Effect of ewe age and high population density on the early nursing behaviour of mouflon

Transcription

1 Ethology Ecology & Evolution 7: , 1995 Effect of ewe age and high population density on the early nursing behaviour of mouflon D. RÉALE and P. BOUSSÈS 1 Laboratoire d Evolution des Systèmes Naturels et Modifiés, Muséum National d Histoire Naturelle - U.R.A. 1853, 36 rue Geoffroy St-Hilaire, Paris, France Received 26 September 1994, accepted 8 September 1995 The population of mouflons (Ovis musimon) inhabiting the Kerguelen archipelago shows marked density fluctuations with periodic winter die-offs. Compared with two European populations, parental care of Kerguelen mouflon females is low at peak population density, when resources are depleted. Mean suckle duration did not differ markedly between Kerguelen and European ewes. In contrast, Kerguelen mothers suckling frequency was lower, and the overall suckling time was also reduced. In the Kerguelen population females rejected most suckling attempts and ended more than 90% of suckling bouts even within the first few days of a lamb s life. Parental care decreased as ewe age increased, with a decline in both suckling frequency and total suckle duration. As a result, lambs of old females devoted much of their time budget to grazing. KEY WORDS: mouflon, nursing behaviour, high population density, resource depletion, ewe age. Introduction Methods Results Comparison with European populations Effects of ewes age Discussion Effects of ecological conditions on suckling patterns Effects of ewes age on suckling patterns Acknowledgements References Present address: Large Animal Research Group, Department of Zoology, University of Cambridge, CB2 3EJ, UK.

2 324 D. Réale and P. Boussès INTRODUCTION Following TRIVERS hypothesis (1972, 1974), it is assumed that parental investment and parent-offspring conflict may vary with ecological conditions (CAR- LISLE 1982, CLUTTON-BROCK 1991). Under poor conditions, parents that invest highly in their current offspring could reduce their residual reproductive value and survival. Thus, parents would benefit from lowering their parental investment under poor conditions (CARLISLE 1982). On the other hand, the residual reproductive value of females usually decreases with age, and females should increase the reproductive effort devoted to their current offspring as they get older (WILLIAMS 1966, STEARNS 1976). Lactation is by far the largest form of energetic expenditure in mammals rearing young (GITTLEMAN & THOMPSON 1988, CLUTTON-BROCK et al. 1989). Among ungulates, milk yield varies with female nutritional status during gestation and lactation, and has important consequences for offspring growth rate (MUNRO 1962, LOU- DON et al. 1983). Thus, as suggested by GEIST (1971), nursing behaviour might be a good indicator of environmental quality for sheep populations. In this paper, we examine the hypothesis that maternal care in mouflon ewes (Ovis musimon Pallas 1811) is lower under poor environmental conditions, based on the Ile Haute population, Kerguelen subantarctic archipelago. To test this hypothesis, early suckling and grazing behaviours of mouflon lambs were examined at high population density and low food availability, and compared with European populations living under conditions of greater food availability. We also investigated to what extent ewe age affects nursing parameters. METHODS Study area and population. The Ile Haute (49 24 S, E) covers an area of 6.5 km 2 and rises to an elevation of 321 m. Climatic data have been collected since 1951 at the Portaux-Français permanent station 20 km away. Annual variation in air temperature is low, with a mean temperature of 1.9 C in the coldest month (July) and a mean of 7.4 C in the warmest (January). Annual rainfall averages 817 mm, with approximately 222 days of rain, including 129 days of snow. Nevertheless, the snow pack rarely exceeds 10 cm. The Ile Haute landscape, by the absence of trees and woody shrubs, is characteristic of tundra habitats. Twelve fodder plant species were introduced in 1973 in order to improve the island s carrying capacity (BOURZAT & MONIÉ 1977). Nevertheless, swards remain localised and represent less than 15% of the area, while fell fields cover more than 60% of the island (BOUSSÈS et al. 1994). Following the introduction of a pair of mouflon in 1957 (CHAPUIS et al. 1994), the population had reached the carrying capacity of the island by the end of the seventies (BOUSSÈS et al. 1992). Since then, due to periodic die-offs, the density has fluctuated from 50 to 110 ind./km 2. The population has not been hunted since 1989, and predators and interspecific competitors are absent. Data collection. The study was carried out at peak population density from October 20th to November 20th The lambing season occurs from mid-october to mid-november. In 1991, births peaked at the end of the month and were highly synchronised. Data were collected using a spotting-scope ( 20-60), from two hides. Observations were performed during the main feeding periods, i.e. the 3 hr following sunrise and the 3 hr before sunset. Females with lambs were selected at random and observed for 10 min using the focal sampling method (ALTMANN 1974).

