UNIVERSITY OF CALIFORNIA

SANTA CRUZ

FUR VERSUS BLUBBER: A COMPARATIVE LOOK AT MARINE MAMMAL INSULATION AND ITS METABOLIC AND BEHAVIORAL CONSEQUENCES

A dissertation submitted in partial satisfaction of the requirements for the degree of

DOCTOR OF PHILOSOPHY

in

ECOLOGY AND EVOLUTIONARY BIOLOGY

by

Heather Elizabeth Mostman Liwanag

December 2008

The dissertation of Heather Elizabeth Mostman Liwanag is approved:

______Professor Terrie M. Williams, Chair

______Professor Daniel P. Costa

______Professor Annalisa Berta

______Lisa C. Sloan Vice Provost and Dean of Graduate Studies

Copyright © by

Heather Elizabeth Mostman Liwanag

2008 TABLE OF CONTENTS

page

LIST OF FIGURES………………………………………………………………….vii

LIST OF TABLES…………………………………………………………………….x

ABSTRACT……………………………………………………………………….....xi

ACKNOWLEDGEMENTS………………………………………………………….xv

INTRODUCTION

Introduction……………………………………………………………………1

References……………………………………………………………………..7

CHAPTER 1. Morphological and thermal properties of mammalian insulation: implications for the evolutionary transition to an aquatic lifestyle.

Abstract………………………………………………………………………11

Introduction…………………………………………………………………..12

Materials and Methods……………………………………………………….14

Fur and Blubber Sampling…………………………………………...14

Fur Characteristics…………………………………………………...15

Blubber Characteristics………………………………………………20

Thermal Conductivity………………………………………………..23

Statistical Analyses…………………………………………………..25

Results………………………………………………………………………..26

Fur Characteristics…………………………………………………...26

Blubber Characteristics………………………………………………32

iii Total insulation………………………………………………………36

Discussion……………………………………………………………………37

Modification of fur for aquatic living………………………………..37

The evolutionary transition to blubber……………………………….41

Ecological and evolutionary implications……………………………46

Tables and Figures…………………………………………………………...51

References…………………………………………………………………..102

Appendix……………………………………………………………………112

CHAPTER 2. The effects of water temperature on the energetic costs of juvenile and adult California sea lions (Zalophus californianus): The importance of recycling endogenous heat during activity.

Abstract……………………………………………………………………..121

Introduction…………………………………………………………………122

Materials and Methods……………………………………………………...125

Animals……………………………………………………………..125

Experimental protocol……………………………………………...126

Oxygen consumption……………………………………………….127

Core temperature……………………………………………………129

Statistical analyses and calculations………………………………..130

Results………………………………………………………………………132

Resting Metabolism and Thermal Neutral Zone……………………132

Core Temperature…………………………………………………..133

Activity Thermogenesis…………………………………………….134 iv

Discussion…………………………………………………………………..136

Thermal capabilities of the California sea lion……………………..136

Thermal limitations in juvenile sea lions…………………………...138

The effect of locomotor activity on thermal responses……………..140

Tables and Figures………………………………………………………….145

References…………………………………………………………………..157

CHAPTER 3. Energetic costs and thermoregulation in northern fur seal (Callorhinus ursinus) pups: The importance of behavioral strategies for thermal balance in furred marine mammals.

Abstract……………………………………………………………………..162

Introduction…………………………………………………………………163

Materials and Methods……………………………………………………...167

Animals……………………………………………………………..167

Protocol……………………………………………………………..167

Oxygen consumption……………………………………………….168

Determination of the thermal neutral zone…………………………169

Behavioral analyses………………………………………………...170

Statistical analyses………………………………………………….171

Results………………………………………………………………………172

Resting Metabolism and Thermal Neutral Zone……………………172

Jughandling Metabolism……………………………………………173

Grooming Metabolism……………………………………………...175

v Effects of water temperature on behavior…………………………..175

Effects of feeding on behavior……………………………………...176

Discussion…………………………………………………………………..177

Thermal capabilities of northern fur seal pups……………………..177

The effects of behavioral strategies on thermal responses...... 180

The effects of ingested prey items on thermal responses…………..183

Conclusions…………………………………………………………185

Tables and Figures………………………………………………………….187

References…………………………………………………………………..203

vi LIST OF FIGURES

page

CHAPTER 1

FIGURE 1. Conductivity chamber, showing placement of thermocouples relative to the sample and standard material…………………………………………….56

FIGURE 2. Carnivore phylogeny used for evolutionary comparisons……………...58

FIGURE 3. Scanning electron micrographs of representative hairs from terrestrial and marine carnivores, showing the cuticular scale patterns………………...60

FIGURE 4. Elongation (maximum length / maximum width) of the cuticular scales of carnivore guard hairs and underhairs……………………………………...71

FIGURE 5. Circularity (minimum diameter / maximum diameter) of carnivore guard hairs and underhairs………………………………………………………….73

FIGURE 6. Lengths of carnivore guard hairs and underhairs………………………75

FIGURE 7. Fur density and blubber thickness among carnivores…………………..77

FIGURE 8. Fur densities among otariid species, with a cladogram illustrating the evolutionary relationships among the species………………………………..79

FIGURE 9. Character evolution of guard hair circularity among pinniped species, with the polar bear as an outgroup…………………………………………...81

FIGURE 10. Character evolution of guard hair length and underhair length among pinniped species, with the polar bear as an outgroup………………………..83

FIGURE 11. Character evolution of fur density among pinniped species, with the polar bear as an outgroup…………………………………………………….86

FIGURE 12. Amount of air trapped in the fur and lost during diving for the pelts of eight pinniped species………………………………………………………..88

FIGURE 13. Height of guard hairs and underhairs for pinniped pelts when dry, immersed, and under hydrostatic pressure…………………………………...90

FIGURE 14. Percent lipid content (by weight) of pinniped blubber………………..92 vii

FIGURE 15. Change in thermal conductivity with lipid content of blubber for phocids, sea lions, and fur seals……………………………………………...94

FIGURE 16. Percent composition of fatty acid types in pinniped blubber…………96

FIGURE 17. Thermal resistance of the total insulation (in m2·°C·W-1) in relation to body mass (in kg) for carnivores………………………………………….....98

FIGURE 18. Schematic comparing relative body size and insulating layers of fur seals and sea lions…………………………………………………………..100

CHAPTER 2

FIGURE 1. Resting metabolism and core body temperature in relation to water temperatures for mature and immature California sea lions………………..147

FIGURE 2. Metabolic rate in relation to temperature for juvenile California sea lions at rest and while swimming at 1 m/s………………………………………..149

FIGURE 3. Metabolic heat production in relation to calculated swim velocity for four juvenile California sea lions…………………………………………...151

FIGURE 4. The effects of body mass on swimming metabolic rates for four juvenile California sea lions…………………………………………………………153

FIGURE 5. Comparison of predicted and measured metabolic heat production in relation to calculated swim velocity for juvenile sea lions…………………155

viii CHAPTER 3

FIGURE 1. Northern fur seal pup in the jughandle position, with one foreflipper held between both hind flippers above the water………………………………..191

FIGURE 2. Representative metabolic trace for a northern fur seal exhibiting the jughandling posture, illustrating the characteristic changes in metabolic rate associated with this behavior……………………………………………….193

FIGURE 3. Resting metabolic rate in relation to water temperature for three northern fur seal pups………………………………………………………………...195

FIGURE 4. Metabolic rate according to behavior for northern fur seal pups……..197

FIGURE 5. Percent time spent grooming in relation to water temperature for three northern fur seal pups……………………………………………………….199

FIGURE 6. Comparison of thermal liability of sea otter and pinniped species…...201

ix LIST OF TABLES

page

INTRODUCTION

TABLE 1. Fur vs. Blubber……………………………………………………………5

CHAPTER 1

TABLE 1. Hair, pelt, and full sculp samples analyzed in the present study………..51

TABLE 2. Thermal conductivities of pelts and blubber layers……………………..54

TABLE A1. Species means ± 1SD for hair circularity from the present study……112

TABLE A2. Species means for hair length from the present study and published sources………………………………………………………………………113

TABLE A3. Species means ± 1 SD for fur density from the present study and published sources…………………………………………………………...115

TABLE A4. Species means ± 1 SD (by age class) for blubber thickness, from the present study and published sources………………………………………..117

TABLE A5. Species means ± 1 SD for blubber and subcutaneous fat lipid content, from the present study and published sources……………………………...120

CHAPTER 2

TABLE 1. Body mass, resting metabolic rate, and lower critical temperatures of the California sea lions used in this study………………………………………145

CHAPTER 3

TABLE 1. Body mass, lower critical temperatures, and metabolic rates for northern fur seal pups………………………………………………………………...187

TABLE 2. Lower critical temperatures (TLC) for sea otters and pinnipeds resting in water………………………………………………………………………...189 x FUR VERSUS BLUBBER:

A COMPARATIVE LOOK AT MARINE MAMMAL INSULATION

AND ITS METABOLIC AND BEHAVIORAL CONSEQUENCES

Heather Elizabeth Mostman Liwanag

ABSTRACT

This research compares the use of fur and blubber as insulation in mammals, with a focus on the otariids (fur seals and sea lions) and the transition of mammals to aquatic living. The otariids represent the only mammalian family to include both types of insulation: fur seals have dense, waterproof fur and a moderate blubber layer, while sea lions rely solely on blubber for insulation in water. To compare the effectiveness of these different mechanisms of insulation, I examined the thermal properties of the fur and blubber of various otariid species, and evaluated their role in the physiological and behavioral responses of the California sea lion (Zalophus californianus) and northern fur seal (Callorhinus ursinus) exposed to different water temperatures.

Chapter 1 presents the morphological and thermal characteristics of fur and blubber as measured on sculps (fur, skin, and blubber) from a wide variety of pinniped species. Values were then compared to those of pelts (fur and skin) from terrestrial and semi-aquatic carnivores. Morphological measurements included hair cuticle shape, hair circularity, hair length, fur density, and blubber thickness. In addition, the composition of blubber fat was determined from lipid and water content, and fatty acid profiles. The effects of hydrostatic pressure on the insulating layer were determined empirically. Measurement of the thermal conductivity of the samples allowed assessment of the overall effectiveness of each insulation type. I found consistent trends in hair morphology associated with aquatic living, which included 1) flattening and shortening of the hairs, 2) elongation of hair cuticle scales, and 3) increases in fur density for species utilizing fur for insulation in water. Such characteristics are considered critical for maintaining an insulating air layer within the fur during submersion. I also observed a secondary loss of these features in species with more developed blubber layers. Comparisons of blubber composition indicated stratification of this layer in species relying on the blubber for insulation. Lipid stratification was consistent with the use of the outer layer for thermoregulation and the inner layer for energy storage. Among otariids, blubber quality (lipid content and thermal conductivity) did not differ between fur seals and sea lions. Rather, blubber quantity (thickness) differentiated each otariid group. Overall, differences in total insulation among carnivore species, both terrestrial and aquatic, were influenced substantially by body size and habitat, and to a lesser extent by latitudinal climate.

Chapters 2 and 3 subsequently address the influence of body size, insulation type, and maturity on whole-animal metabolic responses to water temperature. Once again, California sea lions and northern fur seals were compared as representatives of two different thermal mechanisms. Lower critical temperatures (TLC) and the metabolic consequences of at-sea behaviors were determined for both species. By comparing the TLC determined in the laboratory with routine temperatures encountered in the wild, I was able to assess a relative thermal liability for each species. With a TLC of 5 °C, adult female California sea lions demonstrated thermal competence across the natural range of water temperatures encountered by this species in the wild. In contrast, juvenile California sea lions had a TLC greater than

12 °C, revealing a potential thermal limitation that was mitigated by swimming activity. Northern fur seal pups, small-bodied marine mammals using only fur for insulation, had a TLC of 8 °C, and thus also exhibited thermal limitations within their natural water temperature range. Based on this and the pelagic lifestyle of young fur seals, other mechanisms of thermoregulation would be required for the animals to remain in thermal balance when resting at sea. I propose that young fur seals utilize both the heat increment of feeding as well as a variety of at-sea behaviors, including grooming and a unique jughandle position, to mitigate the observed thermal limitation.

In summary, the results of these studies indicate that fur underwent considerable evolutionary modification for aquatic living. Ultimately, based on an examination of extant species, fur acts as an effective insulator in water for small- bodied endotherms. Fur is the superior insulator in terms of thermal resistance per unit thickness, as long as the air layer can be maintained among the hairs. This requires a metabolic investment by the animal through grooming. With larger body size, marine mammals can develop extraordinarily thick blubber layers, which facilitate thermal balance during swimming and diving. This form of insulation can serve additional roles such as buoyancy control, energy storage, and streamlining. Overall, I find that the most effective form of insulation for aquatic mammals depends on body size as well as habitat. Heat generated by the processing of food and through skeletal muscle thermogenesis during activity helps to mitigate thermal shortfalls in insulation, and this mechanism of thermal substitution appears to be especially important for maintaining thermal balance in the smaller aquatic species and during immature life stages. ACKNOWLEDGEMENTS

There are so many people that I want to thank for their contribution to this dissertation. First and foremost, I want to thank my husband, Leslie Liwanag, for his unwavering support throughout my research. He has contributed so much, from helping me to build equipment and refine figures, to providing emotional support and encouragement throughout the process, and I am enormously grateful for everything he has done for me.

It would have been nearly impossible to complete the laboratory and field experiments without the assistance and dedication of numerous undergraduate and graduate volunteers. I am truly amazed at the level of dedication and commitment that so many people demonstrated, and I cannot express the depth of my gratitude. In order to thank everyone, I have included a section of acknowledgments for each chapter. However, if I missed any names it is not for lack of appreciation.

I would also like to thank my thesis committee, who both supported and challenged me throughout my research. I appreciate the encouragement and the challenges, as they shaped me as a scientist. I want to thank Dan Costa for the many intellectual discussions we’ve had, and his encouragement to pursue new angles and explore new hypotheses. I thank Annalisa Berta for her constant support and for ensuring that I look at the big picture at all times. Finally, I must thank my advisor,

Terrie Williams, for her trust and confidence in me as a scientist, and her encouragement to always push the science forward. She has greatly influenced me as both a researcher and a teacher. xv I want to thank the staff and volunteers of the Marine Mammal Physiology

Program at Long Marine Lab, especially Brett Long, Jen Gafney, Traci Fink, and

Beau Richter. Without their dedication and commitment to the research and the animals, none of the captive animal research would have been possible. I would also like to thank Frances Gulland and the staff and volunteers at the Marine Mammal

Center for their assistance with the transport and release of the fur seals.

In addition, I would like to thank all of my lab mates, who are not only colleagues but friends. They have helped me in both the lab and the field, have offered feedback on countless manuscripts and presentations, and have offered their support along the way. I thank my lab mates in the Williams lab: R. Dunkin, J.

Maresh, D. Monson, S. Noren Kramer, and L. Yeates. I also thank my lab mates in the Costa lab: C. Champagne, M. Fowler, S. Fowler, J. Gedamke, A. Harrison, J.

Hassrick, S. Hayes, L. Huckstadt, M. Kappes, C. Kuhn, C. Lester, S. Maxwell, Y.

Mitani, B. McDonald, P. Robinson, S. Shaffer, S. Simmons, Y. Tremblay, N.

Teutschel, S. Villegas, A. Walli, R. Weber, M. Weise, and K. Yoda. I have been fortunate to work with so many people in a variety of scientific pursuits. Special thanks to Carey Kuhn, whose mentorship over the years gave me the confidence to push forward as a scientist, and Jen Maresh, whose hours of editing and intellectual feedback have improved my science from the initial ideas to the presentation of data.

Finally, I want to thank my family and friends for their constant support and encouragement during these years and throughout my career as a biologist. Their unwavering confidence in me has always been a blessing, and I am so grateful. xvi INTRODUCTION

During the course of mammalian evolutionary history, several lineages secondarily invaded the aquatic environment. Two options were employed to deal with the potentially high rates of heat loss: increased metabolic rates and the development of specialized waterproof insulation to aid in retaining body heat. The former provides a comparatively short-term solution but is energetically expensive if maintained for extended periods. Semi-aquatic mammals such as mink (Williams,

1986), muskrats (Fish, 1979), and water rats (Dawson & Fanning, 1981) solve the problem of these elevated energetic costs by limiting the duration of immersion.

Obligate marine mammals such as the cetaceans do not have this option, and often demonstrate elevated basal metabolic rates that are 1.5 to 2.0 times those predicted for terrestrial mammals of similar body mass (Scholander et al., 1942; Kleiber, 1975;

Williams et al., 2000). Recent studies suggest that the basal metabolic rates of marine mammals may be highly variable and dependent on the individual species and life history patterns. Thus, many species, including cetaceans (Williams et al., 2001), sea lions (Hurley & Costa, 2001), and sea otters (Morrison et al., 1974; Costa &

Kooyman, 1982), have exceptionally high metabolic rates; others appear similar to terrestrial mammals (Lavigne et al., 1982); and still others such as the tropical manatees and dugongs have relatively low metabolic rates that average 25 - 30% of values predicted for terrestrial mammals of similar size (Miculka & Worthy, 1995).

It is likely that evolutionary history and the capacity of the metabolic pathways to 1 change have played key roles in terms of the energetic solutions reached by different mammalian lineages as they re-entered the water.

For semi-aquatic mammals, an increase in metabolic rate through locomotor activity and the processing of food may be used to supplement resting heat production and offset thermoregulatory costs in the short term. The importance of these processes for marine mammals is less well defined. For example, sea otters and pinnipeds may increase metabolic rate 30 - 67% over resting levels following the ingestion of prey, a large proportion of which is used to simply warm cold food

(Costa & Kooyman, 1984; Wilson & Culik, 1991; Markussen et al., 1994; Rosen &

Trites, 1997; Williams et al., 2004). Thermal benefits associated with skeletal muscle thermogenesis will also depend on the balance between elevated heat loss associated with convective cooling and increased heat production with the locomotor activity of the swimmer (Hind & Gurney, 1997).

The second mechanism for preventing excessive heat loss during immersion, specialized insulation, represents the most efficient long-term solution for maintaining a high stable core body temperature while living in water (Williams &

Worthy, 2002). Because both fur and blubber serve this role in marine mammals and have very different evolutionary and developmental origins, there are differences in the relative costs and benefits for each type of insulation.

The transition from terrestrial to aquatic living by mammals was accompanied by numerous modifications in the type of insulation used to retain body heat. Early changes in pelage included flattening of the guard hairs and increased hair density to 2 act as a water-resistant barrier and reduce hydrodynamic drag while swimming (Ling,

1970). As mammals became more specialized for living in water, the reliance on external insulation provided by fur decreased while internal insulation provided by thick blubber layers dominated. Thus, the most recent invaders of the aquatic environment, semi-aquatic mammals such as the platypus, beaver, muskrat, mink, and otters, utilize a dense, non-wettable fur for insulation; the deep-rooted marine mammal lineages including the sirenians (manatees, dugongs) and cetaceans use blubber; and the intermediate group, the pinnipeds (seals, sea lions, and walrus) use some combination of the two (Fish, 2000). Although phocids (true seals) have retained a fur covering, these pinnipeds rely primarily on a thick blubber layer when immersed due to the poor insulating quality of the wetted fur (Kvadsheim & Aarseth,

2002). Interestingly, members of the remaining marine mammal line, the otariids (fur seals and sea lions), demonstrate two distinct strategies for keeping warm in the cold waters in which they forage. Fur seals have both dense fur and a moderate blubber layer, while sea lions rely solely on blubber to prevent excessive heat loss. As a result, the Otariidae is the only mammalian family with closely related species using two very different forms of insulation, and provides an ideal group with which to examine 1) the relative costs and benefits of fur and blubber, and 2) factors influencing the evolutionary pathways by which insulative mechanisms develop.

The use of fur or blubber by aquatic mammals involves several trade-offs

(Table 1). As an insulator, fur is comparatively lighter than blubber and a better insulator in air, but its effectiveness is reduced in water (Scholander et al., 1950; 3 Hammel, 1955; Johansen, 1962; Ling, 1970; Frisch et al., 1974; Morrison et al.,

1974; Doncaster et al., 1990; Fish, 2000; Kvadsheim & Aarseth, 2002). The insulating air layer of fur allows the skin to be maintained at body temperature, a feature especially important for wound healing (T.M. Williams, unpublished data); however, this air layer requires the energetically expensive process of grooming to maintain (Williams, 1989; Yeates et al., 2007). The effectiveness of fur as a thermal barrier is also compromised at depth due to the compression of the air layer during a dive, which will also affect buoyancy control (Repenning, 1976; Lovvorn & Jones,

1991; Pabst et al., 1999). In contrast blubber may not provide the same level of buoyancy relative to fur, and it is relatively incompressible with depth (Repenning,

1976; Lovvorn & Jones, 1991). Furthermore, because blubber is living tissue, it can be bypassed by perfusion when the animal needs to dissipate excess heat (Schmidt-

Nielsen, 1990).

The energetic costs associated with maintaining each type of insulation also differ. Several studies have demonstrated that there is a significant energetic cost to growing and maintaining fur insulation by marine mammals. In sea otters 12 - 40% of the day may be spent in grooming the fur, an activity that increases metabolic rate

64% above resting (Williams, 1989; Yeates et al., 2007). Growing a pelt is also costly, even in pinnipeds with a sparsely furred integument. In northern elephant seal pups, 49 - 85% of the body protein lost during the post-weaning fast is used to replace the epidermis and pelage during the molt, with the remainder of the protein serving as a fuel to support metabolic costs (Noren, 2002). A longitudinal study of resting 4 metabolism in California sea lions found that metabolic rate increased from 8.7 ± 1.4 mlO2·kg-1·min-1 to 11.0 ± 0.8 mlO2·kg-1·min-1 during the 2.5 month molt (Williams et al., 2007). Blubber, as a living tissue, requires continuous metabolic maintenance and is also energetically costly to produce (Schmidt-Nielsen, 1990). Unlike fur, however, blubber can be used as an energy reserve when the animal is fasting or migrating, or when food is scarce (Lindsey, 1966; Worthy & Lavigne, 1983;

Lindstedt & Boyce, 1985; Nordøy & Blix, 1985; Millar & Hickling, 1990; Schmidt-

Nielsen, 1990; Pabst et al., 1999). Additionally, because blubber is subcutaneous, it does not have to be groomed to maintain its effectiveness as an insulator.

Table 1. Fur vs. Blubber. Advantages are shown in bold. (Bryden, 1968; Williams, 1989; Schmidt-Nielsen, 1990; Lovvorn & Jones, 1991; Pabst et al., 1999; Fish, 2000)

FUR BLUBBER light weight heavier better insulator for given volume and mass poorer insulator for given volume and mass skin is maintained at or near body temperature skin is at or near ambient temperature metabolically inert cost to metabolic maintenance requires air layer and grooming no grooming required; can withstand fouling compressed at depth – limits effectiveness relatively incompressible buoyancy changes with depth buoyancy does not change with depth metabolically unavailable can be used as an energy reserve cannot be bypassed by perfusion can be bypassed by perfusion

The diversity in thermal insulation among pinnipeds provides a unique opportunity to evaluate the key characteristics of mammalian insulation that facilitate thermal stability in the aquatic environment. By comparing fur seals and sea lions, I was able to evaluate the effectiveness of two forms of insulation, from the tissue level to whole animal metabolic and behavioral responses. The purpose of this study was 5 to determine the relative effectiveness of fur and blubber as insulating mechanisms in aquatic mammals, in the context of the selective pressures that may have driven the convergent evolution of blubber across mammalian lineages. At the tissue level, I examined the morphological and thermal properties of fur and blubber, including shape of the hair cuticle, hair circularity, hair length, hair density, thermal conductivity, compression of the insulation under pressure, blubber thickness, water and lipid content, and fatty acid composition. The energetic consequences and behavioral strategies associated with the use of each type of insulation in the aquatic environment were assessed at the whole animal level. Chapter 1 describes the morphological and thermal properties of the insulation, and considers the evolutionary pathways by which the different mechanisms for preventing heat loss evolved in aquatic mammals. Chapter 2 compares the thermoregulatory energetics of adult and juvenile California sea lions, and explores the importance of locomotor activity as a behavioral strategy for mitigating thermal limitations. Lastly, in Chapter

3, I assess the thermoregulatory energetics of weaned northern fur seal pups to determine the significance of body position, grooming, and the heat increment of feeding for thermoregulation in furred marine mammals. Overall, this is the first study to determine the relative effectiveness of fur and blubber as insulating layers in water, and how different mammalian lineages were able to take advantage of the specific costs and benefits of each to maintain an aquatic lifestyle.

6 References

Bryden, M. M. (1968). Growth and function of the subcutaneous fat of the elephant seal. Nature 220, 597-599.

Costa, D. P. and Kooyman, G. L. (1982). Oxygen consumption, thermoregulation, and the effect of fur oiling and washing on the sea otter, Enhydra lutris. Can. J. Zool. 60, 2761-2767.

Costa, D. P. and Kooyman, G. L. (1984). Contribution of specific dynamic action to heat balance and thermoregulation in the sea otter Enhydra lutris. Physiol. Zool. 57, 199-203.

Dawson, T. J. and Fanning, F. D. (1981). Thermal and energetic problems of semiaquatic mammals: a study of the Australian water rat, including comparisons with the platypus. Physiol. Zool. 54, 285-296.

Doncaster, C. P., Dumonteil, E., Barre, H. and Jouventin, P. (1990). Temperature regulation of young coypus (Myocastor coypus) in air and water. Am. J. Physiol. 259, R1220-R1227.

Fish, F. E. (1979). Thermoregulation in the muskrat (Ondatra zibethicus): the use of regional heterothermia. Comp. Biochem. Physiol. 64A, 391-397.

Fish, F. E. (2000). Biomechanics and energetics in aquatic and semiaquatic mammals: platypus to whale. Physiol. Biochem. Zool. 73, 683-698.

Frisch, J., Ørtisland, N. A. and Krog, J. (1974). Insulation of furs in water. Comp. Biochem. Physiol. 47A, 403-410.

Hammel, H. T. (1955). Thermal properties of fur. Am. J. Physiol. 182, 369-376.

Hind, A. T. and Gurney, W. S. C. (1997). The metabolic cost of swimming in marine homeotherms. J. Exp. Biol. 200, 531-542.

Hurley, J. A. and Costa, D. P. (2001). Standard metabolic rate at the surface and during trained submersions in adult California sea lions (Zalophus californianus). J. Exp. Biol. 204, 3273-3281.

Johansen, K. (1962). Buoyancy and insulation in the muskrat. J. Mammal. 43, 64-68.

Kleiber, M. (1975). The fire of life: an introduction to animal energetics. Huntington, N.Y.: R.E. Kreiger Publishing Co. 454 pp. 7

Kvadsheim, P. H. and Aarseth, J. J. (2002). Thermal function of phocid seal fur. Mar. Mamm. Sci. 18, 952-962.

Lavigne, D. M., Barchard, W., Innes, S. and Ørtisland, N. A. (1982). Pinniped bioenergetics. In Mammals in the seas: Small cetaceans, seals, sirenians, and others, vol IV (ed. G. S. Clark), pp. 191-235. Rome: FAO Fisheries Series 5.

Lindsey, C. C. (1966). Body sizes of poikilotherm vertebrates at different latitudes. Evolution 20, 456-465.

Lindstedt, S. L. and Boyce, M. S. (1985). Seasonality, fasting endurance, and body size in mammals. Am. Nat. 125, 873-878.

Ling, J. K. (1970). Pelage and molting in wild mammals with special reference to aquatic forms. Quarterly Review of Biology 45, 16-54.

Lovvorn, J. R. and Jones, D. R. (1991). Effects of body size, body fat, and change in pressure with depth on buoyancy and costs of diving in ducks. Can. J. Zool. 69, 2879-2887.

Markussen, N. H., Ryg, M. and Ørtisland, N. A. (1994). The effect of feeding on the metabolic rate in harbour seals (Phoca vitulina). J. Comp. Physiol. B 164, 89-93.

Miculka, T. A. and Worthy, G. A. J. (1995). Metabolic capabilities and the limits to thermoneutrality in juvenile and adult West Indian manatees (Trichechus manatus) (abstract). In Eleventh Biennial Conference on the Biology of Marine Mammals, Orlando Florida, 14-18 December 1995, p.77.

Millar, J. S. and Hickling, G. J. (1990). Fasting endurance and the evolution of mammalian body size. Funct. Ecol. 4, 5-12.

Morrison, P., Rosenmann, M. and Estes, J. A. (1974). Metabolism and thermoregulation in the sea otter. Physiol. Zool. 47, 218-229.

Nordøy, E. S. and Blix, A. S. (1985). Energy sources in fasting grey seal pups evaluated with computed tomography. Am. J. Physiol. 249, R471-R476.

Noren, D. P. (2002). Body energy reserve utilization during the postweaning fast of Northern elephant seals (Mirounga angustirostris): Implications for survival. Ph.D. dissertation, University of California, Santa Cruz. 151pp.

8 Pabst, D. A., Rommel, S. A. and McLellan, W. A. (1999). The functional morphology of marine mammals. In Biology of Marine Mammals (ed. J. E. Reynolds, S. A. Rommel), pp. 15-72. Washington, D.C.: Smithsonian Institution Press.

Repenning, C. A. (1976). Adaptive evolution of sea lions and walruses. Syst. Zool. 25, 375-390.

Rosen, D. A. S. and Trites, A. W. (1997). Heat increment of feeding in Steller sea lions, Eumetopias jubatus. Comparative Biochemistry & Physiology, 118A, 877-881.

Schmidt-Nielsen, K. (1990). Animal Physiology. (4th ed.) New York: Cambridge University Press. 602pp.

Scholander, P. F., Irving, L. and Grinnell, S. W. (1942). Aerobic and anaerobic changes in seal muscle during diving. J. Biol. Chem. 142, 431-440.

Scholander, P. F., Walters, V., Hock, R. and Irving, L. (1950). Body insulation of some arctic and tropical mammals and birds. Biol. Bull. 99, 225-236.

Williams, T. M. (1986). Thermoregulation in the North American mink during rest and activity in the aquatic environment. Physiol. Zool. 59, 293-305.

Williams, T. M. (1989). Swimming by sea otters: adaptations for low energetic cost locomotion. J. Comp. Physiol. 164A, 815-824.

Williams, T. M., Davis, R. W., Fuiman, L. A., Francis, J., Le Boeuf, B. J., Horning, M., Calambokidis, J. and Croll, D. A. (2000). Sink or Swim: Strategies for Cost-Efficient Diving by Marine Mammals. Science 288, 133- 136.

Williams, T. M., Estes, J. A., Doak, D. F. and Springer, A. M. (2004). Killer appetites: assessing the role of predators in ecological communities. Ecology 85, 3373-3384.

Williams, T. M., Haun, J., Davis, R. W., Fuiman, L. A. and Kohin, S. (2001). A killer appetite: metabolic consequences of carnivory in marine mammals. Comp. Biochem. Phys. A 129, 785-796.

Williams, T. M., Rutishauser, M., Long, B., Fink, T., Gafney, J., Mostman- Liwanag, H. and Casper, D. (2007). Seasonal variability in otariid

9 energetics: Implications for the effects of predators on localized prey resources. Physiol. Biochem. Zool. 80, 433-443.

