Energetics and litter size variation in domestic dog Canis familiaris breeds of two sizes

Ž .Comparative Biochemistry and Physiology Part A 129 2001 919�931

Energetics and litter size variation in domestic dogCanis familiaris breeds of two sizes

Michael Scantlebury a,�, Richard Butterwickb, John R. Speakmana

aAberdeen Centre for Energy Regulation and Obesity, Department of Zoology, Uni�ersity of Aberdeen, Aberdeen, Scotland,AB24 2TZ, UK

bWaltham Centre for Pet Nutrition, Waltham-on-the-Wolds, Melton Mowbray, Leicestershire, England, LE14 4RT, UK

Received 29 June 2000; received in revised form 19 March 2001; accepted 22 March 2001

Abstract

Ž . Ž . Ž .We measured resting metabolic rate RMR , daily energy expenditure DEE and metabolisable energy intake MEIin two breeds of dog during peak lactation to test whether litter size differences were a likely consequence of allometric

Ž . Ž .variation in energetics. RMR of Labrador retrievers 30 kg, n�12 and miniature Schnauzers 6 kg, n�4 averagedŽ . Ž .3437 and 1062 kJ�day, respectively. DEE of Labradors n�6 and Schnauzers n�4 averaged 9808 and 2619 kJ�day,

Ž . Ž .respectively. MEI of Labradors n�12 was 22448 kJ�day and of Schnauzers n�7 was 5382 kJ�day. DEE ofŽ . Ž .Labrador pups 2.13 kg, n�19 was 974 kJ�day and Schnauzers 0.89 kg, n�7 were 490 kJ�day. Although Labradors

had higher MEIs than Schnauzers during peak lactation, there was no difference in mass-specific energy expenditurebetween the two breeds. Hence, it is unlikely that litter size variation is a likely consequence of differences in maternalenergy expenditure. Individual offspring were relatively more costly for mothers of the smaller breed to produce.Therefore, litter size variations were consistent with the expectation that smaller offspring should be more costly formothers, but not that smaller mothers should per se invest more resources in reproduction. � 2001 Elsevier Science Inc.All rights reserved.

Keywords: Domestic dog; Doubly labelled water; Isotope; Energetics; Lactation; Litter size; Offspring; Metabolisable energy intake;Resting metabolic rate; Body mass

1. Introduction

Ž .The domestic dog Canis familiaris has beenassociated with humans for at least 12 000 yearsŽDavis and Valla 1978; Clutton-Brock, 1995; Vila

� Corresponding author. Mammal Research Unit, School ofBiological Sciences, University of Bristol, Bristol BS8 1UG,UK. Tel.: �44-117-9288918; fax: �44-117-9257374.

ŽE-mail address: [email protected] M. Scantle-.bury .

.et al., 1997 , during which time selection on bodysize has produced a 40-fold difference in body

Žmass between breeds Kirkwood, 1985; Heusner,. Ž1991 . Litter size Lyngset and Lyngset, 1970;

. ŽRobinson, 1973 , birth weight Altman and Dit-tmer, 1962; Kirkwood, 1985; Evans and White,

. Ž1988 and longevity Bronson, 1982; Kirkwood,

.1985 also differ between breeds of different sizes.There is a positive intraspecific relationshipbetween litter size and adult body mass in dogs,ranging from the 2-kg Chihuahua with an average

1095-6433�01�$ - see front matter � 2001 Elsevier Science Inc. All rights reserved.Ž .PII: S 1 0 9 5 - 6 4 3 3 0 1 0 0 3 5 9 - 2

( )M. Scantlebury et al. � Comparati�e Biochemistry and Physiology Part A 129 2001 919�931920

litter size of 2.3, to the 80-kg St. Bernard with anŽaverage litter size of 6 Robinson, 1973; Evans

.and White, 1988 . In this paper, we addresswhether such variation in litter size betweenbreeds of different sizes is a likely consequence ofallometric variation in energetics.

Energetics may be important at various stagesin the reproductive cycle. Previous studies of fe-male mammals have shown that food intake gen-erally increases throughout gestation and lacta-

Žtion, reaching a maximum at peak lactation Mil-lar, 1978; Oftedal, 1985; Gittleman and Thomp-

.son 1988; Rogowitz, 1996 . Dogs conform to thisŽmammalian pattern Paragon and Grandjean,

.1993 . Therefore, if energetics is important indetermining litter size variation in dogs, peaklactation is likely to be a key period, as it definesthe maximal amount of resources available toproduce offspring.

One reason why peak lactation may be impor-tant is because the lactating female might ap-proach a ceiling in her potential rate of energy

Ž .expenditure Gittleman and Thompson, 1988 .Previous studies have shown that there may be aphysiological limit to sustained energy expendi-

Ž .ture, termed sustained metabolic scope SusMSŽDrent and Daan, 1980; Weiner, 1989; Peterson

.et al., 1990 . This metabolic ceiling has beensuggested to be approximately 6�7 times resting

Žor basal metabolic rate Hammond and Diamond,.1997 and can be approached by exposure to cold

Ž .Hammond et al., 1994; Koteja, 1996 , duringŽpeak lactation Hammond and Diamond, 1992,

1994; Konarzewski and Diamond, 1994; Speak-.man and McQueenie, 1996 or a combination of

Ž .both Hammond et al., 1994 . At present, there issome confusion in the literature over the mea-surement of SusMS. Some authors have used the

Žratio of DEE�RMR where DEE is the dailyaveraged energy expenditure and RMR is the

.metabolic rate of the animal measured at restŽ .Peterson et al., 1990 , while others have chosen

ŽMEI�RMR where MEI is the daily averaged. Žmetabolisable energy intake Hammond and Di-

