- Email: [email protected]
Food consumption in relation to bodyweight in captive snakes
Research in Veterinary Science 1994, 57, 35-38
Food consumption in relation to bodyweight in captive snakes J. K. KIRKWOOD, C. GILI, Institute of Zoology, Regent's Park, London NW1 4RY
therefore often necessary to estimate their food requirements in the absence of any established method for doing so. In this study, the food consumption of 25 individuals of nine species of snakes housed in the Reptile House at the Zoological Society of London was measured in order to examine the variation in requirements in relation to bodyweight. Some of the snakes ate more than they required for maintenance and others ate less and it was therefore also possible to investigate the variation in the rates of change of bodyweight in relation to their food consumption rates. The results are relevant to the management and clinical care of snakes but are also of interest in comparison with the energetics of homeothermic animals.
The weekly food consumption of 25 captive snakes of nine species fed on freshly killed laboratory mice was measured for periods of three to eight weeks. Their ad libitum intake was found to vary in proportion to their initial body weight (w) raised to an exponent close to and not significantly different from 0.75 (mean [SE] 0.79 [0.064]). However, the food consumption in relation to W varied between animals, and the rates of change in bodyweight relative to w °75 during the measurement period were highly correlated with food consumption (r=0.801, P<0.001). From the regression describing the relationship between these variables, the mean food requirement for maintenance was estimated at 4.2 g day -1 kg -°'7s and the mean rate of weight loss when fasting was estimated at 2.0 g day -1 kg -°'7s. The mean weight gain for each gram of food eaten above maintenance level was estimated to be 0.46 g. The limits, in relation to w, within which the ad libitum food consumption of snakes can be predicted from these results with 95 per cent probability are estimated.
Materials and methods The snakes were housed in vivaria in the Reptile House at London Zoo. The air temperature was maintained at between 25 and 28°C, but some of the vivaria were also equipped with heat lamps or heating pads to p r o v i d e 'hot spots' for the snakes. The ranges of ambient temperature included the mean preferred body temperatures of most snakes (Jackson and Cooper 1984). The snakes were of different but uncertain ages and were therefore classed as either adult or juvenile (Table 1). Most of the snakes, both adults and juveniles, were gaining weight during the study. The snakes' bodyweights (w) were measured at the start of each week before they were fed. Freshly killed laboratory mice were offered once a week, and the keepers provided the quantity (and size) of mice that they considered appropriate. The weekly food consumption was measured by weighing these mice and subtracting the weight of any uneaten mice which were removed the following day, each week for periods of three to eight
SNAKES are maintained in captivity for conservation, research and as companion animals. They commonly suffer from anorexia owing to the difficulties of providing appropriate prey items (for example, snakes for the strictly snake-eating coral snakes, Micrurus species), or as a result of inappropriate environmental conditions or disease. Force-feeding snakes with mice or other diets is thus an important aspect of the management of some species and of supportive therapy. Although standard metabolic rates have been measured in variety of species of reptiles (Bennett and Dawson 1976), there appears to be little information on the rates of food consumption of lizards and snakes (Avery 1976, Bennett and Dawson 1976). It is 3~
36
J. K. Kirkwood, C. Gili
TABLE 1 : Initial body weight (w), food consumption (FC), and rate of change of weight (G) of 25 snakes at the Zoological Society of London
Species
Duration of study w FC G Age (weeks) (kg) (g week-1) (g week-t)
