Oryx Leucoryx): Does Water Limitation Drive Heterothermy?

J Comp Physiol B (2010) 180:1111–1119 DOI 10.1007/s00360-010-0480-z

ORIGINAL PAPER

Variation in the daily rhythm of body temperature of free-living Arabian oryx (Oryx leucoryx): does water limitation drive heterothermy?

Robyn Sheila Hetem · Willem Maartin Strauss · Linda Gayle Fick · Shane Kevin Maloney · Leith Carl Rodney Meyer · Mohammed Shobrak · Andrea Fuller · Duncan Mitchell

Received: 10 February 2010 / Revised: 28 April 2010 / Accepted: 3 May 2010 / Published online: 26 May 2010 © Springer-Verlag 2010

Abstract Heterothermy, a variability in body temperature (36.1 § 0.3 vs. 36.8 § 0.2°C, P = 0.04), resulting in a beyond the limits of homeothermy, has been advanced as a larger daily amplitude of the body temperature rhythm key adaptation of Arabian oryx (Oryx leucoryx) to their (5.0 § 0.5 vs. 2.9 § 0.2°C, P = 0.0007), while mean daily arid-zone life. We measured body temperature using body temperature rose by only 0.4°C. The maximum daily implanted data loggers, for a 1-year period, in Wve oryx amplitude of the body temperature rhythm reached 7.7°C free-living in the deserts of Saudi Arabia. As predicted for for two of our oryx during the hot-dry period, the largest adaptive heterothermy, during hot months compared to amplitude ever recorded for a large mammal. Body temper- cooler months, not only were maximum daily body temperature variability was inXuenced not only by ambient tem- atures higher (41.1 § 0.3 vs. 39.7 § 0.1°C, P = 0.0002) perature but also water availability, with oryx displaying but minimum daily body temperatures also were lower larger daily amplitudes of the body temperature rhythm during warm-dry months compared to warm-wet months (3.6 § 0.6 vs. 2.3 § 0.3°C, P = 0.005), even though ambient temperatures were the same. Free-living Arabian oryx Communicated by G. Heldmaier. therefore employ heterothermy greater than that recorded in any other large mammal, but water limitation, rather than R. S. Hetem (&) · W. M. Strauss · L. G. Fick · S. K. Maloney · high ambient temperature, seems to be the primary driver L. C. R. Meyer · A. Fuller · D. Mitchell Brain Function Research Group, School of Physiology, of this heterothermy. Medical School, University of the Witwatersrand, 7 York Road, Parktown, Johannesburg 2193, South Africa Keywords Body temperature · Arid-adapted antelope · e-mail: [email protected] Hyper-arid environment W. M. Strauss · M. Shobrak National Wildlife Research Center, National Commission for Wildlife Conservation Introduction and Development, PO Box 1086, Taif, Saudi Arabia

