Camp. Biurhrm. Ph,vsiof Vol. 69A. pp 177 to 185. 1981 Printed I” Great Bntam All rights reserved
0300-9629 81.060177-09502.00~0 Copyright 0 19x1 Pergamon Press Ltd
PHYSIOLOGICAL INDICES OF ACTIVITY AND METABOLISM IN THE POLAR BEAR ROBIN C. BEST*, K. RONALD Department
and N. A. ORITSLAND~
College of Biological Science, Guelph, Ontario, Canada
(Received 20 August 1980)
Abstract--l. Respiratory frequency, core temperature, cardiac and stride frequencies, were evaluated as indirect indices of activity and metabolism in the polar bear (Llrsus maritimus). 2. Respiratory frequency was found inadequate as an index due to its role as a thermoregulatory effector and its “resonant” nature. 3. The core temperature of the bears was highly correlated with activity or metabolic rate. 4. The relationship was however curvilinear, resulting in only a 0.5”C change in core temperature over a five-fold metabolic range. 5. It was impossible to accurately define different levels of activity. 6. The cardiac frequencies of resting bears were extremely variable and not, as a result, good indices of these activities. 7. In the active or walking animal, however, cardiac frequency is an accurate index of activity or metabolic rate. 8. Stride frequency may also be used to accurately predict walking speed and, thus, metabolic rate in the polar bear.
INTRODUCTION Recent developments in long-range biotelemetry have made the remote measurement of physiological parameters such as cardiac frequency, respiratory frequency, and body temperature of mammals possible (Craighead et a/., 1971; Cupal et al., 1975; Mackay, 1970; Britsland et al., 1976; Pauley, 1971). Such measurements may be directly, or indirectly, associated with the activity level and metabolism of the animal and are therefore potentially useful as indices of these two parameters. The importance of such indices has stemmed from a renewed interest in the ecological energetics of free-ranging homeotherms (Gessaman, 1973; Moen, 1973) and a subsequent need for techniques that will allow the measurement or estimation of the undisturbed metabolism of these animals. Respiratory frequencies, core temperatures, cardiac frequencies and stride frequencies of captive polar bears (Ursus maritimus) were measured under controlled exercise conditions. The relative value of each of these measurements as an indirect index of either activity or metabolism in these bears has been evaluated.
of seal and whale meat and seal blubber nd /ibitum. Each bear was trained, through reinforcements to walk on a large motor-driven treadmill (Fig. 1) at speeds of 1.1, 1.6, 2.0 and 2.2 m/set. A fan generated speeds of 4 and 7 m/set which were directed into the face of the walking bear. The cardiac frequencies and core temperatures of the bears were monitored using surgically implanted telemetry transmitters (Skutt et al., 1973). The ECG transmitter was implanted subcutaneously between the forelegs, while the temperature transmitter was implanted intraperitoneally slightly anterior to the umbilicus. The radio signals from the transmitters were received on a commercial FM radio with an omnidirectional antenna. The calibrated accuracy of the temperature transmitters was within 0.4”C. Respiratory and stride frequencies were counted visually. Due to time restrictions and technical problems with the ECG transmitters, a complete series of measurements were made for certain bears. An energetics model (Wunder, 1975) was used to equate metabolism with activity. The model was based on the metabolic rate in polar bears (Hock, 1968; 0ritsland er al., 1976) and that for other mammals (Fedak et al., 1974; Taylor et al., 1970). The model: M, = a(3.39m,0.75)
1. Physical characteristics bears used in this study
MATERIALS AND METHODS
The polar bears used in the study (Table 1) were captured and held in the vicinity of Fort Churchill, Manitoba. The captive bears were exposed to outdoor ambient temperatures (- 40.0 to 8.O”C) and were fed a near-natural diet
Identification “cub” Nl N2 2250 2021 1750 1708
* Present address: Instituto National de Pesquisas da Amazonas, Caixa Postal 478, 69,000 Manaus, Amazonas, Brazil. t Norsk Polarinstitutt, Rolfstangveien 12, Postboks 158, 1330 Oslo, Norway. 177
F M M M M M M
I1 182 209 218 239 248 325
ROBINC‘. BE%‘, K.
