Behavioral Ecology Vol. 14 No. 1: 10-15
© 2003 International Society for Behavioral Ecology
Sex-specific differences in reindeer calf behavior and predation vulnerability
aDepartment of Biology, Faculty of Science, University of Tromsø, N-9037 Tromsø, Norway bGreenland Institute of Natural Resources, PO Box 570, DK 3900 Nuuk, Greenland cNorwegian Institute for Nature Research, Tungasletta 2, N-7005 Trondheim, Norway
Address correspondence to A. Landa, Greenland Institute of Natural Resources, PO Box 570, DK-3900 Nuuk, Greenland. arild{at}natur.gl.
Received 29 August 2000; revised 10 February 2002; accepted 9 April 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
According to reproductive strategy theory, males in polygamous breeding systems should invest in morphological or behavioral features that increase reproductive success. When the early development of such traits conflicts with predator protection, we expect that male calves will exhibit risk-taking behavior, such as high activity level and increasing distance from mother, to a greater extent than female calves. We investigated sex differences in mother–calf distance, calf activity levels, and calf mortality in a semi-domesticated reindeer (Rangifer tarandus) population. The results show that male calves stray farther away from their mothers, exhibit a higher level of locomotive behavior in terms of play and walking, and are more vulnerable to predation than are female calves. Although mother–calf distance increased over time in 1- to 6-month-old calves, no evidence was apparent for an increase in sex difference in mother–calf distance over this period. The results suggest a trade-off between predation vulnerability and investments in behavioral traits thought to be important for future reproductive success and suggest that these sex-related differences in behavior are apparent as early as 6 months of age.
Key words: Rangifer tarandus, reindeer, calf behavior, sex differences, sexual segregation, predator vulnerability.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Reproductive strategy theory assumes that morphological and behavioral features that increase reproductive success are likely to spread in a population (Clutton-Brock et al., 1982
). In polygamous ungulates the development of large body size, fighting abilities, and consequent high social status is important in male mating access (Clutton-Brock et al., 1982). In contrast, adult females should invest in care and security of their offspring (Geist, 1971; Jakimchuk et al., 1987; Main and Coblentz, 1990). Such sex-specific traits are likely to result in a number of behavioral differences between sexes.
As a consequence of long-distance migratory habits and the general lack of concealing vegetation in their typical habitats, it is crucial for the newly born reindeer calf (Rangifer tarandus) to follow the mother and maternal group as soon as possible. In addition to the benefits of maternal defense, important features reducing predation in group-living animals include increased predator-detection probability and reduced individual vulnerability as group size increases (Dehn, 1990; Pulliam and Caraco, 1984). Reindeer calves moving far from their mother and those exhibiting high levels of locomotive activities are acting contrary to the predator protection inherent in maternal defense and grouping behavior, and these calves are likely to suffer relatively high predation rates. Because it is important for males but not for females to develop fighting skills, we hypothesized that male calves will exhibit risk-taking behavior to a greater extent than female calves. Furthermore, this difference between sexes is likely to increase with age, which eventually might lead to segregation between males and females (Bon and Campan, 1996). These sex-related behavioral differences should affect the spatial relationship and degree of synchronized behavior between reindeer cows and their newborn calves, as well as the individual calves' activity level.
Apart from the distinct rutting behavior, one of the most pronounced features seen in many ungulates is the difference in habitat utilization before the rut season. This spatial sexual segregation has several proposed explanations (Main and Coblentz, 1990; Ruckstuhl and Neuhaus, 2000). The most promising are (1) difference in energy requirements due to body-size dimorphism (Barboza and Bowyer, 2000; Pérez-Barbería and Gordon, 1999), (2) social factors, such as sex-specific intolerance, dominance relationships, and fighting skills (Bon and Campan, 1996), (3) reproductive strategy-related differences in energy requirements and offspring protection (Jakimchuk et al., 1987; Main et al., 1996), and (4) asynchrony in activity budget due to traits inherent in the above hypothesis, such as differences in body size, energy requirements, digestive efficiencies, and predator avoidance (Conradt, 1998; Ruckstuhl, 1998). These explanations, however, are not mutually exclusive and suggest different traits which likely have evolved to increase individual reproductive success.
