ORIGINAL PAPER

D.J. McCafferty Æ S. Moss Æ K. Bennett Æ P.P. Pomeroy

Factors influencing the radiative surface temperature of grey seal(Halichoerus grypus) pups during early and late lactation

Received: 24 February 2005 / Revised: 6 May 2005 / Accepted: 12 May 2005� Springer-Verlag 2005

Abstract The aim of this study was to examine the var-iation in body surface temperature of grey seal (Halic-hoerus grypus) pups throughout lactation in response todifferent environmental conditions. Radiative surfacetemperatures (Tr, �C) of pups were measured on the Isleof May (56�11¢N, 02�33¢W), southeast Scotland from 29October to 25 November 2003. Records were obtainedfrom a total of 60 pups (32 female and 28 male) fromthree different pupping sites during early and late lac-tation. Pups were sheltered from high wind speeds butair temperature, humidity and solar radiation at pup-ping sites were similar to general meteorological condi-tions. The mean Tr of all pups was 15.8�C (range 7.7–29.7�C) at an average air temperature of 10.2�C (range6.5–13.8�C). There was no difference in the mean Tr ofpups between early and late lactation. However, the Tr

varied between different regions of the body with hindflippers on average 2–6�C warmer than all other areasmeasured. There was no difference in mean Tr of maleand female pups and pup body mass did not account forthe variation in Tr during early or late lactation.Throughout the day there was an increase in the Tr ofpups and this explained 20–28% of the variation in Tr

depending on stage of lactation. There was no differencein the mean Tr of pups between pupping sites or asso-ciated with different substrate types. Wind speed and

substrate temperature had no effect on the Tr of pups.However, solar radiation, air temperature and relativehumidity accounted for 48% of the variation in mean Tr

of pups during early lactation. During late lactation airtemperature and solar radiation alone accounted for43% of the variation in Tr. These results indicate thatenvironmental conditions explain only some of thevariation in Tr of grey seal pups in natural conditions.Differences in Tr however indicate that the cost ofthermoregulation for pups will vary throughout lacta-tion. Further studies examining intrinsic factors such asblubber thickness and activity levels are necessary beforedeveloping reliable biophysical models for grey seals.

Keywords Grey seal Æ Halichoerus grypus Æ Radiativetemperature Æ Microclimate Æ Lactation

Introduction

Grey seals (Halichoerus grypus) occur in temperate andsub-arctic coastal waters on both sides of the NorthAtlantic. Western Atlantic and Baltic populations givebirth in the coldest months of the year, with peak pup-ping occurring in mid-January and early March,respectively. In the eastern Atlantic, pupping occursfrom the end of August to early October on the SWcoast of Britain to December on the northern coast ofNorway and in the White Sea, Russia (Hansen andLavigne 1997). Grey seals are well adapted to low airtemperatures. The lower critical temperature of metab-olism of dry adult grey seals in air is �11�C (Folklowand Blix 1987) while for fasting grey seal pups it is �7�C(Hansen and Lavigne 1997). The latter corresponds withthe mean winter air temperature at their northern limit.

It is not clear how climate influences grey seals at thesouthern edge of their range in the UK. Evidence thattheir breeding behaviour may be limited by warm con-ditions comes from observations that access to the sea orfreshwater is an important factor in the selection of

Communicated by G. Heldmaier

D.J. McCafferty (&)Department of Adult & Continuing Education,University of Glasgow, St Andrew’s Building,11 Eldon Street, Glasgow, G3 6NH, UKE-mail: d.mccafferty@educ.gla.ac.ukTel.: +41-141-3301835Fax: +41-141-3301821

S. Moss Æ K. Bennett Æ P.P. PomeroySea Mammal Research Unit, Gatty Marine Laboratory,University of St Andrews, St Andrews, Fife, KY16 8LB, UK

J Comp Physiol B (2005) 175: 423–431DOI 10.1007/s00360-005-0004-4

breeding sites (Anderson et al. 1975; Twiss et al. 2000).Although cooling in water may improve thermal con-ditions for adult grey seals on land, this may not nec-essarily be true for young pups that have limitedmobility. Grey seal pups are suckled for about 3 weeksand for the following 2–4 weeks weaned pups fast onland before going to sea (Hall 2002). At birth, pups havea thick white lanugo (pelage) and relatively thin blubberlayer but by the time they approach weaning they arewell insulated with thick blubber and have lost most ifnot all of their lanugo.

One way of examining the influence of climate ongrey seal pups may be to model the effects of envi-ronmental conditions on metabolic heat production. Incontrast to terrestrial animals, there have been rela-tively few attempts to model metabolic heat exchangeof marine mammals. One approach has been toexamine heat transfer through blubber (Luecke et al.1975; Hokkanen 1990; Lavigne et al. 1990; Folkowand Blix 1992; Watts et al. 1993). An alternative hasbeen to model conduction through the skin and con-vection from the skin to the environment (Gallivanand Ronald 1979; Ryg et al. 1988; Boily 1995) andthese models have then been used to examine heatexchange of pinnipeds in natural conditions (Trites1990; Hind and Gurney 1998).

