The detectability of amphibian population declines is an important consideration in the design of monitoring projects. Previous studies have considered statistical power in the context of intensive studies of single populations and sample size requirements for extensive monitoring of broadscale trends. Both approaches have assumed that the goal is to detect gradual declines in abundance of animals. We suggest that extensive monitoring data can be used to detect declines in the distribution of animals by using the inherent geographical structure of the data collected. Because under natural circumstances the distribution of many amphibian species is less variable than their abundance levels, such declines should be easier to detect.
[ Read this Paper ]