3 Mouflon nursing behaviour at high population density 325 Three female age classes were considered (young: 2 years old; prime-aged: 3-5 years old; old: 6 years old) on the basis of morphological traits and head colouration (PFEFFER 1967, BOUSSÈS & RÉALE 1994). Between 20 to 30% of females twin in any year. As they very rarely raised twins, young females with twins were not considered. Analysis was limited to the first 3 weeks of a lamb s life and three age categories were distinguished: 1-3 days old (29.4% of focal samples), 4-10 days old (52.5%) and days old (18.1%, n = 160). Lamb age was estimated using morphological and behavioural traits, and with reference to some lambs of known age. Most lambs were about 3 weeks old at the end of the study. About 180 females were present in the two observation areas, representing about 2/3 of the females. The lamb sex ratio was determined in March for the three ewe classes during two censuses performed along a fixed route covering the entire island. Data on the Caroux-Espinouse population were obtained by R. Bon during the lambing seasons of (unpublished data). This population numbered about 1000 mouflons distributed over ha (average 8 ind./km 2 ) of low mountains in the south of France (BON & CAMPAN 1989). The climate is under oceanic influence during winter and Mediterranean influence during summer, resulting in vegetation characterised by a mosaic of Mediterranean and oceanic facies. Thirty focal watches ranging from 6 to 300 min, were performed on females with lambs less than 20 days old (30.0% of 1-3 days old, 26.7% of 4-10 days old, and 43.3% of days old). Observations were distributed throughout the day. Ten samples for which total duration were not available were added for the calculation of mean suckle duration and success rate. Data on the Cazorla population are from OBREGÓN et al. (1992a). A herd of 7 females with their young inhabited an 8 ha enclosure in the Parque Natural de Cazorla (Spain). They feed on natural vegetation and benefit from additional food, water and salt supplies. We used the data for lambs in their first 2 weeks of life. Behavioural parameters. A suckling action occurred when the lamb touched its mother s udder with its nose. Successful suckles lasted 5 sec or more, and shorter suckles were considered as unsuccessful attempts (SHACKLETON & HAYWOOD 1985). Successful suckles separated by more than 10 sec were considered as two suckles. We used the following parameters: the suckling frequency and the total suckling time per 10 min of observation, the mean suckle duration, the frequency of unsuccessful suckles and the rate of successful suckles. When several suckles occurred during a focal sample, the mean duration was used in order to ensure independence of data. This explains why the product of suckling frequency and mean suckle duration do not exactly equal total suckling time (Table 1). Females that initiated a suckling bout stood in a specific suckling position, with their hind legs apart, the head up, and sometime bleated to call Table 1. Comparison of nursing indices between two European and the Kerguelen mouflon populations. SF = suckling frequency, MSD = mean suckle duration, TSD = total suckle duration, SR = success rate, and USAF = unsuccessful suckling attempts frequency. Data on Cazorla are from OBREGÓN et al. (1992a), and concern lambs aged 15 days. Data on Caroux-Espinouse population were obtained by R. Bon in (unpublished), and concern lambs of 1 to 20 days of age (for more details see the text). Percentage termination by the mother are from BON (1985). Population SF MSD TSD SR USAF % termina- (nb/10 min) (sec) (sec/10 min) (%) (nb/10 min) tion by the mother Cazorla females males Caroux 0.44 ± ± ± ± ± Kerguelen 0.34 ± ± ± ± ±

