1. Degradation of stream habitat because of anthropogenic activities (e.g. channelisation) has had a dramatic impact on fluvial environments and their biota, and as a consequence, increasing effort has been directed towards the restoration of degraded rivers. However, a major problem is that the success (or failure) of restoration has been rarely tested using a well-designed monitoring programme to allow reliable detection of an impact, if any exists. We used a spatially and temporally replicated, balanced Before-After-Control-Impact design to assess the impact of stream habitat rehabilitation on the densities and growth of brown trout of three age-classes in North Finnish forest streams.
2. Three separate sections in each of six streams were selected for the study. After 3 years of pre-rehabilitation monitoring, two randomly selected sections in each stream were restored; one using large woody debris and boulders and the other using only boulders. A third section remained as an unmodified control. Monitoring of fish densities continued for 3 years after rehabilitation.
3. Rehabilitation clearly increased streambed complexity, but did not have detectable effects on brown trout stocks in either of the rehabilitation schemes (LWD or stones), except for age-2+ and older fish which decreased in abundance compared to control reaches. A severe drought after rehabilitation in late summer 2002 reduced densities of trout to a low level in all streams, overriding any local effects of rehabilitation. Rehabilitation structures seemed to provide some safeguard against drought for age-2 and older, but not for the younger age-classes.
4. Our results add to the growing body of literature suggesting that large-scale regional factors may overwhelm local management efforts. To be successful in the future, stream rehabilitation schemes must include drought refuge areas for fish and other stream biota.