Williams, T. M. and Worthy, G. A. J. (2002). Anatomy and physiology: the challenge of aquatic living. In Marine Mammal Biology: An Evolutionary Approach (ed. A. R. Hoelzel), pp. 73-97. Malden, MA: Blackwell Science Ltd.

Wilson, R. P. and Culik, B. M. (1991). The cost of a hot meal: Facultative specific dynamic action may ensure temperature homeostasis in post-ingestive endotherms. Comp. Biochem. Phys. A 100, 151-154.

Worthy, G. A. J. and Lavigne, D. M. (1983). Energetics of fasting and subsequent growth in weaned harp seal pups, Phoca groenlandica. Can. J. Zool. 61, 447- 456.

Yeates, L. C., Williams, T. M. and Fink, T. L. (2007). Diving and foraging energetics of the smallest marine mammal, the sea otter (Enhydra lutris). J. Exp. Biol. 210, 1960-1970.

10 CHAPTER 1

Morphological and thermal properties of mammalian insulation:

implications for the evolutionary transition to an aquatic lifestyle

Abstract

The Carnivora includes three independent evolutionary transitions to the marine environment: pinnipeds (seals, sea lions, and walruses), sea otters, and polar bears. Among these, only the pinnipeds have retained two forms of insulation, an external fur layer and an internal blubber layer for keeping warm in water. In this study I investigated key factors associated with the use of these two types of insulation in water, comparing fur and blubber characteristics among these lineages with those of strictly terrestrial carnivores. Characteristics included gross morphology (hair cuticle shape, circularity, length and density, blubber thickness), fat composition (fatty acid profiles, percentage lipid and water), and thermal conductivity. I found that marine carnivores have significantly flatter (F3,93=77.216,

P<0.001), shorter (F3,115=76.920, P<0.001), and denser hairs (F10,90=67.268, P<0.001) than terrestrial carnivores. However, sea lions, phocids, and walrus, which have thicker blubber layers than fur seals (F3,434=63.610, P<0.001), have lower fur densities than fur seals (P<0.001). Comparisons of lipid content, water content, and fatty acid composition indicated significant differences in the composition of the inner and outer regions of the blubber between groups (P<0.001), consistent with the

11 hypothesis that phocids and sea lions utilize the outer layer of their blubber primarily for thermal insulation, and the inner layer for energy storage. Fur seals, in contrast, rely more on their fur for thermal insulation, and utilize their blubber layer primarily for energy storage. Differences in the total insulation among carnivore species were influenced substantially by body size and habitat, and to a lesser extent by latitudinal climate. Overall, these results indicate consistent evolutionary modifications in the insulation of mammals and evidence for convergent evolution of thermal traits across lineages.

Introduction

During the course of mammalian evolution, numerous lineages secondarily invaded the aquatic environment. These included the cetaceans (whales and dolphins), sirenians (manatees and dugongs), pinnipeds (seals, sea lions, and walrus), sea otters, and polar bear. Key to the successful transition to living in water was modification of the type of insulation used to retain body heat and maintain a stable, relatively high core body temperature during prolonged submergence. Indeed, thermal stability is considered an evolutionary hallmark for fully aquatic mammals

(Irving, 1969; Irving, 1973).

Rather than a single insulating mechanism, two evolutionary pathways for insulation have been described for these marine living mammalian lineages

(Scholander et al., 1950; Ling, 1970; Pabst, Rommel & McLellan, 1999; Berta,

Sumich & Kovacs, 2006). First, mammals modified the ancestral form of insulation, 12 external fur, to serve as the primary thermal barrier in water. Accordingly, the two most recent mammalian lineages to reinvade the marine environment, the sea otter

(1.6 Ma) and the polar bear (0.5 Ma), still use fur as the primary insulator (Pabst et al., 1999; Berta et al., 2006). Other mammalian lineages developed a second type of insulation in the form of an internal blubber layer, which could also serve as an energy store. This is observed for the Cetacea and Sirenia, which invaded the water during the early Eocene (50 Ma), and have secondarily lost the hair covering that characterizes mammals (Hart & Fisher, 1964; Ling, 1970; Berta et al., 2006).

In general, specialization for aquatic living resulted in a trend in which there is a decreased reliance on external insulation via dense fur and a replacement by internal insulation provided by thick blubber layers. Interestingly, the Pinnipedia, as the group that evolved during the late Oligocene (29-23 Ma) between the Cetacea and sea otters, is the only mammalian lineage that has retained both types of insulation. In particular, the otariids (fur seals and sea lions) show two distinct mechanisms of retaining body heat: fur seals have dense, waterproof fur and a moderate blubber layer, while sea lions rely solely on their blubber for insulation in water. This diversity in thermal insulation among pinnipeds provides a unique opportunity to evaluate the key characteristics of mammalian insulation that facilitate thermal stability in the aquatic environment.

Therefore, the purpose of this study was to (i) investigate the relative importance and effectiveness of internal and external forms of insulation in mammals and (ii) examine the evolutionary pathway for insulation in aquatic living animals. 13 To accomplish this, I measured and compared morphological traits (shape of the hair cuticle, hair circularity, hair length, hair density, and blubber thickness), physical properties (thermal conductivity, compression of the insulation under pressure), and biochemical aspects (water and lipid content, fatty acid composition) of the fur and blubber of carnivore species, with particular focus on the pinnipeds. These traits were then examined in a phylogenetic context to determine which characters were critical for thermoregulation in water, and to identify how each mammalian lineage utilized these characters to facilitate a marine existence.

Materials and Methods

Fur and Blubber Sampling

Samples were collected from 25 species of terrestrial and marine carnivores

(Table 1). Because samples were obtained opportunistically, a range of age classes

(pups, juveniles, subadults, and adults) were available. As young mammals typically exhibit similar fur characteristics to adult conspecifics (Meyer, Schnapper &

Hülmann, 2002), all age classes were utilized for comparisons of fur characteristics; only adult and subadult animals (from 1 year to sexual maturity) were used for blubber comparisons. Fur and/or blubber samples were collected from full sculps

(fur, skin, and blubber, N=96) of deceased marine carnivores, fresh pelts (fur and skin, N=19), tanned pelts (N=10), or hairs removed from live animals (N=5). Fresh tissue samples were collected only from good condition carcasses, as determined by tissue color. Fresh sculp samples were 25 cm × 25 cm pieces taken from the dorsum, 14 just caudal to the shoulders. Fresh pelt samples were also taken from the same location, although the size of each sample varied (range = 5 cm × 5 cm to 25 cm × 25 cm). Hairs removed from tanned pelts or live animals were taken from the mid- dorsum. The tanning process could stretch the skin, resulting in an underestimation of hair density; however, other hair characteristics are not be affected by tanning.

Thus, the only tanned pelt used to determine hair density was that of the ermine, for which no other samples were available. All fresh sculp and pelt samples were wrapped in plastic wrap and stored in heavy-duty freezer bags to prevent desiccation.

All samples were kept frozen at -20 °C until used for analyses. Additional data were taken from the literature, where available. Measurements incorporated from other studies were from adult or subadult animals, and values from known emaciated or poor condition animals were omitted.

Fur characteristics

Shape of the hair cuticle

Using tweezers, three guard hairs and three underhairs were carefully removed from each sample under a dissecting microscope (SMZ645, Nikon

Instruments Inc., Melville, NY, USA). Hairs were washed with alcohol and prepared for scanning electron microscopy by mounting them on aluminum stubs with cyanoacrylic adhesive (Krazy glue®). The stub and hairs were then sputter-coated with gold-palladium. The coated specimens were viewed and digitally photographed with an ISI WB-6 scanning electron microscope (SEM) to visualize cuticular scale 15 patterns. Magnifications were matched as closely as possible, but were ultimately determined by the focal window of the microscope. Scale patterns were visualized near the base of each hair. Using ImageJ software (NIH, Bethesda, MD), scale shape was quantified with the ratio of scale length to scale width, for which elongated scales have larger values.

Hair length and diameter

Three guard hairs and three underhairs were removed from each sample, as described above. Each hair was washed with alcohol, laid flat in a coat of clear acrylic polish on a plastic sheet, and covered again with polish to keep the hair flat.

Hair length was measured to the nearest 0.01 mm with digital calipers (ABSOLUTE

Digimatic Caliper Series 500, Mitutoyo, Aurora, IL). The hair embedded in dried polish was then peeled from the sheet cover, and placed under a dissecting microscope. A cross section perpendicular to the length of the hair was taken at the widest point on the hair, and mounted vertically on a 1.3 cm Styrofoam block. This mount was viewed under a compound microscope (CH-3, Olympus America Inc.,

Center Valley, PA, USA), under 20× magnification for guard hairs and 40× or 100× magnification for underhairs. Using an ocular micrometer, the maximum and minimum diameter of each hair was determined. Note that hairs were easily distinguished from the polish mount under the microscope, so that only the hair diameter was measured. The ratio of the minimum diameter to the maximum diameter gave an index of circularity of the hair shaft, with perfect circularity 16 represented by a ratio of 1.0. For hair lengths and diameters, values for 3 guard hairs and 3 underhairs were averaged to yield a single value per individual for each hair type.

Fur densities

Fur densities were determined following the methods of Scheffer (1964a). A small (5 cm × 5 cm) square of skin with fur was isolated from the original sample, and any subcutaneous fat or blubber removed. The pelt sample was fixed in 10% buffered formalin for a minimum of 5 days, and then flattened in a press. Because skin tends to contract slightly when removed from the animal and even further when dried, the pressing process helped to restore the sample more closely to its living dimensions (Scheffer, 1964a; Fish et al., 2002). After the sample was flattened, the hair was cut with scissors and then shaved nearly flush with the skin using a safety razor. Six circular discs were cut with a trephine from each shaved sample. The trephine was a cylindrical metal tube sharpened at one end, with a 0.88 cm inner diameter. As the diameter of the trephine could change with repeated use, the inner diameter was measured with digital calipers (Mitutoyo, Aurora, IL) prior to use on each sample, and this value was used to calculate the area of each disc cut. Each disc was stored between two glass microscope slides until needed. Although the area of the disc might become distorted after cutting, the number of hair bundles could not change from further treatment of the disc; thus, the original area cut by the trephine was used to calculate disc area and ultimately hair density for each sample disc. 17 Mammalian fur is organized into bundles, each of which consists of a single guard hair accompanied by a number of underhairs (if present). To count the number of bundles present per unit area, I took digital photographs of three discs from each pelt sample, under a dissecting microscope (SZX7, Olympus America Inc., Center

Valley, PA, USA) with a 1× objective. Using ImageJ software (National Institute of

Health, Maryland, USA), every bundle on each disc was marked digitally, and the marks were summed. Mean bundle density was determined from the average of 3 discs per pelt sample. To count the number of underhairs per bundle, underhairs were manually counted for 20 representative bundles under a compound microscope at 10× or 20× magnification (CH-3, Olympus America Inc.). Mean number of underhairs per bundle was determined from the average of 3 discs per pelt sample. This value was combined with the mean number of bundles per unit area to determine the average fur density for each pelt sample.

Character Evolution

To examine character evolution of fur traits, data from the present study were combined with data from the literature. For each species, all available data were averaged and that mean was used to represent the character state for the species.

Evolutionary analyses were conducted with Mesquite 2.01 (Maddison & Maddison,

2007), using a phylogenetic tree constructed from a molecular-based phylogeny for the Pinnipedia (Higdon et al., 2007), with the polar bear (Ursus maritimus) as an outgroup. As branch lengths were not available, all branch lengths were set to 1.0 for 18 the entire tree, to preserve the complete evolutionary relationships. Trees were subsequently pruned to species for which data were present, and pruning was performed separately for each character examined.

Effects of hydrostatic pressure

To examine the effects of hydrostatic pressure on the integrity of the insulation, a sample was isolated from the original sculp for a subset of pinniped species (Table 1). Each sample was cut to exactly 4.0 cm laterally and 5.2 cm dorso- ventrally to fit the dimensions of the experimental chamber. As I determined no detectable changes in blubber under pressure up to 70 m simulated depth, the underlying blubber was removed from sculp samples to facilitate adhesion to the container insert. The fur was cleaned using cold running water, and the air layer restored to the fur using a hairdryer as detailed in Williams et al. (1988) and

Kvadsheim & Aarseth (2002). The pelt sample was then fit into a clear plastic container insert (4.0 cm long × 5.2 cm wide × 10.0 cm high) and the skin adhered to the bottom with silicon adhesive. Dry fur thickness was measured for both guard hairs and underhairs to the nearest 0.01 mm with digital calipers (Mitutoyo) three times on each side, and the mean values were used for calculations. Water was then carefully poured onto the sample, and wet fur thickness was measured for guard hairs and underhairs using the same method.

The plastic container with the immersed sample was placed into a hyperbaric chamber (Trident Systems Inc., TS3) with a viewing window, such that the side of the 19 container was flush against the viewing window. Dives to 70 m (the maximum capability of the chamber) were simulated once on each lateral side of the sample, with the fur dried and air layer restored between trials. During each simulated dive, height of the guard hairs, underhairs, and water level were marked every 10 m for both descent and ascent, on transparent tape adhered to the viewing window. These marks were subsequently measured to the nearest 0.01 mm with the digital calipers

(Mitutoyo). Reductions in water level were attributed to loss of air from the pelt during the dive. To detect any air remaining in the fur after a dive, entrapped air was forcibly removed from the fur with a dissecting probe, and the resulting change in water level (if any) was measured.

Blubber characteristics

Blubber thickness and water content

The thermal properties of blubber change with a variety of parameters, including thickness, water and lipid content, and fatty acid composition. Blubber thickness for each sample was determined by averaging 12 measurements (3 on each perimeter side) to the nearest 0.01 mm, using digital calipers (Mitutoyo). Water content was determined from a 1 cm × 1 cm subsample of the full vertical blubber depth. This subsample was weighed to the nearest 0.001 g using a digital scale

(PR2003 DeltaRange, Mettler-Toledo, Inc., Columbus, OH, USA), placed in a freeze dryer for a minimum of 24 hours, and then re-weighed. Samples were re-weighed again the following day to ensure that the mass of each sample was stable (± 0.005 g). 20 Reduction in mass was attributed to water loss and calculated as a percentage of the original mass of the subsample.

Lipid content and fatty acid analysis

Lipid content and fatty acid composition were determined for a subset of blubber samples (N=35; see Table 1 for species), beginning with 1 cm × 1 cm subsamples of the full vertical blubber depth. For fatty acid composition, the outer region (top portion) of the blubber, immediately below the epidermis, was analyzed separately from the inner region of the blubber, next to the underlying skeletal muscle. Lipids were extracted from the subsamples using a modified Folch, Lees &

Sloane-Stanley (1957) procedure (Budge, Iverson & Koopman, 2006). Briefly, lipids were extracted with 2:1 chloroform: methanol, washed with salt solution, dried over anhydrous sodium sulphate, and evaporated under nitrogen. The extracted lipid was weighed and used to calculate lipid content as a percentage of the original subsample mass.

To analyze fatty acid composition, fatty acid methyl ethers (FAME) were prepared from the extracted lipid, using H2SO4 in methanol (Budge et al., 2006).

Duplicate analyses of FAME were performed using temperature-programmed gas liquid chromatography according to Budge & Iverson (2003). Samples were analyzed on a Perkin Elmer Autosystem II Capillary gas chromatograph with a flame ionization detector using a flexible fused silica column (30 m × 0.25 mm ID) coated with 50% cyanopropyl polysiloxane (0.25 µm film thickness, Agilent Technologies, 21 DB-23; Palo Alto, CA, USA). Helium was used as the carrier gas and the gas line was equipped with an oxygen scrubber. Results were calibrated according to Budge

& Iverson (2003). Briefly, individual peaks associated with each FAME were compared to a series of response factors calculated from FAME standards (Nu-Chek

Prep, Elysian, MN). Up to 69 FAME were identified according to Iverson, Frost &

Lowry (1997). Of these, 49 FAME were used for analysis based on the following criteria: an average presence among all samples of at least 0.1%, and a maximum value among all samples of at least 0.3%. FAME were described using the shorthand nomenclature of A:Bn-X, where A represents the number of carbon atoms, B the number of double bonds, and X the position of the double bond closest to the terminal methyl group.

To examine the extent to which the blubber may have been modified by endogenous processes, the Δ9 desaturation index (Δ9-DI) was calculated. Δ9-DI is the ratio of potentially endogenous monounsaturated fatty acids (MUFA) to the corresponding saturated fatty acids (SFA) from which they could have originated

(Käkelä & Hyvärinen, 1996). Δ9-DI was calculated according to the formula:

Δ9-DI = [(wt% 14:1n-5) + (wt% 16:1n-7) + (wt% 16:1n-9) + (wt% 18:1n-9)

+ (wt% 18:1n-7)] / [(wt% 14:0) + (wt% 16:0) + (wt% 18:0)] (1)

where wt% represents the percentage by weight of the indicated FAME.

22 Thermal conductivity

Thermal conductivity was measured for squares (approx. 10 cm × 10 cm) trimmed from sculp or pelt samples, using the standard material method (Kvadsheim,

Folkow & Blix, 1994; Kvadsheim & Aarseth, 2002; Dunkin et al., 2005). The fur was cleaned using cold running water, and then the air layer was restored to the fur using a hairdryer on the cool setting (Williams et al., 1988; Kvadsheim & Aarseth,

2002). Blubber, skin, and dry fur thickness were measured to the nearest 0.01 mm with digital calipers (Mitutoyo) three times on each side, and the mean values were used for calculations.

Measurements were conducted in a heat flux chamber (162 quart Igloo Marine ice chest, Igloo Commercial, Katy, TX, USA) with a lower, highly insulated compartment and an upper, chilled compartment modeled after Dunkin et al. (2005)

(Fig. 1). The insulated compartment contained the heat source, a sealed aluminum box through which heated water (35 °C) was circulated from a constant-temperature water bath (Lauda RM20, Brinkmann Instruments, Toronto, Ontario, Canada). The upper chamber was cooled with ice packs to create a steady thermal gradient.

An elastomer (Plastisol vinyl, Carolina Biological Supply, Burlington, NC,

USA) was used as the standard material (k = 0.124 ± 0.008 W·m-1·°C-1). The standard material was placed flush against the heat source, and the blubber or skin of the sample was placed in series with the standard, so that the fur was exposed to the cold air. The standard material and sample were surrounded by insulation to ensure unidirectional heat flow through the materials (Fig. 1). 23 Temperatures were measured using copper-constantin (Type T) thermocouples (Physitemp Instruments, Inc., Clifton, NJ, USA) placed between the surface of the heat source and the standard material (probes 1-3), between the standard material and the blubber (probes 4-6), between the blubber and the skin

(probes 7-9), and on top of the fur (probes 10-12; Fig. 1). In addition, two thermocouples (probes 13-14) were placed at the base of the fur. All thermocouples were wired to a Fluke Hydra data logger (model 2625A, Fluke Inc., Everett, WA,

USA), which recorded the outputs every 6 sec onto a laptop computer. Trials lasted a minimum of 2 h to ensure that the apparatus reached steady state, and data were analyzed for the final 30 min of each trial.

Thermal conductivity was calculated across each layer (blubber, full pelt, and fur) and across the entire sample, using the Fourier equation (Kreith, 1958):

H = k · A · ΔT · L-1 (2)

where H is heat transfer in J·s-1, k is thermal conductivity in W·m-1·°C-1, A is the area

(in m2) through which the heat is moving, ΔT is the temperature differential (in °C) across the material, and L is the thickness of the material in m. Assuming that heat transfer is equal across both the standard material and the sample, the equations for both materials can be set equal and solved for the thermal conductivity of the sample.

24 Statistical Analyses

Numerical values for all data are presented as means  1SD. The relationship between each characteristic and body mass was examined on a log-log scale, and any covariance taken into account prior to further statistical analysis. As suggested by

Harvey & Pagel (1991), a nested ANOVA was used to incorporate phylogenetic influences into statistical comparisons. Statistical significances among means were determined for families using a composite phylogeny (Flynn et al., 2005; Higdon et al., 2007; Fig. 2), and species nested in families, using nested ANOVA and the Tukey

Honestly Significant Difference test (JMP Software, SAS Institute, Cary, NC).

ANCOVA was used to incorporate the relative influences of body mass, habitat, and latitude on total insulation (Systat Software, Inc., Chicago, IL). Where unspecified, statistical significance corresponds to P<0.05.

To examine differences in fatty acid (FA) composition, Principal Components

Analysis (PCA) was performed on proportional FA data that were transformed using the following logarithmic function (Aitchison, 1983): xt = log (xi/g(x)), where xi is a given FA expressed as a percentage of the total for that sample, g(x) is the geometric mean of the FA data for all samples, and xt represents the transformed FA data. The resulting principal component factors were analyzed with Discriminant Function

Analysis (DFA), using a stepwise function (Systat Software, Inc., Chicago, IL). PCA and DFA were performed separately for fur seals, sea lions, and phocids.

25 Results

Fur Characteristics

Shape of the hair cuticle

Shape of the hair cuticle was visualized for 8 terrestrial, 2 semi-aquatic, and

14 marine carnivore species (Table 1, Fig. 3). Terrestrial species showed an irregular cuticular scale pattern on both guard hairs (Fig. 3A) and underhairs (Fig. 3B), with few exceptions. Unlike the other terrestrial species examined, two felid species showed distinctive cuticular scaling. The domestic cat (Felis catus) had elongated cuticular scales on both guard hairs and underhairs, while the bobcat (Lynx rufus) had elongated and pointed scales on the underhairs. However, the strictly terrestrial families including the felids had significantly shorter cuticular scales (F7,16=425.4173,

P<0.001) compared to the mustelids and otariids (Fig. 4).

Regardless of aquatic or terrestrial lifestyles, all of the mustelids examined maintained regular cuticular scaling patterns with elongated scales (Fig. 4). Thus, the primarily terrestrial ermine (Mustela erminea) showed extremely regular, elongated scale patterning on both guard hairs (Fig. 3A) and underhairs (Fig. 3B). This was also evident for the river otter (Lontra canadensis) and sea otter (Enhydra lutris).

Similar to the otters, all five fur seal species examined showed regular, elongated scalar patterning on both guard hairs and underhairs (Fig. 3, 4). In contrast, all three sea lion species showed irregular and shortened cuticular scale patterns on both guard hairs and underhairs, which were more characteristic of terrestrial carnivores. The five phocid seal species also demonstrated irregular and shortened 26 cuticular scale patterning on guard hairs and underhairs, as well as a reduction in the prominence of the scales (Fig. 3). In addition, the polar bear (Ursus maritimus) demonstrated a marked reduction in the prominence of cuticular scales on both guard hairs and underhairs (Fig. 3).

Hair lengths and diameters

Circularity of the hair was calculated as the ratio of the minimum diameter to the maximum diameter, and did not correlate consistently with body mass. Guard hairs were significantly flatter (F24,84=84.697, P<0.001) in the aquatic groups compared to the terrestrial species, with some exceptions (Fig. 5). Both pinniped families (otariids and phocids, N=12 species) had significantly shorter guard hairs compared to felids (N=3), canids (N=3), and polar bear (N=1). The raccoon

(Procyon lotor, N=1), skunk (Mephitis mephitis, N=1), and mustelids (N=3) demonstrated intermediate circularities. There were no significant differences in guard hair circularity within families, except for the mustelids. Both the river otter and sea otter had significantly flatter guard hairs compared to the primarily terrestrial ermine (P<0.050).

Underhair circularity showed significant differences among families

(F23,73=6.464, P<0.001), but the pattern was not as marked as that for guard hairs (Fig.

5). Underhairs were significantly flatter in the otariids (N=7 species) compared to felids (N=3), canids (N=3), mustelids (N=3) and phocids (N=5). The circularity of raccoon (N=1), skunk (N=1), and polar bear (N=1) underhairs did not differ 27 significantly from that of any other group. Significant differences in underhair circularity were not detected among species within families.

Hair length did not correlate consistently with body mass. Hair length differed significantly among families (F37,102=10.847, P<0.001), with longer hairs in terrestrial species (Fig. 6). Guard hairs were significantly shorter in otariids (N=10 species), phocids (N=13), and walrus (N=1) compared to felids (N=3), canids (N=3), polar bear (N=1), raccoon (N=1), and skunk (N=1). Mustelids (N=4) also had significantly shorter guard hairs compared to felids, canids, and polar bear; however, mustelid guard hair length did not differ significantly from that of raccoon or skunk, members of the same superfamily (Musteloidea). Within the otariids, fur seals (N=6 species) had significantly longer guard hairs than sea lions (N=6 species; P<0.001).

Within the phocid family, members of the subfamily Phocinae (N=7 species) had significantly longer guard hairs compared to members of the subfamily Monachinae

(N=6 species; P=0.011).

Underhair length also demonstrated a significant trend among families, with shorter hairs in aquatic groups (F24,78=9.218, P<0.001; Fig. 6). Underhairs were significantly shorter in otariids (N=8 species) and phocids (N=4) compared to felids

(N=3), canids (N=3), polar bear (N=1), and skunk (N=1). Mustelid (N=3 species) underhairs were significantly shorter than those of felids and canids, but not significantly different from skunk, polar bear, or pinniped underhairs. Underhair length for the racooon did not differ from that of any other group. Underhairs are completely absent in the walrus. Within the otariid family, fur seals (N=5 species) 28 had significantly longer underhairs than sea lions (N=3 species; P<0.001).

Differences among phocine and monachine underhair length could not be tested, as northern elephant seals lack underhairs, and underhair length was not reported for the species examined in Scheffer (1964b).

Fur densities

As with the other fur characteristics, fur density showed no consistent relationship with body mass. Fur densities varied significantly among families

(F31,69=18.566, P<0.001), with greater fur densities in families with species relying primarily on fur for insulation in water (Fig. 7). Accordingly, the mustelids (N=4 species) had significantly greater fur densities than all other groups. The otariids

(N=10) had significantly greater fur densities than the canid (red fox, Vulpes vulpes) and phocids (N=13), but did not differ significantly from the other groups. There was no significant difference in fur density among the felids (N=2), polar bear, or walrus.

There were significant differences in fur density among species within mustelids, otariids, and phocids. Among mustelids, the sea otter had significantly denser fur than the river otter, mink, and ermine (Fig. 7). The semi-aquatic river otter had a fur density significantly greater than that for the primarily terrestrial ermine, while the semi-aquatic mink had a comparably intermediate fur density. Among the otariids, fur seals (N=6 species) had significantly denser fur (431.5 ± 112.9 hairs·mm-1) than sea lions (23.2 ± 11.2 hairs·mm-1; N=4 species), except for the New Zealand fur seal

(Arctocephalus forsteri) and New Zealand sea lion (Phocarctos hookeri), which were 29 not significantly different from any other otariid species (Fig. 8). Note that the lack of significance here is likely due to the reduced statistical power associated with single samples. Among the phocids, phocines (N=7 species) had significantly greater fur densities (25.8 ± 11.9 hairs·mm-1) than monachines (15.0 ± 7.2 hairs·mm-1; N=6 species; P=0.018).

Character evolution

Patterns of character evolution of fur characteristics among the pinnipeds indicated that phocid seal fur exhibited the most derived traits, and sea lion fur exhibited more derived traits than fur seal fur (Fig. 9, 10, 11). While the musteloids

(mustelids, procyonids, and mephitids) are the sister group to the pinnipeds in the phylogeny used for this study (Fig. 2; Flynn et al., 2005), the polar bear was used as an outgroup here because the wide range of fur characteristics among the musteloids obscured the evolutionary patterns among the pinnipeds. From the terrestrial state, guard hairs became flatter in pinnipeds; and were flattened further in phocids compared to otariids (Fig. 9). From the terrestrial state, both guard hairs and underhairs became shorter in pinnipeds; hairs were shortened further in sea lions compared to fur seals, and in walrus and phocids compared to otariids (Fig. 10).

Finally, fur density increased in fur seals relative to the ancestral state, but was reduced in sea lions and reduced further in walrus and phocids (Fig. 11).

30 Effects of hydrostatic pressure

Of the species examined, only fur seal pelts were able to maintain an air layer in the fur upon immersion; the sea lion and phocid pelts were saturated with water when immersed (Fig. 12). During the simulated 70 m dives, 78.3 – 99.6% of the trapped air in the fur seal pelts bubbled out at 20 m during ascent. For fur seals, guard hairs were compressed 5.6 – 28.9% upon immersion, and another 4.3 – 15.1% during the dive; underhairs were compressed 2.9 – 18.9% upon immersion, and another 5.3 – 25.1% during the dive (Fig. 13). Sea lion and phocid guard hairs were compressed 29.1 – 52.3% upon immersion, and compressed 0.0 – 2.1% further during the dive. Underhairs were either not present or not visible for measurement in the sea lion and phocids.

Thermal conductivity

Thermal conductivity of the fur layer alone (probes 10-12 and 13-14, Fig. 1) was significantly lower than thermal conductivity of the full pelt (fur and skin; probes

7-9 and 10-12) for all samples measured in this study (paired t-test, t=2.730,

P=0.009). Because previous studies used full pelts for similar measurements

(Scholander et al., 1950; Hammel, 1955), thermal conductivity of the full pelt was used for comparisons among species. Note that all conductivity values for terrestrial species are from Scholander et al. (1950) and Hammel (1955). Thermal conductivity of the pelt in air varied among families (F21,82=11.715, P<0.001), with higher conductivities in fully aquatic species (Table 2). The thermal conductivities of 31 phocid (N=3 species) and otariid (N=8) pelts were significantly greater than mustelid

(N=3) and canid (N=4) pelts. The thermal conductivities of procyonid (raccoon and kinkajou) pelts were significantly lower than phocid pelts, but not significantly different from any other group.

Blubber characteristics

Blubber thickness

Blubber thickness ranged from 6.41 mm to 63.50 mm among pinniped species. On average, phocid seals and walrus maintained the thickest blubber layers, which were 3 – 5 times thicker than fur seal blubber and 1.5 – 2 times thicker than sea lion blubber (Fig. 7). Overall, fur seals maintained the thinnest blubber layers.

Blubber thickness varied inversely with fur density among pinniped species (Fig. 7).

Blubber thickness generally increased with body mass among pinnipeds, but the response differed among groups. To account for the differential responses among groups, mass-specific blubber thickness was calculated for each individual.

Comparing blubber thickness on a mass-specific basis revealed a different pattern from that observed for absolute blubber thickness. Phocid seals (N=9 species) had significantly higher mass-specific blubber thicknesses (0.54 ± 0.31 mm·kg-1) compared to otariids (0.22 ± 0.11 mm·kg-1; N=7), which in turn had significantly higher mass-specific blubber thicknesses compared to walrus (0.06 ± 0.04 mm·kg-1;

F15,425=40.875, P<0.001). Within the otariids, fur seals (N=4 species) had greater mass-specific blubber thicknesses (0.32 ± 0.22 mm·kg-1) than sea lions (0.22 ± 0.10 32 mm·kg-1; N=3 species; P<0.001). Within the phocids, phocines (N=7 species) had significantly greater mass-specific blubber thicknesses (0.55 ± 0.31 mm·kg-1) than the northern elephant seal, a monachine (0.09 ± 0.04 mm·kg-1; P=0.011).