4 326 D. Réale and P. Boussès their lamb. A lamb that successfully reached the udder before its mother assumed this position was considered the initiator of the suckle. Females ended the suckle by moving away. Some allosuckling attempts occurred but were not taken into account for the nursing parameters calculation. The behaviour of the lamb was noted at 1 min intervals (11 observations per focal and per lamb). The proportion of time lambs spent grazing was estimated as the number of observations during which each lamb grazed divided by the total number of observations. Data analysis. Most mouflons were not marked so independence of data cannot be assumed. To limit the risk of pseudoreplication, we selected females in different groups or age classes during each focal session. Moreover, we observed about 20 females per day, though more than 100 lactating females usually used the area. Kruskall-Wallis tests followed by post-hoc tests were used for comparisons between ewe classes. Kendall tests were used to analyse the relationships between the mother s age, or the lamb s age, and suckling indices or lambs grazing activity. Chi-square tests were used to test differences between female age classes in the frequency of initiation and termination of suckles by the mothers. The occurrence of zero values due to short sample duration in the Kerguelen observations could have biased between-population comparisons of suckling frequency or overall suckle duration, leading to a type I error with a Mann-Whitney test. Thus, we restricted tests to mean suckle duration, success rate and the frequency of unsuccessful attempts, for comparisons between the Kerguelen and the Caroux populations. RESULTS The presence of one or two lambs did not affect nursing parameters whatever the age of the ewe (prime-aged or old). Thus, data were pooled in regard to these parameters. We found no offspring sex ratio difference between ewe age classes (c 2 = , df = 2, P = 0.96). Comparison with European populations. Observation periods (i.e. main feeding period vs mid-day period) in the Caroux population had no effect on suckling parameters (Mann-Whitney test, n 1 = 17, n 2 = 13, suckling frequency, z = 0.25, P = 0.8, mean suckle duration, z = 0.65, P = 0.5, total suckle duration, z = 0.34, P = 0.7, unsuccessful suckling attempts frequency, z = 0.82, P = 0.4, success rate, z = 0.69, P = 0.5). Thus, differences in sampling procedures are unlikely to affect the comparisons between the Kerguelen and the Caroux populations. Mean suckle duration did not differ between the Caroux and the Kerguelen populations (Table 1, z = 0.86, P = 0.27, n 1 = 34, n 2 = 41). Suckling frequency tended to be higher, and total suckling time was longer in the Caroux than in the Kerguelen population. In Kerguelen, the frequency of unsuccessful suckles was significantly higher, and success rate significantly lower than in the Caroux population (z = 2.87, P < 0.01, n 1 = 30, n 2 = 160 and z = 6.72, P < 0.001, n 1 = 36, n 2 = 105). The same trends were observed compared to the Cazorla population. Kerguelen females rejected about 80% of the suckling attempts compared to less than 40% for European females. Even during the first 10 days of the lambs life, Kerguelen females ended 93% of suckles, compared to only 34% for the Caroux population (BON 1985). In Cazorla, during the first week of a lamb s life, mothers ended 38.6% of the suckles when the lamb was a male and 75.0% of suckles of female lambs (Table 1). Effects of ewes age. The age of sampled lambs did not differ between ewe age classes in the Kerguelen population (c 2 = 5.61, df = 4, P = 0.23). For young females, mean suckle duration, suckling frequency, total suckling time, and success rate