Kvadsheim et al. (1997) showed that conductive heatflux models overestimated metabolic rates in harp seals(Phoca groenlandica) due to their failure to account forvariation in the distribution of blubber. Boily et al.(2000) took temperature measurements and found thatcutaneous heat flux models tended to underestimatetotal heat production in the same species. All currentmodels are particularly sensitive to measurements oftemperature at the skin surface, both in air and water.This is particularly problematic as skin and pelagetemperature in marine mammals can vary according toanatomical position, in relation to the dive cycle or inresponse to drying out (Williams et al. 1999; Mauck etal. 2003). In addition, a relatively small number ofmeasurements from different areas of the body may notaccount for the variation in surface temperature acrossthe entire body surface and thereby provide unreliabledata for modelling metabolic rate.

Given that empirical data on the temperature dis-tribution of the body surface are required for modellingmetabolic heat exchange of seals, the aim of this studywas to examine the variation in body surface temper-ature of grey seal pups in natural conditions. Mea-surements of the radiative surface temperature (Tr, �C)of pups were taken using an infrared thermometer asthis provided a non-invasive method in the field. Theobjectives of this study were to (1) record the envi-ronmental conditions experienced by grey seal pups ondifferent pupping sites, (2) examine the variation in Tr

across the body, (3) determine if the Tr of these areaschanged throughout lactation and (4) establish whichenvironmental factors influence the Tr of grey seal pupsduring lactation.

Theory

A grey seal pup on land will exchange heat by conduc-tion to the ground, by convection and radiation to thesurroundings and by latent heat transfer. As phocidseals do not have sweat glands (Tarasoff and Fisher1970) latent heat transfer mainly occurs through respi-ratory water loss or through evaporation if the coat iswet. Assuming that the seal is inactive and is at steadystate the metabolic heat production, M (W) can be de-scribed using the equation modified from Hind andGurney (1998):

Mð1� cÞ ¼ A1hcðTc � TsaÞ þ A2hcðTc � TsgÞ ð1Þ

where latent heat loss through respiration depends onthe rate of respiration and is proportional to metabolicrate and therefore can be represented as a fraction ofmetabolic heat, c (=0.1, Hind and Gurney 1998). Heatexchange from the area of the body in contact with air,A1 (m

2) is determined by the heat transfer coefficient ofthe blubber and coat, hc (W m�2 �C�1) and is propor-tional to the temperature difference between core andcoat surface in contact with air,Tc�Tsa (�C). Heat lossfrom the area of the body in contact with the ground A2

(m2) is also controlled by the heat transfer coefficient ofthe blubber and coat but is proportional to the tem-perature difference between core and coat in contactwith the ground, Tc�Tsg (�C). For a pup that maintainsa relatively constant core temperature changes in coattemperature will determine the rate of heat loss from thebody. An increase in coat temperature will reduce heatloss and thereby reduce any additional metabolicrequirements for thermoregulation. Conversely a de-crease in coat temperature will increase heat loss and themetabolic cost of thermoregulation.

Seal coat temperatures have been measured usingtemperature sensors that record the temperature at aknown point on or in the coat (e.g. Worthy 1991).Infrared thermometers and thermal imaging camerashave also been used to measure the radiative surfacetemperature, Tr of seals (e.g. Ørtisland et al. 1978;Maucket al. 2003). These instruments measure the amount ofnear infrared radiation emitted by a surface and thenconvert this to a Tr reading according to the Stefan–Boltzmann equation for a blackbody emitter such thatthe energy emitted, R (W m�2) is proportional to thefourth power of its absolute temperature, T (K): R=�rT4 where � is the emissivity of the surface and r is theStefan–Boltzmann constant (5.67 · 10�8 W m�2 K�4).Most animal coats are considered to be blackbodyemitters with emissivity values ranging between 0.98 and1.0 (Campbell and Norman 1998).

The radiant flux emitted by bare-skinned animals isgoverned by the skin surface temperature but theradiation emitted by animals with coats, such as sealsmay originate either from the skin, if this is incom-pletely obscured by fibres, or from the fibre elementsthemselves. The radiating surfaces of the fibres are at

424

a range of temperatures determined by the tempera-ture gradient between the skin and the coat surface.The exchange of radiation may be further complicatedby external fluxes that contribute to the heat balanceof the fibres. For an animal with thick fur, such as aseal pup the surface measured by an infrared ther-mometer is some layer within the coat. This radiativesurface temperature is typically several millimetresbeneath the physical surface of the coat and may beregarded as the site where the various fluxes of energybetween the animal and its environment are parti-tioned. The equilibrium temperature of this surface isdetermined by the loss of heat from radiation andconvection to the surroundings, the conduction of heatthrough the coat and the exchange of thermal andshort wave radiation (Cena 1974). A measurement ofTr therefore provides the temperature of radiant heatexchange and an approximation of the coat tempera-ture that determines convective heat loss between theanimal and the environment.

Previous studies on harp seals have shown thatincreasing air temperature increases the radiative surfacetemperature of the coat (Ørtisland et al. 1978). Solarradiation also increases the temperature of the coat bysurface heating (Cena 1974). Wind brings about areduction in coat surface temperature as wind penetratesthe coat to greater depths and removes heat by forcedconvection (Ørtisland et al. 1978; Treagar 1965). Airspaces within a coat contain water vapour that is inequilibrium with water absorbed or adsorbed on fibresof the coat. Sorption and de-sorption of water are exo-thermic and endothermic processes, respectively. Achange in relative humidity within a coat is thereforeassociated with a transient heat flux that can give rise tosmall changes in coat temperature (Gatenby et al. 1983).However, greater changes in the surface temperature ofanimal coats occur through evaporative cooling whenthe coat is wet (McCarthur and Ousey 1994).

