Dr. Thibaut Rota

1. Intraspecific variability has implications for our understanding of ecosystem functioning. Yet, beyond ontogenetic variation in body size, we know little about how phenotypic differences between individuals modulate ecosystem functioning.
2. We addressed this question empirically within a forested stream, by studying the ecosystem effects triggered by conspecific predatory dragonfly larvae that varied in sex, body shape, behaviour and growth rate. With the help of stream enclosures, we were able to evaluate how different predator phenotypes drive cascading effects on litter-dwelling prey, fungal biomass and leaf litter decomposition, in uncontrolled natural conditions.
3. Some individual predators increased leaf litter standing stocks and fungal biomass compared to controls (without dragonfly larvae added), but others had the opposite effect. These contrasting effects were repeatable between the two runs of our experiment. Phenotypic effects were related to the sex and the position of individuals along a life history—behavioural continuum, and were possibly mediated by differential top-down control of prey, or by inter-individual variability in non-consumptive effects.
4. Several phenotypic dimensions are likely to mediate individual variability in predator effects on ecosystem processes, even in a complex and stochastic natural context such as a forested stream.
5. We invite further investigations that seek to understand the mechanisms (e.g., non-consumptive effects) driving individual variability in predator effects on ecosystem processes.

Little is known of how Neotropical freshwater ecosystems will respond to future climate scenarios. In Neotropical rainforests, a substantial fraction of the freshwater available to the aquatic fauna is found within phytotelmata, plant-held waters that form aquatic islands in a terrestrial matrix. We hypothesized that phytotelmata in close proximity have higher resilience capacity to severe drought than the isolated ones, under the assumption that immigration from nearby sources promotes faster recovery. We used rainshelters to emulate an extreme drought (67 days without rainfall) in tank-forming bromeliads arranged in patches of 1, 3 or 6 plants in a primary forest of French Guiana. Habitat size was a stronger determinant of invertebrate species richness and biomass per bromeliad than patch size. Larger bromeliad patches attenuated the adverse effect of drought on the biomass of predators, probably because short-range migration within dense patches allowed individuals to find moist refuges. However, the recovery of aquatic communities and ecosystem functions was mostly supported by in situ resistance, and a rescue effect of immigration was weak. Whilst environmental management plans tend to focus on dense networks of connected water bodies, our study shows that efforts should not omit the isolated ones.