Paleo-history of the Palmer LTER (Western Antarctic Peninsula) Region a Supplement to the Palmer Long-Term Ecological Research Project R.C. Smith, E.W. Domack, S.D. Emslie & W.R. Fraser Introduction The inherent variability in the state of the atmosphere-ocean-cryosphere-lithosphere system establishes the setting in which ecosystems evolve. Understanding the effects of climate change is thus critically important to the interpretation of data derived from long-term ecological research. There is considerable evidence (Kennett & Barron 1992) suggesting that the Antarctic holds paleoenvironmental information important for understanding the relationships between climate and ecosystem evolution, as well as the linkages between environmental change and ecosystem responses. Indeed, a recent synthesis (see Ross, Hofmann & Quetin 1996) of environmental and ecological processes in the Western Antarctic Peninsula region not only supports this view, but suggests that the Palmer LTER (PAL) program is ideally situated in a climatically sensitive region. This region may thus present unique opportunities to: 1) detect the effects of climate change on Southern Ocean ecosystems, 2) understand the processes that lead to ecosystem change, and 3) define the scope and cause of the natural variability inherent in the system. The latter in particular provides the prerequisite data necessary for examining the effects of human disturbance against the background of natural ecosystem variability. We propose to collect and integrate interdisciplinary data from the paleoecological record of the Antarctic Peninsular region to address hypotheses on natural climatic variability and ecosystem responses to that variability in the Palmer LTER (PAL) region. We will involve non-LTER scientists with paleoecological experience (E.W. Domack, S.D. Emslie and others) to work with PAL PI's to create a paleohistory, with emphasis on the past few thousand years, in order to place the current PAL research into a larger context. Hypotheses on high-trophic species, especially penguins, and how they respond to interannual variations in climate (past and present) will be a key focus of this research. In addition, the sediment record will provide an extended time frame in which to view current and future observations on phytoplankton productivity and carbon flux. This project will draw together scientists from several disciplines who have been actively working in the Peninsular region to focus on specific questions concerning natural versus human-induced factors that have and are impacting productivity of Antarctic living marine resources. The resultant collaboration will allow us to test our data against each other's results, obtain new information from field and laboratory studies, and pose new questions and identify new avenues of research for future investigation. This effort will represent the first time that biological, oceanographic and geological communities will have joined forces in order to address a specific region and problem in Antarctic marine research and this collaborative thrust is a major objective for this proposed supplement. Paleoclimatic Trends in the PAL Region Paleoclimatic data for Antarctica, especially for the past 10,000 years, stems primarily from the collection and analysis of ice cores and marine sediments (Lorius 1984; Aristarain et al. 1986; Clapperton and Sugden 1988; Peel et al. 1988; Jacka 1990; Mosely-Thompson et al. 1990, 1991; Parkinson 1990; Domack et al. 1993, 1995; Shevenell 1966; LoPiccolo 1966). These data indicate that Antarctica has been characterized by considerable climatic variation through time, and that this variation is not consistent (when ice cores and marine sediment cores are compared) throughout the Antarctic. Data based on sediment cores that are specific to the PAL suggest that over the past several thousand years the region has been characterized by distinct climate zones (Domack & McClennen, In Press) and short-term cycles of warming and cooling that recur approximately every 200 years (Leventer et al., In Press). Especially relevant to the current focus of PAL research on the effects of sea ice on ecosystem dynamics (see Smith et al. 1995), is that these periods of warming and cooling appear to have produced coincident changes in the extent and duration of regional ice cover (Domack et al., 1993) and glacial margin fluctuations at the heads of fjords (Domack et al., 1995). In the vicinity of Palmer Station, for example, the Little Ice Age (mid-1500's to the 1850s) sediment record suggests that sea ice was much more of a permanent, as opposed to a seasonal, feature of the marine environment about 330 years ago, the peak of the LIA (Domack et al., 1996). Expression of this feature may have included permanent fast ice or ice that broke out late in the season. Since then, long-term regional ice cover characteristics (permanent vs. seasonal) have reversed one or more times (Leventer et al., In Press), although dates for expression of the modern record, which incudes high interannual variability in seasonal sea ice extent and duration (Fraser et al., 1992; Smith et al., 1996) cannot be estimated with precision with the present data set. To overcome this problem the LTER program in co-operation with Hamilton College's REU initiative collected three critical short cores in late 1995. These records hold the key to integrating the longer term record of paleoenvironmental change with the documented changes of the last 50 years. These sediment records, along with the proposed penguin rookerie excavations, will help to document the most recent time periods. For example, based on relationships between sea ice characteristics and ambient air temperatures determined within the period of instrument records the pattern of high variability in the PAL area appears to have become established during mid-century and was not a feature of the marine environment prior to that time (Fraser et al., 1992; Murphy et al., 1995; Smith et al., 1996a,b). This proposed supplement will help clarify the paleoenvironmental record and place our current work into the context of these longer records. Implications for Ecological Processes What effects this long-term variability in climate has had on ecosystem change in the PAL region has come from two very different components of the marine food web, the diatom fauna present in sediment cores and the species represented in excavations of abandoned penguin rookeries. Both these sources clearly point to variability in sea ice extent and duration at various time and space scales as being one of the major determinants of change in ecosystem structure and function. In the case of the diatom data, for example, warming and cooling periods are associated with sequential changes in the species composition, abundance