Michael M. Bailey Christopher M. Holbrook Michael T. Kinnison Joseph D. Zydlewski Daniel S. Stich 2015 <p><span>We developed a hierarchical multistate model to estimate survival of Atlantic salmon (</span><i>Salmo salar</i><span>) smolts in the Penobscot River, USA, over a decade during which two mainstem dams were removed from the catchment. We investigated effects of (</span><i>i</i><span>) environmental factors, (</span><i>ii</i><span>) rearing history, and (</span><i>iii</i><span>) management actions, including dam removal, turbine shutdown, and installation of new powerhouses. Mean ± SD smolt survival per kilometre was higher through free-flowing reaches of the catchment (0.995 ± 0.004·km</span>⁻¹<span>) than through reaches containing dams that remain in the system (0.970 ± 0.019·km</span>⁻¹<span>). We observed maximum survival between 12 and 17 °C and at intermediate discharges (1200 m</span>³<span>·s</span>⁻¹<span>). Smolt survival increased concurrent with dam removal and decreased following increases in hydropower generation. The greatest increase in smolt survival followed seasonal turbine shutdowns at a dam located on the largest tributary to the Penobscot River, while other shutdowns had little influence. Our model provides a useful tool for assessing changes to survival of migratory species and will be useful for informing stocking plans to maximize numbers of smolts leaving coastal systems.</span></p> application/pdf 10.1139/cjfas-2014-0573 en NRC Research Press Catchment-wide survival of wild- and hatchery-reared Atlantic salmon smolts in a changing system article