Impacts of sea ice retreat, thinning, and melt-pond proliferation on the summer phytoplankton bloom in the Chukchi Sea, Arctic Ocean

摘要

In 2011, a massive phytoplankton bloom was observed in the Chukchi Sea under first-year sea ice (FYI), an environment in which primary productivity (PP) has historically been low. In this paper, we use a 1-D biological model of the Chukchi shelf ecosystem, in conjunction with in situ chemical and physiological data, to better understand the conditions that facilitated the development of such an unprecedented bloom. In addition, to assess the effects of changing Arctic environmental conditions on net PP (NPP), we perform model runs with varying sea ice and snow thickness, timing of melt, melt ponds, and biological parameters. Results from model runs with conditions similar to 2011 indicate that first-year ice (FYI) with at least 10% melt pond coverage transmits sufficient light to support the growth of shade-adapted Arctic phytoplankton. Increasing pond fraction by 20% enhanced peak under-ice NPP by 26% and produced rates more comparable to those measured during the 2011 bloom, but there was no effect of further increasing pond fraction. One of the important consequences of large under-ice blooms is that they consume a substantial fraction of surface nutrients such that NPP is greatly diminished in the marginal ice zone (MIZ) following ice retreat, where NPP has historically been the highest. In contrast, in model runs with <10% ponds, no under-ice bloom formed, and although peak MIZ NPP increased by 18-30%, this did not result in higher total annual NPP. This suggests that under-ice blooms contribute importantly to total annual NPP. Indeed, in all runs exhibiting under-ice blooms, total annual NPP was higher than in runs with the majority of NPP based in open water. Consistent with this, in model runs where ice melted one month earlier, peak under-ice NPP decreased 30%, and annual NPP was lower as well. The only exception was the case with no sea ice in the region: a weak bloom in early May was followed by low but sustained NPP throughout the entire growth season (almost all of which occurred in deep, subsurface layers), resulting in higher total annual NPP than in cases with sea ice present. Our results also show that both ultraviolet radiation and zooplankton grazers reduce peak open water NPP but have little impact on under-ice NPP, which has important implications for the relative proportion of NPP concentrated in pelagic vs. benthic food webs. Finally, the shift in the relative amount of NPP occurring in under-ice vs. open-water environments may affect total ecosystem productivity. (C) 2014 Elsevier Ltd. All rights reserved.