Serena Ceola
The temporal variability of streamflows is a key feature structuring and controlling ecological communities and ecosystem processes. The magnitude, frequency and predictability of streamflows, and thus of velocity and near-bed shear stress fields, control ecological structure and function, particularly of benthic organisms. River ecosystem dynamics is indeed deeply connected to streamflow variability, which chiefly relies on rainfall, climate, land use, and geomorphologic properties. Although alterations of streamflow regime due to climate change, habitat fragmentation or other anthropogenic factors are ubiquitous, their ecological implications remain poorly understood. The present Thesis therefore addresses a quantitative analysis of the implications of hydrological fluctuations on river ecosystems positing that they have major ecological importance. From a food-web perspective, the response of benthic biota to hydrologic change is analyzed through experimental and theoretical approaches. Starting from the theoretical characterization of the probability distribution function of streamflows, flow velocities, water depths and near-bed shear stresses at a site, where a comparison between measured and analytical streamflow statistics across several catchments is outlined, a flume experimental campaign has been carried out in order to analyze how flow variability affects biofilm growth and invertebrate grazing activity. Here, two contrasting flow regimes (a constant and a time-varying discharge sequence) and four different light conditions (from 90% to 27% of incoming light radiation), as a key control on biofilm algal productivity and grazer activity, have been performed. Average grazing rates were significantly enhanced under time-varying flow conditions and highest at intermediate light availability. This result suggests that the stochastic flow regime offers increased opportunities for grazing under more favorable shear stress conditions, with implications for stream ecology and trophic carbon transfer in stream food webs. A spatial generalization of the above results, upscaling them to whole river basins, is made possible by employing scaling relationships that are known to characterize the geometry and topology of natural landforms and riverine patterns. The proposed generalization accounts for hydrological characterizations of the driving streamflow variability where the explicit dependence of invertebrate suitability on geomorphic controls (e.g. river depth, bed shear stress and flow velocity) and on hydrologic conditions (e.g. embedded in the probability of streamflow, and the ensuing temporal correlations that define the persistence of the hydrologic signal in time) has been accounted for. Different invertebrate grazing species have been compared. Suitable dynamic models, describing benthic biota dynamics, in particular stream biofilm growth under both ungrazed and grazed conditions (that is, via the absence and presence of invertebrates, respectively), have been explored in order to assess the relevant processes that controlled biofilm-invertebrate temporal pattern. The present Thesis therefore aims to further our understanding of the linkages between hydrology and ecology, possibly leading to a comprehensive theory of hydrologic drivers and controls of biotic processes in stream benthic environments.
EPFL2012
