During 2005, the modeled winds did not differ significantly between the northern and southern shelves. Water-column stratification varied from north to south with a sharp two-layer structure over the southern middle shelf and more gradual vertical density gradient over the northern middle shelf (Fig. 6). This was largely due to the decrease in tidal mixing energy with increasing latitude. In the spring, the vertical structure of the northern shelf is weakly stratified by temperature and moderately stratified by salinity. By early autumn, the southern shelf is stratified almost completely by temperature and the northern shelf by a combination of temperature and salinity. Following Carmack (2007), there is a temptation to refer to the salinity influenced northern portion of the Bering shelf as a beta ocean and the temperature-dominated southern portion as an alpha ocean, but such a designation would not be appropriate for this region. Carmack (2007) described the ocean basins, where such a differentiation is long lasting. In contrast the salinity-temperature structure on the shelf can be ephemeral and dependent upon the year-to-year variability in sea-ice extent.
The presence/absence of sea ice modifies the physics, chemistry, and biology of the Bering Sea shelf. In the spring, the northern limit of the southern shelf was determined by the southern extent of spring ice. Advection modified the position of the transition zone between the north and south over the middle shelf. In 2005, this resulted in a northward shift of the transition between the northern and southern shelves during late spring and early summer. Other than the position of this transition zone, differences between the north and southern middle shelf, that were set up by ice, persisted through summer. Whether this occurs each year is not known, although we would hypothesize that in other years sea ice would influence the structure over the shelf and a northward shift would occur, since the weak flow patterns over this shelf appear to be consistent from year to year (Stabeno et al., 2007). The position of the cross-shelf flow that occurs north of the Pribilof Islands, however, may change from year to year, since it is probably related to the baroclinic structure over the shelf. This structure is dependent upon maximum ice extent, so this boundary between the northern and southern shelves will vary because of variability in the spatial position and timing of maximum ice extent. It must be noted that climate models predict that the maximum ice extent over the eastern Bering Sea shelf will remain south of St. Matthew Island. Thus the division between the northern shelf and southern shelf, as evidenced by the extent of the cold pool, will likely remain in the vicinity of St. Matthew Island.
In contrast to the stability of the middle domain, water properties resulting from presence of sea ice over the outer shelf did not persist. Flow over the outer shelf (particularly along the 100-m isobath, Fig. 1) is stronger than over the middle domain. In addition, the outer shelf interacts with water from the slope and basin. In particular, instabilities over the slope can result in intrusions of slope water onto the shelf (Stabeno and van Meurs, 1999; Schumacher and Stabeno, 1998; Mizobata et al., 2008). All these can modify the water properties over the outer shelf.
There was a significant reduction (~50%) in nitrate concentrations on the northern shelf during the summer that was likely a result of nutrient uptake by phytoplankton. In addition, there were unprecedented nitrite concentrations in the summer in an area that had been covered by ice and modified through advection. These high concentrations were perhaps a result of uncoupled nitrification.
During spring, the zooplankton communities differed between the northern and southern middle shelf domains. In late summer 2005, biological processes may have acted to erase differences in zooplankton community structure created by the initial conditions during winter and spring.
The work presented here significantly increases our understanding of the patterns and processes over the eastern Bering Sea shelf and complements the focused studies of the Bering Strait region between St. Lawrence Island and Bering Strait (Grebmeier, 1993) and the examinations of how ice influences these ecosystems (Clement et al., 2004, 2005). Historically, northern and southern regions of the Bering Sea shelf have been defined by absolute geographic coordinates, but from an ecosystem perspective, north and south may be more appropriately defined by a physical, chemical, and biological structure that is a response to the presence or absence of seasonal sea ice. In spring 2005, the transition between the northern and southern shelves was evident in the temperature, salinity, vertical density structure (Brunt-Väisälä frequency), surface nutrients, and the zooplankton community (Figs. 4 and 8), while in late summer, it was most evident in the surface salinity, bottom temperatures, and surface nutrients (Fig. 5).
In 2005, the southern shelf ice-free area stretched from the Alaska Peninsula to approximately Station 91, while the transition region extended from Station 91 to just south of St. Matthew Island. The northern portion of the middle shelf stretched from the northern end of the transition to St. Lawrence Island where the middle shelf ends. So, in addition to the standard cross-shelf domains (coastal, middle, and outer), there are, at least over the middle shelf, a northern domain and a southern domain. We suggest the southern part of the middle shelf retain the name of the Middle Shelf Domain, while the northern part be called the Northern Middle Shelf Domain with the transition between them identified as the North-South Transition.
The results reported here are from a single year, 2005, with limited ice extent. They are the patterns one would expect in the early stages of a warming cycle. Results from low, moderate, and high ice years are necessary to understand if the observations and mechanisms proposed here are valid for the full range of ecosystem response. At the time of this writing we are in a period of extensive ice, and cool to extreme cold for the eastern Bering Sea. Two new programs, the Bering Ecosystem Study (BEST) and the Bering Sea Integrated Ecosystem Research Program (BSIERP) are providing the data to examine our conclusions.
Acknowledgments. We thank D. Kachel and N. Kachel for data analysis and S. Salo for providing satellite images. S. Salo, D. Kachel, P. Proctor, A. Jenkins, J. Clark, K. Mier, R. Cartwright, W. Floering, C. DeWitt, D. Righi, M. Dunlap, S. Smith, S. Thornton, and B. Munger provided assistance at sea and were responsible for collecting the majority of our data. C. Harpold provided assistance with chlorophyll sample analyses, and chlorophyll and zooplankton data management. We thank K. Coyle and G.L. Hunt Jr. for their comments on an early draft of this paper. K. Birchfield and K. McKinney provided graphics work and R.L. Whitney did the technical editing. We thank the officers and crews of the NOAA ship Miller Freeman and R/V Thomas G. Thompson for invaluable assistance in making these oceanographic measurements. This research was funded by NOAA's North Pacific Climate Regimes and Ecosystem Productivity Program and the North Pacific Research Board (Grants: #517, 602, 701). This is contribution FOCI-N688 to the Ecosystems and Fisheries Oceanography Coordinated Investigations; 1813 to Joint Institute for the Study of the Atmosphere and Ocean, University of Washington, and 3197 to PMEL.
Go to References