An Evaluation of Random and Systematic Plot Placement for Estimating Frequency

Abstract

A computer simulation study was conducted to evaluate the effects of pattern on the precision of frequency estimates as determined from random and systematic plot placement. Computer graphics were used to generate artificial population maps containing 40 or 80 clumps of differing spatial intensity with known frequencies of 20, 35, and 50%. The maps were repeatedly sampled both randomly and systematically using a 200-plot sample size to obtain frequency estimates. Three systematic plot spacings (4, 8, and 12) along randomly located transects were evaluated. Analysis indicated that frequency means from systematic plot placement were significantly affected by clumping, pattern intensity, and plot spacing. Random sampling resulted in frequency means that were unaffected by clumping or pattern intensity, and more consistently estimated population frequencies. An evaluation of probabilities of occurrence of Type I errors when statistically comparing frequency estimates from systematic plot placement indicated higher Type I error rates as compared to random sampling.