Spatial habitat segregation is one of two processes that result in sexual segregation (Bon and Campan, 1996; Conradt, 1999), the other being gender differentiation in social behavior and social segregation (Bon and Campan, 1989; Festa-Bianchet, 1991; Geist, 1971). Social segregation has a clear ontogenetic character in the development of social partner choice and is likely to appear before the characteristic habitat segregation takes place (Bon and Campan, 1996). An age-specific choice of social partners in juveniles has been reported in reindeer and muskox (Ovibos moschatus) (Espmark, 1971; Lent, 1974). Male calves in elephant (Loxodonta africana) and domestic cattle (Bos taurus) have been found to play more often with calves of same sex and to be less closely tied to their mothers than female calves (Lee, 1986; Lidfors and Jensen, 1988; Veissier et al., 1990). If such behavior, which is thought to be related to the development of male fighting skills (Clutton-Brock et al., 1982), is also associated with substantial costs in terms of increased predator vulnerability, a trade-off between developing fighting skills and predator protection will be apparent.
This study focused on two behavioral characteristics: mother–calf distance and level of locomotive activity reflected in the activity pattern. We expected to find that male calves stay less close to their mothers and have a higher activity level than female calves and that sex difference in spatial mother–calf relationship increases over time. Factors related to antipredator behavior, such as within-group position, group size, and habitat characteristics were also assessed, and we investigated whether male calf predation rate was actually higher than for females.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We observed a semi-domesticated reindeer herd in the Hjerttind reindeer pasture district, Troms County, northern Norway (69°00' N, 18°00' E). The study site is a part of an approximately 1000-km2 area at 370–1380 m elevation, used as summer grazing pasture for herd of about 700–800 adults and their young. A typical group consisted of 32 individuals (geometric mean), but group sizes of up to 300 were observed. The area encompasses zones of subarctic mountain birch (Betula pubescens) forest in the lowlands, heather and grassland higher up, and mountainous alpine terrain at the highest altitudes.
In spring 1997, we marked 20 adult female reindeer with radio collars (142 MHz, SirTrack, New Zealand). We used these radio-marked individuals to locate the herd or groups of reindeer. In August 1997, 180 male and 197 female calves were marked with numbered, expanding collars, and in June 1998, 126 male and 126 female newborn calves were marked with numbered, expanding mortality collars, (142 MHz, SirTrack, New Zealand). During mid November 1998–May 1999, we marked 30 male and 123 female calves with expanding mortality collars. The capturing and marking was carried out in close cooperation with and under the expert supervision of the reindeer owners. We used two circular enclosures of approximately 90 m and 15 m width where the calves were either captured by hand, by lasso, or by using a pole with a snare at the end.
Field observations
We observed behavior during the period 17 August–13 October 1997 and 9 June–5 July 1998 using binoculars (8 x 40) and a 15–45x spotting scope. We identified mother–calf pairs by scanning the herd for individuals showing typical mother–calf associations, such as suckling and close following behavior. As a consequence, only initially active animals were selected for the instantaneous sampling methodology used (Martin and Bateson, 1993). A full sampling sequence on a focal pair consisted of 10.5 min of observation, with behavioral data recorded at 30-s intervals, thus generating 21 plots for each sample.
Distance between mother and calf was estimated using length of an adult female (ca. 2 m) as a reference. We recorded the activity categories grazing, walking, lying, observing, running, suckling/nursing, playing, ruminating, and other activity not specified for both calf and mother.
If high activity levels and great distance from mother is confined to central group-positions or to males being at more peripheral positions more often than female calves, this might confound the importance of distance to mother in determining predator vulnerability. Therefore, we recorded both group size and the calf's initial position in the group, the latter on an ordinal scale: center (1), near center (2), near edge (3), edge (4), or alone (5). A pair was categorized as lone if the distance to a group of more than one animal exceeded 100 m. Calf color (white, light gray, gray, or brown) were recorded to distinguish whether individuals with less cryptic pelage (white and brown) were less precocious (i.e., stayed closer to their mother than other calves).
Because habitat, weather, presence of observers, or diurnal, seasonal, or yearly variation may also influence individual behavior (Elgar, 1989), the variables outlined below were sampled to permit statistical control of any influence. Weather conditions recorded included precipitation type (rain, scattered showers, drizzle, fog, snow and none), cloud cover (%), wind speed (calm conditions, and gentle, moderate, and fresh breeze) and temperature (°C). Vegetation characteristics recorded included habitat-related categories (woodland, swamp, heather, and ridge), and forest density and shrub density were recorded on an ordinal scale from 1 (no coverage) to 5 (ca. 100% coverage). We also recorded terrain inclination angle, observer distance, time of day, date and year.