W. M. Strauss Amongst large desert-dwelling mammals, the species Department of Environmental Sciences, which arguably faces the greatest challenge to homeostasis University of South Africa, Private Bag X6, is the Arabian oryx (Oryx leucoryx). This antelope (body Florida 1709, South Africa mass 70–100 kg) inhabits one of the hottest deserts in the S. K. Maloney world, in Saudi Arabia, and survives most of the year, Physiology: Biomedical, Biomolecular, and Chemical Science, including the hot summer months, without access to drink- University of Western Australia, Stirling Highway, ing water. It is held (e.g. Louw and Seely 1982) that in such Crawley, WA 6009, Australia environments large mammals employ the physiological M. Shobrak mechanism termed “adaptive heterothermy”. During adap- Biology Department, Science College, tive heterothermy body heat is stored during the day, with a Taif University, PO Box 888, Taif, Saudi Arabia consequent rise in body temperature, reducing both heat 123 1112 J Comp Physiol B (2010) 180:1111–1119 gain and evaporative heat loss, and so conserving body Arabian oryx (Ostrowski et al. 2003). The oryx displayed water (Schmidt-Nielsen et al. 1957). This stored heat then an amplitude of the body temperature rhythm which can be dissipated non-evaporatively when heat load is increased from 1.5 § 0.6°C in winter to 4.1 § 1.7°C in lower at night. However, the concept of adaptive hetero- summer (Ostrowski et al. 2003) but the recordings of body thermy also incorporates a reduction of metabolic heat gen- temperature by radiotelemetry, especially at night, were eration and conservation during the night, so that body sporadic, and night recordings and daytime recordings were temperature falls more than normal, and the temperature not contiguous and may not have been made on the same rise of the following day starts from a lower base. Animals animals. Continuous recordings of body temperature have using adaptive heterothermy therefore would display an been obtained for the smaller Arabian sand gazelle (Gazella increase in the daily amplitude of the body temperature subgutturosa marica), living in the same habitat as the rhythm, without the mean daily body temperature necessar- oryx, and the daily amplitude of the body temperature ily changing. rhythm increased from only 1.7 § 0.3°C in winter to The phenomenon of adaptive heterothermy was Wrst 2.6 § 0.8°C in summer (Ostrowski and Williams 2006). So demonstrated more than 50 years ago by Schmidt-Nielsen the amplitude of the daily rhythm of body temperature for et al. (1957). They demonstrated, in experiments on two both of these species, in spite of being exposed to air juvenile dromedary camels (Camelus dromedarius) housed temperatures exceeding 40°C, was substantially lower than in an outdoor yard, that the daily rhythm of rectal tempera- the 6°C variation originally reported for captive camel ture more than doubled when the camels were deprived of (Schmidt-Nielsen et al. 1957), eland and beisa oryx (Taylor drinking water in summer. Indeed, on 1 day, the tempera- 1969). Also, contrary to Schmidt-Nielsen et al. (1957), ture of one camel varied by 6.2°C, increasing from 34.5°C Ostrowski and Williams (2006) presented no evidence that in the morning to 40.7°C in the evening. The ability to the availability of drinking water inXuenced body tempera- employ adaptive heterothermy subsequently has been ture in their animals, concluding that “There exists no attributed to several species of African antelope (Taylor evidence that free-living gazelles or oryx that used hetero- 1970a, b) even when they had adequate water. Despite hav- thermy were suVering from dehydration in either study”. ing free access to water, the rectal temperatures of both However, since neither study measured dehydration state eland (Tragelaphus oryx) and beisa oryx (Oryx gazella during the summer nor winter months, it remains possible beisa) rose by more than 6°C when exposed to an ambient that the animals’ body water status was compromised. temperature of 40°C for 12 h (Taylor and Lyman 1967; Whether hydration state inXuences heterothermy in free- Taylor 1969). All the early studies purporting to demon- living large antelope therefore remains uncertain. As part of strate adaptive heterothermy in antelope, however, were an observational study to further investigate the physiology conducted in artiWcial circumstances, on individual or small of Arabian oryx living free in the hyper-arid desert of Saudi groups of animals in climatic chambers or other conWned Arabia, we obtained the Wrst continuous measurements of spaces. Once technological advances allowed remote mea- body temperatures from this species. We also obtained data surement of body temperature in free-living animals, it from one captive oryx, exposed to similar climatic condi- became apparent that, at the same ambient temperature at tions but with free access to drinking water. Our data, which conWned mammals demonstrated adaptive hetero- obtained over a full year, conWrm the heterothermy reported thermy, Wve species of free-living African ungulates, with by Ostrowski et al. (2003), but support Schmidt-Nielsen free access to water, did not do so, namely the eland (Fuller et al.’s (1957) concept that large amplitudes in the daily et al. 1999), zebra (Equus burchelli; Fuller et al. 2000), rhythm of body temperature arise as a consequence of springbok (Antidorcas marsupialis; Mitchell et al. 1997; water limitation. Such conclusions are based on assessment Fuller et al. 2005; Hetem et al. 2009), black wildebeest of environmental water availability, rather than the mea- (Connochaetes gnou; Jessen et al. 1994) and gemsbok surement of dehydration state of the oryx. (Oryx gazelle gazella; Maloney et al. 2002). A possible explanation for the discrepancy between conWned and free- living ungulates was that the heterothermy measured in Materials and methods conWned ungulates was an artefact of conWnement, in animals deprived of their full suite of behavioural thermoregu- Animals and habitat lation (Fuller et al. 1999). It remains possible that, at ambient temperatures higher The study took place between March 2006 and April 2007 than those experienced by the African ungulates in the Weld, within the 2,200 km2 Mahazat as-Sayd Protected Area large ungulates would demonstrate adaptive heterothermy, (28°15Ј N, 41°40Ј E) in the open steppe desert in Saudi especially when confronted with water shortage. Indeed, a Arabia that is both the historical and current habitat for recent study has reported heterothermy in free-living Arabian oryx (Oryx leucoryx, Pallas 1777). Three male and 123 J Comp Physiol B (2010) 180:1111–1119 1113 two female adult, wild-born oryx were captured in the pro- 2 weeks of recovery, the Wve oryx in the Mahazat as-Sayd tected area in mid-March 2006. The oryx were habituated Protected Area were released into a 2 km2 fenced enclosure in outdoor pens for 2 weeks to reduce potential peri-opera- with natural forage and water available ad libitum. Ten days tive stress. An additional male oryx was obtained from the later they were allowed to enter and range freely within the breeding herd at the National Wildlife Research Center in Mahazat as-Sayd Protected Area where they were left undis- Taif (21°15Ј N, 40°42Ј E). This male oryx remained in Taif turbed, apart from the occasional serendipitous visual con- in a partially open pen throughout the 11-month study tact by rangers. The Wve oryx separated from each other period, with lucerne and water available ad libitum. All shortly after their release, with some joining other oryx. experimental procedures were approved by the Animal One of our oryx died, of unknown causes, 10 months Ethics Screening Committee of the University of the after surgery. A year after surgery, the four remaining oryx Witwatersrand (protocol no. 2005/87/5). were tracked, captured and transported to the holding pens. Those oryx, and the one at Taif, were anaesthetized again Surgery and the data loggers were removed under a surgical proce- dure similar to that used for the original implantation. The The oryx, at both locations, were darted and anaesthetized surgical wounds had healed and there were no signs of in the holding pens with etorphine hydrochloride (2.5 mg infection. Most of the loggers were found in the pelvic intramuscularly (IM), M99, C-Vet, Leyland, UK) and, once canal and were not encapsulated in adhesive tissue. After recumbent, were transported to a temporary operating the- 2 weeks of recovery in pens, the four free-living oryx were atre within 200 m of the pens. There the oryx were placed re-released into the Mahazat as-Sayd Protected Area. in sternal recumbency, with their heads elevated. Oryx were intubated and anaesthesia was maintained with 2–6% Data loggers isoXurane (Aerrane, Astra Zeneca, Johannesburg, South Africa), administered in 100% oxygen. Respiratory rate, The thermometric data loggers (StowAway XTI, Onset heart rate, arterial oxygen saturation, and rectal temperature Computer, Pocasset, MA, USA) used to measure abdomi- were monitored throughout the surgery. nal temperature had outside dimensions of »50 £ Using sterile surgical procedures, we implanted temper- 45 £ 20 mm and a mass of »40 g when covered in inert ature-sensitive data loggers (see below) into the abdominal wax (EXP978 Sasol, Johannesburg, South Africa). The log- cavity of the oryx, via an incision in the paralumbar fossa. gers had a resolution of 0.04°C and measurement range The loggers were dry-sterilized in formaldehyde vapour from +34 to +46°C. The recording interval was set at before implantation. After administering a local anaesthetic 15 min. The loggers were calibrated against a high-accu- [3 ml 2% lignocaine hydrochloride, Bayer Animal Health racy thermometer (Quat 100, Heraeus, Hanau, Germany) in (Pty) Ltd. Isando, South Africa] subcutaneously (SC), we an insulated water bath. After calibration, the loggers shaved the incision site and sterilized it with povidone recorded temperature to an accuracy of better than 0.05°C. iodine antiseptic (Vetedine, Vetoquinol, Lure, France). After implantation of the data logger, wounds were sutured Climatic data measurements and then treated with a topical antiseptic spray (Necrospray, Centaur Labs, Johannesburg, South Africa). The oryx We collected climatic data from a portable weather station received a long-acting antibiotic (450 mg IM, penicillin, erected near the Mahazat as-Sayd Protected Area, at the Norocillin La, Norbrook Laboratories Ltd., Newry, North- Saja/Umm ar-Rimth Protected Area (23°22Ј N, 42°45Ј E), ern Ireland), a non-steroidal anti-inXammatory analgesic and also at the National Wildlife Research Center in Taif. (100 mg IM, phenylbutazone, dexaphenylarthrite injectable We recorded wind speed (m s¡1), solar radiation (W m¡2), solution, Vetoquinol Veterinary Pharmaceuticals, Cedex, dry-bulb temperature (°C) and relative humidity (%). We France), a long-acting parasiticide (2 ml SC, Ivermectin, also recorded black globe temperature (°C) and rainfall on Noromectin, Norbrook Laboratories Ltd., Newry, Northern site at the Mahazat as-Sayd Protected Area for the duration Ireland) and a multivitamin (9 ml IM, Multivit injectable of the study. Photoperiod was calculated as the daily diVer- solution, Univet Ltd., Ireland). Before anaesthesia was ter- ence between sunrise and sunset times, which were minated, we Wtted a neck collar (MOD-500 Telonics, Inc. obtained from the US Naval Observatory website (http:// Mesa, AZ, USA), containing a tracking radio transmitter. aa.usno.navy.mil/data/). Following surgery, the oryx were transported back to their pens, where they became ambulatory within »10 min Data analysis after the eVect of etorphine was reversed with diprenorphine hydrochloride (7.5 mg intravenously, M5050, C-Vet, Ley- We arbitrarily deWned four seasonal periods, based on pre- land, UK). The oryx at Taif remained in its pen. After vailing climatic conditions of ambient temperature and 123 1114 J Comp Physiol B (2010) 180:1111–1119

