Physiology of Diving of Birds and Mammals

PHYSIOLOOICAL REI1E'US Vol 77, No. 3. July I887 mnierl in u.Sn

PATRICK J. BUTLER AND DAVID R. JONES

School of Biological Scimes, The Unidtyof Birmingham, Edgbmlon, Bidnghum, United Kingdom- and Department of Zoology, Univmity of British Columbia, Vancouver, Bitish Columbia, Camui,a

I. Introduction TI. Diving Rehavior A. Birds B. Mammals I11 Metatrolic Rate and Metabolism A. Aerob~cdive limit B. Methods for determining metabolic rate of diving animals C. Birds D. Mammals I!'. I!'. Controlline Metabolism: Cardiores~iratorv" Relationshi~s Durine - Divine- A. ~ircula&yadjustments to di&g B Effiracy of cardiores1,irnror). responses lo d~ung V. Hecoven IFrom D~\ulg.t'ardiontsp~rato~~ Rt!sln)~~srs 113 Sllrf;i(.~llg VI. Control of ~ardiores~iratolyResponses VII. Concluding Comments

Butler, Patrick J., and David R. Jones. Physiology of Diving of Birds and Mammals. Physiol. Rev. 77: 837-899, 1997.-This review concentrates on the physiological responses, and their control, in freely diving birds and mammals that enable them to remain submerged and sometimes quite active for extended periods of time. Recent developments in technology have provided much detailed information on the behavior of these fascinating animals. Unfortunately, the advances in technology have been insufficient to enable physiologists to obtain anything like the same level of detail on the metabolic rate and physiological adjustments that occur during natural diving. This has led to much speculation and calculations based on many assumptions concerning usable oxygen stores and metabolic rate during diving, in an attempt to explain the observed behavior. Despite their shortcomings, these calculations have provided useful insights into the degree of adaptations of various species of aquatic birds and mammals. Many of them, e.g., ducks, smaller penguins, fur seals, and Weddell seals, seem able to metabolize aerobically, when diving, at approximately the same (if not greater) rate as they do at the surface. Their enhanced oxygen stores are able to support aerobic metabolism, at what would not be considered unusually low levels, for the duration of the dives, although there are probably circulatory readjustments to ensure that the oxygen stores are managed judiciously. For other species, such as the larger pengulns, South Georgian shag, and female elephant seals, there is a general- consensus that thev must either be reducing their aerobic metabolic rate when diving. possibly by way of regional L~wotl~emlia,andlor producing ATP, at least pmtly, by anaerobiosis and metabolizing the lactic acid when at the surface (although this is hardly likely in the case of the female elephant seals). Circulation is the proximate regulator of metabolism during aerobic diving, and heart rate is the best single indicator of circulatory adjustment. During voluntary dives, heart rates range from extreme bradycardia to well above resting, rcflccting mctaholic pcrformancc. Effercnt cardiac control is largely parasympathetic. Rcflcx cardiorespiratory responses are modulated by conditioning and habituation, but reflexes predominate during extended dives and during recovery, when gas exchange is maximized.

I. INTRODUCTION hundreds of meten. There are, therefore, two major problems confronting many species of aquatic birds and mm- Diving bids and mammals remain submerged under- mals: relating to their limited oxygen stores and to the water for differing durations, from several seconds to large hydrostatic pressures to which they are exposed. many minutes, and, in general terms, these differing dura- With an increase of 1 atmosphere for approximately every tions relate to the depth to which the animals routinely 10-m descent into the water column, an animal at a depth dive in the water column, from a meter or so to many of 200 m will experience a pressure of 21 atmospheres.

0031-9333197 15.00 Copyright 0 1997 the Amencan Physiological Society m.7 PATRICK J. BUTLER iLND DAVID R. JONES Volume 77

The problems facing these animals are all related to the TABLE 1. Sites of oxygen storage in diving and fact that they have air-filled cavities in their bodies and nondiving l~orneothmr~s that air is compressible. There are a number of anatomic features in marine mammals that enable them to over- Myoglabin, dl00 g hissue come many of these problems, and these were all dis- Muscles cussed by Butler and Jones (53) and Kooyman (218). Be- Hummingbird (pectoralis) cause there has been little research on th's topic during Tufted duck (gastrocnemius) the past 10 years, we do not intend to discuss this aspect Tufted duck jpectoralis) Gentoo penguin (pectoralis) of diving to any great extent in this review. Thoroughbred horse (psoas) Regarding their linriled oxygen stores, the central Beaver question is: What physiological and metabolic processes Ribbon seal Weddell seal enable these air-breathing, homeothcrmic endothenns to Elephant seal remain submerged and, maybe, active for extended peri- Bottlenose dolphin ods? Two vely important aspects of the answer to this Porpoise (psoas) Sperm whale question relate to the amount of oxygen that can be stored Bottlenose whale in the body for use underwater (and, conversely, the amount of carbon dioxide that can be stored during sub- Volume, ml BTPSkg mersion for removal at the surface) and the rate at which Respiratorz~S?istm that oxygen is used during the period of submersion. Re- lated to the latter is how the stored oxygen is distributed Mallard duck 112 Tufted duck (lesser sraup) 180 (355) to the various organs and tissues to meet their different Adhlie penguin 165 requirements and to what extent living andlor diving in Dog 61 cold (maybe around 0°C) water and eating cold food af- Human 74 Beaver 60 fects the metabolic rate of these endothermic homeo- Weddell seal 48 them. Although temperature regulation and adaptations Harbor porpoise 59 Bottlenose whale 26 to low temperature were also discussed in some detail in -- Reference 53 and are not covered to any great extent in Volume, mV Oxygen Capacity, the current review, we intend to highlight some exciting 100 g body Hb, g1100 rnl ml OdlOO rnl recent observations that indicate that regional hypother- mass blood blood mia may be an important factor in reducing overall oxygen Circulatory requirements during diving, in at least some species of Pigeon 9.2 marine birds and mammals. Another important aspect, Mallard 9. I particularly as far as foraging is concerned, is the rapidity Tufted duck 11.4 Chinstrap penguin with which the oxygen stores can be replaced (and the E~rlperorpenguin carbon dioxide removed) when the animals are at the Human 't.7 surface. Beaver 6.6 Weddell seal 21.0 It is possible to identify three distinct phases in the Bladdernose seal evolution of the answer to the above question. Initially, Elephant seal 21.7 data were obtained from restrained animals that were Bottlenose porpoise 7.1 Dall porpoise 14.3 forcihly submerged. These studies go hack well over 100 Sperm whale years, but perhaps the most influential publications were those of Irving (195, 196) and Scholander (346). These Hb, hemoglobin; BTPS, body temperature and pressure, saturated. and subscqucnt publications by thc samc authors and 0th Data are from the following sources: Ref. 49; lesser scaup, Ref. 364; elephant seal, Ref. 218; Weddell seal, Ref. 324. crs using rcstraincd animals have been rcviewed many times (3,28,45,53,69, 135,218), so it is not the intention here to produce another review of this material. vasoconstriction causing a reduction in the perfusion of all The model that emerged from these studies is that, parts of the body except the oxygensensitive tissues, the although the oxygen stored in the body is greater in diving central nervous system, and heart [note, there is no reduc- than in nondiving species (Table I), it is insufficient to tion in coronary flow in forcibly submerged ducks (204), enable aquatic bids and mammals to maintain aerobic me- whereas in seals it does decrease in proportion to the re- tabolism and remain submerged for the durations obtained duction in cardim ontput (27, 30, 424)l. during the experiments. Consequently, there is an overall Thus, in the underperfused tissues, which includes reduction in the level of aerobic metabolism so that the the skeletal muscles in these restrained animals, there is oxygen stores are conserved for those tissues that cannot a net accumulation of lactic acid that is flushed into the survive hypoxia This is achieved by selective peripheral blood soon after the animal is surfaced. As well as the Iz~ly1997 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 839

A shown in Figure 1. Of particular importance is the fact 100 that it takes approximately six times the duration of the forced submersion for the concentration of blood lactate i 60 o to return to the predive level (see also Ref. 53). 4 20 This hypothesis of large-scale peripheral vasocon- .-C striction, reduced aerobic metabolism, and a substantial g -20 c increase in anaerobic metabolism during voluntruy sub- -60 mersion was challenged during the late 1970s and early 8 -100 1980s, initially as a result of studies on two species, the tufted duck, Aythya fuligula, and the Weddell seal, Lep- 8 &'..f&*.$ tonychotes weddeUii (59, 235, 414), although it must be $A6@ said that Scholander, himself, was not too certain of the ,j V. universality of the IrvingScholander hypothesis. In his 1940 paper (346), he says that it would "be of great interest Submew to compare the circulation in a relaxed muscle during a dive with that of a working muscle during a dive and gy -iz mightcorrespondingHe goes be thoughton to conditions write possible that when duringeven theif a "reducedquietanimal dive is notmetabolismthis diving." cannot be the case in an ordinary dive. It is during submersion - :h2 3 that a seal does most of its exercise in hunting fish and s I I k by distance swimming." 30 I b 2'0 40 60 80140 Two years later, Scholander and colleagues (348) noted tlraL harbor (common) seals, Plwcu ,uitutir~u,rarely -- dive for longer than 4-5 min "and thus usually dive under ;.-E ator comfortable acid formation." aerobic Similarly, muscular for conditionthe two species without of needpen- E !ILy @ins that he stndied in 1940, macaroni, Eudgptes 0 chrysolopus, and gentoo, Pygoscelis papua, Scholander 240 noted that most natural dives were of < 1 min in duration. e c He concluded that the feature of these birds is not so 8 20 much prolonged diving as their ability to dive repeatedly I with only afew or single breaths between each dive. These 0 -300 10 20 30 50 70 90 comments and observations on some naturally diving Time (min) birds and mammals are not at all consistent with the idea of the accumulation of lactate and long recovery periods ETG. 1. Cardiov-ular and metabolic changes that occur during and a1 the surface lo remove this lactate. after a period of involuntary submersion of Weddeil seals. A: percentage It was the initial observations and studies on freely change in blood flow to various organs after 8-12 min of involuntary submersion of up to 6 seas, i~odfiedfro,,, zapol et d. (424),], B: diving birds artd nrarrorrals that led Lo the second slage in concentration of lactate in whole arterial blod during involuntary sub- the evolution of the answer to the question, i.e., that most, mersion and subsequent recovery of 1 seal. [Modifled from Murphy et if liOt dl, dives perfOnned nalurally aerobic iIr nature, al. (300).] C. mean ? SD of heart rate of 4 maternal seals during and after involuntary submersion, [Modified from Liggins et al. (257).] Attempting to study the physiological processes, and their control, that are associated with a particular type of behavior of an animal is fraught with diiculties. Scholander selective peripheral vasoconstriction, there is also a re- clearly yearned, on occasions, to be able to make physio- duction in heart rate during submersion. Because cardiac logical measurements on freely diving birds and mam- stroke volume does not change in most birds and mammals, but this was not to be. Nonetheless, a compatriot mals (see Ref. 53 for details), this diving bradycardia of his, Eliassen, published in 1960 what turned out to be causes a reduction in cardiac output of similar magnitude. a very perceptive paper (129). The increase in peripheral resistance and decrease in car- Using largely his own observations on the durations diac output tend to offset one another, so there is no of natural dives of a number of marine birds, calculations change in central blood pressure. The bradycardia is often of oxygen stores in these birds, and data on oxygen usage taken as being indicative of the other physiological and during such dives of normal duration (which incorporated metabolic events taking place during submersion. Some his own laboratory experiments on the drag of the birds of the components of this "classic" response to diving are at a range of underwater speeds), he refuted the idea that 840 PATRICK J. BUTLER AND DAVID R. JONES Volume 77 vasoconstriction and anaerobiosis occur in the locomotor rate of aerobic metabolism during diving andlor metabo- muscles of birds during natural diving. In fact, he com- lize anaerobically and accumulate lactate dnring some pared normally exercising and diving animals, suggesting dives. that the circulatory events are the same in both cases, This review concentrates on the physiological re- with blood being preferentially distributed "in favour of sponses, and their control, that occur in freely diving birds the working organs." He went on to say, "Considering the and mammals. Although some of the data presented are diving animals, the peripheral constriction during diving behavioral, the emphasis is on their physiological implica- should, therefore, not take place in the muscles, but in tions. Similarly, some of the data discussed here have the viscera alone." He did not support the idea of a reduc- been obtained from experiments on restrained animals, tion in (aerobic) metabolism during diving. At the time, but they are discussed in the context of their i~rrplications hi methodology was criticized (197) and his proposals for freely diving animals. were rejected (2). Before one embarks on this review, it is important The first real hint that data obtained from birds un- to understand the terminology used to describe the natu- dergoing involuntary submersion might he different from ral, and unnatural, diving of birds and mammals. Three those obtained during voluntary dives was provided by broad categories, natural, forced, and trained, encompass Millard et al. (285). They used radioLrar~smiLLrrsto record virtually all types of dives, but these categories are some- heart rate from freely exercising and diving Adelie and what arbitraty and, consequently, not exclusive. Natural gentoo penguins (P. adeliae and P. papua, rcspcctively). or voluntary dives are performed by unrestrained animals They noted that "The cardiovascular adjustments re- and are most frequently (>YO%) within the animal's capa- corded during voluntary diving in penguins are. . .not bility to provide energy requirements aerobically. Natural readily compared with earlier reported results in diving dives can be performed in the wild or simulated habitats birds. Such comparison is difficult because previous stud- to which animals are acclimated. Extended dives are a ies have been performed on restrained birds forcefully subset of natural diving and are dives that are temporally dived. . . . Our data have been obtained on penguins per- extended by the animal's own volition and may involve a forming normal swimming and diving, which will give re- switch to anaerobic energy production. sults as a composite of (involuntary) diving and exercise Dives in which the animal is restrained and sub responses." It was the further use of radiotransmitters, merged, confined and submerged, is encouraged to dive withpochards (Aythya ferina) and tufted ducks (59,4131, by presentation of a threatening stimulus (escape dives), arid Llie introduction arid use of tirne-depth recorders or is prevented from surfacing ("trapped dives") all fall VDRs) with freely diving Weddell seals (214, 215), both into the category of forced or involuntruy dives. An nn- in conjunction with more conventional physiological tech- usual circumstance is when an animal is trained to dive, niques, that gave rise to the idea that most, if not all, submerging only the head or the whole body, or trained natural dives are aerobic in nature. to change natural diving behavior. Training implies some As is often the case in biology, it can be folly to psychological restraint, so in one sense these are forced make generalizations on the basis of studies on one or dives, but for reasons that will become clear, these dives two species. Tufted ducks and Weddell seals were chosen will be termed trained dives, with the added qualifier that for particular reasons. The former can be used relatively the animal is either active or passive. easily in laboratory and semi-laboratory situations, while the natural habitat of the latter enables a laboratory to be created in the field. Thus it has been possible to obtain 11. DIVING BEHAVIOR more conventional physiological data from these animals, together with the information obtained from the transmit- The diving behavior of an animal will inevitably be ten or TDRs. However, more recent field studies on other related to the extent to which the animal is adapted to an species of diving birds and mammals, which have relied aquatic existence. The methods of locomotion on land, in entirely on transmitters, data storage devices, andlor the air, and in water have different mechanical require- TDRs, have taken 11s to the third stage in the evolution of ments. Most aqnatir hirds are still able to fly, so even if the answer to the question, and about full circle. Evidence the wings are used for underwater propulsion, they and from species such as thick-billed mwes (Briinnich's guil- the density of the body must be consistent with flight in lemots, Uria lomvia) (96), king and emperor penguins both media. Only the penguins have abandoned aerial (Aptenodytes patagonicus and Aptenodytes forsteri, re- flight altogether, but they still retain their legs for terres- spectively) (221, 232), northern and southern elephant trial locomotion. Similarly, among the aquatic mammals, seals (Mirounga a,ngustirostris and Miroz~ngaleoninn, the four limbs of mink, muskrats, and otters have become respectively) (179, 248), and gray seals (Halichoems gv- only marginally adapted for an aquatic existence, whereas pus) (378) indicates that some species, maybe only at in otariids (fur seals and seal lions) and even more so in some time during the annual cycle, either have avely low phocid (true) seals, the limbs have become more adapted Jz~ly1997 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS for aquatic propulsion and the animals are, thus, more cumbersome on land. In terms of locomotion, the cetaceans are completely adapted to an aquatic life-style. The hindlimbs have disappeared altogether, and the animals never come ashore. Thus, for these animals, all of their activity occurs in water. All of these factors, together with the buoyancy of the animals, will contribute to the rate at which oxygen will need to be consumed to swim a given distance underwater, i.e., to the efficiency of underwater swimming. At one extreme, diving ducks and mink are relatively buoyant, use webbed feet for generating thrust, and have relatively high drag coefficients (363, 364, 405). At the other B extreme, penguins and dolphins are close t,o neutral buoy- Foraging lime O*ygen ancy, use flippers or tail flukes as hydrofoils, and have looding relatively low drag coefficienls (11, 149). WreLhrr or not curve* dolphins possess special mechanisms for reducing drag, or whether body shape is the major factor, is still a conten- tious issue (149, 3421, although Bannasch (11, 12) has proposed such mechanisms for penguins. All of these Surfocc lime I.) facts should be borne in mind when studying the behavior 'I r2-... '.. 7 5y.. and metabolism associated with diving in aquatic birds -. and mammals. Although there may be similarities hetween different species, we should not be surprised to tind differences. A relatively simple way of describing diving behavior in a number of different species of birds and mammals is nc. 2. A: graphical representation illustrating amount of time that a diving bird or mammal should spend at surface (s*) loading its oxygen to present a table of mean and extreme values of dive stores to maximize ratio of foraging time (0 to travel time (7)plus time duration and depth. It is clear, from recent studies, how- at surface (s) [fl(i+ s)]. B: graph illuslratiog opli~ctalsululiurt Iur 3 ever, that diving behavior is influenced by a number of values of travel time, T,, T,, and T,, indicating that corresponding optimal times at surface, sf, st, and SF should increase, whereas resulting factors, many of which, such as light levels and food avail- foraging durations, f,, fi, and fa should first increase and then decrease. ability, are variable. Also, the frequency of particular dive [Modiied fmm Houston and Carhone (192).l depths or durations may not be normally distributed, so to give mean values without an indication of the factors that may influence those means, or whether "mean" is a critically greater proportion of Lhe oxygen is used for appropriate, can be misleading. In fact, some authors (35) traveling to and from the food, the time at the food patch give median values rather than means. begins to decrease (Fig. 2). This prediction is corrobo- There have been recent attempts to model diving be- rated by some independent studies (96, 411) but not by havior in terms of optimal foraging because a number of others (81). dives, if not most for some species, are feeding dives. An Recent evidence from gray seals (334) indicates that interesting feature, from a physiological point of view, is the oxygen-loading curve used by Houston and Carbone that the rate at which oxygen can be loaded when an (192) may not be strictly correct for all aquatic birds and animal is at the surface may strongly influence the time mammals. Certainly, the introduction of a 10- to 20-s it remains feeding at a given depth. Houston and Carbone delay into their model representing the time before oxy- (192) predict that, for animals that remain aerobic during gen is extracted from the inhaled air after surfacing more diving, time in the foraging area at, first increaqes and then accurately predicts the diving behavior of Antarctic fur decreases as travel time from the surface to the foraging seals (37). area (i.e., depth of dive) increases. This, they argue, is Thompson et al. (380) have discussed, among other because at relatively shallow depths the animals should things, the speed at which a diving animal should swim to operate over the steep part of the oxygen-loading curve its food supply from consideration of the cost of transport and thus neither unload nor load the oxygen stores toward (amonnt of energy rwnired to transport a given mass over their extreme levels. At relatively greater depths, however, a given distance) as well as the oxygm-loading curve. It it is more efficient for the animal to spend longer at the appears LliaL, il rnl arlllual wishes Lo ~uaxiini~eits gross food source and to operate over a greater range of the energy intake, it should swim slower to rraclr greater depths oxygen-loading curve. Beyond a certain depth, and when (Fig. 3). When the animal has reached the food, its swim 842 PATRICK J. BUTLER AND DAVID R. JONES Volume 77

A E or 02 tieunderwater. For example, the tufted duck may spend Cast of travelling lo as much as 25% of a day underwater, and most of this deplh D foraging behavior occurs during a 14-h period that spans nighttime (321). Foraging behavior of birds can be divided into bouts during which a number of dives follow one another in relatively quick succession; the pause between Oxygen loading dives is relatively short (-30 s), and the animals spend curve more time underwater than at the swface. Most diving ducks spend -63% of a diving bout underwater, i.e., they have a dive-to-pause ratio of 1.7. Such belravior (it., relatively short periods at the surface between dives) is taken Travel time Surface time IT) IS1 to indicate that the dives are aerobic in naturc with little or no accumulation of anaerobic by-products, although B termination of a bout could be the result of a progressive Costs of travelling to accumulation of such by-products. depths D, and D2 The duration of dives performed by most birds is relatively short. For many birds that feed on benthic or- ganisms, dive duration is related to the depth of the water (64, 117, 124, 421). At a depth of 2 m, tufted ducks dive, on average, for 20-25 s, with a maximum duration of 46 s (124, 3ti7). They swim at a speed of 0.6 m/s during descent (60). It appears from a detailed study on eider ducks, Somateria mollissima, that, for benthic feeding ducks, the surface period increases with dive duration (i.e., the dive-topause ratio is constant). Ydenberg and Travel time Surface time Guillemette (421) interpret this to indicate that these birds 1-1 151 complete their recovery between each dive. A number of birds, such as grebes, divers, cormo- nc. 3. A: graphical representation of relatlonshlp, for a dlvlng bird or mammal, between time traveling to its foraging area (71, energy (or rants, and auks are, to a greater or lesser extent, active oxygen) used during foragng (Ed,and time at surface loading its oxygen predators of fish. These species, therefore, may have to store (s). Vertical distance between intersections of common tangent employ different behavior patterns to exploit their mobile with curve reoresentine enerm, cost of traveiine to a oarticular deoth. prey (420). From the latter study, the divet@pause ratio for western grebes, Aechmophoms occidffntalis, in- tion of common tangent with cost of traveling curve represents 7 and creases from -2 for dives of 20-s duration to >3 for dives hence swimming speed (u), which maximizes rate of delivery of oxygen to foraging area at depth D. In B, this optimal condition is illustrated of 60-s duration. This is taken to indicate that there may bv shallower of 2 deoths. D,. If animal dives to a meater deoth. D,... not be full recovery between each dive. Observations of common rangmt with oxvgcn h>acblierune t~11a.11,st.~,it uf trdwhttg a number of species of cormorant (Phalacrocorax sp.) <.un.cI.) left of itlr,,mecuun wtrh aorted ihne. ~ndraunerhdt I( wlllch ~naxlnwccrate of OY~~CIIdeII\cp to foraging area, dccrensw a5 depth indicate mean dive durations from 7 to 70 s and dive- increases. [Modified from Thompson et al. (380).] to-nause ratios from 1 to >4 187.. . 181. 381. 411). Mean swimming speed during diving varies between 0.8 and 1.2 m/s (411). Dive duration is clearly related to the depth of speed will depend on the density of the prey and the rate the dives (381, 411), and larger animals tend to remain at which they are moving. For low-density, rapidly moving submerged longer than smaller ones, even at the same prey, it may be energetically advantageous for the diving depth (87, 381). The latter authors also noted that the animal to remain stationary. When searching for prey, it frequency distribution of dive durations is positively should swim at the minimum cost of transport speed. skewed. However, studies on the Antarctic blue-eyed shag It is proposed to summarize, where possible, the div- (Imperial cormorant, Phalacrocorax atriceps), using ing behavior of those species that are mentioned in subse back-mounted transmitting or recording devices (100, quent sections, in tabular form (Tables 2 and 3) and briefly 386), indicate that the diving behavior of these birds is to describe the behavior of these and other species where strongly bimodal (Fig. 4, Table 2), with long dives being the influence(s) of variable factors is known. twice as common a5 short dives, hut not ronfinnd tn any A. Birds particular time of day (100). It is assumed that shallow dives do not involve ben- As Jones and Furilla (206) point out, diving birds thic foraging, and it has been suggested that they may be foraging in nature may spend a large proportion of their a more economic way of moving foraging location than July 1997 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 843

TABLE 2. &ing behavior of a number of species of birds

~ean(and anm mum) ~ean(and ~axi~"mj ~ean(and ~aumaummj Reference Species Dive Dwation, E Dive Depth, m Surface Time,P Conditions No.

Tufted duck 22.4 (46.3) 2 (3) Experimental pond 367 (AuWaWigula) Blue-eyed shag Male, freely diving at sea with time- (Phalacrncm atriceps) depth recorder Curo~rto~cguiUeruoL 44 Freely diving at sea will, radio (common mum) transmitter (Uria a@e) Brunnich's guillemot F~eelydiving at sea with dive (thick-billed mme) recorders (Uvia iwnvia) Little penguin 22 2.1 March (20) Freely diving at sea with dive and (Eudllptule minor) 21 1.3 December (50) speed recorders Gentoo penguin 46 6.9 (<21 m) 128 Freely dimg at sea uith time-depth (Pugoscelis papua) 166 (>300) 81 (230 m) (156) 100 recordem Adklie penguin 102 (240)* 9.4 (26.8)* Free diving at sea with time-depth (Pyqoscelis adeliae) 73 (160)t 26 (98)t recorders (>I50 <180)1: King penguin (7.5 min) (304) Freely diving at sea with time-depth (Aptenodytes patagonicus) recorders Emperor penguin 4-5 (15.8 min) 20-40 (534) Freely diving at sea with time-depth 228 (Aptenodytesforsteri) (mode for foragmg (mode for foraging recorders dives) dives) flying or swimming, although it is also accepted that they longer than would be expected in terms of calculated may be used to target prey visible from the surface (100). oxygen stores and rate of oxygen consumption, based on For both long and short dives, the dive-to-pause ratio is data from other species. Thus the blue-eyed shag could, 0.3-0.4, i.e., overall surface times are 2-3 times greater routinely, be engaging in anaerobic metabolism while for- than the preceding dive duration, although this does not aging. Mean swim speed during the descent phase of the seem to be the case for some shorter dives (see Fig. 16). long dives was between 1.7 and 2.2 ink. The maximum Both groups of authorr (100, 386) conclude that most of descent rate quoted was a staggering 5 rnls! the long, deep dives performed by blue-eyed shags are A study on the shag (Phalacrocorm aristolelis) has

TABLE 3. Diving behavior of a number of species of mammals

Mc*, (ad Maxllawrl) Meal (*,d Mau.,,",t,j Meart (and Maur,,ul,,j Species nive mio on, heDepth. m Surface Time, mi" Conditions Reference No.

Northem fur seal Breeding female at sea (CaUorhinus ursinw) with time-depth recorder Antarctic fur seal 0.95 (10) 12.6 (181) 0.57 (0.75) Breeding female at sea (Arctocephalw gazella) Median Medim Mcdim with time-depth recorder Weddell seal Diving from an artificial (Lzplonychotes hole in shelf ice with weddellit) time~depchrecorders Carmnon (harbor) seal Adult males at sea with (Phoca vitdih) radiotransmitters Gray seal Adults al sea isah (Haliclwerus gypus) ~~tlll,l~dlrllllll?~, NoWlern elephant seal .%dulr female a1 sea between lactation and molt with time- depth recorder Southern elephant seal Adulb at sea with time- 178 (Mzrounga leonina) depth recordee 354 and M. A. E'edak, personal couun~uficati.liua Sperm whale Diving in open seas 390 IPhusew eatadon) with acoustic @AS 844 PATRICK J. BUTLER AND DAVID R. JONES

Local time (h) Local time (h) - 30 30 nc. 4. Frequency histograms indicating time of & day, depth, and duration of natural dives performed by E 2 male, Antarctic blue-eyed shags. Far one, of mass = 20 20 % 2.55 kg (A-C), 564 dives were recorded during 11 days; m for the other, of nrass 2.80 kg (D-F), 110 dives were P 10 10 U. recorded durlng 3 days. [ModlAed from Croxall et al. (loo).] O O 5 15 25 35 45 55 65 75 85 85 105 5 15 25 35 45 55 65 75 85 95 105 Depth (m) Depth (m)

Duration (min) Duration (min)

demonstrated how characteristics of the prey may influ- distinctly bimodal (403, 404). The birds spend -2046 of ence diving behavior (385). The study was performed over their total diving time engaged in shallow (<21 m) dives three seasons (1987,1988, and 1989), and during 1989, the and -75% of their diving time is spent on deeper (>30 size of the prey (sand eels) was significantly smaller than m) dives. These deeper dives are thought to be feeding during 1987 and 1988. 'l'he authors estimated that both dives and are to mean depths of 30-40 m at dawn and prey capture rates and the time spent submerged at any dusk and to mean depths 80-90 m at midday (Fig. 5), depth were greater in 1989. whidi is consislenl wiU~Lhe behavior of their main prey, Thick-billed murres (B~nnich'sguiuemots) illustrate krill. Availability of prey also iduences the diving behav- another influence of prey on diving behavior (96). Most (67%) of the dives occur between 20.00 and 04.00 h and are generally <20 m deep. As the sun kses, there are fewer but deeper (>40 m) dives, which suggests that the birds are following the diurnal migration of their prey (the amphipod, Parathemisto). The frequency distributions of hoth dive depth and dive duration are positively skewed. Mean descent speed is 0.9 ds. Little penguins (Eudyptula minor) are, as their name suggests, the smallest of the penguins, and their diving behavior is more like that of ducks than of auks and other penguins (Table 2), except that their mean swimming Time (h) speed is impressively high at 2.4 m/s (168). Foraging activ- nc. 5. Histograms relating depths of dives (>30 m) performed by ity of nonbreeding gentoo penguins during winter is simi- 7 freeranging gentoo penguins with time of day. [Modified from Willims lar to that of chick-rearing birds, when the behavior is et a1 (403).] Jzily 1997 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS ior of penguins. A study on the Adelie penguin during the Antarctic summers of 1990 and 1991 (389) indicates that, in 1991 when food availability was comparatively poor, the mean dive depths (12.3 m) and dive durations (1.9 min) were greater than those in 1990 (7.1 m and 1.5 min, respectively). It is also char from the stndies of Naito et al. (307) and Chappell et al. (81) that diving behavior of a species may vary between locations andlor with the type of TDR used (Table 2). Average underwater swim speed of chinstrap, Adelie, and gentoo penguins is -2.2 mls (410), although it has to be noted that externally attached devices may affect the 0 behavior of these animals (10% see also Ref. 403). The 0 4 8 I2 16 20 26 former authors found that instrumented Adelie penguins B LOC~Itime ~h) in a 21-m-long water channel have a bimodal speed distribution of 1.8 and 2.4 4s.The preferred speed of uninstru- mented birds is 2.0 ds. It is the larger penguins, the kings and emperors, that are the most impressive avian divers. A detailed study on the behavior and energetics of diving in king penguins indicate that they, like the gentoo penguins, dive in response to the vertical movements of their prey (221). It has even been postulated that king penguins may be con- strained at night, because of low light levels, to feed at depths where prey densities are lower (330), At night, dives are never deeper than 30 m, whereas during the da;y they are to 3100 m (Fig. G). There may be a number of Local time (hi shallow dives between some of the longer deeper dives nu. 6. Scatter plots of dive depths with time of day for 2 free- (Fig. 7). Preferred swim speed is 2.1-2.2 ds. This com- ranging king penguins. [Modified from Kooyman et al. (221).] pares with a mean descent rate during dives of 0.6 ds. In emperor penguins, the majority of dives are to <200 m in depth and these occur at any time during the 24-h dive in relatively shallow (13 m) water (309). Details of cycle, whereas those to >300 m tend to occur between the dive behavior of some other species of aquatic man- 04.00 and 20.00 h (228). Rates of descent and ascent were mals are given in Table 3. bctwccn 1.0 and 2.0 m/s for divcs to >-I00 m and ncvcr Of thc pinnipcds, it is thc fcmalc otariids that havc exceeded 2.5 mls. In anothcr study, Kooyrnan ct al. (232) bccn most cxtcnsivcly studicd. Brccding fcmalc otariids, recorded a maximum (burst) swim speed of 7.1 dsfor unlike the phocids, periodically go to sea to feed during this species. It is interesting to note that surface time the lactating period, returning to suckle their pups. Thus it increases progressively following dives in excess of 7 to is relatively easy to deploy and retrieve archival recorders 8min duration (228). with these animals, in much the same way as it is with breeding aquatic birds. In both northern and Antarctic fur seals (Callorhinus B. Mammals z~rsinus,Arctocephnlus gazeUa, respectively), most dives occur at night, and these are shallower than those that A similar progression, from smaller species, which occur during the day. This pattern of behavior suggests tend to perform shorter and shallower dives, to larger that the seals follow the vertical movements of their major species, which Lend to perforn~longer and deeper dives, food items, krill for the Antarctic fur seal, fi~hand cepha- as seen in birds, is also apparent among the aquatic mam- lopods for the northern fur seal (99,163). When prey abun- mals, although cetaceans may be an exception to this dance is low, Antarctic fur seals spend proportionately generalization. For otters, Lutra lutra, feeding at sea, longer time diving and feeding at night (34). A closer anal- mean dive duration is related to whether or not prey are ysis of deep aud shallow dives in the northern fur seal caught (86), with successful dives (13.3 s) being signifi- (323) indicates that, although swim velocity is similar be- cantly shorter than unsuccessful ones (22.7 s). Maximum tween the two (1.65 vs. 1.35 ds, respectively), duration dive duration recorded is 96 s (309). Although otters will is substantially different, being 3.2 min for the former and dive to depths over 10 m, their preference seems to be to 1.2 min for the latter. The proportion of time spent at the 846 PATRICK J. BUTLER AND DAVID R. JONES

A Time of day lhl

'1 I FIG. 7. Time-depth records from 2 free-ranging king - 300 penguins. A: transition from day to night types of diving E and hark to day-tme dives. R: n series of deep dives during - 6 Time of day lhl day, most of which are separated by a few shallow dives. 5 [Modified from Kooyman et al. (221).] a 07 08 09 10 11 12 13 IL 15 16

surface is 1.5-3 times greater during deep diving behavior, per indicating how it was possible to deploy and retrieve and thcsc authors conclude that the deeper, longer dives mechanical recording devices with Weddell seals, because must involve substantial anaerobic metabol'im. these animals dive under Antarctic ice and have to surface For the larger lactating female California sea lions, at specific places where natural cracks in the ice exist. Zulophs cutiJwrr~iunus,the pattern of diving is similar Thus, by transporting the seals long distances from natu- to that of fur seals (145). Mean dive duration is 2.1 min ral cracks, boring an artificial hole in the ice and releasing (maximum, 9.9 min), ruean surface interval is 1.3 ruurr, iurd Lhe seal(s) a1 Llris hole, Kooyrr~anand co-workers had a mean dive depth is 61.8 m (maximum estimated depth, field situation where it was possible to obtain behavioral 274 nlj. ks ir~dicatedpreviously for other species, there and physiological data from a freely diving phocid. is a tendency for dives during daylight hours to be deeper Under these conditions, mean dive duration was 11.5 arrd for those at night to be shallower. The swimming min (maximum 82 rnin), and mean depth was 118 m (maxi- speed at which cost of transport is minimum for this spe- mum 626 m) (74). As these authors point out, there are, as cies is --2 nds (I&), which is similar to the mean value will1 olher species, lrtiuly factors: that can influence the div- for the Galapagos sea lion, Zalophus califoomianus wol- ing behavior of WeddeU seals such as season, geography, lcbael~i,when at sca (328). and location of prey, and merely giving mean and maximum The most intensively studied phocid seal is the Wed- values hides this variability. From Kooyman et al. (235), it dell seal, and although much of the behavior of this animal is clear that these seals, during the austral summer at least, has been reported in many of the previous reviews on perform a number of dives in relatively quick succession diving cited in section I, because of the central importance for 10-12 h around midday and rest for the remainder of of these data in the development of our ideas on the physi- the time (Fig. 8). The surface interval between each of these ology of freely diving mammals, a brief account of the dives is only of Z to 4-min duration. When Kooyman et al. diving behavior of this seal is given here. (235) examined the frequency diitribution of dive durations As long ago as 1965, Kooyman (214) published a pa- (Fig. 9), they found that only 8%of the dives fmm an artifi-

Time of day 0 12 0 12 0 12 0 12 0

RG. 8. Time-depth record of a free-ranging Wed- dell seal over a +day period. [Modifled from Koogman 1101: 1 2301 Jztlg 1987 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS

mi;. 9. Frequency distribution of dive duration for Weddell seals. A: 1,057 dives from 6 seals divlng from a hole cut in Antarctic ice and to which they had to return 20 to breathe. R: 4,GOl dives from 22 completely free-ranging seals. Madmum dive duration recorded during these oh- servations was 73 rnin. [Modh?ed from Koayrnan et al. '0 ,,,,, I

Dive duration lmin 1

cial hole exceeded 26-min duration and only 2.7% of those elephant seals. The female northern elephant seal, at least, from seals using natural breathing holes were longer than returns to the rookery to moult after being at sea for 10 26-min duration. For the few dives that were longer than wk after the nursing of her pup. 20-26 min in duration, the recovely time increased progres- The astonishing fact to emerge from a study of these sively more with dive duration so that when returning from animals (246, 248) is that they, unlike the Weddell seal, a 60-min dive, the seals "were exhausted and usually slept dive more or less continuously for the whole 10-wk period for several hours or more in the ice hole before making any (Fig. 10). There is a marked diurnal pattern to their diving. more dives." These data are important because they relate Only rarely is the continuous diving pattern interrupted. to metabolic data, some of which were obtained from the Less than 0.5% of the dives are followed by extended same seals, which are discussed in section 111 and have aven surface intervals, of mean duration 51.9 min. In fact, there ri.~to a concept., the aerobic dive limit, that is frequently is usually little variation in snrface interval with dive dura- used in diving physiology. tion, even after the longest dives (Fig. ll), which, in south- Despile ltie dificulty of recovering data from most em elephaut seals at least, may be followed by further freely diving phocids, success has been achieved with long dives, with short surface intervals (Fig. 12). The re- somc spccics. By attaching very high frequency (VHF) sult is that during the 10 wk at sea between breeding and radiotransmitters to the animals, Fedak et al. (142) were moult, female northern elephant seals spend 83-90% of able to locate harbor (common) seals in the field and the time underwater. Average swim speed di~ringdiving monitor their diving behavior. Mean dive durdLio11 was 2 is -1.3 IIL/S,arid descer~lswi111 speed is greaLer Iur deeper min, and the maximum recorded dive was -6 min in dura- dives (247). This is counter lo Ute predicliori of Thorrrpso~~ lieu. Must surlace inte~lialswere -30 s long. Time-depth et al. (380). recorders have been used with northern and southern Similar studics havc bccn pcrformcd on both male and

Days 0 10 20 30 LO 50 60 70 80

no. 10. Compressed diving record of a female northern elephant seal during whole of her period at sea between iaulaliur, and moult. Her initial mass was 291 kg, and during 81 days at sea she performed 5,657 dives, the longral of which was 44.4 min and the deepest was to 1,093 m. It is not possihlr to disrem normal surface periods, but those of an hour or morc appear as whit^ vertical lines. Note their paucity and irregularity. Regular variations in depth reflect die1 patteln of diving, with deepest dives occurring at midday. [Modified from Le Boeuf et a1 (248).1 PATRICK J. BUTLER AND DAVID R. JONES

3 - - .-C

~-0. - - - -4 5 m ...... 6.' " . .. u2 - - l/+,iIS -. .,,,,.. F 2 o * c m +;- n Q .-E " , 1 :: &og:"e*o - 35 ' :goo;o~~~o n 2. u-n 25 u - ""1""~~'"1"~~1'~"~~~"~ C 0 ' 15 0 5 10 15 20 25 30 ., ., U Dive duration (min) J 5 2 LO FIG. 13. Relationship between time at surface and duration of preceding dive for 3 Cree-ra~ginygrey seals during traveiing and foraging dives. [Modified from Thompson and Fed& (378).]