Liasis fuscus Liasis fuscus Boa constrictor Drymarchon corals Elaphe radiata Elaphe radiata Hydrodynastes gigas Hydrodynastes gigas Hydrodynastes gigas Hydrodynastes gigas Hydrodynastes gigas Lampropeltis getulus Lampropeltis getulus Lampropeltis getulus Lampropeltis triangulum Lampropeltis triangulum Lampropeltis triangulum Lampropeltis triangulum Lampropeltis triangulum Lampropeltis triangulum Lampropeltis fuliginosus Lampropeltis fuliginosus Lampropeltis fuliginosus Micrurus fulvius Micrurus fulvius
Ad Ad Juv Ad Juv Juv Ad Ad Ad Ad Ad Ad Ad Ad Juv Juv Juv Ad Juv Ad Juv Ad Juv Juv Ad
7 7 7 7 8 8 7 7 7 7 7 7 7 7 8 8 8 5 8 8 3 4 4 7 7
2.405 1.975 0.812 3.564 0.058 0-060 0.608 0-765 1.000 0.681 1.811 0.439 0.424 0.388 0.076 0.080 0.054 0.387 0.085 0.187 0.024 0.224 0.062 0.026 0.103
62.4 32.7 49.0 159.4 7.4 9.7 66.3 65.7 65.5 55.6 85.3 25.9 39.1 32.0 7.5 6.9 4.3 13.7 7.7 8.8 3.0 9.8 7.2 1.7 3.1
17.7 -1.4 13.3 49.5 0.9 2.9 25.2 23.0 13.8 12.4 29.6 1.4 5-0 2-7 2.3 2.0 0.3 -6.9 1.1 -0.2 1.4 3.4 1.0 -0.2 -2.5
Ad Adult, Juv Juvenile
weeks. The mean food consumption was calculated as the total weight of food consumed divided by the duration of the study in weeks. The mean weekly rates of change of bodyweight were calculated by subtracting the bodyweight at the start of the study from that at the end and dividing by the duration of the study in weeks. Snakes weighing less than 500 g, and the mice, were weighed to 0.1 g and the heavier snakes were weighed to the nearest 1 g.
/ 100
:"-
,//
8
o 00 /
/
/ 0
/
/'/
5 2
o
/
/
Bodyweight (kg) FIG 1 : Mean weekly food consumption of captive snakes in relation to bodyweight. The solid line shows the linear regression and the broken lines mark the 95 per cent prediction limits of food consumption for a snake of a given bodyweight (see text)
standard error of the slope was 0.064. This equation transforms to: FC = 49.8w 0'79
(2)
The exponent (0.79 + 0.064) did not differ significandy from 0.75. In view of this, for the subsequent analysis of the relationship between food consumption and change in bodyweight these data were expressed in terms o f the metabolic bodyweight (w°'75) in keeping with convention, to account for the variations due to bodyweight (Fig 2). There was a highly significant correlation between the food consumption per unit metabolic
The species of snakes studied, their initial bodyweights and the results of the measurements of food consumption and change in bodyweight and the period over which food intake was measured, are shown in Table 1. The variation in food consumption (Fc) in relation to the bodyweight of the snakes is shown in Fig 1. Linear regression of the natural logarithm of FC on the natural logarithm of w yielded the equation:
~:
The correlation coefficient was 0.93 and the
/
~- lO
~
(1)
/
/ ~~~#o °/ , •/ o o Do / / o ~ o /o//" J
•~
Results
lnFC = 3.908 + 0.787 lnW
50
,/
~
4o o
0 0 00 ~ ~ 000002 0
20 00
"O
o .=_
o
*~
~ -"' ~ " ~"
40
60
80
100
~ -20 Food consumption (g week 1 kg °7~) FIG 2: Rate of change in bodyweight in captive snakes in relation to the rate of food consumption, both expressed in terms of the metabolic bodyweight (kgO75). The line indicates the linear regression
Food consumption in relation to bodyweight in captive snakes bodyweight (Few) g week -1 kg-°'75 and change in bodyweight per unit metabolic bodyweight (Gw, g week -1 kg -0"75) (r=0.801, P<0.001). Linear regression yielded the equation: Gw =
-13.7 + 0.46 FCw
(3)
Discussion The ad libitum food consumption of the sample of captive snakes in this study is summarised by equation 2. The finding that the food intake of individual snakes increased with the bodyweight raised to an exponent close to, and not significantly different from 0.75, is consistent with the results of many studies of interspecies variation in energy intake and expenditure in both homeothermic and poikilothermic animals (Evans and Miller 1968, Bennett and Dawson 1976, Kirkwood 1981, Calder 1984). Three generalisations can be derived from equation 3. First, when FCw is zero, that is when fasting, the equation predicts a rate of loss of weight of 13.7 g week -1 w ~375, secondly, when G is zero, that is at maintenance conditions when weight is neither being gained nor lost, Fe w is predicted to be 