Table 1 Environmental conditions (mean § SD) during the four seasonal periods in which the oryx were living free in the Mahazat as-Sayd Protected Area Warm-wet Hot-dry Warm-dry Cool-dry (April–May) (June–August) (September–November) (December–February)

Globe temperature (°C) 24-h mean 33.8 § 3.4a 37.9 § 1.7b 32.1 § 3.8a 23.3 § 3.6c 24-h minimum 18.5 § 3.3a 22.9 § 2.5b 16.4 § 3.8c 10.3 § 3.4d 24-h maximum 53.3 § 5.2a 57.1 § 3.9b 52.3 § 5.4a 42.6 § 6.1c Air temperature (°C) 24-h mean 27.7 § 4.0a 34.0 § 1.5b 26.1 § 4.6a 15.4 § 3.7c 24-h minimum 19.9 § 4.2a 25.7 § 2.3b 18.2 § 3.8a 8.9 § 3.6c 24-h maximum 34.1 § 4.0a 40.4 § 1.3b 33.2 § 4.8a 22.0 § 4.3c Mean 24-h vapour pressure (kPa) 0.79 § 0.27a 0.57 § 0.12b 0.68 § 0.30a 0.72 § 0.24a Mean 24-h wet-bulb globe temperature (°C) 20.5 § 1.9a 22.1 § 0.9b 18.7 § 1.6c 13.0 § 2.9d Mean 24-h wind speed (m s¡1)4.4§ 1.4a,b 3.9 § 1.0a 3.9 § 1.1a 4.5 § 1.4b Mean 24-h radiation (W m¡2)287§ 48a 298 § 26a 233 § 33b 200 § 40c Total rainfall (mm) 24 0 0 0 Mean time of sunrise 05:36 § 0:15 05:29 § 0:12 06:16 § 0:16 06:58 § 0:07 Mean time of sunset 18:48 § 0:10 19:04 § 0:11 17:48 § 0:25 17:40 § 0:18 Values with diVerent superscript letters diVered signiWcantly (P <0.05) rainfall (Table 1), which we called “warm-wet” (April– mean minimum body temperature, within each 24-h period, May), “hot-dry” (June–August), “warm-dry” (September– against the concurrent minimum and maximum 24-h air November) and “cool-dry” (December–February). temperature, and maximum 24-h body temperature against Although, we called the April to May period “warm-wet” the concurrent maximum 24-h air temperature. We Wtted only 24 mm of rain fell during this period. No rain fell for simple non-linear regressions, in the form of best-Wt power the remainder of the study. We averaged each climatic vari- curves, where appropriate. To test the eVect of season on able over successive 24-h periods and compared seasonal the amplitude of the body temperature rhythm, we corre- periods using a one-way ANOVA. For successive 24-h lated the amplitude of the 24-h rhythm of body tempera- periods, we calculated the mean, minimum, maximum and ture, averaged for all Wve oryx, against photoperiod. We amplitude of the daily rhythm of body temperature of the used partial correlation coeYcients to test whether the oryx. We averaged the body temperature parameters for eVect of season on the amplitude of the body temperature each oryx, for each seasonal period, and performed rhythm was independent of mean air temperature. repeated-measures ANOVAs to test for diVerences in the Statistical analyses were performed using GraphPad body temperature proWle across the four periods, with New- Prism (version 4.00 for Windows, GraphPad Software, San man–Keuls multiple comparison tests to identify sources of Diego, CA, USA). Values are expressed as mean § SD and diVerences. P < 0.05 was considered signiWcant. To further assess the inXuence of the environment on body temperature patterns, we correlated the various parameters of body temperature with the corresponding Results prevailing environmental conditions. Mean 24-h dry-bulb air temperature was correlated with the mean 24-h globe Climate temperature (r2 =0.85, P < 0.0001). Since oryx seek shade during the heat of the day (Stanley Price 1989), we consid- Environmental conditions over the four seasonal periods ered dry-bulb air temperature to provide a better approxi- are shown in Table 1. The mean, minimum and maximum mate of operative temperature than did globe temperature air and globe temperatures were signiWcantly higher during measured in the sun. We therefore correlated the mean 24-h the hot-dry period, and signiWcantly lower during the cool- body temperature, averaged for all Wve oryx, with mean dry period, than all other periods. These variables did not 24-h air temperature, and the amplitude of 24-h rhythm of signiWcantly diVer between the warm-wet and warm-dry body temperature against both mean 24-h air temperature periods (Table 1). Air and black globe temperature varied and amplitude of 24-h air temperature. We also correlated as a function of time of day, peaking just after solar noon 123 J Comp Physiol B (2010) 180:1111–1119 1115