Fig. 1. A schematic diagram of the rreadmlil and wind tunnel apparatus used to study polar hear energetics and physiological response to rest and exercise.
where: M, = total metabolic rate (W) r = activity coefficient (ND) PI,, = weight (kg) Vti.= walking speed (m.‘sec) The activity coefficient, is used to designate the mctabolic increment due to posture (Taylor ~‘f al.. 1970; Wunder, 1975), and is expressed here as multiple of basal metabolic rate. The following activities and activity coef. ficients were defined for the study: (i) sleeping
(ii) lying down
Respiratory fkequenq The respiratory frequency (fi) of bear 2250 ranged from 5-133 breaths per min (c/min). The lowest frequencies were recorded while the animal siepr, the highest during, or following, strenuous activity. TheJ of the moderately active bear was between IO and 20 cjmin. Thermal panting, characterized bv openmouthed, high-frequency (105.-133 c/mint respirations, occurred in response to heat stress brought on by either heavy exercise or warm ambient conditions. The respiratory frequency of the bear was closely correlated with its core temperature (N = 24; r = 0.94) and to a lesser extent, ambient temperature (N = 24; P = 0.62). Walking speed appeared to influence the animal’s ,f; only through its effect on the core temperature (Fig. 2). The “panting threshold” of the bear was at a core temperature of between 38.0 and 4O.o”C. Core temperature The thermoregulatory responses of bears 2250 and 2021, as exemplified by their core temperatures. were
similar to external factors such as wind, walking speed and ambient temperature. At thermal equilibrium the bears’ core temperature was regulated proportionally to the exercise level and independent of ambient temperature over the wide ( - 70 C) range of the thermoneutral zone. As a result of this relationship (Fig. 3) the core temperature of the bears could be accurately predicted as an exponential function of the waiking speeds :
T, = 36.58 + 0.12 V$” (-C)
(N = 54; r2 = 84. I ; P < 0.001 : SEE = 0.3’C) The relationship between the core temperature the activity coefficient (a = M&n,) was: I = -203.0
+ X4.81 T;.’
(N = 54; rz = 74.4: P < 0.001: SEE = 0.59) The bears showed a marked ability to behaviourally regulate their body tem~ratures to the equilibrium levels for each walking speed. even under ambient temperatures exceeding the thermoneutral zone.
Technical problems. arising from a tendency of the electrode leads to break at their junction with the transmitter, precluded a full series of cardiac frequency/activity measurements. The relatively complete series obtained from bear 1750 have been supplemented with observations from other bears. The ECG of a sleeping polar bear is characterized by a marked respiratory sinus arrhythmia, in contrast to the rhythmic beat of the “active” heart (Fig. 4). Basal cardiac frequency recorded at rest, decreased with increasing body size or age in the bears. The relationship between cardiac frequency (d) and activity in the polar bear is biphasic (Fig. 5) as the resting frequencies are linearly related to metabolism or the activity coefficients (x) whereas the active frequencies are curvilinearly related to either walking speed or metabolic rate.
Activity and metabolism in polar bear
qA 0 Walkmg Speed e-l.1 mls o-l.6 m/r ,5 -2.0 mls 0 -2.2 m/s - -
P,ed,cted 1, (Stahl 1967) Probable Pantmg Threshold
Fig. 2. The respiratory frequency of bear 2250 was primarily dependent on the bear’s core temperature and was independent of walking speed, with the exception of the relationship between core temperature and walking speed. The “resonant” nature of the respiratory frequency is implied by the absence of intermediate frequencies. The “panting threshold” is drawn from the literature.
