Sediment organic carbon integrates changing environmental conditions to predict benthic assemblages in shallow Arctic seas

Abstract

Abstract In marine spatial planning, conserving adequate habitats and the food webs they support requires delineating habitats and projecting future trends. For bottom‐feeding marine birds and mammals, repeated benthic sampling over large areas to document changes and to develop predictive models of prey dispersion is quite costly. More easily monitored variables that relate strongly to the biomass and structure of benthic assemblages, and are more readily predicted from physical models of climate change, would facilitate planning efforts. The organic carbon (OC) content of sediments integrates diverse physical and biotic processes, and can be less variable over time than primary production, salinity, temperature, or position of water masses. Sediment OC further subsumes inputs at the base of food webs that can limit carbon flows to higher taxa. For the northern Bering Sea, this study explored the utility of sediment OC as a predictor of benthic assemblage types. Cluster analysis and multi‐dimensional scaling distinguished three main types along a gradient of sediment OC. The assemblage for highest sediment OC had a much greater biomass of brittlestars, diverse marine worms, and two mid‐sized, thinner‐shelled bivalves selected as prey by diving sea ducks. The assemblage for lowest sediment OC lacked brittlestars, had a much greater biomass of amphipods sought by gray whales (Eschrichtius robustus), and had a much higher biomass of two often larger or thicker‐shelled bivalves commonly targeted by walruses (Odobenus rosmarus). Areas of exceptionally low sediment OC tended towards dominance by sand dollars with low foraging value. Our study shows that sediment OC has promise as a proxy for monitoring and predicting changes in important prey assemblages in a given region. Models that link predicted hydrographic patterns to lateral advection of phytodetritus, and the resulting sediment OC, may further allow the use of physical climate models to project the future dispersion of benthic habitats for endothermic predators.