Palmer Deep Canyon is one of the biological hotspots associated with deep bathymetric features along the Western Antarctic Peninsula. The upwelling of nutrient-rich Upper Circumpolar Deep Water to the surface mixed layer in the submarine canyon has been hypothesized to drive increased phytoplankton biomass productivity, attracting krill, penguins and other top predators to the region. However, observations in Palmer Deep Canyon lack a clear in-situ upwelling signal, lack a physiological response by phytoplankton to Upper Circumpolar Deep Water in laboratory experiments, and surface residence times that are too short for phytoplankton populations to reasonably respond to any locally upwelled nutrients. This suggests that enhanced local upwelling may not be the mechanism that links canyons to increased biological activity. Previous observations of isopycnal doming within the canyon suggested that a subsurface recirculating feature may be present. Here, using in-situ measurements and a circulation model, we demonstrate that the presence of a recirculating eddy may contribute to maintaining the biological hotspot by increasing the residence time at depth and retaining a distinct layer of biological particles. Neutrally buoyant particle simulations showed that residence times increase to upwards of 175 days with depth within the canyon during the austral summer. In-situ particle scattering, flow cytometry, and water samples from within the subsurface eddy suggest that retained particles are detrital in nature. Our results suggest that these seasonal, retentive features of Palmer Deep Canyon are important to the establishment of the biological hotspot.

Diel vertical migration (DVM) is a common behavior in zooplankton populations world-wide. Every day, zooplankton leave the productive surface ocean and migrate to deep, dark waters to avoid visual predators and return to the surface at night to feed. This behavior may also help retain migrating zooplankton in biological hotspots. Compared to fast and variable surface currents, deep ocean currents are sluggish, and can be more consistent. The time spent in the subsurface layer are driven by day length and the depth of surface mixed layer. A subsurface, recirculating eddy has recently been described in Palmer Deep Canyon, a submarine canyon adjacent to a biological hotspot. Previous circulation model simulations have shown that residence times of particles increase with depth within this feature. We hypothesize that DVM into the subsurface eddy increases local retention of migrating zooplankton in this biological hotspot and that shallower mixed layers and longer day length would increase the time in the subsurface layer. We demonstrate that vertically migrating particles have residence times on the order of 30 days, which is significantly greater than residence times of near-surface, non-migrating particles. The interaction of DVM with this subsurface feature may be important to the establishment of the biological hotspot within Palmer Deep Canyon by retaining critical food resources in the region. Similar interactions between DVM behavior and subsurface circulation features, modulated by mixed layer depth and day length, may also increase residence times of local zooplankton populations elsewhere.

A decade (2007-2016) of hourly 6 km resolution maps of the surface currents across the Mid Atlantic Bight (MAB) generated by a regional-scale High Frequency Radar network are used to reveal new insights into the spatial patterns of the annual and seasonal mean surface flows. Across the 10-year time series, temporal means, inter- and intra-annual variability are used to quantify the variability of spatial surface current patterns. The 10-year annual mean surface flows are weaker and mostly cross shelf near the coast, increasing in speed and rotating to more alongshore directions near the shelf break, and increasing in speed and rotating to flow off-shelf in the southern MAB. The annual mean surface current pattern is relatively stable year to year compared to the hourly variations within a year. The ten-year seasonal means exhibit similar current patterns, with winter and summer more cross-shore while spring and fall transitions are more alongshore. Fall and winter mean speeds are larger and correspond to when mean winds are stronger and cross-shore. Summer mean currents are weakest and correspond to a time when the mean wind opposes the alongshore flow. Again, intra-annual variability is much greater than interannual, with the fall season exhibiting the most interannual variability in the surface current patterns. The extreme fall seasons of 2009 and 2011 are related to extremes in the wind and river discharge events caused by different persistent synoptic meteorological conditions, resulting in more or less rapid fall transitions from stratified summer to well-mixed winter conditions.