29.8g week -1 w ~ 7 5 and, thirdly, the equation predicts that each g of food eaten in excess of that required for maintenance tends to result in a gain in bodyweight of 0.46 g. These predictions should be of value in the management and veterinary care of snakes because they provide a basis for the estimation of their food requirements for maintenance and growth. It would be inappropriate, however, to place too much confidence in the figures, especially those for fasting and maintenance metabolism, because they are dependent on extrapolating outside the range of the data, and because a relatively small difference in the slope of the regression could have a marked effect on the value of the intercept (the predicted weight loss when fasting). Furthermore, the regressions were based on only 25 snakes which were of different species, which were not equally represented, and not all of them were at the same stage of maturity. These factors may affect the degree to which the regressions approximate the relationships among all snakes. When using regression equations to predict variables such as the food consumption of an individual of known bodyweight it is useful to have an estimate of the likelihood that the prediction is
37
correct. The 95 per cent prediction limits, which indicate the range within which the food consumption of 95 per cent of the individuals of a given weight, taken from the same population as that on which the regression was based, provide this information. These limits can be wide even when the correlation is high (Kirkwood and Bennett 1992). The 95 per cent prediction limits of the regression of food consumption on bodyweight are shown in Fig 1. The results also enable a preliminary estimate of the rates of energy expenditure of snakes at maintenance and when fasting, and of the cost of tissue deposition in snakes. Mice have a gross energy value of about 8.4 kJ/g -1 freshweight (Kirkwood 198l). Not all of this energy is available for metabolism (some is lost in faeces and urine) and in both birds of prey and carnivorous mammals it has been found that about 75 per cent of the gross energy is available (Moors 1977, Kirkwood 1981). Thus the metabolisable energy (ME) content of a mouse is about 6-3 kJ g - 1 freshweight to carnivorous homeotherms and it is probably similar to snakes. Gehrmann (1971) found that the metabolisability of fish fed to grass snakes (Natrix erythrogaster) was 0.80. There are rather few data on the metabolisability of dietary energy in reptiles (Bennett and Dawson 1976) and no others for snakes as far as the authors are aware. Assuming an ME value of 6.3 kJ g-1 for mice, equation 3 predicts a mean ME requirement for maintenance of 27 kJ day -1 kg ~3"75, and a fasting metabolic rate of 12 kJ day -1 kg -0"75. This value is less than Bennett and Dawson's (1976) measurement of the standard metabolic rate of snakes kept at 30°C of 28 kJ day-1 kg-0"77 and the difference may have been due, partly, to the lower temperatures at which these snakes were kept. From equation 1, the mean ad libitum ME intake can be estimated at 45 kJ day-1 kg -0"79. The relationship between maintenance energy requirement and bodyweight appears not to have been estimated in snakes before. The results indicate that their maintenance requirements, like those previously found for the standard metabolic rate at 30°C (Bennett and Dawson 1976) are an order of magnitude lower than those typical for eutherian mammals of comparable bodyweights. The results also provide an estimate of the efficiency with which any food eaten in excess of a snake's maintenance requirement is used for growth. If it be assumed that the body composi-
38
J. K. Kirlcwood, C. Gili
