Effect of temperature on Flathead Sole (Hippoglossoides elassodon) spawning in the southeastern Bering Sea during warm and cold years

Between 2001 and 2013 the southeastern Bering Sea alternated between multi-year periods of relatively warm (2001-2005) and relatively cold (2007-2013) years. Adult Flathead Sole Hippoglossoides elassodon in the southeastern Bering Sea shift their spatial distribution pattern in relation to bottom water temperature indicating the importance of the thermal environment in determining fish habitat. The objective of this study was to infer Flathead Sole spawning area in the southeastern Bering Sea in May from ichthyoplankton surveys of their eggs using generalized additive models to assess how warm (2002, 2003, and 2005) and cold (2008, 2009, 2010, and 2012) years affected the location of their spawning site. Three mechanisms through which temperature could affect egg distribution were investigated: 1) additively, assuming a similar effect of temperature across the spatial domain, 2) as a spatially variable term, assuming that the effect of temperature changes across locations, and 3) as a threshold effect on the spatial distribution of eggs, assuming the distribution of eggs changes abruptly in relation to the annual average temperature. The model including temperature as a spatially variable effect had the smallest AIC score and therefore was chosen as the best fit. This model showed that the effect of temperature on egg density varied across the study area and indicated a northeastward expansion of the spawning site as temperature increased. The area of expansion may correspond with the outer reach of a cold-water mass (called the cold pool) that extends in the Bering Sea middle domain during late spring and summer, and forms as a result of the previous winter ice coverage. Flathead Sole are known to avoid temperatures colder than 2 degrees C that is also used as the boundary for the cold pool, thus spawning fish may have moved relative to the spatial extent of the cold pool.