.amond 1992, 1994 . In many situations, these willonly differ slightly. However, during lactation,considerable amounts of energy are exported asmilk and SusMS expressed as DEE�RMR will besubstantially lower than SusMS expressed as

Ž .MEI�RMR Speakman, 2000 . Moreover, limitsmay act either on the intake or the expenditureŽthe so-called ‘central’ and ‘peripheral’ limits hy-

. Žpotheses Weiner, 1987, 1992; Daan et al., 1989;.Hammond et al., 1994 , adding further signifi-

cance to the alternative methods of calculatingSusMS. To distinguish these measurements here,we use SusMS to refer only to the time-averaged

Ž .energy expenditure DEE�RMR , and for sus-Ž .tained energy intake MEI�RMR we use the

term SusEI.In a previous study of two different sized breeds

Ž .of domestic dog, Labrador retrievers 30 kg andŽ .miniature Schnauzers 6 kg during peak lacta-

tion, we showed that the demands of the larger,Žfaster-growing Labrador pups compared to

.Schnauzer pups were likely to be met by anŽincrease in MEI of the mothers Scantlebury et

.al., 2000b . However, the reasons why Labradorsmaintained high mass-specific energy intakes andsustained the larger litters remain unexplained.During lactation, MEI of mothers is used to sup-port both themselves and their offspring. If theextra energy intake of Labradors was predomi-nantly due to the costs of maternal metabolism, itmight be predicted that Labrador mothers wouldhave higher mass-specific DEEs than Schnauzers.However, if this energy was required primarily tosupport their litters, it might be predicted thatLabrador mothers would have a greater mass-specific energy investment in offspring thanSchnauzers. Smaller individuals generally expenda greater amount of energy�kg body mass than

Ž .larger individuals Kleiber, 1961 . Therefore, incontrast to the first prediction, the smallerSchnauzer mothers and their pups would each beexpected to expend a greater mass-specific energyon their metabolism than the larger Labradormothers and their pups. Consequently, Schnauzermothers would have fewer resources to export totheir offspring, which would in turn be able toallocate a smaller proportion of those resourcesinto growth. Therefore, we aimed to examine

Ž .whether: a the high MEIs of Labrador motherswere effected by a high mass-specific DEE, or ameans of providing a greater investment into off-

Ž .spring; b the mass-specific costs of individualoffspring metabolism were greater in Schnauzers

Ž .than Labradors; and c the Labrador pups wereable to convert a greater proportion of theirenergy intake into growth.

We present measurements of energy budgets ofLabrador retriever and miniature Schnauzermothers and litters in relation to the aims above.In the context of the theory of limits to energy

( )M. Scantlebury et al. � Comparati�e Biochemistry and Physiology Part A 129 2001 919�931 921

budgets, we examine whether there were differ-ences in energy expenditure between breeds thatcould help explain differences in litter size. Dur-ing these observations, we did not intend to testfor limits to energy expenditure or assimilation.Furthermore, it is not possible to establish theeffects of a continuous independent variable, suchas body mass, by the study of only two species or

Žbreeds that differ in this trait Garland and.Adolph, 1994 . However, in the context of ex-

amining the aims above, concerning the effects ofbody mass, there were logistical constraints instudying more than two breeds in sufficient depthto establish the different patterns of investment.We present the results for two breeds here as anindication of the role that mass differences mightplay in determining patterns of investment as astimulus for future studies.

2. Materials and methods

2.1. Animals

A total of 12 lactating Labradors retrieversŽ . Ž‘Labradors’ mean�30.3, S.D.�3.5, range�

. Ž27�37 kg , 77 Labrador pups mean � 2.13,.S.D.� 0.36 kg , seven lactating miniature

Ž . ŽSchnauzers ‘Schnauzers’ mean�6.8, S.D.�1.0,. Žrange�5.3�7.8 kg and 25 Schnauzer pups mean

.�0.889, S.D.�0.08 kg were used. Six of the 12Labrador litters were used for maternal energy

Žexpenditure measurements litter sizes 5, 7, 7, 6, 7.and 5 and six were used to measure offspring

Ž .litter sizes 6, 6, 6, 7, 8 and 7 . Similarly, fourSchnauzer litters were used for maternal energy

Žexpenditure measurements litter sizes 2, 3, 2 and.4 and three litters for measurements of offspringŽ .litter sizes 4, 4 and 6 . We have previously shownthat recycling of isotopes is unlikely to causesignificant errors in maternal DEE calculations,but may cause an underestimate of pup DEE by

Ž .up to 7% Scantlebury et al., 2000a . Adults andoffspring were labelled in separate experimentsand only half the offspring of any litter werelabelled to correct for any effects of isotope recy-cling.

Observations took place over periods of 7 daysŽ24�30 days post partum, with day 0 of observa-

.tion�day 24 of lactation during the period ofŽpeak lactation Meyer, 1984; Oftedal, 1984a;

.Kienzle et al., 1985 , which was the same in both

Ž .breeds Rivers and Burger, 1989 . Animals werebred and reared at the Waltham Centre for Pet

Ž .Nutrition WCPN , Melton Mowbray, UK, andwere housed in a climate-controlled building at20�25�C with access to outside runs. Details ofthe design of the housing conditions and the pens

Ž .have been described previously Loveridge, 1994 .Synthetic bedding and a heated area of the floorof each pen were provided for the pups andmother to sleep in. Each pen was illuminatednaturally through a window and supplemented by

Ž .electric lighting 12L:12D .

2.2. Food intake

Adults were fed Waltham Formula Expert� ŽGrowth Rivers and Burger, 1989; Burger and

.Johnson, 1991 which consisted of 26.9% crudeprotein, 14.4% crude fat, 44.8% nitrogen-free ex-

Ž .tract including carbohydrate and fibre , 7.4%moisture and 6.5% ash. The metabolisable energyŽ .ME of food was determined by total faecal and

Ž .urine collections WCPN, unpublished data andwas 15.6 kJ�g. Food intake was determined by

Žweighing the food bowl Sartorius F150 balance.�0.1 g, Gottingen, Germany at 08:00 h each day.