(12:00) and reaching a minimum just before sunrise. Solar warm hot warm cool warm hot warm cool radiation showed the expected bell-shaped distribution and 43 wet dry dry dry wet dry dry dry wind speed increased in the late afternoon. Rainfall totalled 41 31 mm over the study period, substantially lower than the 39 10-year average of 100 § 60 mm. April was the wettest 37 month (17 mm) but some rain fell during March (7 mm) 35 and May (7 mm). No rain fell during any of the dry periods. Body temperature (°C) 50 March was not included in our seasonal analysis, because 40 we considered it too soon after surgery and the oryx were 30 still penned. 20 10 0 Body temperature Air temperature (°C) A M J J A S O N D J F A M J J A S O N D J F The daily amplitude of the body temperature rhythm of the Month Month free-living oryx varied over the year, with a substantial Fig. 1 Time courses of body and air temperatures. The upper set pan- increase during the hot-dry period (June–August), which els show the original record of 15-min recordings of body temperature was not evident in the captive male oryx (Fig. 1, top pan- from a single free-living male oryx (left top panel) and the captive els). Indeed, at the same mean daily air temperature male oryx (right top panel), that had access to water ad libitum, over the 11-month study period (April 2006 to February 2007). The lower (30.9°C), over three hottest days in Taif during the hot-dry panels show the air temperature recorded at nearby weather stations, period, the amplitude of the daily rhythm of body tempera- over the same period. The dotted lines separate the data into the four ture of the captive oryx (2.3°C) lay outside the 95% conW- seasonal periods analysed, namely warm-wet, hot-dry, warm-dry and dence interval of the amplitude of the daily rhythm of body cool-dry temperature of the Wve free-living oryx (3.5–4.5°C). Although ambient temperature had an apparent eVect on the 42 daily amplitude of the body temperature rhythm (compare the amplitude during the hot-dry and the cool-dry period in 40 hot dry warm wet Fig. 2, left panels), ambient temperature was not the only 38 X cool dry warm dry factor to in uence that amplitude. During the warm-wet 36 and warm-dry periods, ambient temperatures were not V V 50 di erent (Table 1) but there was a di erence in the daily Body temperature (°C) amplitude of the body temperature rhythm arising primarily 40 hot dry warm wet because the free-living oryx displayed a lower minimum 30 cool dry 20 body temperature during the warm-dry than during the warm dry warm-wet period (Fig. 2, right panels). 10 0 A repeated-measures ANOVA revealed that the mean Air temperature (°C) 0:00 6:00 12:00 18:00 0:00 6:00 12:00 18:00 (F3,9 = 11.2, P = 0.002), minimum (F3,9 = 15.4, P = 0.0007), Time of day Time of day maximum (F3,9 =41.4, P < 0.0001) and amplitude (F3,9 = 23.0, P = 0.0001) of the daily rhythm of body temperature Fig. 2 Daily rhythm of body temperature (mean § SD). Twenty four- W diVered signiWcantly between periods (Fig. 3). Maximum hour body temperature rhythm, averaged for all ve free-living oryx, and the prevailing air temperature, over the four seasonal periods. body temperatures and the amplitude of the body tempera- Black and grey horizontal bars indicate night-time for each period ture rhythm were higher during the hot-dry period than during the other dry periods (Fig. 3c, d; Newman–Keuls: P < 0.05). While ambient temperatures did not diVer ables with air temperature (Fig. 4). Mean daily air tempera- between the warm-wet and warm-dry periods, the mini- ture accounted for 35% of the variability in mean daily mum body temperature was higher during the warm-wet body temperature (P < 0.0001, Fig. 4a). There was a dis- period than during the dry periods (Fig. 3b; Newman– junction in the relationship, however, above a mean air Keuls: P < 0.05). This lower minimum body temperature temperature of about 25°C, because mean daily body tem- during the warm-dry period resulted in a lower mean daily perature was lower during the warm-dry than during the body temperature and a larger amplitude of the body tem- warm-wet period, even though air temperatures were simi- perature rhythm during the warm-dry than during the lar (Fig. 2, right panels). Maximum daily air temperature warm-wet period (Fig. 3d). accounted for 85% of the variability in maximum daily To further assess the inXuence of ambient temperature body temperature, across all 313 days of the study, when a on body temperature, we correlated body temperature vari- best-Wt power curve was Wtted to the data (Fig. 4b). Above 123 1116 J Comp Physiol B (2010) 180:1111–1119