Material and methods

Study site

This study was undertaken on the Isle of May(56�11¢N, 02�33¢W), southeast Scotland from 29October to 25 November 2003. Fieldwork was con-ducted at three main pupping sites, at Rona, Tarbetand Kirkhaven (see Pomeroy et al. 2000 for map oflocations). Rona is a relatively flat region of the islandconsisting of a level area (Rona Top) with rock ledgesand ridges on the seaward slopes (West Rona Geoand Rona Rocks). Tarbet consists of a steep sidedravine with cliffs leading to the sea (Tarbet Hole) andbroken rock and soil-covered slopes (Tarbet Slope).Kirkhaven consists of a sheltered rocky cove with rockand beach and flat surrounding slope (the Loan). Theground surface consists mostly of rock or sand, baresoil and in places thinly covered low vegetation.

Pup measurements

Measurements were made of pups during captures ofmother–pup pairs as part of a long-term study on greyseal breeding success and this provided the mosteffective way of taking detailed measurements of pups.Mother–pup pairs were captured 1–2 days followingbirth (early lactation) and prior to the end of lactation(late lactation). At each stage of lactation, femaleswere caught twice for blood sampling on the same dayat an average interval of 4 h. Females were anaesthe-tised using a pressurised dart for each capture(Pomeroy et al. 1999). Following darting the mother–pup pair was left undisturbed and monitored from adistance until the mother was fully immobilised (usu-ally 10 min). The mother–pup pair was then ap-proached and the Tr of the pup (±0.1�C) wasrecorded, using a hand held infrared thermometer(Raynger ST80, Raytek, Milton Keynes, UK) from amaximum distance of 1 m (usually 20–50 cm). Theinfrared thermometer had a spectral response ofwavelength 8–14 lm and was fitted with a circularlaser that showed the approximate area being mea-sured. This had a distance:diameter ratio of 30:1 sothat at 1 m distance Tr was recorded from a circulararea of 3.3 cm diameter. This method allowed mea-surements to be taken with minimal disturbance to thepup. Tr readings were taken from the head, mid-dor-sal, lower dorsal, fore and hind flippers and upperventral side (Fig. 1) using an emissivity for fur of 0.99(Campbell and Norman 1998). The pup was thenweighed ±0.5 kg on a spring balance (Salter IndustrialMeasurements Ltd, West Bromich, UK), sexed andgiven a numbered plastic tag (Dalton, Henley-on-Thames, UK) in a hind flipper before returning it toits mother. Pup age was estimated by coat classifica-tion for grey seals (Kovacs and Lavigne 1986) and foranalysis pups were divided into two groups: early

Fig. 1 Body regions (and order of measurement) on a grey seal pup(stage II) where surface temperatures were recorded: 1 head, 2 mid-dorsal, 3 lower dorsal, 4 hind flipper, 5 fore flipper and 6 upperventral

425

lactation (stages I and II, age £ 10 days) and latelactation (stages III and IV, age>10 days).

Microclimate measurements

Air temperature (±0.1�C), relative humidity(RH±1%) and wind speed (±0.1 m s�1, minimumspeed 0.3 m s�1) were measured using a hand heldenvironmental meter (Kestrel 3000, Lymington, UK).Global solar radiation (±0.1 W m�2) was measuredusing a pyranometer (SKS1110, Skye Instruments Ltd,Powys, UK), which was recorded on a hand helddigital multimeter (T110B, Beckman Coulter Ltd,High Wycombe, UK). The ground temperature wasrecorded with the IRT using an emissivity of 0.95 fornatural ground surfaces (Campbell and Norman1988). Air temperature, relative humidity and solarradiation were determined within 10–20 m of the pupprior to capture. Wind speed was measured at the sitewhere the pup was lying at approximately 10 cmabove the ground surface, which was equivalent tohalf the height of the animal. The wind speed at themid body height of the female (approximately 0.5 m)was recorded where possible.

General meteorological conditions were recorded onan automatic weather station (Minimet Skye Instru-ments Ltd, Powys, UK) situated at the centre of theisland (50 m a.m.s.l). Air temperature (±0.1�C), relativehumidity (±1%) and global solar radiation(±0.1 W m�2) were measured at 1.2 m and wind speed(±0.1 m s�1, minimum speed 0.1 m s�1) and winddirection (degrees) at 2 m above the ground surface.

Instrument calibration and statistical analysis

The infrared thermometer and weather station instru-ments were newly calibrated by the manufacturer. Thehand held environmental meter was checked againstreadings of temperature, humidity and wind speedfrom the weather station. Statistical analysis was

computed using SPSS Version 9.0 (SPSS Inc, Chicago,USA).

Results

Study animals

Measurements were obtained from 31 pups (17 femalesand 14 males) during early lactation and from 29 pups(15 females and 14 males) at late lactation. Twentyindividuals were measured at both stages of lactation.The mean±SE mass of pups was 19.0±0.63 kg and37.9±0.96 kg during early and late lactation, respec-tively. It was not always possible to obtain measure-ments from each pup when females were captured,however, two records were taken on the same day.Mean values were used for analysis, unless otherwisestated.