So, elephant seals may illustrate the extreme of pho- 15 cid diving belrdviur-. However, recent studies on gray seals, H. grypus (378, 379), indicate that there are similarities 5 . . .. *he. . l:Lo TO 20 30 40 so betweenthe use of elephant radiotransmitters seals and toother locate phocid the animals, species. depth With

Duration of preceding dive (mi") and velocity meters could be attachcd to and retrieved from the seals from an ocean-going yacht. These seals did KG. 11. Scnttcr plots of surface interval in relation to duration of 1101 relrlg~~at sea continuously, but hauled out periodi- preceding dive for 3 free-ranging female nonhem elephant seals. Aster- isks hdicate surface intelvals that exceeded 35 min. [~,,difled from Le calls. However, when at sea, they spent a large proportion Boeuf et al. (246).] of their timc underwater. Even when resting in water next to the haul-out sites, they were submerged for almost 85% of the time (379). Although there is an increase in surface female southern elephant seals, but also extending between internal with increasing dive duration for relatively short moult and the next breeding season (178). Attachment of dives, this is not so for dives in excess of 7-min duration radiotransmitters as well as TDRs aided the location of the (Fig. 13). When traveling, the seals have Vshaped depth animals. The diving behavior of the southern elephant seals versus time dive profiles and swim at 1-2 mls, but when is similar to that of the northern species (Table 3), although feeding, the depth versus time dive profiles have a square the mhum recorded duration is substantially longer. shape and, when at the bottom, the animals swim very From this study, it is clear that the relentless diving behavior slowly, if at all. Similar conclusions were reached by of elephant seals continues for the whole of the 8 mo they Crocker et al. (95) studying a single female northern ele- are at sea between molt and the next breeding season and, phant seal. The minimum cost of transport speed of gray on average, 8996 of this time is spent underwater. Although seals is -1.5 ds(380). these seals, like the northern species, do occasionally re- Although cetaceans are the best-adapted mammals to main at the surface for an how or two, the maximum sur- an aquatic existence, there is, unfortnnately, little informa- face interval recorded for one female was only 5 rnin during tion on their behavior and physiology (218). What data there 40 days at sea! Elephant seals are not divers, they are sw- are indicate that at least some of these species are not facers (236). outstanding divers. For example, a group of humpback

Time of day lhl

21:OO 2L:OO 03:OO 06:OO 09:OO FIG. 12. Selected trace from dive record of a free-ranging southern female I , 12j00 ; - elephant seal (320 kg) showing a dive of E 250 120-mim duration that was followed by - continuous di"ing acti"ity, with no ex^ 5 tended surface periods. [Modified from P a 750 Hindell et al. (179).] July 1997 PHYSIOLOGY OF DIVING 0F BIRDS AND MAMMALS 849 whales, Megaptera novaeangliae, performed most of their dives (85%) to <60 m, with few (<39%) exceeding 120 m. Mean dive duration was 2.8 min, and maximum duration was 21.1 min. Average speed during descent was 1.7 ds, and that during ascent was 1.9 ds(120-122). Mean dive duration of bowhead whales in the Beaufort Sea varied between years. In 1980 and 1981, the values were similar and averaged 3.4 min (maximum 17.4 rnin), whereas in 1982, the mean and maximum values were 12.1rnin and >30 rnin, respectively (417). For gray whales, Eschrichtius robustus, mean dive duration was 3.2 min with a maximum of -7 rnin (275, 418). Further examples of the fact that the natural behavior of an animal may not always be an indication of its full potential are the white or beluga whales (Delphinapterus leucas). These are benthic feeders and, because they often swim into shallow water, they have been considered not to be very deep divers, with maximum depths reported to be only 20 m (see Ref. 336). These authors, however, were able to train a pair of heloga whales to dive to much deeper levels. They were capable of diving to 400 m with apparent ease, although dives of 400-600 In appeared to be more difficult. Nonetheless, the deepest trained dive was to 647 m, and the longest was 16.8 min. In deeper water, Martin and Smith (278) recorded natural dives to a maximum depth of 350 m and for a maximnm duration of 13.7 min. Also, the deepest dives, to 251 m, recorded iJ,,10 from three nanvhales (Monodon monoeeros) off northern 0 10 20 30 40 50 60 70 BaTfu~Isliu~d ax ihouglrl Lo represenl the depLh of the Duration of dive(rnin) water column in which these animals routinely dove (277). Fa. 14. Plot of peak concentration of lactate in arterial blood The maximum dive duration of 15.1 min is close to the against duration of preceding dive for 3 Weddell seals diving from and mode of the frequency distribution of dive duration of rehlrning to a hole cut in Antarctic ire Diamond on ahscis~represents 12-13 min. average concentration of lactate in 3 seals when at rest. [Modfied from Kooyman et al. (235).] The most accomplished diver among the cetaceans appears to be the sperm whale, PILyst.1t.r. culudor&.By attaching acoustic transponder tags to these animals, Wat- is, no dnnbt, an important fact,or in this behavior. kins et al. (390) were able to monitor their diving behavior Whether or not other deep-diving birds and mammals (see Table 3). Average underwater swim speed was adopt lhis behavior has ye1 Lo be determined, but the between 0.75 and 1.2 ds, but in one whale, speeds of implications of this discovery are very important. 2 dsand even up to 4 dswere recorded. An exciting recent development has been the de- ployment of a small (30 cm long, 13 cm diameter, 2.2 111. METABOLIC RATE AND METABOLISM kg in water), self-contained video system and data logger with northern elephant seals and bottlenose dolphins, Tursiops trmncutus (T. M. Williams, B. Le Boeuf, A. Aerobic Dive Limit R. Davis, D. Crocker, and R. Skrovan, unpublished data). The data indicate that, when moving horizontally It is clear from the previous section that the diving underwater, the animals engage in burst and glide swim- performance of an aquatic bird or mammal is not always ming, which may be energy efficient (cf. Ref. 25 for determined by its own physiological limitations. Often, review of studies on fish). Of much more significance environmental factors such as light level andlor the behav- is that, doring descent, both species ceased active pro- ior of the prey influence the behavior of the diver. How- pulsive strokes below certain depths (73 i 22 m for ever, t,here is evidence from adult Weddell seals that -90% elephant seals, 64 % 12 m for the dolphins) and glided or dives (feeding dives) are of <20-1nin duration (Fig. 9) passively down. The compression of gas stores, thus and for those rew that are longer (exploratoly dives), reducing buoyancy as hydrostatic pressure increases, there is a long recovery period (235). These authors also 850 PATRICK J. BUTLER AND DAVID R. JONES Volume 77 found that, after the majority of shorter (<20 min) dives, (31). Also, for the oxygen hound to the myoglobin to be there was no accuinulal.ion (11" 1)1(1odlactate, whereas after of use, the Po, in the blood would have to be exceptionally the longer (>20 min) ones there was (Fig. 141. Peak post- low or, maybe, the muscles have to become ischemic, dive blood lactate conccntration was greater with longer because the Po, at which myoglobin is half-saturated (Pm) dives, and after a dive of 60-min duration, it was 50 times is very low. the resting value. Thus the shorter dives, which consti- To illustrate the general prnhlem, earlier calculations tuted the vast majority of those performed by the seals, of the usable oxygen stores for Weddell seals give a value were aerobic in nature, whereas during the relatively in- of 59 ml'kg (75, 218). Thus a 450 kg animal has 2.5 1 of frequent, longer dives, a proportion of ATP production usable oxygen. Castellini et al. (75) found that the mean was by anaerobiosis. Based on these data, Kooyrnan et rate of oxygen consumption (oxygen consu~r~ptiunmea- al. (220) coined the term aerobic dive limit (ADL), which sured during the surface period, averaged over the dura- in this context denotes the dive duration up to which tion of the previous dive and the time at thc surface) for there is no increase in postdive blood lactate concentra- all dives was 4.5 ml .min-' . kg-'. These authors divided tion. From Figure 14, this is -20 min for fully grown adult the dives into relat,ively short (~14min) and relatively Weddell seals [note, Kooyman et al. (235) quote 26 min long (>14 min) ones. For the former, mean rate of oxygen as the critical time]. consumption was 5 ~rd.min-'.kg-' (see Fig. 191, and this Although until very recently the only information on would give a calculated ADL for a 450 kg seal of only 11.8 postdive blood lactate concentrations was from studies min. This would mcan that there is a substantial discrep- on Weddell seals, many authors, studying other species of ancy (-40%) between it and that obtained from the deter- aquatic birds or mammals, have calculated the theoretical mination of postdive blood lactate concentrations (20 ADL for those species. This is done on the basis of calcu- min). However, Ponganis et al. (324) have more recently lated or, in only one instance, measured oxygen slures. themselves nleasured plasma and blood volumes and he- Packer et al. (312) used ''0, LO lneasure the oxygen stores moglobin and myoglobin concentrations in Weddell seals. of harbor seals on land. All other values come from calcu- From the values obtained, they calculate the usable oxy- lating the oxygen content of various oxygen storage coni- gen stores lo be 86.2 mVkg and, with the use of an oxygen partments in the body. Calcr~latedoxygen stores divided consumption value of 4.5 ml .min-'. kg-', calculate an by calculated or rneasurcd ratc of oxygen consumption ADL of 19.1 min. This is almost identical to the value during diving yields the calculated ADL. Clearly, if both (-20 min) determined from Figure 14. Even if an oxygen of these variables are calculated, there will be anumber of consumption value of 5 1111 min-' kg-' is used, a calculated assumptions which give rise to "risks of embarrassment" ADL of 17.2 minis obtained, which is only 14% less Wan (218). If not measured in some way, oxygen consumption that determined from postdive lactatc concentrations. If during diving may be extrapolated from exercise studies the critical dive duration based on blood lactate concen- performed in a laboratory or determined as some multiple tration is taken as 26 min (235), the discrepancy is some- of resting inetabolic rate, usually from an allometric rela- what greater (-34%). Furthermore, there appears to be a tionship of body mass versus resting metabolic rate. Mellr- revelual of the relationship hetween ADL calculated from ods of determining the volume of the respiratory system O2 stores and during diving and that determined from which involve restraint of the animal in any way may yield postdivc lactatc concentration in younger, smaller Wed- values that are significantly less than those obtained from dell seals, with the calculated value being -6Wh greater unrestrained animals of t,he same speries (364, 'I'able 1). than that determined from postdive lactate concentrations Even if the amount of oxygen stored in the various in pups (44). compartments of the body is assessed accurately, assump- Another aspect of the calculated ADL is that it implies tions still have to be made as to what extent they can be that all of the usable oxygen stores are consumed. How- used by the animal when dctcrmining usable oxygen ever, it is clear from Figure 14 that Weddell seals can stores (211). In birds, for example, a substantial propor- and sometimes do remain submerged for longer durations tion of the available oxygen is in the respiratory system than the ADL based on postdive blood lactate concentra- (-50% in the tufted duck), and there is clear evidence, tion. It is assumed, therefore, that during such extended from the low oxygen content of the first exhalation upon dives there is sufficient oxygen remaining in the stores at surfacing of the Humboldt penguin (62), that at least some least for the central nervous system and heart. This, of of this oxygen is used during shallow, voluntary dives. course, further exaggerates the discrepancy between ADL Eliassen (129) suggested that movements of the limbs obtained by the two different methods. Possible explana- during underwater locomotion could cause movement of tions for the discrepancy will be discussed later, but the air between the anterior and posterior groups of air sacs reason for mentioning ADL at this stage is to highlight the and, therefore, through the gas exchange regions of the likely imprecision of many of the calculated values quoted lungs. Measurements of pressures in these air sacs of the later in the review, particularly in the context of the origi- tufted duck while diving indicate that this is a possibility nal definition of Kooyman et al. (220). .1~tLg1997 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 851

B. Methods for Determining Metabolic Rate Metabolic rate should be given as the rate of energy of Diving Animalfi expenditure, i.e., power (W = 1 .Its). The conversion of rate of oxygen consumption to W depends on the meta- There are two major methods for determining the holir snhstrate; for pnre carbohydrate metabolism, i.e., metabolic rate of a freely diving bird or mammal: one is with an RQ of 1, 1 ml O2Is = 21.1 W, whereas for pure fat direct respiro~rrelryand Lhe oUrer is Lo use doubly la- metabolism, i.e., will an RQ of 0.71, 1 it11 OLIs = 19.6 W beled water, DLW, (HTi80, D2I80)to determine the rate (41, 263). of COy production (256). This is then converted to oxygen consumption either by knowing or assuming the respiratory exchange ratio, which in the long term is C. Birdv equivalent to the respiratory quotient (RQ). Apart from the latter (359), there are a number of assumptions asso- Respirometry has been used to estimate oxygen con- ciated with this technique that have been discussed by sumption during voluntary diving by tufted ducks (414). Nagy (304) and Tatner and Bryant (376), among others. The birds dove from an open-circuit respirometer posi- An even less direct method is to use tritiated water tioned at the water surface of a 1.7-m-deep tank. Oxygen (HTO) to determine water turnover, use this to estimate consumption between dives was measured by a fast-re- food intake and convert this to energy expenditure sponding mass spectrometer, and a linear multiple regres- (225). All of these methods have their limitations with sion analysis was performed between dive duration, the free-ranging animals. With a very few exceptions, direct succeeding duration at the surface, and the rate of oxygen respirometry is not feasible. Even when it is used, it can uptake during the surface period. The regression coeffi- only measure rate of oxygen consumption between dives cients represent the mean rate of oxygen consumption at and not the rate at which 0, is being metabolized during mean dive duration and at mean duration at the surface. dives. With labeled water, it is necessary to recapture So, for six tufted ducks, mean oxygen consumption at a the animal after a sufficient delay following injection of mean dive duration of 14.4 s in water at 13.5"C was esti- the isotopes and release, to enable a reasonable decline mated at 57 ml .kg .min-' at standard temperature and in the enrichment of ''0 (a minimum of one biological pressure, dry (STPD). This is 3.5 times the resting value half-life), but before the enrichment is too low (a maxi- and not significantly different from that obtained from the mum of three biological half-lives) (305). What is more, same birds when swimming at their maximum sustainable biological half-life varies, depending on the mass of the speed at the surface (63 ml .kg-'. min-' STPD). Thus, for animal and hetween the maor taxa (305, 376). It also, tufted ducks, feeding undenvater is a very costly activity. of course, depends on the level of activity of the animal. However, on the basis of usable oxygen stores (45 mVkg; The other problem is Lhat Llre technique only gives an Refs. 211, 414), this species should be able to remain average value for energy expenditure between the two submerged and metabolize aerobically at the level given sampling points. above for -50 s. This is some 2.5 times longer than the A new method that is being tested in the field is to prekemed dive duration of these animals on a 1.9- to 2.8- use heart rate as an indicator of oxygen nptake. I.abora- m-deep pond (367) and 4 s longer than the maximum tory studies have demonstrated that there is a good recorded dive duration. This would suggest that these relationship between these two variables in a number ducks irrelaboli~econrylelely aerobically during most, if of species, including aquatic birds and mammals (22, not all, voluntary dives, using oxygen stored in the body 63, 141, 308, 414), and validation studies, both in the and rapidly replacing it upon surfacing. laboratory and in the field, indicate that heart rate is In two thorough studies, Stephenson (363, 364) mea- at least as accurate as DLW for determining oxygen sured buoyancy and volumes of the respiratoly system and consumption (20, 23, 38). In fact, the study by Boyd et plumage of lesser scaup, AyLhyu ufluruis, actively swurur~i~g al. (38) indicates that the DLW may overestimate fleld to a depth of 1.5 m for their food. He found that buoyancy metabolic rate in marine n~ammals(or at least in otariid accounts for 62% of the rrrechanical cost of descent and seals) by as much as 40%. Thus, with the use of a data- 87% of the cost of remaining at the food patch under these logging system (415), heart rate can be monitored and conditior~s.The birds renrahred subrrrerged for an average oxygen consumption estimated for at least some spe- of 11.9-13.5 s and, during this time, blroyancy fell by -200h cies of aquatic birds and mammals when at sea for as a result ol loss or ak frorrr the plumage and co~upression several weeks. It is also possible with this technique by the hydrostatic pressure. The relative importance of these to estimate metabolic rate associated with particular two factors will vary with dive depth and duration. In Ste- types of behavior (21). Unfortunately, it will only en- phenson's (364) experiments, the respiratory system con- able metabolic rate during a diving bout (i.e., including tributed 52% of the initial buoyancy and 65% of the minimum surface periods) to be determined and not that of diving vzllre. This increme wa.; the resnlt of a los? of almost 56% itself. of the air from the plumage during diving. WiL?on et al. (409) 852 PATRICK J. BUTLER AND DAVID R. JONES Volume 77

and Butler (414), there is a clear negative relationship between mean rate of oxygen consumption at mean dive duration and mean dive duration (Fig. 15). The different energy costs of descent, botto~ntime, and ascent (262, 363) cannot completely explain this phenomenon because of the very small descent distance. With the irnportancc of buoyancy on the energy cost of diving kept in mind, this could mean that those hirds that are less buoyant and thus have a lower energy cost during diving are able to remain submerged longer; it could also mean that for any individual duck, the rate of oxygen consumption decreases with divc duration as buoyancy decreases during a dive. The latter could result from loss of air from the I I I I I I feathers (364) because, even during shallow dives, air can 5 10 15 20 25 30 35 Mean dive duration (s) be seen leaving the plumage. During deeper dives, compression would also be a

I I; llcl3riomhlp bi r\rc..n mcll, mt. 5.f u\)gtr, cunsruul,llmr contributing factor (261, 409) but, as mentioned earlier, 31 l#~t.r~l(Il!k~lllr~l~~11 ,\(*.) :LIIII lllr311 #hvt duri,l~<,l,(I. ol1uf #lo,kr this could also cause an increase in heat loss across the ic,l#d.#r< Ic. ;lrr ira,ttt 7 hinli rr;li~,~.dra, allw for dlffelenl rlumtinns %? Rgrrssiurt rqualiorl Cur lltese &la, represented by dashed line, is y = body (which could, in turn, result in a reduction of tissue 0.80 - 0.017~(12 = 0.47), where y is Vo2, and x is I,. Open circles are metabolism; see below). However, in the experiments of values fro111 6 ducks diviug spurllaneously to a depth of 1.7 m (414). Bevan et al. (22), huoyancy was still sufficient to allow Reg~essionequation far all of data, represented by solid Line, is y = passive surfacing, even after long dives. Another possibil- 0.838 - 0.019~(? = 0.87). [From Bevan et al. (22). Copyright is held by Company of Biologists Ltd.] ity is that during longer dives, aerobic metabolism may decrease as part of the "classic dive response" with, perhaps, increasing anacrobiosis. found that the volume of air tragped in the feathers is lower The counteraction between air trapped in feathers and body density is greater in deeper diving hirds. Less air acting as an insulator and as a substantial contributor to in the feathers would reduce insulation and increase heat the buoyancy of diving ducks may be exacerbated during loss, so deep diving bids iequire Inore Pal for kwirlsulation winter. For the same group of tufted ducks when at rest which, by increasing body mass, increases the energetic cost on water, rate of oxygen consumption during winter (air of flight. Such are the conflicting selective prcssurcs acting temperature 6"C, water temperature 7.5%) was 90% on volant, aquatic birds, which have been avoided by pen- greater than that during summer (air temperature 26"C, gt~ins. water temperature 23°C); however, mean rate of oxygen Regulation of the volume of the respiratory system consumption at mean dive duration was similar under could also have large (>40°h) effects on buoyancy. Cer- the two conditions (18). Although mean rate of oxygen tainly, tufted ducks that had been trained to swim long consumption at mean surface duration was 50% greater distances (mean distance 6 m, mean duration 25 s) under- during winter, there was only a 43%greater rate of oxygen water for their food had a smaller end-expiratory respira- consumption over a total dive cycle (dive plus subsequent tory volume [I65 mYkg at body temperature and pressure, surface interval) during winter than during summer. It saturated (BTPS)] than that (232 mlikg BTPS) in ducks was concluded that, when actually diving, the "wasted" that performed shorter dives (mean distance, 0.6 m; mean energy of locomotion contributed to thermoregulation. duration, 11 s) (373). Blood volune was greater in those Although this nlay no1 have been con~plelelyadequate, as ducks that had to travel farther for their food, so that deep body temperature was 1°C lower after a diving bout total usable oxygen stores were the same in both groups. in winter compared with that in summer, it docs mean The conclusion was that the reduced volume of the respi- that the energy cost of diving in winter is not as great as ratory system, by reducing the buoyancy of the ducks, might be expected from the elevated oxygen consumption caused a decrease in the aerobic cost ofundenvater swim- of ducks at rest on the water surface. These experiments ming in those that had to travel further for their food. were performed with ducks trained to dive in shallow (0.6 It is not easy to test the effect of dive duration on m) water, so the effect of compression of the air trapped the energy cost of diving in these birds because, in the in the feathers during deeper diving in winter has yet to wild, dive duration is related to depth (117,124). However, be determined. Even so, its potential effect may be partly a method has been devised, using a computer-controlled offset by increased tissue insulation as a result of periph- system of lights, whereby tufted ducks have been trained eral vasoconstriction (201). Resting in cold water also to dive for up to 45-s duration in 0.6 m of water (22). With causes a substantial increase in metabolic rate in other the use of data fro111 this study and from that of Woakes birds (see Ref. 18 for references). Jtiiy 1997 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 8.53

Tufted ducks and eider ducks feed on benthic organ- those employed by Stephenson et al. (3731, which used a isms, whereas other species, such as grebes, cormorants, low value for the volume of the respiratory system, and, and auks, are active predators of fish. It has been sug- on the basis of this, calcu1att.d a nsahle oxygen slore of gested that the western grebe alternates periods of search 45 ml/kg. A more accurate value may be 55 mVkg (364). for schools of fish during which they recover fully from It is reported that, when inactive in watcr at 6'C, the each dive with periods of more intense diving while they metabolic rate of thick-billed and common murres is ap- exploit a newly discovered school (420). During the latter proximat,ely t,wice the value of that when they are resting periods, the birds may not spend sufficient time at the at 15-20°C (97). The average rate of oxygen uptake for surface between dives to repay fully the oxygen stores Lhick-billed rnmes over complete dive cycles (dive plus and may, therefore, become increasingly reliant upon an- surface interval) in water at 20°C is 1.03 ml 02.s-' . kg-' aerobiosis. Ydenberg and Clark (419) argue that, for such (2.6 timcs the resting value) which is slightly greater than aquatic predators, anaerobic diving may allow better ex- that for tufted ducks diving for a mean duration of 15 s ploitation of a gronp of mobile prey. A study on forcibly (414). Mean dive duration of the rnmes used in the above submerged Pekin ducks indicates that, as well as glycoly- experiments was 41 s, which compares with 55 s for sis, hydrolysis of phosphocreatine (PCr) is also a substan- murres in the field (96). tial anaerobic source of ATP, at least in the pectoral mus- It is clear from Bevan et al. (22) that, for tufted ducks clc (370). Turnover of ATP of the (inactive) pectoral mus- diving in shallow water, mean rate of oxygen consumption cle did not decline during 6.5 min of forced submersion at mean dive duration declines as mean dive duration in these ducks. However, it should be noted that, in re- increases, so that at a mean dive duration of 30 s, mean sponse to electrical stimulation, the muscles of the hind- rate of oxygen consumption is approximately half of the limb of a Pekin duck are unable to maintain strength of value determined at a mean dive duration of 15 s (Fig. contraction for more than 30 s after blood flow was 15). However, mean rate of oxygen uptake over the com- stopped by occlusion of the ischiadic artery and that a plete dive cycle is < 10% lower for dives of 30-s duration similar situation occumed once extl-eme bradycardia was co~nyaredwith that for dives of 15-s duration. If a similar apparent during forced submersion (207). situation exists for murres, then it is likely that the value Thus the locomotory (leg) muscles of a diving duck of oxygen consu~uptionduring diving used by Croll et al. appear unable to contract for more than several seconds (96) for their birds is too high, perhaps 100% too high, iri the absence of oxygen. Also of possible relevance i11 particularly if the effect of depth 011 buoyancy is taken this context is the fact that, if temporarily unable to sur- into account (262, 409). However, on the basis of the face from a voluntary dive, tufted ducks, upon becoming above value (1.03 ml 02.s-'.kg-'), ADL for these birds aware of the situation, exhibit an immediate and profound works out at 45-55 s but, only just over 50% of recorded bradycardia of similar magnitude to that seen during forc- dives of the murres when foraging at sea were within this ible submersion, when there is also peripheral vasocon- duration. Croll et al. (96) seem to favor the idea that these striction and anaerobiosis. They do not perform another murres may accumulate lactate during a series of feeding dive for an hour or more (54, 367). On the other hand, dives (i.e., when prey have been located) and metabolize it tufted ducks can perform a number of successive dives once the bout has terminated (see also above for western when having to swim long horizontal distances (13 m) to grebes). It is quite feasible, however, that most of the dives and from food, even though they remain submerged for are completely aerobic. Clearly, further physiological data longer than usual (average 3.5 s) and develop aprogressive are required from these birds during their natural diving but mild bradycardia during the dives, indicating a partial activity. switch to the classic dive response (367). It is suggested The energy cost of the complete dive cycle in the here, therefore, that if western grebes do resort to anaero- common murre is less than that in thick-billed murres, at biosis when feeding on a school of fish (see above, Ref. 0.64 ml 0,. s-' . kg-' (1.6 times the resting value; see Ref. 419), this is minimal during any one dive, but may well be 97). With the assumption that the usable oxygen store is cumulative during the bout and that diving is terminated similar in the two species, the calculated ADL for the before the oxygen stores are depleted to such an extent common mume, on the basis of these values, is 70 s and that the bird is unable to maintain an adequate supply of again, -50% of recorded dives exceed this duration (387). oxygen to the locomotoly muscles. If this is the case, then The argument concerning mean rate of oxygen consump appropriate cardiovascular adjustments (bradycardia) tion at mean dive duration being 50"/0 of that used by Croll would be expected to occur during such dives. and McLaren (97) (see above) is even more relevant for A detailed study of thick-billed murres (Briinnich's the common murre, because the mean dive duration in the guullemots) suggests that they behave similarly to western experiments of Croll and McLaren (Y7) was 23 s, whereas grebes (96). The murres also feed on schooling fish that mean dive duration in the field is 67 s (387). Thus, if the rise to the surfxe at night. The authors estimated oxygen ADL is actually in excess of 140 s, the duration of all of storage capacity of the birds using similar assumptions to the dives recorded by Wanless et al. (387) was within 854 PATRICK J. BUTLER AND DAVID R. JONES Volume 77 the aerobic limit of this species. On the basis of their one exception, it has only been possible to determine biochemical study of the pectoral muscles of common average metabolic rate while the birds were at sea. mwTes, pigeons, and pheasants, Davis and Guderley (107, For the little penguin, the rate of energy expenditure 108) conclude that the common murre does not rely on when at sea varies between 21 Wkg in winter to 35 Wikg glycolysis to any great extent during diving, and the latter during the late chick-rearing period (3. l and 5.2 times the calculation and comments would support this. resting, or inactive, value, respectively). This presumably Some physiological data are available for other diving reflects the more intense foraging activity as the chicks birds, notably the penguins, and are currently being ob- grow (157). According to these authors, little penguins tained for the South Georgian shag, Phalacrocom eor- spend -95% of their time at sea swimming and diving, gianus. Heart rate of South Georgian shags when resting so the above values are close to the energetic costs of on its nest is, on average, 104 beatslmin, and abdominal these activities. However, because these birds dive for cavity temperature is -39°C (16). When diving, there is a relatively short durations (158), such a high level of aero- progressive fall in abdominal cavity temperature by an bic metabolism during diving would not pose a problem average of 5°C (Fig. 16A), and after a relatively slow de- and is consistent with the basically aerobic nature of cline, heart rate reaches an average minimum of 65 beats! their pectoral muscles (284). Hy incorporating informa- min during relatively short (<3 min) dives and 43 beats/ tion on time budgets, Nagy et al. (306) estimated that, min during relatively long (>3 min) dives. The combina- when swiin~ningurrderwater at sea at an average speed tion of reduced temperature, in at least part of the body, of -1.75 ds, energy expenditure of jackass (African) and a subresting heart rate during diving indicate that penguins, Sph~niscusdemersus, is 27 Wlkg (5.9 times overall aerobic metabolic rate may be lower than for an the value when it is resting on land). Similarly, Chappell equivalent level of activity in air and/or t,here may be an et al. (82) estimated that the cost of swimming for Adklie increase in anaerobiosis. The latter is certainly thought penguins is 26.5 Wlkg. to be the case by Croxall el al. (100) aitd Wanless et al. To calculate ADL for these, and any other penguins, (386), especially for the longer dives. an estimate of usable oxygen stores is required. From The decline in abdominal cavity tcmpcrature dwing Table 4.5 of Kooyman and Davis (223), a value of -45 diving could be the result of ingestion of cold food and, ml Odkg is given for Adelie and gentoo penguins, al- maybe, water (330, 408), but also the result of increased though this appears to have been obtained only from O2 heat loss to the water (52) and/or reduced blood flow to bound to myoglobin and hemoglobin. Keijer and Butler the viscera during diving (17). The important point is that (211) calculated that 20 ml/kg of usable oxygen resided the aninral does rloL experld erler-&y to ir~ainLairll11e tern- in the respiratory system of tufted ducks. In a later paper, perature of the abdominal cavity at its "normal" value. however, Kooyrnan and Ponganis (230) quote the avail- The extent to which this hypothermia affects other parts able oxygen store of emperor penguins to be 58 ml of the body has yet to be determined. The locomotor OJkg and state that this large store "is due mainly to the musclcs (cithcr thosc of thc lcgs or of the wings) will, relatively large gas volume of the air sacs, which we presumably be at normal body temperature, as will the assumed to be similar to that determined for gentoo and ccntral nervous systc~nand hcart. Noncthclcss, reduced Adklie penguins . . ." Because blood oxygen capacity, temperature in part of the body will undoubtedly lead to blood volume, and myoglobin concentration are also sim- a reduction in the metabolic rate of that region. However, ilar in thcsc three spccics of pcnguins, a usable oxygen the return of abdominal cavity temperature to its normal store of 58 ml O,/kg for all three species may seem rea- value after a diving bout (which may he related to the sonable. A smaller amount of oxygen stored in the respi- heat increment of feeding) may result in a substantial ratory system than in that of ducks (on a mass-specific rise in oxygen consnmption ahove its resting level (408). hasis) is certainly consistent wit,h a. lower hnoyancy in Smaller rises in rate of oxygen consumption may also the penguins. However, the values for the volumes of the occur between individual dives, becanse abdominal tem- respirat,ory systems of gent,oo and Adklie pengnins were perature may increase by up to VC, particularly between obtained from restrained birds (233) and could, there- longer dives (Fig. 16B). On the other hand, the wasted fore, be substantially different from those in unre- heat of any exercise at the surface may be used to return strained, naturally diving penguins (cf. Ref. 364). How body termperature to normal. much of the oxygenin the respiratory system of birds is Penguins have been the most intensively studied really available when birds dive to great depths remains group of aquatic birds, and the most widely used tech- to be seen (230). During shallow dives of Humboldt pen- nique to determine the energy cost of foraging is DLW. guins, S. demersus, there is no doubt that gas exchange As indicated in section IIL!~, this technique only gives an does take place during the submersion period (62). average value of COZ production over the experimental So, with the assumption of a usable oxygen store of period, so to determine the energy cost of a specific activ- 58 mVkg and if 1 ml Oz/s = 20.1 W, the calculated ADL of ity, time budgets must also be determined. Even so, with jackass penguins, if metabolizing at the above rate, is 43 s. Juiy 1997 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS

Time (local)

woo

abdominal temperature, and time on water of a female Sonth Georgian shag (2.1 kg) performing relativelv shallow dives over a 12 h ~criod(Ai and a male shag (2.5 kg) ~erfonniiga sehei of relatively deep dives followed by a series of shallower dives over a 6h period (B). [From Bevan and Butler (19). Copyright is held by Springer-Vcrlag.] 856 PATRICK J. BUTLER AND DAVID R. JONES Volz~me77