Food and water were supplied in deep, stainlesssteel bowls that were too high for the pups tofeed or drink from.

2.3. Indirect calorimetry

Ž .Resting metabolic rate RMR was determinedŽin adults prior to breeding n�12 Labradors and

. Žn�4 Schnauzers by indirect calorimetry Brody,.1945; Weir, 1949; Elia and Livesey, 1992 . The

calorimeter system consisted of an airtight cham-ber of 4.72 m3 internal volume maintained at18�1�C by a propylene glycol cooling systemŽ .Rainbird and Booles, 1990 . Airflow was approxi-mately 100 l�min. Carbon dioxide and oxygenwere continuously monitored in the inlet and the

Žoutlet systems 1490 Infrared CO Analyser and2a 540A O Analyser, Servomex, Crowborough,2

.UK . Energy expenditure was calculated using theŽ .measured respiratory quotient RQ and the

Ž . Ž .equation of Weir 1949 Speakman, 1997a as:

Ž .Energy expenditure E

� Ž .� Ž .� rCO 1.106� 3.941�RQ �4.186 12


where E is in J�day and rCO is in ml�day.2Animals were inside the chambers for approxi-mately 17 h, from 15:00 until 08:00 h, and werecontinuously monitored by an infrared videocamera. Food and water were supplied ad libitumin the metabolic chamber. Animals were con-sidered to be ‘at rest’ when they were stationaryŽfrom approx. 2 h after entering the chamber forapprox. the next 14 h, or until they were disturbed

.the next morning . RMR was calculated using atleast 2 h of stable gas recordings while animalswere observed to be at rest and after they hadbeen in the respirometry chamber for at least 4 h.

2.4. Determination of pup isotope equilibrationcur�es

Although previous work has established theequilibrium curves of 2 H and 18 O in adult dogsŽ .Speakman et al., in press , no data were availablefor 4-week-old offspring. We therefore de-termined the isotope equilibrium curves of 18 Oand 2 H in a litter of Labrador retriever pups. FiveLabrador pups aged 24 days and weighing on

Ž .average 2.36 kg S.D.�0.5 were used. A 2.5-mlŽ . Ž .injection i.v. of doubly labelled water DLW

� 18 Ž2.3 parts 90% enriched O water Enritech Ltd.,. 2Rehovot, Israel and 1 part 99.9% enriched H

Žwater MSD Isotopes Inc., Pointe-Claire, Quebec,. �Canada made isotonic with sodium chloride was

administered to each pup. Serial blood sampleswere taken every 40 min for a total of 200 min.2 H and 18 O equilibration curves were estimatedfrom the increase in isotope enrichment mea-sured in the blood samples. During the equilibra-tion period, the pups were separated from theirmothers and had no access to food or water.

2.5. Doubly labelled water protocol

ŽWe used the DLW technique Lifson et al.,1955; Nagy, 1980; Schoeller and van Santen, 1982;

.Speakman, 1997a to measure daily energy expen-Ž .diture DEE in lactating mothers and offspring.

ŽOn day 1, the animals were weighed �1.0 g. ŽSartorius F150 balance and a blood sample ca.

.1.0 ml was taken from a cephalic vein to estimatethe background isotope enrichments of 2 H and18 O. On day 2, a known mass of DLW was admin-

Ž .istered intravenously ca. 0.3 g�kg body weight .Syringes were weighed before and after adminis-

Žtration �0.0001 g Sartorius AS 200 analytical

.balance . Previous work has established the equi-librium curves of 2 H and 18 O in adult dogsŽ .Speakman et al., in press . Blood samples weretaken after 5 h in adults and at 2.5 h post dose inpups to estimate initial isotope enrichments.Thereafter, blood was taken at 24-h intervals andused to estimate the isotope elimination ratesŽ .Speakman and Racey, 1988 . Blood samples werecollected into 1.5-ml heparinised serological tubes,from which 3�100-�l glass capillaries were im-mediately filled and heat-sealed. Excess bloodcollected was frozen at �80�C. Injectate enrich-ment measurements and isotope analysis methods

Žhave been described previously Krol and Speak-.man, 1999 .

2.6. Calculation of DEE

18 2 Ž .O and H injectate enrichments ppm , elimi-Ž . Žnation constants k and k , dilution spaces No d o

.and N , and initial and final pool sizes weredŽ .calculated as recommended Speakman, 1997a .

The ‘plateau’ method was used to estimate initialisotope enrichments with the ‘percentage mass’method of calculating final pool sizes. k and ko dwere evaluated using the least-squares regression

Žof multiple samples seven samples�animal,.one�day obtained using log -converted enrich-e

Ž .ment ppm values above the measured back-ground for the appropriate individual taken onday 1. The one-pool calculation method was ap-plied to the data set of offspring to calculate CO2

Ž .production Speakman, 1997a, equation i and theŽtwo-pool method used for adults Speakman et

.al., 1993 .Direct measurements of the respiratory

Ž . Žquotient RQ in lactating dogs n � 12.Labradors yielded estimates between 0.81 and

Ž .0.87 unpublished data . Errors converting CO2production measurements to energy metabolismare small for animals that eat daily if an RQ of0.83 is assumed for all herbivores and 0.8�0.83

Žfor mammalian carnivores Gessamen and Nagy,.1988 . An RQ of 0.8 assumes a calorific equiva-

lent of 20.09 kJ�l O consumed, whereas an RQ2of 0.9 has an equivalent of 20.51 kJ�l O con-2

Ž .sumed Hardy, 1972 . Hence, variation in RQbetween 0.8 and 0.9 only produces a difference of1% in the final energy calculation. Therefore, weassumed an RQ of 0.85 in the current experi-ments. DEE was calculated via a dedicated com-puter package that took into account the effects


of small deviations from 24 h for the samplingintervals and the effects of mass changes on bodywater pool size over the experimental period.

