Dady increments in otoliths were used to estimate growth rates and hatch dates of young walleye pollock Therayra chalcogramma from the western Gulf of Alaska, USA. Growth rate increased through the first 70 d after hatchng and was descnbed by a nonlinear, Laird-Gompertz model. There were no interannual differences m growth rates of young larvae from the limited size range collected in 1983 and 1985 to 1987. Growth rates of older larvae and pelagic juveniles increased from southwest to northeast along the Alaskan Peninsula. Hatch dates were calculated from age of fish at time of collection; spawning dates were eshmated after accounting for temperature-dependent egg stage duration. From 1983 to 1987, spawning and hatch dates seem to have shlfted to later In the season. This can indicate an overall change m time of spawning or increased removal, through mortality and dispersal, of eggs and larvae spawned early in the season. Hatch dates were earlier in the southwest near Unimak Island and later in the season to the northeast around Kodiak Island. Mortality was estimated from the difference in abundance of each daily cohort between the May and June-July surveys in 1987 and was age dependent. Mean daily instantaneous mortality rate was 0.11 for young larvae and decreased to 0.02 for older larvae and pelagic l u v e d e s . From these rates, it 1s estimated that about 97 % of the larvae wlll die or be removed from the Shelikof Strait area within 35 d of hatching.

Visual surveys of demersal fishes and their associated habitats are being conducted regularly in deep water (i.e., 30-365 m) off Alaska and the West Coast of North America by numerous research groups using quantitative transect methods from the research submersible Delta. The use of Delta has been applied primarily to the characterization of fish habitats, with increasing applications to improve stock assessments, evaluate gear impacts, and identify new species. Using Delta is no longer an unproven concept, but rather an accepted survey tool as demonstrated in more than 85 peerreviewed publications since 1988. Maps of seafloor substrata and bathymetric data are commonly used to identify and quantify survey sites, which serves both to increase the costeffectiveness of the submersible and provide the frame within which to distribute sampling effort. In turn, direct observations from Delta can be used to validate the interpretation of such habitat maps. Values that are commonly measured or estimated during Delta dives include the number and length of organisms, distance to a fish, and distance along a transect, from which habitat-specific species density, total abundance, and biomass can be calculated. In this paper we review the research that has been conducted using Delta during the last twenty years, and consider some potential sources of methodological bias as well as the advantages and some strategies for conducting visual surveys using Delta.

The Christmas tree coral Antipathes dendrochristos is a recently discovered black coral species that represents a habitat associated with numerous sensitive taxa in the Southern California Bight. We evaluated whether broad-scale oceanographic features influence coral density and size by selecting from among generalized additive models (for density) and generalized linear models (for size) that represented competing hypotheses. We constructed models to predict coral density and size using depth, seafloor slope, surface primary productivity, bottom currents, ocean temperature, salinity, and dissolved oxygen as candidate covariates. Specifically, we evaluated hypothesized links between pelagic production and benthic coral utilization and between bottom currents and larval coral dispersal. Our analysis revealed that high surface primary productivity in combination with depth and January currents are important predictors of Christmas tree coral density. Higher coral density coincided with greater chlorophyll persistence and optimal depths near 400 m. Surface productivity increasingly was associated with Christmas tree corals at shallower depths. Our results supported the hypothesis that ocean currents affect coral density via larval dispersal mechanisms. The selected coral size models responded to similar covariates, corroborating coral density results. Fish and invertebrate ordinations indicated that Christmas tree corals were widely distributed across environmental gradients and that Christmas tree corals cooccurred with several demersal fish and invertebrates. Several predicted coral hotspots remain unprotected from fishing, particularly in areas adjacent to highly populated portions of the Southern California Bight. These regions should be targeted by future studies to confirm the presence of Christmas tree coral communities and to evaluate their vulnerability.