Mean dive duration of this species is -70 s (maximum organs), whereas metabolic rate between bouts and at 180 s), and only 35% of dives recorded are of <43-s dura- the end of diving activity may be higher than expected tion (412). At-sea metabolic rates of macaroni and gentoo as abdominal temperature is returned to normal. This penguins are slightly lower than that for swimming jack- could, at least in part, explain the discrepancy between ass penguins, at 22.2 Wikg and 14- 19 Wlkg, respectively dive durations and estimates of energy expenditure dur- (111). In the field, the average swimming speed of Adelie ing diving based on the average energy expenditure dur- and gentoo penguins is -2.2 and 1.8 mJs, respectively ing total time at sea. Because dive duration scales to (104). These authors found that, when swimming at -1.8 body surface area in gentno penguins and South Geor- mJs in a static water canal, the metabolic rate of gentoo gian shags, Boyd and Croxall (36) suggest that these penguins was 16.1 Wkg. Culik et al. (104) point out that, birds reduce rrielabolic rate during diving, possibly by in their experiments the penguins had to accelerate and reducing body temperature. We await further, more de- decelerate, at least once, as they swam from one end of tailed, analysis of thc field heart rate and body tempera- the canal to the other. By making allowances for this, they ture data to see if this is the case. calculated the metabolic cost of swimming at 1.8 m/s for Kooyman et al. (221) determined the energetic cost gentoo penguins to be 13.7 Wkg, which is very similar to of foraging of the king penguin to be 10 Wkg (30 the lower at-sea metabolic rate for this species deter- rnl.mini. kg'), and this is similar to the energy cost of mined by Davis et al. (111). However, Culik et al. (104) swimming in a static water canal at 2.2 mJs (102). With do point out that even their correct estimate may be too usable 02stores of 58 mVkg, the calculated ADL is, there- high, because of the "wall effects" of the water canal. fore, -2 min. If this is the case, then 40-45s of all dives, With the use of the above values for usable oxygen store and this includes all deep, feeding dives, exceed the ADL. and at-sea metabolic rates, the calculated AULs of maca- Riiitz and 8ost (330) report that there are large reductions roni and gentoo penguins are 53 and 61-83 s, respectively. in stomach temperature of foraging king penguins. Recent For gentoos, these are much shorter than the mean dive studies have indicated that there is also a substantial re- duration of 2.5 min for their longer dives and 45-5596 of duction in abdominal temperature of king penguins during recorded dives exceeded 83 s (403). However, these data diving activity (51) and that it may fall to as low as 11°C should be treated with some caution because, as already (171a). Thus the energetic cost of diving could be some indicated, these mean values of energy expenditure for 6099 of the total cost of foraging. If this is so, ADL would the duration of time at sea, or even for swimming and be 3.5 min. This would still place -25% of all dives of diving at sea, may be substantially greater than the energy king penguins in excess of this duration. Although there cost of diving itself, and usable oxygen stores may be 2096 are no comparable data available for emperor penguins, or more greater than those used to calculate ADL. their relatively long dive durations and high underwater As for the South Georgian shag, there is a decline swim speeds led Kooyman et al. (232) to conclude that in abdominal cavity temperature in gentoo penguins many of the dives performed by this species require a when at sea (52), and for seven birds this averaged over significant contribution from anaerobic glycolysis. 6'C. With the use of heart rate as an indicator of rate of When swimming in a water channel in water at 2- oxygen consumption (23), the average value of oxygen 4"C, three emperor penguins reached an average maxi- consumption for the whole of the diving period was 26.5 mum oxygen uptake of --45 ml 02.kg-'. min-' (231), al- ml. min-I. kg-' (9.0 Wlkg), which is 2.2 times the value though, on the basis of wing beat frequency, these authors recorded when they were sitting on their nests and only conclude that when diving freely in the sea, oxygen con- 18% greater than that obtained from the same birds rest- sumption would be between 7 and 20 ml .kg-' .min-' ing on water at 5°C. This gives a calculated ADL of 2.2 (2.2-6.7 Wkg). This would give an ADL of -3-4.5 min, min. Even so, -40% of recorded dives exceed this dura- which again is less than the duration of >40% of their tion (403). Gentoo penguins do not dive continuously natural dives (228). The latter authors conclude, from the when at sea and when they are not diving, heart rate relationship of surface interval to duration of the previous and, therefore, oxygen consumption are elevated. This dive in these birds (see sect. [A), that the true ADL is may reflect surface swimming andlor t,hermogenesis -7-8 min (the point at which surface interval begins sub- (perhaps by way of heat increment of feeding) and the stantially to increase). If so, this would mean that only return of abdominal temperature toward normal. It ap- 4% of foraging dives exceed the ADL and that most of pears that, when gentoo penguins are resting on water these are to depths of >300 m. (at 5"C), there is a substantial increase in metabolic rate, In an important study, Ponganis et al. (327) have man- presumably to maintain body temperature (52). How- aged tn determine postdive lartate concentrations for nm- ever, the metabolic rate during diving itself may be lower peror penguins, which indicate that the true ADL, as per than expected as a result of the fall in abdominal temper- the original defiriitiori by Kuoyrnan el al. (220), is belweerl ature (assuming a &,, of 2, a 10°C drop in temperature 5.5 and 6.5 min. Thus, for the second species (in addition would reflect a 50%fall in metabolic rate in these tissues/ to the Weddell seal), a discrepancy between the ADL de- July 19.97 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 857 termined from postdive lactate concentrations and the corded dive duration was just over 120 s. One possible calculated ADL is apparent, but the true ADL indicates explanation for these dives of long duration is that, whcn that most of the dives performed by the animals are zro- under ice, the animals are able to recover previously ex- bic. As well as regional hypothermia and a low drag coef- pelled gas (272). It is interesting to note also that during ficient, gliding during descent, as recently reported for all dives, regardless of duration or tcmpcraturc of the elephant seals and dolphins (see sect. IB) may be an water, heart rate fell below the value recorded when the important factor in keeping the energy costs of diving at animals were resting in air (273 beatsln~in)or resting in a minimal level. water (283 beatshin; Ref. 273). During foraging dives it averaged 111 beatslmin, whereas during escape dives it was 73 beatslmin. Diving heart rate was lower in animals D. Mammals with reduced abdominal temperature. So, it appears that whether dives are aerobic or not, there is, unlike the situa- Those aquatic mammals that use fur as a form of tion in most hirds, a dear hradycrartliw rloring voluntary insulation have similar problems to birds, in as much as submersion in this aquatic mammal. the air trapped in the fur increases the buoyancy of the At-sea metabolic rate has bccn determined for three animal, and if the air layer is reduced for any reason, species of otariid seals using the DLW technique, and for e.g., loss andfor compression, conductive heat loss will one of these species, the heart rate method is also being increase. Regardless of the insulative properties of their deployed. These studies have indicated how these differ- fur, being in water, even at 29-30"C, causes a 1.9 tunes ent species respond to differences in prey availability. increase in resting oxygen consumption of mnskrats, On- Rate of energy expenditure while at sea varied between datra zibethicus, above that in air (1.5 vs. 0.8 ml years in lactating northern fur seals (91). In 1981, energy 02.g-' .h-'; Ref. 274). The energy cost of diving under expenditure was 6.6 Wkg (1.4 times the resting value), these conditions is 2.2 ml 02.g-' .h-' (2.75 times the rest- whereas in 1982, it was 9.8 Wkg (twice the resting value). ing value in air). As the usable oxygen store (not the total Duration of the foraging trips was similar for the 2 years, O2 store) in muskrats in summer is -21 ml 0, STPD/kg so the conclusion is that foraging effort was greater in (270), the calculated ADL is 33.7 s. In winter, mainly as a 1982, perhaps in response to changes in prey abundance result of a 32% increase in blood volume, the usable oxy- or distribution (88). The increased energy expenditure in gen store increases to 30 ml O2 STPD/kg (270), thus giving 1982 was matched by increased energy intake so that net rise to an increase in the calculated ADL. energy gain was similar in both years (1.77 and 1.83 W/ It is known that, under laboratoly conditions, volun- kg). It could be that in 1982 there was a relatively low tary immersion in cold water causes a reduction in abdom- abundance of high-energy prey, such as pollock, which inal temperature, and increased postdive oxygen con- occur at depth so that the females had to perform a sumption in muskrats (268). Rate of oxygen consumption greater number of relatively shallow dives to obtain the remains elevated after the animals leave the water, and it low-energy prey, squid (88). is during this period that abdominal temperature returns There is evidence to indicate that deep-diving females to normal. It is likely, therefore, that because of the re- make fewer dives and expend less energy than those that duced abdominal temperature and the fact that nonshiv- perform shallow dives (91, 163). During normal years, ering thern~ogmesisappears to be inhibited when the ani- northern fur seals spend 57% of their time at sea being mals are diving in cold water (269), the energy cost of active at the surface, 26% &ving, and 17% resting (la), diving in cold water may be no greater than that in warm so there is scope for increasing their energy expenditure water, which would give a minimum ADL in winter of while at sea. However, during the El Niiio of 1983, there 49 s. When at the surface, and particularly after leaving was a significant (30%) increase in thc duration of foraging the water, oxygen consumption is elevated as a result of trips of lactating northern fur seals compared with the nonshivering thermogenesis in the brown adipose tissue. preceding and following years (113). In winter, all foraging movements of these animals Lactating Antarctic fur seals spend 60°% of their time may occur under ice between shelters and from the shel- at sea swimming, 35% diving, and only 5% of their time ters to the feeding areas. Movements between shelters resting (224). This species, therefore, has much less scope may acconnt for the longest dives routinely encountered for increasing its foraging effort, and in fact, when prey by muskrats, and MarArthur (271) found t,hat for the shel- wm scarce in 1984, the average duration of their foraging ters located in emergent vegetation, virtually all of the trips was 85% greater (8.4 days) than it was in 1985 (90, transit dives were within the calculated winter ADL (49 s). see also Ref. 34 for the seasons 1988/89 to 1993/94), However, for those shelters constructed in open bays, 2090 whereas at-sea energy expenditure was similar (9.2 and of thcm rcquircd transit times in cxccss of the ADL, and 9.8 Wkg, rcspcctively, which was approximately twice dives of 60- to 95-s duration were documented in four the resting value). From these two data sets, it would muskrats swimming from distant shelters. Maximum re- appear that the energy cost of diving and swimming in 858 PATRICK J. BUTLER AND DAVID R. JONES Volume 77 these two species of fur seal is -9.5 Wikg (28.3 ml 02.min-' . kg-'). If the usable oxygen stores of otariids are -40 mlkg (163, 218), this gives an estimated ADL of 1.4 min. Taken at face value, this would indicate that -8W of Ltre dives perfollned by the Antarctic fur seal (35) are within the ADL and that the shorter, shallower dives perfomled by the northern fur seal (mean duration, 1.2 min) are also within the ADL. However, the longer, deeper dives performed by this species (mean duration, 3.2 rnin) are well in excess of the ADL, thus supporting the proposal by Ponganis et al. (323) that they must involve substantial anaerobic metabolism. With the use of the heart rate method, estimated rate of oxygen consumpt,ion of lactating Antarct,ic fur seals resting on shore (52) is similar to the values quoted by Costa and Gentry (91) for northern fur seals and by Costa and Trillmich (92) for Antarctic fur seals. Both the latter authors used the DLW method. However, when estimated from the heart rat,e method, rate of oxygen cons~impt,ion of Antarctic fur seals foraging at sea is some 75% of the values obtained by the above authors using the DLW method. This should be viewed in the context of the results of Boyd et al. (38), which indicate that the DLW method may overestimate at-sea energy expenditure by as much as 40% in otariids. Taking this lower value of the energy cost of foraging gives an ADL of 1.9 min. This wo~rldmean that 90U/oof the dives performed by Antarctic Dive time(min) fur seals are within their ADL. Another interesting feature of the sludy using the heart rate method is that the mea- FIG. 17. Plot of time taken for concentration of lactate in arterial blood of Weddell seals to return to avalue of 0.65 mM (or when rernvp~y sured metabolic rate of Antarctic fur seals resting in water curve was extrapolated as a straight line to resting level) after dives of at 6.S°C (6.1 Wkg) is not significantly different from that various duration8 (see Fig. 14 for further details). [Modified from Kaoy estimated while they were traveling to and from their man et al. (235).] foraging area (6.5 Wkg). Thus it appears that these animals may as well be active when in water, since it costs just as much to be inactive (see also Ref. 93). For dives -2 rnls (144, 407), which is at the top end of the range of < 100-s duration, heart rate during submersion of Ant- of average speeds determined for four species of otariid arctic fur seals is similar to that for animals resting in air. (329). With a usable 0,store of 40 mhg, the above rate of For dives of longer duration, heart rate falls below the energy expenditure during diving would give an estimated resting value, which may be indicative of a reduction in ADL of 2.5 min, which would mean that -40% of the dives aerobic metabolism. Thus it appears that most of the dives are in excess i~fthis (145). However, if, as with Anlarclic of these animals are well within their ADL. fur seals, the ADL as estimated above is 75% of its real Calionria sea lions adopt a foraging strategy that is value, and it should be 3.3 min, then only 2M of the dives intermediate between those of the northern and Antarctic are in excess of this. fur scals. During the El NiAo of 1983, the mean duration As already indicated, the most complete set of data of each trip was 7.3 days, and the corresponding mean on the metabolism of a diving mammal has been obtained energy expenditure was 7.8 Wfkg, whereas in the more from the Weddell seal. For adult Weddell seals, ADL, normal year of 1984, the values were 4.2 days and 5.4 W/ based on postdive lactate measurements, is 20-26 min kg, respectively (89). Consistent with these dala are the (235), but it is shorter for smaller animals (44, 220). The facts that in 1983, dive bouts and dives themselves were greater the length of time the dive exceeds the ADL, the of longer duration than those in 1084 (146). Thc valuc of greater is the postdive concentration of blood lactate (Fig. 7.8 W/kg (23 ml 0,. min-'. kg-') is close to maximum rate 14) and, not unexpectedly, the longer it takes for the blood of energy expenditure (25-30 ml 0,. min-'. kg-') mea- lactate concentration to return to its resting level (Fig. sured in California sea lions swimming at -3 rnls in a 17). On very rare occasions, Weddell seals may perform water channel (63, 144, 407). The value of 5.4 W/kg (16.12 a series of short dives (i.e., within the ADL) following a ml 0, + ~nin-'. kg-') is achieved at a swirnn~ingspeed of dive exceeding the ADL and the lactate accumulated dur- July 1997 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 859

14 - the blood during dives within the ADL, following one - which exceeded the ADL and led to the accumulation of 5 12 zm the lactate. It should be stated, however, that Davis et al. (110) fo~lndin t,he resting common seal that only 27% of lactate turnover was oxidized and that this fell to 21% when the seals were swimming at 50% of maximu111 oxg- gen consumption. It is clear from Figures 14, 17, and 18 that the rate of _m 4- . u lactate removal in Weddell seals is similar whether or not the animal continues to dive following a long, natural dive. L-e-. After the long (26 min) dive depicted in Figure 18, blood ~a'""'""'50 52 lactate concentration was 12 m~,and it took -90 min to Time (h) return to its non~iallevel. From Figures 14 and 17, a lactate concentration of 12 mM was reached after a dive of nc. 18. Sample of data (from lwur 48 to 12our 54 of 75-h sampling -SO-min duration, and it took -100 rnin to recover from period) indicating concentration of lactate in venous blood of a young Weddell seal (206 kg) following dives, from a hole cut in ice, of vdous such a dive. However, after forcible submersion, it took durations (in -, givcn ngninst lincs dram bciwccn dots). Note Vmt, slightly longer (120 min) for blood lactate to return to its after a 26-min dive at -49 h and a 28-min dive at -51 h, seal continued to perfurrtt dives, despite elevated cuncentratiuru uf lactate in blood. resting value from a lower concentration (6 mM, see Fig. These subsequent dives were of relatively shon duration, and concentra- 1). Fedak and Thompson (143) discuss the implications tion of lactate in blood progressiveiy decreased. [Modified from of oxidizing lactate after an excessively long dive for sub- Castellin1 et al. (73).] sequent diving behavior. They point out that only one molecule of ATP is produced for each molecule of lactate produced by glycolysis, yet 17 molecules of ATP are pro- ing the latter is metabolized during the former (Fig. 18; duced by the subsequent oxidation of one molecule of Ref. 73). More often than not, however, there is no diving lactate. Thus, if an animal metabolized anaerobically for activity following dives exceeding the ADL until the blood 5 rnin during one dive, it would take -85 rnin (17 X 5) of lactate concentration has been restored to its resting aerobic diving activity, using lactate as a substrate, until value. the level of blood lactate had returned to normal. For There are a number of aspects of lactate production an animal with an ADL of 15 rnin, that would require and removal from studies on terrestrial mammals that approximately six dives within the ADL (cf. Fig. 18). may be relevant to mention here. One important observa- As with the Antarctic fur seal, the rate of energy tion is that lactate is produced in fully oxygenated con- expenditure during diving of the Weddell seal is not sub- tracting muscles (202). However, during moderate hyp- stantially greater than when it is resting in water (76). For oxia, even if muscle oxygen consumption and developed tension remain unaltered, lactate production increases (190). It has been concluded that the level of lactate production by active skeletal muscle may not always be a suitable indicator of oxygen lack (362). In terms of the concentration of lactate in the blood, its rate of production by the skeletal muscles is only one side of the story. The other side is its rate of removal, and an increase in the rate of production may be matched by an increase in the rate of removal. There is evidence that in dogs, approximately one- half of the lactate produced at rest is removed by oxidation and, although turnover rate of lactatc incrcascs during exercise, the proportion removed by oxidation increases to -75% (116). Thus there is evidence that oxida- t,ive skeletal muscle fibers use lactate as a substrate for " Rest Dives Dives oxidative metabolism during exercise (42). It has also (44 min) (>I4 rnin) been suggested the high level of lactate dehydrogenase aclivily irt Wle heat of the Weddell seal is more related to its ability to metabolize lactate rathcr than to produce .- . it (80). Thus both oxidative skeletal muscle fibers and the after dives was averaged over duration of dive plus subsequent time at heart could be important in removing excess lactate from surface. [Data hmCasteU'ini et a1 (75)) 860 PATRICK J. BUTLER AND DAVID R. JONES Volume 77

FIG. 20. Selected period of 21 h (from a total of 118 h) of continuous recording of temperature in aortic blood and diving behavior (depth) of an adult male Weddell seal diving from a lrole cul in ice. [Mudfled Crur~Hill el al. (1711.1

Time ihl

relatively short dives (submersible dual-wavelength near-infrared nately, there are some fundamental errors in the calcula- spectrophotometer (753-754 and 812-814 nm) attached tions, e.g., the total blood 0%content of a seal before a to a naturally diving Weddell seal. Data were collected by dive is not likely to be the 0,content of the arterial blood an on-board data logger. The probe was implanted on the J~LE?/1987 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 861 zoo 1 he t,hree explanations for this. 1) Flow continues and the hemoglobin signal dominates the rnyoglobin signal. Inter- estingly, hemoglobin sarwation during long &ves declines at one-half the rate of that during short dives, qualitatively but not quantitatively similar to myoglobi desatnration during long and short dives. 2) Flow continues, but its distribntion is uneven anrl may he shnnt,ed away from exchange regions so the muscle saturation signal is a compromise between hemoglobin sdturatior~and satwaled and unsaturated myoglobin. 3)There is a dynamic equilibrium established between the rate at which the rnusclc oxygen store is utilized and oxygen is supplied by blood flow, and this equilibrium changes as dives progress (169). In any event, even when all the myoglobin store is utilized, it would only be sufficient to power metabolism for 5 min at a muscle utilization rate of 4.5 mmol ATPIkg. In contrast, if muscle PCr stores are in the same range as those of terrestrial mammals (-30 mmol PCrIkg), then PCr will provide about one-third more ATP than myoglobin-bound oxygen. Hence, the aerobic and anaerobic stores axe capable of maintaining diving muscle metabo- m - lism for 12 min. Due to the much higher affinity of myo- .-C - d globin for oxygen compared with hemoglobin, both aero- 3 on 1 bic and anaerobic energy stores will nu1 be depleled ur~lil z perfusion stops. If muscle perfusion is maintained, it -Z - II)r means that the muscle stores of oxygen bound to myoglo- * 0 I I I I I I I bin as well as high-energy phosphates in the form of PCr 10 15 20 25 30 35 40 45 50 must be viewed as only an emergency supply. Time Imin) Unfortunately, PCr stores of diving mammals have rlc. 21. Dive depth, muscle oxygen saturation, swimming velocity, not been reported, but available evidence suggests that

and 810-nrn~ ~ transmission (relative blood volurnel durinc a 27min dive they are in the same range as those of their terrestrial bv d \VeIt nm rr.atsrulasl~n~e~l~n hasrllllr law UI dive ad\r.& hrllc~n.edb). hy~,rrc~~~~aheart, and skeletal muscle of the hooded seal (Cystophora during recovely. Swimming velocity remained relatively constant. [From cristata), eider duck, and sheep and found no differences Guyton et a]. (169)l between them. Unfortunately, not all the creatine in muscle and heart, for instance, is available to form PCr, since it is either bound or stored in discrete and inaccessible surface of the latissimns dorsi miiscle. Four seals had a intracellular compartments (343). Nevertheless, the cre- monotonic decl'me in O2 saturation during diving with atine level reported by Blix (26) for the eider duck is slopes of 7.33%min-' during relatively short dives (I7 and Jones (370) for the Pekin duck, using nuclear mag- min) (Fig. 21). Aside from the fact that Weddcll seals netic resonance technology. utilize their muscle oxygen stores, it is difficult to deter- An added advantage of utilizing the PCr store is that mine what these results tell us about the actual energy each millimole of PCr broken down removes an almost obtained from the myoglobin oxygen store during submer- equivalent quantity of H+,which is derived from lactate gence. metabolism. Hence, 30 mmol PCr will effectively remove A complicating factor with the technique is that the 30 mmol Hi, which is more than enough to ameliorate hemoglobin signal interferes with that from myoglobin the acidifying effects of all but the most severe bout of and may even dominate the myoglobin signal at high anaerobic glycolysis. Unfortunately, muscle lactate levels flows. Becausc partial pressure of oxygen in the blood at the end of dives are unknown. In an immature harbor will never fall below the P, of myoglobin even during the seal, Scholander et al. (348) recorded a muscle lactate longest dives (331), myoglobin should either be saturated concentration of 44 mrnoVkg after 15 min of forced diving. (flow condition) or desaturated (no-flow condition). This Muscle lactate levels during natural or extended diving was not the case, and muscle saturation during dives oc- are a matter for speculation. Postdiving blood levels fall cupied intermediate values (169). There would seem to in the range of 8-26 mM lactate after 30- to 60-min diving, 862 PATRICK J. BUTLER AND DAVID R. JONES Vol~cw~e77

swimming activity (1 10, 112,407), there is only one determination of daily energy expenditure (DEE) in free-living common seals (335). These authors used the DLW method with one male during the mating season, and they estimated DEE to he 52.5 1121. They do not give the body mass of this seal, but from the details of how DEE was estimated, an average mass of 90 kg is assumed. Neither is there any indication of what proportion of time was spent at sea. With the assumption that 1 ml 0, -20.1 J, this male had an average oxygen cons~~mptionof 20.1 ml &.ruin-'.kg-', which is -5Ph of maximum oxygen consumption (110). If this is the rat,e at which oxygen is consumed during normal diving activity in these seals, Time Iminl then with oxygen stores of 56 mVkg (110), ADL would be 2.8 min. Approximately 80% of the dives are of shorter nc. 22. Eslir~naLioc~ol laelale cor~cent~alio~tin rnwle from blood durat,ion than this (142). Thus common seals also seem values at end of a 42.6rnin dive by a Weddell seal of body mass -400 Q (235). A cumfitting program was used, and best fit to blood data to dive within their ADL, most of the time, as do gray of Kooyman et al. (235) was obtained when exponenual wash out from seals. Reed et al. (334) direct,lymeaq~ired hreath by breath muscle to blood was 5 times faster than lactate loss from blood. Muscle gas exchange in freely diving, captive gray seals and found "lass was assumed to be 120 kg, and blood volume was 50 1. Lactate concentration in muscle at end of dive was 2.5 times peak blood concen- that, as wit11 Weddell seals (75), oxygerl consumption av- tration. eraged over the dive and surface period decreases with increasing dive duration. Using thc avcragc oxygen consumption for all dives (5.2 ml 0,. min-' .kg-', STPD) and respectively (168, 235). Kooyrnam (217) est,imated that calculated usahle oxygen stores of 50 mVkg, the average muscle lactate concentration would approach, or slightly value for ADL would be 9.6 min. Only 6% of dives moni- exceed, 40 mM after the lo~rgestdives. tored from freely diving gray seals in the open sea ex- Figure 22 shows an estimate of muscle lactate con- ceeded 10-min duration (378), and for dives of that dura- centration based on a blood wash-out curve of a Wcddcll tion, mcasurcd oxygen consumption was only 3.5 ml seal after a 42-min extended dive. It was assumed that 02.min ' . kg-' (334). total muscle mass was 1'20 kg and blood volume 50 1. The The most enigmatic of all the diving mammals are exchange between the tissues and blood and the time the elephant seals, which dive more or less continuously course for removal of lactate from the bloodstream were while at sea and for which there is no relationship be- set to reproduce the observed wash-out curve (235). The tween dive duration and subsequent time at the surface result is that muscle lactate concentration must be 2-2.5 (Fig. 11). Le Boeuf et al. (246) conclude that there cannot times peak blood levels, in this case 34-42 mM. Hence, be any net accumulation of lactate in diving northern ele- PCr breakdown, allied to the considerable buffering prop- phant seals, and Hindell et al. (179) state that if southern erties of seal muscle (7Y), will greatly ameliorate acidifi- elephant seals do accumulate high levels of lactate, then cation caused by anaerobic glycolysis. they must have developed unique mechanisms for dealing The relatively low aerobic requirements of underwa- with it while continuing to dive, often for relatively long ter activity mean that the rate at which oxygen is used is durations. Both groups of authors came to the inevitable not excessively high. Usually, the animal surfaces while conclusion that the metabolic rate of these animals during there is still sufficient oxygen for aerobic metabolism and long dives must be substantially, maybe 6Wh, below what quickly replenishes the stores (hyperventilation, tachycar- would be considered as their resting metabolic rate. How- dia, vasodilatation). Thus, at the tissue level, the animal ever, for juveniles and females, making short repetitive is in a steady state. The discontinuous nature of lung dives of 15-min duration, with 1.7-min surface intervals, ventilation and uptake of oxygen from the atmosphere, metabolic rate is lO?h above that of an animal resting on comwared with the situation in terrestrial mammals. the beach and 1.7 times the metabolic rate predicted for causes associated fluctuations in circulatory performance, resting animals. Hence, it seems likely that in elephant but this does not necessarily reflect similar oscillations in seals, metabolic rate declines with dive duration (4). the supply of oxygen to the tissues. If these animals ex- Hindell et al. (179) used the allometric equation ceed their true ADL, then, as we have seen, postdive lac- quoted in Schmidt-Nielsen (346) for resting metabolic rate tate concentration increases as does postdive recovery in terrestrial mammals on the basis that it applies equally time. well to marine mammals (244). They calculated rate of Although there are data on oxygen consumption of oxygen consumption for the lean body mass of their ani- harbor (common) seals at rest and at different levels of mals and took it to be the diving oxygen consumption. Jl~ly19.97 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 863

They used the value of usable 0, stores (79 ml Odkg) may be for these animals), organs such as the liver and quoted in Kooyman (218) to calculate the ADL for all the kidneys may not be perfused, and it is during the rest seals they had studied in an earlierpaper (178). Calcl~latrd dives and ext.ended surface periods (-70 min a day for in this way, ADL averaged 28.9 min for the females and southern elephant seals) that these organs "catch up on 44.9 min for the larger males. It should be noted that, if the backlog" (179). Certainly there is evidence that the total body mass is used instead of lean body mass, the liver cells of seals can tolerate hypoxic (hypoperfused) mean ADL values are 30.8 and 48.3 min, respectively. On conditions by means of the mitochondria entering a "re- the basis of their calculations, Hindell et al. (179) report versible functionally protected" stat,e, with reductions in that only a very small percentage (0.6%) of the dives per- both aerobic and anaerobic metabolism (186). Such a formed by the males exceeded the calculated ADL, state appears possible, because the mitochondrial mem- whereas the corresponding values for postbreeding and branes have low permeabilities to K' and Ca2+ and thus postmoult females were 7 and 44%, respectively. So, even have reduced needs for ATP. Whether or not the kidneys with an already low resting oxygen consumption (i.e., of diving mammals are able to withstand periods of hypo- based on that for terrestrial mammals), many of the dives perfusion by similar mechanisms remains to be seen, but performed by the females exceed the calculated ADL. it is interesting to note that one of the major differences However, Hindell et al. (179) state that if a value of 40% between the ectothermic reptiles and the endothermic of the calculated resting metabolic rate is taken as the (terrestrial) mammals is the low ionic permeability of the diving oxygen consumption, then only nine of the dives cell membranes of the former (130). that they monitored would exceed the calculated ADL. Hochachka (185) continues to speculate on the mech- On the other hand, if the metabolic rate of female southern anisms(~)responsible for a low metabolic rate during div- elephant seals during moult is used as the diving oxygen ing and argues that a reduction in the demand for ATP in consumption (6.4 ml min-I. kg-'; Ref. 33), then even with response to a reduction in the supply of oxygen, rather the usable 0, stores recently determined for Weddell seals than an increase in the anaerobic production of ATP, is (86.2 mlkg; see Ref. 324), the ADL of the postmoult fe- the more likely during diving in a number of marine mam- males would only be 13.5 min. mals (see Refs. 9 and 189 for detailed discussions). Thus Le Boeuf et al. (246) discuss ways in which low diving a reduction in perfusion of some tissues would lead to a metabolic rates may be achieved in elephant seals. One is fall in their metabolic rate (cf. Ref. 397). In this context, that the animals may spend part of their time underwater though, it is worth noting that Hogan and Welch (190) asleep. Another is that, as already mentioned for birds, report hypoxia having no effect on oxygen uptake in rest- the muskrat, and Weddell seals, being in cold water and ing or working gastrocnemius muscle of dogs. Hochachka ingesting cold prey may cause areduction of body temper- and Foreman (187) point out that, as well as metabolism ature and thus a fall in metabolic rate. Certainly it appears being relatively low during diving, its efficiency may be that inactivity on land leads to a fall of 2-3°C in deep optimized by there being a high ratio of pyruvate kinase- body temperature, even of an adult bull (281), so inactivity to-lactate dehydrogenase, thus maximizing the ratio of in water may have an even greater effect. In juvenile ele- aerobicfanaerobic metabolism. phant seals making short dives (15 min), subcutaneous In the absence of physiological data, particularly temperature fell to within 1°C of water temperature, and postdive levels of blood lactate, it is not possible to deter- the thermal gradient across the blubber layer averaged mine how different the two species of elephant seals are only 14°C. Core body temperature decreased 1-2°C at the from the Weddell seal in their metabolic aaustments dur- onset of deep diving, whereas temperature in nonactive ing diving. We do, however, have some physiological data muscle fell by as much as 8°C (4). Even so, the impact of from gray seals, which in some respects behave similarly these temperature changes is not intl~itivelyobvious. The to elephant seals when they are at sea (see sect. IB). [all iu subcular~eouslentyeralure will restrict surracact:Ileal Diving heart rate is lower during longer dives in gray seals loss and reduce the metabolic costs of ihennoregulalion. (378), and during some long dives, it may fall to as low On the other hand, a fall in core temperature of 1-Z0C as 4 beatslmin, and these authors suggest that gray seals, might be expected to rcducc mctabolism by a Qlu effect. to conserve energy during long foraging dives, may em- The analysis by Boyd and Croxall (36) indicates that fe- ploy wait and ambush tactics rather than active pursuit. male elephant seals, like gentoo penguins and South Geor- Nonetheless, calculations similar to t,hose used for other gia~shags, have dive durations that scale to body surface phocid seals indicate that, to remain aerobic during the area. As for the birds, they conclude that female elephant longer dives, metabolic rate would have had to be below seals reduce metabolic rate during dives, possibly by rc- thc calculated resting lcvcls and bclow that measured in ducing body temperature, which enables them to perform resting common seals (112). many dives in excess of their calculated ADL. Thompson and Fedak (378) point out that when trav- Annthnr possihle mechanism of metabolic rate reduc- eling to and from the deepest part of the dive, metabolic tion is that during unusually long dives (whatever they rate would be expected to be two to three times above 864 PATRICK ,I. RTJTI.ER AND DAVI~R. JONES Volume 77 the resting level (112,407). However, like elephant seals, mas~)~-'"(244), oxygen consumption during diving of a this species may passively glide below certain depths (see 20,000-kg sperm whale will be 1.8 ml .min-' . kg-'. If we sect. IB).Regional hypothermia may bc at lcast a partial take the highest usable 02store quoted for a marine mam- answer to the problem, but the low (sometimes very low) mal, i.e., 86 mag for Weddell seals (324), then the ADL heart rates during the longer dives do indicate much re- for sperm whales would be 48 min, which is some 13 min duced tissue perfusion and hence oxygen supply to parts longer than mean dive duration and 25 min less than the of the body. So, we are back either to anaerobiosis, which maximum recorded dive duration for this species (390). on the basis of the behavior of these animals and the It should be noted that there are no data on the masses activity of glycolytic enzymes in marine mammals as a of the sperm whales that have been studied in the field. whole (80) is unlikely, or to an unusually low metabolic If they were substantially larger than 20,000 kg, their meta- rate, the nature and mechanism of which are unclear. A bolic rate would be less than the 1.8 ml .min-'. kg-' men- pointer to the latter may be that the aerobic capacity of tioned above. By recording respiratory frequency and a locomotor muscle in gray seals (as indicated by the making assumptions about tidal volume and oxygen ex- mass-specific activity of citrate synthase) is -30% of that traction, Blix and Folkow (29) estimated that a 4,000-kg in the harbor seal (and even 5036 of that in the rat) and rninke whale (Balaer~optera aculwroslrulu) at sea yet the myoglobin concentration is 3096 greater in gray consumes oxygen at a rate of 10.9 Vmin (-2.73 seals than it is in harbor seals (333). Whether or not hydro- ml . kg-' .min-I). This, coincidentally, is exactly the same static pressure affects the function of oxidative enzymes as the value obtained if the above assumptions are applied remains to be seen, although it appears to have no influ- to a 4,000-kg marine mammal. ence on Michaelis constant of cofactor binding of NADH of muscle M, lactate dehydrogenase from emperor penguins, elephant seals, and sperm whales (98). IV. CONTROI.LINCr METAROLISM: Although, unlike Weddell seals, gray seals do not CARDIORESPIRATORY RELATIONSHIPS spend extended times at the surface following long dives, DURING DMNG at least one of those studied by Thompson and Fedak (378) pcrformcd a scrics of relatively short dives after It would appear from the preceding sections that vol- each excessively long dive and may, therefore, have been untary diving is usually aerobic, with oxygen stores and oxidizing any accumulatcd lactate during the series of their utilization governing aerobic dive times. The prefer- shorter dives. It appears that this species and perhaps ence for aerobic metabolic pathways shown by most div- some others have a number of options available to them. ing animals avoids deleterious effects of accunlulatiny XL- Few metabolic studies have been performed on exer- aerobic end products. Consequently, it follows that modu- cising cetaceans. Williams et al. (406) used the heart rate lation of the aerobic metabolic rate will influence diving method to estimate rate of oxygen consumption of bottle- performance so that the higher the rate of aerobic metabo- nose dolphins, Tur.$iops tnmcotus, swimming next to a lism (M), t,he shorter will he the dive time (DT), hence boat in the open sea. If mean swimming speed when they dive is 2 ds(and this seems to be the case) and it is the speed at which the cost of transport is at its lowest, then during diving, oxygen consumption will be -8 ml .kg-' .min-', which is not significantly different from In terrestrial animals, a close relationship exists between the value nleasured for ani~nalsresting in water a1 25°C perfusion and cellular metabolism. At the whole body and may be a reflection of their hydrodynamic efficiency, level, the relation between cardiac output &) and rate as much as anything else. If usable oxygen stores are 33 of oxygen consumption begins at resting levels and moves ml/kg (see also Ref. 218), the ADL for this species is 4.1 up toward maximum aerobic metabolic rates (388). Even min. Maximum dive duration recorded for ajuvenile male though cardiac output drives tissue blood flow, the rela- bottlenose dolphin off the coast of Wales was <2.5 min tionship between the change in cardiac output with (259). Thus, in the absence of further field data all of the change in metabolic rate is seldom 1:l due to variations dives of this species seem to be well within the ADL. With in perfusion as well as oxygen extraction at the tissues the exception of the sperm whale, cetaceans as a whole [arterial (a) minus mixed venous (v) oxygen content

appear to have rather short dive durations, at least com- (CO,); (Cao, - Cv*)]. Nevertheless, there can be no doubt pared with the phocid seals, and may all, therefore, oper- that during exercise, for example, changes in flow and ate within their ADL. metabolism are well matched. In diving birds and ma- Whether or not sperm whales dive aerobically or not mals, maintenance and tight regulation of aerobic metabo- remains to be seen but, if a 355-kg Weddell seal has an lism is even more crucial than in terrestrial animals, not oxygen consumption of 5 ml .min-' . kgLwhen diving aer- only to avoid significant anaerobic contributions to diving obically (75) and if oxygen consumption scales to (body metabolism but also to ensure that, whcn submcrgcd, snf- July 1987 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 865 ficient oxygen remains in the oxygen stores to maintain Furthermore, cardiorespiratory adjustments upon physiological integrity of those tissues (i.e., heart and surfacing cannot be ignored. Fedak et al. (142) character- brain) that cannot survive without it. Hence, cardiac out- ized the surface interval as an "unproductive but neces- put and tissue perfusion will be important regulators of sary state," suggesting that surface time should be re- aerobic metabolism and metabolic rate and perfusion rate duced to the bare minimum to improve foraging success. can be considered to be indicative of oxygen flux under This can be done in two ways: 1) to enhance ventilation both reduced and enhanced metabolic conditions (184). and gas exchange to reduce the interval necessary for Therefore, during diving, circulation becomes the proxi- repaying oxygen deficits and 2) t,o increase dive time hy mate regulator of metabolism, modulating dive time ensuring that aerobic (in the form of oxygen) and anaerobic (in the lur~nof PCr) sonrces uf energy gyroduction are fully in place before the dive. Hence, circulatoly work normally rcquircd for exercise is completed before the dive, thereby reducing energy costs during the dive (142). Chwges in heart rate and stroke volu~rrecontribute In this respect, it has been suggested that variation in the to changes in cardiac output unequally. Heart rate can dive time-to-surface time ratio (see sect. IIA) is the major vary one and, sometimes, even two orders of magnitude regulator of metabolism in diving animals (94, 160). between diving and surfacing in both birds and mammals. This does not necessarily hold, however, for recovery In contrast, stroke volume is usually unchanged or re- from dives that have been extended into the anaerobic duced by one-half (16, 328). The difference in arteriove- metabolic domain, beyond the ADL, for nous oxygen content could change by three to flve times, but the fairly rapid fall in oxygen content during natural EDT x [Lac-] dives will limit achievement of that level of increase. Hence, for divers, heart rate (f,)provides the most reason- where EDT is extended dive time and [Lac-] is lactate able estimate of aerobic metabolism in the absence of concentration. Lactate is the end product of anaerobic recordings of stroke volume and tissue oxygen extraction metabolism. Recovery from extended dives therefore rep- so that resents not only preparation for the dive ahead but "pay- back time" for costs incurred during the preceding dive.

A. Circulatory Adjustments to Diving This relation implies that heart rate is variable, depending on dive time, so tissue perfusion must also be I. Heart mte variable in intensity. It seems plausible that, in highly ac- During natural diving, virtually all divers show tive species, th~Iocornot.nr m~~scleswo~~ld he perfnsed in changes in heart rate, occurring not only during the dive preference to visceral vascular beds, but in more lethargic per se but also during the predive and postdive periods. animals, the situation is less clear. Nevertheless, if it is Exceptions would appear to be IIumboldt penguins sub- assumed that tissue perfusion does indeed regulate me- merging in an artificial pond (62); Adklie penguins sub- tabolism, then it follows that merging in a still-water canal after spending at least 50 s at the siirfare hetween dives (101); rhinoceros anklets 1 fHxMx- (Cerorhinca monocerata) making feeding dives in a tank TPR (368); harbor seals making short feeding dives, also in a tank (205); and dolphins submcrgiug whilc swimming at where TPR is total peripheral resistance and is indicative high, but not low, swimming spccds at sea (406). In diving of the dcgree of vasoconstriction that occurs during sub- ducks, heart rate is usually above the resting rate during mersion. natural dives (59). A similar response has also been oh- The above relations imply that, if they can be substan- served in cormorants, dippers (Cinclus mezicanus), and tiated, an analysis of voluntary diving performance can even dabbling ducks (mallards, Anus plat?/rh?/nchos) that bc aclucvcd by monitoring dive time, metabolic rate, and have been trained to dive for food (153, 210, 301). In heart rate. In fact, it is the substantiation of these relations contrast, heart rate in both large and small mammals dur- with respect to cardiorespiratory function that forms the ing natural dives is usually well below eupneic resting basis for what follows. In addition, the control of circula- rates (6, 162, 219, 273, 279, 378) or at least the resting tory adjustments is important because of the role that the rates derived from the allometric scaling curve of Stahl circulation plays as the proximate controller of metabo- (361). Hence, if any relationship holds between diving lism. metabolism and heart rate, then avian diving metabolic 866 PATRICK J. BUTLER AND DAVID R. JONES Volume 77

TABLE 4. Heart rate ranges of selected diving mammals and birds during natuml and forced diving

Eupne~e Natural Dive Hem Rate, beatslrnin Forced Dive Body Mass, Hean Rare, Heart Rate.' Common Name* k bedmin Pdve Dive: Lowesti beatdmin Reference No.

