Spatial modeling of plant distributions : coupling remote sensing with GIS-based models
Spatial species distributions and the relationship between species and environmental factors have been studied for several years. Climate change and habitat fragmentation can be considered as the factors effective in biodiversity changes. Therefore prediction of species range shifts under climate change and other physical processes is a crucial challenge for the management of natural resources. The major objective of this thesis was to integrate MigClim, SDM and CA-Markov chain models so as to assess the effects of future landscape fragmentation and climate change scenarios on the geographic distributions of three open-land plant species in 21st century. For all target plants, simulations were performed for four dispersal events (full dispersal, no dispersal, regular dispersal (short-distance dispersal), and regular dispersal along with long-distance dispersal), two landscape (static and dynamic change) and two climate change (RCP4.5 and RCP8.5) scenarios (chapter 5). In this investigation, it was shown that the predicted distribution areas for all the three species under RCP8.5 scenario will largely increase in the coming decades. Also, a significant difference appears to be between the simulations of realistic dispersal limitations and those considering full or no dispersal for projected future distributions during the 21st century. Besides, the results obtained by the limited projections of future plant distributions via realistic dispersal restrictions showed to be generally closer to no-dispersal than to full-dispersal scenario when compared with real dispersal scenarios. Overall, the results of this study indicate that dispersal limitations can have an important impact on the outcome of future projections of species distributions under climate change scenarios. Also our findings clearly showed that change in landscape fragmentation is more effective than the climate change impacts on species distributions in our study area.