3. Results

Background levels of isotope enrichment aver-Ž . 18aged 1994 ppm S.D.�5.4 for O and 146 ppm

Ž . 2S.D.�5.2 for H across all individuals over the2 years of study. This compared with local tapwater enrichments of 1983 ppm for 18 O and 141ppm for 2 H. In the pup isotope equilibrium-curveexperiment, there was a rapid increase in enrich-ment of both isotopes in the plasma water of allthe pups following injection. The initial increasein enrichment was followed by a small but sig-nificant decrease at around two hours post dose.After this initial decline in enrichment, the iso-tope levels reached a plateau at approximately 2.5h post dose. From this experiment, we concludedthat the optimal time to sample initial isotopeenrichment values was approximately 2.5 h postdose.

3.1. Dilution spaces, elimination cur�es and CO2production

18 2 Ž .Mean O and H dilution spaces N and N ,o dŽ .elimination rates k and k and CO produc-o d 2

tion of Labrador and Schnauzer mothers andoffspring are shown in Table 1. Across all adultsŽ . Ž .n�10 the average dilution space ratio N �Nd o

Ž .was 1.0312 S.D.�0.036 , and across all pupsŽ . Ž .n�26 it was 1.0157 S.D.�0.081 . In all cases,the log-converted isotope enrichments above

background decreased linearly, and the r 2 valuesaveraged 99.5% for both isotopes in both breeds.

3.2. Energy balance in mothers and offspring

Adult dogs of both breeds did not change sig-nificantly in mass during the experimental periodŽpaired t�0.43, P�0.68 for 12 Labradors and

.t�1.0, P�0.39 for 7 Schnauzers . In contrast,ŽLabrador pups initially weighed 2138 g S.D.�

.368 and increased on average by 35 g�dayŽ .S.D.�12 over the next 7 days. Similarly, the

Ž .889-g S.D.�83 Schnauzer pups increased onŽ .average by 10 g�day S.D.�7 . Labradors had

�larger litter sizes than Schnauzers ANCOVA:Ž .F 1,17 �41.63, P�0.001, adult body mass and

�litter size entered as co-variates and there was atendency for mean pup mass to decrease with

� Ž .increasing litter size ANCOVA: F 1,17 �3.38,�P�0.084, Fig. 1 . Total litter mass and MEI were

not significantly correlated with litter size inŽ .Labradors, but were in Schnauzers Figs. 2 and 3 .

Although Labrador mothers consumed more foodthan Schnauzers because they were larger, therewas still a tendency for them to have a greaterMEI, when body mass was included as a co-variate� Ž . �ANCOVA: F 1,15 �3.87, P�0.068 . In con-trast, RMR and DEE were not significantly dif-

� Ž .ferent between breeds ANCOVA: F 1,6 �0.33,�P�0.587 . Maternal DEE was not significantly

� Ž .correlated with litter size ANCOVA: F 1,6 ��0.04, P�0.840, Fig. 4 . There was a large amount

of individual variation in DEE of the SchnauzerŽ .mothers range 1080�6418 kJ�day . Labrador

pups expended approximately twice the energy ofSchnauzers, but metabolised a smaller proportion

Table 1Measured isotope dilution spaces, turnover rates and calculated energy expenditures for all individuals

n N S.D. N S.D. N �N k S.D. k S.D. k �k rCO S.D. E S.D.o d d o o d o d 2Ž . Ž . Ž . Ž . Ž . Žl l �day �day ml�min kJ�

.day

AdultsLabradors 6 18.45 2.60 18.7 2.39 1.013 0.014 0.003 0.012 0.003 1.198 2876 447 9808 1658Schnauzers 4 3.52 0.49 3.7 0.66 1.039 0.011 0.002 0.008 0.001 1.218 871 608 2619 2105

OffspringLabradors 19 1.537 0.287 1.558 0.282 1.018 0.0072 0.0009 0.0054 0.0084 1.3534 282 97 974 334Schnauzers 7 0.636 0.073 0.637 0.086 1.005 0.0078 0.0018 0.0056 0.0013 1.4513 142 27 490 93

18 Ž . 2 Ž . Ž . 18 Ž . 2Means and S.D. of O dilution space N H dilution space, N , dilution space ratio N �N , O elimination rate k , Ho d o d oŽ . Ž . Ž . Ž .elimination rate k , isotope elimination ratio k �k , carbon dioxide production rCO and calculated energy expenditure E ford o d 2

lactating adults and offspring. n denotes sample sizes.


Ž . Ž . Ž .Fig. 1. Mean mass kg of 24-day-old a Labrador pups and b Schnauzer pups against litter size. Sample sizes are shown as numbersabove each point. Error bars denote �standard errors. There was no significant relationship between mean mass and litter size for

Ž 2 . Ž 2 . Žeither Labradors r �0.102, P�0.150 or Schnauzers r �0.175, P�0.192 least-squares regression on means of values at each.litter size .

Ž .of the energy they received as milk Table 2 .These values were not significantly different when

0.83 Žexpressed relative to offspring mass Oftedal,.1984b , although they were significant when ex-

pressed relative to maternal size or RMR.