Mammals Weddell seal (L. seddellii) 70-85 Elephant seal (M. nn,qustirnstris) 100-120 Gray seal (H. yryw) 120 Harbor seal (P vitulina) 80- 150 California sea lion (Z. cal@~v~iur~us) 90- 120 Dolphins (T cmcafus,T gzllz) 80-120 Manatees (T. inunguis, T. manatus) Muskrat (0. ribeMim) 300 Mink (M vison) 230

Bzrds

Diving ducks (A, fuligula, A, afinis. 300-500 A. ammcana) Dabbling duck (A. platyrkynckos) 150-500 Cormorantishag (.P. georgzanus, 250-30U P. aulilss) Rhinoceros auklet (C. monucerata) 440 45 368 Pengmins (A. folatm:, P. pnp?m, IM-220 20-fin fi2,230,2~2,2%5,%6, Eudpytes chn/solophus, P. adeliae, P. J. Butler. R. M. Spkentcus demwsus) Bevan, and A. J. Woakes, unpublished data

*Several species may be grouped under one common name. +Heart rate was usually observed during natural dives (includes trained dives). $Lowest heart rates were often seen dwing extended dives (not including "trapped dives). $Forced dive heart rate referr to restrained or confined anirnals~ rates should be well above those of mammals, which ap- cardiac responses to free diving in such detail but, in pears to be the case (48, 75). harbor seals diving in a tidal stream, there is no doubt An inspection of Table 4 reveals that for birds and that the lowest heart rates are associated with the imme- ~na~rnnalsthe heart rate observed during natural dives is diate dive period (142). That the lowest heart rate occurs above that seen during forced dives. Dolphins, manatees, early in the dive also appears to be the case in ringed and cormorants are perhaps the most obvious exceptions. seals (Phoca hispida) voluntarily diving under ice on Lluring forced dives, heart rate may be up to an order of the end of an electrocardiogram (ECG) tether (139) and magnitude lower than that duringnatriral diving hnt,,more in captive gray seals submerging in a swim mill (141). usually, the difference is less, being of the order of two However, Weddell seals monitored using the tether tech- to five times. During forced dives, the actual 1erlgU1 or tire uique do not show extended cardiac intervals during the dive is relatively unimportant in governing heart rate, once early dive period (219). Also, naturally diving elephant a full bradycardia is established. The full extent of brady- seals carrying Holter monitors do not usually show this cardia can take a couple of minutes or more to be initial cardiac response (6). achieved in elephant seal pups (170) and manatees (347), In contrast to cardiac anticipation of diving, anticipa- but is usually completed in under a minute in most diving tion of surrzirig, ill which hew&raks increase arid ~rrdy birds and mammals. During natural dives, heart rate at- even reach predive levels before emersion, is a usual fed- tains a stable rate, which may or may not be a bradycardia, Lure of the cardiac response to voluntruy diving UI both in the first few seconds of the dive and even, in some birds and mammals. The smallest change in hcart ratc in diving birds and seals, before the dive commences (anlici- anticipation of surfacing occurs in those animals in which pation of submergence). diving heart rate is high (muskrats, Ref 273; ducks, Ref. Anticipation of submergence, in terms of heart rate, 367). Fedak el al. (142) suggest that the duration of the has only been convincingly recorded from diving ducks period of rapid heart beats before surfacing is related to (GO) as well as harbor and harp seals (66, 205) diving in the length of time it takes the animal to reach the surface. artificial ponds. Typically, the longest cardiac interval In fact, even in ducks swimming long horizontal distances during a dive occurs at or just before submergence. The for food, there is a sustained increase in heart rate for coarse time resolution of most remote heart rabe-moni- ahont the la~t16% of the diving period as the animal swims toring devices (i.e., data loggers) precludes analysis of back to the surface. However, there is notable bradycardia July 2897 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 867 in these extended dives and heart rate only returns to the usually associated with forced dives. In the gray seal, heart resting, but not to the predive, rate (:iti7). rates of 4 beatshin may be sustained for long periods while It has been postulated for harbor and elephant seals the animal sits on the bottom (378). Thompson and Fedak that decompression may be important in causing these (378) claim that this is pm of a seal's usual cardiac reper- anticipatory increases in heart rate (173, 219), although toire during submergence, but they followed their seals by harbor and harp seals surfacing from shallow depths still boat, which may have disturbed them. Extremely low diving show the response (66, 205). Visual orientation appears heart rates have been obtained from swimming harbor seals to be an important component in anticipatory increases when tethered to a boat (299) or ringed seals, tethered to in heart rate in ringed seals, because a blindfolded animal an ECG recording device, submerging on icecovered ponds did not display any anticipatory cardiac increase when (139). In fact, Fedak (141) reports that for gray seals swim- approaching an artificial breathing hole in the ice cover ming in a flume, a forced-dive type of cardiac response of a lake (139). meart rate 40beaWmin) is provoked by any "surprising" If time spent at the surface is truly an "unproductive stimulus, such as the appearance of an inflated weather but necessary state" for marine mammals (142), then any- balloon in the window on the side of the flume. On the thing that prepares the animal and its circulatory physiol- other hand, extreme bradycardia has also been recorded in ogv for emergence might serve to shorten the surface elephant seals diving far from, and long after, intervention intend. Thompson and Fedak (378) have suggested that by the experimenters. Heart rates of 3-5 beatsimin were an increased heart rate before surfacing will circulate recorded on occasion; in the example shown in Figure 23Aii, blood around the musculature and other organs so that this occurred as the seal began to swim toward the surface oxygen-depleted tissues and myoglobi in muscle could (4). Whether or not these seals were disturbed by potential remove residual oxygen from the blood, reducing Po2, predators or boats passing overhead is not known. which would maximize oxygen at the start of breathing For mink (Mustela visson), muskrat, and platypus (Ur- and reduce the time needed at the surface. For animals nithorhynehus anatinus), extremely low heart rates oc- wlrose brains are crucially dependent or1 oxygen, this cur when the mima1 is lirotionless underwater. In musk- seems a dangerous exercise, particularly in view of the rats, this behavior occurs during escape dives induced by fact that anticipatory increases in heart rate revert to div- contact with the experimenters (273, 279, 351). In mink, ing bradycardia if the animal returns to depth without extreme bradycardia occurred after the animal entered an actually surfacing (6, 299). underwater pipe (:jtili). The platypus just seems to "like" Aside from the initial and terminal variations, heart wedging itself under tree roots on the bottom of a pond rate during natural dives seems virtually stable during any and remaining motionless with a heart rate of 4 beatslnlin one dive (Fig. 23). However, for elephant, gray, and harbor (140). Only recently have heart rates been monitored from seals and muskrats during natural dives and ducks during birds foraging in nature. Both cmpcror pcnguins (232) and dives extended by training, bradycardia increases with South Georgian shags (16j display heart rates in the same dive time (Fig. 24). In fact, Kooyman and Campbell (219) range as those seen in forced dives (Table 4). Othenvise, believed it was possible to predict dive time from monitor- for birds, low heart rates have only been observed in ing heart rates early in a dlve by Weddell seals. In contrast, diving ducks trained to make extended dives (367) or Hill et al. (177) fuuncl iL in~pussiblelo PI-edicl long dives when access to the surface was prevented at the end of a from the early dive heart rate. Only short dives (<5 min) natural dive (Fig. 23Eii; Refs. 151,367). This latter cardiac could be predicted from high heart rates (>50 beats/min) response is akin lo Lhal displayed by gray seals in a flume, in the early dive period. The preponderance of short dives when access to the breathing hole is prevented (77, 141). biased the relation between early dive heart rate and e~wu- Bulh seals and ducks continue to swim after being ing dive time; if dives of 4 minor less were excluded, then denied access to the surface, so physical activity has little no signscant relationship was obse~ved.Nevertheless, as influence on the low diving heart rates that occur after in other seals, Weddell seals displayed a strong tendency "trapping" them underwater. Diving heart rate in a num- toward greater bradycardia as dive time was prolonged ber of seals is likewise unaffected by physical exertion. (177). The inverse relation between heart rate and dive Harbor seals swimming in a flume exhibit a bimodal heart time led Thompson and Fedak (378) to suggest that 220 rate pattern (high rate, surface; low ratc, submcrgcd) even heart beats are rationed throughout any dive longer than at high levels of exercise (estimated at 4-5 times resting 7 min in gray seals. They find it difficult to think of a rate of oxygen comu~r~ylion;Re[. 407). Similarly, Kooy- plausible rationale for heart beat rationing and, from their man and Campbell (219) report that a young Weddell seal data, so do gray seals. For instance, in dives of 14- to 18- vigorously swimming whilc diving showed the same cw min duration, the total number of heart beats varied from diac response as adults making "leisurely" shallow dives. 87 to nearly 240. Also, heart rates of naturally diving elephant seals (4) and It is apparent, from Table 4 that the lowest heart rates gray seals (378) appear not to be influenced by swimming that occur during naddives are similar to those more up to speeds of 1-2 rnls. PATRICK J. BUTLER AND DAVID R. JONES Volume 77

D -Hear1 role -- Depih

200 100 80 300 - 60 LOO 500 i C LO

' 20 600

0 L - 120 L g loo I

60 LOO

Time Iminl

Time Iminl E

Time Iminl C

Time I5 s periods1 Time Isl July I997 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 869

On the other hand, high levels of physical activity 297,353,424). Bradycardia would, by itself, be expected to appear to have an effect on diving heart rates in dolphins cause an increase in stroke volrrme, so the decrease in (406), seal lions (407), muskrats (273,352), platypus (140), stroke volume is usually attributed to a fall in myocardial and diving ducks with unimpeded access to the surface inotropic state and, in seals with a caval sphincter, de- (45, 59). Butler (45, 48) has hypothesized that for diving creased venolw rrt~~m[although the evidence for such a ducks at least, there is a conflict between the exercise role for the caval sphincter is not good (136,191,295)l.Only and diving cardiovascular responses. This antagonism is three studies have measured stroke volume in unrestrained flexible and the balance can be tipped toward exercise or divers, and a4 animals were trained. In sea lions, stroke diving-We responses depending on conditions, such as volume is unchanged (134), whereas in tufted ducks, stroke the length of the dive. For instance, during natural dives, volun~edoubled at the start of the dive and then declined, the heart rate of diving ducks is above resting rates (59, being around resting stroke volume by the end of the dive 151), whereas during dives that are extended by training, (17). In the harbor seal, however, stroke volume decreased heart rate falls well below resting rates toward the end (328). Hence, the emerging picture for unrestrained animals of the dive (367). In these situations, exercise intensity is dwing trained dives is similar to that for restrained animals "similar" throughout both extended and natural dives during forced suhmel-sions. (262, 363, 372), so it is the factors related to diving per Unvarying or decreasing stroke volume, allied to the se that modulate the exercise response. 0-Adrenoceptor 111arked cluulges in heart rate that accompany submer- blockade with nadolol prevents the interdive heart rate gence in many diving animals, means that circulatoly ad- from increasing above -300 beatsfmin, although there is justments must be occurring, especially if there is a re- no effect of P-adrenoceptor blockade on heart rate during quirement to maintain arterial blood pressure. Some vari- natural diving (151). In cont,raqt,mnskrats forced to swim ability in blood pressure might be expected due to against water currents in a flume show an increase in changes in "set point" of the barostatic control system diving heart. rale wilh increases in water flow. At flows of induced by activity or asphyxia (355) or by different time 1 m/s, heart rate is almost twice that observed when diving courses of the cardiac output and arterial constrictor rein still watcr (362). Hcncc, in this situation, exercise inten- sponses, but over all, one would expect blood pressure sity is increasing (see Refs. 147, 148) and overrul'mg the to be closely regulated. Unfortunately, there is little intor- diving response. P-Adrenoceptor blockade with nadolol mation about blood pressure in freely diving birds and has no effect on the increase in heart rate with exercise none for freely diving mammals. intensity. This contrasts to the situation when the muskrat Tufted and redhead ducks making trained or escape is exercisil~gill air, Cur ~uucl~of Ule increase UI heal rale dives show little change in arterial blood pressure (19, with increased running speed is prevented by P-adreno- 371). Even irapping ariilals underwater, sometimes in- ceptor blockade (352). ducing a ~narkedbradycardia, has little effect on peak ventricular or mean arterial pressures (371). However, 2. Stroke volume and arterial hlood pressure Bevm and Butler (19) report lhal Ule long cardiac interval During forced dives, stroke volume is either unchanged that occurs upon submergence in tufted ducks is associ- ur decreases ur bull1 bids ad rnarrl~uals(27, 30, 150, 208, ated with a fall in diastolic pressure, indicating that pe-

FIG. 23. A: heart rate and dlve depth of an elephant seal performing natural dives at sea. I: Typical profile for heart rate and dive depth. ii:atypical profile in which heat rate fell to extremely low levels a5 seal swam toward surface (R. D. Andrews, D. R. Jones, J. D. Williams, P. H. Thorson, G. W. Oliver, D. P. Costa, andB. J. Le Boeuf, unpublished data). n: heart rate during Z natural dives by a gray seal on one feedmg trip. Each dot represents a single heatt beat. A downward pointing arrowhead marks Jtart of a dive, and an upward pointing arrowhead represents surfacing. In a tmical foraging.. . dive. . heart rate was ertremelv arrhvthmic. with occasional oain of beats close toeether. Heart rate n\:rrnqt.d io hwrs mln for rnllrr ~~~hn~rr~rnrr:~ncl~~h~n~anrlclpnron mrhyra;d!a hrlolr surfacing &her rnrr 5hoa.s :t ti^\^, fnm SRIIIY tOrae1112rnp +,xh~h~l~~lgnnos1 tktr~lll~, Ihrndyca!tllil I., ~nrdrd11. IIII wn.. For rnllrc tll\r, hcdn ralc avtvaccti u'l heauniln ant1 w35 below 1 bearsnnln for !+I of IIIIIQ IhlodtRrtl fronl Tllolnuson and Fed& 374, l ('. heartrate profiles before, during, and after natural and forced dives of muskrats. Time wasset to zero whekarkal's nose entered water (1 dive) and then reset to zero at end of dive when nose brake surface (t surface). Asterisk indicates values significantly different from heart rate during corresponding period in forced dives. [Flom Signore and Jones (351).] D: heart nite (0) and depth profile (0)duliy a lO~minnatural dive by an emperor penguin (232). E, i:mean heart rate (ZSE) of tufted ducks before, during, and after natural dives (8)and extended dives (A). Downward pointing amowhead (time 0) represents point of immersion, and epwarrl pointing arrowhmds represent mean points of ernemion for natural dives (22.4 s) and extended dives (41.4 s). ii: mean heart rate (?SE) of tufted ducks before, during, and after natural dives (m), natural dives from which ducks were temporarily unable to surface, "trapped" (0)and forced dives (A). [Data from Butler and Woakes (GI).] Downward pointing arrowhead represents point of immersion during natural dives and "trapped" dives (but not forced dives). Upward pointing arrowheads represent points of ememion for all dives. Vertical dashed line represents point at which ducks apparently became aware that they were temporaily unable to surfce during "trapped" dives (after 13.8s mean duration of submemion), and it also represents point of head immemion in rase of forced dives. [Modified from Stephenson et al. (3G7).] PATRICK J. BUTLER AND DAVID R. JONES Volume 71

Brealhing -- bl.A& ,A I IC ' loo

Dive duration (rnin) Dive duralion lrninl

150 D Breathing 160-. 4 -- 1LO - r s -0.55, PcO~OOl 100 -

Diving Oo 2,b ,b 60 io 1; i;o i:o iio 1io zdo 2:0 I I I I I Dive duration Is1 0 - 0123 456 F Dive duration lminl 120

100 - 0 -- .-C 0 E 80- 200 - U) - '00 O 8 e eO* o8o.o 0 3 60- ** . o ** 0 150 - *Y 0 ** -- LO- L 0 100 - Y * 20-

50-""""1 0 I I I I 7.5 10 12.5 15 17.5 20 22.5 25 27.5 30 32.5 0 2 L 6 8 10 Time alter immersion Is1 Dive duration Imin) July 1997 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 871 ri~heralvasoconstriction lags- behind the cardiac re- 2 sponse. Mean arterial blood pressure increased during the - 7W 0.m 5 dive as heart rate rose in anticipation of surfacing, but -2 3 $ 0 ,P fell on emergence despite the marked Increase in heart 0.04 - O - 2% rate accompanying that event. In contrast, rhmoceros auk- .g L mo 3 'Y lets +howed no changes in heart ratp or hlood prewure g f -'D O .E 0.02 2 ?. before, during, or after escape dives (368). xi 2% 2;;E 'ID E - - $ 3. Bloodjlnw distribution .-- 0.00 "l W E. Cardiac output increased by 33% in the transition 2 from rest to diving in tufted ducks, yet blood flow to the '-6 -4 -2 0 2 4 8 8 10 12 14 16 lR 20 22 legs- increased five times. indicating- massive vasodilation Time (s) of this vascular bed (17). The amount of blood flow to FIG. 25. Estimated mean cardiac output (GO, 0) and total penph- the legs during diving was similar to that observed in eral resistance (TPR;0) during trained dives in 3 tufted ducks. Estimates ducks swimming at the surface at 0.75 mls. Interestingly, were made fmm heart rate, blood flow, and mean arterial pressnre blood flow to the leg measured directly by Bevan and measurements of Bevan and Butler (17, 19). Horizontal dotted and dashed lines represent levels of TPR and CO, respectively, in interdive Butler (17) in surface-swimming ducks was a similar pro- intervals. [R-om Bevm and Butler (18). Copyright is lleld by Company portion of cardiac output as that assessed indirectly using of Biologists Ltdl radiological imaging techniqnes by Stephenson and Jones (371). Flow in the brachiocephalic arteries increased by 20% while carotid flow doubled during diving, compared vascular beds were effectively unchanged by diving, with rest. In contrast, in animals swimming at the surface, thereby making no contribution to the increase in TPR. brachiocephalic and carotid flows were significantly less On the other hand, brachiocephalic flow resistance in- than or the same as those at rest. Obviously, there was a creased by 20°h due to an estimated two to three times substantial redistribution of blood flow between resting, increase in resistance of the vascular beds supplied by swimming, and diving (17), with the "rest of the body" the brachial astery, after 12s submergence. Toward the (i.c., that not supplicd by thc ischiadic and brachioccpha end of the dive, blood flow was being directed away from lic arteries) receiving substantially less blood during div- the largely inactive breast muscles. Cardiac output and ing than duriy su~fxeswirrlming or 1h;ui in resting ani- slroke volu~newere strongly correlated to heart rate (17). mals. This selective distribution of blood during natural That is not to say, however, that heart rate is necessarily diving must be viewcd in the context that tufted ducks a good indicator of regional blood flow changes in birds. appear to dive well within their ADL and may indicate Using radiological imaging techniques, Stephenson and a mechanism by which regional hypothermia could be Jones (371) showed that flow distribution in ducka swim- achieved in other species of birds, such as shags and ming at the surface was qualitatively similar to flow distri- penguins (see sect. IIIC). bution in ducks making escape dives [in agreement with Bevan and Butler (19) used their blood pressure data the quantitative data of Bevan and Butler (17) for trained along with the above flow data (17) to assess changes dives]. However, whcn ducks wcrc trappcd undcrwatcr, in TPR during trained dives (Fig. 25). Total peripheral only one in four showed a blood flow pattern characteris resistance increased by 35% after 5-s submergence, and tic of the forced dived animal, i.e., preferential distribution this increase was maintained at least through 12 s of div- of blood flow to the heart and brain in association with ing. Peripheral resistances of the ischiadic and carotid pronounced bradycardia, even though heart rate was

nc. 24. Relationship between heart rate and dive duration in birds and mammals. A: elephant seals; 486 dives from 7 seals, r = U.S3 and probability that slope = U is

A B

loOc fi LOO ,- - C -'Z 300 a u c nc. 26. Relationship of half time (tin) for inulin (A) : and indocyanine green (ICG; B) clearance at rest and dur- :200 -" ing natural dives. M, body mass; ADL, aerobic dive limit. .-c Seals 1.2 Number of dives for each column is shown in parentheses. -= [From Davis et d. (109).] C - 100 3*

Dive duration iminl Dive duration imin)

equally as low in a~iothertrapped bird. In fact, no correla- merse its head on command, Stone et al. (375) found tion could be established between heart rate and relative only a 33% decrease in renal artery blood flow, yet in a blood flow to the hindlimbs, heart, and brain in ducks trained harbor seal, diving caused an immediate and that were "trapped" underwater. complete cessation of urine production (298). Similarly, Ducks use their leg muscles for propulsion during using the clearance rate of inulin, either after a single diving, whereas penguins use their wings, so it is perhaps injection or after equilibration of inulin distribution, not surprising that leg blood flow in Adklie and gentoo Davis et al. (109) found that, in Weddell seals, glomeru- penguins declined during diving (285). In penguins sub- lar filtration rate was unchanged from resting levels in merging voluntarily, heart rate fell to one-half or one-third most natural dives, but fell dramatically if the dive ex- the predive value (-200 beatsimin), which was in the ceeded the ADL (Fig. 26A). Hence, a massive increase same range as changes in leg and carotid blood flow. In the in resistance of the renal vascular bed is not an initial absence of data on arterial hlood pressure, it is difficult to part of the circulatory arsenal of cardiovascnlar adjust- assess the nature of these flow changes, but it is possible ments in diving Weddell seals. Support for the notion that cardiac output, blood pressure, and flows may all that there is little renal function during extended dives decline together. comes from clearance studies using p-[3H]aminohippur- Hence, the picture that emerges for bids is one of ate and innlin (168). During a 24-min dive, there was no subtle rather than massive redistribution of blood flow evidence for kidney clearance of these tracers. In fact, during diving, compared with the predive period. In fact, their eqnilibration within the vascular compartment on a somewhat more ext,ended time course, a similar view was considerably delayed, indicating marked vasocon- could be taken of circulatory adjustments to natural div- striction throughout the dive. ing in mammals. Heart rates, and perhaps cardiac outputs, Hepat,ic functinn tests using indocyanine green (ICG) are frequently at one-half to one-third of resting, if not (log), cholic acid, or galactose clearance (168) again gave predive, levels. Hence, if maintenance of arterial hlood contradictory results. Davis et. al. (109) found that, even pressure is a priority then, as in birds, large changes in during extended dives, hepatic flow was maintained (Fig. '1'PK are unnecessary. The 10 times changes in TPR seen 26B), whereas Guppy et al. (1168) found no hepatic clear- in forced dives might only be expected in those animals ance of metabolites even during short natural dives. How- sl~uwing~uu~sually low heart rates, e.g., gray seals sitting ever, during one short dive, Davis et al. (109) found that on the bottom or elephant seals swimming with heart ICG clearance dropped by 25 times. Davis et al. (109) rates of 3-5 beatsirnin (6, 143). suggest that fright or other forms of stress may initiate a Unfortunately, not much information on blood flow profound dive response, similar to that in forced dives, in voluntarily diving mammals is available, whereas which would cause a marked reduction or even elimina- much of what is available is contradictory. In resting tion of glomerular filtration rate and hepatic blood flow. humans, renal and hepatic flows can account for 50% This seems a reasonable explanation for the discrepancies of cardiac output, and if this is also the case in divers, between the two data sets. then any restriction in thcsc regional flows could par- A rangc of indicators (Evans bluc) and radiolaheled tially compensate for changes in cardiac output conse- metabolites (glucose, palmitate, and lactatc) havc also been quent upon submergence. In a sea lion trained to im- used to assess metabolic function in unrestrained dives by PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 873

28min "exploratory" dive (Fig. 270 than during a 17-min iib,"feeding"during the dive feeding (Fig. dive27B), was, despite on average, the fact one-half that heart that rate in the exploratory dive (168). However, there was a "break" in the bradycardia during the shorter dive after 5 min, which lasted for a minute, and may have considerably 0.05 speeded up the mixing process. In gray seals, diving with a heart rate of <10 beatshin, a complex specific activity 60 120 180 decay curve was observed that diered markedly from the - Time lmin) exponential decay seen at rest or during surface swimming a VI (77). In reality, these curves represent entirely different " things. The exponentially declining curves at rest or during -3 01 exercise describe metaholism of the lahel, whereas during - a diving, the curve solely describes wash-in kinetics into a E circulatory compartment in which flow must be greatly a 0.02 slowed (175). I Muscle circulation during natural diving has been in- e- ' 5 10 15 10 20 30 directly estimated from measurement of temperature :-I!&Dive Recovery (325) and partial pressures of nitrogen in the muscles Time Iminl (340). Both of these methods suggest that muscle circulation remains open during natural diving. In Weddell seals making a 40-min dive, muscle temperature should increase by 0.3-2.2"C, depending on the metabolic rate, if there is no blood flow (326). Because this did not occur in the single Weddell seal in which temperature was recorded, then one of two postulates must hold: 1) exercis- 0.02

0 5 10 15 20 25 10 20 30 ingculatioilwise, seal muscle muscleremained circulation is extremely patent in evendolphins hypometabolic during making long 1.6-minor dives. 2) the divesLike- cir- l!!kDive Recovery also remained patent as judged from nitrogen accumula- Time (minl tion. However, it is interesting- to note that ~erfusionof the blubber layer was not maintained as effectively as that nr..~~ 27. - Clearance~~~~ ~ of .I"Cle1ucose ." from olasma of Weddell seals at real (.A 1, during a relanvrly ilt#,n(1: !am) 'feeding" dvr (B,. tu~dduring of muscle, since the partial pressure of nitrogen was much A lor14rl- (56 1111111 'CX~~O~IOI~~II~YP(i) Seah were dnmg from a hole lower than that in muscle (340). cur in ICC Vvntral lines m B mi C m&c.a[c wlnr~ta~~i~laial *~~Tltl(.t.cI N~lr Muscle circulation is stopped during forced dives due Uference between wein A and curves in B and C. The fact that radiolabeled glucose took some time to reach peak activity during dives to extreme vasoconstriction. Some lactatc appears in the is taken to indicate a slow circulation time resulting from peripheral blood during the dive, but there is a large blood lactate vasoconstriction. DPM,&Integrations per minute (1 DPM = 0.0167Bil). peak during recovery as circulation is restored to the pre- [From Butler (41) after modification from Guppy et al. (1681.1 viously hypoperfused tissues (Fig. 1; Ref. 346). A similar pronounced lactate peak during recovery occurred in the Weddell and gray seals (68, 77, 168). Unfortunately, the blood of Weddell seals after extended &ves (Fig. 14; Ref. resulting decay curves are difficult to analyze and interpret. 235), hplying that muscle blood flow was shut down for For instance, the label may be diluted by endogenous prod- a considerable portion of the dive (217). It does not follow, ucts, e.g., lactate during long dives, and it is also possible however, that muscle blood flow continues during natural that lactate label could come from another source, such dives withiin the ADL, for the muscle oxygen and high- a9 glucose or palmitate, when these are injected along with energy phosphate stores can serve to power the energy labeled lactate. Furthermore, the label may be metabolized, requirements (see sect. IID). e.g., free fatty acids or glucose, depending on the state of One way to use muscle oxygen and fuel stores more the circulation. Hence, it is not surprising that these studies effectively is to perfuse muscles intermittently so that contributed little to our undemtanding of the metabolic oxygen can be stripped from the myoglobin during the biochemistry of seals and instead have been used solely as no-flow condition. Kooyman (217, 218) has proposed an indicators of circulatory restriction during diving. Tn Wed- intermittent flow model based on ideas obtained from dell seals, wash-in and equilibration ties are far slower studies of forced dive animals. During the initial stages during diving than at rest (Fig. 27). Furthermore, wash-in of the dive, muscle flow ceases, but after the myoglobin times for glucose, for instance, were 66% slower during a stores have been depleted, accumulation of anaerobic end 874 PATRICK J. BUTLER AND DAVID R. JONES Volume 77 products causes local vasodilation, and the muscle store is recharged. Vasoconstriction would then occur, thus cut- ting off the muscle circulation again. These cycles could repeat over and over. Certainly, the fact that heart rate may cycle up and down during long dives has led others to suggest an intermittent perfusion model for muscle circulation (168,378). If, during extended dives, the worst came to the wont, then arterial vasoconstriction, heynnd the reach of vasodilator metabolites, would eliminate niuscle blood flow, saving blood oxygen for Ire crucially oxygen-dependent heart and brain. This restriction in flow could be effected neurally or via thc agcncy of circulating catecholamines. Hence, in this model, the fall in cardiac output on submersion due to bradycardia could be compensated by an immediate restriction in muscle blood flow (232). Guyton et al. (169) claim that their recording of myo- Dive duration (minl globin/hemoglobin saturation and relative blood volume in the muscle of diving Weddell seals gives no support to B the pulsatile flow hypothesis (Fig. 21). Relative muscle 15 r blood volume declined immediately on diving and during short dives remained low throughout. During long dives (>I7 min), muscle blood volume returned to normal levels on ascent and indicated hyperemia 1-2 min after surfacing. On occasion, muscle resaturation occurred before surfacing at the end of dives, but this must have been due to restoration of a normal muscle flow pattern when the heart rate increased in anticipation of surfacing. Bursts of resaturation throughout dives were not observed. 2.5 To summarize, the picture that emerges of blood flow 0 200 LOO 600 reditribution in freely diving mammals is either profound Dive time (see1 circulatory restriction throughout the dive (77, 168) or moderate initial changes in flow distribution that become FIG. 28. A: Rrst postdive exhalation end-tidal Pu2 (ETPo2) as a more pronounced as the ADL is approached or exceeded function of dive time in a Weddell seal. [From Ponganis et al. (324j.l B: alveolar Po, (Pk,,, *) and Pca, u), sampled from slngle postdive (109, 218). It is possible for the animal to move rapidly breath, in relation to drve time m 2 Amazoman manatees. Kesults from from the latter to the former type of response depending both manatees did not differ significantly and have therefore been corn- on prevailing conditions. A third type of circulatory acljust- bmed. IModlfied from Galllvan et al. (1621.1 ment, or more correctly nonacljustment, involving little or no change in flow during diving is also possible. At pres- blood glucose levels decreased in diving Weddell seals, ent, however, there is no direct evidence that this lark of presumably due to increased glycolytic metabolism. In response occurs in natural diving, and little support can recent years, Hochachka's view has been criticized on be inferred from published data. both theoretical (218) and empirical grounds (71). Seal red blood cells (RUCs) have a much lower glucose concentration than the surrounding plasma, so hematocrit B. Efficacy of Cardiorespiratory changes during a dive can severely affect whole blood Responses to Diving glucose levels. In fact, although all changes in whole blood glucose concentrations are not due to hematocrit varia- Oxygen is crucial for diving birds and mammals, so tions alone, there is no doubt that plasma levels of glucose the levels of oxygen maintained in arterial blood during are unaffected by breath-holding (71). breath-holding are a measure of the efficacy of cardiore- In tufted ducks making 18s trained dives, there was spiratory responses to diving. That depletion of'oxygen little effect of diving on arterial PO* (Pao2),arterial Pco, stores limits diving performance and metabolism was (P*,,), or arterial pH @Ha. The Pa+ values fell by 2.4 challenged by Hochachka (182), who suggested that kPa, Pko, effectively did not change, whereas pH. fell depletion of glucose stores was the crucial factor. Ho- from 7.55 to 7.48 due to an increase in blood lactate from chachka's (182) view was based on the fact that whole 1.7 mM (at rest) to 3.0 mM (49). Only minor changes in July 1997 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 875

