Warming and top predator loss drive direct and indirect effects on multiple trophic groups within and across ecosystems.

The interspecific interactions within and between adjacent ecosystems strongly depend on the changes in their abiotic and biotic components. However, little is known about how climate change and biodiversity loss in a specific ecosystem can impact the multiple trophic interactions of different biological groups within and across ecosystems. We used natural microecosystems (tank-bromeliads) as a model system to investigate the main and interactive effects of aquatic warming and aquatic top predator loss (i.e. trophic downgrading) on trophic relationships in three integrated food web compartments: (a) aquatic micro-organisms, (b) aquatic macro-organisms and (c) terrestrial predators (i.e. via cross-ecosystem effects). The aquatic top predator loss substantially impacted the three food web compartments. In the aquatic macrofauna compartment, trophic downgrading increased the filter feeder richness and abundance directly and indirectly via an increase in detritivore richness, likely through a facilitative interaction. For the microbiota compartment, aquatic top predator loss had a negative effect on algae richness, probably via decreasing the input of nutrients from predator biological activities. Furthermore, the more active terrestrial predators responded more to aquatic top predator loss, via an increase in some components of aquatic macrofauna, than more stationary terrestrial predators. The aquatic trophic downgrading indirectly altered the richness and abundance of cursorial terrestrial predators, but these effects had different direction according to the aquatic functional group, filter feeder or other detritivores. The web-building predators were indirectly affected by aquatic trophic downgrading due to increased filter feeder richness. Aquatic warming did not affect the aquatic micro- or macro-organisms but did positively affect the abundance of web-building terrestrial predators. These results allow us to raise a predictive framework of how different anthropogenic changes predicted for the next decades, such as aquatic warming and top predator loss, could differentially affect multiple biological groups through interactions within and across ecosystems.