3.3. In�estment in offspring and the cost of lactation

Ž .In theory, the DLW measurement included: 1Ž . Ž .RMR; 2 the energy cost of activity; 3 the

Ž .energy cost of thermoregulation; 4 the heat

Ž .increment of feeding; and 5 the energy cost ofmilk synthesis. During the experiments, the nurs-ing mothers were held in climate-controlled build-ings, did not change significantly in body massand were not engaged in activities other thanfeeding and nursing their litters. Therefore, wemade the assumption that, except for lactation,

Žthe mothers were in energy balance i.e. no.growth, fat storage or use of stored energy .

Hence, the energy allocated�litter was approxi-mated as MEI�DEE, where DEE was the daily


Ž . Ž . Ž .Fig. 2. Total litter mass kg of 24-day-old a Labrador pups and b Schnauzer pups against litter size. Sample sizes are shown asnumbers above each point. Error bars denote �standard errors. There was no significant relationship between mean mass and litter

Ž 2 . Ž 2 .size for Labradors r �0.015, P�0.301 , but there was for Schnauzers r �0.668, P�0.015 .

average DLW measurement. The energy cost oflactation was estimated as the DLW measure-ment during lactation minus the maternal pre-breeding RMR. The latter was likely to be anoverestimate, because the mothers incurred someenergy cost of activity. The percent energymetabolised by offspring was estimated as theproportion of energy allocated�pup that was

Ž .pupDEE Table 2 . The cost of lactation, energyallocated�litter and energy allocated�pup weregreater in Labradors than Schnauzers. However,in mass-specific terms, there were no significantdifferences in these variables between the breeds.

4. Discussion

We have previously shown that litter size varia-tion in dog breeds could, in part, be determined

Žby relative differences in maternal MEI Scantle-.bury et al., 2000b . Smaller animals generally eat

more food and produce a larger amount of milkin proportion to their own body weight than larger

Ž .animals Hanwell and Peaker, 1977 . In contrastto this prediction, Labradors appeared to havegreater energy requirements during lactation andto allocate more energy to their litters thanSchnauzers. Labrador pups were larger than


Ž . Ž . Ž .Fig. 3. Metabolisable energy intake MEI, kJ�day of a Labrador and b Schnauzer mothers against litter size. Sample sizes areshown as numbers above each point. Error bars denote �standard errors. There was no significant relationship between MEI and litter

Ž 2 . Ž 2 .size for Labradors r �0.125, P�0.127 , but there was for Schnauzers r �0.582, P�0.028 .

Schnauzers, and they grew faster, both absolutelyŽ .and relatively as a percent of maternal mass .

Therefore, the Labrador pups had greater main-Žtenance and growth energy requirements ab-

.solutely , which together with the larger litters,probably meant that the Labrador mothers had toallocate much more energy�litter than theSchnauzer mothers. In the current study, MEIexceeded energy expenditure, the primary differ-ence being the energy exported in milk. When

compared to predictions of maximal MEI,Labradors were higher than predictions and

ŽSchnauzers were lower Kirkwood, 1983; Weiner,.1989 . Labradors had a greater SusEI than

Ž .Schnauzers Table 2 , and although there was nosignificant difference in MEI between breedswhen the effects of body mass were removed withan ANCOVA, this only marginally failed to reachsignificance. Our results showed that, whilst MEIand total litter mass increased significantly with


Ž . Ž . Ž .Fig. 4. Daily energy expenditure DEE, kJ�day of a Labrador and b Schnauzer mothers against litter size. Sample sizes are shownas numbers above each point. Error bars denote �standard errors. There was no significant relationship between DEE and litter size

Ž 2 . Ž 2 .in either Labradors r �0.14, P�0.249 or Schnauzers r �0.00, P�0.504 .

litter size in Schnauzers, there were no significantchanges in these variables in Labradors, althoughthe trends were the same. Similarly, mean pupmass was not correlated with litter size in eitherbreed, although there was a tendency for pup

Žmass to decrease with increasing litter size Fig..1 . It is likely that we may not have detected these

changes because of the small sample sizes in-volved. Nevertheless, both breeds appeared to beable to meet increasing energy demands by in-creasing energy intake, and hence were not closeto the limits of their SusEI.

We did not attempt to attain limits of energyexpenditure or assimilation in these dogs, forexample by manipulating litter size or forcinganimals to thermoregulate at low temperatures.This was because of logistical constraints, includ-ing the difficulty of mating dogs synchronouslyand the possibility of rejection by foster mothersŽ .personal observations . Extremes of energy ex-penditure, at approximately 10�RMR, have beenobtained by forcing dogs to exercise at sub-zerotemperatures whilst the fur was wetted and they

Žwere fed iced water Cerretelli et al., 1964; Cha-


Table 2Ž . Ž .Mean�S.D. of body mass, litter size, resting metabolic rate RMR , metabolisable energy intake MEI , and daily energy expenditure

Ž . Ž .of mothers DEE and pups pupDEE for Labradors and Schnauzers

Labrador n Miniature n t PSchnauzer

Ž .Body mass kg 30.3�3.5 12 6.8�1.0 7 30.8 �0.001Ž .Litter size N 6.42�0.86 12 3.57�1.29 7 7.37 �0.001

aŽ .RMR kJ�day 3437�517 12 1062�282 4 16.3 �0.0010.75 bŽ .RMR kJ�kg day 266�21 12 267�43 4 0.05 �0.4

Metabolisable energy intakeaŽ .MEI kJ�day 22 448�3760 12 5382�2340 7 9.36 �0.001

0.75 bŽ .MEI kJ�kg day 1738 1278 3.46 �0.01cMEI�RMR 6.53 5.07 2.32 �0.05

Daily energy expenditure of mothersaŽ .DEE kJ�day 9808�1658 6 2619�2105 4 4.62 �0.01

0.75 bŽ .DEE kJ�kg day 759 621 0.15 �0.4cDEE�RMR 2.85 2.47 0.01 �0.4

Daily energy expenditure of pupsaŽ .PupDEE kJ�day 974�334 19 490�93 7 5.54 �0.001

