Resilience in reaction-diffusion systems

Reaction-diffusion systems with zero-flux Neumann boundariesare widely used to model various kinds of interaction in, forexample, the scientific fields of ecology, biology, chemistry,medicine and industry. The physical systems within these fieldsare often known to be (conditionally or unconditionally) resilientwith respect to shocks, disturbances or catastrophies in theimmediate environment. In order to be good mathematical modelsof such situations the reaction-diffusion systems must havethe same resilient or asymptotic behaviour as that of the physicalsituation. Three fundamentally different kinds of reaction termsare usually distinguished according to the entry signs of thereaction Jacobian: mutualism, mixed (predator-prey) interactionand competition. The asymptotic stability (in the Poincarésense) of mutualistic systems has already been studied extensively,but the results cannot be generalized (globally) to the othertwo fundamental types, which are not order-preserving. A partial(local) generalization is, however given here for these twotypes, involving simple Jacobian inequalities and knowledge(often prompted by the underlying physical situation) of invariantsets in solution space. The return time of resilient systemsand the approach rate of asymptotically stable solutions arealso estimated.