Random, spatially-uncorrelated, distribution patterns of species abundances are rare in nature, yet few benthic ecologists explicitly incorporate spatial patterns into their analysis.  Spatial autocorrelation can be viewed as a statistical problem or as an opportunity. 

            We examined spatial patterns of three snail species, including the threatened Bliss Rapids snail and the invasive New Zealand mudsnail, in a freshwater spring tributary of the Snake River, Idaho.  We used two spatially explicit methods, Spatial Analysis by Distance Indices (SADIE) to formally test for non-random distributions and geostatistical kriging to estimate abundances.  We then compared these abundance estimates with a more commonly used method that did not explicitly incorporate spatial patterns. 

            The spatially explicit abundance estimate of the threatened Bliss Rapids snail was more accurate but less precise than estimates that did not incorporate spatial pattern.   Major differences in abundance estimates for the invasive New Zealand mudsnail occurred, as well. 

            We suggest that accuracy is critically more important than precision when estimating abundances of threatened and endangered species.  This is particularly essential when incorporating abundance estimates into risk assessments, population viability analyses, or status reviews.  When neccesary, spatial patterns and spatial autocorrelation should always be examined and explicitly modeled.