tion, and thus the calorific density, of a snake and a mouse are similar, then for each kJ of mouse eaten in excess of maintenance requirements 0.46 kJ of snake tissue was deposited. Since it is likely that only 0-75 kJ is available (metabolisable) from each kJ of mouse eaten then the proportion of the available energy that is deposited is 0.61 (0.46/0.75). Mice consist largely of fat and protein and the theoretical energetic efficiencies of fat and protein synthesis from dietary fat and protein are 0.99 and 0.85, respectively (Millward et al 1976), but in practice the efficiencies of conversion are usually found to be lower (Blaxter 1971). Acknowledgements The authors are grateful to Mr David Ball, the Curator of Reptiles, and his staff for their assistance with this study and to the referees for helpful comments on the paper. References AVERY, R, A. (1976) Thermoregulation, metabolism and social behaviour. In Morphology and Biology of Reptiles. Eds A. d'A. Bellairs
and B. C. Cox, London, Academic Press. pp 245-259 BENNETT, A. F. & DAWSON, W. R. (1976) Metabolism. In Biology of the Reptilia. VoI 5. Eds C. Gans and W. R. Dawson. London, Academic Press. pp 127-223 BLAXTER, K. L. (1971) Methods of measuring the energy metabolism of animals and interpretation of results obtained. Federation Proceedings 30, 1436-1443 CALDER, W. A. lII. (1984) Size, Function and Life History. Cambridge, Massachusetts, Harvard University Press EVANS, E. & MILLER, D. S. (1968) Comparative nutrition, growth and longevity. Proceedings of the Nutrition Society 27, 121-129 GEHRMANN, W. H. (1971) Food consumption and growth in the immature water snake Natrix erythrogaster transversus. Growth 35, 127-136 JACKSON, O. F. & COOPER, J. E. (1984) Nutritional diseases. In Diseases of the Reptilia. Vol 2. Eds J. E. Cooper and O. F. Jackson. London, Academic Press. pp 409-428 KIRKWOOD, J. K. (1981) Maintenance energy requirements and rate of weight loss during starvation in birds of prey. In Recent Advances in the Study of Raptor Diseases. Eds J. E. Cooper and A. G. Greenwood. Keighley, Chiron Publications. pp 153-157 KIRKWOOD, J. K. & BENNETT, P. M. (1992) Approaches and limitations to the prediction of energy requirements in wild animal husbandry and veterinary care. Proceedings of the Nutrition Society 51, 117-124 MILLWARD, D. J., GARLICK, P. J. & REEDS, P. J. (1976) The energy cost of growth. Proceedings of the Nutrition Society 35, 339-349 MOORS, P. J. (1997) Studies on the metabolism, food consumption and assimilation efficiency of a small carnivore, the weasel (Mustela nivalis, L). Oecologia 37, 185-202
Received September 10,1993 Accepted November 30,1993
Food consumption in relation to bodyweight in captive snakes J. K. KIRKWOOD, C. GILI, Institute of Zoology, Regent's Park, London NW1 4RY
therefore often necessary to estimate their food requirements in the absence of any established method for doing so. In this study, the food consumption of 25 individuals of nine species of snakes housed in the Reptile House at the Zoological Society of London was measured in order to examine the variation in requirements in relation to bodyweight. Some of the snakes ate more than they required for maintenance and others ate less and it was therefore also possible to investigate the variation in the rates of change of bodyweight in relation to their food consumption rates. The results are relevant to the management and clinical care of snakes but are also of interest in comparison with the energetics of homeothermic animals.
The weekly food consumption of 25 captive snakes of nine species fed on freshly killed laboratory mice was measured for periods of three to eight weeks. Their ad libitum intake was found to vary in proportion to their initial body weight (w) raised to an exponent close to and not significantly different from 0.75 (mean [SE] 0.79 [0.064]). However, the food consumption in relation to W varied between animals, and the rates of change in bodyweight relative to w °75 during the measurement period were highly correlated with food consumption (r=0.801, P<0.001). From the regression describing the relationship between these variables, the mean food requirement for maintenance was estimated at 4.2 g day -1 kg -°'7s and the mean rate of weight loss when fasting was estimated at 2.0 g day -1 kg -°'7s. The mean weight gain for each gram of food eaten above maintenance level was estimated to be 0.46 g. The limits, in relation to w, within which the ad libitum food consumption of snakes can be predicted from these results with 95 per cent probability are estimated.
