| A central objective in ecology and conservation biology is to understand processes limiting species' distributions and population connectivity. This is particularly important for amphibians, which are in global decline at rate exceeding other vertebrates. Niche modeling and landscape genetics are well suited for addressing distribution and connectivity respectively. This dissertation addresses four objectives: (i) test whether landscape genetics is possible on spatial and temporal scales relevant to conservation, (ii) examine whether spatial processes, environmental condition or dispersal limited niche best explains the observed differences in Pseudacaris maculata and Bufo boreas distributions in Yellowstone, (iii) test alternative hypotheses of ecological processes driving B. boreas connectivity in Yellowstone, and (iv) develop a new application of gravity models to estimate metapopulation connectivity for Rana luteiventris in central Idaho. In Chapter 1, I developed and evaluated "genetic surfacing", a continuous method for representing multilocus genetic variation. I detected landscape genetic structure on a contemporary time scale relevant to conservation questions (≥5 generations post vicariance, migration probability ≤ 0.10), even when population differentiation was minimal (FST ≥ 0.00015). Using spatial distribution models in Chapter 2, I found environmental conditions limiting species' distributions to be divergent and distance limited niche theory best explained observed species' distributions. In Chapter 3, I implemented a novel algorithmic approach to test alternative hypotheses of processes limiting connectivity in B. boreas. At fine scales, connectivity was limited by cover, precipitation and roads, whereas ridges, temperature, and precipitation limited connectivity at a broad scale. Using newly derived gravity models in Chapter 4, I found R. luteiventris connectivity was a function of both at site and between site landscape processes. Primary productivity and fish presence at sites limited production of potential migrants. Temperature and major topographic complexity between sites limited connectivity. The impact of temperature-moisture regimes on all three species suggests that future climate change may have a dramatic impact on anuran distribution and connectivity. The methods developed in this dissertation could be used to predict species' distributions (niche models) and resulting connectivity (landscape genetics) in future landscapes. |