0.75 bŽ .PupDEE kJ�kg day 75 121 3.77 �0.01cPupDEE�RMR 0.28 0.46 4.2 �0.010.83 dŽ .PupDLW kJ�kg day 518 540 0.37 0.4�P�0.25

Energy cost of lactationaŽ .DEE�RMR kJ�day 6371�1811 6 1557�680 4 3.01 �0.01

Ž 0.75 .bDEE�RMR kJ�kg day 481 370 0.1 �0.4cŽ .DEE�RMR �RMR 1.85 1.47 0.01 �0.4

�

Energy allocated� litteraŽ .MEI�DEE kJ�day 12 640�5588 6 2763�802 4 5.32 �0.001

0.75 bŽ .MEI�DEE kJ�kg day 979 656 1.48 0.1�P�0.05cŽ .MEI�DEE �RMR 3.7 2.6 1.3 0.25�P�0.1

Energy allocated�pupaŽ . Ž .MEI�DEE �N kJ�day 1969�580 6 774�62 4 3.12 �0.05

0.75 bŽ . Ž .MEI�DEE �N kJ�kg day 152 184 0.58 0.4�P�0.25cŽ .MEI�DEE �N�RMR 0.57 0.73 0.74 0.4�P�0.25

Energy metabolised by pupsŽ .% 49 63

Ž . Ž . �Ž . �Energy cost of lactation DEE-RMR , energy allocated�litter MEI-DEE , energy a llocated�pup MEI-DEE �N and % energy�Ž . Ž . �intake metabolised by pups pupDEE � MEI-DEE �N , are also shown. n denotes sample sizes. t and P values show the results of

two-sample t-tests comparing Labradors with Schnauzers.a The absolute value.b Ž 0.75.The value relative to maternal metabolic mass kg .c The value relative to maternal RMRd Ž 0.83.The value relative to offspring metabolic mass kg .

.tonet and Minaire, 1966 . However, it is unlikelythat these activities would be able to be main-

Ž .tained over longer periods several days . Further-more, the SusEI values of Labradors obtained inthe current study, even without such manipula-

tions, were high compared to reported valuesŽ .Hammond and Diamond, 1997 , which demon-strates the large amount of energy invested inreproduction by these dogs.

DEEs of both breeds were well below the allo-


Ž .metric predictions of 14.7 MJ�day LabradorsŽ .and 4.2 MJ�day Schnauzers for similarly sized

Ž .free-ranging eutherians Nagy, 1987 , and wellbelow any predicted ceiling on energy expendi-

Ž .ture Speakman, 1997b . By comparison, 25-kgAfrican wild dogs Lycaon pictus, slightly smallerthan the Labradors in the present study, had

Ž . ŽDEEs averaging 15.3 MJ�day 5.2�RMR Gor-. Ž .man et al., 1998 and Alaskan sled dogs 25 kg

Ž .used 47 MJ�day 7.5�RMR during a 70-hŽ .Arctic race Hinchcliff et al., 1997 . There was no

significant difference in mass-specific DEEbetween breeds. However, the large amount ofindividual variation, especially in the Schnauzermothers, may have obscured any difference. Un-der these conditions, the high mass-specific MEIof Labradors did not appear to be effected bymaternal DEE.

We found no significant difference in themass-specific energy allocated�litter betweenbreeds. This is seemingly at odds with theLabradors having a greater mass-specific energyintake that was apparently not metabolised orstored as fat. It is likely that the sample size wasnot large enough to detect a difference, becausethe faster growing Labrador litters clearly hadgreater mass-specific energy demands. Furtherwork is needed on this topic. DEEs of offspringwere similar when expressed relative to offspring

Ž .metabolic mass Oftedal, 1984b , and theLabrador pups were able to convert a greaterproportion of their energy intake into growth.Consequently, the mass-specific costs of support-ing offspring metabolism were greater for theSchnauzer than the Labrador mothers, which,together with the greater mass-specific energyintake of Labrador mothers, is consistent with theLabradors having the larger litters.

In summary, Labradors appeared to have highermass-specific MEIs than the Schnauzers duringpeak lactation, although the maternal DEEs werenot significantly different between breeds. In bothbreeds, maternal MEI and total litter mass in-clined to increase and individual pup mass de-crease with increasing litter size. DEEs of off-spring were similar when expressed relative tooffspring metabolic mass, and Labrador pups ap-peared to be able to convert more of their energyintake into growth. Hence, litter size variationswere consistent with the notion that the smaller

Ž .Schnauzer offspring should be individually morecostly for their mothers to produce, but not that

smaller mothers should allocate more resourcesto reproduction.

Acknowledgements

We are grateful for the help and support of allthose involved in this project. At WCPN, theseincluded Ivan Burger, Derek Booles, Tom Mc-Cappin, Amanda Hawthorne, Waring Hynds,David Poore, Matthew Gilham, Pat Nugent andDavid Lawson. We would like to thank PeterThomson for the mass spectrometry and SallyWard, Colin Selman and two anonymous refereesfor valuable comments on earlier drafts of themanuscript. Calculations of energy expenditureand turnover were carried out using a DLW cal-culation program: Lemen and Speakman, 1997http:��www.natureware.com�double.htm. Thework was funded by a grant from the WalthamCentre for Pet Nutrition to the University ofAberdeen.

References

Altman, P.L., Dittmer, D.S., 1962. Growth. Federationof American Sciences for Experimental Biology,Washington DC.

Brody, S., 1945. Methods in animal calorimetry: meth-ods of indirect calorimetry. Bioenergetics andGrowth. Reinhold, USA, pp. 313�344. chapter 12

Bronson, R.T., 1982. Variation in age at death of dogsof different sexes and breeds. Am. J. Vet. Res. 43Ž .11 , 2057�2059.

