SEAFWA 2018

AUTHORS: Healy Hamilton, NatureServe; Regan Smyth, NatureServe

ABSTRACT: Understanding species distributions is one of the most fundamental information needs of management and conservation. Particularly for at-risk species, the use of county boundaries or other coarse-scale range maps can result in unnecessary regulatory burdens and inefficiencies in directing conservation and management efforts. Species distribution modeling, an approach widely used in academic research, combines locality data from verified species observations with environmental predictors, such as climate and land cover, to produce maps of probability of suitable habitat. With modern advances in computational capacity, in aggregated species observation databases, in the availability of high-resolution environmental predictors, and in the quantification of uncertainty, we argue that distribution modeling has matured into a tool that can help refine species distributions and improve science-based and transparent management decisions. Many examples exist where distribution modeling has been successful in guiding field inventory to discover new populations, identifying potential habitat for restoration, and assessing climate change vulnerability, among many other relevant applications. However, distribution models also have some drawbacks. They are not applicable to all taxa, such as very wide ranging species, or species whose distributions are largely controlled by poorly known environmental factors. They are also not straitforward to interpret. We advocate for the advancement of distribution modeling as a component of best available scientific information in state and federal agency species management activities. The development of standards for producing decision-quality models, and guidance and best practices for their interpretation, are essential components of expanding the use of species distribution modeling in management and conservation.