Blue grama-buffalograss responses to grazing: a Weibull distribution.
Abstract
Characterization of standing herbaceous biomass on rangeland is complicated by both temporal and spatial variability that results from patchiness in vegetation. These patches often cause nonuniform levels of grazing by livestock. Currently accepted methods for estimation of forage, and its utilization, assume a normal distribution. This assumption may not be appropriate if the frequency distribution of amount of biomass becomes skewed as grazing occurs. We evaluated the 3 parameter Weibull distribution as an alternative to the normal distribution in modeling the frequency distributions of plant height and biomass as a function of grazing intensity over time in a shortgrass steppe. Weibull distributions, estimated by probability weighted moments, fit all observed plant height and biomass data distributions at the alpha = 0.05 level of significance. In contrast, the normal distribution fit only 25% of the data sets.
Keywords
mixed pastures;frequency distribution;Buchloe dactyloides;mathematical models;spatial variation;prediction;temporal variation;Bouteloua gracilis;grazing intensity;cattle;biomass;grazing;plant height