A total of 275 mother-calf sampling sequences were initiated, 198 in 1997 and 77 in 1998. We sampled 15 pairs more than once; these repeated samplings were pooled by averaging the measurements for each pair, thus reducing the samples by 24 to a total of 251 independent observations (112 male calves, 125 female calves, and 14 of unknown sex).
Recording of calf mortality
We obtained data on total calf mortality by calculating the difference between the number of calves released and the number surviving until the traditional roundup during November 1997 and 1998. The calves equipped with mortality collars (1998–1999) were tracked daily during early June–mid-November and two to three times per week during December–April using radio receivers LA12Q (AVM Instrument Company Ltd., USA), RX-8910 and RX-81 (Televilt AB, Sweden). On foot, or in vehicles and snowmobiles, we used a three-element collapsible yagi-antenna (SirTrack, New Zealand). When using aerial tracking (Cessna 172), we used two-strut assembled H-antennas (Televilt AB, Sweden). All dead individuals found in the field were postmortem analyzed following Landa (1999).
Mortality was analyzed as simple proportions of marked calves of the two sexes that were missing at the next census roundup; missing calves, even if not found, were assumed to have died in the intervening period. Mortalities of the two sexes in three periods, summer 1997, summer 1998, and winter 1998–1999, were compared. Assuming that sex-specific behavioral or other traits affecting mortality might dispose one sex to be more vulnerable to mortality than the other by a constant factor, mortality models were simplified where appropriate by fitting rates that had a constant ratio for the two sexes. Such models fitted well. Models were fitted by maximum likelihood, likelihood ratios were used to test statistical significance, and standard errors of fitted parameters were estimated from likelihood support intervals.
Test of hypotheses
The dependent variable mother–calf distance (proximity) was log10 transformed to achieve homogenous variance in the gender groups and approximation to a normal distribution. To obtain the most appropriate statistical model explaining proximity and to test for the expected differences between calf gender, while controlling for the potential confounding components, we performed an ANCOVA. The initial ANCOVA model included proximity as the dependent variable, gender as an independent group factor, and the 10 potential confounding components as covariates. Then, the covariates with least significant linear relationship to proximity were successively excluded from the model in a backward stepwise procedure. The nominally categorized variables (color of calf, year, habitat, and precipitation type) were tested separately (ANOVA or ANCOVA) for relationships with proximity. The position in group was investigated in a two-way ANOVA to detect sex-specific differences in mother–calf distance for different positions. The presence of an increased sex difference in spatial mother–calf relationships over time was tested for by comparing the regression coefficient for the two sexes.
For each individual sampling sequence, the proportion of an activity that occurred during a sample sequence provided scores (relative frequency) for each type of activity (Martin and Bateson, 1993). Occurrences of activities, in sampling sequences containing fewer than 21 plots, were still divided by 21. Thus, a score value for an activity in an incomplete sampling was not overestimated compared to a full sampling sequence of 21 plots. Frequency tables were calculated to obtain total percentage distribution of activities for both calf gender and their respective mothers.
The percentage synchronization of activity between mother and calf was assessed by first assigning a value to locomotive activities (walk, run, and play) and another value to nonlocomotive activities (graze, lie, observe, suckle/nurse, ruminate, and other). The values for the mother and calf were compared for each plot in the sampling sequence; equal values returned the synchronization factor (Sf) = 1 and unequal values returned Sf = 0. The mean Sf for each mother–calf pair (the synchronization index) represents the percentage time when mother and calf were synchronized.