Weather and microclimate experienced by pups

The weather station recorded a mean air temperature of9.1±0.03�C, relative humidity of 87.3±0.09%, windspeed of 6.6±0.05 m s�1, winddirection of 193±1.1� andmean daytime solar radiation of 109.4±2.39 W m�2 .There were 14 days of rain (‡0.2 mm) with a total rainfallof 24.5 mm throughout the 28-day period.

At pupping sites the mean air temperature recordedwas 10.2±0.18�C, ground temperature was 8.6±0.39�C,relative humidity equalled 82.2±0.92%, wind speed atpup and female height were 1.1±0.08 m s�1 and1.5±0.14 m s�1, respectively and solar radiation was75.2±8.4 W m�2. Although measurements of microcli-mate were restricted to daylight hours, there was closeagreement with the mean and range of air temperature,relative humidity and solar radiation recorded by theweather station (Table 1). The maximum wind speedexperienced by pups and females was 3.5 m s�1 and5.1 m s�1, respectively, compared to 18.5 m s�1 re-corded by the weather station.

Table 1 Weather and microclimate experienced by grey seal pups on Isle of May, October–November 2003

Weather station Pupping sites

Mean SE Range Mean SE Range

Relative humidity (%) 87.3 0.09 67.6–100 82.2 0.92 69.0–99.0Air temperature (�C) 9.2 0.03 4.4–14.6 10.2 0.18 6.5–13.8Ground temperature (�C) – – – 8.6 0.39 3.4–25.7Solar radiation (W m�2) 109.4 2.39 0–413.9 75.2 8.4 2.0–312.0Wind speed (m s�1) at 2 m 6.6 0.05 0.0–18.5 – – –Wind speed (m s�1) at 0.1 m – – – 1.1 0.08 0.2–3.5Wind speed (m s�1) at 0.5 m – – – 1.5 0.14 0.6–5.1Wind direction (�) 193 1.1 0–355 – – –Rainfall (mm) 24.48 – – – – –

Values are mean, SE and range. Solar radiation is the mean daytime radiation (Civil Twilight Sunrise 07:50–Sunset 16:00 GMT)

426

Differences in microclimate between pupping sites

During early lactation there was a significant differencein wind speed (ANOVA: F2, 30=9.47, P=0.001) andsolar radiation (F2, 30=3.92, P=0.03) but no differencein relative humidity, air or ground temperature betweenpupping sites (P>0.10 all cases). The mean wind speedat Rona (1.4±0.12 m s�1) was greater than at Tarbet(0.8±0.06 m s�1) or Kirkhaven (0.6±0.08 m s�1 Tukeytest P £ 0.01 both cases) and there was no differencebetween Tarbet and Kirkhaven (P=0.39). Solar radia-tion at Rona (132.0±27.1 W m�2) was greater than atTarbet (36.0±9.90 W m�2, P=0.02) while there wereno differences between these sites and Kirkhaven(66.5±15.7 W m�2, P>0.64 both cases). In late lacta-tion there was no difference in any of the environmentalvariables recorded between sites (P>0.2 all cases).

Mean radiative temperature of pups during lactation

A total of seven pups (three early and four late lactation)were wet when measured. As there was no difference inmean Tr (mean from six measurement sites) between wetand dry pups (ANOVA: F1,60=0.70, P=0.41) data forpups were pooled for analysis. There was no differencein mean Tr of male and female pups either during earlyor late lactation (ANOVA: P>0.43 both cases). Pupbody mass did not account for variation in Tr duringearly (R2=0.01) or late lactation (R2=0.12, F-test,P>0.07 both cases).

The mean Tr of all pups during early and late lacta-tion was 15.7±0.34�C (range 7.7–29.7) and15.8±0.29�C (range 7.7–26.2), respectively. In earlylactation pups the Tr varied across the body (F5,

185=18.3, P<0.001; Fig. 2). Post-hoc comparisonsindicated that the hind flippers were on average 3–5.1�Cwarmer than all other regions of the body (Tukey test:P<0.001 all cases). The fore flipper was on average2.1�C warmer than the lower back and ventral side of

the body (P<0.02 both cases). In late lactation therewas again variation in the Tr of different regions (F5,

172=18.9, P<0.001; Fig. 2); the hind flipper averaged4.2–6.0�C warmer than other areas of the body(P<0.001 in all cases).

For the 20 pups captured at both stages of develop-ment, there was no difference in the mean Tr of pupsbetween early and late lactation (paired t-test: t=0.45,d.f. = 19, P=0.66). None of the body regions differedin Tr between these two stages (P>0.10 all cases).

Radiative temperature and microclimate

There was no difference in the mean Tr of pups betweenpupping sites either in early or late lactation (ANOVA:P>0.20 in both cases). There was also no difference inmean Tr of pups lying on earth, rock or vegetation inearly or late lactation (P>0.54 in both cases) and therewas no difference in the mean Tr of these substrates(P>0.71 both cases).

The mean Tr of pups during early and late lactationshowed a diurnal pattern reaching a maximum in themiddle of the day (quadratic regression: R2=0.20–0.28,P>0.001 both cases; Fig. 3). In early lactation all bodyregions showed a similar diurnal pattern (R2=0.14–0.29,P<0.005 all cases). This was also found during latelactation (R2=0.16–0.18, P<0.02 five cases) with theexception of the hind flippers that showed no change inTr throughout the day (R2=0.05, P=0.17).