Carbon dioxide

FIG. 29. Instantaneous measures of carbon dioxide, oxygcn, and air flow during a single sluface period after Oxygen a natural dive by a gray seal. Scale bar (on left) represents I t E~~~~~~~~~ a 1Osichange in oxygen and carbon dioxide traces and 0 to +40 lis (expiration for flaw trace). Muimun~tidal voluue, zy$$rFlow (Inspiration peak levels for carbon dioxide, and lowest oxygen levels Seal surfaces Seal diver t 1 are not attained until 3rd to 5th breath postdivc. [Rcdmwn I I I. from Reed et al. (334).] 0 45 90 Time Is1 flow distribution occur in diving ducks, with the blood submersion and that alveolar gas equilibrates with that in gascs indicating that more profound adjustments are un- the blood only after surfacing. necessary. Similarly, in rhinoceros auklets, in which no Partial pressures of blood gases have been recorded cardiac acljustments occur upon diving, Pao, only fell to from freely diving Weddell seals (331,423), and the picture 7 kPa, whereas Pkozwas unchanged. Interestingly, hema- that emerged was more complex than the inverse propor- tocrits of blood samples taken during diving (25%) were considerably below predive levels (3796) (368). For diving mammals, data for Weddell seals (227, A 324), manatees (161, 162), and the bottlenosed porpoise t:: (341) support the reasonable notion that the longer the dive, the lower will be the Pao, at the end. The Pao, values .direr < l7min were assessed indirectly from end-tidal Poz of the ht 0 dire,. 17min breath, or in the case of manatees the only breath, termi- ;; 15 nating thc divc (Fig. 28). All these data point to the extreme tolerance to hypoxia displayed by diving mammals with alveolar oxygen levels of 2% not being unusual. In fact, arterial levels may be even lower than this. Hence, it is possible that in Weddell seals returning to the blow hole after a long dive, Pao, could be as low as 1.33 kPa. These field data showing extreme insensitivity to hypoxia Dive , complement observations made many years earlier on Weddell seals subjected to simulated dives or nitrogen 0 10 20 0 10 20 30 LO 50 hreathing (138). Recently, Reed et al. (334) have shown that the lowest end-tidal oxygen is not reached for several breaths after surfacing in gray seals (Fig. 29) and that end-tidal Po2 of the first breath upon surfacing (12.0 % 0.23 Wa) is similar to that of the last breath before submerging (12.9 i 0.21 kPa). Assuming that end-tidal Po, in this study is represen- tative of alveolar Po2, these data indicate that there is little re~novalof 0, from the lungs during submersion. This either means that there are effective pulmonary shunts, thus maintaining alveolar Po,, and that 0, has been preferentially removed from the blood and, possibly, ~ive I Recovery 0, 1 from the muscles, or that effective tissue shunts occur, 0 10 20 0 10 20 30 40 50 whereby tissues are not perfused and rely entirely on the Time lminl oxygen- .. and PCr stored within them, thus maintaining .. I'a,,, as well as alvcolar 1'0.. It must bc assumcd, howc\.cr, nc,. 30 t.1xygv11panl:d 1.rt.wln (1~3. j (:IJ ;and vgmr,.nt (I a,. ,,I ;anen;.l lsl~~~~~l(13) \\ v~l~lell +v:~I. tlonng witwal dntng and af(vr rtwtr- ha^ ~hct~ssucs rt~ut have s~nall 02 stores. such LLS the ra<.,,, F;,,.I\. .I,,,,,~ l.h5,.,,1 I YI,6.n,bli. ,. isil. central nervous system and heart, are uerfused and that reco;ded w& 2.4 CP;P~at'end of 27%" dive. ~izlarlow Pa, valugt they sufficient oxygen red;ce p*,. ~h~~,al- were recorded at end of short dives (i.e., dives that ended ci7 min). During long dives, Cab ~mainedabove resting values (18-19 "01%) for the mechanisms are not clear, it seem 16-17 mi". Rate of decrease thenafter was not significantly different likely atat there Xe effective pulmonary shunts during from that measured during short dives. [From Qvist et al. (331).] 876 PATRICK J. BUTLER AND DAVID R. JONES Volumo 77 tionality of P%, and dive time assessed from measuring solely end-tidal Po2. Values of arterial oxygen content Manne mammais ws juvenile 0 Tenssfrial mammals /. (C%,) and Pa,,, rccordcd close to the end of short dives NESadulf; late (< 17 min) were higher but not significantly different from NESadult: early t,hose recorded at the end of extended dives (>I7 min) ws a,,,, \ (Fig. 30, A and B). Hence, Pao, fell slower during long '. '. than during short dives, indicating greater circulatory restriction during long dives right from the start (Fig. 30A), confirming the conclusions of Guppy et al. (168). Com- pression hyperoxia was apparent at the beginning of a number of relatively short dives, but in spite of this, C*, fell throughout the dive (331). In contrast, during long dives, Cao, wrndined normal or elevated for the flrst Blood volume (%Mb) 15 min or so despite a fall in P%, from 10.5 to 4.25 kPa. nc. 31. Relationship between sple~cmass and blood volume, both C*, was elevated due to an influx of oxygenated RBCs, as a percentage of body mass (%Mh),for several species of marine and with blood hemoglobin reaching 25 g1100 ml blood and terrestrial mammals. Marine mammal mules were calculated from mea- hematocrit approaching 6090. Hematocrit also increased sured M, values and hematoclit (Hct) ranges, known plasma volumes, and calculated maximum blood volumes @cued on maximum Hct) nnd during short dives, but this was insufficient to prevent a include values for northern elephant seal (NES) adults during early and fall in Ca02. At the end of a dive, RBC conce~llrationre- late fasting. Values for adult Weddell seals (WS) were obtained from turned to resting levels within 10-12 min (423). Similar calculat~onsof &st et al. (331). Values for terrestrial mammak were obtained from published studies as indicated: a, Ref. 65; b, Ref. 172; e, RBC dynamics were also observed in elephant seals pups Ref. 286 d, Ref. 291; e, Ref. 322; f,Ref. 382; and g, Ref. 395. [From during periods of apnea and breathing (72). Castellini and Castellini (67).] Hematocrit variations during diving were first attributed to venous pooling of RBCs as the circulation slowed during submergence (235). However, Qvist et al. (331) will contribute to hematocrit change during diving, but showed that not only venous but also arterial hematocrits the frequency and efficacy of such contractions is specula- increased during diving and suggested that the spleen was tive. On the other hand, the Weddell seal spleen has the the source of the RBCs. Bryden and Lim's (43) postulate requisite equipment with many smooth muscle cells in the that the spleen might serve as a reservoir for RBCs in capsule and trabeculae, which are extensively innervated seals was considerably embellished as a result of the re- (349). Spleen size is reduced immediately after injection search of Qvist et al. (331). The spleen was envisaged of epinephrine (194). as a "scuba tank," injecting oxygenated RBCs into the A caveat to the idea of the spleen functioning as a circulation to promote aerobic metabolism while, at the scuba tank in large marine mammals is provided by the same time, assuaging the impact of C02retention on blood diving behavior of many of these animals. Most marine chemistry and diluting N, tensions (183, 422, 423). mammals do not spend long periods at the surface be- Because blood viscosity is proportional to hemato- tween dives. Dives occur in bouts in Weddell seals, with crit, the former increasing exponentially with hematocrit surface intervals of <5 min between each dive of a bout then, in the absence of change in arterial blood pressure, (74), whereas in both northern and southern elephant an increase in hematocrit of some 5@??(from 40 to 60%) seals, surface periods of longer than 3 min are unusual, will almost halve blood flow, purely passively (174, 282, even when the animal is diving continuously for 24 h or 401). This passive flow restriction may be ameliorated to more (178, 246). Hence, hematocrit will rise on the first some extent in the smallest blood vessels, due to the dive of a bout and, in the case of elephant seals, could anomalous viscous properties of blood. Passive flow remain elevated for months. If this is the case, then it is changes may also be offset due to increases in blood pres- not unexpected that the viscosity of northern elephant sure, consequent to increases in blood volume, unless the seal blood is lower than that of Weddell seals at all flow venous storage vessels can accommodate all the in- rates (282). creased circulating volume. So, the question iuust be asked as to what is the Measured and estimated splenic masses of pinnipeds effect on oxygen stores and ADLs of splenic contraction support the idea that the splenic reservoir is large enough in the first dive of a bout compared with maintained high to function as a scuba tank (67,71,331,326). In fact, it has hematocrit in all the ensuing dives. Ponganis et al. (326) been argued that, because splenic mass as a proportion of examined this question in Weddell seals using the splenic body mass is highly correlated (# = 0.9) to mass specific size and blood volume estimates of Qvist et al. (331) and blood volunre then, as diving manimals hhvc such largc thc ADL assumptions i~ndcalculations of Kooyrnan et al. blood volumes, it follows that they have large spleens (220, 227, 235). The RBCs in the spleen were assumed to (Fig. 31; Ref. 67). Hence, splenic contraction or relaxation be fully oxygenated. During the first dive of a bout, splenic July 1997 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 877 contraction provides the major proportion of the blood seal blood is well buffered, the large acid-base disturbance oxygen (67%) during the dive. However, during the next during forced diving will mean that equivalent parlial pres- dive, RBCs continue to circulate if the spleen remains sures of oxygen in arterial blood during forced and volun- contractcd, so thc total blood oxygen store is only re- tary dives will represent very different oxygcn contcnts duced by 10% (1.9 l), which is enough oxygen for 1.2 min due to the prominent Bohr effect of the blood (332, 358, of diving time (326). The slight reduction in the blood 394). Addition of oxygenated RBCs to the blood notwith- oxygen stores comes about because RBCs from the spleen standing, there seems little doubt that forced diving in are fully oxygenated, whereas when mixed in the circula- Weddell seals does not appear to result in the parsimony tion, 67% of the RBCs are on the venous side (326). Similar of oxygen use from the blood that might be expected. calculations for harbor seals yield an even smaller diifer- The properties of the blood in loadimg and unloading ence in oxygen stores between the first and subsequent oxygen, at the lungs and tissues, respectively, are im- dives in a bout (326). portant to underwater survival. Blood with a high afilnity A direct measure of the efficacy of oxygen conserva- for oxygen would favor full utilization of oxygen in the tion would appear to be provided by a comparison of lung, whereas a low affinity would be advantageous in blood gas data obtained from Weddell seals during forced allowing more effective oxygen unloading at the tissues dives (257,424), natural dives (331), and spontaneous ap- (241). A complicating factor here is that there may be neas on land or in water (235). Unfortunately, a compar- an inverse relation between P, and body mass, either son of blood gas data is not straightforward, because it intraspecifically or interspecifically, and many diving is not known how hematocrit changes during forced dives mammals and birds are relatively large (14, 265). How- and apnea. If hematocrit increases during apnea, as oc- ever, Dhindsa et al. (118), summarizing data for 12 ceta- curs in elephant seal pups (72), then a comparison be- ceans and 7 pinnipeds, found no relation between Pboand tween Pao, during voluntruy dives and apnea is justiied. body mass. In fact, Snyder (358) argues that oxygen affin- After 4-8 min of apnea, Pag declined to -3 kPa (235), ity in mammalian divers relates to whether they use the with these levels not being reached in free dives for over lung as an oxygen store or not. Hence, affinity is high in 20 min or so (331). Hence, given the caveats mentioned small cetaceans, manatees, and rodents, in which the above, oxygen would appear to be conserved less well lungs represent an oxygen store (85, 251, 399), and low during apnea than during voluntary diving. in large cetaceans and pinnipeds (242). The extreme circulatory restriction during forced Almost all vertebrates have ~nultiplehemoglobins dives (424) means that oxygen uptake will be greatly re- that have different properties with respect to oxygen equi- duced compared with that during voluntaxy dives. Qvist libria, so it is possible that the predominant hemoglobin et al. (332) estimate that -6 ml 02.1-I. min-I was pro- type and oxygen affinity could change with exposure to vided from the blood in the forced dived animals stud- the diving habit. Most seals, however, only appear to have ied by Liggins et al. (257) which, for a 450-kg animal, two hemoglobins (251, 303, 393). In Weddell seals, "fast" gives an aerobic metabolic rate of somewhat <1 and "slow" hemoglobins were designated according to ml .kg-'. min-'. This valuc for oxygcn uptake during their relative mobilities (394). The oxygen affinity of the forced dives is one-third to one-fifth that estimatcd for fast hemoglobin is higher than that of the slow hemoglo- natural diving (75). Nevertheless, during forced dives by bin, whereas the Bohr shift is greater for the slow hemo- pregnant and nonpregnant Weddell seals, Pao, appeared globin (394). In contrast to the work of Wells and Brennan to fall more rapidly than that during voluntary dives. In (394), Qvist et al. (332) report four hemoglobins in both pregnant seals, Pa,,, fell to 4 kPa after a 4-min submer- fetal and adult Weddell seal blood, with the two major gence and to 3 kPa after a 20-min submergence (257). In components exhibiting practically identical P5~values and nonpregnant animals, Pa,,, fell more slowly, reaching 4.25 Bohr shifts. The P,, of fetal blood, however, is lower than kPa after 8 12 min of diving (424). Heart rate fell in the that for adult blood (252, 332), which results from differ- range of 8-15 beats/min during the forced dives, being ent 2,3-diphosphoglycerate (2,3-DPG) concentrations in 12-2596 of the predive rate. In contrast, heart rate during the fetal and adult erythrocytes. Concentrations of 2,s voluntary dives only halved, yet Pao, was 4.25 kPa after DPG are almost three times higher in adult than fetal 15 min of submergence (331). RBCs, and the rke in concentration, and subsequent fall Acid-base disturbances are much more striking than in oxygen affinity, is complete 4-5 wk after bii (332). Pao, differences between natural diving and spontaneous Lapennas (240) has argued that, in t,he steady state, apnea on the one hand and forced diving on the other. Bohr factors of about one-half the RQ will maximize oxy- During natural diving and apnea, Pk,, rose by <20%, gen delivery to t,he tissues, rather than buffering pH whereas pH, was effectively unchanged or fell by 0.5 units, changes caused by the addition of C02 to the blood in at most (236, 331). In contrast, during forced dives, the tissues. On Lhe oLlrer hand, much larger Bohr shifts, Pko, increased by 26-60% and pH. declined by at least approaching unity, would be optimal to reduce Pcoz and 0.5-1 unit (257,424). Hence, despite the fact that Weddell acidosis (39). Large Bohr shifts may also indicate that the 878 PATRICK J. BUTLER AND DAVlD R. JONES lung is not used as an oxygen store during diving because C02camot be eliminated during a breath-hold dive. Large Bohr shifts may also reflect a dependence on anaerobic 300 -, metabolism, in that lactic acid added to the blood during dives will contribute to maintenance of a high Pao, (203, - 254,288). The minke or lesser roqual whale (Rolam~ptwu~ - 90 acutorostrata) sccms to bc an extreme example, in that 3- 200 o 0 80 lactate plays an important role as a specific allosteric li- gand facilitating oxygen unloading (40). Finally, if the =5 J 0 Bohr factor is not constant but declines with oxygen satu- 100 - o ration,favored. then This oxygen seems toextraction be the case from in boththe lungadult mightand fetal be l&/-/ sb O0 blood of Weddell seals (332) but not in gray seals (242) - 40 or ducks (283). However, in harbor seals, the Bohr effect I I I - 1:. 1 2 3 4 5 is largest around 60% saturation and declines at both low [Mbl and high saturations (402). (g1100g wet weight) The buffering capacities of the blood and tissues prevent large reductions in aH as a result of the accumulation nc. 32. Relationship between bufferingcapacity (D) expressed as of respi.a; lllrli;l,,,li,. :,,.i,lS I1o,,,C\.Cr, high buff. 111lltll I~JLSS..N~~OHI w'llrQd 10 rhi+* 1,11 I s 18f11111\(It. hv I pll !U!,I ,~BI~~IIO!~Swete III~~C I,t,tnt,t.18 ~I~H~~L.~#~A el) pll d pH -,, tsrirlg capacitirs would rend ro reduce PI[flllct~atiol~ aid md 1118li~.l+ I~I?~~IUIIIIICO~I(~~IIL~.~IIOR IJI~ tlte fullou.lngdt\uteI,~Y~S thus oppose the Bohr efIecl(358). Increased protein con- and mammals $otter polpaise (~tenellaatto~uata),norGen1 rir seal, centrations and high hemoglobin levels in penguin blood harbor seal, Weddell seal, sea ocEr adult and pup (Enhydra lutns), elephant seal pup, Ad6lie penguin, California sea lion, and a California increase its buffering characteristics compared with the gray whale calf (Eschrlchcius mbustw). Ratio of 0 to [Mb] is also blood of most birds (302). The buffer value (AHCOJApH) plotted for these speaes to illustrate that correlation between B and of true plasma for the Addie penguin is 33 meqL.pHL, [Mb] does not imply a strongly causal relationship between these 2 variables. [Derived from Castellini and Sornero (7!4).] which is some 50?h higher than that of the muscovy duck (254, 344). In gray and bladdernose seals, the high hemoglobin concentration is solely responsible for elevated the tissues will be favored by a more markedly sigmoidal blood buffer values (55,242), whereas in harbor, northern curve (Itill's number is high). Birds seem to have some- elephant, ribbon, and especially Weddell seals, other Eac- what higher cooperativity than pinnipeds (242,243, tors such as plasma proteins contribute to the buffer ca- 402), which could be viewed as a disadvantage in diving pacity (251-253). Furthermore, the Bohr and Haldane ef- birds which access the lung oxygen store during submer- fects are liked so that the Haldane effect (H' binding by gence (62). However, in virtually all bird species, coopera- hemoglobin on deoxygenatiorr) will alsu buffer pH tivity declines at low oxygen saluraliuns, thereby facilitat- changes. Hence, the adaptive feature of a high Bohr effect ing oxygen extraction from the lung (180, 243, 344, 392). may be to counter the high buffering capacity of the blood Many divers experience regional hypothernlia (see and ensure an appropriate swing in blood oxygen afhity sect. 111, C and D). Temperature has a dual action on the between the lungs and tissues (358). oxygen affinity of blood, a direct effect, which is a func- Similar to the role of hemoglobin in the blood, myo- tion of the intrinsic heat of oxygenation (AH) and an globin contributes to the buffering capacity of tissues, indirect effect caused by temperature-induced changes in especially muscle (79). For the diving birds and mammals blood pH (416). A fall in blood temperature increases invrstigat,ed by Castellini and Somero (1981), the tissne oxygen affinity directly, as well as indirectly, by making buffering capacity (P) is related to myoglobin concentra- the blood more alkaline. These complementmy effects tion ([Mb], g1100 g tissue) as follows on oxygen affinity will tend to compromise both oxygen unloading at the tissues and cellnlar oxygenation. How- ever, oxygen binding in many diving animals has a reduced exothermic oxygenation enthalpy (smaller AH), The low 7' indicates that myoglobin, however, is not a which will offset effects of temperature per se on oxygen major contributor to muscle p in diving animals. Castellini aSEnity. In the harbor seal, for instance, Willford et al. and Somero (79) reached the same conclusion using both (402) report a AH (kcaVmol) almost one-half that of hu- empirical and theoretical arguments (Fig. 32). mans. This low thermal sensitivity has been attributed The shape of the oxygen equilibriom cuwe (Hill's to endothermic, oxygen-linked reactions such as 2,s-DPG number) is also important because oxygen extraction release (391). In fact, 2,3-DPG in harbor seal blood is twice fro111UI~ lungs will be favored if lhe curve is steep (Hill's the concentration necessary to bind hemoglobin fully number is low) at low saturations, whereas unloading at (350) and probably plays an important role in decreasing PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 879

phates on the blood of two divers, blacked necked grebe (Podiceps niyricollis) and com~o~w~l(PI&-rocorax car- bosimis), is markedly less than it is in the blood of the chicken and turkey (164,384). However, unlike the situation with respect to 2,5DPG in mammals, inositol pentaphos phate (IPS)is synthesized at the time of cell differentiation and remains constant throughout the life span of the cell (199,264). Hence, modulatory effects of changing concentm- tions of organic phosphates that occur in mammalian blood are not paralleled in bid blood. Giardina et al. (164) hypoth- esize that other adaptive mechanisms may be irnportamt in diving birds, such as competition between IP, and CO,. In view of the situation in the blood of the lesser roqual whale, this might prove to be a fruitful area of investigation. The properties of hemoglobin and its interactions with allosteric modifiers, allied to subtle regional variations in blood flow that can occur during diving, suggest that it would be futile to describe the metabolic character of a dive in terms of variations in a single physiological variable, such as heart rate. Nevertheless, averaged over whole dive cycles (submergence plus recovery at the surface), diving animals may be in a steady state (N),and heart rate, independent in absence (open symbols) and in prcscncc (closcd symbols) of 20% of changes in tissue oxygen extraction and changes in stroke C02. [From Di Prisco et al. (119).] volume can, with certain provisos, provide an excellent and accurate measure of overall aerobic metabolism (23,38). In the sensitivity of the oxygen dissociation curve to temper- this respect, Fedak (141) demonstrated a linear relationship ature. In this respect, the tripling of 2,3-DPG concentra- between heart rate and oxygen consun~ptiox~of swinui~lg tion in the blood of adult compared with that of fetal gray seals when heart rate aud oxygen uptake were averaged Weddell seals also suggests that 2,3-DPG might play a role over complete dive cycles. IIaving now reviewed the com- in reducing AH as well as decreasing oxygen affinity in plexity of the interaction between the relevant variables, adult blood. Physiologically this makes a lot of sense, then advocating the use of heart rate as a global measure because the fetus is protected from large temperature of diving performance may seem too simplistic, hut. never- variations. However, 2,3-DPG and hemoglobin only occur theless, it is such a convenient simplicity that it is sure to in equimolar concentrations in adult RBCs (332). There be pursued, at least in terms of the relation between aerobic is not an excess of 2,3-DPG as is found in the harbor seal. metabolism and heart rate. In the lesser roqual (minke) whale, the functional properties of hemoglobin are regulated by organic phosphate, GO,, and temperature in a highly sophisticated manner (1 19, V. RECOVERY FROM DMNG: 165). There is a strong effect of temperature on oxygen CARDIORESPIRATORY RESPONSES binding in the absence of C02 and organic phosphate. The TO SURFACING value of AH drops by two-thirds in the presence of the latter and halves again when CO, is added to the hemoglobin (Fig. Gas exchange in diving animals is markedly uneven 33). This minor enthalpy change means that, in the presence as the animal switches between a closed system during of organic phosphates and CO,, oxygen biding is virtually dives to an open system during recovery. During recovery, independent of temperature. Fbrthermore, the CO, effect is, the potential exists to achieve a steady state in gas ex- in itself, temperature dependent. There is no C0,-induced change, although this potential is modulated by the meta- release of oxygen from hemoglobin at 37'C, but there is a bolic rate, the "nature" of the oxygen deficit, tlie level of marked release at 20°C (166). Also, if the animal is breathing ventilation, and the state of the circulation. The oxygen extremely cold air, then the allosteric behavior of Cqmay deficit may he aerobic, requiring only replenishment of promote oxygen uptake at the lungs. Elimination of CO, will oxygen stores, or have anaerobic components (143). increav? the oxygen affinity of hemoglobin provided the lung When a prono~incedanaerobic contrih~itionhas heen surface is cold (165). made to dive metabolism, then recovery is prolonged and In birds, inositol phosphates are the principal alloskric cam even be sLreLched oul over several dive/surFace cycles modulators of oxygen mty, and there is some evidence (73, 227, 331). that the effect of saturating concentrations of organic phos It is not unreasonable to suppose that surface times 880 PATRICK J. BUTLER AND DAVID R. JONES

- . .

nc. 34. Minute volume (Vmin), tidal volume, and respirato'~frequency averaged over first 2 min after :. surfacing against durations of natural dives by 5 Wed- dell seals varying ie mws frurunt 370 to 450 kg. Minute = 160 volume and Udal volume culves were drawn by eye. All volumes are in liters, BTPS. [From Kooyman et al. . Minute volume N.9 N:7 (226). Reprinted with kind permission of Elsevier Sci- o Tidal volume ence-NL,Sara Burgerhatstmat 25, 1066 KV Amster- ' 80 dam, The Netherlands.] * Paling 5.10 10.8 2HO ID I

Dive duration lminl will be proportional to the preceding dive tie. Certainly, the Weddell seal was due to an increase in V, of 1.5-2 in Weddell seals, dives exceeding the ADL are followed times while f,, increased 3 times (Fig. 34). Tidal volume hy extended rerovery periods (236). A 36-min dive was could approach 75% of TLC, which would promote rapid followed by -10 min of recovery, whereas a dive in excess and complete exchange of alveolar gas. In contrast, when of 60 min required 120 rnin for recovery (Fig. 17). Blood .f,, is >20 min-', postdive VT seems to be closer to that lactate concentration was -5 mM at the end of the 35- of tel~estrialmammals of the same size. Both Gallivan min dive and 25 mM at the end of a 61.4min dive (Fig. (160) and Craig and Pkche (94) reported that, at f,,, of 14). Hence, surface interval increases exponentially with -27 breathdmin in harp seals and over35 min-' in harbor dive time (143). In rare instances, diving may continue, seals, VT was between 20 and 3096 of estimated TLC. In but the subsequent dives are short (73). However, long fact, Pkche (318) had remarked earlier that the ability to dives are rare in Weddell seals (Fig. 9) and usually the increase VT and fmSp above resting levels by harbor seals, surface interval varies from 2 to 5 min (Fig. 17; Ref. 235). in response to hypoxia, was less than that in terrestrial In contrast, gray seals show a direct proportionality be- mammals. tween surface time and dive time only up to 7 min of Birds have an exceptional volume of air in their lung- diving (Fig. 13). Above 7 min of submergence, surface air sac system; this volume is three to seven times greater time is independent of dive time (378). Ohvionsly, there than that of most diving animals on a nnit weight hssis is a major change in the diving physiology and metabolism (Table 1). The only diving mammal that comes close to of gray seals after 7 rnin of diving. In this respect, Gallivan the TLC of birds is the sea otter (249, 250, 253). Hence, (160) noted that in unrestrained harp seals, short dives, in birds. VT is probably only a small proportion of TLC usually associated with high metabolic rates, required rel- during postdive hyperpnea Breathing freqnency in- atively longer surface times than long dives, in which met- creased in both pochard (58) and tufted ducks (59) during abolic rates were low. the interdive intervals of a diving bout. Both birds and mammals hyperventilate, not only be- Northern elephant seals at sea make extremely long fore, but also after natural diving (59, 94, 226, 235, 341). dives, frequently in excess of the calculated ADL (see Even animals that usually take single breaths will increase sect. nD), with only short surfaxe intervals between them the frequency of individual breaths after a dive, i.e., (Fig. 11). On land, awake and sleeping seals show pro- whales and dolphins (83, 222, 340, 341). Ventilation is the longed periods of apnea interspersed with short bouts of product of tidal volume (%) and respiratory frequency breathing. In northern elephant seals, respiratory varia- If,,), but there is often a reciprocal relation between tions in heart rate are pronounced, with heart rate increas- these two variables during hyperventilation, the limits being on inspiration and falling during expiration (13, 69, ing set by lung mechanics. Tidal volume as a proportion 70, 76). Surprisingly, respiratory variations in heart rate of total lung capacity (TLC) is exceptionally large in many were not observed in beached southern elephant seals of the single or low-frequency breathers. In animals with (213). In a study with beached subadult northern elephant breathing frequencies of ~20breathdmin, VT may vruy seals, spectral analysis of cardiac intervals and breathing from 40 to 90% of TLC during postdiving hyperpnea (334, patterns showed maximum power at frequencies of be- 341). Kooyman et al. (226) found that hyperventilation in tween 0.1 and 0.2 Hz for both breathing and heart rates JILL?/1991 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS

E - = 0.8 - 3 C 06 r, nc. 35. Spectral analysis of heart rate variability in 0 -. - an elephant seal breathing on beach and breathing at sea 0.6 - 04 bet.ween dives. On heath, frequency of maxirnum power 2 (0.15 Hz) was identical to breathing frequency. Frequency rn S of maximum power at sea (-0.4 Hz) gives a breathing 04 - 0'2 kquency of 24 min-' between dives. A nonlinear band- U - -.-N ... o pass fdter was used to remove baseline trend and vely Q 0.2 0 w low frequency components. (R. D. Andrews, Y. Yeh, D. R. E a Jones, J. IJ. Williams, P. H. Ibo~on,D. P. Costa, and B. J. 0 - Le Boeuf, unpublished data). = 0 -0.2

1 I I I I I I 1-0.1 0 0.1 0.2 0.3 0.L 0.5 0.6 0.7 0.8 Frequency (Hz1

(Fig. 35; R. D. Andrews, Y. Yeh, D. R. Jones, J. D. Williams, Any circulatory changes initiated in anticipation of, P. H. Thorson, D. P. Costa, and B. J. Le Boeuf, unpub- or coincident with, the start of diving are usually reversed lished data). Hence, spectral analysis of cardiac intervals before the end of the dive, in anticipation of surfacing. alone can be used to give breathing rates of animals at Thompson and Fedak (378) suggest that during anticipa- sea When breathing at sea, maximum power occurred at tory tachycardia, blood flow may be restored to previously 0.4 Hz (Fig. 35; Andrews et al., unpublished data). This hypoperfused areas of the body so that the blood will be breathing rate is three ties the maximum rate observed further deoxygenated and maximum oxygen uptakc on in beached juvenile seals (13, 24). surfacing will be ensured. Certainly, gray seals showing The time taken for a single breathing cycle in masine no anticipatory increase in heart rate did not achieve max- mammals is usually short, which is advantageous in reduc- imum rate of oxygen uptake with the first few breaths ing surface time. The extreme case is the dolphin, Tursi- (334). This may be analogous lo the situation at the start ops, whch exhales and inhales in -0.33 s (222,245). Expi- of exercise in humans when ventilation runs ahead of ration in marine mammals is usually briefer than inhala- circulatory changes (398). Heart rates and, presumably, tion, with expiratory air flow rates being higher than cardiac outputs are quickly restored to predive levels and inspiratory rates (229,311,334,360). The hydrostatic pres- are nsr~allywell above resting levels in both birds and sure head when Lhe arlurial approaches the surface, allied mammals at the end of a dive (6, 17, 66, 101, 142, 177, to extremely compliant chest walls, means that the driving 205, 232, 331, 378). In harbor seals, there is no relation pressure for exhalation will be very high (15, 249). To belweer~hear1 rate during breathing and the length of the tale advantage of this large driving pressure, many marinc prcccdiing divc (142). This suggests that in seals, heart mammals exhale through constricted airways ("pmed- rates in the range of 80-150 beatshin observed in the lips" breathing; Refs. 216, 229), and the airways them- early postdive recovery period may be maximal. Seals selves are strongly reinforced to prevent collapse (15,114, contrast with birds, in which interdive heart rate increases 115,234). Nevertheless, peak expiratory flow rates in ma- with each succeeding dive in a bout, up to rates ap- rine mammals are not exceptional in comparison to those proaching 5.00 beatslmin (59). If the surface intesval is for terrestrial mammals (250), but what is exceptional is prolonged, then heart rates usually decline in both birds the ability of marine mammals to maintain high flow rates and mammals. at extremely low lung volumes (15,222,234). Hence, rein- Hyperpnea combined with increased circulation forced airways prevent l~lngcollapse at high diving pres- leads to arapid restoration of the partial pressure of blood sures and permit high flows at all lung volumes, thereby gases. The Pa+ value is usually elevated somewhat during shortening expiratory time. This allows V, to be extrenrely the early postdive recovery period in Weddell seals, reach- large and end-expiratory volume to be low so that there ing values well above those at rest (Fig. 30A). Changes is a high turnover of alveolar gases (250). Reinforced air- in Pat,,, are more complex depending on whether the dive ways also have an additional role in ensuring orderly col- exceeded the ADL, in which case a large lactate load is lapse of the lungs from the smallest to the largest airways added to the blood during the recovery period. After natu- during hydrostatic compression, thereby removing poten- ral dives, P*,, falls rapidly, usually within 1-2 min, tially hazardous nitrogen from the exchange surface dur- whereas after extended dives, it may take 5-10 min for ing deep dives (216, 338, 346). Pea, to be reduced to or below predive values (235,331). 882 PATRICK J. BUTLER AND DAVID R. JONES Volz~me77

Acid-base balance is little disturbed by natural diving but, ml. min-' . kg-'. Thii is within the range of estimated and after extended dives, when large concentrations of lactate recorded n~axilnumfor this species (132, 407). appear in the blood markedly reducing base excess, pH, After dives of 5-20 min, Weddell seals had a maxi- falls and usually takes 30-45 min to regain predive values mum oxygen consumption of >40 ml O,.min-'. kg-' (235, 331). Body temperature of seals may drop during (227). This was 2.5 times oxygen consumption after short diving, particularly during extended dives, so the initial (<5 min) dives and, surprisingly, 25% greater than maxi- blood gas values could be overestimated (331). In birds, mum oxygen consumption after extended dives (20-70 there is little information about blood gas tensions in the mid. The rate of oxygen consumption fell to the resting postdive recovery period, although in diving bouts by rate (5.15 ml O,.min-'. kg-') 5 min after surfacing, al- ducks, pH, declines due to a doubling of blood lactate though hyperventilation persisted for at least 20 min after (48, 49). In contrast, although lactate also increased in extended dives (226). These values can be compared with rhinoceros a~~klet.~making escape dives, the acidosis was estimated rate of oxygen consumption from data averaged partially compensated by lung ventilation between dives, over a dive and surface cycle (75). After 15-min dives, because Pko, did not increase progressively throughout with 2.5-5 min of recovery, oxygen consumption was the diving bout (368). Only minor changes in partial pres- within the range of 16-36 ml 0, .fin-'. kg-', which falls sure of oxygen in the blood and in blood pH wcrc ob- within the range of values given by Kooyrnan et al. (227). served after voluntary diving in penguins (285). For diving ducks, the situation appears to be some- Due to the increased hematocrit and Ph, in Weddell what different. The rate of oxygen consumption during seals after diving, Cao, can attain the re~narkablevalue of the recovery period is around six times resting oxygen 33 vol% (Fig. 30B). But Ch, falls as RBCs are removed consumption (414), which is probably half the minimum from the circulation. Obviously, it is advantageous if RBCs oxygen consumption that occurs during flight (48). Thii are stored early in recovery, when Pao, is highest. Also, may reflect the much higher levels of Pao, at the end of the metabolic acidosis after a long dive could reduce 2,3- a dive in ducks as well as lower V,, as a proportion of DPG content of the RBCs, thereby increasing blood oxy- total lung/air sac volume, during the recovery period. gen affinity, although there was no evidence for this from The above discussion has been about divers that take the in vivo oxygen equilibrium curve (331). However, the multiple breaths during the recovery period. Manatees, on time course for restoration of resting hematocrit levels the other hand, only take a single breath on surfacing, so is controversial. Recent data suggest that hematocrit is there is no possibility that manatees will ever attain a elevated throughout a diving bout in Weddell seals (73) steady state (161). In fact, large imbalances in oxygen and also during apnea-eupnea cycles in elephant seals stores have to be corrected over several dive-surface cy- (70). Hence, only if the interdive interval is prolonged cles. Hence, maximum oxygen consumption values re- (>5 min) will hematocrit return to resting levels between corded in manatees are much lower than those of seals, dives. being in the range of 8.0 ml 02.min' . kg-'. Tidal volumes Elevated hematocrit, rapid circulation, low oxygen in manatees are only -4W of TLC (162). levels in arterial and venous blood, combined with hyper- In both birds and mammals, hyperventilation before pnea, mean that oxygen consumption during the recovery a dive is usually associated with respiratory exchange phase will be high. This will allow rapid repayment of ratios (RE) greater than one, because oxygen stores are oxygen deficits if no anaerobic debt has been incurred. loaded and C02 is blown off (94, 227,414). Hypewentila- In fact, oxygen consumption during the postdive recovery tion, after short dives by seals, does not usually cause phase may represent the highest rate of oxygen consump- much of a change in RE, compared with that at rest (94, tion ever recorded for seals and dolphins. 160,227), whereas during recovery from longer dives, RE Even after forced dives with restrained animals out increases and may exceed unity after 5-6 min of hyper- of water, maximum oxygen consumption of gray seals ventilation (227, 235). During the early phase of recovery, approached 24 ml 02.rnin-' .kg-' (346), which is higher oxygen consumption and carbon dioxide consumption than maximum values reported for exercise (141), but not must increase in step but, even without addition of lactate as high as the oxygen consumption of unrestrained gray to the blood, it is possible for RE to increase due to differ- seals, which approaches 40-45 ml 0,. min-'. kg-' for ent time constants for loading and unloading 0, and CO, breaths associated with the lowest end-tidal Po2 values slores, respeclively. In cu~~lrasllo seals, ducks recovering during the recovery period (calculated from Ref. 334). from dives shorter than their ADL have an RE in excess Unfortunately, the current proclivity for presenting recov- of unity (414). Because lactate increases during a diving ery as a function of a dive and surface cycle means such bout, CO, will be blown off to preserve acid-base homeo- values do not represent maximum during the surface pe- stasis (49). riod. However, this was: not the case with data from har- During recovery, increases in ventilation and circula- bor seals (94). Oxygen uptake in the short surface inter- tion should be matched, although there is some evidence vals (0.25-0.5 min) rcachcd valucs between 30 and 50 to suggest that circulation during recovery may be selec- July 1997 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 883 tively altered as it is during diving. Scholander et al. (348) These afferents and their interact,ions have heen reviewed remarked on the mosaic appearance of muscle during often, and at length (28, 45, 53, 105, 365, 369). recovery from forced dives, indicating that peripheral va- Changes in heart rate in anticipatioli of natural diving, sodilatation was modulated. This observation was con- or surfacing, in birds and mammals suggest that suprabul- hedand extended to include other hypoxia-resistant bar or even cortical inRuences may modulate thc rcflcxo- tissues (e.g., liver, kidney, and stomach) in a detailed mi- genic response. That is not to say, however, that the same crosphere study by Blix et al. (27). In fact, even during mechanisms are invnlved in birds and mammals. A hint recovery from free dives, there is evidence that restora- that the mechanisms are different may be gleaned from tion of kidney and liver function may be delayed. The half- the fact that bradycardia can be conditioned in sea lions time values for indicators of kidney and liver function (337) but in birds the cardiac response has only ever been during recovery from diving in Weddell seals were much overrulcd, not accentuated, by habituation (154). On the longer than those obtained from resting animals (168). other hand, heart rate is lower when birds and small mam- Clearances of the kidney function indicators, p-aminohip- mals make "escape" dives compared with heart rates dur- purate and inulin, during recovery from a 24-min dive ing natural dives, suggesting similar responses to arousal were, respectively, two and approximately flve times (151, 279). slower than those at rest (168). Similarly, cholic acid clear- In muskrats, there was a marked difference in brady- ance by the liver was over three times faster at rest than cardia during natural (115 beatslmin), escape (95 beats1 during recovery following a natural dive that was well min), and forced (60 beatslmin) dives (279). A similar within the ADL (168). pattern of cardiac response as in the muskrat was displayed by the beaver (166,167). Physical removal of -51% of the cortex in muskrats eliminated any differentiation VI. CONTROL OF CARDIORESPIRATOILY in cardiac response to the type of dive, with all heart RESPONSES rates being similar to those during natural dives (279). Interestingly, sham operation had no effect on heart rate In response to forced diving, both birds and mammals during natural dives but accentuated the cardiac response display the same suite of responses: apnea, profound to escape and forced dives. Hence, it is tenlptir~gLO sug- bradycardia, and marked increase in peripheral resis- gest that upper respiratory or pulmonary afferents initiate tance. Similar afferent and efferent neural pathways con- the bradycardia, which is then njodulated by corlical UI- trol the circulation in forcibly submerged birds and mam- fluences. The extent of this modulation in muskrats seems mals. The responses are reflex, being little altered by de- quite small but, in harbor seals, the "reflex" response to cerebration (2, 125) or even brain transection at the submergence can be completely overruled (205). The ex- pontomedullary level (155). Nasal or other upper respira- pression of an extreme reflex-type cardiac response oc- tory tract receptors initiate the cardiac responses in diving curs during natural dives (6, 378, Table 4) and, in capt,ive birds and mammals (125, 127, 151, 313-315), and these animals, when they are presented with unusual stimuli receptors may even play a role in dabbling ducks, such during submergence (141). The rapidity with which seals as the mallard, when predive heart rate is high, well above can switch from one level of heart rate to another led resting. This initial response in dabbling birds is subject Fedak (141) to suggest that marine mammals have incor- to habituation (156). Collapse of the lungs or cessation of porated autonomic function into the realm of behavior. activity in central respiratory neurons may have provoca- In this respect, it is interesting that Elsner et al. (139) tive, permissive, or facilitory effects on the diving cardio- found no anticipation of surfacing in a blindfolded ringed vascular responses (8, 57, 125, 126, 207, 260, 396). seal. Both central and peripheral chemoreceptors will be It seems unlikely that the reflex expression of the stimulated by the progressively developing hypoxia and cardiac response is modulated by higher nervous centers hypercapnia during apnea, and their output causes the in diving birds, because pronounced reflex controls on majority of the cardiovascular response in dabbling ducks heart rate during natural submergence have never been (209), whereas chemoreceptors reinforce the response in demonstrated. [A rider should be added here, because forcibly submerged birds and mammals (61,106,133). The cardiac responses to exhalation per se on submergence nasal and chemoreceptor components can be habituated have never been investigated during voluntary diving in both diving and dabbling birds (154) and, in dabbling (207). However, with respect to pulmonary afferents, their ducks, habituation even occurs after decerebration effect may be minimal because cardiac responses to vol- (G. R. J. Gabbott and D. R. Jones, unpublished data). The untary diving are unaffected by breathing hypercapnic role played by baroreceptors in the circulatory adjust- gases before a dive (55). In birds, pulmonary receptors ments to forced submersion in ducks is disputed (356, are silenced by high levels of C02in the airway.] Abolition 357, 369), but baroreceptors may be important in seals, of nasal receptor input by local anesthesia of the internal due to an increase in gain of the baroreceptor reflex (7). nares, but not the glottis, only reduced heart rate by 10- 884 PATRICK J. BUTLER AND DAVID R. .JONES Volz~ms77

30% during natural submergence in redhead ducks com- argue against conditioning of cardiac adjustments (212). pared with untreated animals diving naturally (151). Dur- The ducklings had been on the water an average of 1.2 ing forced dives, anesthetizing the internal nares elimi- days before their first whole body dives. The average ini- nated up to 80% of the bradycardia (151). tial cardiac responses to beak and head submersions (pcr- Carotid body chemoreceptor denervation (61) or formed before any dives), first-ever dives, and subsequent breathing hyperoxic gas mixtures before diving (55, 151) dives were similar to those seen in adult tufted ducks had no effect on heart rate during natural dives of normal upon natural submersion (59, 367). Although it is clear duration in redhead and tufted ducks. However, carotid that the cardiac response to natural diving is not a hahitu- chemoreceptors were involved in the gradual develop- ated forced dive response, as suggested by Gabbott and ment of bradycardia during extended dives by the tufted Jones (IM), it is uncertain whcthcr the response is ulti- duck (F'ig. 24E; Ref. 55). Chronic barodenewation had mately conditioned on nasal receptor input, because there no effect on heart rate during natural dives of redhead can be no expertation that the reflex response to submer- ducks (152). Interestingly, however, arterial barorecep- gence should be different from the conditioned response; tors play the major role in the cardiac adjustments to in fact, they would be the same. Support that some form whole body submergence in the mallard duck, trained to of "learning" may be involved in cardiac control during dive rather than to dabble for food (153). In intact ducks, natural diving is given by recent work on elephant seal heart rate was aqjusted to -250 beatslmin during dives. neonates and pups (78). Heart rate does not cycle up and If predive heart rate was <250 beatslmin, then heart rate down with breathing (sinus arrhythmia index; Fig. 36D) increased on submergence; if >250 beatslmin, then heart in young pups and during apnea their heart rate can be rate fell on submergence. Cardiac adjustments to lrained in excess of 100 beatdmin (Fig. 3GB). In contrast, in older dives were virtually eliminated by chronic barodenerva- pups, sinus arrhythmia is pron~inenland heart rate during tion; heart rate remained unchanged throughout thc div- apnea is considerably lower, close to the minimum seen ing maneuver (153). during breathing (Fig. 3GC). Durations of apnea increased Many diving birds make the whole gamut of cardiac with age (Fig. 36A). Obviously, tight cardiorespiratory aqjustments to diving (predive tachycardia followed by control is crucial for long-duration apnea, and this control a rapid decline in heart rate) even when no dive is per- matures over time (78). formed (414). On the other hand, the full, forced dive That the vagus is tlre primary efferent neural pathway type bradycardia occurs in both ducks and penguins if for cardiac adjustments to diving in ducks was shown by the normal diving routine is interrupted, usually by trap- Butler and Woakes (GO). After prctreatment with atropine, ping the animal underwater and preventing access to heart rate of tufted ducks was high and did not change the surface (62. 151,367). These obselvations, together, during escape dives. Similar observations have been made suggest that the initial cardiac response in ducks may be on atropiniied muskrats diving on an artificial pond (273, conditioned and that the full bradycardia after trapping 351) and harbor seals, diving at sea (299). A point of inter- results from dishabituation of the conditioned response. est is that atropinized muskrats dive voluntarily beyond Because trapping is usually done sometime after submer- their ADL and can withstand forced submergence for over gence, then it is the chemoreceptor input that will domi- 7 min (351). This contrasts with an atropinized seal diving nate in this phase. Hence, denervation of chemorecep- at sea, which succumbed in 3 min (299). tors should greatly affect the evolution of the profound Furilla and Jones (152) speculated about the efferent bradycardia. As was tlte case with both nasal and barore- neural control of heart rate responses during forced, ceptors (152), the cardiac response to natural diving was trapped, and natural diving in ducks. They observed a unaffected, but subsequent reinforcement and maintc- strong correlation between predive (or pretrapped) heart nance of bradycardia during trapped and extended dives rate and heart rate some 2-5 s after diving (or trapping) in were slowed aft,er long-standing bilateral denervation of redhead ducks and lesser scaup (A. &inis). The relation carotid body chemoreceptors in the tufted duck (55). encompassed natural diving, forced diving in restrained Therefore, llre suggestion that dishabituation after trap- ducks, and even ducks forced to dive while exercising ping allows the full chemoreflex response to be played on a treadmill (Fig. 37B). Fu~thenirore,tlre relatiorr also out is supported. On thc othcr hand, this cannot be the described predive-dive heart rate relations in tufted ducks explanation for the fact that bradycardia during ex- (A. fuligula) studied by Butler and Woakes (59). Finally, tended dives attained the same level as after trapping the relation even applied to diving ducks when they were (66). Gabbott and Jones (154) showed that intense che- dabbling (Fig. 37A). This last observation is significant, rnoreceptor input could not be habituated during forced because metabolic rate of diving ducks during dabbling dives of restrained ducks, and this is probably the case must be well below that when diving, indicating that the during extended dives. level of exercise in ducks diving voluntarily has, per se, Cardiac responses to the first dives ever made by little influence on diving heart rate. ducklings were similar enough to the adult response to Furilla and Jones (152) were able to identify regions July 1997 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS

FIG. 36. Changes in apnea duation (A), averaged heart rates (f") during apnea (B), and eupnea (C) with age in -7 iro northern elelphant seal pups. D: sinus ar- C lythmia index, which is difference between inspiratmy and eqiratory heart rates. [From Castellini et al. (78).]

