Decreased marine survival is identified as a component driver of continued declines of Atlantic Salmon Salmo salar. However, estimates of marine mortality often incorporate loss incurred during estuary migration that may be mechanistically distinct from factors affecting marine mortality. We examined movements and survival of 941 smolts (141 wild and 800 hatchery‐reared fish) released in freshwater during passage through the Penobscot River estuary, Maine, from 2005 to 2013. We related trends in estuary arrival date, movement rate, and survival to fish characteristics, migratory history, and environmental conditions in the estuary. Fish that experienced the warmest thermal history arrived in the estuary 8 d earlier than those experiencing the coolest thermal history during development. Estuary arrival date was 10 d later for fish experiencing high flow than for fish experiencing low flow. Fish released furthest upstream arrived in the estuary 3 d later than those stocked further downstream but moved 0.5 km/h faster through the estuary. Temporally, movement rate and survival in the estuary both peaked in mid‐May. Spatially, movement rate and survival both decreased from freshwater to the ocean. Wild smolts arrived in the estuary later than hatchery fish, but we observed no change in movement rate or survival attributable to rearing history. Fish with the highest gill Na+, K+‐ATPase activity incurred 25% lower mortality through the estuary than fish with the lowest gill Na+, K+‐ATPase activity. Smolt survival decreased (by up to 40%) with the increasing number of dams passed (ranging from two to nine) during freshwater migration. These results underscore the importance of physiological preparedness on performance and the delayed, indirect effects of dams on survival of Atlantic Salmon smolts during estuary migration, ultimately affecting marine survival estimates. Received October 25, 2014; accepted January 8, 2015

Timing of ocean entry is critical for marine survival of both hatchery and wild Atlantic salmon (Salmo salar) smolts. Management practices and barriers to migration such as dams may constrain timing of smolt migrations resulting in suboptimal performance at saltwater entry. We modeled influences of stocking location, smolt development, and environmental conditions on (i) initiation of migration by hatchery-reared smolts and (ii) movement rate of hatchery- and wild-reared Atlantic salmon smolts in the Penobscot River, Maine, USA, from 2005 through 2014 using acoustic telemetry data. We also compared movement rates in free-flowing reaches with rates in reaches with hydropower dams and head ponds. We compared movement rates before and after (1) removal of two mainstem dams and (2) construction of new powerhouses. Initiation of movement by hatchery fish was influenced by smolt development, stocking location, and environmental conditions. Smolts with the greatest gill Na+, K+-ATPase (NKA) activity initiated migration 24 h sooner than fish with the lowest gill NKA activity. Fish with the greatest cumulative thermal experience initiated migration 5 days earlier than those with lowest cumulative thermal experience. Smolts released furthest from the ocean initiated migration earlier than those released downstream, but movement rate increased by fivefold closer to the ocean, indicating behavioral trade-offs between initiation and movement rate. Dams had a strong effect on movement rate. Movement rate increased from 2.8 to 5.4 km·h−1 in reaches where dams were removed, but decreased from 2.1 to 0.1 km·h−1 in reaches where new powerhouses were constructed. Movement rate varied throughout the migratory period and was inversely related to temperature. Fish moved slower at extreme high or low discharge. Responses in fish movement rates to dam removal indicate the potential scope of recovery for these activities.

This study evaluated Atlantic salmon Salmo salar smolt survival through the lower Penobscot River, Maine, U.S.A., and characterized relative differences in proportional use and survival through the main-stem of the river and an alternative migration route, the Stillwater Branch. The work was conducted prior to removal of two main-stem dams and operational changes in hydropower facilities in the Stillwater Branch. Survival and proportional use of migration routes in the lower Penobscot were estimated from multistate (MS) models based on 6 years of acoustic telemetry data from 1669 smolts and 2 years of radio-telemetry data from 190 fish. A small proportion (0·12, 95% c.i. = 0·06-0·25) of smolts used the Stillwater Branch, and mean survival through the two operational dams in this part of the river was relatively high (1·00 and 0·97). Survival at Milford Dam, the dam that will remain in the main-stem of the Penobscot River, was relatively low (0·91), whereas survival through two dams that were removed was relatively high (0·99 and 0·98). Smolt survival could decrease in the Stillwater Branch with the addition of two new powerhouses while continuing to meet fish passage standards. The effects of removing two dams in the main-stem are expected to be negligible for smolt survival based on high survival observed from 2005 to 2012 at those locations. Survival through Milford Dam was been well below current regulatory standards, and thus improvement of passage at this location offers the best opportunity for improving overall smolt survival in the lower river.

Objectives for recovery of alosines commonly involve improving fish passage at dams during migration. However, a quantitative basis for dam passage performance standards is largely absent. We describe development of a stochastic life-history-based simulation model for American shad, Alosa sapidissima, to estimate effects of dam passage and migratory delay on abundance, spatial distribution of spawning adults, and demographic structuring in space and time. We used the Penobscot River, Maine, USA, as a case study to examine sensitivity of modeled population metrics and probability of achieving specific management goals to inputs. Spawner abundance and percentage of repeat spawners were most sensitive to survival and migration delay at dams, marine survival, and temperature cues for migratory events. Recovery objectives related to abundance and spatial distribution of spawners were achievable under multiple scenarios, but high rates of upstream and downstream passage were necessary. The simulation indicated trade-offs between upstream and downstream passage efficacy whereby increased downstream passage was required to maintain or increase population abundance in conjunction with increased upstream passage. This model provides a quantitative support tool for managers to inform ecologically based decisions about a suite of management scenarios to facilitate recovery and sustainability of diadromous fish populations.