Resting cardiac frequencies (SD = 9.8 c/min) and there is overlap between the frequencies resting posture. The relationship the restingf, is:
are highly variable consequently much associated with each between activity and
51= 0.37 + O.oOZf, (ND)
(N = 27; rz = 62.41; P < 0.001; SEE = 0.19) In the walk* bear the cardiac frequency is much less variable (SD = 2.8 c/min) and it is closely related to the walking speed (5) or metabolic rate. f, = 83.34 f
13.97 V,$’ (c/min)
(N = 25; r2 = 98.2; P < 0.001;
SEE = 2.9 c/min)
The maximum and minimum frequencies observed in the study were 192 c/min (1708) and 40c/min (1750). The transition between the “active” and “resting” cardiac phases occurred between 80 and 85 c/min in bear 1750. Stride frequency
The stride frequencies (fs) of the bears (N = 5) increased curvilinearly with walking speed and decreased with increasing body size (Fig. 6). These parameters could be combined to accurately predict the walking speed of a bear of known weight: VW= - 259.46 + 249.64 j’p,“’ + 0.01 mb (m/set) (6)
(N = 30; r:
= 96.1; r,$, = 2.3; P < 0.001;
‘SEE = 0.1 mjsec) The curvilinear nature of thef,/V, relationship is a result of simultaneous changes in the stride length and frequency at the lower walking speeds. A weight-specific energy cost per unit step (mcsk) may be calculated by dividing the total energy cost of walking [M, of equation (I)] by the weight and stride frequency of the animal (Calder & Gold, 1974; Gold, 1973). The resulting value is a negative exponential function (7) of the animal’s walking speed, reaching an asymptote of about 6.2 J/step.kg at speeds exceeding I m/set (Fig. 7). %k
= 6.18(1 - e-2.68”X (J/step.kg)
(N = 30; r* = 76.0; P < 0.001; SEE = 0.28 J,/step. kg) This asymptotic value is apparently independent of either the weight or velocity of the animal while the walking gait is maintained.
The primary function of respiration in the mammal is to obtain sufficient oxygen for metabolic processes and secondarily to aid in thermoregulation. The
ROBIN C. BEST, K. RONALD and
N. A. ~RITSLANII
WALKINGSPEED (m/r) Fig. 3. The core temperatures of the polar bears, within the thermoneutral zone, were regulated as a function of the walking speed (metabolic level). Numbers in parentheses indicate the number of observations and vertical bars represent f 1 SD of the mean.
underlying physiological interrelationship between respiratory frequency and metabolic rate may be expressed in the following form (Hargrove & Gessaman, 1973): M, = %I, = .L’ VT(F,O,- FEo2)(War
Vo, = oxygen consumption 1; = respiratory
V, = tidal volume (l/c) F102 = oxygen content
FEo, = oxygen content
of expired air (10,/l)
It is possible in equation (8) to measure only the frequency by telemetry, although future developments may enable the measurement of tidal volume or oxygen consumption (Mackay, 1970). The respiratory frequencies of the polar bear were similar to those predicted from the empirical equation (9) derived for mammals (Stahl, 1967), as the observed normal frequencies ranged from ,f; = 53.5 m;o,26 t_ 0.4 (c/min)
S-20 c/min while the predicted rates were 8-19 c/min. Thermal panting (J = 110-133 c/min) occurred in response to either strenuous activity, excitement or warm ambient conditions. This breathing mode was characterized by apparently shallow, open-mouthed exhalations. Panting frequencies of 120 and 130 c/min
have been recorded from brown bears (I/. arctos) weighing 222 and 310 kg respectively (Jaczewski et al., 1960). The similarity of both resting and panting frequencies of the polar bear to those of other species of nearly equivalent weight and the apparent lack of intermediate frequencies may be a result of the “resonant” nature of this parameter (Crawford, 1962). The regulation of respiratory frequency was primarily associated with the core temperature of the animal rather than by relative work load or metabolic rate (Fig. 2). Adjustments in the respiratory pattern resulting from an increased metabolic rate were apparently compensated for by an increased tidal volume rather than through frequency changes. These factors combine to render support to Hargrave & Gessaman’s (1973) theory that respiratory frequency is valueless as a metabolic indicator in the polar bear. Core temperature The core temperature of mammals is regulated proportionally to the metabolic rate and is independent of ambient temperature throughout the thermoneutral zone (Bligh, 1973). These characteristics of mammalian thermoregulation have led to the evaluation of core temperature as an index of activity of metabolism (Berggren & Christensen, 1950). In various sized mammals, the core temperature is maintained around 37°C (Bligh, 1973; Morrison & Ryser, 1950), whereas metabolic rate is an exponential function of the animal’s weight (Kleiber, 1975; Poczopoko, 1971). The use of any type of core temperature/
Activity and metabolism in polar bear
b) WALKING Fig. 4. (a) A typical cardiogram from a sleeping polar bear, illustrating the marked cardiac arrhythmia characteristic of this posture. (b) A typical cardiogram from a walking polar bear, showing the typically rhythmic
heart beat of an active animal.
