Research note: Calibrating a disc pastures meter to estimate grass fuel loads on the Zululand coastal plain

Abstract

A disc meter was calibrated to estimate fuel load on the Zululand coastal plain. The regression model was valid for fuel load estimates within a disc height of 1-58cm. The mean calibrated disc height was 11.8cm. Estimated fuel load at mean disc height, with standard error (SE), was 4 593?113kg ha^-1. Using this equation to estimate fuel load at the calibrated mean, the precision of the estimates varied depending on the number of disc readings. As sample size increased there was at first a rapid reduction in the SE and hence in the 95% confidence limits (CL) of the predicted fuel load. There was little reduction in the CL with increasing sample size for more than 100 observations. To achieve a CL?10% of 5 000kg ha^-1, 100 samples are needed. With fuel loads of less than 4 000kg ha^-l, more than 160 samples are required to achieve the same precision. Sampling with more than 100 readings is not, however, recommended owing to the poor reward (precision) per unit of sampling effort

Keywords: coastal forest; disc meter; fire management; fuel load; grass; grasses; herbage mass; pastures; precision; regression model; sample size; south africa; standing crop; thornveld; zululand

African Journal of Range & Forage Science, 13(1): pp. 39-41