Résumé
Current global change is associated with an increase in disturbance
frequency and intensity, with the potential to trigger population
collapses and to cause permanent transitions to new ecosystem states.
However, our understanding of ecosystem responses to disturbances is still
incomplete. Specifically, there is a mismatch between the diversity of
disturbance regimes experienced by ecosystems and the one-dimensional
description of disturbances used in most studies on ecological stability.
To fill this gap, we conducted a full factorial experiment on microbial
communities, where we varied the frequency and intensity of disturbances
affecting species mortality, resulting in twenty different disturbance
regimes. We explored the direct and long-term effects of
these disturbance regimes on community biomass. While most communities
were able to recover biomass and composition states similar to undisturbed
controls after a halt of the disturbances, we identified some disturbance
thresholds that had long-lasting legacies on communities. Using a
model based on logistic growth, we identified qualitatively the
sets of disturbance frequency and intensity that had equivalent
long-term negative impacts on experimental communities.
Our results show that an increase in disturbance intensity is a bigger
threat for biodiversity and biomass recovery than the occurrence of more
frequent but less intense disturbances.