Stability and resilience trajectories of stream ecosystems after two major hurricanes

Understanding how natural ecosystems respond to disturbance events is a major goal in ecology, especially as extreme climatic events become more frequent with climate change. We evaluated the effects of hurricanes Irma and María on multiple aspects of stream ecosystem stability (i.e. resistance, engineering resilience, recovery and temporal stability) over 5 years post-disturbance. We fitted eight models to analyse recovery trajectories and used a Bayesian structural equation model to examine interactions among variables during recovery. Most variables exhibited low resistance, with recovery remaining incomplete and high variability during the return to pre-hurricane conditions. Canopy openness increased sevenfold, and although the initial disturbance caused a fivefold rise in leaf litter fall, subsequent sampling showed a significant drop in organic matter input, while epilithic algae slightly decreased. Decapod and macroinvertebrate responses varied across streams. Recovery trajectories differed by variable, including exponential (epilithic algae), logarithmic (leaf litter), inverted logistic (decapods), linear (canopy openness) and humped curves (macroinvertebrate density). Analyses of ecological interactions revealed differing associations and dynamic feedback among variables. This study underscores the profound, lasting effects of hurricanes on tropical streams and emphasizes the need for long-term monitoring to understand ecosystem responses to increasingly intense and frequent disturbances.