Climatic and anthropogenic influences on aquatic ecosystems in the valley of the Great Lakes, Mongolia.

Abstract

Climate warming and major land use changes have profoundly affected the Mongolian landscape in the past several decades. As in many arid and semi-arid regions, water resources are critically important for ecological, social, and economic viability. In Mongolia, traditional semi-nomadic pastoralism contributes substantially to the national economy as well as to individual subsistence and depends on limited freshwater resources to provide for grazing herds and human needs. Yet, because of substantial variability across this immense region, its remoteness, and recent political transitions, little work has been done to monitor water quality or to set baseline standards against which to measure future changes. Compounding the problem is a lack of well-resolved paleo-ecological and paleo-climatological work. These data are necessary to provide a foundation for understanding the natural variability in the aquatic systems of the region, especially with recent changes in climate and land use. This work is a contribution to developing these records by first, developing diatom-based inference models for total phosphorus and salinity, and second, applying the models to investigate lacustrine sediment records of past changes. The diatom-based inference models were based on a survey of the water chemistry, physical characteristics, and diatom flora of 64 lakes in western Mongolia. The region had a diverse diatom flora with over 300 species, nearly 100 of which had not been previously reported from Mongolia, from lakes ranging from fresh to hypersaline. The many isolated lake basins provided unique ecosystems where multiple unique communities, novel species distributions, and new and endemic flora were found. Three of these new species are described here in a careful examination of the genus Cyclotella in western Mongolia. Canonical correspondence analysis (CCA) was used to identify four variables (specific conductivity, total phosphorus, bicarbonate, and lake morphology) that were significantly related to the distribution of diatoms; predictive models were developed for specific conductivity and total phosphorus using weighted averaging regression and calibration methods. The application of these models to dated lake sediment cores, along with interpretations of other geochemical and sediment characteristics, was then used to develop records of variability in lake salinity and nutrient flux. The interpretation of diatom and sedimentary records demonstrated increases in nutrient fluxes to the lakes related to climate warming and major changes in land-use over the last 20 years. Diatom-inferred lake salinity was correlated with changes in temperature over the past 2000 years, as inferred from tree-ring records, demonstrating a positive relationship between increased warming and increased lake salinity in recent geologic history. Changes in warm-season temperature, as inferred from tree-rings, in the most recent decades were less-well correlated with inferred changes in salinity than over most of the 2000 year record. However, instrumental records of winter temperature were well correlated with recent shifts in inferred salinity, perhaps suggesting recent changes in climate that are unique from those over the past several thousand years.