Pearcy et al Oncorhynchus cisrki cl3rki and O mykiss off Oregon and Washington 



705 



return to spawn in Sand Creek, OR, and Prairie Creek, 

 CA, had no ocean growth on their scales (Sumner 1972, 

 DeWitt 1954). Tomasson (1978) thought that coastal 

 cutthroat did not migrate in large numbers beyond the 

 Rogue River estuary. 



Because coastal cutthroat only spend several months 

 in the ocean, it is assumed that they inhabit waters 

 close to the coast and not far from their homestream 

 (Giger 1972). Johnston (1982) observed that anadro- 

 mous cutthroat usually fr-equented waters less than 3 

 meters in depth in Puget Sound and that migration of 

 tagged wild cutthroat did not extend much beyond 50 

 km from the home stream. However, cutthroat were 

 caught by Loch and Miller (1988) up to 31.5 km off- 

 shore and within the Columbia River plume between 

 Tillamook Bay, OR, and Willapa Bay, WA, suggesting 

 more extensive offshore movements. Sumner (1972) 

 also thought that cutthroat originating from the Co- 

 lumbia River basin may traverse along the coast within 

 the Columbia River plume. Our data also suggest that 

 some cutthroat undertake substantial movements along 

 the coast (>250 km) and are found as far offshore 

 (37-46 km) as the more oceanic steelhead smolts dur- 

 ing the summer. One fish was caught 66 km offshore. 

 Thus some individuals may reside far offshore along 

 an open coast, as opposed to staying close to shore as 

 they presumably do in protected inlets like Puget 

 Sound (Johnston 1982). Straying of returning sea-run 

 hatchery cutthroat trout is connnon (Bulkley 196(i, 

 Giger 1972, Jones 1977). Giger (1972) reported stray- 

 ing of marked hatchery cutthroat between the Nes- 

 tucca, Siuslaw, and Alsea Rivei's of the central Oregon 

 coast. Most of the strays were recovered in rivers south 

 of the river of release, possibly the result of advection 

 by coastal currents during the early summer period of 

 ocean residence, similar to that seen for eoho salmon 

 (Pearcy and Fisher 1988). 



Giger (1972) noted a non-random distribution in the 

 number of cutthroat smolts and kelts in seine samples 

 in the Alsea estuary during the spring and "large 

 schools" of sea-run fish in estuaries in the fall. He 

 thought that anadromous cutthroat trout formed 

 schools while at sea. However, we found no evidence 

 of schooling from our limited ocean catches. Most cut- 

 throat were caught singly. To our knowledge, school- 

 ing of cutthroat or steelhead trout in the ocean has not 

 been documented. 



Steelhead migrate long distances into oceanic waters 

 and are widely distributed in the North Pacific Ocean 

 based on catches of marked and unmarked fish 

 (Sutherland 1973; Pearcy and Masuda 1982, 1987; 

 Okazaki 1983; Hartt and Dell 1986; Light et al. 1988). 

 Miller et al. (1983) found that juvenile steelhead caught 

 in purse seines between Tillamook Bay and Willapa Bay 

 occurred farther offshore than juvenile coho or chinook 



salmon and migrated out of the coastal sampling area 

 early in the summer. Purse seining and tagging studies 

 by Hartt and Dell (1986) from Cape Flattery, WA, to 

 the Aleutian Islands clearly showed that juvenile steel- 

 head migrate directly offshore rather than along a 

 coastal belt where other juvenile salmonids typically 

 migrate. Recovery of steelhead during their first sum- 

 mer in the ocean in the Gulf of Alaska confirms rapid 

 migrations of some fish into oceanic waters far from 

 land. Pearcy and Masuda (1982) report on a steelhead 

 captured over 1600 km from land only a few months 

 after its initial ocean entry. 



These conclusions of immediate migrations of steel- 

 head offshore and into subarctic waters of the North 

 Pacific after ocean entry, and residence in oceanic 

 waters during their first winter in the ocean, do not 

 apply to all steelhead. Most steelhead from the Rogue 

 River in Oregon (the "half-pounder" runs) return to 

 freshwater after only a few months following their ini- 

 tial migration to the ocean (Everest 1973), and likely 

 do not migrate very far in the ocean. The recovery of 

 marked Rogue River summer steelhead south of the 

 Rogue River (Everest 1973), and the rarity of marked 

 steelhead originating from streams south of Cape 

 Blanco in waters to the north of Cape Blanco in our 

 catches, suggest that tliese steelhead from the southern 

 extremity of their range may not migrate to the north 

 after ocean entry. This conclusion is supported by the 

 high seas distribution of tagged steelhead. Although 

 9 tagged steelhead from California and 11 from Oregon 

 streams have been recovered north of 45 °N in the 

 noi'theastern Pacific, only 3 fish with coded-wire tags 

 from California have been recovered at sea of the over 

 1 million coded-wire tagged steelhead released between 

 1980 and 1985, and only one was caught north of 

 California (Light et al. i988. Pacific States Marine 

 Fisheries Comm. unpubl.). Possibly southern steelhead 

 reside in the strong upwelling zone off northern Califor- 

 nia and southern Oregon (Bakun 1975). 



The feeding habits of juvenile steelhead and cutthroat 

 trout in estuaries and in coastal waters are similar. 

 Both species feed intensively on gammarid amphipods 

 and insects in estuaries on their initial migration to the 

 sea, but steelhead smolts also eat the benthic mollusk 

 Corbicula (Loch 1982, McCabe et al. 1983, Bottom et 

 al. 1984). During their early residence in waters along 

 tlie open coast, cutthroat trout feed primarily on fishes 

 (Armstrong 1971, Fresh et al. 1981, Brodeur et al. 

 1987a, Loch and Miller 1988, this paper). We found that 

 steelhead trout also consumed fish but had a more 

 varied diet than cutthroat trout. Euphausiids and other 

 crustaceans were important in the diet of steelhead 

 trout, especially during the strong upwelling years of 

 our study when euphausiids may have been abundant 

 (Brodeur 1986). Many prey species identified in the 