Burger, I.H., Johnson, J.V., 1991. Dogs large and small:the allometry of energy requirements within a single

Ž .species. J. Nutr. 121 11 , S18�S21.Cerretelli, P., Piiper, J., Mangili, F., Ricci, B., 1964.

Aerobic and anaerobic metabolism in exercisingŽ .dogs. J. Appl. Physiol. 19 1 , 25�28.

Chatonet, J., Minaire, Y., 1966. Comparison of energyexpenditure during exercise and cold exposure in thedog. Fed. Proc. 25, 1348�1350.

Clutton-Brock, J., 1995. Origins of the dog: domestica-Ž .tion and early history. In: Serpell, J. Ed. , The

Domestic Dog: Its Evolution, Behaviour and Inter-actions with People. Cambridge University Press,Cambridge, pp. 7�20.

Daan, S., Masman, D., Strikjstra, A., Verhukst, S., 1989.Intraspecific allometry of basal metabolic rate: rela-tions with body size, temperature, composition andcircadian phase in the kestrel, Falco tinnunculus. J.

Ž .Biol. Rhythms 4 2 , 267�283.


Davis, J.M., Valla, F.R., 1978. Evidence for domestica-tion of the dog 12,000 years ago in the Natufian ofIsrael. Nature 276, 608�610.

Drent, R.H., Daan, S., 1980. The prudent parent: ener-getic adjustments in avian breeding. Ardea 68,225�252.

Elia, M., Livesey, G., 1992. Energy expenditure andfuel selection in biological systems: the theory andpractice of calculations based on indirect calorimetry

Ž .and tracer methods. In: Simopoulos, A.P. Ed. ,Metabolic Control of Eating, Energy Expenditureand the Bioenergetics of ObesityWorld Review ofNutrition and Dietetics, 70. Karger, Basel, pp.68�131.

Evans, J.M., White, K., 1988. A Complete Guide toUnderstanding and Caring for Bitches. Henston.

Garland, T., Adolph, S.C., 1994. Why not do 2-speciescomparative-studies? � limitations on inferring

Ž .adaptation. Physiol. Zool. 67 4 , 797�828.Gessamen, J.A., Nagy, K.A., 1988. Energy metabolism

� errors in gas exchange conversion factors. Phys-Ž .iol. Zool. 61 6 , 507�513.

Gittleman, J.L., Thompson, S.D., 1988. Energy alloca-tion in mammalian reproduction. Am. Zool. 28,863�875.

Gorman, M.L., Mills, M.G., Raath, J.P., Speakman,J.R., 1998. High hunting costs make African wilddogs vulnerable to kleptoparasitism by hyenas. Na-ture 391, 479�481.

Hammond, K.A., Diamond, J., 1992. An experimentaltest for a ceiling on sustained metabolic rate in

Ž .lactating mice. Physiol. Zool. 65 5 , 952�977.Hammond, K.A., Diamond, J., 1994. Limits to dietary

nutrient intake and intestinal nutrient uptake inŽ .lactating mice. Physiol. Zool. 67 1 , 282�303.

Hammond, K.A., Diamond, J., 1997. Maximal sustainedenergy budgets in humans and animals. Nature 386,457�462.

Hammond, K.A., Konarzewski, M., Torres, R.M., Dia-mond, J., 1994. Metabolic ceilings under a combina-

Ž .tion of peak energy demands. Physiol. Zool. 67 6 ,1479�1506.

Hanwell, A., Peaker, M., 1977. Physiological effects oflactation on the mother. Symp. Zool. Soc. Lond. 41,297�312.

Hardy, R.N., 1972. Temperature and Animal Life.Arnold, London.

Heusner, A.A., 1991. Body mass maintenance and basalŽ .metabolism in dogs. J. Nutr. 121 11 , S8.

Hinchcliff, K.W., Reinhart, G.A., Burr, J.R., Schreier,C.J., Swenson, R.A., 1997. Metabolizable energy in-take and sustained energy expenditure of Alaskansled dogs during heavy exertion in the cold. Am. J.

Ž .Vet. Res. 58 12 , 1457�1462.Kienzle, E., Meyer, H., Dammers, C., Lohrie, H., 1985.

Milk intake, weight development, milk digestibility

and energy and nutrient retention in suckling pup-Ž .pies. In: Meyer, H. Ed. , Investigations on Nutrient

Requirements in Breeding Bitches and SucklingPups. Advances in Animal Physiology and AnimalNutrition; suppl. to J. Anim. Physiol. Anim. Nutr.Verlag Paul Parey, Hamburg und Berlin, pp. 26�50

Kirkwood, J.K., 1983. A limit to metabolisable energyintake in mammals and birds. Comp. Biochem. Phys-iol. 75A, 1�3.

Kirkwood, J.K., 1985. The influence of size on thebiology of the dog. J. Small Anim. Pract. 26, 97�110.

Kleiber, M., 1961. The Fire of Life. An Introduction toAnimal Energetics. John Wiley & Sons, New York.

Konarzewski, M., Diamond, J., 1994. Peak sustainedmetabolic rate and its individual variation in cold

Ž .stressed mice. Physiol. Zool. 67 5 , 1186�1212.Koteja, P., 1996. Limits to the energy budget in a

rodent, Peromyscus maniculatus � does gut capacityŽ .set the limit? Physiol. Zool. 69 5 , 981�993.

Krol, E., Speakman, J.R., 1999. Isotope dilution spacesof mice injected simultaneously with deuterium, tri-

Ž .tium and oxygen-18. J. Exp. Biol. 202 20 , 2839�2849.Lifson, N., Gordon, G.B., McClintock, R., 1955. Mea-

surement of total carbon dioxide production bymeans of D O18. J. Appl. Physiol. 7, 704�710.2

Loveridge, G., 1994. Provision of environmentally en-Ž .riched housing for dogs. Anim. Technol. 45 1 , 1�20.

