IDENTIFYING EXTINCTION RISK PATTERNS IN POLLINATION NETWORKS AND CONSERVATION PLANNING TO REDUCE THE COST OF MANAGEMENT-NECESSITATED TRAVEL

Extinction rates are currently far above background levels, as a result of human activity. Although conservation spending is substantial, there is a shortfall between the cost of meeting biodiversity targets and the resources available to do so. Consequently, decision-makers are faced with the task of identifying which taxonomic groups and geographical regions are of greatest concern, and prioritizing approaches with the greatest conservation benefit. In this dissertation, I develop novel approaches in both of these areas.

First, I take a stochastic, community dynamic approach to modeling extinction risk patterns in pollination networks. Despite the ecological and agricultural importance of pollination and the expectation that environmental variability will increase under climate change, stochasticity is an under-studied driver of extinction risk in the mutualist modeling literature. Here, I demonstrate that extinction risk in a simplified network increases with environmental variability, and that this increase is more rapid for specialists than for generalists. I then examine this pattern over a range of realistic network sizes and structures, using interaction data from 35 empirical networks. Across all networks, extinction risk declines with increasing number of mutualist partners for poorly-connected species, and approaches an asymptote for well-connected species. I use a nonlinear mixed regression to describe how network connectance and nestedness change the relationship between specialization and extinction risk. Highly connected and nested networks showed a significantly steeper decline in extinction risk with increasing partner number, as well as lower extinction risk for the most specialized species.

Second, I develop a method to reduce the operating costs that conservation organizations incur as a consequence of undertaking management activity. Land protection decision-making is well studied, but the question of where to base conservation staff is not. Office locations, however, affect the cost of management-necessitated travel. Here, I estimate the annual over-road travel cost incurred by the management activity of two organizations, both under their current office configurations and configurations that minimize total travel distance. In each case, the reduction in travel cost from either adding an office or reconfiguring current locations is sufficient to protect 171-360 Ha or employ an additional management officer.