A POWER ANALYSIS OF METHODS FOR ASSESSMENT OF CHANGE IN SEAGRASS COVER

Blade counts and biomass measurements from cores were compared with qu adrat surveys in terms of the minimum sample sizes required to detect changes in seagrass cover with specified statistical power levels. Bla de counts and biomass methods required extremely high replicate number s to detect differences in seagrass cover. Percent cover estimation us ing quadrats required the fewest replicates and the least sampling eff ort for detection of a given magnitude difference, and is therefore th e more cost effective method, Use of a 2 m(2) quadrat balances field e ffort with the number of replicate samples required to detect a differ ence, while allowing the statistical power of the test to remain high.