Death model for tussock perennial grasses: thresholds for grazing-induced mortality of mulga Mitchell grass (Thyridolepis mitchelliana)

Abstract

We investigated the relationship between grazing intensity and death of mulga Mitchell grass [Thyridolepis mitchelliana (Nees) S.T. Blake] plants in semi-arid wooded grasslands of eastern Australia. The study (from July 1993 to March 1997) involved factorial combinations of intensity and duration of sheep grazing before rest from grazing. Grazing intensity varied considerably among plants within the small plots and logit analysis of the binary responses of individual plants (a plant was judged to be alive or not) was used to establish the relationships between grazing history and death. Average residual (shoot) biomass over a moving window of 10?12 months was a good predictor of either general plant death or death during summer drought. Death over summer increased as average residual biomass dropped below 70%, and increased rapidly when the average fell below 50%. Variables based on residual shoot biomass generally provided better predictors of death than variables based on foliar height. However, as a predictor of death over summer, current foliar height was as good as average residual shoot biomass over the extended period. Summer death increased rapidly as foliar height fell below 10�cm. Environmental conditions were much less important than grazing in determining death rates, indicating that grazing management can have important benefits in maintaining productive grasslands even in more or less ?normal' seasons.