Climate change is altering ecology in myriad ways; the distributions, dynamics, phenology, morphology and life-histories of species are changing. Establishing the relative importance of these changes in different ecological traits is important to fully understand how climate and environmental change threatens populations. However, until recently, studies have often focused solely on range-shift responses and have rarely examined multiple forms of ecological change in a single species. Case-studies focusing on multiple aspects of ecology, for example demography, life-history and behaviour, could provide important general insights into how climate change will most threaten the persistence of species and populations. In this thesis, I investigate ecological variation, and the influence of climate and environmental change, in a common species of considerable cultural, economic and ecological importance, the Alpine chamois (Rupicapra rupicapra). Using a 38-year dataset on body masses and demography, augmented with demographic and behavioural data collected during this thesis, I examine variation in the life-history, demography and behaviour of this species, making both species-specific and general inferences. I find variability in reproductive strategies in the study populations, providing considerable insight into poorly understood area of this species’ ecology. Male chamois in neighbouring populations adopt contrasting reproductive strategies, likely due to climatic differences among areas. Interestingly though, males time investment in reproduction similarly in the different populations. However, there is a strong pattern of age-dependent investment; young males delay reproduction until the end of the breeding season, apparently as a strategy to avoid high levels of competition. This is the first clear evidence for such a strategy in a male mammal. Next, focussing on the effect of climatic and environmental factors on chamois ecology, I found a strong influence of climate on body size, demography and foraging behaviour. Climate change appears to be driving declines in body size and climatic variation exerts a strong influence on demographic variation and foraging behaviour. However, the effects of local anthropogenic drivers in some cases exceed these effects. High hunting pressure has a profound effect on survival, with hunting mortality largely replacing natural mortality. Furthermore, the influence of disturbance by livestock grazing on altitudinal migration is very pronounced, forcing chamois to much higher altitudes. This project illustrates that climate change can affect diverse aspects of ecology. Despite the overriding research focus on range-shifts, climate-driven life-history and demographic change could influence populations more strongly. Furthermore, my results show that effects of other drivers, such as harvesting and biotic interactions, can also have profound effects on ecology, at times exceeding the influence of climate.

The environmental factors influencing the population densities of red (Cervus elaphus) and roe deer (Capreolus capreolus) living in Scottish plantation forests were investigated. The work followed from intensive studies of the ecology of red and roe deer within a single forest in Argyll, Scotland. Densities were estimated in twenty forests, based on standing crop dung group counts conducted on c. 100 transects (1.5 x 100 m) per forest, stratified to take account of different habitat types. The figures were converted to deer densities using decay rates estimated from experimental plots and from published values of deer defaecation rates. Estimated densities of both species were inversely related (p 0.05), and ranged from 0.5 to C.35 deer km-2. The relationships of densities with habitat composition, climate, vegetation composition, soil type and culling pressure were investigated. Analyses included pairwise correlation, principal components analysis and stepwise multiple regression. Red deer density was positively correlated with temperature, and negatively correlated with 'chilling conditions' (combined wind-speed and low temperatures) and culling pressure (p 0.05). Pregnancy rates of both species were positively correlated with each other, and with sunshine, windspeed and forb abundance (p 0.05). Red deer pregnancy rates were negatively correlated with snow-depth and population density (p 0.05); roe deer pregnancy rates were negatively correlated with rainfall {p 0.05). It is suggested that deer density variation reflects both the relative climatic adaptations of the two species, and the differential influence of climate, diet quality and competition upon two ruminants with different digestive specialisations.

