Short-term effects of instream habitat restoration on macroinvertebrates and a comparison of sampling approaches

Many streams and rivers worldwide are restored with the intention to mitigate degradation caused by human activities, but these rehabilitation projects often involve physical instream work with diggers or other heavy machinery. The short-term effects of such restoration works on stream ecosystems are rarely investigated. The primary aim of our study was to assess the short-term effects of physical instream restoration works (which likely increased instream fine sediment load temporarily) on the benthic macroinvertebrate community in a lowland soft-bottomed stream, and our second aim was to compare the results of semi-quantitative kick-net and quantitative Surber sampling in this assessment. Invertebrates were collected using these two methods from three Control and three Impact sites, before and two days after recontouring the stream banks and installing instream woody cover features. Three of seven macroinvertebrate community-level indexes suggested positive short-term effects of the restoration works, whereas none indicated negative effects. By contrast, seven of the 14 most common taxa were negatively affected by the restoration works, possibly due to increased fine sediment levels reducing habitat quality for these taxa. These differences imply that taxon-specific invertebrate responses may be more suitable for detecting short-term impacts of instream restoration works than community-level metrics. Community indexes and common taxa were also likely affected by a drought in the spring and summer before restoration, which may have obscured some effects of the restoration works. Finally, kick-net and Surber sampling methods yielded similar findings for the community-level metrics but differed more for the common taxa, probably because the two methods sampled different microhabitats within the stream. Consequently, we recommend that for routine monitoring of macroinvertebrate communities in lowland soft-bottomed streams, the semi-quantitative kick-net sampling method should be used.