LIMITS AND LIMITATIONS OF DATASETS FOR PALAEOCLIMATIC RECONSTRUCTION USING NONMARINE OSTRACODS (Invited Presentation)

Ostracods, widespread in marine and continental aquatic environments, have an excellent fossil record and are valuable palaeoclimate proxies. They are, in effect, multi-proxy microfossils; their shells not only provide whole-animal evidence of past distributions from which to draw palaeoclimate inferences via indicator species, transfer function and mutual climatic range approaches, but also are handy packages of biogenic calcite in which stable isotopes and trace elements can be measured and interpreted in terms of, e.g., temperature and salinity. Calibration of ostracod species as palaeoclimate proxies utilises living distributional datasets in combination with modern climate datasets. Small regional datasets collected during relatively short time intervals may be internally consistent in terms of sampling methods, taxonomy and positional accuracy, but often fall short of capturing the full climatic ranges of taxa. Geographically more extensive, literature-based datasets are more likely to encompass species' distributions in climate space but, since they represent collections by many people over many years, are prone to taxonomic inconsistency as well as poor locational precision and accuracy, and may even represent distributions blurred by shifts in response to contemporary climate change. Regional climatic datasets, too, span different time intervals. The inability to match climate data exactly to the interval during which distributional data were collected weakens an underlying assumption of palaeoclimate proxy methods: that species' distributions are approximately in equilibrium with climate. We explore some implications of these issues with reference to large regional nonmarine ostracod databases for Europe and North America. We suggest that the accuracy of ostracod proxy methods for palaeoclimatic reconstruction may be improved by utilising (1) OMEGA (Ostracod Metadatabase of Environmental and Geographical Attributes), a global metadatabase under development that will facilitate capturing species' full climatic ranges, and (2) the WorldClim global interpolated climate dataset as a standard for species calibrations. Realising the potential of such a global dataset will require harmonisation of the taxonomic schemes of the component regional databases.