The selection of summer forage by three sympatric ungulates in the Damodar Kunda region of upper Mustang in north Nepal was studied to assess the extent of food overlap between them. To compare their diets, a micro-histological technique of faecal analysis was used, adjusted for inherent biases by comparing it with bite-count data obtained in domestic goats. Tibetan argali Ovis ammon hodgsoni, naur (blue sheep or bharal) Pseudois nayaur and domestic goat Capra hircus consumed mostly forbs, graminoids and browse, respectively. The proportions of food items in their diets were significantly different both at the plant species (P<0.02) and at the forage category level (P<0.001). Except for sharing three common plants (Agrostis sp., Stipa sp. and Potentilla fruticosa), dietary overlap at the species level was quite low. At the forage category level, naur and domestic goat overlapped more than the other ungulate pairs. Although all three species were opportunistic, mixed feeders, argali was a more selective forb specialist grazer than the other two ungulates. Owing to some spatial separation and little dietary overlap, interspecific competition for summer forage was low. If animal densities increase, however, goats are expected to compete more with naur than with argali because of their more similar diets. Owing to differences in forage selection by argali and naur throughout their large geographical ranges, reflecting adaptations to local ecological conditions, inferences regarding forage competition between domestic livestock and these two wild caprins need to be made from local, site-specific studies, rather than from general diet comparisons.

In our study three badger Meles meles populations separated by only a few km but subjected to different environmental conditions were compared. Differences are especially marked for climatic factors, the three areas being located at different altitudes, and for intensivity of soil use by people. The diet of the three populations was significantly different, with one or two dominant items in each area: mammals and cereals in the mountain, maize in the mid-mountain and in the lowland areas. In the most intensively cultivated area, maize was the most consumed item in autumn and spring, several months after harvesting. Earthworms had only a secondary importance in the diet in the mountainous area, but were negligible in the mid-mountain and lowland areas. Soil management seemed to play a preponderant role, mostly owing to soil quality and topography. Climate seemed to have a secondary effect only.

Skull variables were analysed for allometry patterns in 56 species of extant carnivores. As previously reported, many skull variables scale near isometrically with either skull length or lower jaw length. The maximal gape angle scales insignificantly (P<0.05) with skull size, but the clearance between the canines shows a significant relationship with skull size and scales near isometrically. Maximal bite forces were estimated from geometrical cross-sectional areas of dried skulls, and the bending strength of the canines was computed by modelling the canines as a cantilevered beam of solid, homogeneous material with an elliptical cross section. Previous hypotheses of large taxon differences in canine bending strengths, so that felids have stronger canines than canids, are corroborated when actual bite forces at the upper canine are ignored. Incorporation of bite force values, however, nullifies the differences in canine bending strength among felids and canids, and ursids seem to have stronger canines than felids. This is probably because of the significantly longer canines of felids compared to canids and ursids, and the generally high bite forces of felids.

Longer-term ecosystem level dynamics are often neglected in conservation studies involving single species. In this study, a retrospective analysis is presented on the feeding performance of African wild dogs Lycaon pictus in the Serengeti in relation to a competing species, the spotted hyena Crocuta crocuta, to test whether hyenas had an effect on feeding performance of wild dogs in this ecosystem. Our analysis is based on observations of over 700 wild dog kills recorded over a 20-year period (from 1964 to 1987) during which time there was a decline in wild dog numbers (ending with their local extinction in 1991) and a twofold increase in hyena density. Overall, the amount of time that dogs had access to the kill (access time) decreased with increasing numbers of hyenas attending kills, but access time increased with increasing hunting-group size of dogs and carcass mass. In addition, in the 1980s, dogs spent longer at kills than in the 1970s for a given set of conditions, including when hyenas were absent. Our analysis demonstrates a greater potential for group benefits than was found in a previous study (Carbone, Du Toit et al., 1997). Hunting-group sizes of between two and six dogs performed best when hyenas attended dog kills because the benefits of increased defence outweighed the costs of having to share the carcass with more dogs. Hunting-group sizes of wild dog and levels of hyena attendance at the kill broadly paralleled the population trends in these species, with hunting-group sizes of wild dog declining, followed by hyena attendance increasing. Despite the combined effects of increased hyena attendance and reduced hunting-group size, dogs in the 1980s typically spent longer feeding and consumed more of the carcass including the poorest sections. This suggests that dogs in the 1980s may have been under greater energetic stress.