5 Mouflon nursing behaviour at high population density 327 remained stable as lambs grew older (Table 2). Unsuccessful suckling attempt frequency was the only parameter that decreased significantly with lamb age. Suckling frequency and total suckling time decreased significantly with lamb age for prime-aged females (Table 2). Lambs of old females performed more suckling attempts as they got older (Table 2). Other indices did not change with lamb age. The data were pooled and compared across ewe age classes. There was no effect of ewe age on mean suckling duration (Fig. 1A, Kruskall-Wallis test, Hc = 2.75, df = 2, P = 0.25). Suckling frequency differed significantly between age classes (Fig. 1B, Hc = 11.89, df = 2, P < 0.01). This was due to the fact that young females suckled their lambs much more frequently than old ones (multiple comparisons, P < 0.05). The overall suckling time differed significantly between female classes (Fig. 1C, Hc = 12.61, df = 2, P < 0.01), the lambs of young females suckling for longer periods than those of old females (15.2 sec vs 3.7 sec per 10 min; P < 0.05). The frequency of unsuccessful suckling attempts did not vary with ewe age classes (Fig. 2A, Hc = 0.53, df = 2, P = 0.75). The apparently higher frequency of suckling attempts for lambs of old ewes was due to some lambs performing a very high number of attempts (up to 62 per 10 min). Success rate differed significantly between female classes (Fig. 2B, Hc = 11.79, df = 2, P < 0.01). It was significantly reduced for old females compared with young females (P < 0.01). There was no difference between young and prime-aged females nor between prime-aged females and old ones (P > 0.05). Females prevented their lamb from suckling by pawing the hind leg or by moving when the lamb tried to establish contact with the udder. We never observed any aggressive behaviour (e.g. head threat or butting) by the female toward its young after a suckling attempt. Whatever their mothers age, lambs initiated and ended the same number of successful suckles (Fig. 3, c 2 = 2.78, df = 2, P = 0.25 and c 2 = 1.71, df = 2, P = 0.43 respectively). During 160 focal samples we observed 23 attempts by the focal lambs to suckle strange females, and 18 suckle attempts by strange lambs on focal fe- Table 2. Development of nursing parameters with lamb s age. Kendall s tau values were used to test the effect of lamb s age on nursing parameters. For abbreviations see Table1. Female classes Nursing parameters t n P Young SF NS MSD NS TSD NS SR NS USAF < 0.01 Prime-aged SF < 0.05 MSD NS TSD < 0.05 SR NS USAF NS Old SF NS MSD NS TSD NS SR NS USAF < 0.05

6 328 D. Réale and P. Boussès Fig. 1. Nursing indices (± SE) for the three age classes of ewes. (A) mean suckle duration, (B) suckling frequency, (C) total suckling time. Number of focal samples is indicated above the bars. Significance levels (* P < 0.05). males. There was no difference among the three ewe age classes (Hc = 0.37, df = 2, P = 0.80, and Hc = 0.06, df = 2, P = 0.98). Allosuckling attempts were always rejected, except in one case: a female, observed with a dying lamb a few hours earlier suckled foreign twins twice on consecutive occasions. The proportion of time spent grazing increased with lamb age for the three classes of females (Fig. 4, Kendall s test, young: t = 0.39, P < 0.001, n = 40, prime-

7 Mouflon nursing behaviour at high population density 329 Fig. 2. (A) Unsuccessful suckling attempt frequency (± SE), and (B) success rate of suckling (± SE), for lambs of three age classes of ewes. Number of focal samples are indicated above the bars. Significance levels (** P < 0.01). Fig. 3. Proportion of successful suckles initiated (black) and ended (shaded) by the mother for the different age classes of ewes. Number of suckles are indicated above the bars.

8 330 D. Réale and P. Boussès Fig. 4. Proportion of time spent grazing (± SE) for lambs of young (white), prime-aged (shaded), and old ewes (black). Number of focal samples are indicated above the bars. aged: t = 0.32, P < 0.01, n = 37, old: t = 0.26, P < 0.01, n = 79). Significant positive relationships were observed between mother s age and the proportion of time devoted to grazing for 1-3 and 4-10 days old lambs (Fig. 4, t = 0.23, P < 0.05, n = 45 and t = 0.16, P < 0.05, n = 82 respectively), but not for 11 to 20 days old lambs (t = 0.12, P = 0.26, n = 29). Finally, the proportion of time spent grazing was positively correlated with number of unsuccessful suckles (t = 0.28, P < 0.001, n = 156), and negatively correlated with success rate (t = 0.16, P < 0.05, n = 104). DISCUSSION Effects of ecological conditions on suckling patterns. During this study, the population was at peak density, and the mouflons experienced very low food availability. Standing crops of live grass on swards in summer were 33% lower than the preceding year, during which time the population increased from about 85 to 110 ind./km 2. Poor environmental conditions resulted in the mortality of more than 25% of the neonates during the 1991 lambing season, compared to less than 10% the preceding year. Moreover, a massive die-off occurred the following winter, during which more than 90% of lambs entering the winter perished (unpublished data). Resource availability, or quality, seems a major factor responsible for the nursing differences between populations, although other factors (genetics, climate, vegetation types, habitat, etc.) cannot be excluded. Maternal care, assessed by suckling frequency and overall suckling time was lower in the Kerguelen population compared to the European ones. However, though it is often considered that total suckling time reflects maternal expenditure (BERGER 1979, 1986; CLUTTON-BROCK et al. 1982; GREEN 1986, 1990; LEE & MOSS 1986; HOGG et al. 1992; OBREGÓN et al.

