The Long‐Term Effects of Created Habitat on Native Salmonids in a Small Headwater Stream in the Boreal Forest

ABSTRACT Stream improvement projects are common across North America. The goal of these projects is usually increased fish production, but long‐term monitoring is rarely conducted to determine if these goals are met. We examined how a constructed pool‐riffle sequence influenced density and biomass of native salmonids (brook trout ( Salvelinus fontinalis ) and Atlantic salmon ( Salmo salar )) over more than two decades in a small headwater stream in Newfoundland and Labrador, Canada. Two of the pools contained artificial undercut bank structures (LUNKERS) that provide additional cover for brook trout. We used a before‐after‐control‐impact (BACI) study design to assess the influence of the constructed sites on stream habitat and fish density and biomass. Sampling occurred 2 years prior to construction (1988–1989), continued for the first 4 years after construction (1991–1994), and then periodically for 26 years (1999, 2007, 2015, 2016). Fish density and biomass were estimated using successive removal electrofishing. Brook trout populations fluctuated over 26 years, but age‐1+ biomass was consistently higher in the constructed habitat compared to control sites. The percentage of large brook trout, that is, with fork length greater than 150 mm, increased in the stream over time. Furthermore, constructed pools with LUNKERS had 86% and 177% higher mean density and biomass, respectively, of age‐1+ brook trout compared to constructed pools that did not. Age‐1+ Atlantic salmon biomass and density increased in the constructed reach while young‐of‐year Atlantic salmon decreased throughout the study area. In this small headwater stream with limited pool habitat, the addition of deep pools with cover significantly improved age‐1+ brook trout biomass and remained structurally stable over more than 20 years.