I studied the population dynamiec and social behaviour of a marked population of roe deer at Chedington, Dorset, form 1966-68, with additional observations and experiments in 1969. The numbers of territorial bucks and adult does, which formed the permanent breeding population, were stable and most were resident for several years. Numbers of non-tor-ritorial bucks showed greater fluctuations. Yearlings for-med nearly a third of the population each spring despite a reduced reproduction rate by does at Chedington when compared with does in other areas. Mortality in the wood was low and many non-territorial bucks and yearling does emigrated in spring when a change in social behaviour occurred. In winter, groups of two to eight deer voro common but breke up in spring when territorial bucks became more aggressive, After March, observations of single adult does and bucks inoreasod. Territories were actively defended from April - August; most aggresstion was directed at non-territorial bucks. Fights between adjacent territorial bucks were rare, and bucks occupied the same worritorios for several years with little overlap between territories. Non-territorial bucks did not defend an area and their home ranges were large, overlapped each other oxtensively, and often passed through territories of several bucks. Non-territorial bucks which failed to secure territories were rarely seen the following year. Adult does occupied the same home ranges for several years and these sometimes coincided with the territories of bucks, but more frequently overlapped the territories of two or three bucks. The androgenic activity of bucks which reached a peal: in July, coincided with the period of territorial behaviour, However, aggression of bucks could not b increased experimentally by testosterone implants, but a buck lost hie territory about eight weeks after castration. Experiments showed that territorial bucks defended territories and not does and their behavior prevented non-territorial bucks from holding territories and limited the number of territorial bucks in the population, but had no effect on numbers of does, Adult does remained in their home ranges despite the absences of territorial bucks and then removed were replaced by yearling does. Thus yearling does seemed to form the surplus doe population and were prevented from establishing themselves by the behaviour of adult does, but the mechanism involved was not clear.

Vertebrate species are managed for many reasons, including their role as economically important predators or as carriers of disease. Successful management depends on the ability to predict the outcome of management actions on a species’ population dynamics. However, uncertainty in the models used to make such predictions can arise from multiple sources, including sampling error in vital rates, intraspecific demographic variation and unknown interspecific interactions. The red fox Vulpes vulpes provides a useful model organism for exploring such uncertainty, because management of this important predator and disease host is often ineffective, despite substantial sampling effort. By explicitly accounting for sampling error in survival and fecundity, confidence intervals for population growth rates were derived from published point estimates of red fox demographic data. Uncertainty in population growth rates was found to be high, requiring a quadrupling of sampling effort to halve the confidence intervals. Given the often poor justification for the choice of distribution used to model litter size, the influence of probability distributions on population model outcomes was tested. In this first comprehensive evaluation, estimates of quasi-extinction and disease control probabilities for three Canid species were found to be robust to litter size distribution choice. Demographic analyses of the red fox revealed a medium to fast life history speed and significant survival and fecundity contributions from juveniles to population growth. Intraspecific variation was detected within these spectra of demographic metrics: the first such demonstration for carnivores. Simulated data substitution between fox populations revealed that geographic proximity and similar levels of anthropogenic disturbance did not infer demographic similarity. Considering the sampling effort expended on the red fox, the species appears well-studied; yet, substantial limitations in data collection were identified. Compartment modelling of a sarcoptic mange outbreak in an urban fox population in Bristol, UK, revealed that disease transmission was frequency-dependent, consistent with contact rates being determined by social interactions rather than by population density. Individual-based modelling suggested that indirect transmission, genetic resistance and long-distance recolonisation were required to replicate the observed rapid spread of mange and subsequent population recovery. Thus, this first attempt to model mange dynamics in this canid provided novel insight into previously uncertain epidemiological and behavioural processes in the transmission of sarcoptic mange in the red fox.

This study attempted to determine if competition was an important influence on the behaviour, distribution and abundance of lion and hyaenas.  Complementary contrasting approaches were adopted, an intensive study in two areas, and a comparison among protected areas.  The behaviour of lions and hyaenas was compared in two contrasting ecosystems.  Matusadona National Park.  MNP (Zimbabwe) and Liuwa Plains National Park.  LPNP (Zambia).  MNP is a closed woodland habitat dominated by resident population of buffalo and impala.  LPNP is an open floodplain system dominated by migratory populations of wildebeest and zebra.  MNP had a greater ratio of lions to hyaenas (2: 1) whereas in LPNP, hyaenas considerably outnumbered lions (18:1). The study also compared data from 13 Protected areas of Africa that differed in terms of habitat, available pre, size and human influence. In MNP there was no evidence of competitive avoidance behaviour by either species despite a considerable overlap in diet.  Both species prey predominantly on buffalo and impala.  There was evidence that each species was extending its niche to utilise prey species that would be uneconomic for the other species to use facilitating co-existence.  The home ranges of the two species overlapped considerably and the location and size of their home ranges were similar. However, the study was not able to measure temporal distribution and the two species may have been avoiding each other by using the same areas of the park at different times.  Further analysis of the responses of the two species to an increase in the population of buffalo, the main prey species of the park indicated that lions might have indirectly affected the hyaenas in the park, as the hyaena population decreased with an increase in the lion population.  In the same park, more evidence of an adverse competitive effect was observed when the response behaviour of the two predators to a simulated prey resource (pre-recorded squealing pigs combined with a smelly bait) was compared.  Lions responded as expected in areas of high prey density, whereas hyaenas responded more in areas with low prey densities and low visibility of prey suggesting that they were avoiding lions.  In contrast, hyaenas in LPNP responded to the simulated resource in areas of high prey densities.  The response behaviour of lions did not differ between the two Protected areas despite the fact that lions were outnumbered in LPNP and numerically dominant in MNP.  This supports the hypothesis that a larger body size may enable lions to offset the effects of competition.

