The peculiarities of the banded mongoose social system provide an opportunity to investigate questions that are fundamental to understanding the evolution of relationships between individuals in a broader social and ecological context. I start by investigating the communication between dependent pups and their carers (Chapters 3 and 4), asking three questions: (a) what information does pup begging provide about the internal state of pups; (b) how does begging influence the behaviour of carers; and (c) do pups change they way they beg based on differences in the way carers respond? I show that the response of individual cares to begging is influenced by the costs and benefits of investing in pups, and I show that pups adjust their begging in relation to the responsiveness of carers. Having determined how individual variation in pups and carers influence the way they interact with each other, I then investigate how the broader social environment influences the relationship (Chapter 5). I show that pups derive direct benefits from begging by their companions. Finally, I investigate how the social structure of banded mongooses influences the intensity of intrasexual competition for reproductive opportunity (Chapter 6), and I ask whether this influences the allocation of resources to male and female pups. I show that reproductive conflict is more intense among females, and that juvenile condition has a stronger effect on female reproductive success. Ultimately, large scale details of the social structure and life history of banded mongooses feed back to influence the fine scale interactions between pups and carers.

The first half of this thesis concentrates on investigating the hormonal mechanisms of paternal care in the common marmoset (<i>Callithrix jacchus).</i> <i>Chapter 3</i> examines the way in which a number of hormones correlate with the expression of paternal care. In <i>Chapters 4 and 5, </i>I discuss the affect of transiently manipulating the circulating level/effectiveness of prolactin and endogenous opioid peptides while fathers were caring for infants, to evaluate whether either of these hormones is involved in stimulating the behaviour. In the second half of the thesis, I look in detail at paternal food sharing behaviour and the energetic cost of infant care. <i>Chapter 6</i> examines how food novelty, food accessibility and infant age determine the willingness with which fathers and mothers share food with infants. <i>Chapter 7</i> concentrates on the benefit infants may get from sharing food, and looks at whether sharing food or watching parents eat are important in the development of long term infant feeding preferences. Both infant carrying and food sharing are energetically costly to both parents. <i>Chapter 8</i> uses weekly body weight measurements to assess whether body size and infant carrying affects the body weight of fathers and mothers during the first 3 months after the infants are born. Finally, in <i>Chapter 9</i> I discuss the results of my experiments within the context of current hypotheses and published data.

This dissertation explores the relationship between metabolic rate, mortality and the foraging behaviour of bees. A basic underlying assumption of some optimal foraging models is that harder working individuals die earlier. A trade-off between foraging activity and longevity may have implications for colony growth, and ultimately for colony reproductive success. Optimality models make quantitative predictions about behaviour from analyses of costs and benefits. When the 'currency' in such models is energy, costs are usually evaluated with metabolic rate data from laboratory studies. Laboratory data often represent pooled results for steady-state activity under standardised conditions, and as such, may bear little resemble to the metabolic costs incurred by individuals in the field. The research presented here uses the non-invasive doubly labelled water method to provide the first direct measurements of field metabolic rate in honeybees and the first direct measurements for bumblebees foraging at natural sources. The results suggest that ecological factors such as ambient temperature or flower-type affect metabolic rate to such an extent that models based on laboratory costs have little or no predictive value. An activity <I>vs.</I> longevity trade-off is tantamount to the 'rate of living' theory of senescence, a theory often invoked to explain mortality in foraging bees. It is a purely mechanistic theory, however, and as such is often opposed on evolutionary grounds: the rate of ageing should be tuned to the pressure of extrinsic mortality (predation, disease and accidents), and senescent death should rarely occur in natural populations.