Materials and methods The snakes were housed in vivaria in the Reptile House at London Zoo. The air temperature was maintained at between 25 and 28°C, but some of the vivaria were also equipped with heat lamps or heating pads to p r o v i d e 'hot spots' for the snakes. The ranges of ambient temperature included the mean preferred body temperatures of most snakes (Jackson and Cooper 1984). The snakes were of different but uncertain ages and were therefore classed as either adult or juvenile (Table 1). Most of the snakes, both adults and juveniles, were gaining weight during the study. The snakes' bodyweights (w) were measured at the start of each week before they were fed. Freshly killed laboratory mice were offered once a week, and the keepers provided the quantity (and size) of mice that they considered appropriate. The weekly food consumption was measured by weighing these mice and subtracting the weight of any uneaten mice which were removed the following day, each week for periods of three to eight
SNAKES are maintained in captivity for conservation, research and as companion animals. They commonly suffer from anorexia owing to the difficulties of providing appropriate prey items (for example, snakes for the strictly snake-eating coral snakes, Micrurus species), or as a result of inappropriate environmental conditions or disease. Force-feeding snakes with mice or other diets is thus an important aspect of the management of some species and of supportive therapy. Although standard metabolic rates have been measured in variety of species of reptiles (Bennett and Dawson 1976), there appears to be little information on the rates of food consumption of lizards and snakes (Avery 1976, Bennett and Dawson 1976). It is 3~
36
J. K. Kirkwood, C. Gili
TABLE 1 : Initial body weight (w), food consumption (FC), and rate of change of weight (G) of 25 snakes at the Zoological Society of London
Species
Duration of study w FC G Age (weeks) (kg) (g week-1) (g week-t)
Liasis fuscus Liasis fuscus Boa constrictor Drymarchon corals Elaphe radiata Elaphe radiata Hydrodynastes gigas Hydrodynastes gigas Hydrodynastes gigas Hydrodynastes gigas Hydrodynastes gigas Lampropeltis getulus Lampropeltis getulus Lampropeltis getulus Lampropeltis triangulum Lampropeltis triangulum Lampropeltis triangulum Lampropeltis triangulum Lampropeltis triangulum Lampropeltis triangulum Lampropeltis fuliginosus Lampropeltis fuliginosus Lampropeltis fuliginosus Micrurus fulvius Micrurus fulvius
Ad Ad Juv Ad Juv Juv Ad Ad Ad Ad Ad Ad Ad Ad Juv Juv Juv Ad Juv Ad Juv Ad Juv Juv Ad
7 7 7 7 8 8 7 7 7 7 7 7 7 7 8 8 8 5 8 8 3 4 4 7 7
2.405 1.975 0.812 3.564 0.058 0-060 0.608 0-765 1.000 0.681 1.811 0.439 0.424 0.388 0.076 0.080 0.054 0.387 0.085 0.187 0.024 0.224 0.062 0.026 0.103
62.4 32.7 49.0 159.4 7.4 9.7 66.3 65.7 65.5 55.6 85.3 25.9 39.1 32.0 7.5 6.9 4.3 13.7 7.7 8.8 3.0 9.8 7.2 1.7 3.1
17.7 -1.4 13.3 49.5 0.9 2.9 25.2 23.0 13.8 12.4 29.6 1.4 5-0 2-7 2.3 2.0 0.3 -6.9 1.1 -0.2 1.4 3.4 1.0 -0.2 -2.5
Ad Adult, Juv Juvenile
weeks. The mean food consumption was calculated as the total weight of food consumed divided by the duration of the study in weeks. The mean weekly rates of change of bodyweight were calculated by subtracting the bodyweight at the start of the study from that at the end and dividing by the duration of the study in weeks. Snakes weighing less than 500 g, and the mice, were weighed to 0.1 g and the heavier snakes were weighed to the nearest 1 g.
