Community-based management of natural resource (CBNRM) projects have commonly failed to deliver conservation and development benefits. This thesis examined how the theoretical assumptions of common pool resource (CPR) theory have contributed to the indifferent performance of CBNRM projects. Evidence was gathered from two CBNRM case studies in Zanzibar to show that CPR institutional design does not sufficiently acknowledge the politics or social relations of project sites. Moreover, these limitations reduce CPR theory's explanatory power and the functionality of CBNRM projects. This is because CPR theory's influence on CBNRM projects is to frame people with fixed identities and related interests as 'rational resource users', rather than people enrolled in multiple network relations with differentiated means of influence, interests and responsibilities. Actor-oriented theory is used to show that CBNRM would benefit from a shift in the correlation with institutional design factors to understanding the operation of power and conflict at project sites. These findings suggest that currently CBNRM projects are too mired in concern about regulating the 'direct' relationship between resource users and conservation objectives, with problematic implications. It is shown that actor-oriented theory is more sensitive to the different capacities, interests and strategies of actors in CBNRM institutional transformation processes. While actor-oriented theory does not offer a parsimonious or predictive theory to reform CPR theory or CBNRM policy, it can provide insights into pre-project conditions and emergent practice useful for explaining project interventions.

Monitoring the distribution, abundance and demography of species is vital to ensure that conservation efforts are appropriate and effective. Monitoring enables evaluation of responses to natural or human disturbance, highlights the need for management interventions and enables these interventions to be assessed and refined. The methods used largely depend on the specific aim of monitoring and behaviour of the target species, as well as the time and spatial scale that monitoring is required to cover. The European pine marten (Martes martes) is one of few remaining mammalian predators native to the UK. Since persecution in the early 19th century resulted in their near extinction, pine martens have recovered part of their former range in Scotland. Their recent recovery and an overlap in territory with vulnerable prey species such as the capercaillie (Tetrao urogallus) make the collection of accurate baseline data and subsequent monitoring of population trends vital for conservation efforts. Faecal counts have traditionally been used to provide a relative measure of population density for pine marten. In most cases, absolute measures of population density require individual identification. Non-invasive genotyping can provide this information but the quality of DNA from these samples is often poor. Here, the process is refined for pine marten faeces (scats) and hair samples. DNA degradation increased significantly for scats exposed to rainfall, with the rate of DNA amplification success reduced by 38% over a 16 day period. Success rates for hair samples were higher when more hair follicles were included in the analysis. Population densities were estimated for three forests in Scotland using a robust combination of non-invasive genotyping of hair samples and spatially explicit capture recapture modelling. Population density estimates of 0.07 (95% CI 0.03 - 0.16) to 0.38 km-2 (95% CI 0.11 - 1.07) were obtained which are within the range of previous estimates for forests elsewhere in Scotland. The first attempt to quantify the relationship between traditional scat counts and home range size was then conducted; a significant negative correlation exists but only when population density is relatively low. Previous studies suggest that pine martens in Western Europe are less forest dependent than those in Eastern Europe. Results from this thesis support this, with populations at the highest density found at sites with intermediate forest cover. This tolerance of lower forest cover is also apparent in the diet. Despite a preference for Myodes voles in populations in Eastern Europe, those in Western Europe show a strong preference for Microtus voles as well as a higher level of frugivory. We assess the autumn diet of four populations in Scotland assess the effect of forest cover and sex on the diet. There was no evidence of differential consumption of Microtus voles or birds between the sexes. Our analysis shows that frugivory is influenced by forest fragmentation, with a 5-fold increase in the occurrence of fruit (from 2% to 10%) as forest cover increased from 4% to 47%. Diversionary feeding has been suggested as a management technique to reduce the depredation of capercaillie by pine martens. This thesis presents the first attempt to quantify the success and cost-efficacy of diversionary feeding for a range of problems (crop damage, threatened safety, livestock depredation) across 30 experimental trials. The strategy proved more effective when targeted towards food-limited populations, and when aiming to alleviate habitat damage or risks to human safety rather than depredation. A novel decision-making framework was developed to aid managers with the initial planning of the strategy and its subsequent implementation within an adaptive format. Further to this, the feasibility of using diversionary feeding with a view to reduce the depredation of capercaillie by pine martens was assessed. Questionnaire responses were collected from people who have provided food for pine martens throughout Scotland. A positive reaction to food was observed, with 58% of respondents reporting that initial visits occurred within one week of placement and 46% reporting that subsequent visits were received daily. These results suggest that diversionary feeding may be a viable option for pine marten management, although testing of its impact on capercaillie productivity would be required.