A COMPUTER-PROGRAM TO ASSIST WITH SAMPLING DESIGN IN THE COMPARISON OF NATURAL AND CONSTRUCTED WETLANDS

Statistically sound comparisons of natural and constructed wetlands ar e complicated by variability between and within wetlands. This paper d emonstrates the use of a SAS(R) program that estimates sampling effort and allocates that effort among number of wetlands, number of station s within wetlands, and number of samples to be collected at each stati on. Estimates are based upon variances known from pre-existing data, d esired statistical power and critical effect, and project budget. The program is used to generate a sampling design for comparison of chiron omid populations of natural and constructed freshwater herbaceous wetl ands in central Florida.