Prioritizing conserved areas threatened by wildfire and fragmentation for monitoring and management.

Abstract

In many parts of the world, the combined effects of habitat fragmentation and altered disturbance regimes pose a significant threat to biodiversity. This is particularly true in Mediterranean-type ecosystems (MTEs), which tend to be fire-prone, species rich, and heavily impacted by human land use. Given the spatial complexity of overlapping threats and species' vulnerability along with limited conservation budgets, methods are needed for prioritizing areas for monitoring and management in these regions. We developed a multi-criteria Pareto ranking methodology for prioritizing spatial units for conservation and applied it to fire threat, habitat fragmentation threat, species richness, and genetic biodiversity criteria in San Diego County, California, USA. We summarized the criteria and Pareto ranking results (from west to east) within the maritime, coastal, transitional, inland climate zones within San Diego County. Fire threat increased from the maritime zone eastward to the transitional zone, then decreased in the mountainous inland climate zone. Number of fires and fire return interval departure were strongly negatively correlated. Fragmentation threats, particularly road density and development density, were highest in the maritime climate zone, declined towards the east, and were positively correlated. Species richness criteria showed distributions among climate zones similar to those of the fire threat variables. When using species richness and fire threat criteria, most lower-ranked (higher conservation priority) units occurred in the coastal and transitional zones. When considering genetic biodiversity, lower-ranked units occurred more often in the mountainous inland zone. With Pareto ranking, there is no need to select criteria weights as part of the decision-making process. However, negative correlations and larger numbers of criteria can result in more units assigned to the same rank. Pareto ranking is broadly applicable and can be used as a standalone decision analysis method or in conjunction with other methods.