Deer have been introduced outside their native ranges numerous times worldwide, causing significant economic and environmental impacts. The scale of problems caused by some introduced deer species is expected to increase, while others are relatively harmless and may be important for conservation as their numbers decline in native regions. This thesis examines the factors promoting invasion in non-native deer, and investigates the population genetics of two deer species introduced to Great Britain. Factors hypothesised to predict invasion success were tested using a comparative analysis of the outcomes of introductions of non-native deer populations. Two modelling approaches were taken to account for confounding effects in species-level and population-level analyses, and different factors were found to explain success at different stages of the invasion process. Populations of roe deer expanding from refugia and reintroductions were examined, and differences in genetic diversity between core and peripheral populations, alongside a decline in diversity with rate of expansion from the core were found. High levels of differentiation between and among core and peripheral populations suggest that genetic drift is the major factor causing these patterns. Using a genome scan of British and European roe deer, loci potentially under selection were identified. Association between some of these loci and environmental variables suggests that climatic extremes may have a role in exerting selective pressures on roe populations. The Chinese water deer is severely declining in range and number in its native habitat, but rapidly expanding after being introduced to Great Britain. Relatively high levels of diversity in native Chinese populations, and significant differentiation between the Chinese and British populations were found. The source population of the British deer is likely to be extinct, and the level of genetic structuring indicates that conserving populations across both ranges is important to maintaining their diversity.

Global biodiversity is under threat from multiple fronts. The Millennium Ecosystem Assessment predicts that climate change (CC) will be the “dominant direct driver of biodiversity loss” by the end of the 21st century. This project studies the impact of CC on the distribution and conservation status of antelopes found in Africa. Africa is an area of high climate variability and high vulnerability to CC, and most of the world’s antelope species are native to the continent. Various threats are already causing the decline in 63% of antelope species with 26% being classified as threatened. Antelopes are a speciose and biologically diverse group and therefore provide an ideal opportunity to reveal more general patterns of the effect of CC across taxa. To assess the impact of climate change on Africa’s antelopes I use species distribution models (SDMs), based on climatic variables, to produce ensemble predictions of species distributions for 2080. Using the SDMs I also establish links between biological traits and the optimal climatic conditions for species. The ensemble predictions incorporate three climate models for three climate scenarios, and I predict the future distributions using three approaches. The first is a pessimistic representation of species’ distributions in a future where they are unable to disperse from their current range to track CC. The second, optimistic approach, permits species to disperse at a given rate based on body mass. Finally, the envelope approach presents a comparison of suitable climatic conditions, which are connected to the existing distribution, between now and the end of the century (i.e. not restricted by current distribution or dispersal). The results indicate that 81-85% of species (59-62 of 73) will exhibit a contraction in range based on suitable climatic conditions, and that the average contraction of those species is 39.4-50.1%. Up to six of 73 species are predicted to be without any climatically suitable areas in 2080 depending on the modelling and forecasting approach taken. Worryingly, there is also a disproportionate reduction in the predicted distribution of threatened antelope, whereas species with broader climatic niche and a preference for warmer temperatures typically perform better. Using Marxan conservation planning software, I produce protected area network solutions that protect Africa’s antelopes based on their predicted distributions in 2080. High and low protection options are presented which require 8.7 and 8.1% respectively of Africa’s land mass in order to protect Africa’s antelopes. These solutions, that are based around the existing IUCN protected area network (categories I-VI), reveal key new areas are required in Somalia, while expansion of existing protected areas (PAs) are required continent-wide, particularly in Liberia, Cameroon, Kenya, Tanzania, and the Central African Republic. I also find that the contribution of community led PAs, and other non-IUCN PAs, to these solutions is currently limited, but in some areas they play an important role in bridging gaps between existing IUCN PAs. Finally, 14 species are either identified to qualify as threatened due to climate change based on IUCN Red List criteria, or can be considered threatened due to having no protected area coverage within their predicted range in 2080. Based on predicted species distributions under the A1B climate scenario, the hirola (Beatragus hunteri) and Aders’ duiker (Cephalophus adersi) have no suitable climatic conditions within, or connected to, their existing range in 2080. However, for all species, climatically suitable conditions are present in Africa in the future, and protected areas are found within those regions. For the species most at risk I provide recommendations for translocation options taking into account the potential for interspecific competition by assessing the number of novel antelope species and interspecific niche overlap in the climatically suitable areas.