A stepwise linear regression model was used toexamine the relationship between mean Tr of pups(dependent variable) and air temperature, wind speed,relative humidity, ground temperature and solar radia-tion (independent variables). In early lactation onlymodels containing solar radiation, air temperature andrelative humidity were significant (Table 2). Models withair temperature, air temperature + solar radiation andair temperature + solar radiation + relative humidity

Fig. 3 Relationship between mean radiative temperature (�C) ofgrey seal pups and time of day (Greenwich Mean Time, GMT) forearly (quadratic regression line R2=0.28, P<0.0001) and latelactation (R2=0.20; P<0.0001). Vertical lines mark the time ofsunrise and sunset

Fig. 2 Mean radiative temperature of different parts of body ongrey seal pups during early and late lactation. Standard error (SE)bars are given

427

accounted for 28, 42 and 48% of the variation in meanTr, respectively. This analysis indicated that during earlylactation the Tr of pups increased with increasing airtemperature (Pearson regression coefficient=0.52�C/�C)and solar radiation (0.009�C/W m�2) and decreasedwith relative humidity (�0.09�C/%RH). In late lactationonly models containing air temperature and air tem-perature + solar radiation were significant (Table 2).Models with air temperature and air tempera-ture + solar radiation accounted for 24% and 43% ofthe variance, respectively. The mean Tr of pups in-creased with an increase in solar radiation (0.012�C/W m�2) and air temperature (0.72�C/�C).

Discussion

Microclimate of pups

The climate experienced by grey seal pups in the UK isrelatively temperate in comparison to conditions formore northerly populations. In this study the mean airtemperature was 9.2�C (range 4.4–14.6�C), which waswithin the range in air temperature for November in thisarea (Leuchars: �0.3 to 11.6�C http://www.met-office.-gov.uk). However, on Sable Island Canada, pups areborn in January when air temperature averages �3.5 to2.8�C (http://www.climate.weatheroffice.ec.gc.ca) and inthe Baltic pups are born on the ice in March wheretemperatures may range from �11�C to �1�C (http://www.bbc.co.uk/weather/world).

The environmental conditions recorded by theweather station were representative of average condi-tions experienced by grey seal pups in pupping areas,with the exception of wind speed. The wind speed re-corded by the weather station was considerably greaterdue to its exposed position at the top of the island andthe height of anemometer above the ground surface.Pups were therefore sheltered from strong winds bybeing close to the ground surface and by the shelterprovided by their mother, behind rocks or other naturalfeatures. In many cases, pups gained additional shelterby placing their head inside the entrance to disused

puffin (Fratercula arctica) or rabbit (Oryctolagus cuni-culus) burrows. The mean wind speed recorded in thisstudy was similar to conditions experienced by Gala-pagos fur seals (Arctocephalus galapagoensis, range 0–3.7 m s�1; Limberger et al. 1986) between rocks andboulders and for northern elephant seals (Miroungaangustirostris, 0.25–3 m s�1; Heath et al. 1977) but lessthan for harp seals (Phoca groenlandica) on ice 0–11 m s�1 (Ørtisland and Ronald 1973; Ørtisland et al.1978).

There was some variation in the microclimate expe-rienced by pups on different areas of the island related tothe topography of different pupping areas. There weredifferences in solar radiation and wind speed betweensites but no consistent pattern throughout the studyperiod. Relative humidity, air temperature and groundtemperature did not differ between the three sites, indi-cating that on average pups experienced air temperatureand humidity determined by general weather conditionsand not significantly modified by local microclimate.These results suggest that overall meteorological condi-tions are able to describe the microclimate experiencedby pups, but that standard measurements of wind speedshould be adjusted by taking into account changes inwind speed with height and the shelter provided bymothers and terrain.

Radiative temperature of pups

The mean Tr of pups was 15.8�C in air temperatureaveraging 10.1�C. In air and in the absence of solarradiation, post-weaned pups reportedly have a thermo-neutral zone of metabolism that extends from �7�C to23�C and have an average core body temperature of38.9�C (Hansen and Lavigne 1997; Worthy 1991). Thereis no suggestion from the Tr recorded that pups showedsigns of overheating for prolonged periods of time. Al-though a maximum temperature of 29.7�C was recordedfrom the hind flippers during the study, there was rela-tively little variation in the radiative temperature be-tween different regions of the body. The hind flippersaveraged only 3–6�C warmer than all other areas, indi-cating that these areas were probably vasoconstrictedand not dilated as they would be if excess body heat wasbeing lost (see review Pabst et al. 1999).

