Density-dependent, landscape, and climate effects on spawning distribution of walleye pollock Theragra chalcogramma

Abstract

Little is known about the ways in which density-dependent and density-independent factors influence the spatial dynamics of fish spawning, despite their importance for egg and larval develop- ment, survival, and, ultimately, recruitment. Using 2 novel approaches, we related the spatial and temporal patterns of 28 yr of walleye pollock Theragra chalco- gramma egg data in Shelikof Strait, Alaska, USA, to density-dependent, landscape, and climate effects. First, we used a spatially explicit variable coefficient generalized additive model (GAM) to show that spawning biomass, water temperature, and ocean transport each had unique spatial effects on walleye pollock egg density. Specifically, increased spawning biomass had a positive effect on egg density at the periphery of the core spawning area, inferring spawn- ing area expansion at high spawning biomass and con- traction at low levels. Moreover, increased Shelikof Strait transport caused pollock egg density to increase at the downstream edge of the core spawning area and decrease at the upstream edge. To test the validity of the variable coefficient GAM we also used a geometric method that related the area and center of mass of the pollock spawning aggregation to spawning biomass, transport, and temperature in annual time steps. The area of the spawning aggregation was positively related to spawning biomass, but not to transport or temperature. Variable coefficient GAMs are a useful tool to disentangle the spatially explicit forces of den- sity-dependent and density-independent processes on fish populations.