and distribution of the diatoms present in the sediment cores (Leventer et al., In Press). A key implication of these findings is that changes in marine productivity and associated carbon flux appear to have been modulated by coincident cyclicity in the timing and extent of sea ice cover during the last several hundred years (Domack & McClennen, In Press; see also Smith & Stammerjohn, In Press). Excavations of abandoned penguin rookeries have demonstrated that variability in paleoclimate and its effect on the spatial and temporal persistence of sea ice may be a key factor determining changes in the biogeography of upper-trophic level predators. Emslie (1995), for example, has shown that during the LIA Chinstrap Penguins (Pygoscelis antarctica) may have occupied rookeries near the tip of the Antarctic Peninsula only during warming periods. These data provide one of the most important lines of evidence in support of an hypothesis advanced by Fraser et al. (1992) on how climate change may affect population trends and biogeography of apex predators. According to these authors, which base their hypothesis on winter habitat studies, Chinstrap Penguins are an ice-intolerant species, while the closely related Adelie Penguin (P. adeliae) is ice-dependent. These authors thus argue that recent increases in the populations of Chinstrap Penguins and decreases in those of Adelie Penguins are better explained by a decrease in the frequency of cold winters with extensive sea ice resulting from environmental warming during the last five decades (see also Smith et al., 1996a) than by the effects of competitive release on krill availability assumed to have occurred when commercial hunting depleted baleen whale stocks (cf. Laws 1985). Significance for PAL The fact that environmental factors suspected of driving recent trends in the populations of these penguins can be coupled to habitat occupation based on the paleoecological record provides a strong theoretical basis for developing future models on how climate change may affect the structure and function of Southern Ocean ecosystems. One model currently being developed (Fraser & Trivelpiece, 1996a) describes a "habitat optimum" for Adelie Penguins that varies in space and time in accordance with changing atmospheric and oceanic processes and their coupled effects on sea ice formation, which ultimately mediates the availability of suitable nesting and foraging habitats (Fraser & Patterson, 1996; Fraser & Trivelpiece, 1996b). According to this model, the habitat optimum for Adelie Penguins in the vicinity of Palmer Station occurred sometime in the last 100-200 years. Supporting evidence includes both currently decreasing populations of Adelie Penguins and large numbers of extinct colonies (Fraser & Patterson, 1996). The surface area covered by the latter suggests that an additional 10,000-15,000 pairs of Adelie Penguins nested within 10 km of the station. Research on abandoned rookeries in the PAL region can add significant new information to test the model presented above. First, this region previously has not been sampled for paleoecologic data of this nature and doing so will allow comparison with similar data from the tip of the Antarctic Peninisula previously published by Emslie (1995). This comparison will be especially important given the high degree of climatic variability that occurs throughout the Antarctic Peninsular region. Second, it will allow integration of data at several levels, including independent paleoclimatic information collected by Domack from sediment cores, and previously published ice-core data, with that from the abandoned rookeries. Finally, it will allow us to test and refine the model on how penguins have responded to climate change in the past in comparison to current responses to global warming and other, human-induced disturbances occurring in Antarctica today. With these results, we can then gain insight on the relative role that natural and human-induced disturbances have on the conservation of Antarctic living marine resources. Methods All sites investigated will be mapped with a Global Positioning System instrument. Excavations will be completed at abandoned rookeries by establishing no more than two 1 m2 or smaller test pits within the rookery. Excavation methods, identification of penguin remains, and radiocarbon analyses will follow the methods of Emslie (1995). All radiocarbon samples (bone, feathers or eggshell) will be processed at the Laboratory for Accelerator Radiocarbon Research, University of Colorado, Boulder, for Accelerator Mass Spectrometry (AMS) dating, and will be calibrated for the marine reservoir effect using the methods of Stuiver et al (1986). Otoliths and squid beaks recovered from the excavations will be identified by B. Walker (NMFS). Budget Justification A key element of this proposed supplement to the PAL is to directly involve paleoenvironmental expertise (Domack & Emslie) in PAL research. The PAL has already established the beginnings of this collaborative effort with Domack. Making use of a Polar Duke cruise, PD95-10 (Dave Karl, Chief Scientist), the LTER project and a REU grant to Hamilton College (Eugene Domack, PI) several cores were obtained in the PAL area. The object of this collaboration was to obtain several sediment cores from the Palmer Deep (and area close to Palmer Station and within the Adelie penguin foraging area studied by the PAL). Preliminary magnetic and radiocarbon analyses of these cores has established a tie betweent the longer term record (where 200 year cyclicity dominates) and the most recent pertubations of the LIA and nineteenth century warming. Thus the necessary sediment cores for collaborative work have already been obtained. The penguin excavation is critical inorder to link current higher trophic level findings and hypotheses to the paleorecord. Two months salary, as well as costs for radiocarbon and mass spectrometry analysis, are included for Steve Emslie so that he can do this proposed research in the PAL area. Travel and per-diem are provided for Domack, Emslie and their collaborators (Leventer, Ishman, Venkatesen) as well as interested PAL PI's to travel to Santa Barbara for a several day workshop. The objective of this workshop will be to not only discuss results and to prepare manuscripts discussing the anticipated multidisciplinary results. Further, the workshop will be used to develop a systematic plan for future collaborative projects on the paleoenvironmental evolution of the PAL area. This plan will include proposed work on biomarker (organic geochemical) components in the marine sediment record which can provide a glimpse on the response of a number of trophic levels to past paleoclimatic fluctuations (Domack, Mashiotta, Venkatesen, 1992) as well as the penguin excavation research.