Disentangling dam impacts in river networks

• Dam density is proposed as a multi-scale indicator distinguishing local and network scale impact.
• Responses in macroinvertebrate and fish communities are strongest for functional trait metrics.
• Macroinvertebrate communities respond strongest to upstream dam density whereas for fish the local context prevails.
• Invertebrate assembly and responses reflect metacommunity structure.
• Upstream–downstream contexts and responses differ and demand for different restoration strategies.

Damming is one of the most widespread impairments of river systems around the world. The formulation of scientifically justified guidelines for restoration and remediation of impairments requires better understanding of the relationships between alteration and stream condition. The present study examines relationships between the presence and density of dams and biological metrics of river health in the context of a variety of environmental drivers over the Loire river basin. We hypothesized that dam density measured at supra-reach level would show more significant influence on river health than the local level density, and further that the impact of dams is best estimated with measures for the functional traits of biotic assemblages. An extensive dataset of fish (169 sites) and invertebrate (211 sites) communities in the Loire river basin, described with metrics of density of ecological guilds, taxonomic richness and life history traits, and coupled with reach hydromorphology and catchment characteristics was constructed. Generalized linear modeling was performed in order to quantify dam impact and investigate the importance of regional- and local-scale measures of dam density to the structure of biotic communities. The analysis showed that community structure at the basin scale responded significant to dam presence and confirmed that the strongest relationships were observed for specific functional trait-based metrics. For the macroinvertebrates the observed impact counts up to 25% of the variance in the trait-based quality indices, whereas for fish communities the dam density only explains up to 12%. Macroinvertebrate responses were stronger at higher scale level, and especially the upstream context explained on its own 70% of the observed impairment. For fish communities, the local context prevails and explained up to 70% of the dam impact. These results can be explained by the biotic processes ruling community assembly in the specific groups, passive dispersal for the invertebrates and migrations between habitats for fish. The geographic context furthermore explains the differentiation in these responses, reflecting the metacommunity structure of invertebrate assembly over the river basin. We conclude that for upstream parts of the river basin, locally based management actions can be successful in restoring biotic integrity, whereas more downstream, dam removal actions require more integrated measures at regional rather than local scale.