Climate variability and ecosystem dynamics at remote alpine and arctic lakes: the last 200 years

19 páginas, 11 tablas, 4 figuras. This paper compares the palaeolimnological evidence for climate change over the last 200 years with instrumental climate data for the same period at seven European remote mountain lakes. The sites are Øvre Neådalsvatn (Norway), Saanajärvi (Finland), Gossenköllesee (...

Full description

Bibliographic Details
Main Authors: Battarbee, Rick, Grytnes, J. A., Thompson, R., Appleby, Peter, Catalán, Jordi, Korhola, A., Birks, H. J. B., Heegaard, Einar, Lami, A.
Format: Article in Journal/Newspaper
Language:English
Published: Springer 2002
Subjects:
Online Access:http://hdl.handle.net/10261/44803
Description
Summary:19 páginas, 11 tablas, 4 figuras. This paper compares the palaeolimnological evidence for climate change over the last 200 years with instrumental climate data for the same period at seven European remote mountain lakes. The sites are Øvre Neådalsvatn (Norway), Saanajärvi (Finland), Gossenköllesee (Austria), Hagelseewli (Switzerland), Jezero v Ledvici (Slovenia), Estany Redó (Spain, Pyrenees), and NiÓné Terianske Pleso (Slovakia). We used multiple regression analysis to transfer homogenised lowland air temperature records to each of the sites, and these reconstructions were validated using data from on-site automatic weather stations. These data showed that mean annual temperature has varied over the last 200 years at each site by between 1 and 2 °C, typical of the high frequency variability found throughout the Holocene, and appropriate, therefore, to test the sensitivity of the various proxy methods used. Sediment cores from each site were radiometrically dated using 210Pb, 137Cs and 241Am and analysed for loss-on-ignition, C, N, S, pigments, diatoms, chrysophytes, Cladocera and chironomids. Comparisons between the proxy data and the instrumental data were based on linear regression analysis with the proxy data treated as response variables and the instrumental data (after smoothing using LOESS regressions) as predictor variables. The results showed few clear or consistent patterns with generally low or very low r2 values. Highest values were found when the data were compared after smoothing using a broad span, indicating that some of the proxy data were capturing climate variability but only at a relatively coarse time resolution. Probable reasons for the weak performance of the methods used include inaccurate dating, especially for earlier time periods, the influence of confounding forcing factors at some sites e.g., air pollution, earthquakes, and the insensitivity of some methods to low amplitude climate forcing. Nevertheless, there were trends in some proxy records at a number of sites that had a ...