Age-related improvements in reproductive performance in seabirds have been well documented, and may be explained by improvements in foraging efficiency or increased experience and reproductive effort with age. The interactive effects of parental age and food supply on reproductive performance, however, remain poorly understood. A widespread mass mortality of pilchards Sardinops sagax in southern Australian waters in 1998 provided a unique opportunity to investigate the effects of a sudden reduction in the availability of amajor prey species on Australasian gannets Morus serrator, an important local marine predator. Age-related differences in the breeding performance of gannets were evident in 1 year of reduced pilchard availability; when food was not limited, both young and experienced parents were equally capable of rearing chicks and had similar levels of breeding success. These data clearly demonstrate the interactive effects of parental age and food supply on breeding performance and suggest that such differences only become apparent when conditions become more stressful.

Natterer's bat Myotis nattereri is one of the least known European bats, using roosts inaccessible to humans and shifting to alternative roosts every few days. The characteristics of day and night roosts used by maternity colonies during summer was assessed. Roosts were located by radio-tracking. Roost sites (which may contain more than one roost) were concentrated mainly in core areas of up to 2 km2, within which roost site density ranged from 7 to 15 per km2. Day roosts included tree cavities, apices of attics, mortises in attics and in barns, soffit boxes at eaves, tops of gable walls, crevices in stone walls and a modern cavity wall. Trees comprised 67% of roost sites. Two maternity colonies each used between 21 and 31 roosts distributed across 15 to 25 roost sites. During late summer, the temperature of roosts adjacent to a roof covering averaged 22.5 °C during the day and were significantly warmer than roosts in trees, which averaged 17.3 °C. Attic mortise roosts were intermediate at 20.0 °C. The maximum ambient roost temperature in which bats remained was 32 °C and the maximum body surface temperature recorded while roosting was 37.6 °C. Thus, bats seem to require access to a large number of roosts with a range of temperatures that they can use depending on their reproductive state and energy requirements. The study colonies also exhibited high inter-annual fidelity to roosts. Attic maternity roosts used during late gestation and early lactation were reused for the same function in successive years, as were heavily timbered barns during late lactation and post-lactation. This continuing dependence upon established roosts indicates that these types of maternity roost should be given high conservation priority. Field boundaries should also be managed to maintain mature trees with cavities for potential day roosts and hedges suitable as night perches.

Little is known of patterns of male reproductive effort and investment, and their relationship with age-related differences in male mating strategy, in mammals. In this study, the age-specific patterns of reproductive expenditure in a male marsupial, the common brushtail possum Trichosurus vulpecula were investigated. Possums were studied for two breeding seasons on Magnetic Island, north Queensland, Australia. Energy expenditure in male possums during the mating season was estimated from changes in mass taken monthly during live-trapping. Male possums lost an average of 4.2% of their mass during the main mating season from January to May, suggesting that mating and consort behaviour is energetically expensive and may constrain foraging activity. The percentage of body mass lost by individual males during the breeding season was negatively related to age; older males directed less somatic investment into reproduction than young males. Older males seemed to target their energy expenditure toward the period when oestrus synchrony was highest. The pattern of age-specific energy expenditure for reproduction in T. vulpecula differs markedly from that reported previously for male cervids, probably owing to the spatial dispersal of females leading to differences in mating strategy. The results are discussed in terms of their relevance to potential age-related differences in reproductive strategy and behaviour.

