Modelling the effect of habitat fragmentation at two scales: consequences for species movement and persistence

nnn In recent years, major advances have been made in understanding the relationship between patterns of habitat fragmentation and movement and persistence of organisms, with simulations in artificial landscapes commonly used to address this question. Yet no studies have investigated how these processes interact when fragmentation is observed at two or more resolutions. This is a key issue because different resolutions of habitat fragmentation may have different impacts on movement and fitness of organisms, with critical consequences for species persistence. Furthermore, different resolutions of habitat fragmentation can interact to cumulatively impact on speciesr movement and fitness. Also, the impact of habitat fragmentation on speciesr movement and fitness may depend on the life-history characteristics of a species. Despite the implications of the resolution at which habitat fragmentation occurs for the long term survival of species in fragmented landscapes, no studies have investigated how different resolutions of habitat fragmentation interact to affect movement and fitness of species with different life-history characteristics, which therefore is a key research priority.nnn The primary aim of this thesis was to investigate the impacts of habitat fragmentation, occurring at two resolutions, on the movement and persistence of species with different life histories. Specifically, I addressed four research questions: (1) How do different land uses drive the resolution at which habitat fragmentation occurs? (2) Does habitat fragmentation at one resolution interact with the amount of habitat, dispersal distance, and habitat fragmentation at other resolutions to affect species' dispersal? (3) What are the consequences of the resolution at which habitat fragmentation occurs for the fitness of species with different life histories? (4) Does the resolution at which fragmentation of existing habitat occurs influence the preferred choice of habitat restoration strategies?nnn I answered the first question using an approach based on fractal theory, to quantify the degree of habitat fragmentation in landscapes subjected to different land uses, using Queensland, Australia, as a case study. To address questions two, three and four, I developed a novel simulation framework that independently varies the intensity of habitat removal (i.e., how much habitat is removed in a single removal event) and the degree of fragmentation at two nested resolutions, while controlling for the total amount of habitat in the landscape. Using this framework, I conducted individual-based simulations to understand the impacts of the resolution at which habitat fragmentation occurs on dispersal and fitness of species. Finally, for question four, I applied the simulation framework under different restoration scenarios to explore whether the preferred choice restoration strategy depends on the resolution at which fragmentation of existing habitat occurs.nnn First, I showed that cropping and grazing land uses drive the degree of habitat fragmentation at different resolutions. In particular, the difference between the degree of habitat fragmentation at the coarse resolution and the degree of habitat fragmentation at the fine resolution is higher in cropping than in grazing land use. Furthermore, I showed that the resolution at which most of the fragmentation occurs is finer in cropping than in grazing landscapes, suggesting that fragmentation due to cropping operates at finer scales than fragmentation due to grazing.nnn Second, I showed that the impact of the amount of habitat, the dispersal distance and the degree of fragmentation on dispersal success depends on the resolution at which habitat fragmentation occurs and the intensity at which habitat is removed. Simulations revealed that these effects are weaker when there is strong habitat selection during dispersal. Third, I showed that individual fitness is more affected by fragmentation at fine resolutions than fragmentation at coarse resolution. However, the fitness of species incurring low foraging costs was more affected by habitat fragmentation at fine resolution than the fitness of species with high foraging costs. Fourth, I demonstrated that the consequences for individual fitness of restoration strategies that increase patch size versus those that increase connectivity between habitat patches depends on the resolution at which fragmentation of existing habitat occurs.nnn This thesis makes an important contribution to our understanding of the drivers of scale-dependent landscape change, and its consequences for the spatial dynamics of species in human-modified landscapes. The findings of this research have also important implications for landscape management and conservation. In particular, the study outcomes suggest that habitat fragmentation associated with cropping land use should be managed at fine resolutions, while habitat fragmentation associated with grazing land use should be managed at coarser resolutions. Furthermore, the resolution at which fragmentation should be managed also depends on the amount of habitat in the landscape, a species' dispersal distance, and the interaction between the degrees of fragmentation at different resolutions. Appropriate use of this information will significantly reduce the risk of mismatches between the scale of ecological processes and the scale of management. Finally, the thesis highlights the importance of considering the resolution at which fragmentation of existing habitat occurs, and the life-history characteristics of species, in developing effective habitat restoration strategies.