Lipid and water content

Percent lipid content ranged from 50% to 97% among pinnipeds. Phocid blubber had the highest lipid content compared to the other pinniped groups

(F11,229=29.7469, P<0.001; Fig. 14). Among the phocids, harp (Pagophilus groenlandica) and ringed seal blubber had significantly greater lipid content compared to northern elephant seal (Mirounga angustirostris) blubber, while the other species did not differ significantly. Walrus blubber had a significantly higher lipid content than otariid blubber, and the otariid species did not differ from each other.

Percent water content varied linearly with lipid content (r2 = 0.72, P<0.001), according to the equation:

% H2O = -0.92 (% Lipid) + 76.84% (3)

where % H2O is the percent water content by mass, and % Lipid is the percent lipid content by mass.

33 Thermal conductivity

Thermal conductivity of the blubber was significantly higher for otariids compared to phocids (F1,65=24.997, P<0.001; Table 2). Blubber conductivity did not differ among species within either family. Thermal conductivity varied linearly with lipid content (r2 = 0.40, P<0.001), according to the equation:

k = -0.004 (% Lipid) + 0.512 (4)

where k is thermal conductivity in W·m-1·°C-1, and % Lipid is the percent lipid content by mass (Fig. 15). This relationship is similar to that reported for cetacean blubber (Worthy & Edwards, 1990).

Fatty acid composition

The inner and outer layers of the blubber of each sample were analyzed separately and for fatty acid composition. For fur seals (N=4 species), seven principal component factors were defined, with the first two factors explaining 59.7% of the variance. However, discriminant function analysis revealed that none of these factors was significantly different between the inner and outer layers of fur seal blubber (F7,14=0.0964, P=0.998). For sea lions (N=1 species), five principal component factors were defined, with the first factor alone explaining 57.6% of the variance. Fatty acid composition did differ between the inner and outer layers of sea lion blubber (F1,20=5.076, P=0.0356), with one principal component factor 34 contributing to the difference. This factor revealed that sea lions tended to have more

14:1n-5 and less 18:0 in the outer layer of the blubber compared to the inner layer.

For phocids (N=3 species), 5 principal component factors were defined, with the first factor alone explaining 51.7% of the variance. Fatty acid composition differed significantly between the inner and outer layers of phocid blubber (F2,21=3.627,

P=0.0444), with 2 principal component factors contributing to the difference. This revealed that phocids tended to have more 17:1 and less 20:0 in the outer layer of the blubber compared to the inner layer, and to a lesser extent more 18:1n-7 and less 24:1 in the outer layer.

Δ9 desaturation index (Δ9-DI) was significantly higher for the outer blubber layer compared to the inner layer in phocids (paired t test, t=-6.292, P<0.001) and sea lions (t=-3.876, P=0.003). However, there was no significant difference in Δ9-DI between the blubber layers for fur seals (t=-2.076, P=0.065). Overall, fur seal blubber had a significantly lower Δ9-DI (1.6 ± 0.5) compared to sea lions

(F2,32=12.031, P=0.045) and phocids (P<0.001). Sea lions had a lower Δ9-DI (2.2 ±

0.7) compared to phocids (2.7 ± 0.4), but the difference was not statistically significant (P=0.069).

When fatty acids were grouped by structural type and examined across the full blubber depth, there were significant differences in composition among pinniped groups (Fig. 16). Fur seal blubber had a higher percentage of saturated fatty acids compared to sea lion or phocid blubber (F2,32=11.814, P<0.001). Phocid blubber had a higher percentage of monounsaturated fatty acids (F2,32=14.974, P<0.001) and a 35 lower percentage of polyunsaturated fatty acids (F2,32=13.318, P<0.001) compared to fur seal or sea lion blubber (Fig. 16).

Total insulation

To examine the overall insulation available to each animal, thermal resistance was calculated according to the equation R = L/k, where R is the thermal resistance in m2·°C·W-1, L is the thickness of the insulation in m, and k is the thermal conductivity in W·m-1·°C-1. Thermal resistance was calculated for the pelts of terrestrial carnivores, polar bear, and river otter, and for full sculps of pinniped species.

Thermal resistance varied (F7,110=67.002, P<0.001) according to habitat (terrestrial, semi-aquatic, marine) and latitudinal climate (tropical, temperate, polar; Fig. 17).

The temperate and polar terrestrial species, along with the polar bear, had significantly greater thermal resistances compared to all other groups (P≤0.007). The pinnipeds, river otter, and tropical terrestrial species had comparatively lower thermal resistances, which were not significantly different from each other (P=1.000).

Principal components analysis revealed that body size and habitat strongly covaried, with larger animals in the aquatic habitat. In contrast, latitudinal climate was independent of body size and habitat among the species sampled.

36 Discussion

Modification of fur for aquatic living

Fur functions as an insulator by maintaining a relatively still air layer between the animal’s skin and the surrounding environment, primarily within the dense underfur (Ling, 1970). However, the insulating value of fur is diminished by the presence of water vapor, and wet fur is further compromised as an insulator

(Scholander et al., 1950; Hammel, 1955; Johansen, 1962; Ling, 1970; Frisch,

Ørtisland & Krog, 1974; Morrison, Rosenmann & Estes, 1974; Costa & Kooyman,

1982; Doncaster et al., 1990). In addition, the physical forces of drag during swimming and hydrodynamic pressure during diving alter the position of the fur relative to the body (Fish, 2000). Under water, the trapped air layer is key, as penetration of the fur by water results in a 3-fold increase in heat loss (Williams,

1986; Williams et al., 1988). Because the insulative effectiveness of fur requires the presence of trapped air, the transition to an aquatic lifestyle required several modifications of fur to maintain its insulative function in water.

To protect the air layer trapped by the fine underhairs, guard hairs became flattened in aquatic species, even in the semi-aquatic river otter (Fig. 5). Unlike most terrestrial mammals, sea otters and pinnipeds lack arrector pili muscles in their hair follicles, and therefore have little physiological control over the positioning of the hairs (Montagna & Harrison, 1957; Scheffer, 1962; Ling, 1965; Ling 1970). When the animal is submerged, this feature increases the extent to which the hairs lie flat against the body (Fig. 13), and the overlapping arrangement of guard hairs protects 37 the underlying air layer by preventing penetration by water. The flat shape of the guard hair facilitates both laying flush when the animal is submerged and natural lifting as the fur dries (Ling, 1970; Mostman-Liwanag, pers. obs.), enhancing its function in both air and water without the need for piloerector muscles.

The flattening of underhairs was not as marked as that observed for guard hairs (Fig. 5). The underhairs are comparatively fine and malleable, and are not typically exposed to the surrounding water. However, the underhairs of pinnipeds were significantly flatter than those of most terrestrial species, demonstrating a consistent pattern for both hair types. The shortening of both guard hairs and underhairs (Fig. 6) likely facilitates streamlining, thus providing an indirect thermal benefit by moving water past the animal. Hence, the flattening and shortening of the hairs in aquatic species not only provides a waterproof barrier to protect the insulating air layer, but also facilitates movement through the water by reducing friction drag

(Noback, 1951; Ling, 1970; Fish, 2000).

A key characteristic associated with preventing the penetration of water into the fur is the ability of underhairs to interlock with each other. As previously described for river otters and sea otters (Williams et al., 1992; Weisel, Nagaswami &

Peterson, 2005), both species demonstrated characteristic geometric scale patterning on the underhairs (Fig. 3B), which enables the underhairs to interlock and more efficiently maintain an air layer even while submerged. Both the river otter and sea otter also showed regular, elongated scale patterning on the guard hairs (Fig. 3A), which may increase the tendency of the hairs to overlap. Interestingly, these same 38 patterns were observed for guard hairs and underhairs in the ermine, a primarily terrestrial mustelid (Fig. 3), and were previously described for the guard hairs of the western polecat (Meyer et al., 2000). The ermine often burrows in snow and does have the ability to swim long distances (Taylor & Tilley, 1984), which could have provided selective pressure for the development of interlocking scales; however, the existence of such patterning on the hairs of the strictly terrestrial polecat suggests a potential exaptation of hairs for water colonization among the mustelids.

Fur seals share the same elongated patterning on both guard hairs and underhairs (Fig. 3, 4), consistent with their use of fur to trap air and insulate against the aquatic environment (Fig. 12). There is still debate as to whether mustelids or ursids are the closest outgroup to the pinnipeds (Berta et al., 2006), so it is unclear whether this shared characteristic represents convergent evolution or retention of a shared common ancestry. However, both sea lions and phocids have clearly lost this patterning secondarily (Fig. 3, 4), which provides morphological evidence that both groups no longer utilize the fur for insulation in water. This was further confirmed by the lack of trapped air in immersed sea lion and phocid pelts compared to fur seal pelts (Fig. 12).

Patterns of fur density also indicate initial changes associated with aquatic living that are reduced or lost in the later diverging species (Fig. 7, 11). Significant differences in fur density were observed within the mustelid family alone (ermine, mink, river otter, and sea otter), such that more aquatic species showed concomitant increases in fur density (Fig. 7). As pelage units, or bundles, are typically spaced 39 evenly across the body surface (Scheffer, 1964b), increases in fur density can be attributed to increases in the number of underhairs per bundle. Accordingly, the secondary reduction in fur density observed in sea lions, phocids, and walrus is associated primarily with a reduction or complete loss of underhairs (Scheffer, 1964b; present study). Interestingly, the spacing of pelage units is no longer uniform in the walrus and most phocid species (Scheffer, 1964b), which may contribute further to a decline in fur density in those species. Both decreases in underfur and non-uniform spacing of pelage units are consistent with the loss of thermal function for the fur in water.

The loss of thermal function in water for sea lion fur represents a clear case of convergent evolution with phocids. While the exact evolutionary relationship among otariid species (fur seals and sea lions) is subject to continued debate (Berta et al.,

2006), both morphological and molecular studies agree that the northern fur seal

(Callorhinus ursinus) is the earliest diverging species of the extant otariids (Berta &

Deméré, 1986; Berta & Wyss, 1994; Lento et al., 1995, 1997; Wynen et al., 2001;

Deméré, Berta & Adam, 2003; Higdon et al., 2007). The placement of a fur seal at the basal position suggests that dense fur was an ancestral characteristic for the otariid family, and that sea lions must have secondarily lost this characteristic (Fig. 8). This may represent a single character change from fur seal to sea lion, as indicated by parsimony analysis of morphological characters (e.g., Berta & Deméré, 1986; Berta

& Wyss, 1994; Deméré et al., 2003), or may be a case of multiple evolutionary transitions, as suggested by the paraphyletic relationships of sea lions and fur seals in 40 recent molecular phylogenies (e.g., Wynen et al., 2001; Higdon et al., 2007). Total evidence analysis, which accounts for both morphology and molecular data, indicates that fur seals and sea lions are paraphyletic groupings (Flynn et al., 2005). In either scenario, the morphological changes observed for sea lion and phocid fur – irregularity of cuticular scales and reduction in fur density – provide further evidence of convergent evolution associated with aquatic specialization (Fig. 3, Fig. 7).

The higher thermal conductivity of the pelt in pinnipeds compared to terrestrial carnivores suggests some loss of functionality in air associated with the morphological adaptations of the fur to the aquatic environment (Table 2, Fig. 17).

While these values were measured in air, the thermal conductivity of the immersed pelt would only increase relative to its conductivity value in air (Scholander et al.,

1950; Williams, 1986; Williams et al., 1988). Thus, pinnipeds with fur that no longer functions in water must rely instead on an alternative form of insulation.

The evolutionary transition to blubber

Subcutaneous fat in mammals is an energy storage tissue that may also serve other functions (Pond, 1978). For aquatic mammals with a blubber layer, the fat acts not only as an energy store but also as an insulator (Scholander et al., 1950), and can contribute to body streamlining as well as buoyancy control (Pabst et al., 1999; Fish,

2000). Blubber is a continuous, subcutaneous layer of adipose tissue, reinforced by collagen and elastic fibers (Parry, 1949; Ling, 1974; Ackman et al., 1975; Lockyer,

McConnell & Waters, 1984, 1985; Pabst et al., 1999). The blubber layer serves as 41 the primary form of insulation for fully aquatic mammals, and evolved independently at least three times among mammals: in cetaceans, sirenians, and pinnipeds. Blubber maintains a large thermal gradient between the animal’s core and body surface, but requires a relatively large thickness (compared to fur) in order to achieve that gradient

(Scholander et al., 1950; Costa & Kooyman, 1982; Worthy, 1991).

Because of the large thickness required to establish a comparable thermal gradient to fur, a larger body size is necessary to maintain maneuverability in blubbered species. Thus, the transition to blubber as the primary insulating layer occurs in the larger-bodied otariids, the sea lions (Fig. 18). Quality and quantity of blubber can change seasonally (Ryg, Smith & Ørtisland, 1988; Aguilar & Borrell,

1990; Worthy & Edwards, 1990; Samuel & Worthy, 2004; Budge et al., 2008) as well as with animal condition (Aguilar & Borrell, 1990; Dunkin et al., 2005). Despite this variation, I was able to observe consistent patterns in blubber characteristics among pinniped groups. While fur seals had thinner blubber than sea lions on average (Fig.

7), the quantity of blubber was greater for fur seals on a mass-specific basis, consistent with previous reports of the relationship between lipid mass and total body mass in otariids (Costa, 1991). The comparatively large quantity of walrus blubber

(Fig. 7) can be attributed primarily to the extremely large body size (over 1000 kg, on average) exhibited by this species (Fay, 1981), such that the differences between walrus and otariid blubber thickness disappeared or rather reversed when corrected for mass. Phocid seals, however, had a larger quantity of blubber compared to otariids on both an absolute and mass-specific basis. Yet thickness alone does not 42 determine thermal characteristics, as changes in quality can produce differences in thermal conductivity of up to 400% (Pabst et al., 1999; Worthy & Edwards, 1990;

Dunkin et al., 2005).

In terms of quality, measurements for the conductivity of animal fat have generally fallen into the range of 0.1 – 0.21 W·m-1·°C-1, with reported values up to

0.28 W·m-1·°C-1 (reviewed in Dunkin et al., 2005). Interestingly, the average thermal conductivity of otariid blubber (0.280 ± 0.069 W·m-1·°C-1, range = 0.161 – 0.449) fell at the high end of the range of known values, though the ranges do overlap considerably. The similar quality of blubber among the otariid species, as measured by both lipid content and thermal conductivity, is consistent with the occurrence of rapid evolution within this group (Higdon et al., 2007).

Phocid blubber appears to be greater in both quantity and quality. Not only was phocid blubber of a greater quantity beyond that associated with large body size, but phocid blubber also had a significantly lower thermal conductivity and higher lipid content compared to otariid and walrus blubber (Table 2, Fig. 14). As the insulating quality of blubber is directly correlated to its lipid content (Fig. 15; Worthy

& Edwards, 1990), this indicates better quality blubber in the later diverging pinniped group. The otariids represent extant examples of the evolutionary transition from fur to blubber among aquatic mammals, and the lower quality of otariid blubber may be a constraint related to the transitional state of this group. Within the otariids, fur seals generally showed a greater variation in lipid content and thus thermal conductivity of the blubber (Fig. 15); this may reflect the use of the blubber for energy storage rather 43 than for insulation. Sea lions appear to compensate for a lack of waterproof fur with a larger body size and associated thicker blubber layer (Fig. 7, 18). Within the phocids, the monachine seals appeared to utilize a similar strategy, with larger body sizes, thicker blubber, and shorter and less dense fur compared to phocines.

Vertical stratification of blubber has been well documented in cetaceans

(Ackman, Eaton & Jangaard, 1965; Ackman, Epstein & Eaton, 1971; Ackman et al.,

1975; Aguilar & Borrell, 1990; Koopman, Iverson & Gaskin, 1996; Hooker et al.,

2001; Samuel & Worthy, 2004; Reynolds, Wetzel & O’Hara, 2006; Budge et al.,

2008), and it has been suggested that the outer and inner layers of the blubber serve different functions (Aguilar & Borrell, 1990). I hypothesized that pinniped species relying more on blubber than fur for insulation would demonstrate stratification of the blubber associated with two different uses: the outer layer of the blubber would function more for insulation, while the inner layer would serve as an energy store.

The differing fatty acid compositions I observed between the blubber layers in phocids and sea lions, but not in fur seals, support this hypothesis. While the minimal blubber layer in fur seals may limit the extent to which stratification occurs, the homogenous nature of the blubber layer combined with significantly greater fur densities (Fig. 7) suggest that fur seals rely on blubber as an energy store and rely more on their fur for thermoregulation in water.

This hypothesis was further supported by the differences in structural groups among the blubber of pinnipeds. In general, unsaturated FAs tend to have lower melting points than saturated FAs, and thus tend to remain more fluid at cold 44 temperatures (Hilditch & Williams, 1964; Hochachka & Somero, 2002). Because of this, unsaturated FAs are generally used for thermoregulation, as the outer layers in particular would be expected to sustain colder temperatures resulting from the thermal gradient across the insulating layer. Phocids and sea lions had significantly higher levels of unsaturated FAs compared to fur seals (Fig. 16), suggesting that fur seal blubber is less suited for colder temperatures and thus is not as effective for the maintenance of a thermal gradient. As homeoviscous theory indicates that changes associated with lipid fluidity may be induced by exposure to certain temperatures

(Hochachka & Somero, 2002), these results are also consistent with fur seal blubber being maintained at warmer temperatures because the fur provides the primary thermal gradient between the subcutaneous tissues and the surrounding environment.

Further evidence was provided by the differences in Δ9 desaturation index

(Δ9-DI) among the groups. Δ9-DI is an indicator of the extent to which the animal has modified saturated fatty acids (SFA) to endogenously form monounsaturated fatty acids (MUFA). In sea lions, Δ9-DI increased from 2.0 in the inner blubber to 2.4 in the outer layer; for phocids, Δ9-DI increased from 2.3 in the inner blubber to 3.1 in the outer layer. The higher Δ9-DI in the outer blubber layer is consistent with the use of the outer layer for thermoregulation, as this layer will experience colder temperatures when generating a thermal gradient across the tissue (Käkelä &

Hyvärinen, 1996). While many unsaturated FAs, and polyunsaturated fatty acids

(PUFA) in particular, are derived largely from the diet of pinnipeds, Δ9-DI is calculated from MUFA that mammals can form endogenously (Käkelä & Hyvärinen, 45 1996; Budge et al., 2006). Thus, higher levels of these FAs in the outer blubber suggest that in sea lions and phocids endogenously derived MUFA may be preferentially deposited into the outer blubber layer to maintain its fluidity while it serves as an insulator. In contrast, fur seals, like terrestrial carnivores (Käkelä &

Hyvärinen, 1996), rely on their fur to create a thermal gradient; this allows fur seals to maintain the blubber layer at warmer temperatures and does not require the endogenous addition of MUFA to maintain blubber fluidity. Interestingly, FA composition did not correlate significantly with thermal conductivity in any way.

This suggests that thermal conductivity is dependent primarily on the amount of fat present (Fig. 15) rather than the type of fat, and that changes in FA composition are driven by diet and the need to maintain tissue fluidity.

Ecological and evolutionary implications

The extent to which fur or blubber is utilized as insulation in aquatic mammals appears to reflect the level of evolutionary adaptation for the aquatic environment. The most recently diverged groups (polar bears and sea otters) have retained fur as the primary insulator, the earliest diverging groups (cetaceans and sirenians) have secondarily lost the fur and now use blubber, and the intermediate groups (pinnipeds) have both retained fur and developed blubber. The evolutionary transition to an aquatic lifestyle was accompanied first by a modification of the ancestral fur state to maintain thermal resistance in water, followed by a gradual replacement of the fur with an internal blubber layer (Fig. 7). Certain morphological 46 changes to the fur associated with aquatic living, including flattening of the hair shaft and shortening of the hairs, seem to reduce the insulating value of the fur in air (Table

2, Fig. 17). For example, the polar bear has fur with more ancestral (terrestrial-like) characteristics (Fig. 3, 4, 5, 6), and this was reflected in the high insulating value of its pelt (Fig. 17). In contrast, the river otter’s fur possesses more derived morphological characteristics, resulting in a reduction in the insulating value of the pelt (Fig. 17).

The selective pressures of the aquatic environment on the evolution of mammalian insulation were likely quite strong, as the total insulation among carnivores varied strongly according to habitat (Fig. 17). Terrestrial carnivores generally showed greater insulating values compared to pinnipeds, suggesting that blubber is a poorer insulator compared to fur. Within groups, the influence of climate was also apparent, as polar and temperate species generally had higher insulating values compared to tropical species (Fig. 17). In addition, phocid species demonstrated higher insulating values compared to otariids, reflecting the greater quantity and quality of blubber in the later divergent pinniped species (Fig. 7, 14;

Table 2). The more developed insulation may also help explain the greater diversity of phocids in polar climates (Berta et al., 2006).

Body size also plays an important role among pinnipeds, as walrus and sea lions appear to rely on the increased blubber thickness associated with large body size to increase their levels of insulation. Large body size alone confers a thermal advantage in cold environments, by reducing surface area to volume ratio and thus 47 relative heat loss to the surroundings. However, it appears that large body size also facilitates the evolutionary switch from fur to blubber as the primary insulator. This changeover is evident within the otariid family, as the generally smaller-bodied fur seals rely on their fur for insulation, while the larger-bodied sea lions have made the transition to the use of blubber as the primary insulator (Fig. 18). Because a substantial blubber thickness is required to achieve a similar thermal gradient to that provided by fur, smaller-bodied marine mammals may be constrained to the retention of fur as the primary insulator. A disproportionately large blubber layer would severely limit the flexibility and maneuverability of the animal, both on land and in the water. Thus, the use of blubber as the primary insulating layer requires a larger body size to allow for greater blubber thicknesses without impeding the animal’s movement.

As blubber is a relatively poor insulator compared to fur (Fig. 17; Scholander et al., 1950), the selective pressures driving the evolutionary transition to blubber may not be due to blubber’s thermal properties. Instead, the hydrodynamic pressures associated with aquatic living and the ability to use blubber as an energy store may have selected for the convergent evolution of blubber across mammalian lineages.

Blubber allows for body streamlining, hence reducing friction drag in the viscous aquatic medium; its insulative properties are not changed with submergence in water; and it is not compressed appreciably by hydrostatic pressure during diving, thus providing a stable source of both insulation and buoyancy to a diving mammal (Pabst et al., 1999; Fish, 2000; present study). In addition, blubber acts as energy storage 48 that can be utilized during fasting periods and has enabled the separation of breeding and feeding in many species (e.g., Costa, 1991). Fur does not confer these advantages to an aquatic mammal, and in fact is more likely to act as a liability when it comes to hydrodynamics and grooming costs. In particular, the compression and loss of the trapped air layer within the fur during a dive will reduce both the insulation and the buoyancy of the animal (Fig. 12, 13; Fish et al., 2002).

Despite the disadvantages, fur is the superior insulator, as long as the air layer is maintained among the hairs. For small-bodied endotherms in the marine environment, fur may be the only option, as the amount of blubber needed to achieve a similar level of insulation would comprise a disproportionate fraction of the animal’s body size and thus inhibit maneuverability. However, both the use of fur as the primary insulator and small body size limit the diving capabilities for these animals. Larger body sizes facilitate the development of a blubber layer, which is advantageous for swimming and diving. Thus, the most effective insulatory mechanism depends on body size as well as habitat, with larger body size and a blubber layer comprising the best configuration for a fully aquatic lifestyle.

Acknowledgements

I wish to thank J. Arnould, M. Ben-David, T. Bernot, D. Casper, D. Costa, M.

Goebel, T. Goldstein, J. Gomez, M. Gray, D. Guertin, B. McDonald, L. Leppert, B.

Long, M. Miller, L. Polasek, P. Robinson, S. Simmons, C. Stephens, P. Tuomi, D.

Verrier, S. Villegas, E. Wheeler, and especially J. Burns and P. Morris for their 49 assistance with obtaining samples for this project. I would also like to thank M.

Abney, P. Dal Ferro, J. Devlahovich, R. Dunkin, L. Fox, R. Franks, J. Krupp, L.

Liwanag, R. Ludwig, T. Lynn, L. Mandell, C. McCord, J. Oraze, I. Parker, R. Patzelt,

S. Perez, D. Potts, A. Roth, M. Ruane, J. Soper, N. Teutschel, J. Webb, and L. Yeates for laboratory assistance, as well as C. Kuhn and J. Maresh for thoughtful feedback on the manuscript. I would also like to express gratitude to Annalisa Berta, Dan

Costa, and Terrie Williams for their mentorship and support throughout the project.

This study was funded by the Friends of Long Marine Laboratory, SEASPACE, the

Janice A. Nowell Memorial Fund, the American Museum of Natural History Lerner-

Gray Fund, and the UCSC Ecology and Evolutionary Biology department. All animal use protocols, including sample collection, were evaluated and approved by the University of California Santa Cruz Institutional Animal Care and Use

Committee.

50

Table 1. Hair, pelt, and full sculp samples analyzed in the present study. Total number of samples (N) is indicated for each species, with sample types indicated by superscripts: A = full sculp; B = fresh pelt; C = tanned pelt; D = hairs removed from live animal. Analyses performed on at least one sample are indicated by X’s in the appropriate columns.

SEM = scanning electron microscopy; k = thermal conductivity; Press. = hydrostatic pressure experiments; H2O = % water content; Lipid = % lipid content; FA = fatty acid analysis.

Common Fur Characteristics Blubber Characteristics Species N Name SEM Length Diameter Density k Press. Thickness H2O Lipid FA k Canis Domestic 3D X X X familiaris dog Canis Coyote 1B X X X latrans Felis Domestic 1D X X X catus cat Felis Mountain B 4 X X X X concolor lion Felis Bobcat 2B X X X X rufus Mephitis Terrestrial Skunk 1B X X mephitis Mustela Ermine 1C X X X X erminea Procyon Raccoon 1B X X X lotor Vulpes Red fox 7B X X X X

vulpes

-

Lontra River otter 3B X X X X X canadensis

Ursus C tional

Transi Polar bear 1 X X X maritimus

Enhydra Sea otter 1C X X X

lutris Sea otter

Table 1 (continued)

Common Fur Characteristics Blubber Characteristics Species N Name SEM Length Diameter Density k Press. Thickness H2O Lipid FA k Arctocephalus Antarctic fur 2A X X X X X X X X X X X gazelle seal Arctocephalus A

pusillus Cape fur seal 6 X X X X X X X X X X X pusillus Arctocephalus Guadalupe 1A X X X X X X X X townsendi fur seal Fur seals Arctocephalus Subantarctic 4A X X X X X X X X X X X tropicalis fur seal Callorhinus Northern fur 5A,C X X X X X X X X X X X ursinus seal Eumetopias Steller A,D

3 X X X X X X jubatus sea lion Zalophus California 32A X X X X X X X X X X X californianus sea lion

Sea lions Zalophus Galapagos 1A X X X X X X X X wollebaeki sea lion Cystophora Hooded seal 15A,C X X X X X X cristata

Pagophilus Harp seal 24A,C X X X X X X groenlandica Phoca Harbor seal 4A,C X X X X X X X X X X vitulina Pusa A Phocidseals Ringed seal 1 X X X X X X X X X X hispida Mirounga Northern 6A X X X X X X X X X X X angustirostris elephant seal

Table 2. Thermal conductivities of pelts and blubber layers for carnivore groups.

Values are presented as means ± 1SD. Canids include the Arctic fox, red fox, wolf, and domestic dog. Ursids include the grizzly bear and polar bear. Procyonids include the raccoon and kinkajou. Mustelids include the marten, least weasel, and river otter.

Otariids are represented by 5 fur seals and 3 sea lions, and phocids are represented by

3 species.

54

Pelt conductivity Blubber conductivity Group Source (W·m-1·°C-1) (W·m-1·°C-1) Canids 0.048 ± 0.006 N/A Scholander et al. (1950) Ursids 0.069 ± 0.015 N/A Scholander et al. (1950) Procyonids 0.034 ± 0.004 N/A Scholander et al. (1950) Mustelids 0.045 ± 0.030 N/A This study; Scholander et al. (1950) Otariids 0.111 ± 0.034 0.280 ± 0.069 This study Phocids 0.158 ± 0.104 0.192 ± 0.040 This study

55

Figure 1. Conductivity chamber, showing placement of thermocouples (numbered circles) relative to the sample and standard material. Size of thermocouples is greatly exaggerated. Diagonal lines represent the fur layer.

Figure 2. Carnivore phylogeny used for evolutionary comparisons. Groups of interest to this study are labeled in black, while additional groups are labeled in gray for reference. Felidae = cat family; Hyaenidae = hyaena; Viverridae = genets and civets; Ailuridae = panda family; Ursidae = bear family; Mustelidae = weasel family;

Procyonidae = raccoon family; Mephitidae = skunk family; Odobenidae = walrus;

Otariidae = eared seals; Phocidae = true seals; Canidae = dog family. Phylogeny was compiled from Bininda-Emonds, Gittleman & Purvis (1999), Flynn et al. (2005), and

Higdon et al. (2007). Note that branch lengths are not to scale.

58

Felidae

Hyaenidae

Viverridae

Ailuridae Ursidae Ailuridae

Mustelidae

Procyonidae Mephitidae Odobenidae

Otariidae

Phocidae

Canidae

Figure 3. Scanning electron micrographs of representative hairs from terrestrial and marine carnivores, showing the cuticular scale patterns of (A) dorsal guard hairs and (B) dorsal underhairs. Each micrograph is labeled with the common name of the species. Note slightly different magnifications, indicated by the scale bars under each micrograph.

Magnifications were matched as closely as possible, within the limitations of the microscope.

A Coyote Domestic dog Red fox

Mountain lion Bobcat Domestic cat

A Raccoon Skunk Ermine

Polar bear River otter Sea otter

A Northern fur seal Antarctic fur seal Subantarctic fur seal

Cape fur seal Guadalupe fur seal

A California sea lion Galapagos sea lion Steller sea lion

Hooded seal Harp seal Northern elephant seal

A Ringed seal Harbor seal

B Coyote Domestic dog Red fox

Mountain lion Bobcat Domestic cat

B Raccoon Skunk Ermine

Polar bear River otter Sea otter

B Northern fur seal Antarctic fur seal Subantarctic fur seal

Cape fur seal Guadalupe fur seal

B California sea lion Galapagos sea lion Steller sea lion

Hooded seal Harp seal

Underhairs are absent in northern elephant seals.

B Ringed seal Harbor seal

Figure 4. Elongation (maximum length / maximum width) of the cuticular scales of carnivore guard hairs (black bars) and underhairs (white bars). Heights of the bars and lines indicate means and standard deviations for the indicated families or groups.

Letters above the bars indicate statistically significant differences among means, with capital letters for guard hairs and lower case letters for underhairs. Felidae includes the domestic cat, bobcat, and mountain lion; Canidae includes the red fox, coyote, and domestic dog; Ursidae includes the polar bear; Mephitidae includes the striped skunk; Procyonidae includes the raccoon; Mustelidae includes the ermine, river otter, and sea otter; Otariidae includes 4 fur seals and 3 sea lions; and Phocidae includes 5 phocid seals. Data are from the present study only.

