Abstract: Linear hierarchies, the classical pecking-order structures, are formed readily in both nature and the laboratory in a great range of species including humans. However, the probability of getting linear structures by chance alone is quite low. In this paper we investigate the two hypotheses that are proposed most often to explain linear hierarchies: they are predetermined by differences in the attributes of animals, or they are produced by the dynamics of social interaction, i.e., they are self-organizing. We evaluate these hypotheses using cichlid fish as model animals, and although differences in attributes play a significant part, we find that social interaction is necessary for high proportions of groups with linear hierarchies. Our results suggest that dominance hierarchy formation is a much richer and more complex phenomenon than previously thought, and we explore the implications of these results for evolutionary biology, the social sciences, and the use of animal models in understanding human social organization.

Abstract: Introduction
Glucocorticoids are present in many biological fluids as a free fraction or bound to Corticoid
Binding Globulins (CBG) (Matteri et al, 2000). There are conflicting claims regarding the validity of
saliva as a biological fluid to measure cortisol in horses (Lebelt et al, 1996; McGreevy and Pell, 1998;
van der Kolk et al, 2001). Measuring changes in salivary cortisol levels in normal horses and horses
with Cushing`s disease van der Kolk and collaborators (2001) demonstrated the validity of saliva to
assess adrenal function. Puzzling results were reported by McGreevy and Pell (1998) who suggested
that plasma and salivary cortisol concentrations in horses showing oral stereotypies were correlated
but this association was non-existent in control animals. Investigating the responses of foals to
branding and foot-trimming Zanella et al (unpublished results) were unable to identify a relationship
between plasma and salivary cortisol levels in foals. In several species, levels of cortisol in plasma and
saliva are tightly correlated (Fenske, 1996). Cortisol found in blood consists of a fraction bound to
corticoid binding globulin (CBG) and a free, unbound fraction. Free cortisol represents the
biologically active fraction of this steroid hormone. Salivary cortisol reflects the unbound fraction
found in plasma or serum and it passes readily through the parotid membrane (Riad-Fahmy, 1983;
Horning Walker et al,1977). Unbound steroids transfer rapidly between plasma and saliva
(Walker,1989; Scott et al 1990). Saliva flow-rate does not appear to influence saliva cortisol levels in
different species (Hubert and de Jong-Meyer, 1989; Walker 1989, Scott et a, 1990). In horses, Lebelt
et al (1996) reported that salivary and plasma total cortisol in stallions were correlated. We
hypothesized that changes in salivary cortisol in foals would show a pattern that is correlated to that of
plasma free and plasma total cortisol concentrations in foals. In addition, we anticipated that the lack
of good sampling techniques provides an explanation for the failure in determining the association
between salivary and plasma cortisol in foals.

Abstract: Animals frequently use signals that travel further than the spacing between individuals. For every intended recipient of a given signal there are likely to be many other individuals that receive information. Eavesdropping on signalling interactions between other individuals provides a relatively cost-free method of assessing future opponents or mates. Male great tits (Parus major) extract relative information from such interactions between individuals unknown to them. Here, we show that male great tits can take information gathering a stage further and obtain more information about a previously unencountered intruder, by the hitherto unknown capability of combining information gathered by eavesdropping with that derived from their own direct interaction with an individual. Prior experience with an intruder (A) was achieved by subjecting a focal male to different levels of intrusion simulated using interactive playback. This intruder (A) then took part in a simulated interaction with an unknown male (B) outside the territorial boundary of the focal males. In response to subsequent intrusion by the second male (B), focal males showed low song output in response to males that had lost to a male that the subject was able to beat. Males of known high quality, or those about which information was ambiguous, elicited a high level of song output by focal males. We discuss the implications of this finding for the evolution of communication and social behaviour.

Abstract: Differences in behavioural characteristics between individuals of the same species are often described as being due to the temperament of the individuals. These differences can have enormous implications for welfare with some individuals apparently being able to adapt to environmental challenge more easily than others. Such differences have resulted in animals often being described as either `active' copers, which try to escape from or remove an aversive stimulus, or `passive' copers, which show no outward signs of a situation being aversive, thus, appearing to be unaffected. Tests previously developed to assess the temperament of animals have been criticised for several reasons. Behaviour is often recorded and categorised using methods that are not objective and tests are generally carried out once with no consideration of whether or not behavioural responses are consistent over time. This study takes these factors into account. The behaviour of 33 horses was recorded in three types of test--an arena test, response to a person and response to an object. In order to test whether or not responses were consistent over time, the tests were repeated three times with an average of 9 days between trials. Test results were validated using responses from questionnaires completed by the farm team leader. The data were analysed using an initial principal component analysis (PCA) and factor analysis. The horses were found to behave consistently over the three trials in their responses in the arena test. The responses to the person test and the object test were similar to each other; however, these responses were not consistent over trials. The behaviour in the arena test was unable to be used to make a prediction of behaviour in the person and object tests and vice versa. The responses shown by the horses did not enable them to be categorised as either active or passive copers. Behavioural responses in the tests were not predictive of the response to a startle test (water spray), nor could they be used to predict status or response to being reintroduced to the group after testing. There was no relationship between the responses in the tests and the ratings given by the farm team leader. It was concluded that horses vary widely in their responses to artificial behavioural tests, with only the responses to an open-field arena test being consistent over time, and therefore, the only type of test which can indicate some core factor of temperament.

Abstract: Experimental analyses of dynamic visual signals have to overcome the technical obstacle of reproducing complex motor patterns such as those found in courtship and threat displays. Video playback offers a potential solution to this problem, but it has recently been criticized because of sensory differences between humans and nonhuman animals, which suggest that video stimuli might be perceived as deficient relative to live conspecifics. Quantitative comparisons are therefore necessary to determine whether video sequences reliably evoke natural responses. Male Jacky dragons, Amphibolurus muricatus, compete for territories using complex displays delivered in a rapid stereotyped sequence. We evaluated video playback as a technique for studying this visual signal. Digital video sequences depicting a life-sized displaying male were indistinguishable from live male conspecifics in the rate and structure of aggressive displays evoked. Other measures of social behaviour suggested that video stimuli were more effective in this context. Lizards produced significantly more appeasement displays and had higher rates of substrate licking and locomotor activity in response to video playback than to confined male opponents, which failed to produce aggressive displays. Lizards tracked temporal changes in the display rate of video stimuli and were also sensitive to individual differences in morphology and behaviour between video exemplars. These results show that video stimuli are appropriate for the experimental analysis of Jacky dragon aggressive displays. We compare the potential shortcomings of video playback with those of other techniques and conclude that no approach offers a panacea, but that several have complementary characteristics. Copyright 2002 The Association for the Study of Animal Behaviour. Published by Elsevier Science Ltd. All rights reserved.