Abstract--There is much interannual variation in numbers of first-year Sebastes that settle near shore off northern California, and when more abundant, they become prey of an increased variety of predators. Among predators that concentrate on these abundances are the black rockfish (Sebastes melanops); the blue rockfish (Sebastes mystinus); and the kelp greenling (Hexagrammos decagrammus). One might expect S. melanops to be involved in these attacks because it is known to be piscivorous, but S. mystinus and H. decagrammus ordinarily feed on invertebrates. The predation on Sebastes juveniles is concentrated during a relatively brief period shortly after they have settled in nearshore habitats. Most of this predation occurs during June and decreases sharply through July and August, and relatively little after that. We argue that the pattern of predation reflects removal of less adapted individuals during the period shortly after settlement.

First-year Sebastes spp. off California vary greatly in number from year to year (Stephens et al., 1984; Love, et al., 1991), as do young-of-the-year of many marine fishes (Hjort, 1926; Cushing, 1973; Ricker, 1954). It has been argued that this variability between years is set during the earliest stages of life history, so that numbers present at settlement can be used to predict ultimate strength (or weakness) of year classes (Bailey and Spring, 1992; Myers and Cadigan, 1993; Ralston and Howard, 1995). But such predictions may need to consider postsettlement mortality, which has been recognized as a density-dependent force that reduces the interannual variation in year-class size evident at settlement (e.g. Sissenwine, 1984).

Early evidence that young-of-the-year (YOY) suffer density-dependent mortality after settlement came from studies of fishes caught by trawlers in temperate seas. Some of these studies identified predation as the main cause of this mortality (e.g. Lockwood, 1980), but most recognized predation as just one of several possibilities (Veer, 1986; Myers and Cadigan, 1993). More recently, investigators using scuba to observe underwater have studied postsettlement juvenile mortality in a variety of tropical and temperate species. Although some have concluded that this mortality is not density dependent (e.g. Victor, 1986; Sale and Ferrell, 1988), others have concluded not only that it is density dependent but also that it acts to regulate populations (Sano, 1997) and is a result of predation (Hixon and Carr, 1997; Steele, 1997). That predation is the major cause of mortality during and after settlement now seems generally accepted, and attempts are being made to identify the mechanisms involved under controlled conditions in laboratories (e.g. Bertram and Leggett, 1994; Witting and Able, 1995).

Although there have been many studies of the relations between predation and recruitment, all have been of short duration and thus limited as examinations of interannual variations. An opportunity to draw on 11 years of data for such a study developed from our work with marine communities off northern California. There, an investigation of trophic relations in fishes overlapped annual assessments of YOY Sebastes spp., and during this period (1977 to 1987) certain predators attacked the recently settled Sebastes only during years when...