Does soil bioengineering benefits aquatic biodiversity? An empirical study of the relative influence of local and regional drivers on benthic macroinvertebrate assemblages

Abstract

Riverine ecosystems form a dendritic network in which landscape and catchment-scale properties influence freshwater community structure. Placed in a restoration framework, this suggests that regional drivers can overrule the benefit of measures aiming at improving local habitat quality. Disentangling the relative influence of local and regional drivers on freshwater communities is thus crucial for ecosystem management and restoration. Along riverbanks, soil bioengineering is often used to both control erosion and improve ecological conditions. Soil bioengineering techniques aim at copying naturally functioning riverbank models and can thus be viewed as riparian ecosystem restoration. Nevertheless, these techniques are mostly designed at the local scale and are implemented in a broad range of rivers. This implies large variations in regional drivers, which may greatly influence the response of freshwater communities to restoration efforts. We studied 37 riverbanks, from civil engineering to soil bioengineering, plus natural willow stands, in the foothills of the Alps and Jura Mountains, and assessed the relative influence of local (terrestrial and aquatic habitat conditions) and regional (water quality, hydrological context and land cover composition) drivers on benthic macroinvertebrate assemblages. Using multivariate GLM and structural equation modelling, we investigated variations in the taxonomic and functional composition and in the diversity of native, exotic, shredder and scraper taxa to both set of drivers. Our results showed that soil bioengineering improved local habitat conditions, with an increase in the vegetation density and in the aquatic habitat quality. These changes directly influenced functional composition but indirectly diversity patterns. Instead, we found that native and shredder species richness increased between civil engineered and soil bioengineered structures, suggesting a positive effect of vegetated riverbanks on other local abiotic factors (i.e. shade, water temperature, organic matter supply). Our results also showed that macroinvertebrates were more influenced by regional than by local drivers. Thus, the hydrological context best explained the composition of taxa feeding habits and variation in taxa diversity, with larger abundance and richness of scrapers and shredders in small headwater streams. Land cover ranked second in explaining variation in functional composition. Also, the diversity of natives, scrapers and shredders increased as the proportion of predominantly urban landscapes decreased. Finally, the abundance of scraper and native species increased with water quality, while the richness of exotic species decreased. Overall, these results highlight the hierarchical structure of local and regional drivers on freshwater communities. Along river networks, catchment-scale properties and landscape attributes are major drivers of macroinvertebrate assemblages. Soil bioengineering improves habitat quality and as such appears to be a good compromise solution to control erosion and support freshwater communities, even though this nature-based solution cannot solve anthropogenic pressures at larger scales. To improve the efficiency of restoration efforts, integrated approaches accounting for both local and regional drivers remains a priority.