71

16 b

14

12

10 B 8 c

6

4 C

Scale elongation (length / width) (length elongation Scale A a A a A a A a A a A a A a 2

0

ae

Felidae

Ursidae

Canidae

Fur Fur seals

Phocidae

Sea lions Sea Mustelidae

Mephitidae Procyonid

Otariidae

72

Figure 5. Circularity (minimum diameter / maximum diameter) of carnivore guard hairs (black bars, top panel) and underhairs (white bars, bottom panel). Heights of the bars and lines indicate means and standard deviations for the indicated families

(no error bars for single samples). Letters above the bars indicate statistically significant differences among means, with capital letters for guard hairs and lower case letters for underhairs. Felidae includes the domestic cat, bobcat, and mountain lion; Canidae includes the red fox, coyote, and domestic dog; Ursidae includes the polar bear; Mephitidae includes the striped skunk; Procyonidae includes the raccoon;

Mustelidae includes the ermine, river otter, and sea otter; Otariidae includes 4 fur seals and 3 sea lions; and Phocidae includes 5 phocid seals. Values for fur seals and sea lions were not significantly different for this character, and are thus presented together in Otariidae. Data are from the present study only.

73 1.0 A A A B

0.8 B

C C

0.6 C D

0.4 D

E Guard hair circularity hair Guard

0.2

0.0

1.0 a a a a a a a b b b b

0.8

0.6

0.4 Underhair circularity Underhair

0.2

0.0

Felidae

Ursidae

Canidae

Otariidae

Phocidae Mustelidae

Mephitidae Procyonidae

74

Figure 6. Lengths of carnivore guard hairs (black bars) and underhairs (white bars).

Heights of the bars and lines indicate means and standard deviations for the indicated families (no error bars for single samples). Letters above the bars indicate statistically significant differences among family means, with capital letters for guard hairs and lower case letters for underhairs. Asterisks indicate significant differences among groups within families. Felidae includes the domestic cat, bobcat, and mountain lion; Canidae includes the red fox, coyote, and domestic dog; Ursidae includes the polar bear; Mephitidae includes the striped skunk; Procyonidae includes the raccoon; Mustelidae includes the ermine, mink, river otter, and sea otter;

Otariidae includes 6 fur seals and 4 sea lions for guard hairs, 5 fur seals and 3 sea lions for underhairs; Phocidae includes 13 species (7 phocines, 6 monachines) for guard hairs and 4 species (3 phocines, 1 monachine) for underhairs; Odobenidae includes the walrus. Data are from the present study, Scheffer (1964b), Frisch et al.

(1974), Hilton & Kutscha (1978), Williams et al. (1992), and Fish et al. (2002). See appendix Table A2 for species-specific values and sources.

75

A a A a b 80

A a A a

60 A a B b B b

B b 40 C c C c C c C

Hair length (mm) length Hair * 20 * * *

0

Felidae

Ursidae

Canidae

Fur Fur seals

Sea lions Sea

Phocinae

Mustelidae

Mephitidae

Monachinae Odobenidae Procyonidae

Otariidae Phocidae

76 Figure 7. Fur density and blubber thickness among carnivores. Heights of the bars indicate means for the indicated species or groups. Whiskers indicate range of species means (no whiskers for single species). Blubber thicknesses are only indicated for pinnipeds. Fur seals: N=6 species for fur density; N=4 species for blubber depth. Sea lions: N=4 species for fur density; N=3 species for blubber depth.

Phocids: N=13 species for fur density; N=9 species for blubber depth. Data for fur density are from the present study, Scheffer (1964b), Sokolov (1962), Kenyon (1969),

Kaszowski, Rust & Shackleford (1970), Tarasoff (1972), Frisch et al. (1974),

Tarasoff (1974), Williams et al. (1992), and Fish et al. (2002). Data for blubber thickness are from the present study, Addison & Smith (1974), West, Burns, &

Modafferi (1979a,b), Stewart & Lavigne (1984), Pitcher (1986), Folkow & Blix

(1987), Wiig (1989), Savelle & Friesen (1996), Cameron et al. (1997), Kleivane et al.

(1997), Krahn et al. (1997), Nilssen et al. (1997), Rosen & Renouf (1997), Addison

& Smith (1998), Hall et al. (1999), Muir et al. (2000), Severinsen, Skaare & Lydersen

(2000), Kucklick et al. (2002), Lydersen et al. (2002), Addison, Ikonomou & Smith

(2005), Willis et al. (2005), Ylitalo et al. (2005), Del Toro et al. (2006), Kucklick et al. (2006), Stapleton et al. (2006), the Central Puget Sound Marine Mammal

Stranding Network, Massey University, the New Zealand Ministry of Fisheries and the New Zealand Department of Conservation.

77

)

2

-

Fur density (hairs · mm · (hairs Fur density

Blubber thickness (mm) thickness Blubber

s

otter otter

Mink

seal

Walrus

Ermine

Bobcat

Red fox Red

Phocids

Sea Sea

Fur Sea lions

River

Polar bear Polar Mountainlion Mustelids Otariids Felids

78

Figure 8. Fur densities among otariid species, with a cladogram illustrating the evolutionary relationships among the species (based on Higdon et al., 2007). Heights of the bars and lines indicate means and standard deviations for the indicated species

(no error bars for single samples). Letters above the bars indicate statistically significant differences among means. Note that fur seals have significantly greater fur densities compared to sea lions, and that the paraphyly of these groups indicates convergent evolution for this trait. Data are from the present study and Scheffer

(1964b).

79 A

600 A A A

A

)

2 - A 400 B

200

Fur density (hairs · mm · (hairs Fur density A B B B B

0

80

Figure 9. Character evolution of guard hair circularity among pinniped species, with the polar bear as an outgroup. Phylogenetic relationships among pinnipeds are based on Higdon et al. (2007). Warmer colors represent more circular guard hairs, while cooler colors represent flatter guard hairs. Colors and character ranges were determined by the Mesquite program. Note that guard hairs are flatter in phocids compared to otariids.

81

Figure 10. Character evolution of (A) guard hair length and (B) underhair length among pinniped species, with the polar bear as an outgroup. Phylogenetic relationships among pinnipeds are based on Higdon et al. (2007). Warmer colors represent longer hairs, while cooler colors represent shorter hairs. Colors and character ranges were determined by the Mesquite program. Note that guard hairs and underhairs (when present) are shorter in walrus, sea lions, and phocids compared to fur seals.

83

A

B

Figure 11. Character evolution of fur density among pinniped species, with the polar bear as an outgroup. Phylogenetic relationships among pinnipeds are based on

Higdon et al. (2007). Warmer colors represent denser fur, while cooler colors represent less dense fur. Colors and character ranges were determined by the

Mesquite program. Note that fur density is greater in otariids compared to walrus and phocids, and greater in fur seal species compared to sea lions.

86

Figure 12. Amount of air trapped in the fur and lost during diving for the pelts of eight pinniped species. Full height of the bars indicates the total height of the air layer trapped in the fur. Gray bars represent the amount of air lost by 20 m depth during a 70 m simulated dive. Black bars represent the amount of air remaining in the fur after a 70 m simulated dive. Pelts of the sea lion and phocid seals were saturated upon immersion.

88

4

3

2

1 Height of air layer in fur (mm) fur in air layer of Height

Ø Ø Ø Ø

0

(pup)

fur seal fur

Harp seal Harp

Hooded seal Hooded

fur seal fur

Cape

ntarctic fur seal fur ntarctic

seal fur Antarctic California sea sea lion California

Suba

Northern Northern elephant seal elephant Northern

89

Figure 13. Height of guard hairs (top panel) and underhairs (bottom panel) for pinniped pelts when dry (black bars), immersed (light gray bars), and under hydrostatic pressure (dark gray bars). Values under pressure represent maximal compression during a simulated 70 m dive. Note that guard hairs are maximally compressed upon immersion for the sea lion and phocid seals. Underhairs (when present) were not visible for the sea lion and phocid seals.

90

Height of guard hairs (mm) hairs guard of Height

Height of underhairs (mm) underhairs of Height

Ø Ø Ø Ø

fur

Cape

ntarctic

furseal furseal

furseal

ant seal ant

sealion

Northern

Antarctic

California

Harpseal

seal(pup)

Northern

Suba

Hoodedseal eleph

91

Figure 14. Percent lipid content (by weight) of pinniped blubber. Heights of the bars and lines indicate means and standard deviations for the indicated groups. Letters above the bars indicate statistically significant differences among means. Otariids include 4 fur seals and 1 sea lion. Phocids include 6 species. Data are from the present study, Addison & Smith (1974), Beck, Smith & Hammill (1993), Kleivane et al. (1997), Krahn et al. (1997), Addison & Smith (1998), Hall et al. (1999), Muir et al. (2000), Severinsen et al. (2000), Kajiwara et al. (2001), Kajiwara et al. (2002),

Kucklick et al. (2002), Lydersen et al. (2002), Kannan et al. (2004), Addison et al.

(2005), Del Toro et al. (2006), Kucklick et al. (2006), and Stapleton et al. (2006).

92

B 100 C

80 A

60

40 Lipid content (%) content Lipid

20

0 Otariid Phocid Walrus

93

Figure 15. Change in thermal conductivity with lipid content of blubber for phocids

(white), sea lions (gray), and fur seals (black). Each point represents an individual sample, with different symbols representing different species within each group. Line represents the best-fit linear regression for all the data, and is described in the text.

The two points with low lipid contents are from poor-condition animals, and serve to demonstrate the increase in thermal conductivity that accompanies a loss of lipid content. Data are from the present study only.

94

0.8

) 0.7

1

-

C



· 1 - 0.6

0.5

0.4

0.3

Thermal conductivity (W·m conductivity Thermal 0.2

0.1 0 20 40 60 80 100

Lipid content (%)

95

Figure 16. Percent composition of fatty acid types in pinniped blubber. Heights of the bars and lines indicate means and standard deviations for the indicated pinniped groups. Asterisks indicate significant differences among groups for each fatty acid type. SFA = saturated fatty acids; MUFA = monounsaturated fatty acids; PUFA = polyunsaturated fatty acids. N=4 species for fur seals, N=1 species for sea lions, and

N=3 species for phocids. Data are from the present study only.

96

80 Fur Seal * Sea Lion

Phocid

60

40 * *

20 Fatty Acid Composition (%) Composition Acid Fatty

0 SFA MUFA PUFA

97 Figure 17. Thermal resistance of the total insulation (in m2·°C·W-1) in relation to body mass (in kg) for carnivores. Symbols and lines indicate species means and standard deviations. Symbol shades indicate habitat, with darker colors for terrestrial species and lighter colors for aquatic and semi-aquatic species. Symbol colors indicate latitude: polar (blue), temperate (orange/yellow), and tropical (red/pink). Symbol shapes indicate groups as follows: terrestrial species are represented by squares, phocids by triangles, otariids by circles, the river otter by the inverted triangle, and the polar bear by the diamond. Polar terrestrial species include the Arctic fox, lynx, marten, and wolf. Temperate terrestrial species include the bobcat, domestic cat, coyote, domestic dog, grizzly bear, raccoon, red fox, and skunk. Tropical terrestrial species include the kinkajou and least weasel. The polar semi- aquatic species is the polar bear. The temperate semi-aquatic species is the river otter. Polar marine species include the Antarctic fur seal and the ringed seal. Temperate marine species include the Cape fur seal, California sea lion, northern elephant seal, Guadalupe fur seal, harbor seal, northern fur seal, Steller sea lion, and subantarctic fur seal. The tropical marine species is the Galápagos sea lion. Data are from the present study, Scholander et al. (1950), and Hammel (1955). Note that terrestrial species (squares) generally show higher insulating values compared to pinnipeds (triangles and circles). The polar bear groups with the terrestrial carnivores, while the river otter groups with the aquatic species. Polar (blue) and temperate species (orange/yellow) generally have higher insulating values compared to tropical species (red/pink). In addition, phocid species (triangles) demonstrate higher insulating values compared to otariids (circles).

98

0.2

0.0

-0.2

-0.4

-0.6

-0.8 Log Thermal Resistance Resistance Thermal Log

-1.0

-1.2 -1 0 1 2 3

Log Body Mass

99

Figure 18. Schematic comparing relative body size and insulating layers of fur seals and sea lions. Fur seals are smaller- bodied, and rely more on their fur for insulation in water. The dashed circle indicates the stratification of blubber in sea lions, which rely more on the outer layer of blubber for thermoregulation.

Fur seal Sea lion

Fur

Blubber

Core Core

References

Ackman, R. G., Eaton, C. A. and Jangaard, P. M. (1965). Lipids of the fin whale (Balaenoptera physalus) from north Atlantic waters. Can. J. Biochem. 43, 1513-1520.

Ackman, R. G., Epstein, S. and Eaton, C. A. (1971). Differences in the fatty acid compositions of blubber fats from northwestern Atlantic fin whales (Balaenoptera physalus) and harp seals (Pagophilus groenlandica). Comp. Biochem. Physiol. B 40, 683-697.

Ackman, R. G., Hingley, J. H., Eaton, C. A., Logan, V. H. and Odense, P. H. (1975). Layering and tissue composition in the blubber of the northwest Atlantic sei whale (Balaenoptera borealis). Can. J. Zool. 53, 1340-1344.

Addison, R. F., Ikonomou, M. G. and Smith, T.G. (2005). PCDD/F and PCB in harbour seals (Phoca vitulina) from British Columbia: response to exposure to pulp mill effluents. Mar. Environ. Res. 59, 165-176.

Addison R. F. and Smith, T. G. (1974). Organochlorine residue levels in Arctic ringed seals: variation with age and sex. Oikos 25, 335-337.

Addison, R. F. and Smith, T.G. (1998). Trends in organochlorine residue concentrations in ringed seal (Phoca hispida) from Holman, Northwest Territories, 1972-91. Arctic 51, 253-261.

Aguilar, A. and Borrell, A. (1990). Patterns of lipid content and stratification in the blubber of fin whales (Balaenoptera physalus). J. Mammal. 71, 544-554.

Aitchison, J. (1983). Principal component analysis of compositional data. Biometrika 70, 57-65.

Beck, G. G., Smith, T. G. and Hammill, M. O. (1993). Evaluation of body condition in the northwest Atlantic harp seal (Phoca groenlandica). Can. J. Fish. Aquat. Sci. 50, 1372-1381.

Berta, A. and Deméré, T. A. (1986). Callorhinus gilmorei n. sp., (Carnivora: Otariidae) from the San Diego Formation (Blancan) and its implications for otariid phylogeny. Trans. San Diego Soc. Nat. Hist. 21, 111-126.

Berta, A., Sumich, J. L. and Kovacs, K. M. (2006). Marine Mammals: Evolutionary Biology. San Francisco: Academic Press.

102 Berta, A. and Wyss, A. R. (1994). Pinniped phylogeny. Proc. San Diego Soc. Nat. Hist. 29, 33-56.

Bininda-Emonds, O. R. P., Gittleman, J. L. and Purvis, A. (1999). Building large trees by combining phylogenetic information: a complete phylogeny of the extant Carnivora (Mammalia). Biol. Rev. 74, 143-175.

Budge, S. M. and Iverson, S. J. (2003). Quantitative analysis of fatty acid precursors in marine samples: Direct conversion of wax ester alcohols and dimethylacetals to fatty acid methyl esters. J. Lipid Res. 44, 1802-1807.

Budge, S. M., Iverson, S. J. and Koopman, H. N. (2006). Studying trophic ecology in marine ecosystems using fatty acids: a primer on analysis and interpretation. Mar. Mamm. Sci. 22, 759-801.

Budge, S. M., Springer, A. M., Iverson, S. J., Sheffield, G. and Rosa, C. (2008). Blubber fatty acid composition of bowhead whales, Balaena mysticetus: Implications for diet assessment and ecosystem monitoring. J. Exp. Mar. Biol. Ecol. 359, 40-46.

Cameron, M. E., Metcalfe, T. L., Metcalfe, C. D. and Macdonald, C. R. (1997). Persistent organochlorine compounds in the blubber of ringed seals (Phoca hispida) from the Belcher Islands, Northwest Territories, Canada. Mar. Environ. Res. 43, 99-116.

Castellini, M. A., Davis, R. W., Loughlin, T. R. and Williams, T. M. (1993). Blood chemistries and body condition of Steller sea lion pups at Marmot Island, Alaska. Mar. Mamm. Sci. 9, 202-208.

Costa, D. P. (1991). Reproductive and foraging energetics of pinnipeds: implications for life history patterns. In: Behaviour of Pinnipeds (ed. D. Renouf), pp. 300- 344. New York: Chapman and Hall.

Costa, D. P. and Kooyman, G. L. (1982). Oxygen consumption, thermoregulation, and the effect of fur oiling and washing on the sea otter, Enhydra lutris. Can. J. Zool. 60, 2761-2767.

Del Toro, L., Heckel, G., Camacho-Ibar, V. F. and Schramm, Y. (2006). California sea lions (Zalophus californianus californianus) have lower chlorinated hydrocarbon contents in northern Baja California, México than in California, USA. Environ. Pollut. 142, 83-92.

103 Deméré, T. A., Berta, A. and Adam, P. J. (2003). Pinnipedimorph evolutionary biology. Bull. Amer. Mus. Nat. Hist. 279, 32-76.

Doncaster, C. P., Dumonteil, E., Barre, H. and Jouventin, P. (1990). Temperature regulation of young coypus (Myocastor coypus) in air and water. Am. J. Physiol. 259, R1220-R1227.

Dunkin, R. C., McLellan, W. A., Blum, J. E. and Pabst, D. A. (2005). The ontogenetic changes in the thermal properties of blubber from Atlantic bottlenose dolphin Tursiops truncatus. J. Exp. Biol. 208, 469-480.

Fay, F. H. (1981). Walrus – Odobenus rosmarus. In: Handbook of Marine Mammals, Volume 1: The Walrus, Sea Lions, Fur Seals and Sea Otter (ed. S. H. Ridgway, R. J. Harrison), pp. 1-23. San Diego: Academic Press.

Fish, F. E. (2000). Biomechanics and energetics in aquatic and semiaquatic mammals: platypus to whale. Physiol. Biochem. Zool. 73, 683-698.

Fish, F. E., Smelstoys, J., Baudinette, R. V. and Reynolds, P. S. (2002). Fur does not fly, it floats: buoyancy of pelage in semi-aquatic mammals. Aquat. Mamm. 28, 103-112.

Flynn, J. J., Finarelli, J. A., Zehr, S., Hsu. J. and Nedbal, M. A. (2005). Molecular phylogeny of the Carnivora (Mammalia): assessing the impact of increased sampling on resolving enigmatic relationships. Syst. Biol. 54, 317-337.

Folch, J., Lees, M. and Sloane-Stanley, G. H. (1957). A simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem. 226, 497-509.

Folkow, L. P. and Blix, A. S. (1987). Nasal heat and water exchange in gray seals. Am. J. Physiol. Regulatory Integrative Comp. Physiol. 253, 883-889.

Frisch, J., Ørtisland, N. A. and Krog, J. (1974). Insulation of furs in water. Comp. Biochem. Physiol. 47A, 403-410.

Hall, A. J., Duck, C. D., Law, R. J., Allchin, C. R., Wilson, S. and Eybator, T. (1999). Organochlorine contaminants in Caspian and harbour seal blubber. Environ. Pollut. 106, 203-212.

Hammel, H. T. (1955). Thermal properties of fur. Am. J. Physiol. 182, 369-376.

104 Hart, J. S. and Fisher, H. D. (1964). The question of adaptations to polar environments in marine mammals. Fed. Proc. 23, 1207-1214.

Harvey, P. H. and Pagel, M. D. (1991). The Comparative Method in Evolutionary Biology. Oxford: Oxford University Press.

Higdon, J. W., Bininda-Emonds, O. R. P., Beck, R. M. D. and Ferguson, S. H. (2007). Phylogeny and divergence of the pinnipeds (Carnivora: Mammalia) assessed using a multigene dataset. BMC Evol. Biol. 7, 216-234.

Hilditch, T. P. and Williams, P. N. (1964). Chemical constitution of natural fats. New York: John Wiley.

Hilton, H. and Kutscha, N. P. (1978). Distinguishing characteristics of the hairs of the eastern coyote, domestic dog, red fox and bobcat in Maine. Am. Midl. Nat. 100, 223-227.

Hochachka, P. W. and Somero, G. N. (2002). Biochemical Adaptation: Mechanism and Process in Physiological Evolution. New York: Oxford University Press.

Hooker, S. K., Iverson, S. J., Ostrom, P. J. and Smith, S. C. (2001). Diet of northern bottlenose whales inferred from fatty-acid and stable-isotope analyses of biopsy samples. Can. J. Zool. 79, 1442-1454.

Irving, L. (1969). Temperature regulation in marine mammals. In: The Biology of Marine Mammals (ed. H. T. Andersen), pp. 147-174. San Francisco: Academic Press.

Irving, L. (1973). Aquatic mammals. In: Comparative Physiology of Thermoregulation, III. Special Aspects of Thermoregulation (ed. G. C. Whittow), pp. 47-96. New York: Academic Press.

Iverson, S. J., Frost, K. J. and Lowry, L. F. (1997). Fatty acid signatures reveal fine scale structure of foraging distribution of harbor seals and their prey in Prince William Sound, Alaska. Mar. Ecol. Progr. Ser. 151, 255-271.

Johansen, K. (1962). Buoyancy and insulation in the muskrat. J. Mammal. 43, 64-68.

Kajiwara, N., Kannan, K., Muraoka, M., Watanabe, M., Takahashi, S., Gulland, F., Olsen, H., Blankenship, A. L., Jones, P. D., Tanabe, S. and Giesy, J. P. (2001). Organochlorine pesticides, polychlorinated biphenyls, and butyltin compounds in blubber and livers of stranded California sea lions, elephant

105 seals, and harbor seals from coastal California, USA. Arch. Environ. Contam.Toxicol. 41, 90-99.

Kajiwara, N., Niimi, S., Watanabe, M., Ito, Y., Takahashi, S., Tanabe, S., Khuraskin, L. S. and Miyazaki, N. (2002). Organochlorine and organotin compounds in Caspian seals (Phoca caspica) collected during an unusual mortality event in the Caspian Sea in 2000. Environ. Pollut. 117, 391-402.

Käkelä, R. and Hyvärinen, H. (1996). Site-specific fatty acid composition in adipose tissues of several northern aquatic and terrestrial mammals. Comp. Biochem. Physiol. B 115, 501-514.

Kannan, K., Kajiwara, N., Le Boeuf, B. J. and Tanabe, S. (2004). Organochlorine pesticides and polychlorinated biphenyls in California sea lions. Environ. Pollut. 131, 425-434.

Kaszowski, S., Rust, C. C. and Shackleford, R. M. (1970). Determination of hair density in the mink. J. Mamm. 51, 27-34.

Kenyon, K. W. (1969). The sea otter in the eastern Pacific Ocean. N. Amer. Fauna 68, 1-352.

Kleivane, L., Espeland, O., Fagerheim, K. A., Hylland, K., Polder, A. and Skaare, J. U. (1997). Organochlorine pesticides and PBCs in the east ice harp seal (Phoca groendlandica) population. Mar. Environ. Res. 43, 117-130.

Koopman, H. N., Iverson, S. J. and Gaskin, D. E. (1996). Stratification and age- related differences in blubber fatty acids of the male harbour porpoise (Phocoena phocoena). J. Comp. Physiol. B 165, 628-639.

Krahn, M. M., Becker, P. R., Tilbury, K. L. and Stein, J. E. (1997). Organochlorine contaminants in blubber of four seal species: integrating biomonitoring and specimen banking. Chemosphere 34, 2109-2121.

Kreith, F. (1958). Principles of heat transfer. New York: Intext Educational Publishers.

Kucklick, J. R., Krahn, M. M., Becker, P. R., Porter, B. J., Schantz, M. M., York, G. S., O’Hara, T. M. and Wise, S. A. (2006). Persistent organic pollutants in Alaskan ringed seal (Phoca hispida) and walrus (Odobenus rosmarus) blubber. J. Environ. Monit. 8, 848-854.

106 Kucklick, J. R., Struntz, W. D. J., Becker, P. R., York, G. W., O’Hara, T. M. and Bohonowych, J. E. (2002). Persistent organochlorine pollutants in ringed seals and polar bears collected from northern Alaska. Sci. Total Environ. 287, 45-59.

Kvadsheim, P. H. and Aarseth, J. J. (2002). Thermal function of phocid seal fur. Mar. Mamm. Sci. 18, 952-962.

Kvadsheim, P. H., Folkow, L. P. and Blix, A. S. (1994). A new device for measurement of the thermal conductivity of fur and blubber. J. Therm. Biol. 19, 431-435.

Lento, G. M., Haddon, M., Chambers, G. K. and Baker, C. S. (1997). Genetic variation, population structure, and species identity of southern hemisphere fur seals, Arctocephalus spp. J. Heredity 88, 28-34.

Lento, G. M., Hickson, R. E., Chambers, G. K. and Penny, D. (1995). Use of spectral analysis to test hypotheses on the origin of pinnipeds. Mol. Biol. Evol. 12, 28-52.

Ling, J. K. (1965). Hair growth and moulting in the southern elephant seal, Mirounga leonina (Linn.). In: Biology of the Skin and Hair Growth (ed. A. G. Lyne, B. F. Short), pp. 525-544. Sydney: Angus and Robertson.

Ling, J. K. (1970). Pelage and molting in wild mammals with special reference to aquatic forms. Quart. Rev. Biol. 45, 16-54.

Ling, J. K. (1974). The integument of marine mammals. In: Functional Anatomy of Marine Mammals, Vol. 3 (ed. R. J. Harrison), pp. 1-44. London: Academic Press.

Lockyer, C. H., McConnell, L. C. and Waters, T. D. (1984). The biochemical composition of fin whale blubber. Can. J. Zool. 62, 2553-2562.

Lockyer, C. H., McConnell, L. C. and Waters, T. D. (1985). Body condition in terms of anatomical and biochemical assessment of body fat in north Atlantic fin and sei whales. Can. J. Zool. 63, 2328-2338.

Lydersen, C., Wolkers, H., Severinsen, T., Kleivane, L., Nordøy, E. S. and Skaare, J. U. (2002). Blood is a poor substrate for monitoring pollution burdens in phocid seals. Sci. Total Environ. 292, 193-203.

107 Maddison, W. P., Maddison, D. R. (2007). Mesquite: a modular system for evolutionary analysis. Version 2.01 http://mesquiteproject.org.

Meyer, W., Schnapper, A. and Hülmann, G. (2002). The hair cuticle of mammals and its relationship to functions of the hair coat. J. Zool. Lond. 256, 489-494.

Montagna, W. and Harrison, R. J. (1957). Specializations in the skin of the seal (Phoca vitulina). Am. J. Anat. 100, 81-114.

Morrison, P., Rosenmann, M. and Estes, J. A. (1974). Metabolism and thermoregulation in the sea otter. Physiol. Zool. 47, 218-229.

Muir, D., Riget, F., Cleemann, M., Skaare, J., Kleivane, L., Nakata, H., Dietz, R., Severinsen, T. and Tanabe, S. (2000). Circumpolar trends of PCBs and organochlorine pesticides in the Arctic marine environment inferred from levels in ringed seals. Environ. Sci. Technol. 34, 2431-2438.

Nilssen, K. T., Haug, T., Grotnes, P. E. and Potelov, V. (1997). Seasonal variation in body condition of adult Barents Sea harp seals (Phoca groenlandica). J. Northw. Atl. Fish. Sci. 22, 17-25.

Noback, C. R. (1951). Morphology and phylogeny of hair. Am. N. Y. Acad. Sci. 53, 476-492.

Nordoy, E. S. and Blix, A. S. (1985). Energy sources in fasting grey seal pups evaluated with computed tomography. Am. J. Physiol. Regulatory Integrative Comp. Physiol. 249, 471-476.

Pabst, D. A., Rommel, S. A. and McLellan, W. A. (1999). The functional morphology of marine mammals. In: Biology of Marine Mammals (ed. J. E. Reynolds, S. A. Rommel), pp. 15-72. Washington: Smithsonian Institution Press.

Parry, D. A. (1949). The structure of whale blubber, and a description of its thermal properties. Quart. J. Micr. Sci. 90, 13-25.

Pitcher, K. W. (1986). Variation in blubber thickness of harbor seals in southern Alaska. J. Wildl. Manage. 50, 463-466.

Pond, C. M. (1978). Morphological aspects and the ecological and mechanical consequences of fat deposition in wild vertebrates. Annu. Rev. Ecol. Syst. 9, 519-570.

108 Reynolds, J. E., Wetzel, D. L. and O’Hara, T. M. (2006). Human health implications of omega-3 and omega-6 fatty acids in blubber of the bowhead whale (Balaena mysticetus). Arctic 59, 155-164.

Rosen, D. A. S. and Renouf, D. (1997). Seasonal changes in blubber distribution in Atlantic harp seals: indications of thermodynamic considerations. Mar. Mamm. Sci. 13, 229-240.

Ryg, M., Smith, T. G. and Ørtisland, N. A. (1988). Thermal significance of the topographical distribution of blubber in ringed seals (Phoca hispida). Can. J. Fish. Aquat. Sci. 45, 985-992.

Samuel, A. M. and Worthy, G. A. J. (2004). Variability in fatty acid composition of bottlenose dolphin (Tursiops truncatus) blubber as a function of body site, season, and reproductive state. Can. J. Zool. 82, 1933-1942.

Savelle, J. M. and Friesen, T. M. (1996). Derivation and application of an otariid utility index. J. Archaeol. Sci. 23, 705-712.

Scheffer, V. B. (1962). Pelage and surface topography of the northern fur seal. N. Amer. Fauna 64.

Scheffer, V. B. (1964a). Estimating abundance of pelage fibres on fur seal skin. Proc. Zool. Soc. London 143, 37-41.

Scheffer, V. B. (1964b). Hair patterns in seals (Pinnipedia). J. Morph. 115, 291-304.

Scholander, P. F., Walters, V., Hock, R. and Irving, L. (1950). Body insulation of some Arctic and tropical mammals and birds. Biol. Bull. 99, 225-236.

Severinsen, T., Skaare, J. U. and Lydersen, C. (2000). Spatial distribution of persistent organochlorines in ringed seal (Phoca hispida) blubber. Mar. Environ. Res. 49, 291-302.

Sokolov, W. (1962). Adaptations of the mammalian skin to the aquatic mode of life. Nature 195, 464-466.

Stapleton, H. M., Dodder, N. G., Kucklick, J. R., Reddy, C. M., Schantz, M. M., Becker, P. R., Gulland, F., Porter, B. J. and Wise, S. A. (2006). Determination of HBCD, PBDEs and MeO-BDEs in California sea lions (Zalophus californianus) stranded between 1993 and 2003. Mar. Pollut. Bull. 52, 522-531.

109 Stewart, R. E. A. and Lavigne, D. M. (1980). Neonatal growth of Northwest Atlantic harp seals, Pagophilus groenlandicus. J. Mamm. 61, 670-680.

Stewart, R. E. A. and Lavigne, D. M. (1984). Energy transfer and female condition in nursing harp seals Phoca groenlandica. Holarctic Ecol. 7, 182-194.

Tarasoff, F. J. (1972). Comparative aspects of the hind limbs of the river otter, sea otter and seals. In: Functional Anatomy of Marine Mammals, Vol. 1 (ed. R. J. Harrison), pp. 333-359. London: Academic Press.

Tarasoff, F. J. (1974). Anatomical adaptations in the river otter, sea otter and harp seal with reference to thermal regulation. In: Functional Anatomy of Marine Mammals, Vol. 2 (ed. R. J. Harrison), pp. 111-141. London: Academic Press.

