Abstract: Recent attention has been drawn to the existence of individual differences in correlated behaviour across contexts, animal `personality' (Gosling 2001, Psychological Bulletin, 127, 45-86) and behavioural syndromes (Sih et al. 2004b, Quarterly Review of Biology, 79, 241-277). The causes of these patterns of behaviour are subjects of debate. Here, we present a very simple model of how adaptively managing noisy information, combined with differences in individual state, can lead to evolutionarily stable differences in how individuals respond to environmental cues. When information is very noisy, behavioural syndromes are most likely, but as long as there is some error, some types of individuals display the same behaviour in all contexts. In extreme cases, very few individuals display flexible behaviour, and different stable behavioural types dominate the population.

Abstract: Studies of ape tool use have been conducted in captivity since the early 1900s and in the wild since the 1960s. Chimpanzees are the most prolific tool users among the apes, and are known to use more tools than any other nonhuman animal. In contrast, reports of gorilla tool use are rare both in wild and captive settings. Studies of the processes involved in tool use learning have been limited in the wild by the lack of ability to control several unpredictable variables, and in captivity by tool use opportunities that are often presented in non-naturalistic contexts. We attempted to address both of these limitations by providing naïve subjects with a naturalistic tool use device (built to simulate a termite mound) while housed in a more natural social setting to approximate how learning would occur in the wild. Both gorillas and chimpanzees participated in the experiment to allow comparative analyses of acquisition of tool behaviour and the factors that may affect acquisition. Both species showed low frequencies of interaction with the mound in the baseline condition, before baiting with a food reward. Once baited, chimpanzees both attempted and succeeded to extract the reward more quickly than did gorillas. The number of social group members at the mound was significantly higher for chimpanzees than for gorillas and may have affected skill acquisition. We advocate that comparative approaches to skill acquisition and learning are valuable, but that researchers need to be cognizant of species differences in social structure that may affect results.

Abstract: Mother-infant interactions and patterns of foal behaviour in the Grevy's zebra, Equus grevyi, differe from those reported for other equids. Grevy's zebra foals exhibit longer intervals between suckling bouts, do not drink water until they are 3 months old, and reach independence from the mare sooner than other equids. Furthermore, Grevy's zebra foals advance their acquisition of adult feeding behaviour. A 6-week-old Grevy's zebra foal spends as much time feeding as a 5-month-old wild horse foal. From the time their foals are born until the foals reach an age of 3 months, females form small groups (three females and their foals). These groups are never found further than 2·0 km from surface water and are usually associated with a territorial male. Unlike other equids, the foals of which always follow their mares, when female Grevy's zebra go to drink, they leave their foals in “kindergartens”, which are guarded by a single adult animal, usually a territorial male. It is proposed that many of these differences in behaviour and rates of juvenile development are the result of adaptation to an arid environment. Water requirements during early lactation appear to influence strongly the social behaviour of the Grevy's zebra and should also be a strong influence on the mother-infant behaviour of other arid-living ungulates.

Abstract: Foragers can often show a broad range of strategies when searching for resources. The simplest foraging strategy is to search randomly within a habitat; however, foragers can often assess habitat quality over various spatial scales and use this information to keep themselves in, or direct themselves to, regions of high resource abundance or low predation risk. We investigated models that describe a population of consumers competing for a renewable resource that is distributed among discrete patches. Our aim was to identify what foraging strategy or strategies are expected to persist within a population, where strategies differ in the degree of habitat assessment (i.e. none, local, or global). We were interested in how the optimal strategies are dependent on the cost of assessment and habitat structure (i.e. the variation in renewal rates and predation risks among patches). The models showed that the simple random foraging strategy (i.e. make no habitat assessments) often persisted even when the cost of habitat assessment was low. Persistence could occur when habitat assessment and population dynamics generated an ideal free distribution because it could be exploited by the random foragers. Habitat assessment was more advantageous when consumers could not achieve ideal free distributions, which was more likely as patches became less productive. When productivity was low we sometimes observed the situation where different foraging strategies generated resource heterogeneities that promoted their coexistence, and this could occur even when all patches were intrinsically identical.

Abstract: I develop guidelines for assessing the precision and power of statistical techniques that are frequently used to study nonhuman social systems using observed dyadic associations. Association indexes estimate the proportion of time that two individuals are associated. Binomial approximation and nonparametric bootstrap methods produce similar estimates of the precision of association indexes. For a mid-range (0.4-0.9) association index to have a standard error of less than 0.1 requires about 15 observations of the pair associated, and for it to be less than 0.05, this rises to 50 observations. The coefficient of variation among dyads of the proportion of time that pairs of individuals are actually associated describes social differentiation (S), and this may be estimated from association data using maximum likelihood. With a poorly differentiated population (S~0.2), a data set needs about five observed associations per dyad to achieve a correlation between true and estimated association indexes of r=~0.4. It requires about 10 times as much data to achieve a representation with r=~0.8. Permutation tests usually reject the null hypothesis that individuals have no preferred associates when S2H>5, where H is the mean number of observed associations per individual. Thus most situations require substantial numbers of observations of associations to give useful portrayals of social systems, and sparse association data inform only when social differentiation is high.