Characterizing predation risk during Atlantic salmon smolt migration

Salmon populations across the globe have incurred substantial declines over the last 50 years with many on the verge of extinction. Among these, the Gulf of Maine Distinct Population Segment of Atlantic salmon (Salmo salar) has declined by > 95%, and was subsequently listed under the Endangered Species Act in 2000. Once inhabiting rivers throughout New England, Maine is the last place in the United States where sea-run adults return to spawn each spring. The majority of these fish return to the Penobscot River, which has been the focus of intense restoration and research efforts over the last two decades. The smolt stage of the salmon life cycle, when salmon migrate from freshwater to the marine environment, is a period of low survival. During migration, smolts incur high mortality as they negotiate impoundments, salinity differences, and novel predators in route to marine entry. Since 2005, researchers from the National Oceanic and Atmospheric Administration, University of Maine, and U.S. Geological Survey have tagged over 5,000 smolts to better characterize the risks incurred by smolts that move through the Penobscot River. These studies have revealed areas of elevated mortality and movement patterns that may be indicative of predation. High levels of predation, especially if beyond natural levels, can be a limiting factor in population recovery. However, beyond anecdotal accounts and evidence from other systems, there remains a critical knowledge gap in the understanding of smolt predation risk in this system. In this study, we used novel methods to quantify predation rates, identify smolt predators, and describe relationships between predation risk and environmental, seasonal, and individual influences during smolt migration in the Penobscot River. We tagged smolts with acid-sensitive acoustic predation transmitters with temperature sensors (n=270) from 2021–2023 to quantify predation rates and identify predator taxa. Tagging studies were paired with the use of Predation Event Recorders to capture predation on tethered smolts (n=660) in effort to link acute environmental conditions with predator activity. Results from these studies were then used to inform retrospective assessments of smolt survival through the lower Penoboscot River from 2005–2023. First, we used known fates from predated and surviving smolts to develop a machine learning model to characterize annual variation in bird and mammal predation in the Penobscot River Estuary. Then we used detection data from over 3,600 acoustic-tagged smolts to describe survival through the Penobscot River Estuary, Penobscot Bay, and the greater Gulf of Maine. Our results indicate that smolts incur predation throughout migration. Through smolt tagging and tethering studies, we observed 290 predation events from 2021–2023. In the main-stem Penobscot River, we estimated that at least 46% of all smolts were predated before entering Penobscot Bay which contributed to at least 55% of all smolt mortality. Relative predation risk was greatest through the impoundments of hydropower dams and the lower estuary, where on average, predation rates were 5- and 9-fold greater than free-flowing reaches, respectively. In freshwater, predation on tethered smolts was dominated by smallmouth bass (Micropterus dolomieu, accounting for 71% of predated smolts), but we also documented chain pickerel (Esox niger, 27%) and brown bullhead (Ameiurus nebulosus, 2%) as potential smolt predators. We found that predator activity in these areas was primarily influenced by water temperature and river bathymetry. Temperature sensor-equipped acoustic transmitters indicated that species-specific risk shifted to marine mammals once smolts entered tidal waters, while avian predation was documented throughout the watershed. Time series assessments indicated that predation risk from marine mammals varies annually, and that bird predation in the estuary has decreased since 2008. Overall, smolt survival through the estuary was relatively high (91%), where we attributed among-individual variation in survival to differences in migration history, arrival time, and body condition. Once smolts leave the estuary, we identified the primary migration route and estimated survival (71%) through Penobscot Bay. From there, at least 34% of smolts reached the Scotian Shelf off of Halifax, Nova Scotia, Canada as they traveled to foraging grounds off West Greenland. This body of research represents as significant advancement in our understanding of smolt mortality risk for this and other salmonid populations. We demonstrate that multiple predator species have the potential to consume smolts in the Penobscot River. Collectively, these studies suggest that anthropogenic actions may increase predation risk for smolts and other migratory species by impounding waterways, warming rivers via climate change, and introducing non-native predators (e.g., smallmouth bass). However, our results indicate potential opportunities to mitigate smolt losses by facilitating safe movement through dams, managing predator populations, and adjusting stocking locations. We anticipate these findings will be used by salmon mangers at local, regional, and global levels in effort to conserve and restore endangered populations.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.