* maximum daily body temperature increasing because the a * * period was hot, and minimum daily body temperature 39 * decreasing because it was dry. One classic hallmark of adaptive heterothermy is a posi- tive correlation between the amplitude of the daily rhythm of body temperature and the amplitude of air temperature Mean 24-hour 38 rhythm. We found no such correlation (Fig. 4d). On the

body temperature (°C) other hand, mean daily air temperature accounted for 63% b * of the variability in the amplitude of the daily rhythm of 39 * W W ** body temperature when a best- t power curve was tted to the data (Fig. 4e). The amplitude of the daily rhythm of W 37 body temperature also was correlated signi cantly with photoperiod (P < 0.0001, Fig. 4f). Although the linear correlation between the amplitude of the daily rhythm of body Minimum body temperature (°C) 35 temperature and photoperiod was signiWcant, the relation- * ship was not straightforward. The amplitude of the daily c ** rhythm of body temperature was »2°C during the warm- 42 ** ** wet season, when photoperiod was »13 h, and increased to 41 »6°C as conditions became progressively hotter and drier, * which coincided with an increasing photoperiod. However, 40 the amplitude of the daily rhythm of body temperature did

Maximum body not return to the expected 2°C as day length shortened and temperature (°C) 39 conditions became cooler, even when photoperiod was less than 11 h during the cool-dry period near the end of the 9 * d ** study. A partial correlation coeYcient revealed that the * 7 ** amplitude of the daily rhythm of body temperature was * more dependent on mean daily air temperature than on pho- 5 toperiod, with the latter correlation disappearing when the 3 confounding eVect of mean daily air temperature was elim-

temperature (°C) inated (rxz.y = 0.061, t310 = 1.08, P = 0.28). Amplitude of body 1

warm wet hot dry warm dry cool dry Seasonal period Discussion Fig. 3 Characteristics of body temperature. a Mean, b minimum, c maximum, and d amplitude of the daily rhythm of body temperature Our study provides the Wrst continuous measurements of (mean § SD, n = 5) over the four seasonal periods, namely warm-wet body temperature, extending over many successive 24-h (black bar), hot-dry (hatched bar), warm-dry (white bar) and cool-dry cycles, of free-living Arabian oryx in their natural habitat, (hatched bars). *P < 0.05; **P < 0.001, repeated-measures ANOVA the deserts of Saudi Arabia. The oryx displayed not only higher maximum daily body temperatures (41.1 § 0.3 vs. a maximum daily air temperature of 36°C, maximum daily 39.7 § 0.1°C) but also lower minimum daily body temper- body temperature increased by 0.25°C per 1°C increase in atures (36.1 § 0.3 vs. 36.8 § 0.2°C), resulting in larger maximum daily air temperature. Conversely, over the full daily amplitudes of the body temperature rhythm (5.0 § 0.5 range of air temperatures, minimum daily body temperature vs. 2.9 § 0.2°C), during the hot-dry period than during the showed a weak negative correlation with both minimum cool-dry period, a trend reported previously by Ostrowski daily air temperature (r2 =0.06, P < 0.0001) and maximum et al. (2003). The maximum amplitude of the daily rhythm daily air temperature (r2 =0.07, P < 0.0001). Minimum of body temperature reached 7.7°C for two of our oryx dur- daily body temperature increased with increasing maxi- ing the hot-dry period, the largest ever reported for any mum daily air temperature up to 30°C (r2 =0.47, large mammal. The oryx therefore displayed heterothermy, P < 0.0001, Fig. 4c) but above 30°C there was a disjunction but, contrary to the results obtained by Ostrowski et al. in the relationship as minimum body temperatures during (2003), with our continuous rather than intermittent mea- the warm-dry period were lower than during the warm-wet surements of body temperature, we found no correlation period. Thus, relative to other times of the year, homeo- between the amplitude of the daily rhythm of body temper- thermy was abandoned during the hot-dry period, with ature, and therefore the quantity of heat stored, and the 123 J Comp Physiol B (2010) 180:1111–1119 1117

40 a 42 b 41 39 40 38 39 warm-wet hot-dry Maximum body temperature (°C) temperature (°C) warm-dry

Mean 24-hour body 37 38 cool-dry

0 10 20 30 40 10 20 30 40 50 Mean air temperature ( °C) Maximum air temperature ( °C)

39 c d 38 6

37 4

36 2 Minimum body temperature (°C) temperature (°C)

