Species distribution ranges and conservation : spatial structure, contraction patterns, global change impacts, and bias in species data

Abstract

The impact of human activities is causing an accelerated loss of biodiversity. The two main human impacts, habitat loss through land use modifications and climate change are predicted to be widespread and more intense in the future, increasing the risk of extinction of many species. Therefore, there is an urgent necessity to improve our understanding of how extinctions occur in order to improve predictions and to maximize the effectiveness of conservation policies. The complete extinction of a species is usually preceded by a process of extinction of its local populations which causes a reduction of the geographic extension of its distribution range, in a process called range contraction. Distribution ranges are naturally dynamic, but nowadays, the impact of human activities is the main factor driving colonizations and local extinctions. The spatial structure and dynamics of species¿ ranges are related to their ecology, their coexistence with other species and other historical, and evolutionary factors, affecting the overall probability of extinction and recolonization. These factors and processes show variability among taxa due to differences in the spatial characteristics of the range including the prevalence of human impacts. Studying the structure and dynamics of geographic ranges including how they contract offers the opportunity to assess the vulnerability of populations to human impacts and understand the processes leading to global extinction. In the present thesis I explore several aspects of range dynamics including whether we can define a null model of range contraction (Chapter 1), how the spatial structure of the range determines the vulnerability of species (Chapter 2), the distinct and interacting effects of climate change and land use in driving distribution range contractions (Chapter 3), and the existing bias in data availability and how these biases influence our ability to assess vulnerability to extinction (Chapter 4). In the first chapter, I show that traditional null models of range contraction fail to explain range contractions in terrestrial vertebrates and propose new multifactorial models that combine land use and spatial configuration to generate baseline expectations of range contraction. In the second chapter, I identify several spatial configuration descriptors, in addition to the widely use total range area, which are associated with vulnerability to extinction. In particular, fragment size heterogeneity and shape are the most important predictors, but there are also multiple interactions that illustrate the complex ways in which spatial configuration influences extinction risk. In the third chapter, I show the importance of both land use and climate change in currently observed biodiversity loses. The combination of both threats leads to severe range contraction, with species remaining in refuge areas which may not be representative of their original niche breadth. In the fourth chapter, I explore possible bias in species data availability and their effect on the findings of macroecological studies of extinction risk.