The Sampling Unit and Its Effect on Saltbush Yield Estimates

Abstract

A 60- by 60-ft plot of Nuttall saltbush (Atriplex nuttallii) was completely harvested using 1- by 1-ft sampling units. Yield data for sampling units of various sizes and shapes were obtained by combining contiguous 1- by 1-ft plots. Sampling unit size and shape both had an important effect on sampling efficiency. The coefficient of variation decreased rapidly, as sampling unit size increased, until a sampling unit of about 60 ft2 was reached. For the smaller sampling units, the rectangular units generally had a lower coefficient of variation than those that were nearly square. Data are presented showing the relationship between sampling units of various sizes and shapes and (1) their coefficient of variation, (2) the number of samples needed to obtain a yield estimate that is within 20% of the population mean 90% of the time, and (3) the total area needed to be sampled to obtain this yield estimate.