/ 100
:"-
,//
8
o 00 /
/
/ 0
/
/'/
5 2
o
/
/
Bodyweight (kg) FIG 1 : Mean weekly food consumption of captive snakes in relation to bodyweight. The solid line shows the linear regression and the broken lines mark the 95 per cent prediction limits of food consumption for a snake of a given bodyweight (see text)
standard error of the slope was 0.064. This equation transforms to: FC = 49.8w 0'79
(2)
The exponent (0.79 + 0.064) did not differ significandy from 0.75. In view of this, for the subsequent analysis of the relationship between food consumption and change in bodyweight these data were expressed in terms o f the metabolic bodyweight (w°'75) in keeping with convention, to account for the variations due to bodyweight (Fig 2). There was a highly significant correlation between the food consumption per unit metabolic
The species of snakes studied, their initial bodyweights and the results of the measurements of food consumption and change in bodyweight and the period over which food intake was measured, are shown in Table 1. The variation in food consumption (Fc) in relation to the bodyweight of the snakes is shown in Fig 1. Linear regression of the natural logarithm of FC on the natural logarithm of w yielded the equation:
~:
The correlation coefficient was 0.93 and the
/
~- lO
~
(1)
/
/ ~~~#o °/ , •/ o o Do / / o ~ o /o//" J
•~
Results
lnFC = 3.908 + 0.787 lnW
50
,/
~
4o o
0 0 00 ~ ~ 000002 0
20 00
"O
o .=_
o
*~
~ -"' ~ " ~"
40
60
80
100
~ -20 Food consumption (g week 1 kg °7~) FIG 2: Rate of change in bodyweight in captive snakes in relation to the rate of food consumption, both expressed in terms of the metabolic bodyweight (kgO75). The line indicates the linear regression
Food consumption in relation to bodyweight in captive snakes bodyweight (Few) g week -1 kg-°'75 and change in bodyweight per unit metabolic bodyweight (Gw, g week -1 kg -0"75) (r=0.801, P<0.001). Linear regression yielded the equation: Gw =
-13.7 + 0.46 FCw
(3)
Discussion The ad libitum food consumption of the sample of captive snakes in this study is summarised by equation 2. The finding that the food intake of individual snakes increased with the bodyweight raised to an exponent close to, and not significantly different from 0.75, is consistent with the results of many studies of interspecies variation in energy intake and expenditure in both homeothermic and poikilothermic animals (Evans and Miller 1968, Bennett and Dawson 1976, Kirkwood 1981, Calder 1984). Three generalisations can be derived from equation 3. First, when FCw is zero, that is when fasting, the equation predicts a rate of loss of weight of 13.7 g week -1 w ~375, secondly, when G is zero, that is at maintenance conditions when weight is neither being gained nor lost, Fe w is predicted to be 29.8g week -1 w ~ 7 5 and, thirdly, the equation predicts that each g of food eaten in excess of that required for maintenance tends to result in a gain in bodyweight of 0.46 g. These predictions should be of value in the management and veterinary care of snakes because they provide a basis for the estimation of their food requirements for maintenance and growth. It would be inappropriate, however, to place too much confidence in the figures, especially those for fasting and maintenance metabolism, because they are dependent on extrapolating outside the range of the data, and because a relatively small difference in the slope of the regression could have a marked effect on the value of the intercept (the predicted weight loss when fasting). Furthermore, the regressions were based on only 25 snakes which were of different species, which were not equally represented, and not all of them were at the same stage of maturity. These factors may affect the degree to which the regressions approximate the relationships among all snakes. When using regression equations to predict variables such as the food consumption of an individual of known bodyweight it is useful to have an estimate of the likelihood that the prediction is
37
