West Antarctic Peninsula: An Ice-Dependent Coastal Marine Ecosystem in Transition

The extent, duration, and seasonality of sea ice and glacial discharge strongly influence Antarctic marine ecosystems. Most organisms' life cycles in this region are attuned to ice seasonality. The annual retreat and melting of sea ice in the austral spring stratifies the upper ocean, triggering large phytoplankton blooms. The magnitude of the blooms is proportional to the winter extent of ice cover, which can act as a barrier to wind mixing. Antarctic krill, one of the most abundant metazoan populations on Earth, consume phytoplankton blooms dominated by large diatoms. Krill, in turn, support a large biomass of predators, including penguins, seals, and whales. Human activity has altered even these remote ecosystems. The western Antarctic Peninsula region has warmed by 7°C over the past 50 years, and sea ice duration has declined by almost 100 days since 1978, causing a decrease in phytoplankton productivity in the northern peninsula region. Besides climate change, Antarctic marine systems have been greatly altered by harvesting of the great whales and now krill. It is unclear to what extent the ecosystems we observe today differ from the pristine state.

As in other coastal systems, geomorphology and bathymetry play critical roles in structuring the WAP ecosystem (Schofield et al., 2013, in this issue, address some aspects of this). In common with most of the world's coastal regions, the WAP system is profoundly affected by anthropogenic influences, including climate change, pollution, past exploitation of upper trophic level species such as whales and seals, and current exploitation of fish and krill stocks.
There is a rich history of oceanographic research in the region, starting with the Discovery Investigations of 1924-1951(Hardy, 1967. With a strong focus on Euphausia superba, the Antarctic krill, these studies formed the foundation of all subsequent research in the region and, indeed, throughout the aBStr act. The extent, duration, and seasonality of sea ice and glacial discharge strongly influence Antarctic marine ecosystems. Most organisms' life cycles in this region are attuned to ice seasonality. The annual retreat and melting of sea ice in the austral spring stratifies the upper ocean, triggering large phytoplankton blooms. The magnitude of the blooms is proportional to the winter extent of ice cover, which can act as a barrier to wind mixing. Antarctic krill, one of the most abundant metazoan populations on Earth, consume phytoplankton blooms dominated by large diatoms. Krill, in turn, support a large biomass of predators, including penguins, seals, and whales. Human activity has altered even these remote ecosystems. The western Antarctic Peninsula region has warmed by 7°C over the past 50 years, and sea ice duration has declined by almost 100 days since 1978, causing a decrease in phytoplankton productivity in the northern peninsula region. Besides climate change, Antarctic marine systems have been greatly altered by harvesting of the great whales and now krill. It is unclear to what extent the ecosystems we observe today differ from the pristine state.  (Ducklow et al., 2007;Steinberg et al., 2012) by recording semiweekly observations of nearshore processes at Palmer Station (64.8°S, 64.1°W) between October and April, and by conducting a regional-scale cruise in January (Figure 1) each austral summer. PAL was built on intensive studies of Adélie penguin demography and feeding ecology carried out since the mid-1970s (Fraser and Trivelpiece, 1996) and on related oceanographic research (Ross et al., 1996)   and the BAS (Meredith et al., 2004;Clarke et al., 2007).

OceaNOgr aphy aND climate
The WAP's coastal region  is punctuated by islands, promontories, and small peninsulas, and includes a complex network of straits, bays, and passages between the islands and the continental mainland ( Figure 1). A complex coastal circulation is associated with the irregular coastline and nearshore bathymetry, and it includes the recently described Antarctic Peninsula Coastal Current, which appears to be driven by winds and glacial meltwater inputs in the austral summer (Moffat et al., 2008). The coastal circulation may serve to retain or transport plankton within the coastal region, but the spatial and temporal distributions of these effects are not well established.
Along the peninsula, the seafloor deepens abruptly to 200-300 m or deeper within a few kilometers of shore. It is bisected by the landward ends of several glacial-erosion submarine troughs and canyons that exceed 750 m in depth and extend across the continental shelf (Anderson, 1999 (Fraser and Trivelpiece, 1996;Schofield et al., 2013, in this issue). The changing regional climate is discussed in detail elsewhere (Turner et al., 2009), but it should be noted here that the WAP exhibits among the most rapid rates of regional warming anywhere, especially in winter (+7°C since 1950, or five times the global annual mean). The average annual winter (JJA) and summer (DJF) air temperatures are -1.5, -4.9, and +1.4 °C, respectively, for the period  Figure 1. map of study region along the western antarctic peninsula. The small dots are regular hydrographic stations, colored to indicate the percent meteoric water (predominantly glacial melt) in the water column in january 2011 (meredith et al., 2013). red letters show locations of palmer (p) and rothera (r) stations. kg = king george island. c = charcot island. hydrographic lines (colored dots) are 100 km apart (north to south) and 20 km apart (cross shelf). The continental shelf break is to the left.
The ocean in the region is also warming greatly, with a rise in surface ocean temperature in excess of 1°C measured during the second half of the twentieth century (Meredith and King, 2005). Part of this upper-ocean warming is thought to be of atmospheric origin, with the transfer of heat facilitated by greater amounts of ice-free waters from spring to autumn.
The deeper ocean has warmed tremendously as well . A strong source of the heat input to the WAP region is the inflow of warm, mid-depth UCDW from the ACC, where warmer intrusions along the glacially scoured canyons impinge on the inner shelf regions (Martinson, 2012;Martinson and McKee, 2012).
The warming from above and below has resulted in the rapid retreat of the majority of glaciers along the peninsula (Cook et al., 2005), with significant consequences for the coastal ecosystem. (± 41 days; or -2.7 ± 1.2 days/year, p = 0.02). These seasonal sea ice changes are largely wind driven (Holland and Kwok, 2012;Maksym et al., 2012).
Strong northerly winds drive the ice edge southward, delaying ice edge advance in autumn and accelerating its retreat in spring, often synoptically with each passing storm (Stammerjohn et al., 2003;Massom et al., 2008). Increased solar ocean warming in summer (due to earlier and longer ice-free conditions) is also contributing to the sea ice changes, acting as a positive feedback to enhance and sustain the rate of warming and sea ice retreat (Meredith and King, 2005;Stammerjohn et al., 2011). The

WAP and southern Bellingshausen Sea
show the largest and fastest Antarctic sea ice decreases, on a par with the largest regional decreases in Arctic sea ice .

SecONDary prODuctiON aND tOp preDatOrS
Traditionally, Antarctic marine ecosystems are believed to be dominated by the Antarctic krill Euphausia superba and its predators (Murphy et al., 2013).      (Fraser and Trivelpiece, 1996;Fraser and Hofmann, 2003). Cohorts of Antarctic krill year classes can be followed through four-to five-year cycles in Adélie penguin diet samples (Fraser and Hoffman, 2003). Significant krill recruitment events occurred in 1991-1992, 1995-1997, and 2000, 2006, and 2010; data derived after Fraser and Hoffman, 2003). Krill recruitment success is related to heavy winter sea ice (Fraser and Hoffman, 2003), and declining sea ice extent and duration in the Palmer Station region may be contributing to the decline of the krill (Atkinson et al., 2004), as well as the Adélie population, as discussed further below.
Once immobilized in polar regions, POPs enter and concentrate in phytoplankton and krill (Chiuchiolo et al., 2004) and in penguins, giant petrels, and skuas (Geisz, 2010 in the late 1970s, even before the whale recovery began to take off, implicating climate change as an additional factor in Antarctic population dynamics (Fraser et al., 1992). This controversy rages on.