metabolic index is thus restricted to the situation where metabolic increments are expressed as dimensionless activity coefficients (a = M,/m,). For polar bears this relationship was strongly curvilinear, reflecting the regulation of core temperature and walking speed (Fig. 3). The essentially linear increments in metabolic rates due to increasing walking speed (Oritsland et al., 1976) cause a curvilinear change of body temperature. This precludes its use as an index of either activity or metabolism in this species. The depressed core temperature’s characteristic of the denning animal (Folk et al., 1972) creates further difficulties, the use of such an index is not recommended. Cardiac frequency Mammalian metabolism and cardiac frequency shows the following physiological interrelationship: M, = ii02 = fc. V,(Ca,, - Cu,,) (W or 1 O*/min) (10) where: V, = stroke volume (l/c) Cao, = oxygen (10,/l)
Cv,, = oxygen (10,/l)
The above relationship has led a number of researchers to use cardiac frequency as a metabolic index for mammals (Johnson 8.1 Gessaman, 1973). The cardiac frequency of the polar bear in relation to activity is biphasic, as in other mammals (Rushmer, 1970) with one component being representative of the quiescent animal and other typical of the active situation (Fig. 5). Minimal cardiac frequencies occurred during rest and were related to the weight (Fig. 7) or age, of the bear (Folk et al., 1972; Hock, 1966, 1968). They are predictable using an empirical equation (11) derived for mammals (Stahl, 1967). Even lower frequencies (27-35c/min) have been recorded from adult polar bears (26O310 kg) during the “winter sleep” associated with denning (Folk et al., 1970, 1972). f, = 241.0m; o.25 _+ 0.34 (c/min)
The extreme variability of the frequencies observed for each resting posture (Fig. 5) likely resulted from interactions between the stroke volume and frequency (Rushmer, 1970). This variability makes fc a poor index of activity or metabolism in the resting polar bear (N = 2.5; r = 0.43). A similar conclusion was reached for man (Booyens & Hervey, 1960) and for horses (Ehrlein et al., 1970). Only the respiratory sinus arrhythmia, characteristic of the sleeping polar bear,
BEST, K. RONALD and N. A. ~RITSLAND
0 Sloeping A Lying Dorm Sitting 0 Standing 0 Walking
Fig. 5. The cardiac frequency of bear 1750 changed with activity level or metabolic rate [estimated from equation (l)]. This relationship was biphasic due to simultaneous changes in cardiac frequency and stroke volume for different postures. Numbers in parentheses indicate the number of observations and the vertical bars represent k 1SD of the mean.
is distinctive enough to allow discrimination of this activity in this species. The “active” cardiac frequencies are not greatly variable and increase curvilinearly with walking speed in the polar bear (Fig. 5). The curvilinear nature of this relationship may be due to the effects of body temperature and its regulation on f, (Mostardi et al., 1974; Vogt et al., 1973). The high degree of correlation between VWand f, (N = 23; I = 0.98) implies that the cardiac frequency is an excellent index of either walking speed or metabolism in the active bear. The use of cardiac frequency as a general index within a population is limited due to the individual variability caused by factors such as age, sex and physiological condition (Andrews, 1971; Johnson & Gessaman, 1973). To circumvent this variability the use of the “net cardiac frequency” (f,,) may be necessary (Andrews, 1971; Vogt et al., 1973). The&, may be determined by dividing an individual’s total cardiac frequency by its observed basal frequency, with the net increment being used to characterize the level of activity or metabolism for each individual. This technique has not been used to our knowledge in mammals other than man, but it may allow intraspecific comparisons without necessitating the laboratory “calibration” of each individual that is to be monitored.
