Juvenile Pacific salmon (Oncorhynchus spp.) and associated biophysical data were collected along a primary marine migration corridor in the northern region of southeastern Alaska. Thirteen stations were sampled over six time periods (32 sampling days) from May to August 2003. This survey marks the seventh consecutive year of systematic monitoring, and was implemented to identify the relationships among biophysical parameters that influence the habitat use, marine growth, predation, stock interactions, year-class strength, and ocean carrying capacity of juvenile salmon. Habitats sampled included stations in inshore (Auke Bay), strait (four stations each in Chatham Strait and Icy Strait), and coastal (four stations off Icy Point) localities. At each station, fish, zooplankton, surface water samples, and physical profile data were collected using a surface rope trawl (fish), conical and bongo nets (zooplankton), and a conductivity-temperature-depth profiler (physical data), usually during daylight. Surface (2-m) temperatures and salinities ranged from 7.6 to 15.8ΕC and 15.5 to 32.0 PSU from May to August. A total of 10,724 fish and squid, representing 23 taxa, were captured in 64 rope trawl hauls from June to August. Juvenile salmon comprised 29% of the total catch and occurred frequently in the trawl hauls, with chum (O. keta) occurring in 66% of the trawls, pink (O. gorbuscha) in 56%, coho (O. kisutch) in 55%, sockeye (O. nerka) in 50%, and chinook salmon
(O. tshawytscha) in 2%. Of the 3,254 salmonids caught, 98% were juveniles. Walleye pollock (Theragra chalcogramma) and Pacific herring (Clupea pallasi) were the only non-salmonid species that comprised more than 1% of the total catch. Temporal and spatial differences were observed in the catch rates, size, condition, and stock of origin of juvenile salmon species. Catch rates of juvenile salmon were highest for chinook and sockeye salmon in June, highest for chum and pink salmon in July, and highest for coho salmon in August. By habitat type, juvenile salmon catch rates for pink, chum and sockeye were highest in the coastal habitat, whereas catch rates of coho and chinook were highest in the strait habitat. Size of juvenile salmon increased steadily throughout the season; mean fork lengths in June and early August were, respectively: 105 and 133 mm for pink, 116 and 138 mm for chum, 120 and 145 mm for sockeye, 173 and 215 mm for coho, and 169 mm (June only) for chinook salmon. Coded-wire tags were recovered from two juvenile coho and one immature chinook salmon; all were from hatchery and wild stocks of southeastern Alaska origin. Alaska hatchery stocks were also identified by thermal otolith marks from 32% of the chum, 45% of the sockeye, 11% of the coho, and 100% of the chinook salmon. Onboard stomach analysis of 248 potential predators, representing 10 species, indicated one predation instance on juvenile salmon by a spiny dogfish (Squalus acanthias) in the coastal habitat in July. This research suggests that in southeastern Alaska, juvenile salmon exhibit seasonal patterns of habitat use synchronous with environmental change, and display species- and stock-dependent migration patterns. Long-term monitoring of key stocks of juvenile salmon, on both intra- and interannual bases, will enable researchers to understand how growth, abundance, and ecological interactions affect year-class strength and ocean carrying capacity for salmon.