I: 110 C 0

D 0 m 90 0 0 oO .U 3 80

Age (days I Age Idaycl

on a perspective plot of heart rates, resulting from vagal after the first dive of a bout, the heart rate will fall from and cardiac sympathetic nerve stimulation in domestic 100 beatdmin (Fig. 38, point C) to 20 beatsimin (Fig. 38, ducks (A. platyrhynchos), which corresponded to their point D), and vagal activation will again be maximal. empirically established predive-dive heart rate relation Other predive-dive heart rates can also be recognized on (Fig. 38). In essence, Furilla and Jones (152) argue that, Figure 38. In the laboratory, predive heart rate is low regardless of the level of cardiac sympathetic excitation, (-100 beatdmin) and is unaffected by P-adrenoceptor all diving (or trapping) is accompanied by an increase in blockade, so the vagal-sympathetic interplay producing vagal activity to the heart of some 50% of the maximum this point must be at point C (Fig. 38). Hence, a 50% possible activation (Fig. 38). increase in vagal activity will give a rate of 20 beatslmin For example, in the first dive of a series, heart rate (Fig. 38, point D), which is common during laboratory increases from the resting rate (100-120 beatdmin; Fig. dives. Finally, predive heart rate in P-adrenoceptor- 38, point 0 to 275 beatsimin (Fig. 38, point B). This blocked ducks diving voluntarily never exceeds 300 beats/ increase is rapid, unaffected by P-adrenoceptor blockade, min and is described by point B (Fig. 38). During dives, and must be due to withdrawal of parasympathetic nerve heart rate in 0-adrenoceptor-blocked ducks falls to 100 activity (152). When a duck dives naturally, heart rate falls beatdmin (Fig. 38, point C) which, again, represents an to 100 beatdmin, representing an increase in vagal activity increase in vagal activation of 500% of maximum. of some 5M. At the end of a voluntary diving bout, predive There is little information about the role of the sym- heart rate, due to increased sypathetic outflow to the pathetic nervous system in regulating circulatory re- heart, will be 500 beatdmin (Fig. 38, point A) and during sponses to free diving. 0-Adrenoceptor blockade in ducks dives, heart rate will fall between 200 and 250 beatdmin prevents the interdive heart rate from going over 300 (Fig. 38, point F), representing an increase in vagal activ- beaidmin during a diving bout (152). Normally, interdive ity of 50% of maximurn. If this animal is prevented from heart rate will approach 500 beawmin toward the end of resurfacing, heart rate will fall from 250 to 100 beaWmin a bout (59), and this increase in heart rate, therefore, must (Fig. 38, point E), and vagal activity will now be maximal. be sympathetically driven. However, the increase in heart On the other hand, if the bird is prevented from surfacing rate, from rest to 300 beatdmin, before the first dive of a 886 PATRICK J. BUTLER AND DAVID R. JONES VO/~I.">IP77

A.. (255). lntensc vagal activity during diving blocks sympa-. . t the& inputs to the heart, despite the persistence of sym- 111 oivins papalhetic lone, so corr~plelepharmacological sympathetic blockade will not result in a significant decrease in heart 4 f rate. This means that the parasympathetic system takes t4 over cardiac control durhg diving and that sympathetic I21 Dabbling input to the heart is ineffective while the animal is underwater. In fact, this would be a very effective way of rapidly suppressing sympathetic influences, because a response i ti t to changes- in svmuathetic- activation occurs much more (31 Face immersion slowly than to changes due to parasympathetic activity (1, 152, 200). A cholinergically mediated reduction of the response of cardiac cells to adrenergic stimulation during -5 s diving may also explain why forced-dive harbor seals de velop a diving bradycardia, despite a large increase in B circulating epinephrine and norepinephrine levels (171). 300 - Also, mallard ducks forced to dive for several millutes o display the most remarkable increase in circulating cate- -- cholamine levels known, yet diving heart rate remains .-C E low (237, 238, 239). During forced dives, vagal output is - 200 - 4-Y assumed to be maximal and is obviously able co~upletely Z to ovenule 0-adrenergic activation. & . L Interestingly, in the naturally diving muskrat, P-adre- 0 a = 100 - noceptor blockade with nadolol has no effect on diving U heart rate, whereas bradycardia is reduced by propranolol .-> m 0

0 I Ij 50 100 500 Pre-dive heart rate (min-'I

FIG.37. A: electrocardiogram traces from a redhead duck dlvlng voluntarily (I), dabbling (2),and immersing its head into a beaker of water to retrieve food (3). Downward arrows are appmrdmate points of submergence, and npwud arrows are points at which dock s~rfared. E relationship between dive (or tsapped) heart rate and logarithm of predive (or pretrap) heart rate for all dives. 0,Restrained dives including those after exercise; o, trapped dives; A, natural dives including 8-adre noreceptor blacked dives. [From Furilla and Jones (152). Copyright is held by Company of Biologists LM.] bout is probably due entirely to vagal withdrawal. The initial increase in heart rate is associated with a period of hyperventilation, but whether hyperventilation is the causd factor is unknown. Thc incrcasc in intcrdive hcast rate during a bout (-200 beatslmin) is somewhat greater tlm the increase in diving heart rate (-120-150 beats/ FI~;:In Rc~lala~tvl~il,uf Iw:&n ~rlc 1.1 llila!rr:,l .;rltmllaliut~of ilr,r;rl min; Fig. 38), suggesting that a 50% of maximal increase cur rnds ol and cardwr nmpallrrra. nrrvrh of Pcklli duck i.4 in vagal activation on submersion is not effective in over- plat~i~yl$clzos);10096 represents frequency of stimulation above which no further changes in heart rate occurred. Heart rate resulting from a ruling the increased sympathetic activation of the heart mven level of vaeal and svmaathetlc stlrnulatlon was dotted on 'ber- in a bout of diving. ipeetive" gmph paper and sirface was drawn, by eye, to encamiass ~k~internlay between the two branches of the auto. all heart rates obtained in stimulation experiments. Point B represents complete canlit detiervatiun, and puinl E represents n-lal vagrr nomic nervous system in modulating diving bradycardia sympathetic stimulation. ~ffectsof sympathetic stimulation at minimal has been investigated in naturally and forced diving musk- and maximal vagal activity are represented by lines A-B and D-E, effect rats in some detail, signore and jones(3,511 propose that respectively. Similarly, of vagal stimulation at minimal and maximal sympathetic activity are represented by Lines B-D and A-E, respec- accentuated antagonism Occurs between the tively. See text for an explanation ofpointsA-Frelative to diving. [From branches of the autonomic nervous system dluing diving Mllaand Jones (152). Copyright is held by Company of Biologists Ltd.] July 1997 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 887 treatment. Propranolol, unlike nadolol, crosses the blood- not be reduced to predive levels, even though heart rate brain barrier, which suggests that diving bradycardia may may be approaching predive rates. The accentuated antag- be partially regulated by a central catecholaminergic path- onism between the parasympathetic and sympathetic way (351). This serendipitous observation gives insight branches of the autonomic nervous system will be re- into the pharmacological basis of natural diving bradycar- duced by a fall in vagal parasympathetic activity, allowing dia and complements studies on forced-dived muskrats the cardiac chronotropic response to adrenergic stimula- and rats which have shown the importance of excitatory tion to be expressed (352). amino acids as transmitters in early parts of the trigeminal Breathing rates are high during the postdive period neural circuit (280, 316). and in elephant seals, for example, appear to approach In muskrats, diving bradycardia is also unaffected by maximum rates (Andrews et al., unpublished data). Cen- a-adrenoceptor blockade with phentolamine (351). This tral neural interactions between respiratory and circula- suggests that the cardiac response to submergence is in- tory neurons, as well as afferent neural input from lung dependent of peripheral vasoconstriction. Dissociation of receptors, cause heart rate to wax and wane with each bradycardia and arterial constriction was previously breathing cycle. When hreat,hingis slow, these oscillations shown in the seal forced to dive (296). Even in the absence in heart rate are particularly obvious. Even so, the highest of vasoconstriction, muskrats may submerge voluntarily heart rates reached on iirlralalio~rare generally below for over a minute in excess of the calculated ADL (270), those attained in the interval between dives, when heart but underwater survival of forced dived, unrestrained rates may approach the maximum ratc for thc species. In muskrats is reduced to 5 min. Underwater survid time elephant seals, heart rate was almost twice as high at sea, of forcihly snhmerged animals is not fi~rtherreduced by in the interdive periods, than during eupnea on land (5, treatment with a combination of phentolamine and atro- 70). An increase in sympathetic influence on the heart, pine (351). either neurally or humorally, accentuates the increase in Circulatory catecholamines have been measured in heart rate in the immediate postdive period initially set Wcddcll scals on rctum from natural divcs (188). Aftcr in train by a decline or cessation of vagal parasympathetic short dives that are <5 min, there was no increase in activity (351,352). Restoration of normal blood gas values epinephrine or norepinephrine during the postdive pe- may make a contribution to the elevation in heart rate riod. However, after dives longer than this, epinephrine because in ducks, postdive heart rate was usually higher and norepinephrine increased in direct proportion to after carotid body denervation (55). Also, postdive tachy- dive tinre, reaching levels at least two- or threefold above cardia was sig~~ilicanllyrecluced, cornpared with arrinmls resting levels after dives longer than 15 min. Interest- surfacing into air, when ducks breathed a hypercapnic ingly, these increases were in Lhe sanle range as those gas ~r~ixtureafter surfacing, whetlrer or riot ducks had from seals forced to dive for up to 15 min (424). The innervated carotid bodies (55). Because lung receptors in forced-dive rrteasurenrerrts were itlade near the end or birds are sile~tcedby high levels of carbon dioxide in Lhe the dive and not postdive as with the observations from airway, their input ensures full expression of postdive naturally diving seals. If circulating catecholamines re- tachycardia. Nevertheless, the major contributor to post- ally do increase during diving and not just during the dive tachycardia remains to be demonstrated, since post- postdivc period, thcn it is possiblc that they could con- divc hcart ratcs were around twice thc diving ratc cvcn tribute to circulatory adjustments (including contraction in ducks with intact carotid bodies breathing hypercapnic of the spleen; see sect. IVB)during submergence. Cate- gases (55). cholamines remained elevated for a prolonged period in Control of breathing before and after diving has re- Weddell seals during recovery (188). ceived scant attention in freely diving animals compared During voluntary diving, heart rate starts to increase with studies on forced-dived birds and mammals (53,365, in anticipation of surfacing. In larger diving birds and 369). Diving ducks and Wcddcll scals hyperventilate be- mammals (emperor penguins, Ref. 232; seals, Refs. 4,378), fore submergence. In Weddell seals, Pk, increases and heart rate may reach predive rates before surfacing, al- Pa,,, decreases, which has led to the suggestion that Wed- though in smaller species (ducks, Ref. 367; muskrats, Ref. dell seals have two settings to their respiratory controller 351) heart rate seldom exceeds resting rates in the imme- (227). The P*, value is lower and the Pko, value higher diate period before surfacing and is well below predive in resting seals. How the settings are changed just before rates. Obviously, efferent vagal parasympathetic activity diving is unknown. In this respect, it could be argued that must decline during this period, but how this decline is predive hyperventilation is part of a "feed-forward or hrolrght ahont is specolative. A blindfolded ringed seal "central command" component of respiratory control showed no increasr in heart rate when approaching a hole (128). in L11e ice cover oTa lake, which suggesls that suprabulbar All diving birds and mammals are sensitive to low influences may be involved (139). IT circulating catechol- oxygen and high GO2 in terms of their respiratory re- amines increase during the dive, then vagal activity need sponses (53, 377). The sensitivity (I. min-' . kPa-') of the 888 PATRICK J. BUTLER AND DAVID R. JONES Volume 77 ventilatory responses to CO, in diving birds and mammals study, there was no doubt that hypercapnia had more is similar to that in terrestrial species, alUrough the thresh- significant effects in reducing dive time than hypoxia. This old for the response may be raised (32, 56, 94, 159, 289, is similar to the situation in the tufted duck in which 317). The response to oxygen lack may be "blunted" in hypercapnia reduced dive tiemuch more than hypoxia adult seals (317, 318), although newborn gray seals have (55). Interestingly, breathing hyperoxic gas nlixtures did a brisk hypoxic response, indicating that any blunting of not affect dive time at all in manatees (159) or tufted the ventilatory response occurs postnatally (290). ducks (55) and actually shortened dive time in phocids Hence, at the end of a dive, hyperventilation is driven (318). In phocids, hypercapnia also shortened dive dura- by the low Pk, and high Pko, of the blood. However, in tion, and CO, relenlion could explain the shortened dive Weddell seals, hyperventilation is related to the length of time after breathing hyperoxic gases (319). the previous dive and, aCLer am exknded dive, may persist In tufted ducks, denervation of carotid body chemo- for 70 min, yet Pk, and Pko, are usually above and below receptors significantly increased natural dive time (61). predive levels, respectively, after only 5 min of breathing Furthermore, in intact ducks, virtually the whole of the (236,331). Consequently, the prolonged phase of hypenren- reduction in dive time caused by breathing hypoxic gas tilation is driven not by O2 or C02 but by the change in hefnre diving can he attribnted to stimulation of carotid acid-base status of the blood consequent upon the release body chemoreceptors (55). Looked at in this light, the fact of lactate into the blood (227, 2J5, 331). A similar situation that Weddell seals are claimed to anticipate the duration holds during recovery from long forced dives of domestic of a dive and adjust heart rate accordingly (219) may ducks (M. Shimizu and D. R. Jones, unpublished data). only be a reflection of a lowered rate of hloorl oxygen Single-breath breathers (159) and even Weddell seals utilization when heart rate is low; that is, the seal adjusts on occasion (73) reduce succeeding dive times when re- dive duration corresponding to heart rate, and rate of fall covering from an extended dive. Therefore, control of in Pa,,, established early in the dive (205). The obvious apnea and control of dive time can be viewed in the same complex behavioral components associated with surfac- context. Apneas can result from reflexogenic contribu- ing to breathe led Fedak and Thompson (143) to warn that it is unrealistic to expect a single variable alone to tions from upper airway receptors (10) or cessation of terminate breath-holding. activity in central respiratory neurons (76), but motivation Of course, both behavioral and physiological factors must also be a powerful factor in determining the length can modify the integration of reflex mechanisms, a most of the breath-hold in diving animals, as it is in voluntarily obvious one is sleep. Many diving mammals display both diving hnmans (268). In fact, data from animals forced to slow-wave sleep (SWS) and rapideyemovement (REM) dive suggest that all three factors make important contri- sleep when on land and in water (76, 294, 339). However, butions, because the lungs of diving animals that have the periods of REM sleep are usually reduced when in water died during submergence often do not contain water. (76, 293) and, at the extreme, to the extent that REM sleep C:onsiderable behavioral evidence suggests that central has never been observed in gray seals and boLllenoseed dol- and peripheral chemoreceptor; are important in controlling phins in water (293, 339). In terrestrial mammals, change in the length of the breath-hold. Breathing hypoxic or hyper- heart and breathing periodicities are associated with sleep oxic gas mixtures before natural diving shortened and state. In REM sleep, breatkhg is irregular in harp seal pups lengthened the normal dive tie, respectively, in redhead (267), is regulas in gray seak (339), and does not occur ducks (151). Breathing hyperoxic gas caused Weddell seals at all in elephant seal pups (76, 287). Elephant seal pups, to make unusually long dives, one for 87 rnin (226). However, submerged in shallow water, surface to breathe while sleep- hypoxic gas mixtures, down to a Po, of 5.3 kPa, did not ing (76). On the other hand, Caspian seals (P. caspica) afiect, the diving patterns of Weddell seak nor did it signifi- arnnse to sruface and breathe (292). The northern fur seal cantly reduce dive time, although dive times for these ani- sleeps with its nostrils above the water surface but with mals, even after breathing air, were unusually short (317). only one brain hemisphere at a time showing SWS (292). Both male and female harbor seals markedly reduced dive Unihemispheric sleep also occurs in the cape fur seal (Arcto- tie when inspired Po, was ~13.3kPa, refusiig to dive cephalus pusillus) (266) and, par excellence, in the bottle- when inspired Po, fell below 8 kPa (94). Siarly, hypoxic nosed dolphin (292). Slow-wave sleep can occur in one hooded seals also reduced dive time (318). hemisphere for over 2 h. According to Mukhametov and c* Even in the manatee, which seems exceptionally in- workers (292, 293, 310), unihemispheric SWS and the lack sensitive to hypoxia, there is an association between of REM are adaptations that allow the dolphin to swim breathing hypoxic gases before diving and dive time (162). continuously and to breathe. The manatee, however, is unusual among mammalian divers in that partial pressures of alveolar C02at end dive VII. CONCLUDING COMMENTS are ~nuchhigher Lkran Ulose in phocids divi~~gfor Lhe sanlr For thousands of years, the recorded and pictorial length of time (162). In fact, in Gallivan's (159) earlier histories of widespread civilizations and cultures have at- July 1897 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 889 tested to the remarkable feats of those animals that ex- dell seals. The vagaries of calculating ADL have already tend their territory to environments which appear to be been discussed, but one of the Iu~lllarr~entaldifferences totally hostile. In life, these animals are largely hidden between calculated ADLs and experimental ADLs based from scrutiny, and early art, myth, and ceremony at- on the measurement of blood lactate concentration is that, tempted to portray the invisible aspects of their behavior. for the former, the assumption is made that all of the Consequent,ly,the large diving vert,ebrates, being so well wable oxygen is exhausted, whereas this cannot be the adapted to the aquatic environment as to be almost un- case in the latter situation. Weddell seals are able to re- seen, have lrequenLly been invested with supernatural main submerged for periods two to three times the dura- powers. Early scientiic endeavor, in the form of observation of their experimental ADL, which would not be possi- tion and analysis of visible behavior, merely empowered ble if they had used all of the oxygen available to them the hunter-gatherers who literally followed the animals' at ADL. life cycle. Determined, empirically based scientific en- We suggest, therefore, that the dive duration up to quiry over the last 100 years, although narrow and fo- which there is no increase in postdive blood lactate con- cused, has not abated the fascination and allure of diving centration is renamed the diving lactate threshold (DLT) vertebrates, whereas overexploitation and greed have and that ADL determined by a calculation of total usable added rarity and the need for conservation to our con- oxygen stores (plus PCr stores?) and rate of oxygen sciousness. The need for continuing innovation and effort, (ATP?) utilization is called calculated ADL (cADL). This both technically and financially, to reveal the mysteries may lead to some confusion at first but, because these of diving birds and mammals is obvious. The path forward terms more accurately reflect the differences between the is illdefined but veers toward concerted field studies of two values, it is hoped that they will add some clarity to the type pioneered by Kooyman and Zapol and their many what has become a somewhat confusing situation. One co-workers. The role of all the wonderful reflexes, uncov- important point to note is that, if the estimates of total ered in laboratory studies over the past 60 years, in con- usable oxygen stores and rate of oxygen utilization during tributing to the success and survival of diving birds and a dive are correct, cADL should be greater than DLT for mammals in nature remains obscure. any particular species. By incorporating in this review data on the behavior Despite these criticisms, cADLs have provided osefiil of aquatic birds and mammals when free ranging in their insights into the degl-ee of adaptations of various species natural environment, it is evident that it is not only the of aquatic birds and mannnals to their submerged exh- extent of their physiological and biochemical capabilities tence. It would appear that many of these are able to that determine their diving behavior. The behavior oftheir metabolize aerobically when diving at approximately the prey and the physical characteristics of their environment same (if not greater) rate as when they are at the surface, (e.g., depth of the water column) are also important. e.g., ducks, smaller penguins, fur seals, and Weddell seals. Equally, it is clear from the behavior of some animals In other words, their enhanced oxygen stores are able to (e.g., female elephant seals) that we have insufficient support aerobic metabolism at what would not be consid- knowledge of their physiological and biochemical adjust crcd to bc unusually low lcvcls for thc duration of thc ments to enable us to explain fully how they are able to dives. There arc probably circulatoly rcadjustmcnts to perform such behavior. ensure that the oxygen stores are managed judiciously, Recent developments in technology have provided and it has to be said, we are not too clear what the details much detailed information on the behavior and ecology of these readjustments might be. On the other hand, some of these fascinating and important animals. Not surpris- species, such as the larger penguins, South Georgian shag, ingly, the advances in technology have been insufficient and female elephant seals, exceed their cADL by such a to enable physiologists to obtain anything like as much large extent that the general consensus is they must either detail on the metabolic rate and physiological adjustments be reducing their aerobic metabolic rate when diving (cf. that occur during the time that the animals are submerged Ref. 36) andlor producing ATP, at least partly, by anaero- underwater. This has led to much speculation and calcula- biosis and metabolizing the lactic acid when at the surface tions based on many assumptions to explain the observed (although this is hardly likely in the case of the female behavior. As indicated in section I, the total usable body elephant seals). oxygen stores and the rate at which they are used are the An indication that these animals do indeed reduce basic aspects of any explanation. Together, these have aerobic metabolism when diving is the recent evidence been used by many authors to calculate what Kooyman from data loggers that, in at least some parts of the body, et al. (220) deiined as the ADL based on the measurements there is a reduction in temperature. Whether t,his is t,he of lactate concentration in the blood of Weddell seals cause of or the result of a reduction in metabolism re- upon surfacing at the end of a dive. However, the calcu- mains to be seen, but it indicates the need for inany Inure lated versions are not the same as the original values physiological data from these animals in their natural en- obtained by Kooyman and co-workers (235, 220) for Wed- vironment. We do not begin to understand the subtleties 890 PATRICK J. BUTLER AND DAVID R. JONES Volume 77 of the cardiovascular adjustments in these animals associ- Nachtigall and A. Wisser. Stuttgart, Gemany: Fischer, 1996, p. 151- 176, ated with feeding, digestion, thermoregulation, locomo- 13. BARTHOLOMEW, G. A,, Jx. Body temperature and respiratory and tion, etc., while underwater, let done how they are con- 11eul rates ill lhe nolthern elephant seal. J. Mammal. 35 211-218, trolled. The unusually low drag coefficients determined 1954. 14. BAUMANN, F. H., AND R. BAUMANN. A comparative study of the in penguins and cetaceans and the discovery that elephant respiratoly properties of bird blood. Respir ~h~~i~i.31: 333 843, seals and dolphins may glide for a substantial proportion 1977. of the descent of their dives are other explanations 1" BERGEY, M., AND H. BAIER. Lung a~eclranicalpruperties in the West India Malatee (Trichechus nlanatus). Rqi?: Physiol. 68. for the lower than expected metabolic rates postulated 63-75, 1987. for some aquatic birds and mammals during diving. 16. BEVAN, R. M., I. L. BOYD, P. J. BUTLER, K R. REID, A. J. WOAKES, AND J. P. CROXALL. Heart rates and abdominal tempera- We are grateful to Jackie Hanis, Pauline Hill, Than Ngu- tures of free-ranging South Georgian shags, Phalacrocorm geor- yen, A. M. A. Lacombe, and Roger and Pauline Jones for assis- gian,us. J. E.w. Biol 200 661-675, 1997. tmce in prep~ngthis ,.,,view. ~i~h~~dB~~~,~~~~ll~~d~~~~, 17. BEVAN, R. M., AND P. J. BUTLER. Cardiac output and blood flow and Pierre Siwore read drafts of the manuscript and provided distxibution dwng swimming and voluntary diving of the tufted duck

REFERENCES line the feedine behaviour of aauatic birds: heart rate and oxveen I. AKSELROD, S., D. GORDON, J. B. MADWED, N. C. SNIDMAN, cokumption during dives of different duration. J Em. ~io1.162: D. C. SHANNON, AND R. J. COHEN. Hernodynamic regulation: in- 91-106, 1992. 23. BEVAN, R. M.,A. J. WOAKES, P. J. BUTLER, AND J. P. CR0XN.L. vestiaation bv s~ectralanalysis. Am. J Phusiol." 249 (fieart ~irc. Heart rate and oxygen consumption of exercising gentao penguins. P~.~J.?~oI.~--~-~--,18~"~~7-~~75. ---~ . 1k5. .. . . 2. ANDERSEN, H. T. The reflex nature of the physiological adjust- Physiol. Zool. 68:855-877, 1995. menb and their &rent AC~phwiol. sand. 58: 263- 24. BLACKWELL, S. B., ANU B. J. LE BOEUF. Developmental aspects 273, 1963. of sleep apnoea in nolthem elephant seas, Mirounga angustir- 3. ANDERSEN, H. T. Physiological adaptations in diving vertebrates. ostris. J. Zool. Lond. 231: 437 ,447, 1093. Phusiol. Rev. 40: 212-243. 1966. 25. BLAKE, K. W.fish Locomotion. Cambridge, UK:Cambridge Univ. 4. ANDREWS, R. D., D. R. JONES, J. D. WILLIAMS, D. E. CROCKER, Press, 1983. D. P. COSTA, iwn 8. J. LE BOEUF. Metabolic and cardiovascular 26. BLIX, A. S. Creatine in diving animals: a comparative study. Cos~p. adjustments to diving in northern elephant seals (Mimunga an- Biochem. Physiol. 40: 805-807, 1971. gustimstris) (Abstract). Fhysiol. Zool. 68: 105, 1995. 27. BLIX, A. S., R. ELSNER, AND J. K. KJEKSHUS. Cardiac output and 5. ANDREWY, R. D.,D. R. JONES, J. D. WILLIAMS, P.H. THUHSON, its distribution through cnpill&es and AT shunffi in dlving seals. Acta Physiol. Scand. 118: 109-116, 1983. G. W. OLIVER. D. P. COSTA. AND B. J. LE BOEUF.~ ~ Heart rates of~ ~ NorUtrnn elephant seals while diving at sea and resting on the 28. BLIX, A. S., AND B. FOLKOW. Cardiovascular adjustments to diving beach. J. Exp. Biol. In press. In mammals and birds. In: Handbook ofphysiology nLe Ca~diovas- 6. ANDREWS, R. D., J. D. WLLIAMS, D. R. JONES, AND B. J. LE nlar System. Pm'pheral Cimlation and Organ Blood Row. BOEUF. Heart rate responses to apnea on land and at sea in Now- Bethesda, MD: h11. Pl~ysiol.Soc., 1983, sed. 2, vul. 111, pl. 2, cilapl. em elephant seals (Abstract). Am. Zool. 32: 314 1992. 25, p. 917-945. 7. ANGELWAMES. J. E.. M. DE B. DALY. AND R. ELSNER. Arterial 29. BLIX, A. S., AND L P. FOLKOW. Daily cncrgy cxpcnditure in free baroreceptor reflexes in the seal and their modification during ex- living minke whales. Acta Physiol. Scand. 153: 61-66, 1995. perimental dives. Am. J Physiol 234 (Heart Circ Phgsiol. 3): 30. BLIX, A. S., J. K. KJEKSHUS, I. ENGE, AND A BERGAN. Myocardial H730-H739, 1978. blood flow in the diving seal. Acta Physiol. Scand. 96: 277-280, 8. ANGELWAMES, J. E., R. ELSNER, ANU M. DE B. DALY. Lung ..1976~ . .. inflation: effects on heart rate, respiration, and vagal afferent activ- 31. BOGGS, D. F., P. J. BUTLER, AND M. A. WARNER. Fluctuations in ity in seals. Am. J. Physiol. 240 (Heart Cire. Physiol. 9): H190- differential pressure between the anterior and posterior alr sacs of H198, 1981. tufted ducks, Agthya fnligula, during breathhold dives. Physiolo~ 9. ARTHUR, P. G., M. C. HOGAN, D. E. DEBOUT, P. D. WAGNER, oist 39: 27. 1996. AND P. W. HOCHACHKA. Modeling the effects of hypoxia on ATP 32. BOUVEROT, P. Control of breathing in birds compared with rnm~ turnover in exercising muscle. J. Appl Physiol. 73: 737-742, 1992. rnals. Physiol. Rev. 58: 604-655, 1978. 10. BAMFORD, 0. S., AND D. R. JONES. On the initiationof apnoeaand 33. BOYD, 1. L., T. ARNBOM. AND M. FEDAK. Water flux, body compo- some cardiovascular responses to submergence in ducks. Respir sition, and metabolic rate during malt in female southemelephkt Physiol. 22: 199-216, 1974. seals (Mirounga leonina). Physiol. Zool. 66: 43-60, 1993. 11. BANNASCH, R. Hydrodynamics of ~enguins:an experimental ap- 34. BOYD. I. L.. J. P. Y. ARNOULD, T. BARTON, AND J. P. CROXALL. proach. In: '77~Penguin, edited by P. Dann, 1. Norman, and P. Foraging behaviour of Antarctic fur seals during periods of con- Reilly. Australia: Surrey Beatty, 1995, p. 141-176. lrastirtg prey abundance. J. Anim. Ecol. 63: 703-713. 1994. 12. BANNASCH, R. Widerstandsame svemungskelper: optimalfor- 35. BOYD, I. L., AND J. P. CROXALL. Diving behaviour of lactating Ant- men nach patenten der natur. In: BIONA: Report 10, edited by W. arctic fur seals. Can. J Zool. 70: 919-928, 1992. July 1997 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 891