35 Amplitude of body 0

10 20 30 40 50 5 10 15 20 25 Maximum air temperature ( °C) Amplitude of air temperature ( °C)

C) e f ° 6 6

4 4

2 2 temperature (°C) temperature ( Amplitude of body Amplitude of body 0 0

0 10 20 30 40 10 11 12 13 14 Mean air temperature ( °C) Photoperiod (h)

Fig. 4 Correlations of body temperatures with ambient temperatures body temperature rhythm did not correlate with the daily amplitude of and photoperiod (mean, n = 5, 313 days). a Mean daily body tempera- air temperature (r2 = 0.07) but correlated with both e mean daily air ture correlated positively and linearly with mean daily air temperature temperature by means of a best-Wt power curve (r2 = 0.63, (r2 = 0.35, y =0.02x + 38.0). b Maximum daily body temperature cor- y =2.8+1.3£10¡12x8) and f linearly with photoperiod (r2 = 0.31, related positively with maximum daily air temperature by means of a y =0.50x ¡ 2.5). However, there were two discrete relationships best-Wt power curve (r2 =0.85, y = 39.7 + 9.2£10¡14x8.2). c Minimum between body temperature and photoperiod at moderate air tempera- daily body temperature correlated weakly and inversely with maxi- tures and long photoperiods; curved arrows indicate the progression mum daily air temperature (r2 =0.07). d The daily amplitude of the of time amplitude of daily ambient temperature (Fig. 4c), implying minimum daily body temperatures (36.5 § 0.6 vs. that the degree of heterothermy was not driven by ambient 37.8 § 0.2°C), despite ambient temperature being similar. temperature but by other factors. Our data do not allow us to discern whether the low mini- It is clear from the results that ambient temperature alone mum daily body temperatures, which occurred in the early was not responsible for the heterothermy we observed in mornings, were incidental, resulting from water and food the oryx. Another variable that could account for the het- limitations in the desert environment, or adaptive, resulting erothermy in our oryx was the lack of water in a hot envi- from suppression of the set point for body temperature reg- ronment, leading both to a reduction in availability of ulation. Elucidating the exact role of water in determining drinking water and a decline in food quality and availabil- body temperature proWles of the oryx will require measure- ity. Under captive conditions, with food and water available ment of water loss, metabolic rate, or osmolarity. However, ad libitum, our single male oryx maintained an amplitude the presence of humans disrupts normal thermoregulation of the daily rhythm of body temperature of 2.6 § 0.5°C in wild animals (Recarte et al. 1998) and the technology to throughout the year, despite being exposed to ambient tem- measure these variables remotely in free-living animals peratures very similar to those of the free-living oryx. The does not yet exist. Though we could not quantify dehydra- free-living oryx displayed greater daily heterothermy dur- tion in our free-living oryx, previous studies showed that ing the warm-dry period than during the warm-wet period free-living oryx generally have low rates of water inXux (3.6 § 0.6 vs. 2.3 § 0.3°C), mostly as a result of lower during summer (Williams et al. 2001; Ostrowski et al. 123 1118 J Comp Physiol B (2010) 180:1111–1119