Previous studies on juvenile grey seals (in air) haveindicated that there is no evidence of differences betweenbody regions, other than the head was 8–9�C aboveambient (Worthy 1991) and no difference in surfacetemperature between the dorsal region and hind flippersof adult grey seals over a range of air temperatures (Fol-kow andBlix 1987). Infrared images of grey seals howeverindicated that the head is one of the main areas of highestTr and there can be large temporal differences in the Tr ofdifferent body regions, including hind flippers in responseto drying out in air (Mauck et al. 2003). In this study hindflippers may have been warmer because some pups be-came alert during measurements and moved away or

Table 2 Stepwise linear regression models of mean radiative tem-perature of grey seal pups against air temperature (Ta), solarradiation (Sr) and relative humidity (RH) during early and latelactation

Linear regression model R2 P

Early lactationTr=0.84Ta+7.4 0.28 <0.001Tr=0.64Ta+0.01Sr+8.4 0.42 <0.001Tr=0.52Ta+0.009Sr � 0.09RH+17.5 0.48 <0.001Late lactationTr=0.75Ta+8.7 0.24 <0.001Tr=0.72Ta+0.012Sr+8.0 0.43 <0.001

The coefficient of variation (R2) and significance of F test (P) areshown

428

opened their hind flippers to expose warmer regions whenmeasurements were taken. The lack of difference in thehead region probably reflects the site of thick fur wheremeasurements were made rather than warm regionsaround eyes as observed in infrared images.

Body size did not account for the variation in meanTr of pups and no difference was found in Tr betweenearly and late lactation, or between male and femalepups. During early lactation grey seal pups have a rel-atively thick coat with relatively thin blubber layer incomparison to late lactation pups that are often moult-ing but have thick blubber (Kovacs and Lavigne 1986).It was expected that changes in blubber thickness withage would have influenced mean Tr, as previous studieson post weaning grey seal pups reported that surfacetemperatures were lower when blubber was thickest(Worthy 1991). Instead the Tr of pups averaged 5–6�Cabove air temperature throughout lactation, suggestingthat pups were able to adjust their conductance, eitherthrough changes in blood circulation through the blub-ber layer or by altering their behaviour to control heatloss (Limberger et al. 1986).

Microclimate and radiative temperature of pups

The mean Tr of pups was not related to ground temper-ature or type of substrate. In order to minimise distur-bance, no measurements were made of ventral regionswhere the body was in direct contact with the groundsurface. The importance of ground temperature andsubstrate type may therefore have been overlooked as thesurface temperature of the body in contact with theground is likely to be close to the temperature of the skinas the fur becomes compressed (Hind and Gurney 1998).Measurements of an adult harbour seal (Phoca vitulina)and a California sea lion (Zalophus californianus) indicatethat seals lose 10–26% of metabolic heat productionthrough conduction (Ohata and Whittow 1974).

The mean Tr of pups increased throughout the day.Although the core body temperatures of Galapagos furseal pups vary more during the day than at night, there isno evidence for endogenous diel rhythms in core tem-perature of pups during lactation (Limberger et al.1986). Microclimate variables did however explain agreater proportion of the variation in mean Tr than timeof day alone. This suggests that the increase in Tr ofpups throughout the day was not due to changes in coretemperature but was related to diurnal variation inenvironmental conditions. In early lactation, solarradiation, air temperature and relative humidity ex-plained 48% of the variation in mean Tr while duringlate lactation air temperature and solar radiation aloneexplained 43% of the variation in mean Tr. Previously,no significant relationship was found between the sur-face temperature of juvenile grey seals and solar radia-tion, air temperature or wind speed in captivity (Worthy1991). However, these were post-weaned animals withdark, adult-type pelage. For white-coated pups the

thick, well-insulated fur with high reflectance has beenshown to provide a significant heat trapping action whenirradiated by sun in comparison to dark fur (Ørtislandand Ronald 1973; Ørtisland et al. 1978).

During early lactation, an increase in relativehumidity reduced the Tr of pups, whereas at late lacta-tion no effect was found. Heat transfer theory wouldsuggest that if an increase in ambient relative humiditybrought about an increase in the relative humiditywithin the coat there would be a small transient heatingeffect due to the sorption of water by the fur (Gatenby etal. 1983). This process will go undetected when the rel-ative humidity of the air changes over long time periods(hours to days) and would therefore not account forchanges in Tr in this study. The high relative humiditiesprobably indicate that it had been raining recently and/or the ground was wet. Although coat condition wasvisually checked some of the pups may not have beenfully dry, especially in damp conditions when high rel-ative humidities were recorded. Evaporative coolingfrom damp fur therefore probably explains why Tr wasnegatively correlated with relative humidity.

Pups had wet coats as a result of rain or lying on wetground or in water. Water had been absorbed by thecoat and there was no free standing water on the surface.The Tr of wet pups was around 1–3�C less than all drypups indicating a small evaporative cooling effect at thesurface (although this difference was not statisticallysignificant). As newborn pups rely on insulation pro-vided by the lanugo, any increase in the thermal con-ductivity of fur due to wetting is likely to have a greatereffect on heat loss than at later stages of lactation wheninsulation is provided by both lanugo and blubber (Blixet al. 1979). The 2003 breeding season was relatively drywith only 24.5 mm recorded in the 28-day period com-pared with an average November rainfall of 57.3 mm forthe area (http://www.met-office.gov.uk). The heat fluxfrom wet northern elephant seals was around 50%greater than when they were dry (Heath et al. 1977),suggesting that during years of high rainfall, wettingmay play an important role in the thermal balance ofpups at early stages of development.

Although wind increased the rate of heat loss innorthern elephant seal pups (Heath et al. 1977) there wasno effect of wind speed on the mean Tr of pups in thisstudy. Previously Ørtisland et al. (1978) found that airtemperature rather than wind speed exerted the greatesteffect on Tr of harp seal pups. In this study, grey sealpups experienced average wind speeds of 1.1 m s�1 andat these low wind speeds there is likely to be minimalwind penetration into dense fur (Tregear 1965). Forcedconvection therefore would not appear to exert a strongeffect on coat temperature in natural conditions.