Calculating mean mother-calf distance
Because identifications of mother–calf pairs relied on their close and intimate association, a biasing effect of this methodology was expected to be evident in the measure of mother–calf distance. This effect was reflected in an overall positive relationship between the dependent variable (log10 mother–calf distance) and the independent variable plot number in the sampling sequence, when all plot numbers (plot1-plot21) were included in the model (F1,4977 = 50.472, r2 =.010, p <.001). By successively excluding initial plots, a nonsignificant relationship occurred once the first five plots were excluded (F1,3600 = 3.141, r2 <.001, p =.077). Thus, the threshold beyond which sampling time ceased to provide significant contribution to explain the mother–calf distance was at 3.0 min of observation. Therefore the first five plots (2.5 min) in the sampling sequences were excluded from the analyses. As a consequence, 10 samples with less than 3 min of observation were also excluded from the analysis. We calculated mother–calf proximity by averaging the mother–calf distance for the remaining 8.0 min, thus providing independent data for each pair.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Mother–calf proximity
Using ANCOVA with gender as factor, proximity as independent variable, and controlling for the effects of terrain inclination, forest density, and group-size, male calves were found to be significantly farther from their mothers (geometric values: mean ± SE = 4.55 ± 1.08 m, adjusted mean = 4.56 m, n = 106) than were female calves (geometric values: mean ± SE = 3.30 ± 1.06 m, adjusted mean = 3.29 m, n = 121; F1,222 = 12.538, p <.001; Figure 1). The covariates terrain inclination, forest density, and herd size each showed a significant linear relationship with mother–calf proximity (Table 1) and together explained a significant amount of variation in proximity (F3,222 = 8.983, r2 =.108, p <.001). The common regression coefficient for these relationships did not differ between gender (F3,219 = 1.627, p =.184); hence male calves were farther from their mothers for any given value of these three covariates. When the calf was lying down, the difference in proximity to mothers of male calf (geometric values: mean ± SE = 7.08 ± 1.32 m, 95% CI = 3.91–12.71 m, n = 17) and that to mothers of female calf (geometric values: mean ± SE = 3.78 ± 1.23 m, 95% CI = 2.44–5.85 m, n = 21) showed no significant difference (t = 1.843, df = 36, p =.074).
|
|
In a two-way ANOVA with gender and position in group as factors and proximity as the dependent variable, there was no effect of either position (F4,208 = 1.301, p =.268) or the interaction of gender and position, (F4,208 = 0.593, p =.668; Figure 2). When the other categorical variables were tested for relationships with the dependent variable proximity, no relationships were apparent for either weather (F3,235 = 0.986, p =.400), calf color (F3,234 = 1.110, p =.345), habitat type (F3,233 = 1.981, p =.118), or year when controlled for effect of development over time (F1,236 = 3.241, p =.073).
|
Differences in activity pattern
Male calves displayed a higher frequency of walking and playing activity than female calves, with no significant gender differences apparent for other activities (Table 2). No significant differences in activity pattern were found between cows with male and cows with female calves, but for the two predominant activities, a nearly significant difference was apparent (grazing: mother of male, median = 0.71, 95% CI = 0.57–0.86; mother of female, median = 0.81, 95% CI = 0.67–0.91; walking: mother of male, median = 0.10, 95% CI = 0.05–0.14; mother of female, median = 0.05, 95% CI = 0.00–0.10; Table 3).
|
|
Paired-sample comparisons showed that, for the two predominant activities, the male calf grazed less but walked more frequently than its mother, whereas the female calves showed no difference from their mothers in these two activities (Table 4). Still there was no significant difference in the mother–calf pair synchronization index between pairs with male calves (median = 0.95, 95% CI: 0.95–1.00, n = 108) and pairs with female calves (median = 1.00, 95% CI: 0.95–1.00, n = 125) (Mann-Whitney U test, Z = 1.066, p =.286)
|
Development of mother–calf spatial relationship over time
There was a positive linear relationship between mother–calf distance and date of observation (F1,225 = 5.954, r2 =.026, p =.015). But in an ANCOVA model with gender as a factor, proximity as dependent variable, and date of observation as a covariant, the interaction by covariates (gender x date) showed no significant difference in the regression coefficients for the proximity–date relationships for the two genders (F1,223 = 0.076, p =.783). Thus, contrary to expectation, the analysis did not show a significant sex difference in the development of mother–calf distance over time.
Calf mortality and gender vulnerability
A total of 32 calves were found dead based on the mortality collars, 16 during summer (May–October) and 16 during winter (November–April). Predation accounted for 75% (n = 24) of the losses. Only one death (3%) could be attributed to other causes (drowning), while 22% (n = 7) were found too late to ascertain the cause of death. Thus, predation accounted for 24/25 cases in which the cause of death could be identified. Too much was consumed of eight carcasses to determine predator species. For the remaining 16 carcasses, lynx (Lynx lynx) was the main predator (55%, n = 13), followed by wolverine (Gulo gulo; 8%, n = 2) and golden eagle (Aquila chrysaetos; 4%, n = 1).