Lyngset, A., Lyngset, O., 1970. Kullstørrelse hos hundŽ . Ž .Litter size in the dog . Norges Vet. Løgskole Oslo22, 186�191.

Meyer, H., 1984. Mineral metabolism and requirementsin bitches and suckling pups. In: Anderson, R.S.Ž .Ed. , Nutrition and Behaviour in Cats and Dogs.Pergamon Press, pp. 13�24.

Millar, J.S., 1978. Energetics of lactation in PeromyscusŽ .leucopus: the cost of lactation. Ecology 59 5 ,

1055�1061.Nagy, K.A., 1980. CO production in animals: an analy-2

sis of potential errors in the doubly labelled watertechnique. Am. J. Physiol. 238, R466�R473.

Nagy, K.A., 1987. Field metabolic rate and food re-quirement scaling in birds and mammals. Ecol.

Ž .Monogr. 57 2 , 111�118.Oftedal, O.T., 1984a. Lactation in the dog: milk compo-

sition and intake by puppies. J. Nutr. 114, 803�812.Oftedal, O.T., 1984b. Milk composition, milk yield and

energy output at peak lactation: a comparative re-view. Symp. Zool. Soc. Lond. 51, 33�85.

Oftedal, O.T., 1985. Pregnancy and lactation. In: Hud-Ž .son, R.J., White, R.G. Eds. , Bioenergetics of Wild

Herbivores. CRC Press, Boca Raton, FL, pp. 215�238chapter 10.

Paragon, B.M., Grandjean, D., 1993. Conduite du ra-tionnement de la chienne en periode de reproduc-

Ž .tion. Recueil Med. Vet. 169 4 , 223�230.


Peterson, C.C., Nagy, K.A., Diamond, J., 1990. Sus-tained metabolic scope. Proc. Natl. Acad. Sci. USA87, 2324�2328.

Rainbird, A., Booles, D., 1990. Construction and initialoperation of a whole body calorimeter for adult

Ž .dogs. In: Edney, A.T.B., Meyer, H., Kienzle, E. Eds. ,Nutrition, Malnutrition and Dietetics in the Dog andCat. British Veterinary Association, London, pp.42�43.

Rivers, J.P.W., Burger, I.H., 1989. Allometric consider-ations in the nutrition of dogs. In: Burger, I.H.,

Ž .Rivers, J.P.W. Eds. , Nutrition of the Dog and Cat.Cambridge University Press, Cambridge.

Robinson, R., 1973. Relationship between litter sizeand weight of dam in the dog. Vet. Rec. 92, 221�223.

Rogowitz, G.L., 1996. Trade-offs in energy allocationduring lactation. Am. Zool. 36, 197�204.

Scantlebury, M., Booles, D., Hynds, W., Speakman,J.R., 2000a. Isotope recycling in lactating dogs Canisfamiliaris. Am. J. Physiol. 278, R669�R676.

Scantlebury, M., Butterwick, R., Speakman, J.R., 2000b.ŽEnergetics of lactation in domestic dog Canis famil-

.iaris breeds of two sizes. Comp. Biochem. Physiol.Ž .125 2 , 197�210.

Schoeller, D.A., van Santen, E., 1982. Measurement ofenergy expenditure in humans by doubly labelled

Ž .water method. J. Appl. Physiol. 53 4 , 955�959.Speakman, J.R., 1997a. Doubly Labelled Water, Theory

and Practice. Chapman and Hall, London.Speakman, J.R., 1997b. Factors affecting the daily en-

ergy expenditure of small mammals. Proc. Nutr. Soc.56, 1119�1136.

Speakman, J.R., 2000. The cost of living: field metabolicrates of small mammals. In: Fitter, A.H., Raffaelli,

Ž .D.G. Eds. , Advances in Ecological Research.Academic Press.

Speakman, J.R., McQueenie, J., 1996. Limits to sus-tained metabolic rate, and morphology in reproduc-

Ž .ing mice, Mus musculus. Physiol. Zool. 69 4 ,746�769.

Speakman, J.R., Racey, P.A., 1988. Consequences ofnon-steady-state CO production for accuracy of the2doubly labelled water technique: the importance ofthe recapture interval. Comp. Biochem. Physiol. 90AŽ .2 , 337�340.

Speakman, J.R., Nair, K.S., Goran, M.I., 1993. Revisedequations for calculating CO production from2doubly labelled water in humans. Am. J. Physiol. 264Ž .6, pt. 1 , E912�E917.

Speakman, J.R., Booles, D., Burger, I.H. Validation ofthe doubly labelled water technique in domestic

Ž .dogs Canis familiaris by comparison with indirectcalorimetry. Br. J. Nutr., in press.

Vila, C., Savolaninen, P., Maldonado, J.E. et al., 1997.Multiple and ancient origins of the domestic dog.Science 276, 1687�1689.

Weiner, J., 1987. Maximum energy assimilation rates inŽ .the Djungarian hamster Phodopus sungorus . Oe-

Ž .cologia 72 2 , 297�302.Weiner, J., 1989. Metabolic constraints to mammalian

Ž .energy budgets. Acta Theriol. 34 1 , 3�35.Weiner, J., 1992. Physiological limits to sustainable

energy budgets in birds and mammals: ecologicalŽ .implications. Trends Evol. Ecol. 7 11 , 384�388.

Weir, J.B.deV., 1949. New methods for calculatingmetabolic rate with special reference to proteinmetabolism. J. Physiol. 109, 1�9.

Documents

Energetics and litter size variation in domestic dog Canis familiaris breeds of two sizes