2002), and display higher haematocrit, plasma protein low body temperatures in the morning. Previous studies concentration and plasma osmolality during summer than proposed that the morning decrease in body temperature is during winter (Ostrowski et al. 2003). the result of vasodilation, which mixes cool peripheral The latent heat of evaporation of water is suYciently blood with that from the core (Schmidt-Nielsen et al. 1957; high, and the water vapour pressure of desert air suY- Zervanos and Hadley 1973; Brown and Dawson 1977; ciently low, for desert mammals to be able to dissipate Fuller et al. 1999; Maloney et al. 2004). Such a decrease in resting metabolic heat and environmental heat (even in body temperature may be programmed, in anticipation of direct solar radiation) by evaporation, and so avoid hyper- hot ambient conditions (Maloney et al. 2004), thus pre- thermia. However, the evaporation that is necessary emptively permitting additional storage of heat. However, depletes body water. At our experimental site, the oryx in our free-living oryx, minimum daily body temperatures had access to drinking water only in the ephemeral pools were low not only during the hot-dry period, but throughout that formed when the 31 mm of rain fell between March all dry periods, favouring the view that the exaggerated and May. During the hot-dry period (June–August) the minimum body temperature in the morning was related not oryx had no access to drinking water, and would have had to anticipated diurnal heat load, but was the consequence of to rely on metabolic water and the pre-formed water in the limited food and water availability during the dry seasons, sparse desert vegetation, itself desiccating at that time. which would reduce the metabolic rate of desert animals Consequently, the oryx would have beneWtted from trans- (Brosh et al. 1986; Merkt and Taylor 1994; Williams et al. ferring the responsibility for heat dissipation to non-evap- 2001; Ahmed and El Kheir 2004; Ostrowski et al. 2006a, orative avenues. Several antelope species reduce the rate b). Indeed, the low morning body temperatures in both kan- of evaporative water loss as they become dehydrated garoos (Dawson et al. 2007) and camels (Grigg et al. 2009) (Taylor 1970a; Maloiy 1973; Finch and Robertshaw 1979; probably were more related to compromised energy Baker 1989; Nijland and Baker 1992; Silanikove 1994; balance than to anticipating heat storage. Jessen et al. 1998), which results in an increased body We propose that, in hot-dry environments, the degree of temperature during heat exposure (Taylor 1969, 1970a, b; heterothermy employed is determined not by ambient tem- Finch and Robertshaw 1979; Jessen et al. 1998; Alamer peratures alone, but also by aridity. When our oryx had 2006). access to suYcient water, they maintained homeothermy One of the processes for transferring heat loss to non- rather than implementing heterothermy, even in environ- evaporative avenues is the employment of heterothermy. ments in which globe temperature exceeded 50°C. Indeed, The capacity to store heat, and therefore reduce evaporative when given the opportunity, other oryx increased their cooling, holds whether the heterothermy is a controlled water consumption during the summer months (Stanley thermoregulatory event that constitutes adaptive hetero- Price 1989) and preferentially selected areas where new thermy, or whether it results from failure of homeothermy rain had fallen (Corp et al. 1998). However, when our oryx (Mitchell et al. 2002). But failure of homeothermy places did not have access to suYcient water, they abandoned the animal at risk of damaging daytime hyperthermia in the homeothermy and displayed heterothermy, which presum- absence of a lower nadir of body temperature (a hallmark of ably would have allowed them to conserve body water. adaptive heterothermy). What distinguishes adaptive het- Whether such heterothermy is a programmed active pro- erothermy from dehydration-induced suppression of evapo- cess, implied by the term “adaptive heterothermy”, or rative heat loss is the amplitude of the body temperature whether it is fortuitous, resulting from failure of the energy rhythm increasing without the mean daily body temperature sources necessary for thermoregulation, remains to be necessarily increasing, because minimum body temperature investigated. is depressed (Mitchell et al. 2002). The term “adaptive” carries the implication that the depression of minimum Acknowledgments We thank the National Commission for Wildlife body temperature is an adaptive thermoregulatory event. Conservation and Development (NCWCD), Riyadh, Saudi Arabia, in particular the director His Royal Highness Prince Saud Al Faisal, the But neither our data, nor that of others, can distinguish current secretary-general, HH prince Bander Bin Saud, and the secre- between an active event and failure of cold defences in con- tary-general at the time the study was conducted, Professor AH Abuzi- ditions of energy, or perhaps water, deWcits. nada, for supporting the research. From the National Wildlife Research Under captive conditions of water deprivation, the drom- Center (NWRC), we are grateful to Dr Saud Anagariyah for his support in capturing the oryx and the current director, Ahmad Al Bouq. In edary camel (Schmidt-Nielsen et al. 1957; Schroter et al. addition, we thank the Mahazat as-Sayd Protected Area rangers for 1987), hartebeest (Alcelaphus buselaphus; Harthoorn et al. monitoring the animals, and the mammal keepers at NWRC for their 1970; Maloiy and Hopcraft 1971), Thomson’s gazelle help with animal handling and assistance during surgery. This research (Gazella thomsonii) and zebu steer (Bos primigenius indi- was funded by the National Research Foundation, South Africa, the University of the Witwatersrand Medical Faculty Research cus; Taylor 1970b) displayed an increased amplitude of the Endowment Fund and START/PACOM African PhD fellowship body temperature rhythm, at least partially resulting from awarded to RS Hetem. 123 J Comp Physiol B (2010) 180:1111–1119 1119