Taylor, R. H. and Tilley, J. A. V. (1984). Stoats (Mustela erminea) on Adele and Fisherman Islands, Abel Tasman National Park, and other offshore islands in New Zealand. N. Z. J. Ecol. 7, 139-145.

Weisel, J. W., Nagaswami, C. and Peterson, R. O. (2005). River otter hair structure facilitates interlocking to impede penetration of water and allow trapping of air. Can. J. Zool. 83, 649-655.

West, G. C., Burns, J. J. and Modafferi, M. (1979a). Fatty acid composition of blubber from the four species of Bering Sea phocid seals. Can. J. Zool. 57, 189-195.

West, G. C., Burns, J. J. and Modafferi, M. (1979b). Fatty acid composition of Pacific walrus skin and blubber fats. Can. J. Zool. 57, 1249-1255.

Wiig, O. (1989). A description of common seals, Phoca vitulina L. 1758, from Svalbard. Mar. Mamm. Sci. 5, 149-158.

Williams, T. D., Allen, D. D., Groff, J. M. and Glass, R. L. (1992). An analysis of California sea otter (Enhydra lutris) pelage and integument. Mar. Mamm. Sci. 8, 1-18.

Williams, T. M. (1986). Thermoregulation of the North American mink (Mustela vison) during rest and activity in the aquatic environment. Physiol. Zool. 59, 293-305.

Williams, T. M., Kastelein, R. A., Davis, R. W. and Thomas, J. A. (1988). The effects of oil contamination and cleaning in sea otters: thermoregulatory implications based on pelt studies. Can. J. Zool. 66, 2776-2781. 110

Willis, K., Horning, M., Rosen, D. A. S. and Trites, A. W. (2005). Spatial variation of heat flux in Steller sea lions: evidence for consistent avenues of heat exchange along the body trunk. J. Exp. Mar. Biol. Ecol. 315, 163-175.

Worthy, G. A. J. (1991). Insulation and thermal balance of fasting harp and gray seal pups. Comp. Biochem. Physiol. 100A, 845-851.

Worthy, G. A. J. and Edwards, E. F. (1990). Morphometric and biochemical factors affecting heat loss in a small temperate cetacean (Phocoena phocoena) and a small tropical cetacean (Stenella attenuata). Physiol. Zool. 63, 432-442.

Wynen, L. P., Goldsworthy, S. D., Insley, S., Adams, M., Bickham, J., Gallo, J. P., Hoelzel, A. R., Majluf, P., White, R. P. G. and Slade, R. (2001). Phylogenetic relationships within the family Otariidae (Carnivora). Mol. Phylogenet. Evol. 21, 270-284.

Ylitalo, G. M., Stein, J. E., Hom, T., Johnson, L. L., Tilbury, K. L., Hall, A. J., Rowles, T., Greig, D., Lowenstine, L. J. and Gulland, F. M. D. (2005). The role of organochlorines in cancer-associated mortality in California sea lions (Zalophus californianus). Mar. Pollut. Bull. 50, 30-39.

111 Appendix

Table A1. Species means ± 1SD for hair circularity from the present study. N = number of individuals examined for that species. For N=1, means and SD are from three representative hairs. Note that underhairs are absent in the northern elephant seal (Mirounga angustirostris).

Guard Underhair Habitat Group Species hair circularity N circularity Canis familiaris 0.70 ± 0.08 0.81 ± 0.05 3 Canids Canis latrans 0.80 ± 0.10 0.78 ± 0.20 1

Vulpes vulpes 0.82 ± 0.06 0.84 ± 0.04 7 Felis catus 0.88 ± 0.05 0.82 ± 0.06 1 Felids Lynx rufus 0.81 ± 0.02 0.78 ± 0.09 2 Puma concolor 0.71 ± 0.12 0.66 ± 0.07 4 Terrestrial Mustelid Mustela erminea 0.84 ± 0.04 0.75 ± 0.03 1 Raccoon Procyon lotor 0.60 ± 0.20 0.70 ± 0.08 1 Skunk Mephitis mephitis 0.41 ± 0.03 0.55 ± 0.11 1 Semi- Mustelid Lontra canadensis 0.36 ± 0.03 0.79 ± 0.03 3 aquatic Polar bear Ursus maritimus 0.74 ± 0.10 0.66 ± 0.10 1 Mustelid Enhydra lutris 0.39 ± 0.05 0.79 ± 0.04 2 Arctocephalus gazella 0.29 ± 0.01 0.42 ± 0.15 2 Arctocephalus pusillus 0.27 ± 0.03 0.52 ± 0.12 6 Arctocephalus 0.25 ± 0.02 0.47 ± 0.07 1 Fur seals townsendi Arctocepahlus tropicalis 0.26 ± 0.04 0.38 ± 0.09 4

Callorhinus ursinus 0.26 ± 0.04 0.47 ± 0.09 5 Eumetopias jubatus 0.30 ± 0.09 0.43 ± 0.14 3 Sea lions Zalophus californianus 0.28 ± 0.03 0.60 ± 0.10 22 Aquatic Zalophus wollebaeki 0.27 ± 0.02 0.47 ± 0.06 1 Cystophora cristata 0.22 ± 0.01 0.84 ± 0.05 1 Mirounga angustirostris 0.19 ± 0.03 N/A 7 Phocid Pagophilus 0.24 ± 0.03 0.53 ± 0.16 4 seals groenlandica Phoca vitulina 0.21 ± 0.03 0.60 ± 0.07 3 Pusa hispida 0.20 ± 0.05 0.50 ± 0.11 1

112

Table A2. Species means for hair length from the present study and published sources. Dashes indicate missing values. N = number of individuals examined. For N=1 from the present study, means and SD are from three representative hairs. *Guard hair data from Hilton & Kutscha (1978) are weighted averages; no standard deviations were reported.

Guard hair Underhair Habitat Group Species N Source length (cm) length (cm) Canis familiaris 5.224 ± 5.148 3.194 ± 3.389 3 This study 3.392* - 15 Hilton & Kutscha (1978) Canis latrans 3.640 ± 0.251 1.828 ± 0.346 1 This study Canids 7.868* 2.4 32 Hilton & Kutscha (1978)

Vulpes vulpes 5.983 ± 0.575 2.739 ± 0.585 7 This study 3.632* - 8 Hilton & Kutscha (1978) Felis catus 6.671 ± 0.268 5.599 ± 0.259 1 This study Lynx rufus 2.389 ± 0.088 1.384 ± 0.390 2 This study

Terrestrial Felids 3.956* - 5 Hilton & Kutscha (1978) Puma concolor 3.830 ± 0.646 1.921 ± 0.288 4 This study Mustelid Mustela erminea 1.857 ± 0.303 0.869 ± 0.149 1 This study Raccoon Procyon lotor 5.166 ± 0.536 1.993 ± 0.258 1 This study Skunk Mephitis mephitis 5.346 ± 0.057 3.207 ± 0.184 1 This study

Lontra canadensis 1.995 ± 0.308 1.319 ± 0.227 3 This study

- Mustelids 2.03 1.40 1 Fish et al. (2002)

Mustela vison 2.43 1.40 1 Fish et al. (2002)

quatic Semi

a Polar bear Ursus maritimus 8.46 3.40 1 This study 5.0 3.0 2 Frisch et al. (1974) Mustelid Enhydra lutris 3.136 ± 0.401 2.045 ± 0.898 2 This study 2.74 ± 0.09 1.81 ± 0.08 3 Fish et al. (2002) 2.69 ± 0.76 1.28 ± 0.62 10 Williams et al. (1992)

Arctocephalus forsteri 2.5 - 1 Scheffer (1964b) Arctocephalus gazella 2.549 ± 0.352 1.651 ± 0.075 2 This study Arctocephalus pusillus 1.877 ± 0.212 1.096 ± 0.099 6 This study Aquatic Fur seals Arctocephalus townsendi 1.541 ± 0.070 1.161 ± 0.082 1 This study Arctocephalus tropicalis 1.725 ± 0.053 1.149 ± 0.132 4 This study Callorhinus ursinus 2.002 ± 0.284 1.194 ± 0.073 5 This study 2.1 - 1 Scheffer (1964b)

Guard hair Underhair Habitat Group Species N Source length (cm) length (cm) Eumetopias jubatus 1.387 ± 0.241 0.672 ± 0.133 3 This study 3.1 - 1 Scheffer (1964b) Otaria bryonia 1.1 - 1 Scheffer (1964b) Sea lions Phocarctos hookeri 3.0 - 1 Scheffer (1964b) Zalophus californianus 1.026 ± 0.175 0.401 ± 0.117 22 This study 0.9 - 1 Scheffer (1964b) Zalophus wollebaeki 0.868 ± 0.051 0.382 ± 0.028 1 This study Walrus Odobenus rosmarus 0.7 - 1 Scheffer (1964b) Cystophora cristata 1.808 ± 0.182 0.690 ± 0.088 1 This study 1.5 - 1 Scheffer (1964b) Erignathus barbatus 1.0 - 1 Scheffer (1964b)

Halichoerus grypus 1.4 - 1 Scheffer (1964b)

Histriophoca fasciata 0.8 - 1 Scheffer (1964b) Hydrurga leptonyx 0.8 - 1 Scheffer (1964b)

Aquatic Leptonychotes weddellii 1.3 - 1 Scheffer (1964b) Lobodon carcinophagus 1.1 - 1 Scheffer (1964b) Mirounga angustirostris 0.797 ± 0.114 N/A 7 This study Phocid seals 0.7 N/A 1 Scheffer (1964b) Monachus shauinslandi 0.7 N/A 1 Scheffer (1964b) Ommatophoca rossi 1.3 - 1 Scheffer (1964b) Pagophilus groenlandica 1.642 ± 0.218 0.842 ± 0.039 4 This study 1.0 0.3 1 Frisch et al. (1974) 1.3 - 1 Scheffer (1964b) Phoca vitulina 1.027 ± 0.504 0.461 ± 0.075 3 This study 0.9 - 1 Scheffer (1964b) Pusa hispida 1.249 ± 0.127 0.843 ± 0.024 1 This study 1.6 - 1 Scheffer (1964b)

Table A2 (continued)

Table A3. Species means ± 1 SD for fur density from the present study and published sources. N = number of individuals examined. *Values from Frisch et al. (1974) are median values.

Habitat Group Species Fur Density (hairs·mm-1) N Source Canid Vulpes vulpes 61.09 ± 14.21 5 This study Felis concolor 57.33 1 This study Terrestrial Felids Felis rufus 48.09 1 This study Mustelid Mustela erminea 96.59 1 This study Lontra canadensis 585.26 2 This study 803.12 1 Fish et al. (2002) 578.33 ? Tarasoff (1974) Mustelids Semi-aquatic 669.37 ? Tarasoff (1972) Mustela vison 338.45 1 Fish et al. (2002) 241.58 ? Kaszowski et al. (1970) Polar bear Ursus maritimus 12.25* 2 Frisch et al. (1974) Mustelid Enhydra lutris 1188.77 ± 70.37 3 Fish et al. (2002) 775.26 1 Williams et al. (1992) 1253.33 ? Tarasoff (1974) 1340.52 ? Tarasoff (1972) 1008 ? Kenyon (1969) 328.75 ? Sokolov (1962) Aquatic Arctocephalus forsteri 335.8 1 Scheffer (1964b) Arctocephalus gazella 421.94 ± 85.65 2 This study Arctocephalus pusillus 359.06 ± 109.88 6 This study Fur seals Arctocephalus townsendi 540.31 1 This study Arctocephalus tropicalis 417.60 ± 91.03 4 This study Callorhinus ursinus 525.79 ± 130.03 4 This study 491.3 ± 51.32 2 Scheffer (1964b)

Habitat Group Species Fur Density (hairs·mm-1) N Source Eumetopias jubatus 48.16 ± 16.77 2 This study 7.6 1 Scheffer (1964b) Otaria bryonia 10.6 1 Scheffer (1964b) Sea lions Phocarctos hookeri 21.6 1 Scheffer (1964b) Zalophus californianus 24.32 ± 11.39 30 This study 10.8 1 Scheffer (1964b) Walrus Odobenus rosmarus 1.8 1 Scheffer (1964b) Cystophora cristata 13.8 1 Scheffer (1964b) Erignathus barbatus 28.0 1 Scheffer (1964b) Halichoerus grypus 22.2 1 Scheffer (1964b) Histriophoca fasciata 14.5 1 Scheffer (1964b) Hydrurga leptonyx 19.8 1 Scheffer (1964b) Aquatic Leptonychotes weddellii 13.6 1 Scheffer (1964b) Lobodon carcinophagus 28.2 1 Scheffer (1964b) Mirounga angustirostris 14.81 ± 3.32 5 This study Phocid seals 3.7 1 Scheffer (1964b) Monachus shauinslandi 4.5 1 Scheffer (1964b) Ommatophoca rossi 20.6 1 Scheffer (1964b) Phoca groenlandica 17* 1 Frisch et al. (1974) 16.8 1 Scheffer (1964b) Phoca vitulina 42.77 ± 2.24 2 This study 21.6 1 Scheffer (1964b) Pusa hispida 44.19 1 This study 20 1 Scheffer (1964b)

Table A3 (continued)

Table A4. Species means ± 1 SD (by age class) for blubber thickness, from the present study and published sources. For N=1 from the present study, means and SD are from four perimeter sides of a single sample. Only good condition animals were used to calculate means. Note that only adult and subadult values were used for comparisons in the present study.

Group Species Age Class Blubber thickness (mm) N Source Arctocephalus gazella Adult female 24.38 ± 4.96 2 This study Arctocephalus pusillus Adult female 11.20 ± 5.76 6 This study Arctocephalus townsendi Yearling 2.68 ± 0.54 1 This study Fur seals Arctocephalus tropicalis Adult female 6.41 ± 0.66 4 This study Callorhinus ursinus Pup 8.10 ± 3.61 1 This study Subadult 9.33 ± 1.86 6 Krahn et al. (1997) Eumetopias jubatus Pup 4.96 ± 1.85 2 This study 3 ± 1 9 Castellini et al. (1993) Adult female 18.65 ± 1.48 2 Willis et al. (2005) Adult male 50 1 Central Puget Sound Marine Mammal Stranding Network Zalophus californianus Pup 3.038 ± 0.65 1 This study Yearling 12.56 ± 0.06 2 This study Subadult 13.09 ± 3.25 18 This study Sea lions Subadult & adult 20.50 ± 10.85 22 Stapleton et al. (2006) Adult female 9.70 ± 1.77 4 This study 15 ± 11 60 Ylitalo et al. (2005) Adult male 25.46 ± 3.34 3 This study 23.13 ± 6.29 8 Del Toro et al. (2006) 20 ± 14 16 Ylitalo et al. (2005) 42.5 ± 7.07 2 Savelle & Friesen (1996) Zalophus wollebaeki Yearling 14.64 ± 3.66 1 This study Odobenus rosmarus Adult female 57.5 ± 9.19 2 West et al. (1979b) 61.75 ± 17.25 4 Kucklick et al. (2006) Walrus Adult male 63.0 ± 18.38 2 West et al. (1979b) 53.63 ± 48.83 8 Kucklick et al. (2006) Erignathus barbatus Unknown 46.4 ± 14.66 5 Krahn et al. (1997) Juvenile 64 1 West et al. (1979a) Phocids Adult male 51 1 West et al. (1979a) Halichoerus grypus Pup 33.5 ± 4.95 2 Nordøy & Blix (1985) Subadult 39.33 ± 4.51 3 Folkow & Blix (1987)

Table A4 (continued)

Blubber thickness Group Species Age Class N Source (mm) Cystophora cristata Neonate 24.08 ± 10.85 1 This study Pup 50.06 ± 4.79 3 This study Adult female 51.42 ± 17.49 2 This study Adult male 55.27 ± 3.21 3 This study Histriophoca fasciata Subadult 30 1 West et al. (1979a) Adult male 44 1 West et al. (1979a) Mirounga angustirostris Pup 55.73 ± 2.00 2 This study Yearling 29.01 ± 0.63 1 This study Adult female 52.01 ± 0.87 2 This study Adult male 75 1 Central Puget Sound Marine Mammal Stranding Network Phoca groenlandica Neonate 6.8 ± 2.3 7 Stewart & Lavigne (1980) Yellowcoat 17.5 ± 5.6 8 Stewart & Lavigne (1980) Thin whitecoat 18.3 ± 5.5 6 Stewart & Lavigne (1980) Fat whitecoat 32.0 ± 4.5 12 Stewart & Lavigne (1980) Graycoat 36.9 ± 2.5 28 Stewart & Lavigne (1980) Phocids Ragged jacket 45.83 ± 6.22 5 This study 43.6 ± 3.6 17 Stewart & Lavigne (1980) Beater 38.61 ± 10.37 3 This study 41.9 ± 3.7 24 Stewart & Lavigne (1980) Pup 23.90 ± 11.33 80 Stewart & Lavigne (1984) Juvenile 31.5 ± 2.12 16 Kleivane et al. (1997) Adult female 40.05 ± 7.58 4 This study 54.5 ± 6.7 10 Lydersen et al. (2002) 27 9 Kleivane et al. (1997) 48.5 ± 9.03 102 Stewart & Lavigne (1984) Adult male 39 ± 11.3 16 Kleivane et al. (1997) Unknown 38.05 ± 27.32 6 Nilssen et al. (1997) Phoca vitulina Neonate 10 ± 1.4 10 Pitcher (1986) Juvenile 27.8 ± 3.11 1 Rosen & Renouf (1997) 22 1 Wiig (1989) 21.3 ± 6.95 51 Pitcher (1986)

Table A4 (continued) Group Species Age Class Blubber thickness Source N (mm) Phoca vitulina Subadult 24.84 ± 1.33 2 This study 18.2 ± 5.6 11 Addison et al. (2005) 19 1 Wiig (1989) 25.5 ± 6.5 116 Pitcher (1986) Adult female 36.11 ± 1.87 1 This study 38.7 ± 16.3 3 Addison et al. (2005) 33.7 ± 8.0 1 Rosen & Renouf (1997) 28 1 Wiig (1989) 29.8 ± 5.4 212 Pitcher (1986) 49 1 West et al. (1979a) Adult male 22.8 ± 10.3 12 Addison et al. (2005) 30.8 ± 2.9 4 Rosen & Renouf (1997) 26.6 ± 3.9 179 Pitcher (1986) 34 1 West et al. (1979a) Unknown 27.7 ± 10.0 102 Hall et al. (1999) Pusa caspica Adult female 52.42 ± 11.24 12 Kajiwara et al. (2002) Adult male 38.75 ± 11.44 4 Kajiwara et al. (2002) Phocids Pusa hispida Subadult 57.62 ± 2.78 1 This study 38.86 ± 10.65 7 Krahn et al. (1997) 30.50 ± 2.08 4 Muir et al. (2000) 31.29 ± 6.23 14 Kucklick et al. (2006) Adult female 32.22 ±16.99 9 Kucklick et al. (2006) 19.33 ± 1.16 3 Kucklick et al. (2002) 45.11 ± 6.60 9 Muir et al. (2000) 39.85 ± 9.03 49 Addison & Smith (1998) 30 ± 5 6 Cameron et al. (1997) 34 ± 5.7 13 Addison & Smith (1974) Adult male 31.56 ± 12.63 9 Kucklick et al. (2006) 26.50 ± 6.32 6 Kucklick et al. (2002) 44.13 ± 8.73 9 Muir et al. (2000) 35.08 ± 8.88 58 Addison & Smith (1998) 34.5 ± 6.3 31 Cameron et al. (1997) 53 1 West et al. (1979a) 32 ± 7.2 15 Addison & Smith (1974) All 36 ± 5 11 Severinsen et al. (2000)

Table A5. Species means ± 1 SD for blubber and subcutaneous fat lipid content, from the present study and published sources. Note that only good condition animals were used to calculate means (where possible). *Value from Addison & Smith (1974) is a median value.

Group Species Lipid Content (%) N Source Arctocephalus gazella 62.3 ± 16.0 2 This study Arctocephalus pusillus 50.4 ± 19.2 2 This study Fur seals Arctocephalus tropicalis 61.7 ± 7.1 2 This study Callorhinus ursinus 54.0 ± 14.7 7 Krahn et al. (1997) Zalophus californianus 56.1 ± 12.7 6 This study 61.8 ± 19.1 18 Del Toro et al. (2006) Sea lion 53.0 ± 15.4 20 Stapleton et al. (2006) 50 ± 24 36 Kannan et al. (2004) 58.0 ± 23.5 9 Kajiwara et al. (2001) Walrus Odobenus rosmarus 71.7 ± 12.0 15 Kucklick et al. (2006) Erignathus barbatus 79.7 ± 5.2 6 Krahn et al. (1997) Mirounga angustirostris 70.0 ± 6.1 5 This study 85.0 ± 1.4 2 Kannan et al. (2004) 73.8 ± 37.2 4 Kajiwara et al. (2001) Phoca groenlandica 91.1 ± 5.0 17 Lydersen et al. (2002) 93.4 ± 0.9 41 Kleivane et al. (1997) 97.0 ± 1.5 13 Beck et al. (1993) Phoca vitulina 76.5 ± 3.4 3 This study 91.6 ± 9.4 26 Addison et al. (2005) 87 1 Kannan et al. (2004) 77.9 ± 18.7 15 Krahn et al. (1997) Phocids Pusa caspica 85.0 ± 6.8 23 Kajiwara et al. (2002) 78.2 ± 17.0 10 Hall et al. (1999) Pusa hispida 78.2 1 This study 85.1 ± 3.9 33 Kucklick et al. (2006) 85.7 ± 1.6 8 Kucklick et al. (2002) 92.2 ± 2.4 26 Muir et al. (2000) 87.3 ± 4.8 11 Severinsen et al. (2000) 89.6 ± 3.2 17 Addison & Smith (1998) 88.8 ± 3.5 8 Krahn et al. (1997) 94* 28 Addison & Smith (1974) Mustelid Enhydra lutris 4.6 1 Kannan et al. (2004) Ursid Ursus maritimus 83.2 ± 1.4 5 Kucklick et al. (2002)

120

CHAPTER 2

The effects of water temperature on the energetic costs of

juvenile and adult California sea lions (Zalophus californianus):

The importance of recycling endogenous heat during activity

Abstract

As highly mobile marine predators, many pinniped species routinely encounter a wide range of water temperatures during foraging as well as in association with seasonal, geographic, and climatic changes. To determine how such variation in environmental temperature may impact energetic costs in otariids, I determined the thermal neutral zone of mature and immature California sea lions

(Zalophus californianus) by measuring resting metabolic rate using open-flow respirometry. Five adult female (body mass range = 82.2 - 107.2 kg) and four juvenile (body mass = 26.2 - 36.5 kg) sea lions were examined over experimental water temperatures ranging from 0 to 20 C (adults) or 5 to 20 C (juveniles). I found

-1 -1 that the metabolic rate of adult sea lions averaged 6.36  0.64 mLO2kg min when resting within the thermal neutral zone. The lower critical temperature of adults was

6.4  2.2 C, approximately 4 C lower than sea surface temperatures routinely encountered off coastal California. In comparison, juvenile sea lions showed a narrower range of physiologically neutral water temperatures. Resting metabolic rate

-1 -1 of the younger animals, 6.33  0.53 mLO2kg min , increased as water temperature 121 approached 12C, and suggested a potential thermal limitation in the wild. To determine whether muscle thermogenesis during activity could mitigate this limitation, I measured the active metabolic rate of juveniles swimming at Twater = 5,

12, and 20 C. No statistical difference (F=0.063, P=0.939) was found among water temperatures, suggesting that thermal disadvantages due to small body size in immature sea lions may be circumvented by recycling endogenous heat during locomotor activity.

Introduction

For endotherms, maintaining thermal balance is especially challenging when in water (Williams, 1983; Dejours, 1987; Hind and Gurney, 1997). The high thermal conductivity and heat capacity of water compared to air promotes elevated rates of heat loss through conductive and convective pathways. Consequently, marine endotherms have developed a variety of mechanisms to counteract excessive heat transfer when the animals are immersed. These include increased insulation

(Scholander et al., 1950; Irving and Hart, 1957; Williams and Worthy, 2002), elevated basal metabolic rates (Scholander et al., 1942; Williams et al., 2000), and increased metabolic heat production associated with the digestion of food (Costa and

Kooyman, 1984; Yeates, 2006) and locomotor activity (Williams, 1999).

Numerous studies have demonstrated elevated metabolic costs in aquatic and semi-aquatic mammals (Irving and Hart, 1957; Fish, 1979; Dawson and Fanning,

1981; Costa and Kooyman, 1982; Williams, 1986; Williams et al., 2001). While 122 raising heat production through increased metabolism appears to be an effective mechanism by which to counteract heat loss, it also represents an added energetic expense to the animal, especially if employed over long periods. This problem can be exacerbated in young mammals, due both to their smaller body size compared to adults and to an immature physiology. Smaller body size results in a larger surface area to volume ratio, such that the animal has proportionately more surface area through which heat can be transferred. In addition, physiological thermal competence is slow to develop in otariids (e.g., Donohue et al., 2000; Rutishauser et al., 2004).

As a result, young otariids are more likely to experience elevated thermoregulatory costs compared to adults.

The natural history of many species of pinnipeds requires independent living and foraging early in life (e.g., days to 1 year). In view of the thermal challenges presented above, it is likely that these young animals encounter higher thermal costs than adults experience. For example, California sea lion (Zalophus californianus,

Esson) pups wean between six and twelve months of age, at a body size approximately one-third of adult mass (Heath, 1989). During the weaning process, pups do not learn foraging skills from their mothers (Melin et al., 2000); the increased hunting time in the water associated with learning to find food will expose the young animals to extended periods of elevated thermal costs. Depending on the duration and magnitude of these costs, the insulating blubber layer will be depleted as fat stores are converted to energy, resulting in a reduction of thermal stability in the water and a spiraling decline in energy balance (Rutishauser et al., 2004; Rosen et al., 123 2007). Because thermoregulatory processes increase the costs of swimming and diving in pinnipeds (Hind and Gurney, 1997), the thermal limitations of immature animals will impact the transition to the ocean and independent foraging, a phase that is critical to survival into adulthood.

One strategy that young otariids may utilize to offset thermal limitations during foraging is the recycling of heat generated by swimming activity. An increase in metabolic heat production associated with swimming exercise has been demonstrated for pinnipeds and cetaceans (Williams, 1999). This metabolic increase has been shown to provide thermal energetic benefits for a variety of aquatic and semi-aquatic mammals including mink (Williams, 1983), muskrat (Fish, 1979), and the sea otter (Yeates et al., 2007). Similarly, sustained activity may help to reduce the energetic costs of thermoregulation in young sea lions.

The purpose of this study was to investigate the thermal capabilities and limitations of young otariids. To accomplish this, I examined the effects of water temperature on energetic costs of California sea lions. The metabolic responses of adult and juvenile California sea lions resting over a range of water temperatures were measured, and the results used to develop a thermal neutral zone (TNZ) for each age class (Bartholomew, 1977). Below the lower critical temperature (TLC) or above the upper critical temperature (TUC), an animal must increase its metabolism (and thus energetic cost) to offset heat loss or facilitate heat dissipation, respectively. As noted above, because immature animals are likely to experience greater thermal challenges than adults, I also explored one mechanism for mitigating potential thermal 124 limitations in juvenile sea lions. The importance of skeletal muscle thermogenesis for thermal balance was evaluated by comparing the metabolic rates of juvenile sea lions swimming in a flume at different water temperatures. By including temperatures outside the TNZ, I was able to examine the effects of locomotor activity on total energetic responses.

Materials and Methods

Animals

Nine California sea lions were used in this study (Table 1). Five adult females were housed at Long Marine Laboratory (Santa Cruz, CA). Three of the adults were captive sea lions on loan from the Brookfield Zoo (Brookfield, IL), and two were wild-caught from San Nicolas Island (CA). Four immature (less than 1 year old)

California sea lions (3 male, 1 female), on loan from Sea World (San Diego, CA) were housed at the Physiological Research Laboratory (Scripps Institution of

Oceanography, La Jolla, CA). All animals were maintained in saltwater pools and fed a daily diet of herring, capelin, and/or mackerel supplemented with vitamins.

Water temperatures in the holding pools reflected ambient coastal water conditions and averaged 15 °C for the adults and 18 C for the juveniles during the experimental period.

125 Experimental protocol

Adults. Oxygen consumption (VO2) was determined by training the sea lions to rest quietly beneath a Plexiglas metabolic hood (approx. 72 cm high x 40 cm wide x 138 cm long) mounted above a temperature controlled saltwater pool. Animals were conditioned for several months prior to the tests to remain quiescent during the trials.

Water temperature was randomly varied between metabolic trials, from 0 °C to 20 °C, in approximately 5 °C increments. Temperature of the seawater was adjusted with ice or heated water, and was monitored throughout each trial. Behavior of the animal was recorded continuously and only quiescent periods were used in the analyses.

Each experimental session consisted of a 5 - 10 min acclimation period at the trial water temperature, followed by a 30 - 120 min data acquisition period under the metabolic dome. Each animal was tested only once per day, and care was taken to avoid testing the same animal on sequential days. Resting oxygen consumption was determined for 2 - 3 separate trials at each temperature ( 1 C) for each animal.

Rectal temperatures were recorded for the animals as described below.

Juveniles. Oxygen consumption and stomach temperature (Tstomach) were measured simultaneously as juvenile sea lions rested or swam at 1.0 ms-1 against a current in a water channel (Hydraulics Laboratory, Scripps Institution of Oceanography, La Jolla,

CA). During swimming the animals were attached to a weighted line set at three different work levels (25 - 50%, 75%, and 80 - 100% of maximum effort). Loads were determined from the predicted metabolic scope of swimming sea lions

(Feldkamp, 1987; Williams et al., 1991) and followed the procedures of di Prampero 126 et al. (1974). Exercise load was increased by adding weights to a line connected to a frictionless pulley system that was attached to a nylon loop carried on the back of the animal. Weights ranged from 1 to 5 kg depending on the animal, and were adjusted to submaximal exercise levels as determined in Williams et al. (1991). The channel was filled with fresh water maintained at 5, 12, or 20 C. Additional details of the water channel and the experimental test section have been described in Davis et al.

(1985) and Williams et al. (1991).

Four weeks prior to experimentation the sea lions were trained to swim consistently in the water channel at 1.0 ms-1 for 20 - 30 min. Experiments were delayed until reproducible values (within 5%) for oxygen consumption were obtained. Each experimental session consisted of a 20 - 60 min rest period followed by 20 - 30 min swim bouts at each of the three work loads. Swim bouts were separated by rest periods of at least 15 min. Resting values for oxygen consumption and Tstomach were determined during 10 minute quiescent periods before the swimming sessions as the animals floated in the water channel.