9 Mouflon nursing behaviour at high population density a), several studies failed to relate suckling duration to milk transfer or offspring growth rate (FLETCHER 1971, LOUDON et al. 1983, ROBERTSON et al. 1992). It is thus also useful to consider behavioural indices, such as the frequency of suckling attempts by the young and the mother s rejection rate, in order to estimate parentoffspring conflict in mammals (LEE 1987, REDONDO et al. 1992). Among wild sheep, GEIST (1971) compared nursing behaviour and growth rate between a bighorn sheep (Ovis canadensis) and a Stone s sheep (Ovis dalli) populations. Bighorn lambs had lower suckling frequency and mean suckle duration, and shorter overall suckling time than Stone s lambs. They devoted more time than Stone s lambs to grazing, attempted to suckle more frequently, were rejected more often, and in contrast to Stone s lambs, never ended suckling bouts. Finally, bighorn lambs had a lower growth rate than Stone s lambs. Our data revealed the same behavioural trends between the Kerguelen and the two European populations, and may indicate a reduced milk transfer in the Kerguelen population. The high number of suckling attempts by the lambs and the importance of the mother s rejection rate suggest an intense mother-lamb conflict in regard to suckling. Decrease in rate of milk transfer is assumed to indicate the gradual weaning process, very often accompanied by some changes in mother and young behaviour (MARTIN 1984). Suckling frequency is usually responsible for the decrease in suckling time during weaning (GAUTHIER & BARRETTE 1985, BABBITT & PACKARD 1990, HOGG et al. 1992). ARNOLD et al. (1979) found that weaning occurs under a certain milk yield threshold in domestic sheep, and occurs sooner if ewes are maintained under a low nutritional plane. Commonly, ewes let their young suckle as long as they want to during the first 2 weeks of lactation (EWBANK 1967, BON 1985, OBREGÓN et al. 1992b). The high frequency of unsuccessful suckles and termination of suckles by the ewes in Kerguelen reveals the earlier occurrence of the period of weaning conflict compared with European populations. Young ungulates often increase their foraging activity during the weaning process (FESTA-BIANCHET 1988, BABBIT & PACKARD 1990, FAICHNEY 1992), or when milk production is reduced (GEIST 1971, ROBBINS & MOEN 1975, SADLEIR 1980, DUNCAN et al. 1984). Grazing begins when lambs are 2 weeks old among mouflon and bighorn, and coincides with an increase in suckle rejection rate by the mother (BON 1985, SHACKLETON & HAYWOOD 1985, OBREGÓN et al. 1992a). The Kerguelen lambs grazed earlier in life, and devoted more time to grazing when 15 days old. These results support the hypothesis of a reduced milk production by ewes in the Kerguelen population. Even if an efficient rumen is developed by 8 weeks of age (LYFORD 1988), grazing activity may be related to the nutrition level of lambs and that the earlier the mother rejects its lamb s suckling attempts the earlier it begins to graze. Effects of ewes age on suckling patterns. Total suckling duration decreased with age of ewes in the Kerguelen population. Moreover, old females were less tolerant of their offspring, their rejection rate being particularly high. Lambs of old ewes spent more time grazing than lambs of younger ewes, across all lamb age classes. Furthermore, the more a lamb tried to suckle, and the less successful it was in its attempts, the more it grazed. These results suggest a reduced milk production and a more intense mother-offspring conflict for old than for younger ewes, and do not support the hypothesis of increased maternal investment by old females. The origin of this phenomenon could be the rapid teeth-wearing of females, due to feeding on overgrazed pastures, such that older females may not be able to nurse their young very efficiently. In fact, tooth wear decreases foraging

