Analysis of the production stability of mixed grasslands. I. A conceptual framework for the qualification of production stability in grassland ecosystems

The increased use of white clover in grasslands has led to new management challenges, as mixed pastures have been associated with unstable herbage production. The stability of mixed pastures depends on a complex of intrinsic ecosystem properties, on the variability of the environment, and on the spatial and temporal scale at which it is studied. In this paper, intrinsic ecosystem properties and processes are explored and illustrated using a simple dynamic simulation model of grass and clover interactions. Competition for light, the delayed availability of nitrogen fixed by white clover, grazing per se, preferential grazing for clover, and cutting were identified as destabilising processes. Instead, niche-differentiation, nitrogen dependence of grass and clover, and plant mechanisms to `escape¿ from grazing, were identified as stabilising processes. The intrinsic stability of mixed swards depended on the balance of, and the interactions between these processes. Including the temperature as an environmental variable into the model, unstable ecosystems were stabilised by seasonal temperature fluctuations, and were either further destabilised, or stabilised by stochastic temperature fluctuations. Stable ecosystems were always destabilised by these stochastic fluctuations. It is explained how spatial heterogeneity can stabilise ecosystems, which oscillate at patch scale. Heterogeneity can be maximised by increasing the incidence of small-scale disturbances and by minimising large-scale disturbances. Finally, three concepts of stability are presented. The actual stability is defined as the stability of grasslands as measured in the field, i.e. subjected to both seasonal and stochastic environmental fluctuations. The extrinsic stability presumes the presence of seasonal, yet the absence of stochastic environmental fluctuations. The intrinsic stability represents the stability of yields in a hypothetical constant environment. It is explained how these concepts of stability can bridge the gap between experimental and theoretical studies. It is demonstrated that long-term experiments are required for the experimental analysis of grassland stability, and it is argued that the development of spatially and environmentally explicit simulation models is a prerequisite for the prototyping of management systems for mixed grasslands.