Transitory cardiac phenomena, such as the tachycardia associated with excitement or stress, and the bradycardia associated with immersion or the ingestion of food (Folk et al., 1972, 1973), are likely of too short duration to affect the overall use of cardiac frequency as an index. The technical problems encountered with the flexible electrodes of the Skutt et a[. (1973) transmitters breaking at their insertion to the capsule may be overcome through the use of fixed electrodes such as those of the “Iowa implantable radio-capsule” (Folk & Copping, 1973). Until this problem is solved for the polar bear transmitters, the use of this parameter will be restricted to short-term laboratory studies.
Stride frequency Stride frequency and metabolic rate are proportional to walking speed and body weight in mammals (Fedak et al., 1974; Heglund et al., 1974; Taylor et al., 1970). These relationships have led some authors to propose that the weight-specific cost per unit step (m,,,) may be constant for geometrically similar animals (Calder & Gold, 1974; Gold, 1973; Herschman, 1974). The relationship between mesk and walking speed in the polar bear is negative-exponential due to simultaneous changes in stride length and frequency at the
A -2250 0
0 -Nl + -N2
Fig. 6. Stride
lower walking speeds. Once the asymptotic value of 6.2 J/step.kg has been reached at speeds exceeding 1 m/set, however, this value is apparently independent of either the speed or weight of the bear (Fig. 7). The use of mcsk as an index of metabolic rate in the polar bear is restricted to situations where the walking speed of the animal is known. An apparently more versatile method of using stride frequency as an index is through the use of the relationship between the weight, stride frequency and walking speed of the animal [equation (6)]. The weight of the bear, measured at the time of capture, and the recorded stride frequency may be used to accurately predict the walking speed. The walking speed and weight may, in turn, be used in an energetics model [equation (I)] if the metabolic rate is desired. This latter technique has the advantage that it does not require the comparison of curvilinear and linear variables as does the mesk/Mt relationship. The stride frequencies used in this study were measured visually, although suitable telemetry transmitters for the measurement of the electromyogram (EMG) of active muscle are available (Ko, 1965; Kuck et al.. 1963). The use of stride frequency as a
speed and decreasing
metabolic index has not been tested in any other studies that we are aware of and as a result, many of the potential advantages or disadvantages are not readily discernible. Its use as an index of walking speed or metabolic rate in the polar bear seems practical, although the fact that it has no use under resting conditions may be a serious limitation. It may, however, be possible to characterize the EMG patterns for each resting posture so that they can be differentiated and thus extend the usefulness of this technique throughout the full spectrum of activities. Such measurements may also allow the discrimination of shivering thermogenesis, activities such as swimming, and may reflect the increased energy cost of locomotion in deep snow through the resulting changes in the locomotory pattern (Heinonen et al.. 1959; Ramaswami er a/., 1966). SUMMARY
Of the four physiological indices evaluated as indices of activity and metabolism, respiratory frequency and core temperature were not suitable for such a purpose, whereas both cardiac frequency and
BEST, K. RONALDand N. A. ORITSLANU
0 Thisaudy. A Folknt 91.119731 0 Hock (19681 +f- 0ritsland(1970) - Stahi(1967)
g ; 200 11 I
Fig. 7. The weight-specific energy cost per unit step (w,~) of the polar bear increased non-linearly with walking speed, reaching an asymptote of approximately 6.2 J/step.kg at walking speed of nearly t .Om/see.
stride frequency could be used as indices for the polar bear. The inherent characteristics of the latter two physiological parameters made them poor indices of the resting postures. This could restrict the value of
these techniques as polar bears may spend from 70-85% of the day resting (Knudsen, i973; Stirling, 1974). The relatively narrow range of metabolic levels encompassed by these activities (%= _ 1.0-1.7) may allow the use of a mean activity coefficient @) for all resting levels without introducing large errors into the estimation of the daily energy budget. The use of such a value, 5 = 1.1, calculated from Stirling’s (1974) data, resulted in a 2.4”/, underestimate of the calculated daily energy budget. Cardiac frequency and stride frequency are both accurate and versatile indices of the walking speed and active metaboIism in the poIar bear. The cardiac frequency may reflect metabolic increments resulting from non-exercising processes (Johnson & Gessaman, 1973), whereas, due to inherent individual variability, its use, without detailed analytical techniques, is presently restricted to individuals whose cardiac responses to activity have been previously determined in the laboratory. Stride frequency, although useful only as an index in the moving animal, is scaleable to body size (Heglund et al., 1974) and is therefore a relatively standard index within a species. The simultaneous use of both of these parameters on an individual animal is recommended, as the characteristic limitations of each technique are to an extent
mutually exclusive and their dual use could reduce the Iikeliho~ of misinterpretation of the data.
