Dynamical effects of the statistical structure of annual rainfall on dryland vegetation

In this study, we extend a model of daily dryland dynamics by parameterizing a modified version of a minimalistic annual model to examine how the statistical structure of annual rainfall and grazing intensity interact to influence dryland vegetation. With a Monte Carlo approach, an ensemble outcome provides a statistical description of likely dryland vegetation dynamics responding to variations in rainfall structure and grazing intensity. Results suggest that increased rainfall variability decreases the average and increases the variability of grass cover leading to more frequent degradation of the grass resource. Vegetation of drier regions is found to be more sensitive to interannual variability in rainfall. Concentrating this variability into an organized periodic mode further decreases the mean and increases the variability of grass cover. Hence, a shift toward lower, more variable, or more inter-annually correlated annual rainfall will likely lead to a general decrease in the grass resource and increased dryland vulnerability to degradation. Higher grazing intensity or lower annual rainfall both lead to more frequent and longer duration degradation of the grass condition. We note an interesting interaction in the response of grass biomass to grazing intensity and rainfall variability, where increased rainfall variability leads to longer duration degradation for low grazing, but shorter periods of degradation for high grazing. Once grass reaches a degraded condition, we find that woody vegetation strongly suppresses recovery even if successive rainfall is high. Overall, these findings suggest that the projected increase in interannual rainfall variability will likely decrease grass cover and potentially lead to more frequent, longer lasting degradation of dryland vegetation, particularly if enhanced rainfall variability is concentrated in long period (e.g. decadal) modes.