Four studies were conducted to investigate behavioural and physiological indicators of welfare in western lowland gorillas (<i>Gorilla gorilla gorilla</i>) housed at six European zoos. In the first two studies, behavioural responses of gorillas to changes in housing were measured, namely relocations to new enclosures (Study 1) and feeding-related environmental enrichment efforts (Study 2). As enrichment efforts do not necessarily equate with successful enrichment, I hypothesised that these housing changes would have either an enriching, aversive or neutral effect on animals. The results of Study 1 have shown that the behavioural repertoire of most individuals did not change across the entire post-relocation period compared with the baseline. Greater behavioural responses were observed month-by-month and, for some individuals, conclusions can be drawn about changes in welfare. In Study 2, behavioural data were collected before, during and after the provision of feeding-related environmental enrichment efforts designed to provide a more complex feeding environment. For most individuals, these efforts were associated with a reduction in time spent performing behaviours that indicate poorer welfare, compared with during the baseline. Faecal samples collected opportunistically during Studies 1 and 2 were used in Study 3 for glucocorticoid metabolite analyses. The results of a validation procedure showed that an enzyme immunoassay (EIA) measuring 5β-3α,11β-diol corticoid metabolites was likely to have practical applications for this purpose . Gorillas’ adrenal responses to the housing changes suggested that there has been an enriching effect overall. Results of a storage experiment showed that faecal samples can be left at environmental temperature (25 °C) for up to 8 hours after defecation with no significant effect of bacterial enzymes on metabolite concentrations.

There are differences in the walking behaviour of solitarious and gregarious locusts. Gregarious locusts walk 2.5 times faster; steps are longer and higher; the excursion of the femoro-tibial angle of the hind leg is larger; the body is held up and at an angle off the floor and the antennae are held up. Solitarious locusts hold the thorax and abdomen close to and parallel with the ground and hold the antennae down. There are also differences in the firing rate of the slow extensor tibiae motor neuron (SETi) in response to simulated flexion and extension of the leg, achieved by the controlled movement of the apodeme of the femoral chordotonal organ, a proprioceptor that monitors the movement of the leg around the femoro-tibial joint. At all simulated leg angles, the firing rate of SETi is higher in solitarious locusts than in gregarious. Muscle forces of the extensor tibiae (ETi) muscle of the hind leg were measured in response to extracellular stimulation of SETi and the fast extensor tibiae motor neuron (FETi) of the muscle. When SETi is stimulated at 1 Hz, the ETi of gregarious locusts produces twitches that are 3 times the amplitude, and that contract and relax faster than in solitarious. From 7 Hz to 50 Hz there no difference in the amplitude of contradictions produced. Solitarious locusts maintain greater catch tension in the presence of the neuromodulator octopamine, which otherwise potentiates twitch tension and increases the rate of contraction and relaxation of muscle contractions in both phases. When FETi is stimulated at 50 Hz for 10 seconds, there is a decline in tetanus of 32% over the stimulation period in solitarious locusts but 66% in gregarious. Spikes in FETi elicit excitatory post synaptic potentials (EPSPs) in F1Ti motor neurons (Hoyle and Burrows, 1973). The relationship between FETi and FlTi motor neurons was examined using double intracellular recording. At stimulation of 0.5-2 Hz, antidromically induced FETi spikes in solitarious locusts cause a greater degree of depression of flexor EPSPs than in gregarious. At 5-20 Hz there is no difference in EPSP depression, for example 20 Hz depression reaches 20% original amplitude in both phases after a stimulus train of 10 pulses. At all frequencies EPSP amplitude recovers less well in solitarious locusts. When a single stimulus pulse is given 10 seconds after the stimulus train at 20 Hz, EPSP amplitude in gregarious locusts recovers to 76% original height, but only to 63% in solitarious. Firing patterns recorded in FETi and flexors during kicking in solitarious show some similarities and some differences to those previously obtained in gregarious locusts.

No description available.

This thesis describes studies of the development of the circadian system in the Siberian hamster. It concentrates on the ontogeny of the photic entrainment system, the functionality of maternal-postnatal entrainment, and the role of dopaminergic mechanisms underlying circadian entrainment. The induction of the immediate early gene <I>c-fos </I>was used as a phase-dependant marker of cellular activation to study the ontogeny of a photic response in the suprachiasmatic nucleus (SCN). Induction of <I>c-fos </I>is correlated with light-induced phase shifts of the biological clock in the adult. These studies established that <I>c-fos </I>was first inducible by a light pulse on postnatal day (PD) 3, and that the number of cells expressing light-induced <I>c-fos </I>increased from PD3 to PD5. The innervations of the SCN by the retinohypothalamic tract (RHT) was examined using tract-tracing agents, and the growth of fibres into the SCN was correlated with the development of the light induced <I>c-fos</I> response. Previous studies indicate that glutamate is the principal neurotransmitter of the RHT and mediates light entrainment of the SCN, and is also associated with neuronal plasticity and synapse formation in other brain regions. The expression of NMDAR1 splice variants was studied at different developmental stages. It was discovered that the SCN expressed a conserved isoform throughout the postnatal period. Experiments were undertaken to establish the existence and relevance of maternal entrainment during the postnatal period. Mothers maintained in constant dark from early pregnancy produced litters with synchronised circadian rhythms of locomotor activity at weaning. This suggests that despite an absence of light-dark cues, the developing animals are entrained by maternally-derived cues. In a subsequent experiment, new-born litters fostered to mothers exposed to a reversed light-dark cycle established a maternal influence on the circadian phase of the pups.