REFERENCES ANDREWS R.
B. (1971) Net heart rate as a substitute for calorimetry. Am. f. C&n. Nutr. 24,
t139-1147. BEST R. C.
(1976) Ecological energetics of the polar bear (Ursus mnritimus Phipps 1774). M.Sc. thesis. University of Guelph, Guelph, Ontario. 136 pp. BO~YENSJ. & HERVEYG. R. (1960) Pulse rate as a means of measuring metabolic rate in man. Can. J. Biochem. Phvsiol. 38, 1301-1309. BLIGH J. (1973) Temperature regulation in mammals and other vertebrates. pp. l-436. North-Holland, London. BERGGRENG. & CHRISTENSEN E. H. (1950) Heart rate and body temperature as indices of metabolic rate during work. Arbei~sphys~~~ogie 14, 25$X0. CALDER W. A. & GOLD A. (2974) Energy cost of animal locomotion. Science, N.Y. 184, 1098. CRAIGHEADJ. J., CRAIGHEADF. C.. VARNEY J. R. & Co~ri C. E. (1971) Satellite monitoring of black bear. Bio-
science 21, 12061212. CRAWFORD E. C. (1962)
Mechanical aspects of panting in
dogs. J. appl. Physiot.
17, 249-251. CUPAL J. J., WARD A. L. &WEEKS R. W. (1975) A
repeatertype biotelemetry system for use on wild big game animals. Instrum. Sot. Am. Trans. 14, 101-108.
EHRLEINH. J., ENGELHARDTW. V., HORNICKE H., TOLKMITTG. & DUSK J. (1970) (Studies on the relationships
between heart rate performance in the horse.) Zentbl. VetMed. 17A, 577-591. (German.)
and metat ,olism in polar
FEDAK M. A.. PINSHOW B. & SCHMIDT-NIELSEN K. (1974) Energy cost of bipedal running. Am. J. Phqjsiol. 227, 1038~1044. FOLK G. E. BERBERICH J. J. & SAUNDERS D. K. (1973) Bradycardia of the polar bear. Arcric 26, 78-79. FOLK G. E.. BREWER M. D. & SAUNIIERS D. K. (1970) Cardiac physiology of polar bears in winter dens. Arctic 23, 130. FOLK G. E. &COPING J. R. (1973) Telemetry in biometeorological sciences. Biometeoroloyy 5, I53- 170. FOLK G. E., FOLK M. A. & MINOR J. J. (1972) Physiological condition of three species of bears in winter dens. In Bears: their biology a,*d-mclnagente,lr (Edited by HERRERO S.), pp. 107-124. Int. Union Conserv. Nature Pub]., New Ser. 23. GESSAMAN J. A. (1973a) Ecological eneryetics of homeotherms. p. 155. Utah State University Press, Utah. GOLD A. (1973) Energy expenditure in animal locomotion. Science N. Y. 181, 275-276. HARGROVE J. L. & GESSAMAN J. A. (1973) An evaluation of respiratory rate as an indirect monitor of free-living metabolism. In Eco/oyical erwrgetics of homeotherms (Edited bv GESSAMANJ. A.), DD. . . 77-85. Utah State University Piess, Utah. HEGLUNII N. C.. TAYLOR C. R. & MCMAHON T. A. (1974) Scaling stride frequency and gait to animal size: mice to horses. Scierlce N. Y. 186. 1112-I 113. HE~NONENA. 0.. KARVONEN M. & RLJOSTEENOJAR. (1959) The energy expenditure of walking on snow at various depths. Eryotlomics 2. 389-393. HERSCHMAN A. (1974) Animal locomotion as evidence for the universal constancy of muscle tension. Am. J. Physiol. 42, 778-779.