10 332 D. Réale and P. Boussès efficiency and could have important consequences for milk production (CLUTTON- BROCK et al. 1982, SKOGLAND 1988). Moreover, due to their inexperience, primiparous females might not be able to control their lambs demands and hence overinvest in their young (TRIVERS 1974, HARPER 1981). Finally, a differential mother-offspring conflict according to the sex of the young has been observed in sheep (OBREGÓN et al. 1992a, REDONDO et al. 1992). Thus, as we ignored the sex of the lambs during the study period, our results could simply reflect variation in the lamb sex ratio between ewe age classes. However, in March, when the development of horns allowed the determination of lamb sex, we did not observe variation in the sex ratio of offspring with different female age. It is thus unlikely that the sample composition explains the variations observed in nursing parameters between different female age classes. The above hypotheses are not exclusive, and differential nursing behaviour between females seems to result from multiple factors. The rapid changes in population density and ecological factors should have important consequences on the nursing parameters of the Kerguelen mouflon population. ACKNOWLEDGEMENTS This study could not have been carried out without the help of Thierry Sonnier in collecting data. We also are indebted to Richard Bon who generously offered us data on the Caroux-Espinouse population. We thank Richard Bon, Peter Brotherton, Patrick Duncan, Marco Festa-Bianchet, Jean-Michel Gaillard, Valerius Geist and three anonymous referees for helpful comments on the manuscript. The research was supported by the Institut Français pour la Recherche et la Technologie Polaires, the C.N.R.S. (U.R.A. 1853), and the Office National de la Chasse. Denis Réale received grants from the Ministère de l Enseignement Supérieur et de la Recherche. REFERENCES ALTMANN J Observational study of behaviour: sampling methods. Behaviour 49: ARNOLD G.W., WALLACE S.R. & MALLER R.A Some factors involved in natural weaning processes in sheep. Applied Animal Ethology 5: BABBITT J. & PACKARD J.M Suckling behavior of the collared peccary (Tayassu tajacu). Ethology 86: BERGER J Weaning conflict in desert and mountain bighorn sheep (Ovis canandensis): an ecological interpretation. Zeitschrift für Tierpsychologie 50: BERGER J Wild horses of the great basin; social competition and population size. Chicago: University of Chicago Press. BON R Approche dynamique de l éco-éthologie d une population de mouflons de Corse (Ovis musimon) dans la Reserve National de Chasse du Caroux-Espinouse. Mémoire de Diplôme d Étude Approfondie, Université Paul Sabatier, Toulouse. BON R. & CAMPAN R Social tendency of the Corsican mouflon Ovis ammon musimon in the Caroux-Espinouse massif. Behavioural Proccesses 19: BOURZAT D. & MONIÉ J.P Evolution des pâturages aux îles Kerguelen. Mémoire Ingénieur des Techniques Agricoles, Institut National de Promotion Supérieure Agricole, Dijon. BOUSSÈS P., BARBANSON B. & CHAPUIS J.L The Corsican mouflon (Ovis ammon musimon) on Kerguelen archipelago: structure and dynamics of the population, pp In: Spitz F. et al., Edits. Ongulés/Ungulates 91. Toulouse: Société Française d Étude et de

