You are here:
Using attributable risk to assess the regional-scale impacts of environmental stressors
Citation:
VAN SICKLE, J. AND S. G. PAULSEN. Using attributable risk to assess the regional-scale impacts of environmental stressors. Presented at 2008 National Monitoring Conference, Atlantic City, NJ, May 18 - 22, 2008.
Impact/Purpose:
e describe the application of population attributable risk (AR) for assessing the relative importance of aquatic stressors across large regions.
Description:
We describe the application of population attributable risk (AR) for assessing the relative importance of aquatic stressors across large regions. A stressor's importance depends on its regional extent (e.g., the total length of stream with elevated stressor levels), and also on its relative risk to biological condition (e.g., as indicated by a macroinvertebrate multimetric index). AR combines a stressor's extent and relative risk into a single measure of that stressor's impact on the regional population of streams or lakes. AR is defined as the proportional reduction in the extent of poor biological condition that could be achieved if the stressor were eliminated from the population. The AR concept is borrowed from human epidemiology, and we describe its assumptions and limitations for assessing ecosystems. We describe how AR and its confidence interval can be estimated for unequal-probability sample surveys. In addition, the AR estimate for any one stressor can be adjusted for the effects of other, correlated stressors. We show example applications to data from the 2004 National Wadeable Streams Assessment.