Recognizing cross-ecosystem responses to changing temperatures: soil warming impacts pelagic food webs

The energy and materials that move across ecosystem boundaries influence food web structure and key ecosystem functions. Despite the acknowledged importance of such ecological subsidies, surprisingly little information is available regarding the role of environmental temperature in influencing subsidy quality and the response of the recipient ecosystem. We evaluated the impacts of temperature-mediated changes in leaves from deciduous trees, an important subsidy from terrestrial to freshwater ecosystems, on both the producer-based and detritivore-based components of a pelagic pond food web in a field mesocosm experiment. We hypothesized that variation in leaf chemistry driven by increased soil temperature would alter both the quality of leaf subsidies and the pond response. We collected red maple Acer rubrum leaves from heated and ambient temperature plots from the long-term soil warming experiment at the Harvard Experimental Forest and added them to 167-l field mesocosms containing established plankton communities, creating ‘no leaf’, ‘ambient leaf’ and ‘heated leaf’ treatments during autumn 2012. We then monitored physical, chemical, and biological responses to treatments until the mesocosms froze six weeks later. Experimental soil warming altered the chemical composition of deciduous leaves, the physical and chemical environment of the aquatic ecosystems to which leaves were added, and the pelagic pond food webs as measured by community composition. Compared to leaves from ambient-temperature soils, leaves from warmed soils initially resulted in lower water column phosphorus and dissolved organic carbon, reducing bacterial densities. However, the diminished carbon and phosphorus resulting from soil warming also increased light availability that ultimately stimulated cladoceran zooplankton relative to ambient-temperature leaves. Our results suggest that changes in temperature can alter ecological subsidies in unanticipated ways, and suggest that accurately predicting the potential consequences of climate change will require conducting research across ecosystem boundaries.