Metabolic heat production

Conductive heat flux models for marine mammals tendto overestimate metabolic heat production in compari-

429

son to those determined by oxygen consumption(Kvadsheim et al. 1997). Nevertheless these modelsprovide a means of increasing our understanding of themechanisms of heat loss and the costs of thermoregu-lation in marine mammals. For this purpose Eq. 1 wasused to examine how total heat loss of grey seal pupschanged throughout lactation and how heat loss variedaccording to the diurnal change in Tr (Fig. 3). Themodel was parameterised using previously publishedvalues for grey seals or similar phocids (Table 3). Tr

provided an approximation of the surface temperatureof the coat in contact with air (see theory section) andtemperature of the coat in contact with ground was as-sumed to be the mean of Tr and ground temperature.The model predicted that the total metabolic heat pro-duction for a 19 kg and 38 kg grey seal pup in early andlate lactation was 141 W and 104 W, respectively. Thiscompared with 56 W and 91 W for early and late lac-tation pups as determined by the relationship betweenresting metabolic rate for growing phocids (Lavigne etal. 1986). The large difference for early lactation pupsmay indicate the inadequacy of the model or that theallometric relationship for young, rapidly growing pupscannot be applied. The metabolic heat production ofgrey seal pups was predicted to be on average 20% lessduring the middle of the day than at dawn or dusk dueto a 5�C change in Tr (Fig. 3). The model also indicatedthat the cost of thermoregulation for grey seal pups mayvary by as much as 60% due to differences in Tr from10�C to 22�C throughout the pupping season. Thesefigures are imprecise estimates of metabolic heat pro-duction but are intended to demonstrate how heat lossand the cost of thermoregulation for grey seal pups isstrongly dependent on the coat surface temperature.

Conclusion

Air temperature and solar radiation were importantfactors influencing the Tr of grey seal pups during lac-tation. However environmental variables explained lessthan half of the variation in Tr of pups. Future studiesshould therefore focus on examining how core bodytemperature, blubber thickness, coat condition andactivity patterns influence surface body temperature. Bydetermining how these intrinsic factors influence the

surface temperature of pups in natural conditions wewill be in a better position to develop heat transfermodels to explain the behaviour of breeding grey seals.

Acknowledgements Fieldwork and handling procedures were con-ducted under UK Home Office Licence. We wish to thank J. Dale,J. Mould and C. Sparling for help with fieldwork. We are gratefulfor the assistance of Scottish Natural Heritage and crew of the MayPrincess and Breadwinner for transport to the Isle of May. Thiswork was funded by the Natural Environment Research Council tothe Sea Mammal Research Unit, University of St Andrews; theFaculty of Education, University of Glasgow; and the BlodwenLloyd-Binns Bequest, Glasgow Natural History Society.

References

Anderson SS, Burton RW, Summers CF (1975) Behaviour of greyseals Halichoerus grypus during a breeding season at NorthRona. J Zool (London) 177:179–195

Blix AS, Miller LK, Keyes MC, Grav HJ, Elsner R (1979) New-born northern fur seals (Callorhinus ursinus)—do they sufferfrom cold? Am J Physiol 236:R322–R327

Boily P (1995) Theoretical heat flux in water and habitat selectionof phocid seals and beluga whales during the annual molt. JTheoret Biol 172:235–244

Boily P, Kvadsheim PH, Folkow LP (2000) Cutaneous heat fluxmodels do not reliably predict metabolic rates of marinemammals. J Theoret Biol 207:317–323

Campbell GS, Norman JM (1998) An introduction to environ-mental biophysics. Springer, Berlin Heidelberg New York

Cena K (1974) Radiative heat loss from animals and man. In:Monteith JL, Mount LE (eds) Heat loss from animals and manassessment and control. Butterworths, London, pp 32–58

Folklow LP, Blix AS (1987) Nasal heat and water exchange in grayseals. Am J Physiol 253:R883–R889

Folkow LP, Blix AS (1992) Metabolic rates of minke whales(Balaenoptera acutorostrata) in cold water. Acta Physiol Scand146:141–150

Frisch J, Ørtisland NA (1968) Insulative changes in the harp sealpup during moulting. Acta Physiol Scand 74:637–638

Gallivan GJ, Ronald K (1979) Temperature regulation in freelydiving harp seals (Phoca groenlandica). Can J Zool 57:2256–2263

Gatenby RM, Monteith JL, Clark JA (1983) Temperature andhumidity gradients in sheep’s fleec. II. The energetic significanceof transients. Agric Meteorol 29:83–101

Hall A (2002) Gray seal (Halichoerus grypus). In: Perrin WF,Wursig B, Thewissen JGM (eds) Encyclopedia of marinemammals. New York, Academic

Hansen S, Lavigne DM (1997) Temperature effects on the breedingdistribution of grey seals (Halichoerus grypus). Physiol Zool70:436–443

Table 3 Model parameters used to estimate total metabolic heat production for grey seal pups during early and late lactation