11 Mouflon nursing behaviour at high population density 333 Protection des Mammifères, Institut de la Recherche sur les Grands Mammifères. BOUSSÈS P. & RÉALE D Valeur du masque facial comme critère d âge des femelles de mouflons de Corse (Ovis musimon). Mammalia 58: BOUSSÈS P., RÉALE D. & CHAPUIS J.L Mortalité hivernale massive dans la population de mouflons de Corse (Ovis musimon) de l archipel subantarctique de Kerguelen. Mammalia 58: CARLISLE T Brood success in variable environments: implication for parental care allocation. Animal Behaviour 30: CHAPUIS J.L., BOUSSÈS P. & BARNAUD G Alien mammals, impact and management in the French subantarctic islands. Biological Conservation 67: CLUTTON-BROCK T.H The evolution of parental care. Princeton, New Jersey: Princeton University Press. CLUTTON-BROCK T.H., ALBON S.D. & GUINNESS F.E Fitness costs of gestation and lactation in wild mammals. Nature 337: CLUTTON-BROCK T.H., GUINNESS F.E. & ALBON S.D Red deer; behavior and ecology of two sexes. Chicago: University of Chicago Press. DUNCAN P., HARVEY P.H. & WELLS S.M On lactation and associated behaviour in a natural herd of horses. Animal Behaviour 32: EWBANK R Nursing and suckling behaviour amongst clun forest ewes and lambs. Animal Behaviour 15: FAICHNEY G.H Consumption of solid feed by lambs during their transition from preruminant to full ruminant function. Applied Animal Behaviour Sciences 34: FESTA-BIANCHET M Nursing behaviour of bighorn sheep: correlates of ewe age, parasitism, lamb age, birthdate and sex. Animal Behaviour 36: FLETCHER I.C Relationship between frequency of suckling, lamb growth and post-partum oestrus behaviour in ewes. Animal Behaviour 19: GAUTHIER D. & BARRETTE C Suckling and weaning in captive white-tailed and fallow deer. Behaviour 94: GEIST V Mountain sheep; a study in behavior and evolution. Chicago: University of Chicago Press. GITTLEMAN J.L. & THOMPSON S.D Energy allocation in mammalian reproduction. American Zoologist 28: GREEN W.C.H Age-related differences in nursing behavior among American bison cows (Bison bison). Journal of Mammalogy 67: GREEN W.C.H Reproductive effort and associated costs in bison (Bison bison): do older mothers try harder? Behavioral Ecology 1: HARPER V. L Offspring effects upon parents, pp In: Gubernick D.J. & Klopfer P.H., Edits. Parental care in mammals. New York: Plenum. HOGG J.T., HASS C.C. & JENNI D.A Sex-biased expenditure in rocky mountain bighorn sheep. Behavioral Ecology and Sociobiology 31: LEE P.C Nutrition, fertility and maternal investment in primates. Journal of Zoology, London 213: LEE P.C. & MOSS C.J Early maternal investment in male and female African elephant calves. Behavioral Ecology and Sociobiology 18: LOUDON A.S.I., MCNEILLY A.S. & MILNER J.A Nutrition and lactational control of fertility in red deer. Nature 302: LYFORD S.J. JR Growth and development of the ruminant digestive system, pp In: Curch D.C., Edits. The ruminant animal. Digestive physiology and nutrition. Englewood Cliffs, NJ: Prentice Hall. MARTIN P The meaning of weaning. Animal Behaviour 32: MUNRO J A study of milk yield of three strains of Scottish blackface ewes in two environments. Animal Production 4: OBREGÓN F., ARIAS DE REYNA L. & RECUERDA P. 1992a. Maternal expenditure during lactation in mouflon (Ovis orientalis musimon). Etología 2: OBREGÓN F., ARIAS DE REYNA L. & RECUERDA P. 1992b. Nursing and suckling behaviour in the

12 334 D. Réale and P. Boussès mouflon. Ethology Ecology & Evolution 4: PFEFFER P Le mouflon de Corse (Ovis ammon musimon Schreber, 1782); position systématique, écologie et éthologie comparées. Mammalia 31: REDONDO T., GOMENDIO M. & MEDINA R Sex-biased parent-offspring conflict. Behaviour 123: ROBBINS C.T. & MOENS A.N Milk consumption and weight gain of white-tailed deer. Journal of Wildlife Management 39: ROBERTSON A., HIRAIWA-HASEGAWA M., ALBON S.D. & CLUTTON-BROCK T.H Early growth and suckling behaviour of Soay sheep in a fluctuating population. Journal of Zoology, London 227: SADLEIR R.F.M.S Energy and protein intake in relation to growth of suckling black-tailed deer fawns. Canadian Journal of Zoology 58: SCHACKLETON D.M. & HAYWOOD J Early mother-young interactions in California bighorn sheep, Ovis canadensis californiana. Canadian Journal of Zoology 63: SKOGLAND T Tooth wearing by food limitation and its life history consequences in wild reindeer. Oikos 51: STEARNS S.C Life-history tactics: a review of the ideas. Quarterly Review of Biology 51: TRIVERS R.L Parental investment and sexual selection, pp In: Campbell B., Edit. Sexual selection and the descent of man; Chicago: Aldine. TRIVERS R.L Parent-offspring conflict. American Zoologist 14: WILLIAMS G.C Natural selection, the cost of reproduction, and a refinement of Lack s principle. American Naturalist 100: