Volitional Migration of Atlantic Salmon from Seasonal Holding Ponds

For 5 years we used seawater tolerance tests (seawater challenges) to identify smolts, and tunnel fish counters to record the time of migration, of Atlantic salmon Salmo salar that volitionally left holding ponds on the Merrimack River near Litchfield, New Hampshire. We compared timing of Atlantic salmon movements with environmental conditions and blood composition of migrating and nonmigrating fish. None of the pan from the Nashua National Fish Hatchery (NFH) tested in seawater in 1984–1986 and from the North Attleboro NFH tested in 1987–1988 smoltified in the ponds. Yet, smolts were seemingly produced because 50 of 135 Atlantic salmon with radio transmitters from these ponds were traced to the ocean, and 9 of the 20,680 Atlantic salmon marked in 1984 from Litchfield subsequently returned to spawn. In all years except 1987, we forced Atlantic salmon out of the ponds by mid‐May because of high water temperatures. In 1987, all Atlantic salmon from the North Attleboro NFH left the smolt‐holding ponds by April 10. Only 25% of these fish that were caught in the river passed a seawater tolerance test. In 1988, seawater tolerance tests revealed that salmon held at the North Attleboro NFH smoltified in May after fish in the Litchfield ponds had either been forced out or emigrated. Rearing at a constant temperature in the hatchery may have delayed the time of smoltification. Peak migrations were associated with periods of sustained rising temperatures, river flow, and turbidity. Volitional migration occurred before smoltification was detectable and may be the best predictor of smoltification's early stages.