This study develops a fractographic method to diagnose hammerstone- and carnivore induced fracture. This is important because interpretations of hominin entry into the carnivore guild and evolution of meat-eating are based on rare tool and tooth marks in Oldowan (2.5-1.8mya) fossil assemblages. Consequently, estimating hominin and carnivore involvement is difficult, and questions remain about Oldowan hominin’s position in the carnivore guild and socioecology. One aspect of bone damage, fracture surfaces, is ubiquitous, but largely unstudied. The fractographic (study of fracture surfaces) method is based on fracture principals, particularly how differences in static- and impact-loading affect material response and fracture features resulting from loading extremes. The method is applied to analysis of fracture features in a) the Amboseli Hyaena Den assemblage, b) an experimental hammerstone-broken assemblage, c) a Plio-Pleistocene assemblage previously interpreted as a carnivore accumulation, FLK-NN2 (Olduvai Gorge), and d) the zooarchaeological assemblage from FLK-Zinj, (Olduvai Gorge). This is the first zooarchaeological/taphonomic study to demonstrate that a) static and impact fracture differ fundamentally in applied load size and material responses to loading extremes, b) impact-forces are significantly greater than the maximum carnivore bite-force, c) cones, incipient flakes, radiating cracks, and lateral stress features are characteristic of impact fracture, and e) Oldowan hominins at FLK-Zinj were responsible for breakage of 54% of the limb assemblage (a 37% – 40% increase over estimates based on percussion marks). The socioecological implications of the habitual transport of food from death and/or kill sites to secondary locations are explored by examining reasons why social carnivores transport food. Aspects of modern carnivore behaviour suggest general mammalian constraints that may have predicated food transport by early Homo. Early Homo food transport behaviour was structured by anti-predator defense strategies associated with a) foraging in an open habitat rich with competing predators, b) the lack of masticatory and digestive apparatus to quickly consume animal tissue, and c) the presence of altricial young in the hominin group.

Studies of the evolution of the major histocompatibility complex (MHC) have been central to the understanding sexual selection and pathogen-mediated selection. The European badger Meles meles is well suited for exploring such questions because of its life history characteristics, reproductive biology and mating system. In this thesis, I examined both MHC class I and class II genes. Seven putatively functional sequences were found for class I genes and four for class II DRB genes. Evidence of past balancing selection of both genes was demonstrated by the d_Nd_S ratio, by positive selection at the antigen-binding site (ABS) and by trans-species polymorphism of alleles within other mustelids and carnivores. MHC class I genes also showed evidence of concerted evolution, but domains showed different evolutionary histories. MHC genes may influence microbiota and odour of an individual and influence mating preferences. I examined the bacterial community of the subcaudal gland secretion and demonstrated a high number of bacterial species (56 operational taxonomic units), which cubs exhibited a higher diversity than adults. The microbiota may lead to an individual-specific odour as a cue signaling the MHC genotype of potential mating partners. I report the first evidence for a MHC- based mating preference in carnivores. Female badgers showed a MHC-assortative mate choice towards breeding with males that had functionally similar MHC genes, for MHC class II DRB genes. This applied to neighbouring-group matings. I also found considerable annual fluctuation in the occurrence of MHC-based mate choice. Based on genome-wide background in the same mating randomizations I found no evidence of inbreeding, which indicated that MHC similarity was apparently the actual target of mate choice. In line with MHC-assortative mate choice, MHC heterozygosity had no influence on the co-infection status. Individual MHC alleles did, however, associate with resistance and susceptibility to specific pathogens, suggesting that MHC diversity may be driven and maintained by pathogen-mediated selection through rare-allele advantages and/or fluctuating selection. My study of genetic characteristics, mate choice and pathogen pressures in a wild population revealed past and contemporary evolutionary process of the MHC genes. This increases knowledge of how the MHC may affect mating behaviour and sexual selection, ultimately influencing population processes.