Antarctic lake systems and climate change

Abstract

Approximately 98% of the Antarctic continent is currently ice-covered and, except for subglacial lakes (Priscu and Christner 2004), possesses no liquid water environments. In contrast, the other 2% contains an extraordinary array of aquatic environments including ice-covered freshwater and saline lakes and ephemeral streams. Lakes are found mostly in coastal, ice marginal regions in the Antarctic (Doran et al. 1994). Because of the great differences in mean-annual temperature related to the locations of these aquatic systems (Fig. 1), the temporal extent of ice covers on the lakes varies greatly. There are lakes that are ice-covered for part of each year, while there are perennially ice-covered systems in the dry valleys region of Southern Victoria Land where only 'moats' (ie ice-free littoral zones) form during the warmest summers. Because ice cover greatly influences both the physical and biological processes occurring within lakes, the extent and thickness of ice cover is an extremely important parameter in the biogeochemistry of Antarctic lakes (Wharton et al. 1993, Fritsen and Priscu 1999). Paleolimnological investigations demonstrate that climate variations have greatly impacted the extent of ice cover, in addition to the overall hydrology of Antarctic lakes throughout the Holocene back into the Last Glacial Maximum (LGM) (Wilson 1964, Bird et al. 1991, Bjorck et al. 1996, Fulford-Smith and Sikes 1996, Gore et al. 1996, Lyons et al. 1998b, Hendy 2000). Much of the paleolimnological work from Antarctic lakes has recently been summarized by Hodgson et al. (2004) and Doran et al. (2004) and will not be repeated here. Some of the earlier work on Antarctic lakes illustrated its value in delineating changes in the hydrologic response to climate (Wilson 1964). The integration of paleolimnological data with more recent observations has provided important insights into how these lakes respond to climate change (eg Gibson and Burton 1996, allied with the work of Roberts et al. 2001 in the Vestfold Hills and Poreda et al. 2004, allied with Hendy 2000 in Taylor Valley, Victoria Land). Understanding the impact of climate on the hydrologic balance of Antarctic lakes, and in turn, the influence of hydrologic changes on the overall ecology of these systems is a major challenge to Antarctic limnologists. This will be especially true in the future period of anthropogenically-induced climate change. Kejna (2003) has recently reviewed the air temperature records from 34 stations around the Antarctic continent. The data span from 1958-2000 for 21 stations and for 1981-2000 for all the stations. In general, warming has occurred on the Antarctic Peninsula and in interior West Antarctica, with Faraday Station, on the west coast of the Peninsula having increased at a rate of 0.67°C per decade over the period 1958- 2000. Many of the coastal stations in East Antarctica also demonstrated an increase in temperature over the longer time interval. However, since 1981, many regions of the continent, especially in East Antarctica, have shown a cooling. For example, Casey shows a 0.82°C per decade decrease (Kejna 2003). There has also been a weakening of the warming rate on the Peninsula during the last 20 years. Because Antarctic lakes are so sensitive to both increases and decreases in temperature (ie Wilson 1964, Gibson and Barton 1996, Foreman et al. 2004), the direct monitoring of Antarctic lakes provides an excellent sentinel of climate change, especially as climate impacts the local hydrologic cycle. Climate variation and change are not the only factors affecting Antarctic lakes. The activities of mammal and bird populations also exert considerable influence on the physical, chemical and biological evolution of many Antarctic limnetic systems. Human activities can also be important in certain situations. For example, Heywood Lake, on Signy Island, has undergone eutrophication in the last 30 years because of input of nutrients from an expanding fur seal population within the catchment (Butler 1999). This has led to increased microbial abundance and changes in the structure of the ecosystem, with phytoplankton taxa more typical of polluted waters, as opposed to the more oligotrophic waters that existed in the 1970s and early 1980s (Butler 1999) before the fur seal population explosion. Eutrophication has led to longer periods of lake anoxia during winter thermal stratification and changes in seasonal biological patterns within the lake. Recent work in the Larsemann Hills has demonstrated that lakes impacted by human activities such as grey water and human waste discharge and even rock crushing by tracked vehicles have enhanced nutrient and total dissolved solid loads (Kaup and Burgess 2002). A comparison of human impacted catchments with catchments with little direct human activity indicates that the levels of dissolved nitrogen compounds were generally much higher in the human-influenced catchments. Salinities were also up to an order of magnitude higher in the human impacted catchments (Kaup and Burgess 2002). Monitoring the response of lake dynamics to changing climate has long been recognized as an important task (Wilson 1981). The linkage between changing climate and lake dynamics becomes even more complicated in lakes that are influenced by direct human impacts. Most investigations of Antarctic lakes have been conducted over limited time periods (1-3 years). Although these short-term studies have been extremely important in establishing base-line conditions, determining the taxa that are present, and understanding biogeochemical processes in the lakes, they are not conducive to establishing long-term limnological trends. Our paper focuses on studies resulting in long-term data comparisons that have produced information about biological and/or physical and chemical trends thought to be driven by changing climatic parameters. This focus greatly limits the resources available to compare long-term trends that do exist and historic information on Antarctic limnology because it eliminates numerous one-time studies of specific limnetic systems in various parts of the continent. (We differentiate between longterm and historic data by defining 'long-term' as relatively continuous records through time and 'historic' as data collected with time gaps between collections). In addition, this paper will not address trends in the epishelf lakes that exist on the continent especially in the Bunger Hills and Schirmacher Oasis regions (Bormann and Fritzsche 1995, Doran et al. 2000, Gibson and Andersen 2002). For information on epishelf lakes, see Gibson et al. this volume. Because of the important differences in the climate regimes within the Antarctic, we have separated our discussion into geographic regimes that have been used previously (Convey 2001). © 2006 Springer.