Search

International audience; Soil erosion in freshwaters induces important changes in lake metabolism. The organic matter and the nutrients supplied by soil inputs can change internal biogeochemical cycles and subsidize the whole food web from basal organisms to top-predators. Since the last two decades, the role of allochthonous organic matter as a basal resource for aquatic food webs in natural and controlled conditions has received a growing attention. We studied the impact of soil on the functioning of pond ecosystems by performing monthly additions of soil in freshwater mesocosms. In addition, the food-web structure was manipulated by addition of omnivorous fish to study interactions between the bottom-up effect of soil addition and the top-down effect of fish. The effects of soil and fish addition on the, the elemental and the biochemical compositions of pelagic compartments and recent sediment, on the biomass of seston and zooplankton and on the sediment rates were studied. Soil inputs had no effect on biomass, stoichiometry and lipid composition of seston and zooplankton but fish growth was enhanced by soil addition. Soil treatment had several (but idiosyncratic) effects on the stoichiometry and on the lipid composition of recent sediment. However, the sedimentation rates and the potential biodegradability of recent sediment were not affected by soil inputs. Fish addition affected chlorophyll-a concentration of the water column, seston biomass, sedimentation rates and stoichiometry of seston, zooplankton and short-term sediment. The lipid composition of recent sediment was also influenced by fish addition. However, fish addition did not change the biodegradability of recent sediment. Finally, we did not observe any significant interaction between soil and fish treatments. Our results suggest that the addition of soil as allochthonous inputs to aquatic ecosystem induced a subsidize of the food web only on fish, probably due to direct foraging on bottom sediment. Nevertheless, this process does not induce an effect on the biodegradability of recent sediments. Therefore, allochthonous inputs appear to have little effects on nutrient cycling. The predominance of the effects of food-web structure over allochthonous inputs and the absence of interaction between these top-down and bottom-up controls suggest that the bottom-up control was minor compared to the top-down one exerted by top-predators.

1. Sediment plays a key role in internal nutrient cycling and eutrophication in lakes. However, studies focusing on the efficiency of the biomanipulation techniques for improving the control of primary producers have rarely examined the effects of changes in food-web structure on the sediment biochemical composition and biodegradability. 2. In a 1-year experiment conducted in large replicated mesocosms, we tested how the absence or presence of a zooplanktivorous fish (roach, Rutilus rutilus) affected the elemental composition and the potential biodegradability of recently deposited sediment in a eutrophic system. The potential biodegradability of these sediments was assessed in laboratory microcosms by measuring the production of CO2 during 44-day incubations. 3. The potential biodegradability of recently deposited sediment from the fish treatment was 60% higher than that from the fishless treatment. This higher biodegradability was corroborated by a higher annual loss of sediment in fish enclosures (36%) than in fishless ones (16%). Annual losses of carbon, nitrogen and organic phosphorous were higher for sediment from fish enclosures. 4. Carbon and nitrogen contents of sediment were higher for the fish treatment. In contrast, the sediment C/N ratio, one of the proxies used to estimate sediment biodegradability, did not differ between treatments. No relationship was observed between elemental composition of sediment and its potential biodegradability. This latter appeared to be more probably dependent on the biochemical composition of the sediment and especially on the content of labile compounds such as proteins, sugars and polyunsaturated fatty acids. The use of sterols as biomarkers revealed an important degradation by microorganisms of 1-year-old sediment from both fish and fishless treatments. 5. Our results revealed that fish biomanipulations might favour clear water states not only through a stronger topdown control on phytoplankton but also through a lower biodegradability of sediment reducing internal nutrient cycling.

International audience; 1. Declines in availability of plant resources to pollinators are a major cause of pollinator loss. The management of plant communities to enhance floral resources is often proposed as a way to sustain pollinator populations. Nectar, the main energetic resource for pollinators, plays a central role in behaviour and composition of pollinator communities. Abiotic and biotic factors are known to influence nectar traits at both the species and community levels, but the impact of plant community composition itself has never been investigated. 2. Below-ground interactions in plant communities can induce changes in plant development through (i) plant-derived litter amendment of the soil and (ii) competition for soil resources between plants. We tested how plant below-ground interactions affect above-ground nectar traits involved in plant attractiveness to pollinators. 3. A short-term pot experiment was carried out with three temperate grassland species Mimulus guttatus, Lamium amplexicaule, and Medicago sativa, showing distinct litter stoichiometry and competitive abilities for soil resources. Litter amendment (none, mono and tri-specific litter) and plant interaction treatments (monocultures, two-and three-species mixtures) were crossed in a factorial design. 4. Litter amendment to the soil led to an increase in total nectar sugar content in L. amplexicaule plants but not in the two other species. We also found that the presence of M. guttatus, a competitive species, reduced the total nectar sugar content in L. amplexicaule through a concomitant decrease in nectar volume per flower and in floral display size, but not in other species. Species-specific responses of nectar traits to variation in soil nitrogen availability were thus observed, suggesting consequences for plant species and community attractiveness to pollinators. However, we did not find evidence that the legume M. sativa affected nectar traits of any neighbouring plants. 5. Synthesis. Our results demonstrate that litter inputs and competition between plants for soil resources can alter nectar traits linked to plant attractiveness to pollinators. This supports the idea that below-ground plant-plant interactions for soil resources can influence above-ground plant-plant interactions for pollination services. This offers promising perspectives in studying the role of below-ground-above-ground interactions on higher trophic levels.

International audience