All comparisons concurred that males had higher mortality, and overall this result was highly significant: male mortality was 84% (SE 37.3%), higher than female (likelihood ratio test, p <.01; Table 5). The hypothesis that male mortality was a constant factor higher than female mortality could not be rejected. During the winter 1998–1999, so few mortality collars were fitted that this difference was not significant for that experiment, but also mortality ratio at that season did not differ from other seasons. There was a big difference between summer mortality in 1997 and summer 1998 (
2 = 6.31, df = 1, p =.012, n = 629) and the mortality in summer 1997 was 2.54 (SE = 0.684) times higher (likelihood ratio test, p <.01) than in the following summer. Because mortality was so different in the two summers, we could not draw a general conclusion at to whether winter or summer mortality is larger.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Sex differences in calf behavior
Reindeer mother–calf distance was greater for male than for female calves, and male calves exhibited higher levels of locomotor activity than female calves (more walking and playing). These results indicate that the spatial relationship between the adult female and its neonate is affected by the calf's gender and that behavioral gender divergence begins early, well before the habitat sexual segregation typical of polygamous ungulates (Bon and Campan, 1996
; Geist, 1971).
Development of spatial relationships over time
There was an overall increase in mother–calf distance throughout the summer season, though no gender difference in the rate of increase was apparent. Several factors may have limited the ability to detect such a difference. The male-biased predation observed, as well as the considerable amount of unexplained variation in the data, could have affected whether time-dependent increase in sex difference was seen. This variation could be reduced by controlling for factors not possible in this study, such as individual calf's precise developmental stage. Aditionally, dominance rank and age of mother have been reported to influence mother–calf associations, in red deer (Cervus elaphus), for example (Clutton-Brock et al., 1982).
Mother–calf distance in reindeer as well as in bison (Bison bison) has been found to increase rapidly during the first 3 weeks of life and thereafter to remain relatively constant (Espmark, 1971; Green, 1992a). Thus, the most dramatic acceleration in the calves' maternal independence may have already occurred by the time field observations in this study began.
Group size and habitat characteristics
The negative relationship between group size and mother–calf distance contrasts with findings in bison, where mother and calf were more proximate in small groups than in larger groups (Green, 1992b). Green (1992b) linked this to the general pattern of increased predation risk in small herds due to reduced efficiency in antipredator strategies. However, group size as an antipredation strategy is likely to be modified by vegetation cover and topographic relief gradient (Helle, 1979; Skogland, 1981). The reindeer in our study area exploit many habitat types, and it is possible that underestimation of group size in habitats where visibility is low (e.g., forest cover) could affect our results.
Calf behavior was apparently not affected by position within the group, and males did not show a preponderant presence in lone mother–calf pairs. These observations strengthen the significance of distance to mother and activity level as determinants of sex-biased predation vulnerability.
Sex bias in neonatal predation
Although juvenile predation mortality in ungulates shows somewhat equivocal results regarding sex bias in calf mortality (Linnell et al., 1995), seasonal differences and other factors need to be better assessed. In studies of ungulates where bias in neonatal predation is apparent, it is usually male biased, and sex difference in calf behavior is generally suggested as a potential explanation (Aanes and Andersen, 1996; Bergerud, 1971; Jackson et al., 1972). The sex-specific divergence in calf behavior and predation vulnerability found in the present study are suggestive of a trade-off between behavioral traits related to reproductive success and predation vulnerability early in life. We hypothesize that the strong selection for highly competitive males in polygamous mating systems, such as in reindeer, favors males who begin to develop fighting skills early in life, despite the exposure to predation that this behavior brings.
|
ACKNOWLEDGEMENTS |
|---|
We express gratitude for assistance provided by the field personnel, Ingmund Halgunset and Knut Morten Vangen. Many thanks are also extended to the reindeer owners at Andsvatnet in Troms County and the Norwegian Forest Crown, Troms. In addition to the Norwegian Institute for Nature Research's "Wolverine Project," the Department of Biology, Faculty of Science, University of Tromsø (UiTø), The Roald Amundsen Center for Arctic Research (UiTø), and the Norwegian Directorate for Nature Management, Trondheim, provided financial support to the fieldwork. We also thank Michael Kingsley for valuable statistical advice.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Aanes R, Andersen R, 1996. The effects of sex, time of birth, and habitat on the vulnerability of roe deer fawns to red fox predation. Can J Zool 74:1857-1865.
Barboza PS, Bowyer RT, 2000. Sexual segregation in dimorphic deer: a new gastrocentric hypothesis. J Mammal 81:473-489.
Bergerud AT, 1971. The population dynamics of Newfoundland caribou. Wildl Monogr 25:5-55.
Bon R, Campan R, 1989. Social tendencies of the Corsican muflon (Ovis ammon musimon) in the Caroux-Espinouse (South of France). Behav Process 19:57-78.
