UC Berkeley

Global-scale climate change is an extensively studied topic: researchers monitor climatic indicators from sea surface temperature to atmospheric CO₂ concentrations. However, historical records do not exist on timescales sufficiently extensive to permit projection of future events. To remedy this situation, researchers utilize data collected from sediments and rock formations deposited in pre-historical times in order to refine the predictive capabilities of current climate models. One can see from the relatively small number of datasets comprising samples from both the northern and southern hemispheres that climate change through time has not been uniform across the globe.

Complicating the interpretation of this variability is the dearth of data sets from the southern hemisphere. Many of the data sets generated from samples collected in the southern hemisphere are derived from strata deposited in marine environments. Therefore, while we can, with reasonable certainty, reconstruct temperature at the sea surface and conditions in the atmosphere, scenarios that quantify the magnitude and geographic extent of climatic change occurring on southern continents are less clear. This problem is due in part to sampling bias and to the smaller land mass in the southern hemisphere.

This research provides empiric (as opposed to model-based) constraints on regional climate shifts in southeastern Australia through multiple glacial-interglacial cycles over the past ~400 ka. Because lake sediment records have the potential to provide a contiguous history of temperature and aridity in a terrestrial environment, we collected cores from the lakebed and through the lunette of Lake Tyrrell in northern Victoria, Australia. Lake Tyrrell is hypersaline and ephemeral; consequently, deflation of sediments has created gaps in the sediment sequence, and many sedimentary elements traditionally used for palaeoenvironmental reconstruction are poorly-preserved or absent. Since all of the lakes in the region share these limitations, it was necessary to develop methods to extract useful information from the available sampling sites.

As a new approach, we combined the analyses of the identities, quantities and isotopic signatures of preserved lipid biomarkers with traditional proxies such as pollen, charcoal, mineral content and sedimentary textures. Together, these data describe changes in the chemistry of the lake, in the community composition of microbes within the lake, and in the higher plant ecosystem around the lake and within the catchment.

Evidence of at least three glacial-interglacial cycles is preserved in Tyrrell lakebed and lunette sediments, though the temporal sequence is uncertain due to the lack of dateable materials within the Tyrrell sediments. During what we interpret as the two most recent glacial maxima (~20 ka and 140 ka), the lake dried completely and higher plants colonized the surface following leaching of salt from the surface sediment. At one of the two preceding glacial maxima (either 260 ka or 340 ka), the lake dried to <1m depth, became groundwater-fed, and supported the growth of microbial mats. These mats colonized the interface between the halite crust and the sediment surface. Halophilic archaea and algae lived within the water column. Both the chemistry and microbiology of the lake were comparable to the modern system. During interglacial periods, increased precipitation and runoff increased lake depth to several meters. Lower salinity promoted the growth of dinoflagellates and eustigmatophyte algae in the water column, and of anaerobic bacteria within sediments.

In contrast to the shifts in lake water levels and chemistry, the plant community within the catchment stayed relatively constant through the end of the Pleistocene. It was composed of a dryland casuarina wood with a grassy understory. During the Holocene, increased precipitation following the last glacial maxima (LGM) caused the casuarina woodland to be replaced by Callitris thickets and grasses. Two shifts in the isotopic signature of plant waxes indicate changes in the relative abundance of C₃ vs C₄ grasses over this period. Increased dominance of C₄ grasses following the last glacial maximum points to a strengthening in the Australian summer monsoon. Although direct dating was not possible, periods of increased precipitation between 7.5 &ndash 9.5 ka and 11.5 &ndash 13.5 ka are recorded at proximitous sites. Lake Tyrrell is currently the southernmost terrestrial site from which this precipitation increase has been reported. Over the past 200 years, land use practices following European settlement have caused C₃ grasses to become increasingly dominant.

The data presented here demonstrate that it is possible to obtain information about the timing and extent of climate change in Australia through the Pleistocene and Holocene from ephemeral salt lake sediments. Lipid biomarkers are a particularly useful proxy, as they are preserved even in oxidized, saline sediments. Moreover, they provide taxa-specific information on organisms living within and around lakes, and can be used to reconstruct palaeo-temperatures and lake levels. Isotopic analyses of individual biomarkers, when interpreted in conjunction with charcoal and pollen, generate a picture of the changes in the terrestrial plant ecosystem of the catchment. By employing lipid biomarker analysis alongside traditional proxies, researchers can more accurately quantify changes in palaeo-climate at the regional scale, in order to refine predictive models of continental and global scale climate change.