HOCK R. J. (1966) Heart rate of black bears in relation to age. J. Mammal. 47. 328-329. HOCK R. J. (1968) Some physiological measurements of polar bears. Z. Siiugetirrk. 33, 57-62. JACZEWSKI Z., GILL J. & KOZNIEWSKI S. (1960) Regulation of blood pressure in the brown bear. Int. Union Game Biol., Trans. 4th Congress. Arnheim. Netherland: 131-145. JOHNSON S. F. & GESSAMAN J. A. (1973) An evaluation of heart rate as an indirect monitor of free-living energy metabolism. In Ecologicu/ energetics of homeotherms (Edited by GESSAMAF;J. A.), pp. 44-54. Utah State University Press, Utah. KLEIBER M. (1975) Thefire qf/ije. 2nd edn. 456 pp. R. E. Krieger, N.Y. KNUDSEN B. M. (1973) The ecology of polar bears on North Twin Island, N.W.T. University of Montana, Missoula, Montana. MSc. thesis. 70 p. KO W. H. (1965) Progress in telemetering muscle potentials. Biomed. Sci. I~lstrum. 2. 267-274. KUCK A., LIEBMAN F. M. & KUSSICK L. (1963) A miniature
bear for telemetering
Electron. BME-10, 1i 7-l 19. MACKAY R. S. (1970) Biomedical telemetry. 2nd Trans. Bio-Med.
edn. 533 pp. J. Wiley, Toronto. MOEN A. N. (I 973) Wildlife ecology; au a~lalytical approach. 458 pp. W. H. Freeman. San Francisco. MORRISON P. R. & RYSER R. A. (1952) Weight and body temperature in mammals. Sciewe N. Y. 116. 23 I-232. MOSTARD~ R., KURICA R., VEICSTEINASA. & MARGARIA R. (1974) The effect of increased body temperature due to exercise on the heart rate and maximal aerobic power. Eur. J. appl. Phqasiol. 33, 237-245.
~RITSLAND N. A. (1970) Temperature regulation of the polar bear (Thalarctos maritimus). Camp. Biochem. Physiol. 37. 225-233.
ORITSLAND N. A.. JONKEL C. & RONALI) K. (1976) A respiration chamber for exercising polar bears. Now. J. Zoo/. 24. 65-67.
PAULEY J. D. (1971) A pulse modulated biotelemetry system for monitoring heart rate and body temperature in free-ranging vertebrates. University of Kansas, Lawrence, Kansas. Ph.D. thesis. 107 pp. POCZOPKO P. (1971) Metabolic levels of adult homeotherms. Acta theriol. 16, I-20. RAMASWAMYS. S.. DUA G. L.. RAIZDA V. K., DIMRI G. P.. VISWANATHAN K. R., MADHAVIAH J. & SRIVASTAVA T. (1966) Effect of looseness of snow on energy expenditure in marching on snow-covered ground. J. appl. Pkysiol. 21, 1747-1749. RILEY J. L. (1972) Transmitter for recording respiration rate information by radiotelemetry. J. uppl. Physiol. 32. 259-260. RUSHMER R. F. (1970) Cardioaascular dyrmmics. 3rd edn. 559 pp. W. B. Saunders, Toronto. SKLJTT H. R., BOCK F. M., HAUGSYADP.. HOLTEK J. B., HAYESH. H. & SILVERH. (1973) Low power implantable transmitters for telemetry of heart rate and temperature from white-tailed deer. J. Wildl. Mymt 37, 413- 4 17. STAHL W. R. (1967) Scaling of respiratory variables in mammals. J. appl. Physiol. 22, 453-460. STIRLING I. (1974) Midsummer observations on the behaviour of wild polar bears (Ursus maritimus). Can. J. Zool. 52, 1191-l 195. TAYLOR C. R., SCHMIDT-NIELSEN K. & RAAB J. L. (1970) Scaling energetic cost of running to body weight of animals. Am. J. Phvsiol. 219, 1104-I 107. Vocr J. J., MEYER-SCHWERTZ M. Th.. METZ B. & FOEHR R. (1973) Motor thermal and sensory factors in heart rate variation: a methodology for indirect estimation of intermittant muscular work and environmental heat loads. Ergonomics 16, 45-60. WUNDER B. A. (1975) A model for estimating metabolic rate of active or resting mammals. J. theor. Biol. 49, 345-354.