Oxygen consumption

The rate of oxygen consumption during all trials was measured by open flow respirometry following the protocols of Williams et al. (2002). Adult sea lions were fasted overnight (≥ 12 hours). Conditions for basal metabolic rate (BMR) were followed for the adults, such that animals were non-pregnant, non-lactating, post- absorptive, and sedentary (Kleiber, 1975). Ambient air was drawn through the 127 Plexiglas chamber at 155 L·min-1, with flow rates maintained and monitored continuously with a mass flow controller (Flowkit 500, Sable Systems International,

Las Vegas, NV). Samples of air from the exhaust port of the chamber were dried

(Drierite; W.A. Hammond Drierite Co., Xenia, OH) and scrubbed of carbon dioxide

(Sodasorb; Chemetron, St. Louis, MO) before entering an oxygen analyzer (FC-1,

Sable Systems). The percentage of oxygen in the expired air was monitored continuously and recorded once per second with a personal computer using Sable

Systems software. Oxygen consumption was then calculated using equation 4b from

Withers (1977) and an assumed respiratory quotient of 0.77 (Feldkamp, 1987; Boyd et al., 1995; Arnould et al., 2001). All values were corrected to STPD. The lowest continuous oxygen consumption that was maintained for 10 - 20 min and corresponded to quiescent behavior was used for the analyses. The entire system was calibrated daily with dry ambient air (20.94% O2) and every 3 - 4 days with dry span gases (16.0% O2) and N2 gas according to Fedak et al. (1981). The theoretical fraction of O2 leaving the chamber was calculated from Davis et al. (1985) and compared to measured values from the analyzer.

VO2 measurements for the juvenile sea lions were the same as for the adults, with the following differences. Oxygen consumption measurements were performed with animals resting or swimming in a flume. Details of the system have been reported previously (Davis et al., 1985; Williams et al., 1991). Briefly, the sea lions swam in a glass paneled test section that was 2.5 m long and had a cross-sectional area of 1.1 m2. A plywood cover restricted breathing in the test section to a 1.1 m 128 long  0.6 m wide  0.3 m high Plexiglass skylight dome. Ambient air was drawn into the dome at approximately 80 - 90 Lmin-1 (100 Flo-Sen; McMillan Co.). Flow rates were adjusted to ensure an O2 content in the dome in excess of 20% over 1 min intervals. Samples of the outflow of the metabolic dome were dried (Drierite) and scrubbed of carbon dioxide (Baralyme) before entering an oxygen analyzer (Amatek

S3-A, Inc.). Voltage output of the analyzer was monitored by an A/D converter

(Keithley 500) connected to a personal computer. Calculations and calibrations were as described above for the adult measurements.

Core temperature

Rectal temperature of the adult sea lions was measured before and after metabolic trials with a flexible thermocouple probe inserted approximately 20 cm into the rectum. Values were recorded ( 0.1 ºC) from a digital thermometer (Physitemp

Instruments, Inc., Clifton, NJ). For the juveniles, core body temperature was determined from stomach temperature during rest and swimming. Stomach temperature was monitored continuously with an ingested temperature sensitive radio transmitter pill (Model L-M; Mini-Mitter, Inc.). Prior to use each pill was calibrated in a temperature controlled waterbath over a range of 32 - 42 C. Calibrated pills were then fed to each sea lion in a fish. Pulse signals from the pills were received on a handheld radio receiver (Model CH-3, Realistic) and averaged for 1 min intervals for every 5 min of rest or exercise. Rectal temperature of the juvenile sea lions was

129 also measured before and after each experimental period using a thermocouple probe and digital thermometer (Physitemp Instruments, Inc.). These values were compared to Tstomach and used to assess stability of the temperature pills. To avoid drift after calibration, individual pills were used for less than 10 consecutive days during the experimental period.

Statistical analyses and calculations

Values for oxygen consumption for adult sea lions are presented as means 

1SD. To determine the TLC of each adult sea lion, a breakpoint for the relationship between mass-specific resting metabolic rate (RMR) and water temperature (Twater) was defined from a segmented regression analysis minimizing the sum of the residual sum of squares (Sokal and Rohlf, 1981; Nickerson et al., 1989). The temperature at which the two lines of best fit intersected was used to define the lower critical temperature. The TLC was determined for each individual sea lion (Table 1).

Statistical significance between rectal temperatures before and after metabolic trials was determined by a randomized block design, with individuals as blocks (Systat

10.2). Linear regressions of rectal temperatures with water temperatures were calculated by least-squares procedures (Zar, 1974).

For the juvenile sea lions values for both oxygen consumption and stomach temperature are presented as means  1SD. Statistical significance between resting metabolism, stomach temperature, and swimming metabolism at different water temperatures was determined by 1-way ANOVAs using a Bonferroni pairwise 130 multiple comparison test (Systat Software, Inc., Chicago, IL). Linear relationships were calculated by least-squares procedures (Zar, 1974). Statistical significance among slopes was determined by ANCOVA, followed by comparison of means and confidence intervals among individual regressions (Systat).

Theoretical swimming velocities for juvenile sea lions were calculated from load values, following the methods of Williams et al. (1991). Briefly, each load was converted into a drag force (Newtons) by multiplying the load mass (kg) by the acceleration due to gravity (9.8 m·s-2). This value was added to a calculated value for body drag, based on parameters measured for California sea lions according to

Drag = ½  v2 A Cd (1)

where  is the density of water (10 kg·m-3), v is forward velocity (1 m·s-1), A is the surface area of the sea lion calculated from a mass-based regression using data from

Feldkamp (1987), and Cd is the drag coefficient (0.004067) averaged from Feldkamp

(1987). Theoretical swimming velocity was calculated for each load by solving the drag equation for v, to find the velocity necessary to overcome the calculated total drag force. Metabolic rates were converted to Watts, assuming an average caloric equivalent of 20.1 kJ per LO2 (Schmidt-Nielsen, 1997). Exponential relationships for oxygen consumption in relation to swimming velocity were calculated by least- squares procedures (Zar, 1974).

131 Results

Resting Metabolism and Thermal Neutral Zone

Lower critical temperature (TLC) was determined for each individual adult sea lion by a rise in RMR (Table 1). The average TLC for the adult sea lions was 6.4 ± 2.2

°C (Fig. 1a). This indicates a thermal neutral zone (TNZ) of at least 16 ºC, as metabolic rate did not change significantly between 6.4 ºC and 22.4 ºC (F=3.429,

P=0.072). The average metabolic rate (RMR) of adult California sea lions resting in

-1 -1 water within the TNZ was 6.36  0.64 mLO2kg min . Below the TNZ, resting

-1 -1 metabolism increased to a maximum of 10.17  1.73 mLO2kg min at 0.9  0.5 °C, which represented a 30 - 94% increase in RMR for individual sea lions. The average linear increase in RMR below the TNZ (Fig. 1a) was described by the equation

2 RMR = 10.80 – 0.63(Twater) (P<0.0001, r =0.34) (2)

-1 -1 where RMR is in mLO2kg min and Twater is in ºC.

The metabolic rate of immature California sea lions resting in water ranged

-1 -1 from 6.33  0.53 to 8.73  0.52 mLO2kg min and was dependent on water temperature (Twater). Resting metabolism of the animals increased significantly

(n=12, F=23.432, P=0.001) as Twater decreased over the range of 5 - 20 C (Fig. 1b) and was described by the equation

132 2 RMR = 9.59 – 0.16(Twater) (P=0.001, r =0.67) (3)

-1 -1 where RMR is in mLO2kg min and Twater is in ºC. Resting metabolism did not differ significantly between 5 and 12 C (P=0.29), but approached significance between 12 and 20 C (P=0.068). Resting metabolism did differ significantly between 5 and 20 C (P=0.004), and individual sea lions demonstrated a 26 - 68% increase in resting metabolic rate over the 15 C decrease in water temperature.

-1 -1 Average metabolic rate at Twater = 20 C was 6.33  0.53 mLO2kg min and was comparable to that reported by Feldkamp (1987) for sea lions resting at similar acclimation temperatures. Interestingly, the mean value for RMR in the juveniles was similar to RMR within the TNZ for the adults in this study (Table 1). Note, however, that the two smallest juveniles exhibited significantly higher RMR values than the adults or larger juveniles (P=0.03; Table 1).

Core Temperature

Adult sea lions maintained rectal temperatures between 35.8 °C and 38.2 °C

(average = 37.0  0.5 °C) during the metabolic trials (Fig. 1c). There was no significant difference between pre-trial and post-trial rectal temperatures for any sea lion (F=0.35, P=0.56), although rectal temperature did differ among individuals

(F=16.14, P<0.001). Post-trial rectal temperature was not significantly correlated with experimental water temperature (F=0.76, P=0.39; Fig. 1c).

133 Whether resting or swimming, all juvenile sea lions maintained stomach temperatures (Tstomach) between 34 C and 40 C. Average Tstomach for the immature animals resting in the water channel was 37.3  1.3 C (n=11) for Twater ranging from

5 C to 20 C (Fig. 1d). Although there was a slight trend indicating a decrease in

Tstomach with increasing water temperature for the resting sea lions, the response was not significant (F=0.956, P=0.354). Swimming activity had no consistent effect on stomach temperature of the juvenile sea lions. Mean Tstomach following 20 min of

-1 swimming at 1 ms with no load was 37.8  0.1 C at Twater = 5 C, 38.0  0.7 C at

12 C, and 36.7  0.6 C at 20 C. Tstomach for sea lions swimming with maximum loads was within 0.9 C of these values at comparable water temperatures.

Activity Thermogenesis

The swimming metabolism of juvenile sea lions was independent of Twater over the range of exercise loads examined (Fig. 2, Fig. 3). Thus, no statistical difference in metabolic rate was apparent among the three water temperatures examined when swimming at 1 ms-1 only (n=4, F=0.063, P=0.939), or swimming at this speed with additional loads of 3 kg (n=4, F=0.211, P=0.814) or 4 - 5 kg (n=3,

-1 F=0.384, P=0.695). Mean VO2 for sea lions swimming freely at 1 ms (i.e., with no

-1 -1 load) was 11.17  0.36 mLO2kg min for all three water temperatures. This value was within 10% of that reported by Feldkamp (1987) and Williams et al. (1991) for similarly sized sea lions swimming at the same speed at Twater = 15 - 18 C.

134 Swimming metabolism (MR in Watts) of the sea lions increased exponentially with calculated velocity (Fig. 3) and was described by the equation

MR = 59.85 + 38.65 e0.4244v (P<0.0001, r2=0.86) (4)

for all four sea lions exercising across all water temperatures, where velocity (v) is in ms-1. This non-linear pattern in energetic cost would be expected due to the exponential increase in hydrodynamic drag with increases in swimming velocity

(Feldkamp, 1987).

Swimming metabolism of the sea lions increased linearly with work load (Fig.

4) and was described by

2 VO2 = 11.9 + 4.2(load) (P<0.00001, r =0.81) (5)

for all four sea lions exercising across all water temperatures, where VO2 is in

-1 -1 mLO2kg min and load is in kg. The slope of the metabolic response to exercise load differed significantly among three of the four individuals (ANCOVA, P=0.023;

Fig. 4), and generally corresponded to differences in body mass. The smallest animal, in particular, demonstrated the largest metabolic response with increased load and hence swimming effort. The relationships for VO2 in relation to load for each juvenile were as follows.

135 Juv1 MR = 10.398 + 4.748(load) (P<0.001, r2=0.97, CI=4.038-5.413)

(6a)

Juv2 MR = 8.833 + 3.970(load) (P<0.001, r2=0.97, CI=3.418-4.522)

(6b)

Juv3 MR = 10.430 + 5.062(load) (P<0.001, r2=0.85, CI=3.278-6.847)

(6c)

Juv4 MR = 12.695 + 6.555(load) (P<0.001, r2=0.97, CI=5.685-7.425)

(6d)

Differences among slopes were determined by comparing means and 95% confidence intervals. Because of high variance in the data, the slope of the response for Juv3 was not significantly different from that of the other animals. However, the trend still corresponded to body mass among all four animals.

Discussion

Thermal capabilities of the California sea lion

The range of movements available to endotherms within oceanic environments would appear unlimited in terms of spatial access. An important caveat, however, is the effect of water temperature on energetic costs to endotherms 136 and subsequent selection of geographic ranges. To maintain energy balance, many species of marine mammal must balance the benefits of a large range with potential energetic disadvantages that occur when thermal limits are exceeded. A relationship between lower critical temperature and oceanic movements exists for phocid seals, such that species occupy ranges that enable them to maintain minimum resting metabolic rates in the wild (Hart and Irving, 1959; Gallivan and Ronald, 1979). The specific temperature ranges depend on species and relative adherence to thermal safety factors. Thus, with a TLC below 0 ºC, harp seals maintain a safety factor of several degrees (Gallivan and Ronald, 1979), while harbor seals alter movements seasonally to remain in water temperatures close to their exact TLC (Hart and Irving,

1959).

Less is known about the thermal territories and tolerances of otariids. In the present study, the TLC values measured for adult female California sea lions indicate that this age class does not incur an additional thermoregulatory energetic cost within the routine range of water temperatures encountered off of the California coast, even while resting at sea. The average TLC for adult female California sea lions, 6.4 ± 2.2

°C (Fig. 1a), is approximately 4 C lower than the range of sea surface temperatures routinely encountered by this species in California (NASA MODIS database). This has important implications for the ability of these animals to rest at sea. With a relatively wide TNZ of at least 16 ºC as determined in the present study, the

California sea lion has a large thermal territory available for foraging and movement.

This thermal flexibility could allow for continued range expansion and consequent 137 tolerance to climatic changes. Admittedly, lower water temperatures will be encountered as the sea lions move below the thermocline during diving. However, both the relatively short duration of dives (Feldkamp et al., 1989) and the potential for increased heat production through skeletal muscle thermogenesis (see below) may circumvent any thermal energetic disadvantages associated with lower water temperatures at depth.

California sea lions may also utilize vascular control to improve body insulation when acutely exposed to cold water. In the present study, individual variability in RMR increased as water temperatures decreased below the TLC (Fig.

1a). Such a response would occur with modifications in the level of perfusion and concomitant thermal conductance of the insulating blubber layer. This has been observed for phocid seals, which use a blubber-encased trunk as an important site for heat dissipation on land (Mauck et al., 2003). Although the superficial blood vessel arrangements required for such heat dissipation in air are comparatively sparse in

California sea lions, the deeper location of such vascular structures in sea lion skin provides the potential for significant heat dissipation in warm water as well as heat conservation when cold (Bryden and Molyneux, 1978).

Thermal limitations in juvenile sea lions

During the transition to independence, young pinnipeds theoretically have access to the same oceanic foraging grounds as adult conspecifics. However, the thermal energetic costs associated with immersion are greater in smaller, immature 138 animals compared to adults and may limit the range for foraging by juveniles at a critical point in their life history. The implications for young phocids and otariids are likely quite different due to differences in their insulation. Young phocids have a significantly thicker and better quality blubber layer compared to young otariids

(Mostman-Liwanag, Chapter 1). Accordingly, juvenile harp seals and harbor seals appear to be thermally competent in the range of water temperatures routinely encountered in the wild (Hart and Irving, 1959; Gallivan and Ronald, 1979). This contrasts with young otariids, which exhibit thermal limitations during the period of transition to independent living. Both pup and juvenile Antarctic fur seals demonstrate thermal energetic costs that limit the range of foraging and prevent these age classes from utilizing adult foraging grounds (Rutishauser et al., 2004). Northern fur seal pups also show elevated metabolic rates at water temperatures regularly encountered in the wild, even in coastal regions close to the rookery (Donohue et al.,

2000; Mostman-Liwanag, Chapter 3).

A major consequence of the higher TLC of juvenile California sea lions is the potential effect on resting at sea. For the juveniles in this study, rest periods were the most thermally challenging, and manifested in an increase in metabolic rate as Twater decreased (Fig. 1b). With an estimated TLC above 12 °C and below 20 °C, juvenile

California sea lions will experience added energetic costs when resting within their natural range of water temperatures. Although RMR at 20 C (within the expected

- TNZ) was significantly higher in the two smaller juveniles (9.15  0.83 mLO2kg

1 -1 -1 -1 min ) compared to the larger juveniles (5.47  0.53 mLO2kg min ), all four 139 animals exhibited increases in metabolism at 12 C (Fig. 1b). This marked difference in mass-specific metabolism may represent a body-size threshold for RMR between

32 and 35 kg (Table 1). Interestingly, these differences in RMR were not reflected in the estimated TLC, indicating that body size and RMR are not the only factors driving the observed differences in thermal competence between juvenile and adult California sea lions.

The effect of locomotor activity on thermal responses in juvenile sea lions

The heat generated by swimming activity could be used to offset the thermal limitations experienced by juvenile California sea lions at sea. Thermal benefits associated with skeletal muscle thermogenesis depend on the balance between elevated heat loss associated with convective cooling and increased heat production that occurs with locomotor activity by the swimmer (Hind and Gurney, 1997). The concept of thermal substitution of heat produced by exercise is not new, and has been demonstrated in several species of birds and rodents (reviewed by Lovvorn, 2007).

Hind and Gurney (1997) modeled the metabolic cost of swimming in several marine mammal species, including juvenile California sea lions. This model accounted for both hydrodynamic and thermal interactions between the animals and the surrounding medium. By incorporating my measured values for resting metabolic rate above and below the TLC into the Hind and Gurney (1997) model, I was able to compare the model thermal predictions with the measured heat production at different swimming velocities (Fig. 5). While my data and the model predictions agree at 140 velocities up to 3 m·s-1, large differences occur at higher swimming velocities. Above

3 m·s-1, the model markedly overestimated metabolic heat production.

One explanation for the difference between predicted and measured swimming metabolism may be vascular control of the blubber. A major assumption of the Hind and Gurney (1997) model is that blubber perfusion is greatly reduced during swimming, such that heat conductance through the insulation is minimized.

However, rates of heat production at high swimming velocities may exceed rates of heat loss through unperfused blubber. In this situation, heat dissipation would be necessary to maintain thermal balance during exercise. Indeed it has been shown that

Steller sea lions use the trunk for heat dissipation during heavy exercise in water, demonstrating that exercising sea lions will balance heat production by increasing heat loss across the blubber at high work loads (Willis et al., 2005).

I found that activity served an important role for maintaining thermal energetic balance in young sea lions. When resting, immature sea lions demonstrated an increase in metabolism with declining water temperatures below 20 ºC (Fig. 1b).

With swimming activity and the skeletal muscle thermogenesis associated with flipper movements this change in metabolism was mitigated.

In mammals, the conversion of stored energy into mechanical energy during exercise is relatively inefficient, and approximately 80% of the energy is lost as heat

(Hodgson et al., 1994). Sea lions appear to use this “wasted” heat to compensate for elevated thermal demands at low water temperatures. Thus, I found that water temperature had no effect on metabolic rate during swimming regardless of work load 141 over the range of 5 to 20 C (Fig. 3). Rather than being additive, increased heat production associated with activity counterbalanced the increased thermal demands associated with lower water temperatures. This metabolic compensation occurred at relatively low levels of effort. Swimming speeds as low as 1 ms-1, approximately

50% of the routine speed of free-ranging sea lions (Ponganis et al., 1990), resulted in sufficient heat to compensate for the thermal demands associated with a 15 C decrease in Twater (Fig. 2).

Therefore, by remaining active when at sea, young sea lions can avoid prolonged energetic thermal demands that would otherwise occur when resting in water below their lower critical temperature. Although affording the ability to forage over a larger range, the disadvantage of this strategy over the long term will be an increase in energetic cost due to activity metabolism, at least relative to adults (Fig.

1).

In summary, swimming activity effectively compensated for the elevated demands in heat production with lower water temperatures in juvenile sea lions, by enabling the recycling of heat generated by activity to meet thermal energetic demands. The heat generated by locomotor thermogenesis mitigated thermoregulatory costs, even at relatively low levels of activity. At high levels of activity, sea lions exhibited metabolic rates lower than predicted by hydrodynamic and thermal principles alone. Sea lions also demonstrate a high level of physiological control over their insulation, which allows energetic costs to be minimized over a

142 wide range of water temperatures and activity levels. Together, the wide range of thermal tolerances exhibited by adult California sea lions and the ability of juvenile sea lions to utilize thermal substitution indicate a level of physiological competence that provides the potential for this species to persist even through times of marked climate change.

Acknowledgements

I wish to thank the Brookfield Zoo of Brookfield, IL for the use of its sea lions, as well as support of the research. I would also like to thank Terrie Williams,

Shane Kanatous, Randy Davis, and Ian Boyd for their work with the juvenile sea lions; and Scripps Institution of Oceanography, UCSD for support of the research. I thank Brett Long, Traci Fink, Jen Gafney, and the volunteers of UCSC Long Marine

Laboratory for training and animal care assistance. In addition, I wish to thank S.

Guest, K. Johnson, M. Rutishauser, and S. Seganti for laboratory assistance, as well as C. Kuhn and members of the Costa and Williams labs at UCSC for thoughtful feedback on the manuscript. I would also like to express gratitude to Terrie Williams and Dan Costa for their mentorship and support throughout the project. This study was funded by the Alaska SeaLife Center, the Friends of Long Marine Laboratory, the Earl H. Meyers and Ethel M. Meyers Oceanographic and Marine Biology Trust, the American Cetacean Society Monterey Bay, the American Museum of Natural

History Lerner-Gray Fund, and SEASPACE. All animal care and use protocols were evaluated and approved by the University of California Santa Cruz Institutional 143 Animal Care and Use Committee. Captive animal work was conducted under a US

Fish and Wildlife permit to Terrie Williams.

144

Table 1. Body mass, resting metabolic rate, and lower critical temperatures of the California sea lions used in this study. Five adult female and four juvenile (3 male, 1 female) California sea lions were examined. Body mass is the average for the experimental period. Resting metabolic rate is the average level measured within the thermal neutral zone for adults, and the measurement at 20 C for juveniles. Lower critical temperatures for the adult animals were determined using segmented regression analysis (see text for details). The lower critical temperatures of juvenile animals occurred between 12 C and 20 C.

145

Animal Body Mass (kg) Resting Metabolic Rate Lower Critical -1 (mean ± 1SD) (mLO2 · min · kg) Temperature (C) Adult 1 99.8 ± 1.9 6.16 ± 0.79 5.3 Adult 2 82.2 ± 0.9 7.13 ± 0.70 5.1 Adult 3 95.3 ± 1.2 6.88 ± 0.23 5.4 Adult 4 107.2 ± 6.6 5.55 ± 0.75 5.9 Adult 5 88.8 ± 1.5 6.10 ± 0.49 10.4 Juv 1 (♂) 36.5 ± 2.6 5.09 ± 1.26 > 12 Juv 2 (♂) 35.6 ± 0.8 5.84 ± 0.89 > 12 Juv 3 (♂) 31.9 ± 0.4 9.74 ± 1.06 > 12 Juv 4 (♀) 26.2 ± 0.1 8.56 ± 0.26 > 12

146

Figure 1. Resting metabolism and core body temperature in relation to water temperatures for mature (a,c) and immature (b,d) California sea lions. In (a) each point represents the average of 2 - 3 independent trials for the same animal at similar temperatures, with error bars indicating ± 1SD. A different symbol is used for each individual. Lines represent the average results of segmented regression analysis and are presented in the text. The temperature at the breakpoint between the two lines indicates the lower critical temperature for adult sea lions. In (b) each point represents a single independent trial for each juvenile sea lion, with error bars indicating ± 1SD for individual temperatures. Different symbols are used for different individuals. The solid line represents the best fit linear regression for the data and is described in the text. Dashed lines represent the average results for adult female California sea lions, for comparison. Core body temperature for adult (c) and juvenile (d) sea lions was measured immediately following or during each metabolic trial, respectively. Points in (c) represent individual post-trial measurements; in (d) mean values ± 1SD for each trial are represented by points and error bars. Each line denotes the least squares linear regression as described in the text.

) 1 - a b 14 14

·kg Juv1

1 Adult1 - Adult2 Juv2

12 Adult3 12 Juv3 ·min 2 Adult4 Juv4 Adult5 10 10

8 8

6 6

4 4

2 2

RestingMetabolic Rate (mLO 0 5 10 15 20 0 5 10 15 20

Water Temperature (C) Water Temperature (C)

40 c 40 d

39 39

C)

 38 38

37 37

36 36

35 35

CoreTemperature ( 34 34

33 33 0 5 10 15 20 0 5 10 15 20

Water Temperature (C) Water Temperature (C)

Figure 2. Metabolic rate in relation to temperature for juvenile California sea lions at rest (white circles) and while swimming at 1 m/s (gray squares). Symbols and lines indicate mean values ± 1SD. Note that resting metabolic rate decreased with increasing water temperature, while swimming metabolic rate did not differ significantly among the water temperatures.

149

14

)

1 -

12

min

ּ

1

-

kg ּ 2 10

8

6 Metabolic Rate (mLO Rate Metabolic

4 0 5 10 15 20 25

Water Temperature (°C)

150

Figure 3. Metabolic heat production (W) in relation to calculated swim velocity for four juvenile California sea lions. Three water temperatures are included: 5 ºC (filled circle), 12 ºC (open circle), and 20 ºC (downward facing triangle). Points represent mean values  1SD (error bars), or single measurements (no error bars). The line represents the least squares exponential relationship through the data points for swimming metabolism and is described in the text.

151

400

350

300

250

200

150

100

Metabolic Heat Production (W) Production Heat Metabolic 50

0 0 1 2 3 4 5

Swim Velocity (m/s)

152

Figure 4. The effects of body mass on swimming metabolic rates for four juvenile California sea lions. Each point represents an individual swimming trial. Solid lines indicate least squares linear regressions for individual sea lions, as described in the text. Metabolic rate increased linearly with load level for all animals. However, the slope of the regressions differed among the sea lions. Individual slopes are significantly different, except for Juv3 (due to high variation in the data).

153

40

)

1

-

min

35 ּ

1

-

kg

ּ 2 30

25

20

15 –– Juv1 (34.6-38.3 kg) –– Juv2 (36.2-35.1 kg) 10 –– Juv3 (31.6-32.2 kg)

Swimming Metabolic Rate (mLO Rate Metabolic Swimming –– Juv4 (26.1-26.2 kg)

5

0 1 2 3 4 5 6

Load (kg) at 1 m/s

154

Figure 5. Comparison of predicted (open circles) and measured (filled circles) metabolic heat production in relation to calculated swim velocity for juvenile sea lions. Predicted rates are based on the model proposed by Hind and Gurney (1997). The solid line is the least-squares exponential relationship for measured swimming metabolism as described in the text; the dashed line is the least-squares exponential relationship for the predicted swimming metabolism of the animals.

155

1200

1000

800

600

400

Metabolic Heat Production (W) Heat Production Metabolic 200

0 0 1 2 3 4 5 Swim Velocity (m/s)

156 References

Arnould, J. P. Y., Green, J. A. and Rawlins, D. R. (2001). Fasting metabolism in Antarctic fur seal (Arctocephalus gazella) pups. Comp. Biochem. Physiol. A 129, 829-841.

Bartholomew, G. A. (1977). Body temperature and energy metabolism. In Animal Physiology: Principles and Adaptations (ed. M.S. Gordon), pp. 364-449. New York, NY: Macmillan.

Boyd, I. L., Woakes, A. J., Butler, P. J., Davis, R. W. and Williams, T. M. (1995). Validation of heart rate and doubly-labeled water as measures of metabolic rate during swimming in California sea lions. Funct. Ecol. 9, 151-160.

Bryden, M. M. and Molyneux, G. S. (1978). Arteriovenous anastamoses in the skin of seals II. The California sea lion Zalophus californianus and the northern fur seal Callorhinus ursinus (Pinnipedia: Otariidae). Anat. Rec. 191, 253-260.

Costa, D. P., and Kooyman, G. L. (1982). Oxygen consumption, thermoregulation, and the effect of fur oiling and washing on the sea otter, Enhydra lutris. Can. J. Zool. 60, 2761-2767.

Costa, D. P., and Kooyman, G. L. (1984). Contribution of specific dynamic action to heat balance and thermoregulation in the sea otter Enhydra lutris. Physiol. Zool. 57, 199-203.

Davis, R. W., Williams, T. M. and Kooyman, G. L. (1985). Swimming metabolism of yearling and adult harbor seals Phoca vitulina. Physiol. Zool. 58, 590-596.

Dawson, T. J. and Fanning, F. D. (1981). Thermal and energetic problems of semiaquatic mammals: a study of the Australian water rat, including comparisons with the platypus. Physiol. Zool. 54, 285-296.

Dejours, P. (1987). Water and air. Physical characteristics and their physiological consequences. In Comparative Physiology: Life in Water and on Land (ed. P. Dejours, L. Bolis, C.R. Taylor and E.R. Weibel), pp. 3-11. New York: Springer-Verlag. di Prampero, P. E., Pendergast, D. R., Wilson, D. W. and Rennie, D. W. (1974). Energetics of swimming in man. J. Appl. Physiol. 37, 1-5.

157 Donohue, M. J., Costa, D. P., Goebel, M. E. and Baker, J. D. (2000). The ontogeny of metabolic rate and thermoregulatory capabilities of northern fur seal, Callorhinus ursinus, pups in air and water. J. Exp. Biol. 203, 1003-1016.

Fedak, M. A., Rome, L., and Seeherman, H. J. (1981). One-step nitrogen dilution technique for calibrating open-circuit oxygen volume flow rate measuring systems. J. Appl. Physiol. 51, 772-776.

Feldkamp, S. D. (1987). Swimming in the California sea lion: morphometrics, drag and energetics. J. Exp. Biol. 131, 117-135.

Feldkamp, S. D., Delong, R. L. and Antonelis, G. A. (1989). Diving patterns of California sea lions, Zalophus californianus. Can. J. Zool. 67, 872-883.

Fish, F. E. (1979). Thermoregulation in the muskrat (Ondatra zibethicus): the use of regional heterothermia. Comp. Biochem. Physiol. 64A, 391-397.

Gallivan, G. and Ronald, K. (1979). Temperature regulation in freely diving harp seals, Phoca groendlandica. Can. J. Zool. 57, 2256-2263.

Hart, J. S. and Irving, L. (1959). The energetics of harbor seals in air and in water with special consideration of seasonal changes. Can. J. Zool. 37, 447-457.

Heath, C. B. (1989). The behavioral ecology of the California sea lion, Zalophus californianus. Ph.D. thesis, University of California, Santa Cruz, California. 255pp.

Hind, A. T. and Gurney, W. S. C. (1997). The metabolic cost of swimming in marine homeotherms. J. Exp. Biol. 200, 531-542.

Hodgson, D. R., Davis, R. E. and McConaghy, F. F. (1994). Thermoregulation in the horse in response to exercise. Br. Vet. J. 150, 211-213.

Irving, L. and Hart, J. S. (1957). The metabolism and insulation of seals as bare- skinned mammals in cold water. Can. J. Zool. 35, 497-511.

Kleiber, M. (1975). The Fire of Life: An Introduction to Animal Energetics. Huntington, N.Y.: R.E. Kreiger Publishing Co. 454 pp.

Lovvorn, J. R. (2007) Thermal substitution and aerobic efficiency: measuring and predicting effects of heat balance on endotherm diving energetics. Phil. Trans. R. Soc. B 362, 2079-2093.

158 Mauck, B., Bilgmann, K., Jones, D. D., Eysel, U. and Dehnhardt, G. (2003). Thermal windows on the trunk of hauled-out seals: hot spots for thermoregulatory evaporation? J. Exp. Biol. 206, 1727-1738.

Melin, S. R., DeLong, R. L. and Thomason, J. R. (2000). Attendance patterns of California sea lion (Zalophus californianus) females and pups during the non- breeding season at San Miguel Island. Mar. Mamm. Sci. 16, 169-185.

