Abstract

Little is known regarding the importance of early-life transport and dispersion mechanisms in determining recruitment variability for Pacific ocean perch (POP) in the Gulf of Alaska (GOA). These mechanisms influence the degree of, and variability in, connectivity between offshore natal areas (parturition sites) and inshore demersal nursery habitats for young-of-the year juveniles, and may thus play an important role in the "gauntlet" that individuals must survive from parturition to recruitment. As a first attempt to assess interannual variability in connectivity between natal and nursery areas for POP in the GOA in a synthetic manner, we developed a coupled biophysical individual-based model (IBM) for POP early life history and dispersal with simple representations of active vertical movement, passive horizontal movement, growth, and settlement in appropriate nursery habitat to integrate known early-life traits with variability in environmental forcing. We used an oceanographic model for the GOA based on the Regional Ocean Modeling system (ROMS) to provide the underlying daily physical environment to force the IBM for 1996-2011 and simulated hundreds of thousands of individual POP from parturition along the shelf break to settlement in inshore demersal nursery habitats as young-of-the-year. We used the IBM results to assess spatial patterns of annual "maximum potential" connectivity between presumed natal and nursery areas at along shore scales of similar to 150 km, as well as the interannual variability in these patterns. Results showed that, even in the absence of mortality, most ( > 70%) individuals were unsuccessful in dispersing from presumed natal areas along the continental shelf break to inshore nursery areas. For those that were successful, connectivity was directed in a counter clockwise fashion (southeast to northwest) around the GOA following prevailing current patterns. Typical dispersion distances were on the order of 100's of km along shore, much larger than those inferred from genetic sampling. Natal areas from which the highest fractions of successful individuals originated were in the southeast GOA, while the nursery areas most frequently reached by those successful individuals were in the central GOA. POP from natal areas in the western GOA were consistently exported from the system and likely contribute little to the GOA population, although they may contribute to populations in the Aleutian Islands and eastern Bering Sea. We also found that annual indices derived from the connectivity matrices were not very strongly related to any of a suite of basin- and regional-scale environmental indices, reflecting the overall complexity and scale of the pathways POP in the GOA may undertake during their early life stages and suggesting that multiple drivers operating at different spatial and temporal scales influence connectivity patterns. Finally, while our results indicate that interannual variability in physical transport may have a substantial impact on connectivity, we found little support for the hypothesis that this alone drives variability in POP recruitment.