Cold water woodland streams, where terrestrially derived organic matter fuels aquatic food webs, can be affected by increases in atmospheric CO 2 concentrations, as these are predicted to lead to increases in water temperature and decreases in organic matter quality. In fact, elevated CO 2 (580 ppm) decreased the initial phosphorus concentration of birch litter by 30% compared with litter grown under ambient conditions (380 ppm). Here, we first assessed the effect of differences in litter quality on mass loss, microbial colonization and conditioned litter quality after submersion in a mountain stream for 2 weeks. Leaching did not change the relative differences between litter types, while fungal biomass was two fold higher in elevated litter. We then offered this litter (conditioned ambient and elevated) to a stream detritivore that was kept at 10 and 15 1C to assess the individual and interactive effects of increased temperature and decreased litter quality on invertebrate performance. When given a choice, the detritivore preferred elevated litter, but only at 10 1C. When fed litter types singularly, there was no effect of litter quality on consumption rates; however, the effect of temperature depended on individual size and time of collection. Growth rates were higher in individuals fed ambient litter at 10 1C when compared with individuals fed elevated litter at 15 1C. Mortality did not differ between litter types, but was higher at 15 1C than at 10 1C. Increases in temperature led to alterations in the individual body elemental composition and interacted with litter type. The performance of the detritivore was therefore more affected by increases in temperature than by small decreases in litter quality. However, it seems conceivable that in a future global warming scenario the simultaneous increases in water temperature and decreases in litter quality might affect detritivores performance more than predicted from the effects of both factors considered individually.

No abstract

International audience1. Detritivores preferentially consume certain aquatic hyphomycete species while rejecting others. Fungal identity may therefore be a crucial factor determining stream food-web structure and complexity and extend the impact of microbial diversity to effects up through the food web. 2. In this study, we examined if shredder feeding is affected by the identity of fungi on leaves (bottom-up effects) and if preferences of shredders for particular fungi affect the composition of fungal assemblages (top-down effects). Oak leaf discs were conditioned in microcosms with six individual fungal species previously reported as highly palatable (P), unpalatable/rejected (R) and intermediate (I). Additionally, three microcosms were inoculated with three mixtures of four fungal species, each consisting of a different subset of the six species. 3. Colonised discs were offered to three detritivores with different feeding strategies: Proasellus sp. (Isopoda), Echinogammarus meridionalis (Amphipoda) and Schizopelex festiva (Trichoptera). When offered leaves colonised by single fungal species, consumption rates by E. meridionalis and S. festiva were higher on I, followed by P and, finally, R species. Consumption rates by Proasellus sp. were sim- ilar across fungal treatments. Consumption rates by the three invertebrates were also similar across all fungal multispecies treatments, suggesting that invertebrate preferences for, or rejection of, a given fungal species may be masked when it grows in proximity to other fungi. Composition and structure of fungal communities were not significantly affected by the feeding of any of the three invertebrates. 4. Our results suggest that certain combinations of fungal and detritivore species result in unpredictable bottom-up and top-down effects in stream food webs

We examine the relative importance of substrate quality and temperature in the establishment of aquatic hyphomycete assemblages and in their ability to decompose leaves. We used leaves of alder (Alnus glutinosa) and oak (Quercus robur) and we tested four temperatures (5 , 10 , 15 and 20 C). Differences in decomposition rates and fungal assemblages were higher substrata than across temperatures. In both species, decomposition efficiency measured as the ratio of decay rate to fungal biomass, was greater at higher temperatures. Oak leaves were colonized by fewer aquatic hyphomycete species than was alder. Decomposition rates of oak increased with temperature but that of alder was not affected. We conclude that the substratum is a key driver of aquatic hyphomycete assemblages and can attenuate the effects of temperature differences on litter decomposition.

No abstract