Nickerson, D. M., Facey, D. E. and Grossman, G. D. (1989). Estimating physiological thresholds with continuous two-phase regression. Physiol. Zool. 62, 866-887.

Ponganis, P. J., Ponganis, E. P., Ponganis, K. V., Kooyman, G. L., Gentry, R. L. and Trillmich, F. (1990). Swimming velocity in otariids. Can. J. Zool. 68, 2105-2112.

Rosen, D. A. S., Winship, A. J. and Hoopes, L. A. (2007). Thermal and digestive constraints to foraging behaviour in marine mammals. Phil. Trans. R. Soc. Lond. B 362, 2151-2168.

Rutishauser, M. R., Costa, D. P., Goebel, M. E. and Williams, T. M. (2004). Ecological implications of body composition and thermal capabilities in young Antarctic fur seals (Arctocephalus gazella). Physiol. Biochem. Zool. 77, 669-81.

Schmidt-Nielsen, K. (1997). Animal Physiology: Adaptation and Environment. New York: Cambridge University Press.

Scholander, P. F., Irving, L. and Grinnell, S. W. (1942). Aerobic and anaerobic changes in seal muscle during diving. J. Biol. Chem. 142, 431-440.

Scholander, P. F., Walters, V., Hock, R. and Irving, L. (1950). Body insulation of some arctic and tropical mammals and birds. Biol. Bull. 99, 225-236.

Sokal, R. R. and Rohlf, F. J. (1981). Biometry. San Francisco: Freeman.

Williams, T. M. (1983). Locomotion in the North American mink, a semi-aquatic mammal. I. Swimming energetics and body drag. J. Exp. Biol. 103, 155-168.

Williams, T. M. (1986). Thermoregulation in the North American mink during rest and activity in the aquatic environment. Physiol. Zool. 59, 293-305.

159 Williams, T. M. (1999). The evolution of cost-efficient swimming in marine mammals: limits to energetic optimization. Phil. Trans. R. Soc. Lond. B 354, 193-201.

Williams, T. M., Ben-David, M., Noren, S., Rutishauser, M., McDonald, K. and Heyward, W. (2002). Running energetics of the North American river otter: Do short legs necessarily reduce efficiency on land? Comp. Biochem. Physiol. A 133, 203-212.

Williams, T. M., Davis, R. W., Fuiman, L. A., Francis, J., Le Boeuf, B. J., Horning, M., Calambokidis, J. and Croll, D. A. (2000). Sink or Swim: Strategies for Cost-Efficient Diving by Marine Mammals. Science 288, 133- 136.

Williams, T. M., Haun, J., Davis, R. W., Fuiman, L. A., and Kohin, S. (2001). A killer appetite: metabolic consequences of carnivory in marine mammals. Comp. Biochem. Phys. A 129, 785-796.

Williams, T. M., Kooyman, G. L. and Croll, D. A. (1991). The effect of submergence on heart rate and oxygen consumption of swimming seals and sea lions. J. Comp. Physiol. B 160, 637-644.

Williams, T. M. and Worthy, G. A. J. (2002). Anatomy and physiology: the challenge of aquatic living. In Marine Mammal Biology: An Evolutionary Approach (ed. A.R. Hoelzel), pp. 73-97. Malden, MA: Blackwell Science Ltd: Malden.

Willis, K., Horning, M., Rosen, D. A. S. and Trites, A. W. (2005). Spatial variation of heat flux in Steller sea lions: evidence for consistent avenues of heat exchange along the body trunk. J. Exp. Biol. 315, 163-175.

Withers, P. C. (1977). Measurement of oxygen consumption, carbon dioxide production, and evaporative water loss with a flow-through mask. J. Appl. Physiol. 42, 120-123.

Yeates, L. C. (2006) Physiological capabilities and behavioral strategies for marine living by the smallest marine mammal, the sea otter (Enhydra lutris). PhD dissertation, University of California Santa Cruz, USA.

Yeates, L. C., Williams, T. M. and Fink, T. L. (2007). Diving and foraging energetics of the smallest marine mammal, the sea otter (Enhydra lutris). J. Exp. Biol. 210, 1960-1970.

160 Zar J. H. (1974). Biostatistical Analysis. Englewood Cliffs, N.J.: Prentice-Hall.

161 CHAPTER 3

Energetic costs and thermoregulation in

northern fur seal (Callorhinus ursinus) pups:

The importance of behavioral strategies for

thermal balance in furred marine mammals

Abstract

Behavioral thermoregulation represents an important strategy for reducing energetic costs in thermally challenging environments, particularly among terrestrial vertebrates. Because of the cryptic lifestyle of aquatic species, the energetic benefits of such behaviors in marine endotherms have been much more difficult to demonstrate. In this study I examined the importance of behavioral thermoregulation in the northern fur seal (Callorhinus ursinus) pup, a small-bodied endotherm that spends prolonged periods at sea. Thermal capabilities were determined from the thermal neutral zone of three weaned male northern fur seal pups (body mass range =

11.8 - 12.8 kg) by measuring resting metabolic rate using open-flow respirometry, at water temperatures ranging from 2.5 to 25.0 °C. Metabolic rate averaged 10.03 ±

-1 -1 2.26 mlO2·kg ·min for pups resting within their thermal neutral zone; lower critical temperature was 8.3 ± 2.5 °C, approximately 8 °C higher than sea surface temperatures routinely encountered in northern Pacific waters. To determine whether behavioral strategies could mitigate this potential thermal limitation, I measured 162 metabolic rate during grooming activities and the unique jughandling behavior of fur seals. Both sedentary grooming and active grooming resulted in significant increases in metabolic rate relative to rest (F=29.31, P=0.001), and percent time spent grooming increased significantly at colder water temperatures (F=36.092, P<0.001).

-1 -1 Jughandling metabolic rate (12.71  2.73 mlO2·kg ·min ) was significantly greater than resting rates at water temperatures within the thermal neutral zone (F=25.86,

P<0.05), but less than resting metabolic rate at colder water temperatures. These data indicate that behavioral strategies, while representing an energetic cost to northern fur seal pups, may help to mitigate thermal challenges while resting at sea through a mechanism of thermal substitution.

Introduction

The high thermal conductivity and heat capacity of water compared to air makes the ocean especially thermally challenging to endotherms (Dejours, 1987).

The result of this thermal challenge to marine endotherms is often manifested as a metabolic response. Numerous studies have shown elevated resting metabolic costs for aquatic and semi-aquatic mammals (Fish, 1979; Dawson and Fanning, 1981;

Costa and Kooyman, 1982; Williams, 1986; Williams et al., 2001). While raising metabolic rate is an effective mechanism by which to counteract heat loss, it represents an added energetic expense to the animal if employed over the long term.

In terrestrial environments, the importance of behavioral strategies for maintaining a favorable body temperature and reducing the energetic costs of 163 thermoregulation has been demonstrated for a wide variety of endotherms.

Thermoregulatory behaviors include postural changes, the selection of microhabitats, and huddling behavior (reviewed by Hafez, 1964). These behaviors often have energetic benefits, and can impart a significant metabolic savings by reducing thermoregulatory costs and promoting survival in harsh environments (Sealander,

1952; Le Maho et al., 1976; Pinshow et al., 1976; Bazin and MacArthur, 1992; Perret,

1998; McKechnie and Lovegrove, 2001; Long et al., 2005; Gilbert et al., 2007). One would expect in marine systems, where endotherms are faced with unique thermal challenges, that similar behavioral strategies may be used to offset the costs of thermoregulation. As such, behavioral modification may represent an important strategy for reducing thermoregulatory costs in the ocean. However, due to the cryptic habits of marine endotherms, the extent to which behavioral thermoregulation is utilized and its relation to overall energetic costs or savings are unknown for most species.

One strategy for offsetting thermoregulatory costs is the recycling of metabolic heat produced by activity, termed thermal substitution. The concept of thermal substitution of heat produced by exercise is not new, and has been demonstrated in several species of birds and rodents (reviewed by Lovvorn, 2007). In marine systems, an increase in metabolic heat production associated with swimming exercise has been demonstrated for sea otters, pinnipeds and cetaceans (Costa and

Kooyman, 1982, 1984; Williams, 1999; Mostman-Liwanag, Chapter 2). This exercise-induced increase in metabolism provides a thermal energetic benefit for a 164 variety of aquatic and semi-aquatic mammals including mink (Williams, 1983), muskrats (Fish, 1979), sea otters (Costa and Kooyman, 1982, 1984; Yeates, 2006), and sea lions (Mostman-Liwanag, Chapter 2).

The use of thermal substitution to mitigate thermal stress has also been demonstrated for other at-sea behaviors. Grooming behavior has been shown to be an important part of the thermal budget of the sea otter (Costa and Kooyman, 1982;

Yeates, 2006). The sea otter also uses the metabolic increase associated with newly ingested food (heat increment of feeding, HIF) for thermoregulation (Costa and

Kooyman, 1984). It has been argued that the sea otter requires these behavioral strategies to maintain thermal balance because of its small body size and lack of a blubber layer (Costa and Kooyman, 1982, 1984; Yeates et al., 2007).

While at sea, fur seal species display an unusual posture known as the jughandle position, floating with both hind flippers and one foreflipper tucked and held above the water surface in an arc (Fig. 1). The purpose of this behavior is largely unknown, but popular hypotheses indicate opposing thermoregulatory consequences: the animals may remove the flippers from the water to avoid excessive heat loss to cold water, or they may position the flippers in the air to increase convective heat loss when warm (Bartholomew and Wilke, 1956). Because the physiological consequences of jughandling behavior have yet to be measured, the significance of this behavior for thermoregulation in fur seals has not yet been determined.

165 Northern fur seals (Callorhinus ursinus L.), like sea otters, rely primarily on dense, waterproof fur for thermal insulation. The pups of this species wean at four months of age, at a body size approximately 85% that of an adult sea otter. Upon weaning, northern fur seal pups remain at sea on a pelagic migration lasting up to nine months (Gentry and Holt, 1986; Reeves et al., 2002). Both their small body size and immature physiology make fur seal pups especially vulnerable to thermal stress in the water (Donohue et al., 2000; Rutishauser et al., 2004).

In view of these physiological limitations, I examined how behavioral strategies might help to mitigate thermal challenges encountered by northern fur seal pups as they transition to a life at sea. This was accomplished by assessing the metabolic responses of weaned northern fur seal pups to different water temperatures.

By determining the size of the thermal neutral zone (TNZ) and comparing the results to other marine mammals, I was able to quantify the relative thermal challenge experienced by post-weaning fur seal pups. To evaluate the importance of behavioral strategies for thermoregulation, I measured the energetic consequences of resting, jughandling positions, and grooming behaviors on thermal metabolic responses in northern fur seals. Together these tests provided quantitative measurements of the energetic costs and benefits associated with thermoregulatory behaviors in a small- bodied marine endotherm.

166 Materials and Methods

Animals

Three weaned, male northern fur seal pups were used in this study (Table 1).

The animals were acquired by the Marine Mammal Center (TMMC) in Sausalito, CA, at approximately six months of age (two months past the average weaning date for these animals), rehabilitated, and subsequently transferred to Long Marine Laboratory

(UC Santa Cruz) for research. Fur seals were fully rehabilitated and determined by veterinary staff to be in releasable condition prior to experimentation. Fur seals were maintained in seawater pools and fed a daily diet of herring, capelin, and squid supplemented with vitamins. Water temperatures in the holding pools reflected ambient coastal water conditions and averaged 14.2 °C (range: 10.7 – 17.9 °C) during the experimental period.

Protocol

For metabolic measurements, fur seals were conditioned over several weeks to enter a seawater pool (2.3 m × 2.3 m wide × 1.0 m deep) of known water temperature, over which a Plexiglas metabolic hood (approx. 114 cm wide × 175 cm long × 25 cm high) was mounted. The animals’ movements were limited to an 84 cm × 86 cm basket (36 cm deep) comprised of PVC and netting, positioned directly beneath the metabolic hood. All three animals voluntarily and consistently entered the experimental apparatus for measurements. Water temperature was randomly varied between metabolic trials in approximately 2.5 °C increments, and ranged from 2.5 °C 167 to 25.0 °C. Temperature of the seawater was adjusted with ice or heated saltwater, and was monitored throughout each trial. An experimental session consisted of a 60 -

120 min data acquisition period under the metabolic dome, with each animal tested only once per day. Behaviors were continuously monitored and recorded throughout each trial, to synchronize metabolic rates with behavioral events.

Oxygen consumption

Metabolic rate was determined from oxygen consumption measured by open flow respirometry, using the protocols of Williams et al. (2002). Fur seals were fasted overnight (≥ 12 hours), to ensure postabsorptive status at the time of measurement (Kleiber, 1975), except for trials specifically assessing postprandial metabolism. Ambient air was drawn through the Plexiglas chamber at 60 L·min-1, with flow rates maintained and monitored continuously by a mass flow controller

(Flowkit 500, Sable Systems International, Inc., Las Vegas, NV). Samples of air from the exhaust port of the chamber were dried (Drierite; W.A. Hammond Drierite

Co., Xenia, OH) and scrubbed of carbon dioxide (Sodasorb; Chemetron, St. Louis,

MO) before entering an oxygen analyzer (FC-1, Sable Systems International). The percentage of oxygen in the expired air was monitored continuously and recorded once per second with a personal computer (Toshiba or Acer laptop) using Sable

Systems software. Oxygen consumption (VO2) was then calculated using ExpeData software (Sable Systems International, Inc., Las Vegas, NV), modifying equation 4b from Withers (1977) for this experimental setup. An assumed respiratory quotient of 168 0.77 was used in the calculations (Feldkamp, 1987; Boyd et al., 1995; Arnould et al.,

2001). Lastly, all values were corrected to STPD. The entire system was calibrated daily with dry ambient air (20.94% O2) and every 3 - 4 days with dry span gases

(16.0% O2) and N2 gas according to Fedak et al. (1981). In addition, the theoretical fraction of O2 leaving the chamber was calculated from Davis et al. (1985) and compared to measured values from the analyzer.

Determination of the thermal neutral zone

The thermal neutral zone (TNZ) is defined as the range of environmental temperatures in which an animal does not have to increase its metabolism above resting levels to maintain body temperature (Bartholomew, 1977). Below the lower critical temperature (TLC) or above the upper critical temperature (TUC), an animal increases metabolism (and thus energetic cost) to offset heat loss or facilitate heat dissipation, respectively. The TLC of each fur seal was determined from the breakpoint for the relationship between mass-specific resting metabolic rate (RMR) and water temperature, from a segmented regression analysis minimizing the sum of the residual sum of squares (Sokal and Rohlf, 1982; Nickerson et al., 1989). A computer program created for the determination of critical points (Yeager and Ultsch,

1989) was used. The breakpoint between the two lines of best fit defined the lower critical temperature. TLC was determined separately for each individual (Table 1).

169 Behavioral analyses

Behaviors during each metabolic trial were classified as resting, jughandling, grooming, or other (active), for the entire duration of the metabolic trial. Time spent performing each behavior was calculated as a percentage of total time for each metabolic trial. Behaviors such as shivering or sleeping, which were performed during another categorized behavior (i.e., resting or jughandling), were also noted.

Resting was defined as the animal floating quiescently with its flippers in the water while awake. Resting metabolism used for TNZ assessments was determined from fasting metabolic trials in which distinguishable resting bouts occurred. The lowest continuous oxygen consumption maintained for 5 – 10 min and corresponding to quiescent behavior was used in the analyses.

Because the ingestion and digestion of food increases metabolic rate

(Webster, 1983), two separate metabolic trials per fur seal were conducted immediately following 0.5 – 1.5 kg meals of herring, using the same experimental protocol as described above. Fur seals were fasted overnight (≥ 12 hours) prior to the time of feeding. All feeding trials were conducted in water temperatures within the thermal neutral zone (20.2 ± 0.3 °C). The lowest oxygen consumption rate maintained for 10 min and corresponding to quiescent behavior was used to determine postprandial resting metabolic rate.

Jughandling was defined as the animal floating with both hind flippers and one foreflipper held above the water surface in an arc (Fig. 1). Jughandling metabolism was analyzed for metabolic trials in which the animal consistently 170 maintained the jughandle position for 30 minutes or more. Because the metabolic response of the animal changed during this behavior (Fig. 2), jughandling metabolic rate (JMR) was defined as the lowest, consistent oxygen consumption maintained for

10 min. JMR was calculated only for periods during which the animal was awake.

Grooming was defined as behavior directed toward maintenance or restoration of the air layer in the fur, and included both rolling on the surface of the water and rubbing the fur with the flippers. Grooming was considered “active” if the animal was rolling or otherwise changing positions. Grooming was classified as “sedentary” if the animal was rubbing the fur with its flippers, but otherwise maintaining a resting position on the water’s surface. Grooming metabolism was determined for continuous grooming bouts lasting 5 – 10 min. Sedentary and active grooming were combined into a single category for activity budgets. Active periods not associated with grooming were not included in the metabolic analyses.

Statistical analyses

All numerical values for oxygen consumption are presented as mean  1SD.

Jughandling metabolic rate (JMR) was determined from the lowest level oxygen consumption during this behavior, and served as the baseline for further jughandling analyses (Fig. 2). The total energetic cost of jughandling was determined by integrating under the oxygen consumption-time curve (ExpeData, Sable Systems

International), and subtracting the baseline JMR for that trial. Thus, the resulting value represents the added energetic cost (in mlO2) associated with the jughandle 171 3 position. This value was converted to Joules, assuming 20.1×10 J per LO2 (Schmidt-

Nielsen, 1997).

Statistical significance among levels of metabolism were determined by 1- way ANOVA using a Bonferroni pairwise multiple comparison test (Systat Software,

Inc., Chicago, IL). Linear relationships between metabolic rates and water temperatures, and between activity budgets and water temperatures, were calculated by least-squares procedures (Zar, 1974). Relative durations of individual activities performed during postabsorptive and postprandial trials were compared using

Student’s t tests.

Results

Resting Metabolism and Thermal Neutral Zone

Individual lower critical temperatures (TLC) ranged from 6.6 to 11.1 °C (Table

1). The mean TLC for the fur seal pups was 8.3 ± 2.5 °C (N=3 pups; Fig. 3; Table 1).

This indicates a total thermal neutral zone (TNZ) of at least 16 ºC, as metabolic rate did not change significantly between 8.3 ºC and 24.3 ºC (F1,22=0.172, P=0.682).

The average metabolic rate of weaned northern fur seal pups resting in water

-1 -1 within the TNZ was 10.03  2.26 mlO2·kg ·min (N=22 trials; Table 1). Below the

TNZ, resting metabolism increased as water temperature decreased, with individual fur seals exhibiting a 27 – 168% increase in metabolism at the lowest experimental water temperatures (Fig. 3). The average linear increase in RMR below the TNZ was described by the equation 172

2 RMR = 20.74 – 1.16(Twater) (P=0.180, r =0.067) (1)

-1 -1 where RMR is in mlO2kg min and Twater is in ºC.

Postprandial resting metabolic rate was measured for two of the three pups,

-1 -1 and averaged 16.25  2.99 mlO2·kg ·min (N=4 trials). This value was significantly higher than RMR within the TNZ (F4,78=29.312, P=0.01; Table 1; Fig. 4). The ingestion of fish did not elicit a defined peak in metabolic rate as previously reported for sea otters (Costa and Kooyman, 1984) and pinnipeds (Markussenet al., 1994;

Rosen and Trites, 1997). Rather, feeding induced an increase in RMR relative to fasting levels within the TNZ, even at thermally neutral water temperatures. An average intake of 1.0  0.6 kg herring elicited an increase in resting energy output above baseline RMR levels that represented 9.7  0.9% of energy intake. Note that this value does not represent the full HIF, but rather indicates the average resting metabolic increase elicited by prey ingestion in these animals.

Jughandling Metabolism

Two of the three fur seals exhibited jughandling behavior under the metabolic dome. For both animals, jughandling behavior elicited a characteristic metabolic response, in which metabolic rate first peaked and then decreased in an approximately first-order recovery curve (Fig. 2). On average, this metabolic elevation lasted

173 approximately 30 min before reaching asymptote and represented a 103% increase over the equilibrium jughandling metabolism at the zenith. JMR was calculated from the asymptotic portion of the recovery curve.

No breakpoint could be defined for the relationship between JMR and water temperature, and JMR did not change significantly across the range of experimental water temperatures (F1,15=0.868, P=0.366). One pup jughandled across nearly the entire range of experimental temperatures (FS1, Twater = 5.4 – 25.0 ºC), while one jughandled at warmer temperatures (FS2, Twater = 15.6 – 22.8 ºC). The average JMR

-1 -1 across all water temperatures was 12.71  2.73 mlO2·kg ·min (N=18 trials), and was significantly higher than RMR within the TNZ (F4,68=25.860, P<0.05; Table 1; Fig.

4). This result is consistent with a previous report of elevated JMR relative to RMR in pre-weaned northern fur seal pups (Donohue et al., 2000).

Integration under the jughandling curves revealed that the animals expended

42.5  33.8 kJ of additional energy above the equilibrium JMR, during an average of

48.7  14.8 min in the jughandle position. Once the metabolism reached equilibrium, jughandling still represented a 39% increase over RMR on average.

Postprandial jughandling metabolic rate was measured for one pup (FS1), and

-1 -1 averaged 16.73  0.73 mlO2·kg ·min (N=2 trials). Postprandial JMR was not significantly different from fasting JMR for this animal (t=2.14, P=0.23; Table 1).

174 Grooming Metabolism

-1 -1 Metabolism during sedentary grooming, 15.15 ± 3.58 mlO2·kg ·min (N=11 trials), was significantly greater than RMR within the TNZ (F4,78=29.312, P=0.001;

Table 1; Fig. 4), and represented an increase of 1.3 – 2.3 times RMR for individual fur seals. Sedentary grooming metabolism did not change significantly with water temperature (F1,9=0.051, P=0.827; Twater = 3.1 – 17.5 ºC). In comparison, metabolism

-1 -1 during active grooming, 19.98 ± 4.32 mlO2·kg ·min (N=29 trials), was significantly greater than sedentary grooming metabolism (F4,78=29.312, P=0.001; Table 1; Fig. 3).

As found for sedentary grooming, active grooming metabolism did not change significantly with water temperature (F1,27=1.374, P=0.251; Twater = 2.9 – 24.3 ºC).

Effects of water temperature on behavior

Each fur seal exhibited individual behavioral patterns in response to water temperature during the metabolic trials. For all three fur seals, the percentage time resting did not change significantly with temperature (F1,19=0.035, P=0.853;

F1,15=0.041, P=0.841; F1,16=2.995, P=0.103). Percentage time jughandling increased significantly with temperature for one fur seal (FS1; F1,19=14.247, P=0.001), and also occurred primarily at warmer temperatures for another (FS2; F1,15=3.092, P=0.099).

All three fur seals exhibited a significant increase in percent time grooming as temperature decreased (F1,19=37.570, P<0.001; F1,15=7.068, P=0.018; F1,16=9.554,

P=0.007; Fig. 5); this was the primary behavioral pattern that was consistent among all three individuals. Conversely, other activities were more variable. For one fur 175 seal (FS3), activity increased with increasing temperature (F1,16=7.115, P=0.017). In contrast, the percentage time spent performing other activities did not change significantly across temperatures for the other two fur seals (FS1, F1,19=0.060,

P=0.810; FS2, F1,15=0.340, P=0.568). Shivering occurred either during resting or jughandling behavior, across the range of experimental temperatures, and did not change significantly with temperature for any of the fur seals (F1,12=0.115, P=0.740;

F1,7=0.184, P=0.681; F1,1=0.353, P=0.659).

Effects of feeding on behavior

Percent time resting did not change significantly between postprandial and fasting trials at warm water (20 – 25 ºC) for any of the pups (t=3.18, P=0.135; t=4.30,

P=0.656; t=3.18, P=0.981). Percent time jughandling decreased significantly during postprandial trials compared to warm water fasting trials for one fur seal (FS1; t=3.18, P=0.004), and this was accompanied by a concomitant increase in other activities (t=3.18, P=0.001). A similar trend was evident for the other fur seal that jughandled (FS2), but the results were not statistically significant (jughandling: t=4.30, P=0.389; activity: t=4.30, P=0.074). For the third fur seal (FS3), percent time active was not significantly different between postprandial and fasting warm water trials (t=3.18, P=0.811).

176 Discussion

Thermal capabilities of northern fur seal pups

During the transition to independence, young pinnipeds will eventually access the same range of oceanic foraging grounds as adult conspecifics. Several physiological challenges must be overcome for the young animals to successfully accomplish this transition. First, young pinnipeds do not exhibit the same physiological and behavioral diving capacities as adults. Limitations include higher mass-specific metabolic rates (Fowler et al., 2006; Richmond et al., 2006), lower oxygen stores (Horning and Trillmich, 1997; Noren et al., 2001; Burns et al., 2004;

Fowler et al., 2006; Noren et al., 2005; Richmond et al., 2006), and inexperience constraining dive behavior (Le Boeuf et al., 1996; Merrick and Loughlin, 1997;

Burns, 1999; Jørgensen et al., 2001; Fowler et al., 2006). A second factor comprises thermal energetic costs associated with immersion, which are greater in smaller, immature animals compared to adults. As a result of these factors, the range for foraging by pups may be limited at a critical point in their life history.

Thermal limitations have been reported previously for young otariids (Table

2). For example, the TLC of juvenile California sea lions (>12 ºC) is several degrees higher than the range of water temperatures typically encountered, indicating a thermal challenge for the young of this species (Mostman-Liwanag, Chapter 2).

Likewise, both pup and juvenile Antarctic fur seals demonstrate thermal energetic costs that potentially limit the range of foraging and may prevent these age classes from utilizing adult foraging grounds (Rutishauser et al., 2004). 177 I have also found that thermal limitations may occur in northern fur seal pups

(Table 2). While the molting of the natal pelage helps to increase thermal insulation of northern fur seal pups immediately prior to weaning, the pre-weaned pups still show elevated metabolic rates at water temperatures regularly encountered in the wild, even in coastal regions close to the rookery (Donohue et al., 2000). The present study shows that the thermal capabilities of weaned northern fur seal pups have changed relative to pre-weaned, post-molt pups. Resting metabolic rate within the thermal neutral zone (Table 1) was lower than what Donohue et al. (2000) reported for post-molt pups prior to weaning, and is consistent with the expected reduction in mass-specific metabolism with age (Kleiber, 1975). This effect appears to be due to age rather than body size per se, as the animals in this study were smaller than the pups in Donohue et al. (2000) (Table 2), though within the range of weaning masses for this species (Boltnev et al., 1998). In addition, the TLC of 8.3  2.5 C measured here for weaned pups is lower than estimated for pups prior to weaning (Table 2;

Donohue et al., 2000). Both of these changes indicate a developmental improvement in thermal capabilities as the animals approach independence.

The comparatively low TLC of northern fur seal pups indicates a remarkable tolerance for cold water temperatures despite the small body size of these young animals, particularly in comparison to other otariid species (Table 2). The TLC of weaned northern fur seal pups is also considerably lower than that of the sea otter, a similarly sized marine mammal that also relies on fur for its primary insulation.

Despite a denser pelage (Williams et al., 1992), adult female sea otters exhibit a TLC 178 approaching 20 C (Costa and Kooyman, 1984), more than 10 °C higher than measured here for northern fur seal pups. Overall, the relatively low RMR and TLC measured in the present study indicate that northern fur seal pups exhibit exceptional thermal capabilities for their young age and small body size.

Developmental improvements notwithstanding, the TLC of weaned northern fur seal pups remains higher than the routine sea surface temperatures found in the northern Pacific Ocean (Table 2), where newly weaned pups spend up to nine months diving in pelagic waters (Reeves et al., 2002). This has profound implications for rest periods at sea. While heat generated by activity may offset excessive heat loss when swimming, resting periods may easily result in a thermal deficit. For example, the fur seals in this study demonstrated increases in resting metabolism between 27% and

168% above baseline values at colder water temperatures (Fig. 3), suggesting that rest periods are likely to be energetically costly to these animals even within their natural water temperature range.

These high thermal energetic costs for northern fur seal pups may be associated with the use of fur as the primary form of insulation in water, as a similar thermal liability during rest is observed for many small-bodied, furred marine mammal species (Fig. 6). Here I define “thermal liability” as the difference between the animal’s TLC in water and the lowest sea surface temperature routinely encountered by the species and age class of interest. Comparing thermal liabilities among species reveals that the use of blubber as the primary form of insulation, along with an associated increase in body size, reduces the level of liability in young 179 pinnipeds and eliminates it for adult sea lions and phocids (Fig. 6). Thus, while sea otters and immature fur seals exhibit thermal liabilities of at least 8 C while resting at sea, immature harbor seals and California sea lions, which rely on blubber for their insulation, have thermal liabilities of only 5 C. Young harp seals, an ice-dwelling species that relies on a thick blubber layer for insulation, have a TLC below 0 C and thus no thermal liability even when living above the Arctic circle. Interestingly, adult sea lions and harp seals maintain negative thermal liabilities; in other words, the TLC is lower than the coldest sea surface temperatures encountered, thereby affording the animals the benefit of low thermal energetic costs during rest periods. Overall, these comparisons indicate a thermal advantage for larger-bodied marine mammals using blubber for insulation, compared to small-bodied marine mammals using fur.

The effects of behavioral strategies on thermal responses in northern fur seal pups

The use of behavioral strategies to mitigate thermal limitations has been demonstrated for terrestrial species (Hafez, 1964; Hart, 1971; Whittow, 1971) and for pinniped species on haulouts (Bartholomew and Wilke, 1956; Irving et al. 1962;

Peterson and Bartholomew, 1967; Peterson et al., 1968; Whittow et al., 1971; Gentry,

1972; Odell, 1974; Ohata and Miller, 1977b; Bonner, 1984; Trillmich, 1984; Fleisher,

1987; Pierson, 1987; Heath, 1989; Trites, 1990). However, far less is known about the behavioral strategies employed by marine mammals at sea. I have previously reported that young otariids utilize the recycling of heat generated by swimming

180 activity to offset thermal limitations during swimming (Mostman-Liwanag, Chapter

2). However, a question remains regarding thermal costs during periods of rest.

Coastal sea lions may haul out on shore during these periods (Reeves at al., 2002).

This is not possible for young northern fur seals, which are pelagic and must rest at sea. Consequently, the use of locomotor activity to offset excessive heat loss is not a sustainable strategy. To compensate, young fur seals utilize several other behavioral strategies.

Jughandling behavior

Thermoregulatory explanations for jughandling behavior suggest that removal of the flippers from the water may decrease heat loss in cold water or facilitate heat dissipation in warm water (Bartholomew and Wilke, 1956). The occurrence of jughandling behavior across the full range of experimental water temperatures in this study indicates that these hypotheses may not be mutually exclusive. The large, poorly insulated flippers of fur seals serve as important thermal windows – areas through which the animals can easily dissipate heat (Bryden and Molyneux, 1978).