Parameter Units Early lactation Late lactation

Body mass kg 19 38 This studyMean surface area, A m2 0.57 0.81 Innes et al. (1990)Proportion of seal in contact with air None 0.8 0.8 Leucke et al. (1975)Core body temperature, Tc �C 38.9 38.9 Worthy (1991)Coat conductance W m�2 �C�1 0.341 0.341 Frisch and Ørtisland (1968)Blubber conductance W m�2 �C�1 9 4.5 Worthy (1991a)Total heat transfer coefficient, hc W m�2 �C�1 9.341 4.841 Calculatedb

aThermal conductance of blubber was calculated using a thermal conductivity of grey seal blubber of 0.18 W m�1 �C�1 and estimatedblubber depth of 0.02 and 0.04 m for early and late lactation pups respectivelybThe total heat transfer coefficient is the sum of coat andblubber conductance

430

Heath ME, McGinnis SM, Alcorn D (1977) Comparative ther-moregulation of suckling and weaned pups of the northernelephant seal, Mirounga angustirostris. Comp Biochem PhysiolA 57:203–206

Hind AT, Gurney WSC (1998) Are there thermoregulatory con-straints on the timing of pupping for harbour seals? Can J Zool76:2245-225

Hokkanen JEI (1990) Temperature regulation of marine mammals.J Theoret Biol 145:465–485

Innes S, Worthy GAJ, Lavigne DM, Ronald K (1990) Surfaceareas of phocid seals. Can J Zool 68:2531–2538

Kovacs KM, Lavigne DM (1986) Growth of gray seal (Halichoerusgrypus) neonates—differential maternal investment in the sexes.Can J Zool 64:1937–1943

Kvadsheim PH, Gotaas ARL, Folklow LP, Blix AS (1997) Anexperimental validation of heat loss models for marine mam-mals. J Theoret Biol 184:15–23

Lavigne DM, Innes S, Worthy GAJ, Kovacs KM, Schmitz O,Hickie J (1986) Metabolic rates of seals and whales. Can J Zool64:279–284

Lavigne DM, Innes S, Worthy GAJ, Edwards EF (1990) Lowercritical temperature of blue whales, Balaenoptera musculus. JTheoret Biol 144:249–257

Limberger D, Trillmich F, Biebach H, Stevenson RD (1986)Temperature regulation and microhabitat choice by free-rang-ing Galapagos fur seal pups (Arctocephalus galapagoensis).Oecologia 69:53–59

Luecke RH, Natarajan V, South FE (1975) A mathematical bio-thermal model of the California sea lion. J Thermal Biol 1:35–45

Mauck B, Bilgmann K, Jones DD, Eysel U, Dehnhardt G (2003)Thermal windows on the trunk of hauled-out seals: hot spotsfor thermoregulatory evaporation? J Exp Biol 206:1727–1738

McArthur AJ, Ousey JC (1994) Heat loss from a wet animal:changes with time in the heat balance of a physical modelrepresenting a newborn homeotherm. J Thermal Biol 19:81–89

Ohata CA, Whittow GC (1974) Conductive heat loss to sand inCalifornia sea lions and a harbor seal. Comp Biochem PhysiolA 47:23–26

Ørtisland NA, Ronald K (1973) Effects of solar radiation andwindchill on skin temperature of the harp seal, Pagophilusgroenlandicus (Erxleben, 1777). Comp Biochem Physiol A44:519–525

Ørtisland NA, Lavigne DM, Ronald K (1978) Radiative surfacetemperatures of harp seals. Comp Biochem Physiol A 61:9–12

Pabst DA, Rommel SA, McLellan WA (1999) The functionalmorphology of marine mammals. In: Reynolds J, Rommel SA(eds) Biology of marine mammals. Smithsonian Institution,Washington

Pomeroy PP, Fedak M, Rothery P, Anderson S (1999) Conse-quences of maternal size for reproductive expenditure andpupping success of grey seals at North Rona, Scotland. J AnimEcol 68:235–253

Pomeroy PP, Twiss SD, Duck CD (2000) Expansion of grey seal(Halichoerus grypus) breeding colony: changes in pupping siteuse at the Isle of May, Scotland. J Zool (London) 250:1–12

Ryg M, Smith TG, Ørtisland NA (1988) Thermal significance ofthe topographical distribution of blubber in ringed seals (Phocahispida). Can J Fish Aquat Sci 45:985–982

Tregear RT (1965) Hair density, wind speed, and heat loss inmammals. J Appl Physiol 20:796–801

Trites AW (1990) Thermal budgets and climate spaces: the impactof weather on the survival of Galapagos (Arctocephalus gala-pagoensis Heller) and northern fur seal pups (Callorhinus ursi-nus L.). Funct Ecol 4:753–768

Twiss SD, Caudron A, Pomeroy PP, Thomas CJ, Mills JP (2000)Fine scale topographic correlates of behavioural investment inoffspring by female grey seals (Halichoerus grypus). Anim Be-hav 59:327–328

Watts P, Hansen S, Lavigne DM (1993) Models of heat loss bymarine mammals: thermoregulation below the zone of irrele-vance. J Theoret Biol 163:505–525

Williams TM, Noren D, Berry P, Estes JA, Allison C, Kirtland J(1999) The diving physiology of bottlenose dolphins (Tursiopstruncatus). III. Thermoregulation at depth. J Exp Biol202:2763–2769

Worthy GAJ (1991) Insulation and thermal balance of fasting harpand gray seal pups. Comp Biochem Physiol A–Physiol 100:845–851

431