Modeling Atlantic salmon (Salmo salar) and brown trout (S. trutta) population responses and interactions under increased minimum flow in a regulated river

Abstract

Describing and understanding the relationship between streamflow and ecological processes is a classic problem in stream ecology and river management. We applied the individual-based model inSTREAM to describe the relationship between the dynamic river habitat and emergent population responses in sympatric landlocked Atlantic salmon (Salmo salar) and lake-migrating brown trout (Salmo trutta). This application explicitly describes the environmental conditions in the Gullspång Rapids, a residual flow stretch in the hydropower-regulated Gullspång River, Sweden (Svenskt elfiskeregister – SERS, 2019) etween September 2008 and September 2018. We simulated three static minimum flow scenarios and three variable natural flow regimes, contrasting highly artificial conditions with more natural dynamics. Our main response variable was the number of large (≥ 12 cm) out-migrants of salmon and trout, a proxy for successful population recruitment. The baseline model predicted an average production of 455 salmon and 532 trout out-migrants per year during 2008–2018 in this 11,700 m2 spawning and rearing area. The only alternative scenario producing more out-migrants was when the minimum flow was raised by a factor of three, as this led to a modest increase in trout out-migrants. Interestingly, none of the flow alternatives produced more salmon out-migrants than the baseline model, suggesting a competitive disadvantage originating from spawning later than trout. Density-dependent population regulation, a well-known phenomenon in salmonids, was reproduced by the model. Both intra- and interspecific competition was evident. While the number of out-migrants varied with flow regime, sensitivity analyses showed that other model input, specifically velocity shelter availability and stream temperature, were just as important. Increased availability of velocity shelters (in-stream structures that reduces the swimming speed of drift-feeding fish) was the only environmental factor that increased production of both salmon and trout in silico. We conclude that in this system, flow restoration based on simplistic flow scenarios will have limited effect, unless complemented by an increase of instream structural complexity.