correct. The 95 per cent prediction limits, which indicate the range within which the food consumption of 95 per cent of the individuals of a given weight, taken from the same population as that on which the regression was based, provide this information. These limits can be wide even when the correlation is high (Kirkwood and Bennett 1992). The 95 per cent prediction limits of the regression of food consumption on bodyweight are shown in Fig 1. The results also enable a preliminary estimate of the rates of energy expenditure of snakes at maintenance and when fasting, and of the cost of tissue deposition in snakes. Mice have a gross energy value of about 8.4 kJ/g -1 freshweight (Kirkwood 198l). Not all of this energy is available for metabolism (some is lost in faeces and urine) and in both birds of prey and carnivorous mammals it has been found that about 75 per cent of the gross energy is available (Moors 1977, Kirkwood 1981). Thus the metabolisable energy (ME) content of a mouse is about 6-3 kJ g - 1 freshweight to carnivorous homeotherms and it is probably similar to snakes. Gehrmann (1971) found that the metabolisability of fish fed to grass snakes (Natrix erythrogaster) was 0.80. There are rather few data on the metabolisability of dietary energy in reptiles (Bennett and Dawson 1976) and no others for snakes as far as the authors are aware. Assuming an ME value of 6.3 kJ g-1 for mice, equation 3 predicts a mean ME requirement for maintenance of 27 kJ day -1 kg ~3"75, and a fasting metabolic rate of 12 kJ day -1 kg -0"75. This value is less than Bennett and Dawson's (1976) measurement of the standard metabolic rate of snakes kept at 30°C of 28 kJ day-1 kg-0"77 and the difference may have been due, partly, to the lower temperatures at which these snakes were kept. From equation 1, the mean ad libitum ME intake can be estimated at 45 kJ day-1 kg -0"79. The relationship between maintenance energy requirement and bodyweight appears not to have been estimated in snakes before. The results indicate that their maintenance requirements, like those previously found for the standard metabolic rate at 30°C (Bennett and Dawson 1976) are an order of magnitude lower than those typical for eutherian mammals of comparable bodyweights. The results also provide an estimate of the efficiency with which any food eaten in excess of a snake's maintenance requirement is used for growth. If it be assumed that the body composi-
38
J. K. Kirlcwood, C. Gili
tion, and thus the calorific density, of a snake and a mouse are similar, then for each kJ of mouse eaten in excess of maintenance requirements 0.46 kJ of snake tissue was deposited. Since it is likely that only 0-75 kJ is available (metabolisable) from each kJ of mouse eaten then the proportion of the available energy that is deposited is 0.61 (0.46/0.75). Mice consist largely of fat and protein and the theoretical energetic efficiencies of fat and protein synthesis from dietary fat and protein are 0.99 and 0.85, respectively (Millward et al 1976), but in practice the efficiencies of conversion are usually found to be lower (Blaxter 1971). Acknowledgements The authors are grateful to Mr David Ball, the Curator of Reptiles, and his staff for their assistance with this study and to the referees for helpful comments on the paper. References AVERY, R, A. (1976) Thermoregulation, metabolism and social behaviour. In Morphology and Biology of Reptiles. Eds A. d'A. Bellairs
and B. C. Cox, London, Academic Press. pp 245-259 BENNETT, A. F. & DAWSON, W. R. (1976) Metabolism. In Biology of the Reptilia. VoI 5. Eds C. Gans and W. R. Dawson. London, Academic Press. pp 127-223 BLAXTER, K. L. (1971) Methods of measuring the energy metabolism of animals and interpretation of results obtained. Federation Proceedings 30, 1436-1443 CALDER, W. A. lII. (1984) Size, Function and Life History. Cambridge, Massachusetts, Harvard University Press EVANS, E. & MILLER, D. S. (1968) Comparative nutrition, growth and longevity. Proceedings of the Nutrition Society 27, 121-129 GEHRMANN, W. H. (1971) Food consumption and growth in the immature water snake Natrix erythrogaster transversus. Growth 35, 127-136 JACKSON, O. F. & COOPER, J. E. (1984) Nutritional diseases. In Diseases of the Reptilia. Vol 2. Eds J. E. Cooper and O. F. Jackson. London, Academic Press. pp 409-428 KIRKWOOD, J. K. (1981) Maintenance energy requirements and rate of weight loss during starvation in birds of prey. In Recent Advances in the Study of Raptor Diseases. Eds J. E. Cooper and A. G. Greenwood. Keighley, Chiron Publications. pp 153-157 KIRKWOOD, J. K. & BENNETT, P. M. (1992) Approaches and limitations to the prediction of energy requirements in wild animal husbandry and veterinary care. Proceedings of the Nutrition Society 51, 117-124 MILLWARD, D. J., GARLICK, P. J. & REEDS, P. J. (1976) The energy cost of growth. Proceedings of the Nutrition Society 35, 339-349 MOORS, P. J. (1997) Studies on the metabolism, food consumption and assimilation efficiency of a small carnivore, the weasel (Mustela nivalis, L). Oecologia 37, 185-202
Received September 10,1993 Accepted November 30,1993