Within the Australian wet tropics bioregion, only 900 000 hectares of once continuous rainforest habitat between Townsville and Cooktown now remains. While on the Atherton Tableland, only 4% of the rainforest that once occurred there remains today with remnant vegetation now forming a matrix of rainforest dispersed within agricultural land (sugarcane, banana, orchard crops, townships and pastoral land). Some biologists have suggested that remnants often support both faunal and floral communities that differ significantly from remaining continuous forest. Australian tropical forests possess a relatively high diversity of native small mammal species particularly rodents, which unlike larger mammalian and avian frugivores elsewhere, have been shown to be resilient to the effects of fragmentation, patch isolation and reduction in patch size. While small mammals often become the dominant mammalian frugivores, in terms of their relative abundance, the relationship that exists between habitat diversity and structure, and the impacts of small mammal foraging within fragmented habitat patches in Australia, is still poorly understood. The relationship between foraging behaviour and demography of two small mammal species, Rattus fuscipes and Melomys cervinipes, and food resources in fragmented rainforest sites, were investigated in the current study. Population densities of both species were strongly related with overall density of seed resources in all rainforest fragments. The distribution of both mammal species however, was found to be independent of the distribution of seed resources. Seed utilisation trials indicated that M.cervinipes and R.fuscipes had less impact on seed resources (extent of seed harvesting) than did other rainforest frugivores. Experimental feeding trials demonstrated that in 85% of fruit species tested, rodent feeding increased seed germination by a factor of 3.5 suggesting that in Australian tropical rainforest remnants, small mammals may play a significant role in enhancing germination of large seeded fruits. This study has emphasised the role of small mammals in tropical rainforest systems in north eastern Australia, in particular, the role that they play within isolated forest fragments where larger frugivorous species may be absent.

In this study I examine the effect of seasonal resource limitation on the behaviour and population dynamics of white-eared kob, Kobus kob leucotis, in the Boma region of the southern Sudan. This population, numbering over 800,000, migrates seasonally between savannah grasslands in areas with low rainfall and ephemerally swamped grasslands in areas with high rainfall. The aims of the study were: (1) to examine whether kob migration tracks ephemeral distributions of food or water resources, (2) to test the hypothesis that the Boma kob population is limited by food availability, (3) to determine if calf production is cued to seasonal peaks in food abundance, and (4) to evaluate the effect of breeding synchrony on lekking behavior and male competition. Seasonal climatic changes produced pronounced changes in the distribution and abundance of both green forage and water supplies. Dry season migration primarily tracked limited supplies of water. Within the dry season range, kob aggregated at high densities (over 1,000 per km²) in low-lying meadows that supported grass re-growth when little green grass was available elsewhere in the ecosystem. However, southerly movements in the wet season were not explainable by the resource hypothesis, since both food and water were widely distributed during the wet season. I suggest that kob may move southward in order to avoid surface flooding during the wet season. Kob mortality during the dry seasons of 1982 and 1983 was considerably higher than estimated mortality during the wet season. Unusual rainfall during the dry season of 1982 provided a "natural experiment" to test the food limitation hypothesis. Adult mortality was significantly lower during the dry season of 1982 than during the more typical dry season of 1983. Calf mortality did not vary significantly between years. Adult mortality rates were related to dry season duration. Dry season mortality was related to sub-maintenance food intake and declining fat reserves. The age structure of the kob population in 1983 suggests that large-scale mortality (ca. 40%) occurred in the 1980 drought that immediately preceded this study. These findings support the food limitation hypothesis. Kob exhibited a 4 month period of calf production during the late wet season, when food availability was highest. As a consequence, females continued lactation through the dry season period of food scarcity. I suggest that kob reproductive phenology may result from an obligatory delay during which females restore their fat reserves prior to calving or selection pressures imposed by predation during the vulnerable post-partum period. Synchronous breeding in the Boma kob was related to increased rates of aggression between males and increased color dimorphism, in comparison to the asynchronous breeding Uganda kob, Kobus kob thomasi. Male aggression served not only to establish dominance relations between males on leks, but also disrupted the mating activities of neighboring males. Young adult males suffered higher age-specific mortality than females, possibly resulting from injuries incurred during strenuous fighting on leks. In order to analyze the age structure of the kob population, I devised a new method for estimating age-specific mortality rates that is free of the restrictive assumptions that underlie most conventional techniques. The proposed method has somewhat greater sampling variation, but is considerably more robust, than two current methods. Moreover, the proposed method permits calculation of age-specific mortality at frequent intervals during periods of population fluctuation and, under some circumstances, population numerical trends may be directly determined from age structure. / Science, Faculty of / Zoology, Department of / Graduate

Increased afforestation and protective legislation in the latter half of the 20th Century allowed the British pine marten Martes martes population to recover from near extinction. Although still largely confined to northern Scotland, the marten population is expanding its size and range by utlising coniferous plantation forests which have become increasingly available. However, little is known about the marten’s ecology in plantation forestry, and less about how they may adapt to changing silvicultural trends. This study investigated aspects of pine marten ecology within Morangie forest, a managed plantation in NE Scotland, with the ultimate aim of formulating management guidelines for modern plantation forests. During the course of the study 11 pine marten were radiotracked and their home ranges mapped to examine marten-habitat associations at several spatial scales. Compositional analysis of habitat based on dominant vegetation type showed that martens established their home ranges in areas dominated by mature forest, whilst showing relative avoidance for open heath moor and grazed pasture. Within home ranges, foraging martens utilised patches of graminoid vegetation, such as those typically associated with Microtus voles, in areas with little or no tree canopy cover. These findings provide unequivocal evidence that fine-scale patches of non-forested habitat provide crucial foraging resources for marten, and therefore ought to be provisioned for in forest management plans. To assist the implementation of these requirements in forest planning, a model was developed to predict the fine scale distribution of Microtus-rich foraging habitat for marten using GIS-based habitat variables that are routinely available to forest managers: topographic wetness index, stand tree height and stand basal area. Management recommendations of ways to improve wind-firm plantation forests as habitats for pine marten are provided. To augment the investigation of marten spatial ecology, the diet of martens was examined seasonally through the analysis of contents from c. 2450 scats, 86 % of which were genetically identified as being pine marten in origin. Marten diets displayed marked seasonality, but small mammals, berries and small birds were the principal foods consumed based on both frequency of occurrence and estimated weight of biomass ingested. Comparison of the relative composition of small mammal species in the diet with those available in the environment revealed that marten displayed an indisputable preference for Microtus voles. Such habits demonstrate that the niche of Scottish martens has diverged from those in mainland populations which predominantly prey upon Clethrionomys voles. A comparison of the marten’s winter diet with those found in studies at similar latitudes (58°N) demonstrated that the Scottish diet was more similar to diets at more southerly latitudes as they contained more fruit and fewer large mammals than typically boreal diets. Investigation of inter-annual variation of the marten’s spring diet from five successive years revealed that Microtus were consistently the most important prey species in the diet each year. Indirect evidence of the relative abundance of Microtus suggested that Microtus populations were non-cyclic. Findings are discussed with reference to the unique ecological circumstances confronting marten in the Scotland; typical of insular populations the UK has a depauperate native fauna in comparison with mainland Europe, in addition to this, mild climatic conditions, particularly over winter, and a historically fragmented landscape appear to have allowed the niche of the Scottish marten to diverge from that considered typical elsewhere in its range. The Scottish marten is dependent on both forested and open habitats, and is both a Microtus specialist and trophic generalist.