Identification of management thresholds of urban development in support of aquatic biodiversity conservation

Abstract

Urbanization degrades stream ecosystems and causes loss of bodiversity. Using benthic macroinvertebrates as a surrogate for overall aquatic diversity, we conducted a series of analytical approaches to derive management thresholds of urban development designed to link ecological responses to the primary management goal of protecting aquatic diversity in streams within the Delaware Water Gap National Recreation Area (USA). We were particularly interested in identifying urban thresholds that represent the early phases of biological impact to support cost-effect management and mitigation interventions. We used taxa-specific modeling approaches within a spatially-explicit framework to develop sensitive thresholds that anticipate and demark the onset of taxa loss and provide a foundation for investigating alternative mechanisms driving biological change. We identified an early-warning threshold of 1.5% urban development in the contributing watershed where 15% of the 107 taxa evaluated exhibited significant declines in abundance but prior to any evidence of extirpation, and an extirpation threshold of 6% urban development where nearly 3% of taxa are likely to be lost locally. These thresholds of urban development are substantially lower than response thresholds typically reported based upon traditional modeling approaches that rely on spatially-implicit summaries of land cover and univariate metrics or composite indices. An analysis of ecological and functional trait composition of taxa determined to be sensitive suggests that reduced storage of benthic organic matter caused by flashier hydrographs may be the primary mechanism driving biological changes observed at relatively low levels of urbanization. Although the extent to which stream communities respond to stressor gradients in a non-linear fashion continues to be debated, we show that threshold approaches can be applied in support of aquatic resource management irrespective of whether or not stress-response functions are non-linear.