Many species of marine mammal use poorly insulated areas as thermal windows, including sea otters (Tarasoff, 1972, 1974), otariids (Ohata and Miller, 1977a; Bryden and Molyneux, 1978), phocids (Molyneux and Bryden, 1978; Kvadsheim and

Folkow, 1997; Rommel et al., 1995), and cetaceans (Hampton and Whittow, 1976;

Meagher et al., 2000). Peripheral sites such as the flippers remain the only option for heat transfer in furred species such as fur seals (Ohata and Miller, 1977a; Bryden and 181 Molyneux, 1978; Schmidt-Nielsen, 1997). Because the flippers are highly vascularized and contain countercurrent arrangements of blood vessels (Tarasoff and

Fisher, 1970; Bryden and Molyneux, 1978), fur seals have control over the extent to which heat is moved across the flippers. In view of this, the thermoregulatory consequences of jughandling may depend largely on blood flow. With reduced blood flow to the flippers, heat loss in cold water will be minimized; conversely, jughandling coupled with perfusion of the flippers will facilitate heat loss when the animals are resting in warm water. Thus, the combination of physiological control of blood flow with the behavioral posture of jughandling allows fur seals to exhibit a dynamic response to a range of environmental temperatures.

Despite or perhaps because of the elevated metabolic costs associated with the jughandle position (Fig. 2), it appears that jughandling behavior affords an extension of the effective thermal neutral zone for northern fur seal pups. Equilibrium JMR is significantly higher than RMR and thus represents elevated costs within the TNZ

(Fig. 4). However, for the fur seal (FS1) that exhibited jughandling behavior in cold water, JMR was significantly lower than RMR at water temperatures below the TLC

(t=2.64, P=0.018). These findings indicate a potential energetic benefit associated with jughandling behavior at low water temperatures, in this case a savings of 45

J·min-1 at equilibrium. Because jughandling metabolic rate typically took at least 30 min to reach equilibrium, the energetic savings associated with this behavior require it to be used as a long-term strategy, conferring thermal benefits over time that may counterbalance the initial costs. 182

Grooming behavior

Grooming behavior restores the air layer to the fur, and is therefore necessary to ensure insulation in furred endotherms living in aquatic environments. As reported for sea otters (Costa and Kooyman, 1982; Yeates et al., 2007), I found a significant increase in metabolic rate associated with grooming behavior for northern fur seal pups (Fig. 4). In addition, grooming behavior represented a greater proportion of fur seals’ activity budgets at colder water temperatures (Fig. 5). Similar to the use of thermal substitution observed in exercising birds and mammals (Fish, 1979;

Williams, 1983; Lovvorn, 2007; Yeates et al., 2007; Mostman-Liwanag, Chapter 2), fur seals can use the increased heat production associated with grooming to mitigate thermal costs at sea. Thus, grooming behavior in furred marine mammals not only restores the effectiveness of the insulation, but also serves as an effective strategy to counteract heat loss in cold water with increased heat production.

The effects of ingested prey items on thermal responses in northern fur seal pups

Increases in metabolism associated with the ingestion and digestion of food have been demonstrated for a variety of vertebrates, including fish (Beamish and

MacMahon, 1988), reptiles (Andrade et al., 1997), birds (Hawkins et al., 1997;

Schieltz and Murphy, 1997), and mammals (Markussen et al., 1994; Rosen and

Trites, 1997; Nespolo et al., 2003). The importance of this heat increment of feeding

(HIF) for thermal substitution has been shown both for birds (Chappell et al., 1997; 183 Kaseloo and Lovvorn, 2003; Bech and Præsteng, 2004) and for mammals (Costa and

Kooyman, 1984; Jensen et al., 1999). For sea otters in particular, HIF is an important part of the thermal budget, enabling increased rest periods (Costa and Kooyman,

1984; Yeates et al., 2007). I did not observe the same change in behavior with feeding for northern fur seal pups; rather, the animals increased time spent performing other activities instead of resting following the ingestion of prey.

Feldman and Parrott (1996) suggested that fur seals increase grooming behavior following feeding, due to the presence of fish oils in the water. This was not observed in the present study (Fig. 5; t=3.18, P=0.745; t=4.30, P=0.573; t=3.18,

P=0.426). Although quantifiable changes in behavior were not as expected, all three fur seals qualitatively appeared more quiescent and reduced the intensity of activities following food ingestion. A possible explanation for the differences in these studies is that the meal size (0.5 – 1.5 kg herring) in the present study was not large enough to induce behavioral changes, despite significant metabolic changes.

Postprandial resting metabolic rate for the fur seals in this study was 58% higher than postabsorptive RMR, representing 9.7% of ingested energy. These values are consistent with the magnitudes reported for other marine mammals, including the sea otter (Costa and Kooyman, 1984), harbor seal (Markussen et al., 1994), and

Steller sea lion (Rosen and Trites, 1997). Interestingly, postprandial jughandling metabolism was not significantly different from postabsorptive jughandling metabolism for one fur seal (FS1, Table 1), indicating that, as with sea otters (Costa

184 and Kooyman, 1982), thermal substitution in fur seals functions in a synergistic manner rather than being simply additive.

Conclusions

In summary, behavioral thermoregulation represents a critical mechanism for ensuring thermal stability in small-bodied marine mammals that rely primarily on fur for insulation. The thermal energetic costs associated with immersion are greater in smaller, immature animals compared to adults, but from this species it appears that increases in the range of thermally neutral temperatures occur rapidly with age. In addition, the use of thermal substitution from a variety of sources, including swimming activity and digestion of food, can help to mitigate thermal costs in young fur seals. Although behavioral patterns vary considerably among individuals, grooming behavior consistently provides significant thermal benefits for furred marine species, particularly at colder water temperatures. Jughandling behavior as seen in fur seals may help to expand the overall TNZ. Despite the insulative effectiveness of fur in terrestrial mammals, the use of fur as a primary insulator in the marine environment appears to require the mitigation of thermal costs (Fig. 6).

Additionally, the grooming behavior required to maintain the insulative properties of fur in water is energetically costly (Fig. 4). Overall, the high cost of thermoregulation for furred mammals in the aquatic environment may have contributed to a progression towards the increased use of blubber in marine mammals over evolutionary time.

185 Acknowledgements

I wish to thank Frances Gulland and the Marine Mammal Center of Sausalito,

CA for assistance with the rehabilitation and release of the animals, as well as support of the research. I would also like to thank Traci Fink, Beau Richter, and the volunteers of UCSC Long Marine Laboratory for training and animal care assistance.

In addition, I wish to thank M. Abney, E. Bastiaans, A. Davis, J. Devlahovich, A.

Fahlman, K. Kunkel, L. Mandell, J. Maresh, K. McCully, J. Sapp, and M. Slater for laboratory assistance, as well as C. Kuhn for thoughtful feedback on the manuscript.

I would also like to express gratitude to Terrie Williams and Dan Costa for their mentorship and support throughout the project. This study was funded by the Alaska

SeaLife Center, the Friends of Long Marine Laboratory, SEASPACE, the STEPS

Institute at UCSC, Sigma Xi, the American Cetacean Society Monterey Bay, the

American Museum of Natural History Lerner-Gray Fund, the Society for Integrative and Comparative Biology, and the UCSC Institute for Geophysics and Planetary

Physics. All animal care and use protocols were evaluated and approved by the

University of California Santa Cruz Institutional Animal Care and Use Committee.

Captive animal work was conducted under a US Fish and Wildlife permit to Teri

Rowles.

186

Table 1. Body mass, lower critical temperatures, and metabolic rates for northern fur seal pups. Values indicate mean ±

1SD. Resting metabolic rate was measured within the thermal neutral zone; other metabolic rates are for the indicated behaviors. *Postprandial metabolic rate was calculated during jughandling behavior for FS1, and is not included in the mean for postprandial RMR.

-1 -1 Lower Metabolic Rate (mlO2·kg ·min ) Body Mass Critical Animal (kg) Temperature Resting Grooming Jughandling (C) Postabsorptive Postprandial Sedentary Active FS1 12.8  2.4 6.6 7.75  1.37 15.73  0.73* 13.09  2.87 17.59  1.21 17.83  4.53 FS2 11.8  2.2 7.2 9.15  1.18 14.01  0.61 10.93  0.74 12.33  1.39 21.03  3.45 FS3 12.5  2.9 11.1 11.37  2.51 18.49  2.53 not measured 14.96  3.47 22.95  2.65 Mean 12.4  0.5 8.3  2.5 10.03  2.26 16.25  2.99 12.71  2.73 15.15  3.58 19.98  4.32 (N=3) (N=3) (N=22) (N=4) (N=18) (N=11) (N=29)

Table 2. Lower critical temperatures (TLC) for sea otters and pinnipeds resting in water. Values are compared to the range of water temperatures typically encountered by each species in the wild. Body mass and TLC values are estimated from the sources indicated. Water temperatures are the routine sea surface temperatures encountered by each species and age class

(NASA MODIS database).

Average Water T Species Age Class Body LC Temperatures Source (ºC) Mass (kg) Encountered (ºC) Sea otter, Costa and Kooyman Adult 17.3 20 0 – 20 Enhydra lutris (1982) Northern fur seal, Pup 8.3 > 10 0 – 10 Donohue et al. (2000) Callorhinus ursinus (pre-weaning, pre-molt) Northern fur seal, Pup 14.8 ~ 10 0 – 10 Donohue et al. (2000) Callorhinus ursinus (pre-weaning, post-molt) Northern fur seal, Pup 12.4 8.3 0 – 17 This study Callorhinus ursinus (post-weaning) Antarctic fur seal, Pup 13.6 14.4 0 – 5 Rutishauser et al. (2004) Arctocephalus gazella Antarctic fur seal, Juvenile 17.3 14.4 0 – 10 Rutishauser et al. (2004) Arctocephalus gazella California sea lion, Mostman-Liwanag Juvenile 32.6 > 12 10 – 20 Zalophus californianus (Chapter 2) California sea lion, Mostman-Liwanag Adult 94.7 6.4 10 – 20 Zalophus californianus (Chapter 2) Harbor seal, 27.4 20 15 – 30 (summer) Juvenile Hart and Irving (1959) Phoca vitulina concolor 33.7 13 8 – 23 (winter) Harp seal, Juvenile 38.6 < 0 -2 – 5 Irving and Hart (1957) Phoca groenlandica Harp seal, Gallivan and Ronald Adult 140.7 < 0 -2 – 10 Phoca groenlandica (1979)

Figure 1. Northern fur seal pup in the jughandle position, with one foreflipper held between both hind flippers above the water. The other foreflipper is in the water, and is assumed to help stabilize the animal in this position.

191

192

Figure 2. Representative metabolic trace for a northern fur seal exhibiting the jughandling posture, illustrating the characteristic changes in metabolic rate associated with this behavior. The initial portion of the trace corresponds to mild activity after the animal entered the metabolic chamber. Once the animal began jughandling, metabolism increased above the level associated with activity, and then decreased in an approximately first-order recovery curve. The lowest, steady-state portion of the recovery curve was used to calculate jughandling metabolic rate (JMR), which then served as the baseline for integrating under the curve.

193

400

)

1 -

300

· min · 2

200 activity begin jughandling

100

etabolic rate (mlO rate etabolic M JMR

0 0 10 20 30 40 Time (min)

194

Figure 3. Resting metabolic rate in relation to water temperature for three northern fur seal pups. Each point represents the average of 1-3 independent trials for the same animal at similar temperatures, with error bars indicating ± 1SD. Different symbols are used for different individuals. Lines represent the average results of segmented regression analysis and are described in the text. The temperature at the breakpoint of the two lines represents lower critical temperature.

195

25

)

1

-

·min 1

- 20

·kg 2

15 tabolic rate (mlO rate tabolic

10 Resting me Resting

5 0 5 10 15 20 25 Water Temperature (C)

196

Figure 4. Metabolic rate according to behavior for northern fur seal pups.

Heights of the bars and lines indicate means ± 1SD. Letters above the bars indicate statistically significant differences among means. Resting metabolism is for values within the thermal neutral zone.

197

30

C

) 25

1 - B C

·min B 1

- 20 ·kg

2 B

15 A

10 Metabolic rate (mlO rate Metabolic 5

0 Resting Postprandial Jughandling Sedentary Active Resting Grooming Grooming (N=22) (N=4) (N=18) (N=11) (N=29)

198

Figure 5. Percent time spent grooming in relation to water temperature for three northern fur seal pups. Final bar (Postprandial) indicates time spent grooming after the ingestion of food in warm water (20.2 ± 0.3 °C). Numbers in parentheses above the bars indicate the number of independent trials included in each temperature bin.

199

100 (6)

80

60 (17)

(9) 40 (16)

Percent time spent grooming spent time Percent 20 (8) (6)

0 0-5 5-10 10-15 15-20 20-25 PostHIF - prandial Water Temperature (C)

200

Figure 6. Comparison of the thermal liability of sea otter and pinniped species.

Thermal liability is calculated as the difference between the lower critical temperature

(TLC) and the lowest sea surface temperature routinely encountered. Species are listed in order of increasing reliance on blubber relative to fur. Black bars indicate species with waterproof fur as the primary insulating layer, and gray bars indicate species with blubber as the primary insulating layer. Plain bars include adult animals, and dotted bars indicate pups or juveniles.

201

Primary Insulation Fur Sea otter (adult) Northern fur seal (pre-weaned pup)

Northern fur seal (post-weaned pup) Antarctic fur seal (pup and juvenile) California sea lion (juvenile) California sea lion (adult) Harbor seal (juvenile) Harp seal (juvenile and adult)

Blubber -5 0 5 10 15

Thermal Liability (ºC)

202 References

Andrade, D. V., Cruz-Neto, A. P. and Abe, A. S. (1997). Meal size and specific dynamic action in the rattlesnake Crotalus durissus (Serpentes: Viperidae). Herpetologica 53, 485-493.

Arnould, J. P. Y., Green, J. A. and Rawlins, D. R. (2001). Fasting metabolism in Antarctic fur seal (Arctocephalus gazella) pups. Comp. Biochem. Physiol. A 129, 829-841.

Bartholomew, G. A. (1977). Body temperature and energy metabolism. In Animal Physiology: Principles and Adaptations (ed. M. S. Gordon), pp. 364-449. New York: Macmillan.

Bartholomew, G. A. and Wilke, F. (1956). Body temperature in the northern fur seal, Callorhinus ursinus. J. Mammal. 37, 327-337.

Bazin, R. C. and MacArthur, R. A. (1992). Thermal benefits of huddling in the muskrat (Ondatra zibethicus). J. Mammal. 73, 559-564.

Beamish, F. W. H. and MacMahon, P. D. (1988). Apparent heat increment and feeding strategy in walleye, Stizostedion vitreum vitreum. Aquaculture 68, 73- 82.

Bech, C. and Præsteng, K. E. (2004). Thermoregulatory use of heat increment of feeding in the tawny owl (Strix aluco). J. Therm. Biol. 29, 649-654.

Boltnev, A. I., York, A. E. and Antonelis, G. A. (1998). Northern fur seal young: interrelationships among birth size, growth, and survival. Can. J. Zool. 76, 843-854.

Bonner, W. N. (1984). Seals of the Galapagos Islands. Biol. J. Linnean Soc. 21, 177- 184.

Boyd, I. L., Woakes, A. J., Butler, P. J., Davis, R. W. and Williams, T. M. (1995). Validation of heart rate and doubly labelled water as measures of metabolic rate during swimming in California sea lions. Funct. Ecol. 9, 151-160.

Bryden, M. M. and Molyneux, G. S. (1978). Arterio venous anastomoses in the skin of seals, Part 2: the California sea lion, Zalophus californianus, and the northern fur seal, Callorhinus ursinus (Pinnipedia: Otariidae). Anat. Rec. 191, 253-260.

203 Burns, J. M. (1999). The development of diving behavior in juvenile Weddell seals: pushing physiological limits in order to survive. Can. J. Zool. 77, 737-747.

Burns, J. M., Clark, C. A. and Richmond, J. P. (2004). The impact of lactation strategy on physiological development of juvenile marine mammals: implications for the transition to independent foraging. Int. Cong. Ser. 1275, 341-350.

Chappell, M. A., Bachman, G. C. and Hammond, K. A. (1997). The heat increment of feeding in house wren chicks: magnitude, duration, and substitution for thermostatic costs. J. Comp. Physiol. B 167, 313-318.

Costa, D. P. and Kooyman, G. L. (1982). Oxygen consumption, thermoregulation, and the effect of fur oiling and washing on the sea otter, Enhydra lutris. Can. J. Zool. 60, 2761-2767.

Costa, D. P. and Kooyman, G. L. (1984). Contribution of specific dynamic action to heat balance and thermoregulation in the sea otter Enhydra lutris. Physiol. Zool. 57, 199-203.

Davis, R. W., Williams, T. M. and Kooyman, G. L. (1985). Swimming metabolism of yearling and adult harbor seals Phoca vitulina. Physiol. Zool. 58, 590-596.

Dawson, T. J. and Fanning, F. D. (1981). Thermal and energetic problems of semiaquatic mammals: a study of the Australian water rat, including comparisons with the platypus. Physiol. Zool. 54, 285-296.

Dejours, P. (1987). Water and air. Physical characteristics and their physiological consequences. In Comparative Physiology: Life in Water and on Land (ed. P. Dejours, L. Bolis, C. R. Taylor, E. R. Weibel), pp. 3-11. New York: Springer- Verlag.

Donohue, M. J., Costa, D. P., Goebel, M. E. and Baker, J. D. (2000). The ontogeny of metabolic rate and thermoregulatory capabilities of northern fur seal, Callorhinus ursinus, pups in air and water. J. Exp. Biol. 203, 1003-1016.

Fedak, M. A., Rome, L. and Seeherman, H. J. (1981). One-step nitrogen dilution technique for calibrating open-circuit oxygen volume flow rate measuring systems. J. Appl. Physiol. 51, 772-776.

Feldkamp, S. D. (1987). Swimming in the California sea lion: morphometrics, drag and energetics. J. Exp. Biol. 131, 117-135.

204 Feldman, H. M. and Parrott, K. M. (1996). Grooming in a captive Guadalupe fur seal. Mar. Mamm. Sci. 12, 147-153.

Fish, F. E. (1979). Thermoregulation in the muskrat (Ondatra zibethicus): the use of regional heterothermia. Comp. Biochem. Physiol. A 64, 391-397.

Fleischer, L. A. (1987). Guadalupe fur seal, Arctocephalus townsendi. In Status, Biology, and Ecology of Fur Seals (ed. J.P. Croxall, R.L. Gentry), pp. 43-48. US Department of Commerce, NOAA Technical Report NMFS 51.

Fowler, S. L., Costa, D. P., Arnould, J. P. Y., Gales, N. J. and Kuhn, C. E. (2006) Ontogeny of diving behavior in the Australian sea lion: trials of adolescence in a late bloomer. J. Anim. Ecol. 75, 358-367.

Gallivan, G. J. and Ronald, K. (1979). Temperature regulation in freely diving harp seals (Phoca groenlandica). Can. J. Zool. 57, 2256-2263.

Gentry, R. L. (1972). Thermoregulatory behavior of eared seals. Behaviour 46, 73- 93.

Gentry, R. L. and Holt, J. L. (1986). Attendance behavior of northern fur seals. In Fur Seals: Maternal Strategies on Land and at Sea (ed. R. L. Gentry, G. L. Kooyman), pp. 41-60. Princeton, NJ: Princeton University Press.

Gilbert, C., Le Maho, Y., Perret, M. and Ancel, A. (2007). Body temperature changes induced by huddling in breeding male emperor penguins. Am. J. Physiol. Regul. Integr. Comp. Physiol. 292, R176-R185.

Hafez, E. S. E. (1964). Behavioral thermoregulation in mammals and birds. Internat. J. Biometeor. 7, 231-240.

Hampton, I. F. G. and Whittow, G. C. (1976). Body temperature and heat exchange in the Hawaiian spinner dolphin, Stenella longirostris. Comp. Biochem. Physiol. A 55, 195-197.

Hart, J. S. (1971). Rodents. In Comparative Physiology of Thermoregulation, Volume II: Mammals (ed. G. C. Whittow), pp. 1-149. New York: Academic Press.

Hart, J. S. and Irving, L. (1959). The energetics of harbor seals in air and in water with special consideration of seasonal changes. Can. J. Zool. 37, 447-457.

205 Hawkins, P. A. J., Butler, P. J., Woakes, A. J. and Gabrielsen, G. W. (1997). Heat increment of feeding in Brünnich’s guillemot Uria lomvia. J. Exp. Biol. 200, 1757-1763.

Heath, C. B. (1989). The behavioral ecology of the California sea lion, Zalophus californianus. PhD thesis, University of California, Santa Cruz.

Horning, M. and Trillmich, F. (1997). Development of hemoglobin, hematocrit, and erythrocyte values in Galapagos fur seals. Mar. Mamm. Sci. 13, 100-113.

Irving, L. and Hart, J. S. (1957). The metabolism and insulation of seals as bare- skinned mammals in cold water. Can. J. Zool. 35, 497-511.

Irving, L., Peyton, L. J., Bahn, C. H. and Peterson, R. S. (1962). Regulation of temperature in fur seals. Physiol. Zool. 35, 275-284.

Jensen, P. G., Pekins, P. J. and Holter, J. B. (1999). Compensatory effect of the heat increment of feeding on thermoregulation costs of white-tailed deer fawns in winter. Can. J. Zool. 77, 1474-1485.

Jørgensen, C., Lydersen, C., Brix, O. and Kovacs, K. M. (2001). Diving development in nursing harbor seal pups. J. Exp. Biol. 204, 3993-4004.

Kaseloo, P. A. and Lovvorn, J. R. (2003). Heat increment of feeding and thermal substitution in mallard ducks feeding voluntarily on grain. J. Comp. Physiol. B 173, 207-213.

Kleiber, M. (1975). The Fire of Life: An Introduction to Animal Energetics. Huntington, NY: R.E. Kreiger Publishing Co.

Kvadsheim, P. H. and Folkow, L. P. (1997). Blubber and flipper heat transfer in harp seals. Acta Physiol. Scand. 161, 385-395.

Le Boeuf, B. J., Morris, P. A., Blackwell, S. B., Crocker, D. E. and Costa, D. P. (1996) Diving behavior of juvenile northern elephant seals. Can. J. Zool. 74, 1632-1644.

Le Maho, Y., Delclitte, P. and Chatonnet, J. (1976). Thermoregulation in fasting emperor penguins under natural conditions. Am. J. Physiol. 231, 913-922.

Long, R. A., Martin, T. J. and Barnes, B. M. (2005). Body temperature and activity patterns in free-living Arctic ground squirrels. J. Mammal. 86, 314-322.

206 Lovvorn, J. R. (2007). Thermal substitution and aerobic efficiency: measuring and predicting effects of heat balance on endotherm diving energetics. Phil. Trans. R. Soc. B 362, 2079-2093.

Markussen, N. H., Ryg, M. and Ørtisland, N. A. (1994). The effect of feeding on the metabolic rate in harbour seals (Phoca vitulina). J. Comp. Physiol. B 164, 89-93.

McKechnie, A. E. and Lovegrove, B. G. (2001). Thermoregulation and the energetic significance of clustering behavior in the white-backed mousebird (Colius colius). Physiol. Biochem. Zool. 74, 238-249.

Meagher, E., Pabst, D. A., McLellan, W., Westgate, A. and Wells, R. (2000). Heart rate, respiration and heat flux across the dorsal fin in wild bottlenose dolphins, Tursiops truncatus. Am. Zool., 40, 1128.

Merrick, R. L. and Loughlin, T. R. (1997). Foraging behavior of adult female and young-of-the-year Steller sea lions in Alaskan waters. Can. J. Zool. 75, 776- 786.

Molyneux, G. S. and Bryden, M. M. (1978). Arterio venous anastomoses in the skin of seals, Part 1: the Weddell seal, Leptonychotes weddelli, and the elephant seal, Mirounga leonina (Pinnipedia: Phocidae). Anat. Rec. 191, 239-252.

Nespolo, R. F., Bacigalupe, L. D. and Bozinovic, F. (2003). The influence of heat increment of feeding on basal metabolic rate in Phyllotis darwini (Muridae). Comp. Biochem. Physiol. A, 134, 139-145.

Nickerson, D. M., Facey, D. E. and Grossman, G. D. (1989). Estimating physiological thresholds with continuous two-phase regression. Physiol. Zool. 62, 866-887.

Noren, S. R., Iverson, S. J. and Boness, D. J. (2005). Development of the blood and muscle oxygen stores in gray seals (Halichoerus grypus): Implications for juvenile diving capacity and the necessity of a terrestrial postweaning fast. Physiol. Biochem. Zool. 78, 482-490.

Noren, S. R., Williams, T. M., Pabst, D. A., McLellan, W. A. and Dearolf, J. L. (2001). The development of diving in marine endotherms: preparing the skeletal muscles of dolphins, penguins, and seals for activity during submergence. J. Comp. Physiol. B 171, 127-134.

207 Odell, D. L. (1974). Behavioral thermoregulation in the California sea lion. Behav. Biol. 10, 231-237.

Ohata, C. A. and Miller, L. K. (1977a). Northern fur seal thermoregulation: thermal responses to forced activity on land. J. Therm. Biol. 2, 135-140.

Ohata, C. A. and Miller, L. K. (1977b). Some temperature responses of northern fur seal (Callorhinus ursinus) pups. J. Mammal. 58, 438-440.

Perret, M. (1998). Energetic advantage of nest-sharing in a solitary primate, the lesser mouse lemur (Microcebus murinus). J. Mammal. 79, 1093-1102.

Peterson, R. S. and Bartholomew, G. A. (1967). The Natural History and Behavior of the California Sea Lion. Special Publication No.1, The American Society of Mammologists.

Peterson, R. S., Le Boeuf, B. J. and DeLong, R. L. (1968). Fur seals from the Bering Sea breeding in California. Nature 219, 899-901.

Pierson, M. O. (1987). Breeding behavior of the Guadalupe fur seal, Arctocephalus townsendi. Status, Biology, and Ecology of Fur Seals (ed. J. P. Croxall, R. L. Gentry), pp. 83-93. US Department of Commerce, NOAA Technical Report NMFS 51.

Pinshow, B., Fedak, M. A., Battles, D. R. and Schmidt-Nielsen, K. (1976). Energy expenditure for thermoregulation and locomotion in emperor penguins. Am. J. Physiol. 231, 903-912.

Reeves, R. R., Stewart, B. S., Clapham, P. J. and Powell, J. A. (2002). Guide to Marine Mammals of the World. New York: National Audobon Society, Alfred J Knopf.

Richmond, J. P., Burns, J. M. and Rea, L. D. (2006). Ontogeny of total body oxygen stores and aerobic dive potential in Steller sea lions (Eumetopias jubatus). J. Comp. Physiol. B 176, 535-545.

Rommel, S. A., Early, G. A., Matassa, K. A., Pabst, D. A. and McLellan, W. A. (1995). Venous structures associated with thermoregulation of phocid seal reproductive organs. Anat. Rec. 243, 390-402.

Rosen, D. A. S. and Trites, A. W. (1997). Heat increment of feeding in Steller sea lions, Eumetopias jubatus. Comp. Biochem. Physiol. A 118, 877-881.

208 Rutishauser, M. R., Costa, D. P., Goebel, M. E. and Williams, T. M. (2004). Ecological implications of body composition and thermal capabilities in young Antarctic fur seals (Arctocephalus gazella). Physiol. Biochem. Zool. 77, 669-681.

Schieltz, P. C. and Murphy, M. E. (1997). Heat increment of feeding in adult white- crowned sparrows. Comp. Biochem. Physiol. A 118, 737-743.

Schmidt-Nielsen, K. (1997). Animal Physiology: Adaptation and Environment. New York: Cambridge University Press.

Sealander, J. A., Jr. (1952). The relationship of nest protection and huddling to survival of Peromyscus at low temperature. Ecology 33, 63-71.

Sokal, R. R. and Rohlf, F. J. (1981). Biometry. San Francisco: Freeman.

Tarasoff, F. J. (1972). Comparative aspects of the hind limbs of the river otter, sea otter and seal. In Functional Anatomy of Marine Mammals, Vol. 1 (ed. R. J. Harrison), pp. 333-359. London: Academic Press.

Tarasoff, F. J. (1974). Anatomical adaptations in the river otter, sea otter and harp seal with reference to thermal regulation. Functional Anatomy of Marine Mammals, Vol. 2 (ed. R. J. Harrison), pp. 111-141. London: Academic Press.

Tarasoff, F. J. and Fisher, H. D. (1970). Anatomy of the hind flippers of two species of seals with reference to thermoregulation. Can. J. Zool. 48, 821-829.

Trillmich, F. (1984). Natural history of the Galapagos fur seal (Arctocephalus galapagoensis Heller). Key Environments: Galapagos (ed. R. Perry), pp. 215- 223. Oxford: Pergamon Press.

Trites, A. W. (1990). Thermal budgets and climate spaces: the impact of weather on the survival of Galapagos (Arctocephalus galapagoensis Heller) and northern fur seal pups (Callorhinus ursinus L.). Funct. Ecol. 4, 753-768.

Webster, A. J. F. (1983). Energetics of maintenance and growth. Mammalian thermogenesis (ed. L.Girardier, M. J. Stock), pp. 178-207. New York: Chapman and Hall.

Whittow, G. C. (1971). Ungulates. Comparative Physiology of Thermoregulation, Volume II: Mammals (ed. G. C. Whittow), pp. 191-281. New York: Academic Press.

209 Whittow, G. C., Ohata, F. A. and Matsuura, D. T. (1971). Behavioral control of body temperature in the unrestrained California sea lion. Commun. Behav. Biol. 6, 87-91.

Williams, T. D., Allen, D. D., Groff, J. M. and Glass, R. L. (1992). An analysis of California sea otter (Enhydra lutris) pelage and integument. Mar. Mamm. Sci. 8, 1-18.

Williams, T. M. (1983). Locomotion in the North American mink, a semi-aquatic mammal. I. Swimming energetics and body drag. J. Exp. Biol. 103, 155-168.

Williams, T. M. (1986). Thermoregulation in the North American mink during rest and activity in the aquatic environment. Physiol. Zool. 59, 293-305.

Williams, T. M. (1999). The evolution of cost-efficient swimming in marine mammals: limits to energetic optimization. Phil. Trans. R. Soc. London B 354, 193-201.

Williams, T. M., Ben-David, M., Noren, S., Rutishauser, M., McDonald, K. and Heyward, W. (2002). Running energetics of the North American river otter: do short legs necessarily reduce efficiency on land? Comp. Biochem. Physiol. A 133, 203-212.

Williams, T. M., Haun, J., Davis, R. W., Fuiman, L. A. and Kohin, S. (2001). A killer appetite: metabolic consequences of carnivory in marine mammals. Comp. Biochem. Physiol. A 129, 785-796.

Withers, P. C. (1977). Measurement of oxygen consumption, carbon dioxide production, and evaporative water loss with a flow-through mask. J. Appl. Physiol. 42, 120-123.

Yeager, D. P. and Ultsch, G. R. (1989). Physiological regulation and conformation: A BASIC program for determination of critical points. Physiol. Zool. 62, 888- 907.

Yeates, L. C. (2006). Physiological capabilities and behavioral strategies for marine living by the smallest marine mammal, the sea otter (Enhydra lutris). PhD dissertation, University of California Santa Cruz, USA.

Yeates, L. C., Williams, T. M. and Fink, T. L. (2007). Diving and foraging energetics of the smallest marine mammal, the sea otter (Enhydra lutris). J. Exp. Biol. 210, 1960-1970.

210 Zar J. H. (1974). Biostatistical Analysis. Englewood Cliffs, NJ: Prentice-Hall.

211