A simple variance estimation of herbage mass based on two assessments per pasture

Abstract

Grassland managers should consider the mean and variance of herbage mass for appropriate pasture management. Herein, a time and labor saving method to estimate variance in herbage mass was developed through extending an existing simple estimation method for mean mass. Herein, the difference between the highest and lowest masses in a given paddock and the quotient of pasture size by plot size were substituted for the range and sample size in an existing equation relating range to standard deviation (SD). The applicability of the method was evaluated using Monte Carlo simulations representing virtual pastures with a variety of conditions, and field experiments were conducted to confirm the simulation results. There was a clear relation between the ratio of the range to the SD and volume of generated random numbers in the simulation study, and the curve of the existing equation passed very close to the data points. Thus, the SD of herbage mass could be estimated using this equation. Although the accuracy of the estimate, which was expressed as a ratio of the actual variance of the random numbers, varied widely among simulated datasets, the average accuracy of the estimate was high; a similar result was produced under the field experiments that compares the new estimation method to the random sampling method.