water requirements of the monk parakeet, Myiopsitta monachus. Oecologia (Berl.) 18,329-343 Weathers, W.W. and D.S. Schoenbaechler (1976). Regulation of body temperature in the budgerygah, Me/opsittacus undulatus. Aust. J. Zoo/. 24,39-47 Webster, M.D., G.S. Campbell and J.R. King (1985). Cutaneous resistance to water-vapor diffusion in pigeons and the role of the plumage. Physiol. Zoo/. 58,58-70 Webster, M.D. and J.R. King (1987). Temperature and humidity dynamics of cutaneous and respiratory evaporation in pigeons, Columba Iivia. J. Comp. Physiol. B. 157,253-260 201

Whittow, G.C. (1986). Regulation of body temperature. In 'Avian Physiology 4Ed.' (Ed. P.O. Sturkie) (Springer-Verlag: New York.) Winston, P.W. and D.H. Bates (1960). Saturated solutions for the control of humidity in biological research. Ecology 41,232- 237 Withers, P.C. (1983). Energy, water and solute balance of the ostrich, Struthio came/us. Physiol. Zoo/. 56,568-579. Withers, P.C. and T.M. Crowe (1980). Brain temperature fluctuations in helmeted guineafowl under semi-natural conditions. Condor 82,99-100 Withers, P .C., R.B Forbes, and M.s:· Hedrick. (1987). Metabolic, water and thermal relations of the Chilean tinamou. Condor 89,424-426. Withers, P.C. and J.B. Williams (1990). Metabolic and respiratory physiology of an arid-adapted Australian bird, the spinifex pigeon. Condor 92,961-969 Wunder, B.A. (1979). Evaporative water loss from birds: effects of artificial radiation. Comp. Biochem. Physiol. 63A,493-494 Yates, F.E. (1981 ). Comparative physiology of energy production:

homeotherms and poikilotherms. Am. J. Physio/. 240, R 1-R2 Young, B.A., T.W. Fenton, and J.A. Mclean. (1984). Calibration methods in respiratory calorimetry. J. Appl. Physiol. 56,1120-1125. Zar, J.H. (1968). Calculation and miscalculation of the allometric equation as a a model in biological data. Bioscience 18,1118-1120 202

Appendix 1

List of Abbreviations

Ar effective radiating surface area BAT brown adipose tissue BMR basal metabolic rate BTPS body temperature, ambient pressure, saturated c thermal conductance Ca air boundary layer thermal conductance CEWL cutaneous evaporative water loss Cp plumage thermal conductance

Ctot thermal conductance of whole plumage sample (skin + feathers + air boundary layer) df degrees of freedom EHL evaporative heat loss ED2 oxygen extraction evaporative water loss partial C02 content of air leaving the metabolism chamber

FE~o partial water content of air leaving the metabolism chamber partial 02 content of air leaving the metabolism chamber

Fie~ partial C02 content of air entering the metabolism chamber

Flo2 partial 02 content of air entering the metabolism chamber fR ventilation frequency (breaths/minute) g grams ru= heat flow HLR heat load from radiation 203

kg kilograms LCT lower critical temperature M Mass ml millilitres fVR metabolic rate at a particular T a m/s metres per second NST non-shivering thermogenesis p the probability of radiation striking a fur or feather element within a layer of thickness dz Paco.z partial pressure of C02 in arterial blood PC102 partial pressure of 02 in arterial blood p~ partial pressure of C02 in air leaving the alveolus p~ partial pressure of 02 in air leaving the alveolus PC02 partial pressure of C02 P02 partial pressure of 02 Q heat flow 01 o the ratio of the values of any variable at two temperatures 1 0°C apart r2 coefficient of determination rc coat insulation re air boundary layer insulation REWL respiratory evaporative water loss A-1 relative humidity

FO respiratory quotient (VC02 1 V02) rz insulation between the point of radiation absorption and the coat surface SA surface Area S8v1 standard error of the mean SEslope standard error of the slope 204

SNK student Newnam-Keuls multiple range test STPD standard temperature (0°C), pressure (1 atm), dry t students t value ambient temperature body temperature Te plumage surface temperature Tex expired air temperature TNZ thermoneutral zone Tp plate temperature

Trh temperature of air in humidity sensor Ts skin temperature v flow rate of dry air into the chamber VCD2 carbon dioxide production rate VI ventilation rate (mL/minute) oxygen consumption rate ventilation I perfusion ratio VT tidal volume z average depth of radiation penetration into an animal's coat a. the fraction of intercepted radiation that is absorbed by the fur or feather element. e surface emmissivity cr stefan-Boltzman constant p the fraction of intercepted radiation that is reflected at an angle greater than goo, that is back out of the coat radiative heat flux (watts)

't the fraction of intercepted radiation that is reflected at an angle less than goo, that is it travels further into the coat. 205

Appendix 2

Summary of the Model of Cena and Monteith (1975) relating Radiative Heat Load to Coat Structure and Colour

The fate of radiation incident on an animals coat depends on;

p = the probability of radiation striking a fur or feather element within a layer of thickness dz (this factor is essentially a function of coat structure but is also influenced by the animals orientation toward the incoming radiation, see Walsberg 1gao). p = the fraction of intercepted radiation that is reflected at an angle greater than goo, that is back out of the coat (this is the radiation that is perceived as the animals colour).

't = the fraction of intercepted radiation that is reflected at an angle less than goo, that is it travels further into the coat.

a = the fraction of intercepted radiation that is absorbed by the fur or feather element.

If a coat is divided into layers, each of thickness dz (see Diagram), then at each layer there will be a radiation flux travelling toward the skin (-), and a reflected flux travelling out of the coat(+). The strength of the outward flux, +, will be attenuated by absorption and reflection ([a+p].+.p.dz) but augmented by reflection of part of the inward stream (p.-.p.dz). Thus: 206

Radiation }· ~--~------Te .. ~ ...

~:--- -::------~~ ~ ~ ~: ~-~-~--=~ ~ ~ . ~+dz

rc- rz ... ~· Ts ------Tp Diagram Illustrating the Variables Involved in the Calculation of Heat Load from Radiation (abbreviations explained in text)

Similarly the inward flux (

The radiation absorbed and converted to heat within a layer dz equals the change in total radiant flux density going from z to z+dz: 207

(Walsberg et a/. 1978). This heat then alters the temperature gradient from the animal's skin to the enviroment, such that heat flow out from the skin is reduced or reversed. The heat load at skin level can be determined because the absorbed radiation can either flow toward the skin or toward the environment and does so in inverse proportion to the resistance (insulation) in each direction. For example for radiation absorbed at layer z (Figure 6.6), if the insulation of rz+re is half that of r0 -rz, then twice as much heat will go to the environment as to the skin. The fracton of absorbed radiation that will act as a heat load at skin level is:

(Walsberg 1983). Some confusion exists in the literature because occasionally rz is defined as the distance from the skin to level z and the fraction of heat absorbed acting as a heat load on the skin is given as:

(Walsberg 1988b, Walsberg and Schmidt 1989). This should be:

The total heat load at skin level is then calculated by summing the heat flow toward the skin for each dz:

Os(z) = Q(z) (rz+re) I (rc+re) CO~- Lf6