Prioritizing management options for conservation, with applications in avian landscape ecology

To prioritize competing conservation management options ecologists and conservation biologists must estimate their relative potential benefits to biodiversity. Quantitative estimates of these relative potential benefits, or similarly, the relative effects of ecological factors on biodiversity, are uncommon in the literature. In addition, studies have used a wide range of disparate statistical methods, which has limited our ability to synthesize their results. In this thesis, I address the following questions: 1) Does our current understanding of the relative effects of habitat loss and habitat fragmentation on biodiversity depend on the choice of statistical methods used to account for their correlation; and which methods are the least biased by this collinearity? 2) Are the relative effects of habitat amount, habitat fragmentation, and matrix quality scale dependent? 3) Of the disparate statistical metrics that might be used measure the relative importance of conservation management options, what are the theoretical connections among them and which of them meet five criteria that are necessary to compare results across studies? 4) What is the relative importance of managing residential vegetation (yards and street trees) versus non-residential vegetation (parks and large greenspace) on the species richness and conservation value of an urban bird community?;In Chapter 2, I have shown that, except for standardized partial regression coefficients, the statistical methods used to account for the correlation between habitat amount and fragmentation are biased by both collinearity and the direction of their effects on biodiversity. In Chapter 3, I have shown a scale dependency in the relative effects of forest amount and forest fragmentation on bird species that are considered sensitive to forest fragmentation, which is largely due to the higher variability in fragmentation effects across scales. In Chapter 4, I have shown that the best statistical metric for prioritizing competing management options is a partial regression coefficient, re-scaled as a proportion of the practical or theoretical limits to management. I have also proposed a theoretical framework and categorization of the statistical metrics of relative importance as well as practical criteria by which to judge other metrics. In Chapter 5, I have shown that the relative effects of residential and non-residential vegetation area on the richness and conservation value of an urban bird community are approximately equal at a local scale, but as human population density increases, the relative importance of non-residential vegetation declines.