Human Impact on Lake Ecosystems (LIMPACS)

Although on a global scale lakes play only a minor role in hydrological and biogeochemical cycles, they have a special importance in many continental areas owing to their landscape value, their use as a natural resource (for water supply, recreation etc) and their value as natural ecosystems and centres of biodiversity. However, lakes are under increasing threat due to the separate, but often combined impact of: • nutrient enrichment from domestic and agricultural pollution; • acid deposition from fossil fuel combustion; • salinisation from over extraction of freshwater; • pollution from toxic metals, persistent organic pollutants and radionuclides; • accelerated infi ll from catchment soil erosion; • habitat disturbance from engineering projects; • ecological disruption from species introductions and invasions. • climate change from greenhouse gases; In developing a global strategy for the protection or sustainable use of lakes it is crucial that we understand how lakes function on different time-scales in response to both natural forces and human impact and that we are able to predict how lake ecosystems will change in the future as stresses are altered. The central theme of LIMPACS then is “understanding the past variability of lake ecosystems in order to predict better their future”. It requires collaborative research between paleolimnologists and limnologists to bring together interdependent approaches for studying lake status, past, present and future. This interdependence involves modelling, measuring and reconstruction of lake attributes (Figure 1). Modelling, especially process-based dynamic modelling, is needed to develop a system-level understanding of lake functioning and to make predictions of future status by scenario testing. Reconstruction, using paleolimnological techniques is needed to understand past variability on inter-annual and decadal time-scales and to provide model verifi cation, whilst direct measurements from observation and experiment are needed to understand the nature of shortterm variability and to paramaterise and calibrate models and transfer functions. Where direct observations of an individual water body have taken place over several years to provide longer-term records of seasonal to decadal variability, time-series are created that can be used to evaluate the performance of both dynamic models and paleolimnological transfer functions (e.g. Figure 2). For some lake types, e.g. mountain lakes, such time-series are comparatively rare but as monitoring networks expand the data available for this purpose are increasing. Sites with long-term records will consequently play a key role in LIMPACS and one task within the programme will be to develop an inventory of such sites suitable for model testing.