Severe declines in biodiversity have been attributed to anthropogenic changes in the composition and structure of our landscapes. Predicting the impact of landscape change on biodiversity is essential to halt further declines. In this thesis butterflies were used as indicators of biodiversity, and spatial assessments of butterflies were summarised at 1 km scale across Warwickshire to assess whether landscape characteristics can be used as surrogate measures of butterfly distribution and community measurements. When determining the optimal scale (grain size) for capturing landscape patterns, a grain size of 25 m was found most appropriate for maximising landscape discrimination and detecting landscape patterns which occur within the perceptual range of butterfly species. Utilising a grain size of 25 m landscape metrics measuring the composition, connectivity and structure of the 1 km landscapes, were extracted from the Land Cover Map 2000 (LCM) and the Warwickshire Phase 1 Habitat map (PH1). Logistic regression analysis based on landscape metrics created predictive models of butterfly distribution for all species and species grouped by their ecological attributes (EAGs). Model performance was improved when the landscape metrics were considered in a combined landscape model, and different combinations of landscape parameters were important for the EAGs. Models derived from the PH1 were most accurate in predicting observed presence-absence and were successfully transferred when tested using temporally independent data. The models were also successfully transferred to collected butterfly data which was spatially and temporally independent. This data was also collected alongside information on the local habitat such as vegetation composition. Probability of butterfly occurrence derived from the presence-absence models was successfully related to butterfly community characteristics and measures of local habitat quality. To conclude developed models provide indications of habitat suitability, which together with successful transfer demonstrates their potential for identifying biodiversity hotspots and facilitating targeted conservation efforts.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:640949 |
| Date | January 2014 |
| Creators | Carter, Charlotte Emily |
| Publisher | University of Warwick |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://wrap.warwick.ac.uk/66477/ |
Page generated in 0.0046 seconds