In this thesis, I use a large scale study on the dwarf angelfish <i>Centropyge bicolor </i>to investigate social conflict resolution in a linear hierarchy. Body-size based linear dominance hierarchies in social fishes have great potential to provide insight into within-group conflict resolution, with recent work demonstrating that group stability may be achieved by subordinate regulating their growth rates to avoid punishment by eviction. How subordinates can assess the imminence of punishment remains a crucial outstanding question: despite suggestions that dominant aggression rates might provide the crucial mechanism, the expected aggressive patterns have yet to be empirically demonstrated. I begin by laying the groundwork for social conflict studies in <i>Centropyge bicolor </i>by quantifying multiple behavioural correlates of social rank. I first show that spatial segregation has confounding effects on the behavioural patterns between dominants and subordinates, with important consequences for dyadic relationships and overall group stability. By focusing on spatially overlapping groups, I then show that dominant aggression rates increase gradually as subordinates get relatively larger; subordinates that are too close in size to dominants are evicted from the group. Subordinates respond to the aggressive signal of imminent eviction by reducing their foraging and growth rates, resulting in a clearly-defined size hierarchy and continued group stability. The necessary maintenance of a well-separated size hierarchy places limits on maximum group size. Well-regulated subordinate size results in a neutral effect of subordinates on dominant fitness. My results present a uniquely complete view of the causes and consequences of conflict resolution in linear dominance.

I investigate the role of specific social interactions in the transmission of <i>Mycobacterium bovis</i> (bovine tuberculosis; bTB) in a free-living population of wild meerkats (<i>Suricata suricatta)</i> in the Kalahari Desert of South Africa. During a prospective cohort study lasting two and a half years, I use detailed pathologic investigations, repeated live sampling of individuals and social network analysis (SNA) of observed inter- and intergroup social interactions, to elucidate the epidemiology of bTB in a habituated population of 300 meerkats living in 14 social groups. I show that bTB in meerkats is a disseminated disease which spreads to multiple organs via haematogenous and lymphatic routes. The results suggest that meerkats acquire <i>M. bovis</i> infection principally via the respiratory and oral routes, and excrete it mainly from the respiratory tract and suppurating skin wounds. I find that intergroup transmission appears to occur via roving males that are infected whilst visiting other groups and subsequently return to their original groups, and not via aggressive intergroup encounters involving entire groups. I develop an individual-based dynamic network model to elucidate the relative importance of grooming and aggression in the transmission of <i>M. bovis</i> between meerkats. I found grooming to be more important than aggression in transmitting <i>M. bovis</i> between meerkats within a social group, with groomers at higher risk of infection then groomees. The short lifespan of most meerkats relative to the latent period of <i>M. bovis</i> in this species mean that few individuals reach the infectious stage. This is likely to be a limiting factor in the spread of bTB in populations of free-ranging wild meerkats.

There are a number of processes which can result in social transmission of behaviour patterns; so much effort in social learning research has gone into devising experimental procedures that can isolate imitation from other social learning processes. In this thesis I develop method for distinguishing “simple” social learning processes and test these methods using the domestic fowl as a model organism. Strong evidence is presented for a response facilitation effect on a number of behaviour patterns, a process which might function to the same ends as imitation in animal populations. Recent models have suggested that an ability to imitate might be dependent on prior experience rather than specialised learning mechanisms. A neural network model is used to investigate these hypotheses, and generate predictions as to the conditions under which a capacity for imitation should arise. The model predicts that processes such as behavioural synchrony between individuals might result in the formation of appropriate neural links for an imitative ability. These predictions are tested using experimental and observational data on the domestic fowl. The model also suggests that similar mechanisms and developmental processes might underlie imitation and simpler social learning processes, such as response facilitation and observational conditioning. This suggests that a process of positive feedback might operate, with social learning promoting behavioural synchrony, which in turn promotes the development of social learning mechanisms.