References Maloney SK, Fuller A, Mitchell G, Mitchell D (2002) Brain and arterial blood temperatures of free-ranging oryx (Oryx gazella). X Ahmed MMM, El Kheir IM (2004) Thermoregulation and water bal- P ügers Arch Eur J Phys 443:437–445 ance as aVected by water and food restrictions in Sudanese desert Maloney SK, Fuller A, Kamerman PR, Mitchell G, Mitchell D (2004) goats fed good-quality and poor-quality diets. Trop Anim Health Variation in body temperature in free-ranging western grey kan- Prod 36:191–204 garoo Macropus fuliginosus. Aust Mammal 26:135–144 Alamer M (2006) Physiological responses of Saudi Arabia indigenous Merkt JR, Taylor CR (1994) “Metabolic switch” for desert survival. goats to water deprivation. Small Rumin Res 63:100–109 Proc Natl Acad Sci USA 91:12313–12316 Baker MA (1989) EVects of dehydration and rehydration on thermo- Mitchell D, Maloney SK, Laburn HP, Knight MH, Kuhnen G, Jessen regulatory sweating in goats. J Physiol 417:421–435 C (1997) Activity, blood temperature and brain temperature of Brosh A, Shkolnik A, Choshniak I (1986) Metabolic eVects of infre- free-ranging springbok. J Comp Physiol B Biochem Syst Environ quent drinking and low quality feed on Bedouin goats. Ecology Physiol 167:335–343 67:1086–1090 Mitchell D, Maloney SK, Jessen C, Laburn HP, Kamerman PR, Mitch- Brown GD, Dawson TJ (1977) Seasonal variation in the body temper- ell G, Fuller A (2002) Adaptive heterothermy and selective brain atures of unrestrained kangaroos (Macropodidae: Marsupialia). cooling in arid-zone mammals. Comp Biochem Physiol B Bio- Comp Biochem Phys A Mol Integr Physio 56:59–67 chem Mol Biol 131:571–585 V Corp N, Spalton A, Gorman ML (1998) The inXuence of rainfall on Nijland MJ, Baker MA (1992) E ects of hydration state on exercise range in a female desert ungulate: the Arabian oryx (Oryx leuc- thermoregulation in goats. Am J Physiol Regul Integr Comp oryx) in the Sultanate of Oman. J Zool 246:369–377 Physiol 263:R201–R205 Dawson TJ, Blaney CE, McCarron HCK, Maloney SK (2007) Dehy- Ostrowski S, Williams JB (2006) Heterothermy of free-living Arabian dration, with and without heat, in kangaroos from mesic and arid sand gazelles (Gazella subgutturosa marica) in a desert environ- habitats: diVerent thermal responses including varying patterns in ment. J Exp Biol 209:1421–1429 X heterothermy in the Weld and laboratory. J Comp Physiol B Bio- Ostrowski S, Williams JB, Bedin E, Ismail K (2002) Water in ux and chem Syst Environ Physiol 177:797–807 food consumption of free-living oryxes (Oryx leucoryx) in the Finch VA, Robertshaw D (1979) EVect of dehydration on thermoreg- Arabian desert in summer. J Mammal 83:665–673 ulation in eland and hartebeest. Am J Physiol Regul Integr Comp Ostrowski S, Williams JB, Ismael K (2003) Heterothermy and the Physiol 237:R192–R196 water economy of free-living Arabian oryx (Oryx leucoryx). Fuller A, Moss DG, Skinner JD, Jessen PT, Mitchell G, Mitchell D J Exp Biol 206:1471–1478 (1999) Brain, abdominal and arterial blood temperatures of free- Ostrowski S, Mésochina P, Williams JB (2006a) Physiological adjust- ranging eland in their natural habitat. PXügers Arch Eur J Physiol ments of sand gazelles (Gazella subgutturosa) to a boom or bust 438:671–680 economy: standard fasting metabolic rate, total evaporative water Fuller A, Maloney SK, Kamerman PR, Mitchell G, Mitchell D (2000) loss, and changes in the sizes of organs during food and water Absence of selective brain cooling in free-ranging zebras in their restriction. Physiol Biochem Zool 79:810–819 natural habitat. Exp Physiol 85:209–217 Ostrowski S, Williams JB, Mésochina P, Sauerwein H (2006b) Physi- Fuller A, Kamerman PR, Maloney SK, Matthee A, Mitchell G, ological acclimation of a desert antelope, Arabian oryx (Oryx Mitchell D (2005) A year in the thermal life of a free-ranging leucoryx), to long-term food and water restriction. J Comp Phys- herd of springbok Antidorcas marsupialis. J Exp Biol iol B Biochem Syst Environ Physiol 176:191–201 208:2855–2864 Recarte JM, Vincent JP, Hewison AJM (1998) Flight responses of park Grigg G, Beard L, Dorges B, Heucke J, Coventry J, Coppock A, Blom- fallow deer to the human observer. Behav Process 44:65–72 berg S (2009) Strategic (adaptive) hypothermia in bull dromedary Schmidt-Nielsen K, Schmidt-Nielsen B, Jarnum SA, Houpt TR (1957) camels during rut; could it increase reproductive success? Biol Body temperature of the camel and its relation to water economy. Lett 5:853–856 Am J Physiol 188:103–112 Harthoorn AM, Finch VA, Hopcraft D, McGinnis SM (1970) Adapta- Schroter RC, Robertshaw D, Baker MA, Shoemaker VH, Holmes R, tion to solar radiation by African large herbivores. J S Afr Vet Schmidt-Nielsen K (1987) Respiration in heat stressed camels. Med Ass 41:17–24 Resp Physiol 70:97–112 Hetem RS, de Witt BA, Fick LG, Fuller A, Kerley GIH, Meyer LCR, Silanikove N (1994) The struggle to maintain hydration and osmoreg- Mitchell D, Maloney SK (2009) Body temperature, thermoregu- ulation in animals experiencing severe dehydration and rapid latory behaviour and pelt characteristics of three colour morphs of rehydration: the story of ruminants. Exp Physiol 79:281–300 springbok (Antidorcas marsupialis). Comp Biochem Phys A Mol Stanley Price MR (1989) Animal reintroductions: the Arabian oryx in Integr Physiol 152:379–388 Oman. Cambridge University Press, Cambridge Jessen C, Laburn HP, Knight MH, Kuhnen G, Goelst K, Mitchell D Taylor CR (1969) The eland and the oryx. Sci Am 220:88–95 V (1994) Blood and brain temperatures of free-ranging black wilde- Taylor CR (1970a) Dehydration and heat: e ects on temperature regu- beest in their natural environment. Am J Physiol Regul Integr lation of East African ungulates. Am J Physiol 219:1136–1139 V Comp Physiol 267:R1528–R1536 Taylor CR (1970b) Strategies of temperature regulation: e ect on evap- Jessen C, Dmi’el R, Choshniak I, Ezra D, Kuhnen G (1998) EVects of oration in East African ungulates. Am J Physiol 219:1131–1135 dehydration and rehydration on body temperatures in the black Taylor CR, Lyman CP (1967) A comparative study of the environmen- Bedouin goat. PXügers Arch Eur J Phys 436:659–666 tal physiology of an East African antelope, the eland and the Her- Louw GN, Seely MK (1982) Ecology of desert organisms. Longman eford steer. Physiol Zool 40:280–295 Sci & Tech, New York Williams JB, Ostrowski S, Bedin E, Ismail K (2001) Seasonal variation X Maloiy GM (1973) The water metabolism of a small East African ante- in energy expenditure, water ux and food consumption of Ara- lope: the dik–dik. Proc R Soc Lond B Biol Sci 184:167–178 bian oryx Oryx leucoryx. J Exp Biol 204:2301–2311 Maloiy GM, Hopcraft D (1971) Thermoregulation and water relations Zervanos SM, Hadley NF (1973) Adaptational biology and energy of two East African antelopes: the hartebeest and impala. Comp relationships of the collared peccary (Tayassu tajacu). Ecology Biochem Physiol A Mol Integr Physiol 38:525–534 54:759–774

123