Comparison of Distance-Measurement Techniques for Sampling Tussock Grassland Species in New Zealand

Abstract

Four distance-measurement techniques were tested in 1969 near the Craigieburn Research Station of the Forest and Range Experiment Station, New Zealand Forest Service. Two methods (closest individual and point-centered-quarter) are known to result in biased estimates of density in nonrandom plant populations, and the other two methods (angle-order and corrected-point-distance) have been reported to permit unbiased estimates in nonrandom populations. Sampled were two tussock grassland species in areas where true densities were determined by total counts. The corrected-point-distance method resulted in estimates within 20% of actual density for all populations, and appeared to be the most promising measurement method tested. The closest individual and point-centered-quarter methods gave badly biased (low) estimates of aggregated populations. Reasonably good estimates for some populations were obtained by the angle-order method, but it is time consuming. All methods appear to need additional theoretical work.