36. BOYD, I. L., AND J. P. CROXALL. Dive durations in pinnipeds and 59. BUTLER, P. J., AND A. J. WOAKES. Changes in heart rate and respi- seabirds. Can. J. Zool. 74: 1696-1705, 1996. ratory frequency during natural behaviour of ducks, with particular 37. BOYD, I. L., K. REID, AND R. M. BEVAN. Swimming speed and allo- reference to diving. J. Exp. Biol. 79: 283-300, 1979. cation of time during the dive cycle in Antarctic fur seals. Anim. 60. BUTLER, P. J., AND A. J. WOAKES. Telemetry of physiological vari- Behav. 50: 769-784, 1995. ables from diving and flying birds. Symp. 2001. Sot. Lond. 49: 107- 38. BOYD, I. L., A. J. WOAKES, P. J. BUTLER, R. W. DAVIS, AND T. M. 128, 1982. WILLIAMS. Validation of heart rate and doubly labelled water as 61. BUTLER, P. J., AND A. J. WOAKES. Control of heart rate by carotid measures of metabolic rate during swimming in California sea lions. body chemoreceptors during diving in tufted ducks. J. Appl. Phys- Funct. Ecol. 9: 151-160, 1995b. iol. 53: 1405-1410, 1982. 39. BRIX, O., S. G. CONDO, G. LAZZARINO, M. E. CLEMENTI, R. SCA- 62. BUTLER, P. J., AND A. J. WOAKES. Heart rate and aerobic metabo- TEN.A, AND B. GIARDINA. Arctic life adaptation.111. The function of lism in humboldt penguins, Spheniscus humboldti, during volun- whale (Balaenop tera acu toras tra ta) hemoglobin. Comp. Biochem. tary dives. J. Exp. Biol. 108: 419-428, 1984. Physiol. B Bichem. Mol. Physiol. 4: 139-142, 1989. 63. BUTLER, P. J., A. J. WOAKES, I. L. BOYD, AND S. KANATOUS. Rela- 40. BRIX, O., M. EKKER, S. G. CONDO, R. SCATENA, M. E. CLEM- tionship between heart rate and oxygen consumption during ENTI, AND B. GIARDINA. Lactate does facilitate oxygen unloading steady-state swimming in California sea lions. J. Exp. Biol. 170: from the hemoglobin of the whale, Balaenoptera acutorostrata, 35-42, 1992. after diving. Arct. Med. Res. 49: 39-42, 1990. 64. CARBONE, C., AND A. I. HOUSTON. Patterns in the diving behav- 41. BROBECK, J. R., AND A. B. DUBOIS. Energy exchange. In: Medical iour of the pochard Aythya ferina: a test of an optimality model. Physiology, edited by V. B. Mountcastle. St. Louis, MO: Mosby, Anim. Behav. 48: 457-465, 1994. 1980, p. 1351- 1365. 65. CARNEIRO, J. J., AND D. E. DONALD. Blood reservoir function of 42. BROOKS, G. A., C. M. DONOVAN, AND T. P. WHITE. Estimation of dog spleen, liver and intestine. Am. J. Physiol. 232 (Heart Circ. anaerobic energy production and efficiency in rats during exercise. Physiol. 1): H67-H72, 1977. J. Appl. Physiol. 56: 520-525, 1984. 66. CASSON, D. M., AND K. RONALD. The harp seal, Pagophilus groen- 43. BRYDEN, M. M., AND G. H. K. LIM. Blood parameters of the south- landicus (Erxleben, 1777). XIV. Cardiac arrythmias. Comp. Bio- ern elephant seal (Mirounga leonina, Linn) in relation to diving. them. Physiol. A Physiol. 50: 307-314, 1975. Comp. Biochem. Physiol. 28: 139-148, 1969. 67. CASTELLINI, J. M., AND M. A. CASTELLINI. Estimation of splenic 44. BURNS, J. M., AND M. A. CASTELLINI. Physiological and behavioral volume and its relationship to long-duration apnea in seals. Physiol. determinants of the aerobic dive limit in Weddell seal (Leptony- 2001. 66: 619-627, 1993. chotes weddellii) pups. J. Comp. Physiol. B Biochem. Syst. Envi- 68. CASTELLINI, M. A. Visualizing metabolic transitions in aquatic ron. Physiol. 166: 473-483, 1996. mammals: does apnea plus swimming equal “diving”? Can. J. 2001. 45. BUTLER, P. J. Respiratory and cardiovascular control during diving 66: 40-44, 1988. in birds and mammals. J. Exp. Biol. 100: 195-221, 1982. 69. CASTELLINI, M. A. The biology of diving mammals: behavioural, 46. BUTLER, P. J. The exercise response and the “classical” diving physiological, and biochemical limits. In: Advances in Compara- response during natural submersion in birds and mammals. Can. tive and Environmental Physiology, edited by R. Gilles. Berlin: J. 2001. 66: 29-39, 1988. Springer-Verlag, 1991, p. 106- 134. 47. BUTLER, P. J. Metabolic adjustments to breath holding in higher 70. CASTELLINI, M. A. Apnea tolerance in the elephant seal during vertebrates. Can. J. 2001. 67: 3024-3031, 1989. sleeping and diving: physiological mechanisms and correlations. 48. BUTLER, P. J. Exercise in birds. J. Exp. Biol. 160: 233-262, 1991. In: Elephant Seals: Population Ecology, Behavior and Physiology, 49. BUTLER, P. J. Respiratory adaptations to limited oxygen supply edited by J. Burley, B. J. Le Boeuf, and R. M. Laws. Berkeley: Univ. during diving in birds and mammals. In: Physiological Strategies of California Press, 1994, p. 343-353. for Gas Exchange and Metabolism, edited by A. J. Woakes, M. K. 71. CASTELLINI, M. A. AND J. M. CASTELLINI. Influence of hematocrit Grieshaber, and C. R. Bridges. Cambridge, UK: Cambridge Univ. on whole blood glucose levels: new evidence from marine mam- Press, 1991, p. 235-257. mals. Am. J. Physiol. 256 (Regulatory Integrative Comp. Physiol. 50. BUTLER, P. J. To what extent can heart rate be used as an indicator 25): R1220-R1224, 1989. of metabolic rate in free-living marine mammals? Symp. 2001. Sot. 72. CASTELLINI, M. A., D. P. COSTA, AND A. HUNTLEY. Hematocrit Lond. 66: 317-332, 1993. variation during sleep apnea in elephant seal pups. Am. J. Physiol. 51. BUTLER, P. J., R. M. BEVAN, AND A. J. WOAKES. The energetics 251 (Regulatory Integrative Comp. Physiol. 20): R429-R431, 1986. of free-ranging penguins using the heart rate technique. Proceed- 73. CASTELLINI, M. A., R. W. DAVIS, AND G. L. KOOYMAN. Blood ings of the 5th European Conference on Wildlife Telemetry, edited chemistry regulation during repetitive diving in Weddell seals. by Y. Le Maho and T. Zorn. Strasbourg, Germany: Centre National Physiol. 2001. 61: 379-386, 1988. de la Recherche Scientifique, 1997. 74. CASTELLINI, M. A., R. W. DAVIS AND G. L. KOOYMAN. Annual cy- 52. BUTLER, P. J., R. M. BEVAN, A. J. WOAKES, J. P. CROXALL, AND cles of diving behavior and ecology of the Weddell seal. In: Bulletin I. L. BOYD. The use of data loggers to determine the energetics of the Scripps Institution of Oceanography, edited by C. S. Cox, and physiology of aquatic birds and mammals. Brax. J. Med. Biol. G. L. Kooyman, and R. H. Rosenblatt. Berkeley: Univ. of California 28: 1307- 1317, 1995. Press, 1992, vol. 28, p. l-54. 53. BUTLER, P. J., AND D. R. JONES. The comparative physiology of 75. CASTELLINI, M. A., G. L. KOOYMAN, AND P. J. PONGANIS. Meta- diving in vertebrates. In: Advances in Comparative Physiology and bolic rates of freely diving Weddell seals: correlations with oxygen Biochemistry, edited by 0. E. Lowenstein. New York: Academic, stores, swim velocity and diving duration. J. Exp. Biol. 165: 181- 1982, vol. 8, p. 179-364. 194, 1992. 54. BUTLER, P. J., AND R. STEPHENSON. Physiology of breath-hold 76. CASTELLINI, M. A., W. K. MILSOM, R. J. BERGER, D. P. COSTA, diving: a bird’s eye view. Sci. Prog. 71: 439-458, 1987. D. R. JONES, J. M. CASTELLINI, L. D. REA, S. BHARMA, AND M. 55. BUTLER, P. J., AND R. STEPHENSON. Chemoreceptor control of HARRIS. Patterns of respiration and heart rate during wakefulness heart rate and behaviour during diving in the tufted duck (Aythya and sleep in elephant seal pups. Am. J. Physiol. 266 (Regulatory fuligula). J. Physiol. Lond. 397: 63-80, 1988. Integrative Comp. Physiol. 35): R863-R869, 1994. 56. BUTLER, P. J., AND E. W. TAYLOR. The effect of hypercapnic hyp- 77. CASTELLINI, M. A., B. J. MURPHY, M. FEDAK, K. RONALD, N. oxia, accompanied by different levels of lung ventilation, on heart GOFTON, AND P. W. HOCHACHKA. Potentially conflicting meta- rate in the duck. Respir. Physiol. 19: 176-187, 1973. bolic demands of diving and exercise in seals. J. Appl. Physiol. 58: 57. BUTLER, P. J., AND E. W. TAYLOR. Factors affecting the respira- 392-399, 1985. tory and cardiovascular responses to hypercapnic hypoxia, in mal- 78. CASTELLINI, M. A., L. D. REA, J. L. SANDERS, J. M. CASTELLINI, lard ducks. Respir. Physiol. 53: 109-127, 1983. AND T. ZENTENO-SAVIN. Developmental changes in cardiorespira- 58. BUTLER, P. J., AND A. J. WOAKES. Changes in heart rate and respi- tory patterns of sleep-associated apnea in northern elephant seals. ratory frequency associated with natural submersion of ducks (Ab- Am. J. Physiol. 267 (Regulatory Integrative Comp. Physiol. 36): stract). J. Physiol. Lond. 256: 73P-74P, 1976. R1294-R1301, 1994. 892 PATRICK J. HUTLEH AND DAVID R. JONES Volume 77

79. CASTELLINI, M. A, AND G. N. SOMERO. Buffering capacity of ver- 101. CULIK, B. Diving heart rates in Adelie penguins (Pygoscelis ade- tebrate i~kuscle:cuneltatiu~ts will, pukcdiala fur arrarrulril: Iuncliu~t. liw). CVI~.B~VO~BIIL. Pl~ysi~l. A Pkysiol. 102: 487-490, 1992. J. Comp. Physiol. 143: 191-198, 1981. 102. CULlK B. M., K. PUTZ. R. P. WILSON. D. ALLERS. J. LAGE. C. A. 80. CASTELLINI. M. A. G. N. SOMERO. AND G. L. KOOYMAN. Glvco- BOST AND Y. k ~0.Divine enereetics in kine okneuins (~otm-. . lacenme activities in tissues of marhe~ and~ t,errestrial~ ~~ ~ mammals.~~~~~~~ ~~~~ odytes patayonicus). 'Em. ~iol.j99: 973-9~$,'1996. ihh~siol."~oil.54: 242-252, 1981. 103. CULIK, B., AND R. P. WILSON. Energetics of under-water swimming 81. CHAPPEI.1,. M. A., V.H. SHOEMAKER, D. N. .FANES, S. K. MALO- in Adblie ~enguins(figoscelis o&l,iae). .I Cmnn P)>;o.vi,ol.R Rio- NEY AND T. L. BUCHER. Diving behavionr during foraging in breed- chem. Syst Envimn. Pkysiol. 161: 285-291, 1991. ine AdClic ocneuins. Ecoloou 74: 120441215, 1093. 104. CULIK, B.M., R. M. WILSON, AND R. BANNASCH. Underwater swimming at low energetic cost by pygoscelid penguins. J. Em. Biol. 197: 65-78, 1994. penguins. Ecolugy 74: 2460-2461, 1993. 106. DALY, M. DE 5. Breath-hold diving: mechanisms of cardiovascular 83. CHU. K. C. Dive times andventilation oatterns of sineine-- hnmoback . adjustments in the mammal. In: Recmt Advances in Physiology, whales (Megaplera novaeangliae). Can. J Zool. 66: 1322-1327, edited...... by P. F. Baker. Edinburgh: Churchill Livingstone, 1984, p. 1988 2U1-245. 84. CLAUSEN, G., AND A. ERSLAND. The respiratory properties of 106. DALY, M. DE B., R. ELSNER, AND J. E. ANGELLJAMES. Cardiore- Lhtl blood uf the bladden~useseal (Cystophwa cristatu). Respir spiratory control by carotid chemoreceptors during experimental Physiol. 7: 1-6, 1060. dwes in the seal. Am. J Physhl. 232 [Heart Circ. Ph'hysi~l.1): 85. CLAUSEN, G., AND A. ERSLAND. Blood Or and acid-base changes H608-H516, 1977. in the beaver during submersion. Respzr Pkysiol. 11: 104-112, 107. DAVIS. M. B.. AND H. GUDERLEY. Enerev metabolism in the loco- 1970. ,,,,>l<,r,,,,I.,< I,> <,idl? CO,Il,"<," "18,r"- (/ '*?" ,, >1q.. a,,, lv37 Scotland. VI. Divine times and hunting success of otters iI.vfm 105 nh\lF, hl. R, \Xr. ll. ivlIIERLE1 Kl~r.ht~nr~caladalrnt~~o~ns tu dl!- lutm) at Dinnet ~ochhs,Aberdeenshire &d in Yell Sound, hitl land. ing m the common murre, Un:a oalge, and the atlantic puffin, Fru- .I Zool. Lond. 209: 341-346, 1986. tercula arctica. J. Em. Biol. 253: 235-244, 1990. 87. COOPER, J. Diving patterns of eomorantsPhalamcomeidac, Ibis 109. DAVIS, R. W., M. A. CASTEWNI. G. L. KOOYMAN ANn R. MAIJE. 128: 562-570, 1986. Renal glomerular tiltration rate and hepatic blood flow during vol- XU. COSTA, D. P. Methods for studying the energetics of freely diving unmy diving in Weddell seals. Am. J. Physiol. 245 (Regulatov animals. Can. J. Zool. 66: 45-52, 1988. Integrative Cow.Phyaiol. 14): R743 R748, 1983. 89. COSTA, D. P., G. A. ANTONELIS, AND R. L. DELONG. Effects of El 'I0. R. W., M. A. CASTELLIN[, T. M. AND G. L. NiAo on the foraang energetics of the California sea lion. In: EjJecls MAN. Fuel horneosiasis in the harbor seal during submerged swim- ofEl Nilio on Pinnineds. Resaonses to Environmental Stress. ed- ming. J. Comp. Plcysiol. B Bioclcmr~.Syst Olluirun Pl~ysivL160: iied hy F. ~rillrnich-andK. &o. Berlin: Springer-Verlag, 1991, p~ 627-635, 1991. 111. DAVIS, R. W., J. P. CROXALLANDM.J. O'CONNELL. The ~prodnc- A16iiLlA4 A - - . COSTA, D. P., J. P. CROXALL, AND C. D. DUCK. Foraging energetics tive energetics of gentoo (Pygoscelis pupua) and macaroni (Eu- of antarctic fur seals in relation to changes in prey availability. dyptes chrysolopkus) penguins at South Georgia. J Anim. Eeol. Ecology 70: G9G-COG, 1989. 58: 59-74, 1989. DAVIS, R. W., T. M. WILLIAMS AND L. KOOYMAN. Swimming COSTA, D. P., AND R. L. GENTRY. Fi-eerangingenergetics of north- G. em fur seals. In: Fur Seals; Maternal Strategies on Land and at metabolism of yearling and adult harbor seals Phoca vitulim. Physiol. ZooL 58: 520-526, 1985. Sea, edited by R. L. Gentry and G. L. Kooyrnan. Princeton, NJ: Plinceton Univ. Press, 1986, p. 79-101. DELONG, R. L., AND G. A ANTONELIS. Impact of the 1982-1983 El Nina on the northem firr spa1 population at San Miguel Island, COSTA. D. P., AND F. TRILLMICH. Mass changes and metabollsm during the perinatal fast: a comparison between antarctic (Arcto- Cdifvruia. In: Pircrcipeds crbd El Ni~k.Responses to Emiranmen- tal Stress, edited by F. Trillmich and K. Ono. Berlin: Springer-Ier~ cephalus gazella) and galapagos fur seals (A~ztocepkalusgalapu- lag, 1991, p. 75-83. ooavuis). Pkvsiol. Zool. 6: 160-16Li. 1988. DENISON, D. M., AND G. L. KOOYMAN The structure and fundion ~OSTE~O,'R.R., AND G. C. WHITTOW. Oxygen cost of swimming of the small airways in pinniped and sea otten. Respir. Physiol. in a trained California sea lion. Comp. Biockem. Physiol. APkysiol. 17. 1-ln 1071 ,~ - .., ., , 50: 645-647, 1375...... ~ 115. DENISON, D. M., D. A. WARRELL, AND J. B. WEST. Airway struc- 94. CRAIG, A. B., AND A. P~CHE.Respiratov physiology of freely ture and alveolar emptying in the lungs of sea lion and dogs. Respir, diving harbour seals (Phoca vitulina). Phasiol. Zoo1 53: 419-432,

Phusiol. 15: 253-260. 1871.~ 1980. ~, 116. DE~CAS,F., Y. MINAIRE, AND J. CHATONNET. Rates of forma- CROCKER, D. E., 5. J. LE BOEUF, Y. NAITO, T. ASAGA, AND D. P. tion of lactic acid in dogs at rest and during moderate exercise. COSTA. Swim speed and dive function in a female northern rle- Can. J Physiol Pannacol. 47: 603-610, 1969. pim~tseal. In: Elephant Seals: Population Ecology, Behavior, and 117 nEWAR, .J M. The Bird as m Div~r.London: Witherby, 1924. Physiology, edited by R J 1.e Roeuf and R M I.am Rwkeley 118. DIIINDSA, D. S., J. METCm, A. S. IIOVERSLAND, AND R. A. Llniv. of California Press, 1994, p. 328-339. HARTMAN. Comparative studies of the respiratory functions of CROLL, D. A, A J. GASTON, A E. BURGER AND D. KONNOFF. mammalian blood. X. Killer whale (Ordnus orca Linnaeus) and Fomging behavior and physiological adaptation for diving in thick^ beluga whale (Dulplci~-ptenu.laucas). Respir Pkysiol. 20: 93- billed murres. Ecology 73: 344-356, 1992. 103, 1974. CROLL, D. A,, AND E. McLAREN. Diving metabolism and thermo- 119. Dl PRISCO, G., S. G. CONODD, M. TAMBURRINI, AND B. GIAR- rrgulaliun in conl~nonand thick-billed murres. J Curr~p.Physiol. DINA Oxygen lransporl m extreme envirunments. Trer~dsBio- B Biochem. Syst. Envimn. Physiol. 163: 160-166, 1993. chem. Sci. 16: 471-474, 1991. CROLL, D. A,, M. K. NISHIGUCHI AND S. KAUPP. Pressure and 120. DOLPHIN, W. F. Ventilation and dive patterns of hunpback whales, lacwre dehydragenase function in diving mammals and birds. Phys- M~gapteranouaeangliae, on the~rAlaskan feeding grounds. Can. iol. Zoo1 65: 1022-1027, 1992. J Zoo1 65 83-00. 1087. CROW,J. P., 1. EVERSON, G. L. KOOYMAN, C. RICKETTS AND 121. DOLPHIN. W. F. ~ivebehawor and estimated energy expendltwe R. W. DAVIS. Fur seal diving behaviour in relation to venical distri- of foraging humpback whales in southeast Alaska. Can. 3. Zool. bution of luill. J Avrim. Eeol. 54: 1-8, 1985. 66:..~ ..~354-3G2 ..-, -..1987 . CROXALL, J. P., Y. NAITO, A. KATO, P. ROTHERY AND D. R. 122. DOLPHIN, W. F. Foraging dive patterns of humpback whales, Meg~ BRIGGS. Divlng pvttcrns and pcrformancc in thc Antarctic bluc- aptera novacangliae, in southevst Alvska: a cost-benefit analysis. eyed shag Phalacracoraz atnceps. J. 2001. Lond. 225: 177-199, Can. J Zoo1 66: 243-2441. 1988. 1991. 123. DORMER, K. J., M. J. DENS, AND H. L. STONE. Cerebral blood flow Jz~lyI997 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 893

in the sea lion (Zalopkus califoomianws) during voluntary dives. ming muskrat (Ondatra zibetkim). J Em. Biol. 110: 183-201, Comp. Biockem. Pkysiol. A Physiol. 58: 11-18, 1977. 1984. 124. DRAULANS, D. Foraging and size selection of mussels by the tufted 149. FISH, F. E.. AND C. k HUI. Dolphin swimming: a review. Mammal duck, Aytkyafuligula. J. Anim. Ecol. 51: 943-956, 1982. Rm. 21: 181-195. 1991. 125. DRUMMOND, P. C., AND D. R. JONES. The initiation and maintenance of bradycardia in a diving mammal, the muskrat, Ondatra zibethim. J. Pkysiol. Lond. 290: 253-271, 1979. 126. DYKES, R. W. Facton related to the dive reflex in habor seals: tors to the cardiac response to diving in restrained and unrestrained respiration, immersion, bradycardia, and lability of the heart rate. redhead ducks (Aytkya amm-icana). J. Exp. Btol. 121: 227-238, Can. J. Pkysiol. Phannacol. 52: 248-258, 1974. 1986. 127. DYKES, R. W. Factors related to the dive reflex in harbor seals: 152. FURILLA, R. A., AND D. R. JONES. The relationship between dive sensory contributions from the trigeminal region. Can. J. Physiol. and pre-dive heart rates in restrained and free dives by diving Phannacol. 52: 259-265, 1974. ducks. J. Em. Biol 127: 333-348, 1987. 128. ELDRIDGE, F. L., D. E. MIUHORN, J. P. KELLY, AND T. G. WAL 153 FIIRII.I.A, R A, aNn n R .JONES. Cardiac responses to dahhling DROP. Stimulation by central command of locomotion, respiration and diving in the mallard, Anas platyrkynclws. Phy.rio1. Zool. 60: and circulation during exercise. Respi?: Physiol. 59: 315-3'57, 1985. 406-412, I987 129. ELIASSEN, E. Cardiovascular responses to submersion asphyxia 154. GABBOTT, G. R. J., AND D. R. JONES. Habituation of the cardiac in avian divem. Arhok [Jnili. RPC~PIIM0.t.-Nntunrit~nsk. Spr 2 1- response to involuntary diving in diving and dabbling ducks. J. 76, 1960. Em. Biol 131: 403-415, 1987. 130. ELSE, P. A, AND A. J. HUBERT. Evolution of mammalian endo- 155. GABBOTT, G. R. J., AND D. R. JONES. The effect of brain transec- thermic metabolism: "leaky" membranes as a source of heat. Am. tion on the response to forced submergence in duckq. .I. Aulon. J. Pkysiol. 263 (Regulatory Integrative Comp. Pkysiol. 22): R1- ~VemSysl 36: 65-74, 1991~ R6, 1987. 156. GABRIELSEN, G. W. Free and farced diving in ducks: habituation 131. ELSNER, R. Heart rate response in forced Venus trained experi- of the initial dive response. Acta Physiol. Scand. 123: 67-72, 1985. mental dives in pinnipeds. Hvalradets Sk?. 48: 24-29, 1965. 167. GALES, R., AND 8. GREEN. Tlic ilrlnutll cr~~.rgcticscycle of liltle 132. ELSNER, R. Limitsto exercise performance: some ideas from com~ penguins (Eudyplula minor). Ecology 7: 2297-2312, 1990. oarative studles. Acla Pkwsiol. Scand. 128: 45-51. 1986. 158. GALES, R., C. WILLIAMS AND D. RITZ. Foraging behaviour of the 133. ELSNER. R., J. E. ANGE~MAMES,AND M. DE B. DALY. Carotid little pengum, EudyptUla mznol: initial results and assessment of body chemarecepton retiexes and their interactions in the seal. instlument effect. J Zool. Lond. 220: 61-85, 1990. Am. J. Pkysiol. 232 (Hmrt Cire. Physiol. I): H517VH525, 1977. 159. GALLNAN, G. J. Hypoxia and 11ypercapr"a in tile respiratoly con^ 134. ELSNER.. R.. . D. L. FRANKLIN. AND R. L. VAN CITTERS. Cardiac trol of the Amazonian manatee (Trichechws inwnguis). Pkysiol. output duling diving in an unrestrained sea lion. Nature 202: 809- Zool~~ 53:~~ 254-261.- ~ -~~~IRRO 160. GWIVAN, G. J. Ventilation and gas exhange in unresVained harp 810.~-~, -~~-1964. se& 135. ELSNER, R., AND B. GOODEN. Diving and Asphyzia. Cambridge, (Phoca grodandica). Comp Bioclwn. Physiol. A Pkysiol. UK: Cambridge Univ. Press, 1983. 69: 809-813, 1981. J., nNu R. C. of 136. ELSNER. R. W., W. N. HANAFEE, AND D. D. HAMMOND. Angiopra- 161. GALWAN, G. BEST. Metabolism and respiration phy of inferiorvena cava of the harbor seal during simulated di;ng. the Amazonian manatee (Trichechus inunrmisi.".- Phwsiol. Zool. 53: Am. J Physzol. 220: 1155-1157, 1971. 245-253, 1980. 137. ELSNER, R., D. W. KENNEDY, AND K. BURGESS. Diving bradycar- 162. GALLNAN, G. J., J. KANWISHER, AND R. C. BEST. Heart rates and dia in the trained dolphin. Nature 212: 407-408, 1966. gm cxchange in the Amvzonian manatee (Tricheckus inunguis) in relation to diving. J. Comp. Physiol. B Biockm. Syst. Environ. Physiol. 156: 415-423, 1986. Cerebral roleraice lo h\yuxemia in aspli!ulared \\'eddell seals Rc 163. GENTRY. R. L.. G. L. KOOYMAN, AND M. E. GOEBEL. Feeding and A,,,,' /'I,"<><,/ :I 2b:-Pv7, 1;s-u diving behavior of northern fur seals. In: firSea1.s: Maternal Strat- l.jB ELSUER. R D \VARTZOK. Y B SOYFRiNK, .ui. LI F' KELLY egies on Land and at Sea, edited by R. L. Gentry and G. L. Kooy- Behavioral and physiologjcal reactions of arctic seals during under- man. Princeton, NJ: Frlnceton Unlv. Press, 1986, p. 61-77. ice pilotage. Can.J Zool. 67: 2506-2513, 1989. 164. tiIAKUINA, B., M. COKUA, M. ti. PtiLLEtiKINI, S. G. CONDO, AND 140. EVANS, B. K., D. R. JONES, J. BALDWIN, AND G. R. J. GABBOTT. M. BRUNORI. Functional properties of the hemoglobin system of Diving ability of the platypus. Aust. Zool. 42: 17-27, 1994. J two diving birds (Podicms niorieollis and Phalacrocom carbo 141. FEDAK M. A. Diving and exercise in seals: a benthic pempective. In: Diving in Animals and Man, edited by A. 0. Bxubakk, J. W. Kanwisher, and G. Sundnes. Trondheim, Noway: Tapir, 1986, p. 11-32. interplay of the carbon dioxide and temperature in the oxygen 142. FEDAK, M. A,, M. R. PULLEN AND J. KANWISHER. Circulatory re- unloading. Axt. Med. Res. 49: 93-97, 1990. soonses of seals to veriodic breathing: hean rate and breaIhing 166. GILBERI', F. K, AND N. Whrl'VN. Heart rate values far beaver, during exercise and diving in the laboratory and open sea. Can. j mink and muskrat. Comp. Biochem. Physiol. 73: 249-251, 1982. Zool. 66: 53-60, 1988. 167. GILBERT, F. F., AND N. GOFPON. Terminal dives in mink, muskrat 143. FEDAK, M. A,, AND D. THOMPSON. Behavioural and physiological and beaver. Pkysiol. Bekav. 28: 835-840, 1982. outions in divine seals. Swmn.". Zoo1 Soc. Lond. 66: 333-348. 1883. 168. GUPPY, M., R. D. HILL, R. C. SCHNEIDER, J. QVIST, G. C. LIG- 144. F'ELDKAMP, S. ti. Swimming m the Calrfornlasea hon: morphomet- GINS. W. M. ZAPOL AND P. W. HOCHACHKA. Microcomvuter~as rics, drag and energetics. J Em. Biol 131: 117-135, 1987. 145. FELDKAMP, S. D., R. L. DELONG AND G. A. ANTONELIS. Diving oatterns of California sea lions., Zalonhus. culifomianz~s.Can. J. , ~ool.67: 872-883, 1989. 169. GUYTON, G. P., K.S. STANEY R. C. SCHNEIDER, P. W. HO- 146. FELDKAMP. S. D., R. L. DELONG AND G. A ANTONELIS. Effects CHACHKA, W. E. HURFORD, D. K ZAPOL, G. C. LIGGINS, AND of El Nitio 1983 on the foraqinp;. .. patterns of California sea lions W. M. ZAPOL. Myoglobin saturation in freediving Weddell seals. (Zalophus califomianus) near San Miguel Island, California. In: J. Appl. Physiol. 79: 1148-1155, 1995. Pinnipeds and El Nirio. Responses lo Environmrnfnl Stress, edited 170. HAMMOND, D. D., R. EISNER, G. SIMISON, AND R. HUBBARD. by F. Trillmich and K. Ono. Berlin: Sprhger-Verlag, 1991. p. 146- Submersion bradycardia in the newborn elephant seal Mhunga 155 angustirostris. Am. J. Phgsiol. 216: 220-222, 1969. 171. HANCE, A. J., E. D. ROBIN, J. B. HALTER, N. LEWISTON, D. A. ROBIN, L. CORNELL, M. CALIGIURI, AND J. THEODORE. Hor- monal changes and enforced diving in the harbor seal Phoea vitul- 894 PATRICK J. BUTLER AND DAVID R. JONES Volume 77

ma. 11. Plasma catecholamines. Am. J. Phvsiol. 242 (Reaulaton,. . GUYTON, K. S. STANEK, I). G. ZAPOL, G. C. LIGGINS, AND W. M. Inltqr~tiite1 'g o,p l'hyqic,f I I) R52A-It33". 1982 ZAPOL. Splenic contraction, catecholamine release, and blwd vol- I l.ARI1,Y R \I. RE\AI\. J-R CK\RR:\SSIN, P.I. RITLER ume redistribution during diving in the Weddell seal. J. Appl. Wys- ti l'l'l"/., .\..I \Ul.\KKS, I. l..\l;K, jhl Y. l.b:hl.\llt!. lly~~c~~l,t'n1~i.3 Wl. 80: 298-306, 1996. Ln foraglng klng penguins. Nature In press. 195. IRVING, L. On the ability of warmblooded animals to survive with- 172. HANNON, J P., C A BOSSONE, mu W. G. RODKEY Splenic red out breath@ Sri. Mon 38: 422-428, 1934 cell sequestration and blood volume measurements in conscious 196. IRVING, L. Respiration in diving mammals. Physiol. Rev. 19: 112- pigs. Am. J Physiol. 248 (Regulatory 1nte.qrative Cmp. Physiol. -134 - . , -19:lQ - - - . 17): RZ93-R301, 1985. 197. IRVING, L. Elective regulation of the circulation in diving animals. 173. HARRISON, R. J., AND J. D. W. TOMLINSON. Normal and experi- In: Whdes, Dolphins, and Porpoises edited by K S. Norris. Cam- mental diving in the common seal (Phoca vilulina). Mammalia bridge, UK Cambridge Univ Press, 1966, p. 381-396. 24: 386-399, 1960. 198. IRVING, L., P. F. SCHOLANDER, AND S. W. GRINNELL. Therespira- 174. HEDRICK, M. S., D. A. DUFFIELD, AND L. H. CORNELL. Blood vis- tion of the porpoise Tunsiops truncatus. J. Cell. Comp. Physiol. cosity and optimal hematocrit in a deepdiving mammal, the north- 17: 145-167, 1941. em elephant seal (Mimunga angustimstris). Can. J. Zool. 64: 199. ISAACKS, R. E., AND D. R. HARKNESS. Erythrocyte organic phos- 2081-2085, 1986. phates and hemoglobin function in birds, reptiles and fishes. Am. 175. HEIEIS. M. R. A. AND D. R. JONES. Blood flow andvolumcdrstribu- 2001. 20: 115-120. 1080. tion during forced submergence in Pekin ducks (Anas platy- 2~.JAPUNDZIC, N.. 'M.-L. GRICHOIS, P ZITOUN, D LAUDE, mu rhychas). Can. J 2001. 66: 1589-1596, 1988, J.-L. ELGHOZI. Spectral analysis of blood pressure and heat rate 176. HILL, R. D. Microcomputer monitor and blood sampler for free- in conscious rats: effects of autonomic blockers. J. Auton. Nm. divine Weddell seals. J Aml. Phusiol. 61: 15701576. 1986. Sysi 30: S1-100, 1990. 177. HILL,R. D., R. C. SCHNE~~ER,6. C. LIGGINS, A. H. SCHUETTE, 201. JENSSEN, B. M., M. EKKER AND C. BECH. Themoregulatlon in R. L. ELLIOTT, M. GUPPY, P. W. HOCHACHKA, J. QVTST, K. J. winter~acclimatizedcommon eiders (Somateria mollissima) in air FALKE AND W. M. ZAPOL. He&. rate and body temperature during and water. Cnn. .I. Zool. 67: 669-673, 19119. free diving of Weddell sealo. Am. J Pkysiol. 253 (Regulatory Inte- 202. JOBIS, F. F., AND W. N. STAINSBY. Oxidation of NADH during con grative Comp. Phgsiol. 22): R344-R351, 1987. tractions of circulated mammalian skeletal muscle. Respir Phys- 178. HINDELL, M, A, D. J. SLIP, AND H. R. BURTON. The diving behav- iol. 4: 292-300, 1968. iour of adult male and female southern elephant seals, Milimur~ya 203. JOHANSEN, K, C. LENFANT, AN" G. C. GRIGG. Respiratory prop- leonina (Pinnipedla: Phocidae). Awt. J. Zool. 39: 595-619, 1991. erties of blood and responses to diving of the platypm Ornithv~ 179. HINDELL, M. A,, D. J. SLIP, H. R. BURTON AND M. M. BRYDEN. rhynchus anatinus (Shaw). Camp Biochem. Physiol. 18: 697-606, Physiological implications of continuous, prolonged, and deep 1966. dives of the southern elephant seal (Mirozmga konina). Cnn. J. 204. JONES, D. R., R. M. BRYAN, JR., N. H. WEST, R. H. LORD, AND B. Zool. 70: 370-379, 1992. CLARK. Regional distribution of blood flow during diving in the 180. HIRSOWITZ, L. A,, K. FELL, AND J. A. TORRANCE. Oxygen affinity duck (Anas platyrhynchs). Can. J. Zool. 57: 995-1002, 1979. of avian blood. Respir Physiol. 31: 51-62, 1977. 205. JONES, D. R., H. D. FISHER, S. McTAGGART, AND N. WEST. Heart 181. HOBSON, K. A,, AND S. G. SEALY. Dlving rhythms and diumai most- rate dluing breath-holding and diving in the unrestrained harbor ing timps of pelagic connorants. Wil.?on B71.11. 97: 1 If;-119, 1985. seal (Phrlcn ~iitulinnrichardi). Cm. .I 2001. 51: 671-6110, 1973. 182. IIOCIIACIIKA, P. W. Brain, lung, and heart functions during diving 206. JONES, D. R., AND R. A. FURILLA. The anatomical, physiological, and recovery. Science 212: 5W-514, 1981. behavioral, and metabolic consequences of voluntary and forced 183. HOCHACHKA, P. W. Balancing conflicting metabolic demands of divine. In: Bivd Resairation. edited bv T. J. Seller. Boca Raton. K: exercise and diving. Fedzratior~Proc. 45: 2948-2952, 1986. C~C,"1987.vol. 2, ;. 7512i. 184. HOCHACHKA, P. W. Seals: diving to the edge of hypoxia crisis. 207. JONES, D. R., R. A. FURILLA, M. R. A. HEIEIS, G. R. J. GABBOTT. Tmns. R Soc. Can. 1: 463-469, 1990. AND F. M. SMPPH. Forced and voluntary diving in ducks: cardiovas- 186. HOCHACHKA, P. W. Metabolic biochemistry and the making of a cular austments and their control. Can. J. Zool. 66: 75-83, 1988.

mesonelaeic~~~~~~ mammal. Emerientia 48: 570-575. 1992. 208. JONES. D. R.. AND G. F. HOLETON. Cardiac outout of ducks durine .~~~~"~ ~ ~ . - 186. HOCHACHKA, P. W., J. M. CASTELLINI, R. D. HILL, R. C. SCHNEI~ diving. 'Con& Bioehe~n.Physiol. 41: 639-645, i972. UER, J. L. BENGTSON, S. E. HILL, G. C. LlGGlNS AND W. M. ZA- 209. JONES, D. R., W. K. MILSOM, AND ti. K. J. CiABBVl"C. Kole of cen- POL. Protcctivc metabolic mechanisms during liver ischemia: trans^ tral and peripheral chemoreceptors in diving responses of ducks. ferahle lessons from longdiving animals. Mol. Cell. Biocllem. 84: Am. J Physiol. 243 (Regulatory Integrative Cump. Ptcysiol. 12): 77-85. 1988. R537PR545, 1982.

