Effect of rainfall interannual variability on the stability and resilience of dryland plant ecosystems

Positive feedbacks between vegetation and soil moisture may induce, in arid ecosystems, the emergence of two alternative stable states, corresponding to bare and completely vegetated soil. A new analytical model is developed to investigate this behavior and to study the effects of interannual rainfall variability on the dynamics of bistable ecosystems. When dichotomic Markov noise is used to account for the effect of random rainfall fluctuations, the long‐term dynamics of the system can be investigated through an analytical solution of the model. It is found that in a broad class of bistable ecosystems, random rainfall fluctuations may induce an ordered state in the dynamics, i.e., by turning the bistable deterministic system into a stochastic system with only one statistically stable state. This effect is enhanced by increases in noise intensity, whereas the stochastic dynamics become bistable (i.e., the noise‐induced ordered state disappears) as the noise intensity decreases below a critical (nonnull) value. This effect of noise‐induced stability is found in association with an enhancement of ecosystem resilience, indicating that the likelihood of catastrophic shifts to the desert state decreases as the noise intensity increases.