The feeding habits of feral domestic cats Felis catus (n=264), wild cats Felis silvestris (n=22) and their hybrids (n=30) were investigated in Hungary. Cat groups were identified taxonomically by morphological and molecular methods. Diet components were identified in stomach contents and faeces collected from the recta. In each cat group, abundant small mammals were dominant in the diet (relative frequency of occurrence: feral domestic cat, 61–82%, depending on regions; wild cat, 70%; hybrid, 59%). Birds were the second most important quarry (2–7%, 16% and 20%, respectively in the three cat groups), while the contribution of hares (1–2%, 5% and 3%, respectively) and other taxa was not significant. Every cat group preyed on small-sized animals (<50 g; 89–96%, 80% and 80%, respectively), terrestrial (91–98%, 84% and 86%, respectively) and wild (71–73%, 87% and 77%, respectively) prey. Standardized trophic niche breadth was typically very narrow (BA=0.07–0.16, 0.13 and 0.17, respectively). Feral domestic cats occasionally consumed household food (2–7%) and domestic animals (4–8%). This could mean that feral domestic cats have an advantage over wild cats that are food specialists. The trophic niche overlap between cat groups was high (77–88%). Food composition and feeding habits, (weight, zonation and environmental association of consumed prey) of feral domestic cats, however, was different compared to wild cats, which indicated the possibility of partial resource partitioning. The values for hybrids were between the two groups. As well as the stable presence of feral domestic cats (mean population density, D=1.34 individuals/1000 ha) based on field live-trapping, hybrids are also present (D=0.10), leading to continuous hybridization. This can threaten the population of wild cats, which are present at a low density (D=0.17).

The soldier crab Mictyris brevidactylus breeds in winter. Large females spawned over a few days in early winter after most had decalcified their vulvar opercula. Some females spawned in their first winter, less than 1 year after settlement. The crabs probably copulate during a few days before spawning while the vulvar opercula are decalcified. During the daytime low tide in the breeding season, large males emerged and fed on the sediment surface. Females, on the other hand, made a sand roof above themselves using sand pellets from which they had sorted out food particles, and they fed while hiding under it. After the males had ceased their surface activities, they entered the sand tunnel made by the female. The roof was connected with a hollow shaft, which had been made when the crab had ascended to the surface. Before the habitat was submerged, the shaft was plugged from within with sand to form an air chamber, which suggests that the male and female descend together in the air chamber and copulate there. It is possible that the crab uses the sand roof as a visible signal to attract mates.

Wolf–ungulate interactions were studied in a mountainous region of the western Alps (Italy) from December 1999 to November 2002. Analysis of 848 scats of wolves Canis lupus showed that their most important prey was wild ungulates (87.2%). Cervids were the preferred prey (74.2%) and constituted predominant food items both in winter (84.2%) and summer (54.3%). Wolves preyed most intensively upon young ungulates. Presence of domestic ungulates on high-altitude pastures during summer (May–October) influenced wolf diet (summer 19.0%, winter 0.3%), but, despite the major density of domestic livestock, wolves still preferred wild ungulates. A strong trophic relationship between wolves and red deer Cervus elaphus emerged from the present study. During three winters, 177 ungulate carcasses were recorded. Excluding hunting, the most important cause of death in red deer was predation by wolves (51.5%), followed by disease and starvation (36.4%) and traffic accidents (12.1%). The majority of red deer killed were hinds (58.8%) and calves (29.4%); stag kills represented only 11.8%. For roe deer Capreolus capreolus, starvation and diseases were the lowest factor (11.5%), slightly higher than wolf predation (10.8%), while traffic accidents were the main cause of mortality (77.7%). Chamois Rupicapra rupicapra seemed to be less important than red deer and roe deer in the diet of wolves. The high susceptibility of red deer to wolf predation could be a result of the strong overlap of habitat and altitude use with the wolf, and to their more conspicuous herding.

Baylisascaris procyonis is a common parasitic nematode of the raccoon Procyon lotor. In intermediate or accidental hosts, including humans, B. procyonis can cause severe disease and mortality, and the parasite is increasingly viewed as an important wildlife conservation and public health concern. The prevalence of B. procyonis was assessed over a 4-year period in a population of raccoons in a forested region of lower New York. Prevalence ranged from 0% (undetected) to 21%, which was relatively low compared to values reported in other studies from the north-east. During year three of the study, a subset of the raccoon population was experimentally manipulated through altered resource distribution to enhance the contact of individuals. Within the manipulated subpopulation, prevalence of B. procyonis increased to 54% which was significantly greater than prevalence in the same subpopulation before perturbation or than prevalence among raccoons that were not subject to increased intraspecific contact. These observations suggest that altered resource distributions that directly influence raccoon behaviour may indirectly play a role in the ecology of B. procyonis.