Bon R, Campan R, 1996. Unexplained sexual segregation in polygamous ungulates: a defense of an ontogenetic approach. Behav Process 38:131-154.[CrossRef]
Clutton-Brock TH, Guinness FE, Albon SD, 1982. Red deer: behavior and ecology of two sexes. Chicago: University of Chicago Press.
Conradt L, 1998. Could asynchrony in activity between the sexes cause intersexual social segregation in ruminants? Proc R Soc Lond B 265:1359-1363.[Medline]
Conradt L, 1999. Social segregation is not a consequence of habitat segregation in red deer and feral soay sheep. Anim Behav 57:1151-1157.[Medline]
Dehn MM, 1990. Vigilance for predators: detection and dilution effect. Behav Ecol Sociobiol 26:337-342.
Elgar MA, 1989. Predator vigilance and group size in mammals and birds: a critical review of the empirical evidence. Biol Rev 64:13-33.[Medline]
Espmark Y, 1971. Mother-young relationship and ontogeny of behaviour in reindeer (Rangifer tarandus L.). Z Tierpsycol 29:42-81.
Festa-Bianchet M, 1991. The social system of bighorn sheep: grouping pattern, kinship and female dominance rank. Anim Behav 42:71-82.[CrossRef]
Geist V, 1971. Mountain sheep: a study in behavior and evolution. Chicago: University of Chicago Press.
Green WCH, 1992a. The development of independence in bison: pre-weaning spatial relations between mother and calves. Anim Behav 43:759-773.
Green WCH, 1992b. Social influences on contact maintenance interactions of bison mothers and calves: group-size and nearest-neighbour distance. Anim Behav 43:775-785.
Helle T, 1979. Observations of group size and composition of wild forest reindeer (Rangifer tarandus fennicus Lönn.) during the calving and summer periods in eastern Finland. Aquilo Ser Zool 19:5-11.
Jackson RM, White M, Knowlton FF, 1972. Activity patterns of young white-tailed deer fawns in south Texas. Ecology 53:262-270.
Jakimchuk RD, Ferguson SH, Sopuck LG, 1987. Differential habitat use and sexual segregation in the Central Arctic caribou herd. Can J Zool 65:534-541.
Landa A, 1999. Spor og tegn, et hefte til bestemmelse av store rovdyr. Trondheim: NINA*NIKU, Norwegian Institute for Nature Research.
Lee CP, 1986. Early social development among African elephant (Loxodonta africana) calves. Natl Geogr Res 2:388-401.
Lent PC, 1974. Mother-infant relationships in ungulates. In: The behaviour of ungulates and its relation to management: International symposium at The University of Calgary, Alberta, Canada 2-5 November 1971 (Geist V, Walter F, eds). Morges, Switzerland: International Union for Conservation of Nature and Natural Resources 14-55.
Lidfors L, Jensen P, 1988. Behaviour of free-ranging beef cows and calves. Appl Anim Behav Sci 20:237-248.[CrossRef]
Linnell JDC, Aanes R, Andersen R, 1995. Who killed Bambi? The role of predation in the neonatal mortality of temperate ungulates. Wildl Biol 1:209-223.
Main MB, Coblentz BE, 1990. Sexual segregation among ungulates: a critique. Wildl Soc Bull 18:204-210.
Main MB, Weckerly FW, Bleich VC, 1996. Sexual segregation in ungulates: new directions for research. J Mammal 77:449-461.
Martin P, Bateson P, 1993. Measuring behaviour, 2nd ed. Cambridge: Cambridge University Press.
Pérez-Barbería FJ, Gordon IJ, 1999. Body size dimorphism and sexual segregation in polygynous ungulates: an experimental test with Soay sheep. Oecologia 120:258-267.[CrossRef]
Pulliam HR, Caraco T, 1984. Living in groups: is there an optimal group size? In: Behavioural ecology: an evolutionary approach, 2nd ed (Krebs JR, Davies NB, eds). Oxford: Blackwell 122-147.
Ruckstuhl KE, 1998. Foraging behaviour and sexual segregation in bighorn sheep. Anim Behav 56:99-106.[CrossRef][Web of Science][Medline]
Ruckstuhl KE, Neuhaus P, 2000. Sexual segregation in ungulates: a new approach. Behaviour 137:361-337.[CrossRef]
Skogland T, 1981. Comparative aspects of social organization. In: Tundra ecosystems: a comparative analysis (Bliss LC, Heal OW, Moore JJ, eds). Cambridge: Cambridge University Press 452-483.
Veissier I, LeNeindre P, Garel JP, 1990. Decrease in cow-calf attachment after weaning. Behav Process 21:95-106.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||