Estimating Sufficient Sample Sizes to Detect Changes in Eelgrass Density

Categories:

Published: 2006

Pages: 7

Author(s): Kirk Krueger

Abstract

Eelgrass (Zostera marina) provides important ecological services in the nearshore, has been designated as critical habitat, and is protected by a no-net-loss policy in Washington State. When projects are proposed that may impact eelgrass, such as installing docks or underwater cables, the state of Washington may require surveys to monitor changes in eelgrass density. The guidelines for surveys required by the Washington Department of Fish and Wildlife are sufficiently general to allow rigorous monitoring. However, specific guidance describing sufficient sample sizes and analytical methods has not yet been developed. We used data from a previous study to estimate the number of samples required to reliably detect a range of declines in shoot density using a before-after-control-impact (BACI) survey design. Sufficient sample size to detect a change in eelgrass density is a function of the magnitude of the change in density (effect size), sampling design (including number of samples), variation of eelgrass density among samples, and selected probabilities of Type I and Type II errors. Our study demonstrates the need to consider these factors, with particular reference to Type II errors, when developing sampling designs. Our results have implications for most eelgrass monitoring projects.