Predicting the long-term acidification trends in small subarctic lakes using diatoms

1. A considerable proportion of the total deposition of sulphur in northern most Europe originates from the large non-ferrous smelters of the Kola Peni nsula, Russia. Potential long-term effects of this point source pollution o n sensitive subarctic lakes were evaluated using palaeolimnological techniq ues. 2. Multivariate analysis of a diatom and water chemistry data set from 45 s mall headwater lakes located in north-eastern Finnish Lapland demonstrated that pH, calcium and silica were the three most powerful chemical variables in explaining the variance in the diatom data. From these, lake water pH w as shown to be the strongest determinator by variance partitioning. 3. Weighted averaging partial least squares regression (WA-PLS) was used to develop a diatom-based prediction model for inferring lake water pH from s ediment core diatom assemblages. The performance of the model was assessed by leave-one-out cross-validation. 4. The prediction model was applied to radiometrically dated sediment cores taken from three headwater lakes receiving different amounts of acid fallo ut from the Kola Peninsula smelter industry. 5. Stable diatom assemblages and results of pH reconstructions suggested th at no substantial changes in the acidification status of the lakes have occ urred within the last century despite the very high local acid deposition. 6. The pollution levels in the study area have not increased to the point w here the biology of the lakes has been influenced significantly.