~ AND Free 1R7 HOCH~CHKA.--~ ---- .~P. W.., AND R. A. FOREMAN. Phocid and cetacean 210. KANWISHER, J. W., G. CABRIELSEN, N. KANWISHER. blueprintsaf muselemetabolism. Can. J. Zoo1 71: 2089-2W8,lW5. and forced divine in birds. Science 211: 717-719. 1981. 188. HOCHACHKA, P. W., G. C. LIGGINS, G. P. GUYTON, R. C. SCHNEI- 211. KEUER, E., ANU"P. J. BUTLER. Volumes of the respiratory and DER, K. S. STANEK, W. E. HURFORD, R. K. CREASY, D. G. ZAPOL, cimulalury systems in tufted and mallard ducks. J. Em. Biol. 101: AND W. M. ZAPOL. Hormonal remJatory adiustments dmng volun- 213-220, 1982. tary diving in Weddell seals. ~ioch&. Physiol. BBkehem. 212. KEIJER, E., P. J. BUTLER, AND A. J. WOAKES. Cardiac response Mol. Physiol. 112: 361-375, 1995. to voluntary hving in tufted ducklings (Aythya juligula). J. Exp. 189. HOGAN, M. C., P. G. ARTHER, D. E. BEBOUT, P. W. HOCHACHKA Biol. 138: 195-203, 1988. AND P. D. WAGNER. Rule 01 O2 ill regulating tissue respiration in 213. KENNY, R. Breathing and heat rates of the southem elephant seal, dog muscle working in sit,^^ .J. A~ppl.Physiol. 73: 728-736, 1992. Miruunya leonina (L.). Pap. hc. R. Soc. Ihsman. 11k 21-27, 190. HOGAN. M. C.. am H. G. WELCH. Effect of altered arterial Oo 1979. tensiond on muscle metabohm in dog skeletal muscle during fa- 214. KOOYMAN, G. L. Techniques used in measuring diving capacities tiguing work. Am. J Physiol. 251 (Cell Physiol. 20): C216-C222, of Weddell seals. Polar Rec. 12: 391-394, 1965. 1986. 215. KOOYMAN, G. L. An analysis of some behavioural and physiologi- 191. HOL, R., k S. BLIX, AND H. 0.MWRE. Selective redistribution of cal characteristics related to diving in the Weddell seal. In: B

Ram.~A. P-V. Rhn. Cons. Int Emlor Mer. 169: 423-432. 1975. Washington, DC: Am. Geophys. Union, 1967, vol. 11, p. 227-261. 192. HOUSTON, A L, AND C. CARBONE. The optimal allocation of time 216. KOOYMAN. G. L. Respiratow adaptations in marine mammals. Am. dulng the diving cycle. Behov. Ecol. 3: 255-265, 1992. Zoo1 13: 457-468, 1973. 193. HUDSON, D. M., AND D. R. JONES. The influence of bod,, mass on 217. KOOYMAN. G. L. Physiology without restraint in diving mammals.

fh~~~~~ endurance~~~~~~~~~~~~~ to~~ restrained submergence in the Pekin duck. J. Mar Mam. Sci. 1: 166-178, 1985. Em. Bid. 120: 351-367. 1986. 218. KOOYMAN, G. L. Diuews Dlows, Plcyaiuloyg w?ulBzl~auior:Berlin: Springer-Vedag, 1Y8Y. July 1997 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 895

219. KOOYMAN, G. L., AND W. B. CAMPBELL. Heart rates in freely div- 241. LAPENNAS, G. N., AND P. L. LUTZ. Oxygen affinity of sea turtle ing Weddell seals, Leptonychotes weddelli. Comp. Biochem. Phys- blood. Respir. Physiol. 48: 59-74, 1982. iol. A Physiol. 43: 31-36, 1972. 242. LAPENNAS, G. N., AND R. B. REEVES. Respiratory properties of 220. KOOYMAN, G. L., M. A. CASTELLINI, R. W. DAVIS AND R. A. blood of the Gray Seal Halichoerus grypus. J. Comp. Physiol. 149: MAUE. Aerobic diving limits of immature Weddell seals. J. Comp. 49-56, 1982. Physiol. 151: 171-174, 1983. 243. LAPENNAS, G. N., AND R. B. REEVES. Oxygen affinity of blood of 221. KOOYMAN, G. L., Y. CHEREL, Y. LEMAHO, J. P. CROXALL, P. H. adult domestic chicken and Red Jungle fowl. Respir. Physiol. 52: THORSON, V. RIDOUX AND C. A. KOOYMAN. Diving behavior and 27-39, 1983. energetics during foraging cycles in king penguins. Ecol. Monogr. 244. LAVIGNE, D. M., S. INNES, G. A. J. WORTHY, K. M. KOVACS, 0. J. 62: 143-163, 1992. SCHMITZ AND J. P. HICKIE. Metabolic rates of seals and whales. 222. KOOYMAN, G. L., AND L. H. CORNELL. Flow properties of expira- Can. J. 2001. 64: 279-284, 1986. tion and inspiration in a trained bottlenosed porpoise. Physiol. 245. LAWRENCE, B., AND W. E. SCHEVILL. The functional anatomy of 2001. 54: 55-61, 1981. the delphinid nose. Bull. Mus. Comp. Anat. 114: 103-151, 1956. 223. KOOYMAN, G. L., AND R. W. DAVIS. Diving behavior and perfor- 246. LE BOEUF, B. J., D. P. COSTA, A. C. HUNTLEY, AND S. D. FELD- mance, with special reference to penguins. In: Seabirds: Feeding KAMP. Continuous, deep diving in female northern elephant seals, Biology and Role in Marine Ecosystems edited by J. P. Croxall. Mirounga angustirostris. Can. J. 2001. 66: 446-458, 1988. Cambridge, UK: Cambridge Univ. Press, 1987, p. 64-75. 247. LE BOEUF, B. J., Y. NAITO, T. ASAGA, D. CROCKER, AND D. P. 224. KOOYMAN, G. L., R. W. DAVIS AND J. P. CROXALL. Diving behavior COSTA. Swim speed in a female northern elephant seal: metabolic of Antarctic fur seals. In: Fur Seals: Maternal Strategies on Land and foraging implications. Can. J. 2001. 70: 786-795, 1992. and at Sea, edited by R. L. Gentry and G. L. Kooyman. Princeton, 248. LE BOEUF, B. J., Y. NAITO, A. C. HUNTLEY, AND T. ASAGA. Pro- NJ: Princeton Univ. Press, 1986, p. 115-125. longed, continuous, deep diving by northern elephant seals. Can. 225. KOOYMAN, G. L., R. W. DAVIS, J. P. CROXALL AND D. P. COSTA. J. 2001. 67: 2514-2519, 1989. Diving depths and energy requirements of king penguins. Science 249. LEITH, D. E. Comparative mammalian respiratory mechanics. 217: 726-727, 1982. Physiologist 19: 485-510, 1976. 226. KOOYMAN, G. L., D. H. KEREM, W. B. CAMPBELL, AND J. J. 250. LEITH, D. E. Adaptations to deep breath-hold diving: respiratory WRIGHT. Pulmonary function in freely diving Weddell seals, Lep- and circulatory mechanics. Undersea Biomed. Res. 16: 345-353, tonychotes weddelli. Respir. Physiol. 12: 271-282, 1971. 1989. 227. KOOYMAN, G. L., D. H. KEREM, W. B. CAMPBELL, AND J. J. 251. LENFANT, C. Physiological properties of blood of marine mam- WRIGHT. Pulmonary gas exchange in freely diving Weddell seals, mals. In: Biology of Marine Mammals, edited by H. T. Anderson. Leptonychotes weddelli. Respir. Physiol. 17: 283-290, 1973. New York: Academic, 1969, p. 95-116. 228. KOOYMAN, G. L., AND T. G. KOOYMAN. Diving behavior of em- 252. LENFANT, C., R. ELSNER, G. L. KOOYMAN, AND C. M. DRABEK. peror pengins nurturing chicks at Coulman Island, Antarctica. Con- Respiratory function of blood of the adult and fetus Weddell seal dor 97: 536-549, 1995. Leptonychotes weddelli. Am. J. Physiol. 216: 1595-1597, 1969. 229. KOOYMAN, G. L., K. S. NORRIS, AND R. L. GENTRY. Spout of the 253. LENFANT, C., K. JOHANSEN, AND J. D. TORRANCE. Gas transport Gray Whale: its physical characteristics. Science 190: 908-910, and oxygen storage capacity in some pinnipeds and the sea otter. 1975. Respir. Physiol. 9: 277-286, 1970. 254. LENFANT, C., G. L. KOOYMAN, R. ELSNER, AND C. M. DRABEK. 230. KOOYMAN, G. L., AND P. J. PONGANIS. Behavior and physiology of diving in emperor and king penguins. In: Penguin Biology, edited Respiratory function of blood of the Adelie penguin Pygoscelis by L. S. Davis and J. T. Darby. New York: Academic, 1990, p. 229- adeliae. Am. J. Physiol. 216: 1598-1600, 1969. 242. 255. LEVY, M. N. Sympathetic-parasympthetic interactions in the heart. Circ. Res. 29: 437-445, 1971. 231. KOOYMAN, G. L., AND P. J. PONGANIS. Emperor penguin oxygen consumption, heart rate, blood lactic acid and stroke frequencies 256. LIFSON, N., AND R. McCLINTOCK. Theory of use of the turnover during graded swimming exercise. J. Exp. Biol. 195: 199-209, 1994. rates of body water for measuring energy and material balance. J. Theor. Biol. 12: 46- 74, 1966. 232. KOOYMAN, G. L., P. J. PONGANIS, M. A. CASTELLINI, E. P. PON- 257. LIGGINS, G. C., J. QVIST, P. W. HOCHACHKA, B. J. MURPHY, R. K. GANIS, K. V. PONGANIS, P. H. THORSON, S. A. ECKERT, AND Y. CREASY, R. C. SCHNEIDER, M. T. SNIDER, AND W. M. ZAPOL. LEMAHO. Heart rates and swim speeds of emperor penguins diving Fetal cardiovascular and metabolic responses to simulated diving under sea ice. J. Exp. Biol. 165: 161-180, 1992. in the Weddell seal. J. Appl. Physiol. 49: 424-430, 1980. 233. KOOYMAN, G. L., J. P. SCHROEDER, D. G. GREENE, AND V. A. 258. LIN, Y. C. Breath-hold diving in terrestrial mammals. In: Exercise SMITH. Gas exchange in penguins during simulated dives to 30 and and Sport Science Reviews, edited by R. L. Terjung. Philadelphia, 68 m. Am. J. Physiol. 225: 1467-1471, 1973. PA: Franklin, 1982, vol. 10, p. 270-307. 234. KOOYl!vIAN, G. L., AND E. E. SINNETT. Mechanical properties of 259. LOCKYER, C., AND R. MORRIS. Observations on diving behaviour the harbor porpoise lung, Phocoena phocoena. Respir. Physiol. 36: and swimming speeds in a wild juvenile Tursiops truncatus. Aquat. 287-300, 1979. Mamm. 13: 31-35, 1987. 235. KOOYMAN, G. L., E. A. WAHRENBROCK, M. A. CASTELLINI, R. W. 260. LOPES, 0. U., AND J. F. PALMER. Proposed respiratory “gating” DAVIS AND E. E. SINNETT. Aerobic and anaerobic metabolism dur- mechanism for cardiac slowing. Nature 264: 454-456, 1976. ing voluntary diving in Weddell seals: evidence of preferred path- 261. LOWORN, J. R., AND D. R. JONES. Body mass, volume, and buoy- ways from blood chemistry and behavior. J. Comp. Physiol. 138: ancy of some aquatic birds, and their relation to locomotor strate- 335-346, 1980. gies. Can. J. 2001. 69: 2888 -2892, 1991. 236. KRAMER, D. L. The behavioral ecology of air breathing by aquatic 262. LOWORN, J. R., D. R. JONES, AND R. W. BLAKE. Mechanics of animals. Can. J. 2001. 66: 89-94, 1988. underwater locomotion in diving ducks: drag, buoyancy and accel- 237. LACOMBE, A. M. A., AND D. R. JONES. The source of circulating eration in a size gradient of species. J. Exp. Biol. 159: 89-108, catecholamines in forced dived ducks. Gen. Comp. Endocrinol. 80: 1991. 41-47, 1990. 263. LUSK, G. The Elements of the Science of Nutrition. Philadelphia, 238. LACOMBE, A. M. A., AND D. R. JONES. Role of adrenal catechol- PA: Saunders, 1919. amines during forced submergence in ducks. Am. J. Phys,iol. 261 264. LUTZ, P. L. On the oxygen affinity of bird blood. Am. J. 2001. 20: (Regulatory Integrative Comp. Physiol. 30): R1364-R1372, 1991. 187- 198, 1980. 239. LACOMBE, A. M. A., AND D. R. JONES. Neural and humoral effects 265. LUTZ, P. L., I. S. LONGMUIR, AND K. SCHMIDT-NIELSEN. Oxygen on hindlimb vascular resistance of ducks during forced submer- affinity of bird blood. Respir. Physiol. 20: 325-330, 1974. gence. Am. J. Physiol. 261 (Regulatory Integrative Comp. Physiol. 266. LYAMIN, 0. I., AND I. S. CHETYBROK. Unilateral EEG activation 30): R1579-R1586, 1991. during sleep in the cape fur seal, Arctocephalus pusillus. Neurosci. 240. LAPENNAS, G. N. The magnitude of the Bohr coefficient: optimal Lett. 143: 263-266, 1992. for oxygen delivery. Respir. Physiol. 54: 161- 172, 1983. 267. LYAMIN, 0. I., A. I. OLEKSENKO, AND I. G. POLIAKOVA. Sleep and 896 PATRICK J. BUTLER AND DAVID R. JONES Volume 77

wakefulness in Greenland seal pups. Zh. Vyssh. Nervn. Degat. Im. 291. MOSS, A. A., M. A. FRIEDMAN, AND A. BRITO. Determination of I. P. Pav. 39: 1061-1069, 1989. liver, kidney and spleen volumes by computed tomography: an 268. MACARTHUR, R. A. Aquatic thermoregulation in the muskrat (On- experimental study in dogs. J. Comput. Assist. Tomogr. 5: 12-14, datra xibethicus): energy demands of swimming and diving. Can. 1981. J. 2001. 62: 241-248, 1984. 292. MUKHAMETOV, L. M. Unihemispheric slow wave sleep in the brain 269. MACARTHUR, R. A. Brown fat and aquatic temperature regulation of dolphins and seals. In: Endogenous Sleep Substances and Sleep in muskrats, Ondatra xibethicus. Physiol. 2001. 59: 306-317, 1986. Regulation, edited by S. Inoue and A. A. Borbely. Tokyo, Japan: 270. MACARTHUR, R. A. Seasonal changes in the oxygen storage capac- Jpn. Sci. Sot. Press, 1985, p. 67-75. ity and aerobic dive limits of the muskrat (Ondatra xibethicus). J. 293. MUKHAMETOV, L. M. The absence of paradoxical sleep in dol- Comp. Physiol. B Biochem. Syst. Environ. Physiol. 160: 593-599, phins. In: Sleep ‘86, edited by W. P. Koella, F. Obal, H. Schulz, and 1990. P. Visser. New York: Gustav Fischer Verlag, 1988, p. 154-156. 271. MACARTHUR, R. A. Foraging range and aerobic endurance of 294. MUKHAMETOV, L. M., 0. I. LYAMIN, I. S. CHETYRBOK, A. A. VAS- muskrats diving under ice. J. Mammal. 73: 565-569, 1992. SILYEV, AND R. DIAS. Sleep and wakefulness in an Amazonian 272. MACARTHUR, R. A. Gas bubble release by muskrats diving under manatee. In: Sleep ‘90, edited by J. Home. Bochum, Germany: Pon- ice: lost gas or a potential oxygen pool? J. 2001. Lond. 226: 151- tenagel, 1990, p. 119- 122. 164, 1992. 295. MURDAUGH, H. V. J., J. K. BRENNAN, W. W. PYRON, AND J. W. 273. MACARTHUR, R. A., AND C. M. KARPAN. Heart rates of muskrats WOOD. Function of the inferior vena cava valve of the harbour diving under simulated field conditions: persistence of the brady- seal. Nature 194: 700-701, 1962. cardia response and factors modifying its expression. Can. J. 2001. 296. MURDAUGH, H. V. J., C. E. CROSS, J. E. MILLEN, J. B. L. GEE, AND 67: 1783-1792, 1989. E. D. ROBIN. Dissociation of bradycardia and arterial constriction 274 MACARTHUR, R. A., AND R. E. KRAUSE. Energy requirements of during diving in the seal Phoca vitulina. Science 162: 364-365, freely diving muskrats (Ondatra xibethicus). Can. J. 2001. 67: 1968. 2194-2200, 1989. 297 MURDAUGH, H. V. J., E. D. ROBIN, J. E. MILLEN, W. F. DREWRY, 275 MALLONEE, J. S. Behaviour of gray whales (Eschrichtius ro- AND E. WEISS. Adaptations to diving in the harbor seal: cardiac bustus) summering off the northern California coast, from Patrick’s output during diving. Am. J. Physiol. 210: 176-180, 1966. Point to Crescent City. Can. J. 2001. 69: 681-690, 1991. 298 MURDAUGH, H. V. J., B. SCHMIDT-NIELSEN, J. W. WOOD, AND 276 MANGALAM, H. J., AND D. R. JONES. The effects of breathing dif- W. L. MITCHELL. Cessation of renal function during diving in the ferent levels of O2 and CO2 on the diving responses of ducks (Anas trained seal. J. Cell. Comp. Physiol. 58: 261-265, 1961. platyrhynchos) and cormorants (Phalacrocorax auritus). J. Comp. 299. MURDAUGH, H. V. J., J. C. SEABURY, AND W. L. MITCHELL. Elec- Physiol. B Biochem. Syst. Environ. Physiol. 154: 243-247, 1984. trocardiogram of the diving seal. Circ. Res. 9: 358-361, 1961. 277. MARTIN, A. R., M. C. S. KINGSLEY, AND M. A. RAMSAY. Diving 300. MURPHY, B., W. M. ZAPOL, AND P. W. HOCHACHKA. Metabolic behaviour of narwhals (Monodon monoceros) on their summer activites of heart, lung, and brain during diving and recovery in the grounds. Can. J. 2001. 72: 118- 125, 1994. Weddell seal. J. Appl. Physiol. 48: 596-605, 1980. 278. MARTIN, A. R., AND T. G. SMITH. Deep diving in wild, free-ranging 301. MURRISH, D. E. Responses to diving in the dipper, Cinclus mexi- beluga whales, Delphinapterus leucas. Can. J. Fish Aquat. Sci. 49: canus. Comp. Biochem. Physiol. 34: 853-858, 1970. 462-466, 1992. 302. MURRISH, D. E. Acid-base balance in three species of Antarctic 279. McCULLOCH, P. F., AND D. R. JONES. Cortical infhrences on diving penguins exposed to thermal stress. Physiol. 2001. 55: 137-143, bradycardia in muskrats (Ondatra xibethicus). Physiol. 2001. 63: 1982. 1098-1117, 1990. 303. NAEVDAL, G. Protein polymorphism used for identification of harp 280. McCULLOCH, P. F., I. A. PATERSON, AND N. H. WEST. An intact seal populations. Arbok Univ. Bergen Mat.-Naturvitensk. Ser. 9: glutamatergic trigeminal pathway is essential for the cardiac re- 3-20, 1966. sponse to simulated diving. Am. J. Physiol. 269 (Regulatory Inte- 304. NAGY, K. A. CO2 production in animals: analysis of potential errors grative Comp. Physiol. 38): R669-R677, 1995. in the doubly labeled water method. Am. J. Physiol. 238 (Regula- 281 MCGINNIS, S. M., AND T. P. SOUTHWORTH. Therrnoregulation in tory Integrative Comp. Physiol. 7): R466-R473, 1980. the northern elephant seal, Mirounga angustirostris. Comp. Bio- 305. NAGY, K. A. The Doubly Labeled Water Method: A Guide to Its them. Physiol. A Physiol. 40: 893-898, 1971. Use. Los Angeles: Univ. of California, 1983. (Publication no. 12- 282 MEISELMAN, H. J., M. A. CASTELLINI, AND R. ELSNER. Hemo- 1417) rheological behaviour of seal blood. Clin. Hemorheol. 12: 657-675, 306. NAGY, K. A., W. R. SIEGFRIED, AND R. P. WILSON. Energy utiliza- 1992. tion by free-ranging jackass penguins, Spheniscus demersus. Ecol- 283 MEYER, M., J. P. HOLLE, AND P. SCHEID. Bohr effect induced by ogy 65: 1648-1655, 1984. CO2 and fixed acid at various levels of Oz saturation in duck blood. 307, NAITO, Y., T. ASAGA, AND Y. OHYAMA. Diving behavior of Adelie Pjlugers Arch. 376: 237-240, 1978. penguins determined by time-depth recorder. Condor 92: 583-586, 284, MILL, G. K., AND J. BALDWIN. Biochemical correlates of swimming and diving behavior in the little penguin Eudyptula minor. Physiol. 1990. 2001. 56: 242-254, 1983. 308 NOLET, B. A., P. J. BUTLER, D. MASMAN, AND A. J. WOAKES. Estimation of daily energy expenditure from heart rate and doubly 285 MILLARD, R. W., K. JOHANSEN AND W. K. MILSOM. Radioteleme- try of cardiovascular responses to exercise and diving in penguins. labeled water in exercising geese. Physiol. 2001. 65: 1188-1216, Comp. Biochem. Physiol. A Physiol. 46: 227-240, 1973. 1992. 286. MILLER, B. F., AND C. BROCKMAN KEANE. Encyclopedia and Dic- 309. NOLET, B. A., D. E. H. WANSINK, AND H. KRUUK. Diving of otters tionary of Medicine, Nursing, and Allied Health. Philadelphia, PA: (Lutra lutra) in a marine habitat: use of depths by a single-prey Saunders, 1978. loader. J. Anim. Ecol. 62: 22-32, 1993. 287. MILSOM, W. K., M. CASTELLINI, M. HARRIS, J. CASTELLINI, D. R. 310. OLEKSENKO, A. I., L. M. MUKHAMETOV, I. G. POLYAKOVA, A. Y. JONES, R. BERGER, S. BAHRMA, L. REA, AND D. COSTA. Effects SUPIN, AND V. M. KOVALZON. Unihemispheric sleep deprivation of hypoxia and hypercapnia on patterns of sleep-associated apnea in bottle-nosed dolphins. J. Sleep Res. 1: 40-44, 1992. in elephant seal pups. Am. J. Physiol. 271 (Regulatory Integrative 311. OLSEN, C. R., F. C. HALE, AND R. ELSNER. Mechanics of ventila- Comp. Physiol. 40): RlOl7-R1024, 1996. tion in the pilot whale. Respir. Physiol. 7: 137-149, 1969. 288. MILSOM, W. K., K. JOHANSEN, AND R. W. MILLARD. Blood respira- 312. PACKER, B. S., M. ALTMAN, C. E. CROSS, H. V. MURDAUGH, JR., tory properties in some Antarctic birds. Condor 75: 472-474, 1973. J. M. LINTA, AND E. D. ROBIN. Adaptations to diving in the harbor 289. MILSOM, W. K., D. R. JONES, AND G. R. J. GABBOTT. On chemore- seal: oxygen stores and supply. Am. J. Physiol. 217: 903-906,1969. ceptor control of the ventilatory response to CO2 in the unanaesthe- 313. PANNETON, W. M. Primary afferent projections from the upper tized duck. J. Appl. Physiol. 50: 1121-1128, 1981. respiratory tract in the muskrat. J. Comp. Neurol. 308: 51-65, 1991. 290. MORTOLA, J. P., AND C. LANTHIER. Normoxic and hypoxic breath- 314. PANNETON, W. M. Trigeminal mediation of the diving response in ing pattern in newborn erev seals. Can. J. 2001. 67: 483-487. 1989. the muskrat. Brain Res. 560: 321-325, 1991. Jzdy 1997 PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 897

315. PANNETON, W. M. Trigemino-autonomic connections in the hclin- 338 RIDGWAY, S. H, AND R. J. HARRISON. Diving dolphins In: re^ stem of the mush~at.Soc. Neurosci. Abstr 19: 321, 1993. search on DolphCw, edited by M. M. Brydu~tand R. Hiurkun. Ox- 316. PANNETON, W. M., AND P. YAVARI. A medullary dorsal horn relay ford, UK: Clarendon, 1986, p. 33-58. for the cardiorespiratow responses evoked by stimulation of the 339. RIDGWAY, S. H., R. J. HARRISON, AND P. L. JOYCE. Slccp and nasal mucosa in the muskrat Ondat?a zibethim: evidence for cardiac rhythm in the gray seal. Science 187: 553-555, 1975. excitatory amino acid transmission. Brain Res. 691: 37-45, 1995. 340. RIDGWAY, S H , AN^ R. HOWARD, Dolphin lung collapse and intra- 317. PARKOS. C. A., AND E. A. WAHRENBROCK. Acute effects of hmer- capnia and hypoxia on minute ventilation in unrestrained Weddell seals. Respir Physiol. 67: 197-207, 1987. 3lR. PASCHE, A. Hypnxia in freely diving hooded seal, Cgstuplwra eris- utd deep diving in the bottlenose porpoise. Scierze 166: 1651- tap Camp. Biochem. Physiol. A Phgsiol. 55: 319-322, 1976. 1654, 1969. 319. PASCHE, A. The effect of hypercapnia on respimtory charactcris- 342. ROMANENKO, E. V. Swimming of dolphins: experiments and mod- tics and diving bellaviour uf krly diving seals. Resyir. Pl&ysiol. elling. In: Biological FZuid Dynamics, edited by C. P. Ellington and 26 183-194, 1976. T. J. Pedley. Cambridge, UK: SEB, 1995, p. 21-33. 320. I'ASCIIE, A,, AND J. KROG. Ileart rate in resting seals on land and 343. SAVABI, F. Free creatiue available to the creatine phosphate energy in water. Comp Biochem. Physiol. A Physwl. 67: 77-83, 1980. shuttle in Isolated rat atria. Proc. Natl Acad. Sci. USA 85: 7476- 321. PEDROLI, LC.ActiviQ and time budget of tufted duck on Swiss 7480. 19M~.. lakes dMng winter. Wild.fofou,l93: 105-1 12, 1R82. 4 IFRI: 'I h\\\.\'illll<~~.A\#* I' >c 1it:iIl t8qgema .ant(? 322. PERSSON, S. G. B., L. EKMAN, G. LYDEN, AND G. TUFVESSON. <.:~rIm!~~I~uxldk, d~~~~~clal~ In cjid&hk IBI!+h,,.l 2i i- Circulation effects of splenectomy in the hone I-IV. 11. Effect on 13, 1975. plasma volume and total circulating red cell volume. Zcntralbl. 345. SCHMIDT-NIELSEN, K, AnimalPhysi01ogy:Ada~ptationand Envi- Vctevinaamed. 20: 456-468, 1973. ronment (3rd ed.). Cambridge, UK: Cambridge Univ. Press, 1983,

323. PONGANIS. P. J.. R. L. GENTRY. E. P. PONGANTS~ ~ aNI, K~~ V~. PON-.~ n, RIR- (i.\hlS .\n:iI)sIs of s!nm vrInc111~4t111111~1! IIcI.~, ittnd ..li.~llu\~111vs~ '(lu ~~'iIIJL\Ul~t.i<.1' I. I.\~,~~IIuI~I1i8l II \C.VI~~JIIOIIS011 IIIC rr*lglr;i a,f lw,, ~la,lrhcn~fur srols (irlh.,.l,ii.,c ,r,.,.,,.,\ .11<,r .11~1~,t~~~~,I.~~~~1c.n hm

334. REED., J. Z.., C. CHAMBERS. M. A. FEDAK AND P. J. BUTLER. Gas~ ~ ~ 358. S~DER,G. K Respiralow adaplatiunb irr diving mluomals. ReayLr: exchange of captive freely diving grey seais (~alichoe- grypus). Physiol. 54: 269-294, 1983. J. Eq. Biol. 191: 1-18, 1994. 369. SPEAKMAIi, J. R.. AND P. A. RACEY. The equilibriuni concentration 335. REILLY, J. J., AND M. A. FEUAK. Rates of water turnover and enerm of oxygen-18 in body water: implications ior the accuracy of the expenditure of free~livingmale common seals (Phoca vitulina).J. doubly-labelled water technique and a potential new method of Zoo1 Lond. 223: 461-468, 1991. measuring RQ in free-living animals. J. Theor. Biol. 127: 79-95,

336. RIDGWAY. S. H., C. A. BOWERS, D. MILLER, M. L. SCHULTZ, C. A. -~-1987.~ JACOBS, AND C. A. DOOLEY. Diving and blood oxygen in the white 300. SPENCER. M. P., T. A. GORNALL 111, AND T. C. POULTER. Respira whale. Can. J Zool. 62: 2349-2351, 1984. tory and cardiac activity of killer whales. J. Appl. PhgsioL 22: 974- 337. RIDGWAY, S. H., D. A. CARDER, AND W. CLARK Conditioned 981, 1867. bradycardia m the sea llon Zalophus calz@?rnzanus. Nature 256: 361. STAHL, W. R. Scaillng of resplratoryvanables in mammals. J. Appl. 37-38, 1975. Physioi. 22: 453-460, 1967. 898 PATRICK J. BUTLER AND DAVID R. JONES Volume 77

362. STAINSBY, W. N., AND G. A. BROOKS. Control of lactate metabo- Oxygen aflinity of avian hemoglobins. Comp. Biockem. Physzol. lism in contracting muscles and during exercise. In: Exercise and 44: 711-718, 1973. Sport Sciences Reviews, edited by K. B. Pandolfand J. 0. Holloszy. 385. WANLESS, S., T. CORFIELD, M. P. HARRIS, S. T. BUCKLAND, AND Baltimore, MD: Williams & Wilkins, 1990, p. 29-63. J. A. MORRIS. Diving behaviour of the shag Phalacroco,uz aristo~ 3G3. STEPHENSON, R. Diving energetics in lesser scaup (Aythyu uf: tel& (Aues: Pelecu~cfJv).rrws) ir~relatior, Lu water depth and prey finis. evton). J. Em. Biol. 190: 155-178. 1994. size. .I. Zool. Load. 231: 11-25. 1993.

--., . .. . --, 365. STEPHENSON, R., AND P. J. BUTLER. Nelvouscontrolofthedivlng 387. WANLESS, S., J. k MORRIS, AND M. P. HARRIS. Diving behaviour response in birds and mammals In: The Neurobiology of the Car- of guillemot Uria aalge puffin Fratemla arctica and razorbill Alco diorespimtory System, edited by E. W. Taylor. Manchester, UK: torda as shown by radio-telemetry. J Zool. Lond. 216: 73-81, 1988.

Manchester~~ lJniv.~ Press.~ ~ 1987.~ . .~o. 369-393. 388. WASSERMAN, K., W. L. BEAVER, AND B. J. WHIPP. Gas exchange 'ItX Sl't'Pllt'hil~\, tt I' .I l!l '1'I.t:lt h llIlN~l'~1kt;.\hi) .\ .I theow and the lactic acidosis (anaerobic) threshold. Cim~lalion \Vu.US llvisn rat<, and gu, rxinangv In irecl.! d.\ing .\nrrncact 81, Suppl. 11: II-14-1130, 1990. mink [Mustela uison). J Em.Baal. 134: 435-442, 1988. 38Y. WATANUKI, Y., A. KATO, Y. MORI, AND Y. NAITO. Diving perfor- 367. STEPHENSON. R.. P. J. BUTLER,ANDA.J. WOAKES. Divim behw mance of Ad6lic pcnguins in rclation to food availability in fast iour and heart rat; in tufted ducks (Aythyafuligula). J E&. Biol. sea-ice areas: comparison between years. J Anim. Ecol. 62: 634-

126: 341-359, 1986. fi46~~ -~~ IWA , ~~~ 368. STEPHENSON, R., M. S. HEDRICY AND D. R. JONES. Cardiovascu- 390. WATKINS, W. A,, M. A. UAHER, K. M. FKISTRUP, AND T. J. HO- lar responses to diving and involuntary submergence in the rhinoc- WALD Spem whales tagged with transponders and tracked under- eros anklet (Cerorninca mnocwata Pallas). Can J 2001. 70: water by sonar. Mar. Mammal Sci. 9: 55-67, 1993. 2303-2310, 1992. 391. WEBER, R. E., T. KLEINSCHMIDT, AND G. RRAUNITZER. Embry~ 369. STEPHENSON, R., AND D. R. JONES. Diving physiology-birds. In: onic pig hemoglobins Gower 1, Gower 11, Heide I, and Heide 11: Comporatizir Pvlmonan~Physiology: Current Comqts, cdited by S. C. Wood. New York: Deliker, 1989, p. 735-786. 370. STEPHENSON,R., ANDDR. JONES. Metabolic responses to forced .. dives in Pekin duck rneasured by indirect calurin~etryandmP-MRS. wltult. iluud iuud erytltrucyk susprrtsiurw. Rrsyir: physiuL 27: 21- Am. J. Pkysiol. 263 (Regulatory Integmtiue Comp Pkysiol. 32): 31, 1976. R1309-R1317, 1992. 393. WELLS, R. M. G. Observations on the hacmatology and oxygcn 371. STEPHENSON, R., AND D. R. JONES. Blood flow dktributlon in transport of the New Zealand fur seal, Arctocephalus forsteri. snhmerged and snrfarr-swimming d~~cks..J~ E,w Riol 166 2R5- N Z .J. Zool. 5: 421-424, 1978. 296, 1992. 394. WELLS, R. M. G., AND S. 0.BRENNAN. Oxygen equiliblium proper- 372. STEPHENSON, R., J. R. LOWORN, M. R, k HEIEIS, D. R. JONES, ties of isolated haemoglobins from the Weddell seal Leptonyeholes AND R. W. BLAKE. A hydromechanical estimate of the power re- weddelli. Comp. Biochm. Physiol. A Physiol. 63; 365-368, 1979. 395. WEST, G. P. ErLc?/clope&ia of Animal Care. Baltimore, MD: Wii- liams & Wilkins, 1975. 396. WEST, N. H., AND B. N. VAN VLIET. Facton inlluencing the onset snip between almng acuwty and oxygen storage capacity in the and maintenance of bradycardia in mink. k'hysiol. Zool. 59: 451- tufted duck (Authua fulioula) J Em. Biol. 140: 265-275. 1989. 463, 1986. 397. WHALEN, W. J., D. BUERK, AND C. A THUNING. Blood flow-linr- ited oxygen consumption in resting cat skeletal muscle. Am. J. .I,!, .St,,, Zc.~,l:Id 14.Y 1:nIu Phgsiol. 224: 763768, 1973. .I;; il'l,'iE,tI I.. k GRAY R STAI3l<,.%NDJhl ill.4Nr,I.El2..l~ I(vnrrl 398. WHIPP, B. J., AND S. A. WARD. Cardiopulmonary coupling during l>loodtlou ~na ,,.l ..I l%y*irl 5; I1 1- C. R. Bridges and P. J. Butler. Cambridge, UK: Cambridge Univ. 417, 1976. Press, 1989, p. 77-112. 4W. WHITEHEAD, M. D. Maximum diving depths of the Adklie penguin, 377. TENNEY, S. M., AND D. F. BOGGS. Comparative mammalian respi- Pygoscelis adeliae during the chick rearing period, in Prydz Bay, ratorv control. In: Handbook of Phus~oloou.Tne Remiraton, Sus- Antarctica. Polar Biol. 9: 329-332, 1989. tm. Control of Breathing. ~ethescie,~6-h. ~hysibl. ~oc.;1986, 401. WICKHAM, L. L., R. ELSNER, F. C. WHITE, AND L. H. CORNELL. sect. 3, vol. 11, pt. 2, chapt. 27, p. 833-865. Blood viscosity in phocid seals: possible adaptations to diving. J 378. THOMPSON, D., AND M. A. FEDAK Cardlac responses of grey seals Comp. Physiol. BBiochem. Sysl. Environ. Physiol. 159: 153-158, during divine at sea. J Em. Biol. 174 139-164. 1993. 1989. 379. THOMPSON; D., P. S. H~MMOND,K. S. NICHOLAS AND M. A. 402. WILLFORD, D. C., A. T. GRAY, S. C. HEMPLEMAN, R. W. DAVIS, FEDAK. Movements. divine and foraeine"" behaviour of "ere" " seals iuuo E. P. HILL. Temperature and the oxygen-hemoglobin dissacia (Halid~oemsgrypud). J. Go1 Lond 224: 223-232, 1991. tian curve of the harbor seal, Phoca vitulina. Respir. Physiol. 79. 380. THOMPSON, D., k R. HIBY, AND M. A. FEDAK How fast should I 137-144, IYYU. swim? Behaviourd implications of diving physiology. Symp. Zool. 403. WILLIAMS, T. D., D. R. BRIGGS, J. P. CROXALL, Y. NAITO, AND A. Sue. Lor~d.66: 349-368. 1993. KATO. Divine vattern and verformance in relation to foraging ecol- 381. TRAYLER, K. M., D. J. BROTIIERS,R. D. WOOLLER, AND 1. C. POT- ogy in the &too penguin, ~goscelispapua. J. Zool. ~bnz.227: TER. Opportunistic foraging by three species of cormorants in an -211-2sn .. -.. , .11992. . - . Australian estuary. d. Zoo1 Land 218: 87-98, lY8Y. 4 11.l.l. I I : I'I I l:.\I.l. \ \II,i( 382. TURNER, A. W., AND V. E. HODGETTS. The dynamic red cell stor- BRIC.1:S. S K(JI)\I'CIJ.. .\kt, T R B.UTi,N brvttt~patiem and aee function of the soleen in sheen. I. Relationshin to fluctuations oerformance in nonbreedine zentoa oenzuins (Puooscelis vama)

...... - - ~iol 155-168, 1983. chum. Physiol. i~:267-276, 1965. 406. WILLIAMS, T. M., W. A. FRIEDL, a~oJ. E. HAUN. The physiology 384. VANDECASSERIE, C., C. PAUL, A. G. SCHNEK, AND J. LEONIS. of botuenose dolphins (1Ursiops tmneatus): heart rate, metabolic July xw’ PHYSIOLOGY OF DIVING OF BIRDS AND MAMMALS 899

rate and plasma lactate concentration during exercise. J. Exp. Biol. cation to the estimation of field metabolic rates in Antarctic preda- 179: 31-46, 1993. tors. Med. Biol. Eng. Comput. 33: 145-151, 1995. 407. WILLIAMS, T. M., G. L. KOOYMAN, AND D. A. CROLL. The effect 416. WOOD, S. C. Adaptation of red blood cell function to hypoxia and of submergence on heart rate and oxygen consumption of swim- temperature in ectothermic vertebra&es. Am. Zool. 20: 163- 172,198O. ming seals and sea lions. J. Comp. Physiol. B Biochem. Syst. Envi- 417. W-URSIG, B., E. M. DORSEY, M. A. FRAKER, R. S. PAYNE, W. J. ron. PhysioZ. 160: 637-644, 1991. RICHARDSON, AND R. S. WELLS. Behavior of bowhead whales, 408. WILSON, R. P., AND B. M. CULIK. The cost of a hot meal: facultative Balaena mysticetus, summering in the Beaufort Sea: surfacing, specific dynamic action may ensure temperature homeostasis in respiration, and dive characteristics. Can. J. Zool. 62: 1910- 192 1, post-ingestive endotherms. Comp. Biochem. Physiol. A Physiol. 1984. 100: 151-154, 1991. 418. WURSIG, B., R. S. WELLS, AND D. A. CROLL. Behavior of gray 409. WILSON, R. P., K. HUSTLER, P. G. RYAN, A. E. BURGER, AND E. C. whales summering near St. Lawrence Island, Bering Sea. Can. J. NOLDEKE. Diving birds in cold water: do Archimedes and Boyle Zoo,?. 64: 611-621, 1986. determine energetic costs. Am. Nut. 140: 179-200, 1992. 419. YDENBERG, R. C., AND C. W. CLARK. Aerobiosis and anaerobiosis 410. WILSON, R. P., K. A. NAGY, AND B. S. OBST. Foraging ranges of during diving by western grebes: an optimal foraging approach. J. penguins. Polar Rec. 25: 303-307, 1989. Theor. Biol. 139: 437-449, 1989. 411. WILSON, R. P., AND M.-P. T. WILSON. Foraging behaviour in four 420. YDENBERG, R. C., AND L. S. FORBES. Diving and foraging in the sympatric cormorants. J. Anim. Ecol. 57: 943-955, 1988. western grebe. Omzis Stand. 19: 129-133, 1988. 412. WILSON, R. P., AND M.-P. T. WILSON. The foraging behaviour of 421. YDENBERG, R., AND M. GUILLEMETTE. Diving and foraging in the African Penguin Spheniscus demersus. In: The Penguins: Ecol- the common eider. Ornis Stand. 22: 349-352, 1991. ogy and Management, edited by P. Dar-m, I. Norman, and P. Reilly. 422. ZAPOL, W. M. Diving adaptations of the Weddell seal. Sci. Am. 255: Chipping Norton, NSW, Australia: Surrey Beatty, 1995, p. 244-265. lOO- 105, 1987. 413. WOAKES, A. J., AND P. J. BUTLER. An implantable transmitter for 423. ZAPOL, W. M., R. D. HILL, J. QVIST, K. FALKE, R. C. SCHNEIDER, monitoring heart rate and respiratory frequency in diving ducks. G. C. LIGGINS, AND P. W. HOCHACHKA. Arterial gas tensions and Biotelemetry 2: 153- 160, 1975. hemoglobin concentrations of the freely diving Weddell seal. Un- 414. WOAKES, A. J., AND P. J. BUTLER. Swimming and diving in tufted dersea Biomed. Res. 16: 363-373, 1989. ducks, Aythyafuligula with particular reference to heart rate and 424. ZAPOL, W. M., G. C. LIGGINS, R. C. SCHNEIDER, J. QVIST, M. T. gas exchange. J. Exp. Biol. 107: 311-329, 1983. SNIDER, R. K. CREASY, AND P. W. HOCHACHKA. Regional blood 415. WOAKES, A. J., P. J. BUTLER, AND R. M. BEVAN. An implantable flow during simulated diving in the conscious Weddell seal. J. Appl. data logging system for heart rate and body temperature: its appli- Physiol. 47: 968-973, 1979.