Assessing the accuracy of diatom-based transfer functions in defining reference pH conditions for acidified lakes in the United Kingdom

A key concern in the assessment of the recovery of acidified lakes from the effect of acid deposition is the reference pH of acidified lakes and how reference values can be used to establish targets for restoration. In this paper we evaluated the accuracy of three different, although overlapping, diatom-pH transfer functions using UK Acid Waters Monitoring Network (AWMN) data to compare inferred pH from diatoms collected in annually exposed sediment traps with measured water column pH for 11 acid-sensitive lakes. We then used these transfer functions to infer lake-water pH from approximately AD 1850 (taken as the reference date) to AD 1990 (the core-top date) for radiometrically dated sediment cores from each site. When applied to the sediment-trap samples all three models tended to underpredict measured pH in the less acidic sites, with the SWAP model being the most biased. All three models were in good agreement in reconstructing pH for the AD 1850 reference period. Reference pH varied between sites from approximately 4.9 to 6.4, and was correlated with modern basecation status. In the acidification phases of the cores discrepancies in pH inference between models appear to be linked to differences in pH optima for a minority of taxa. Current (2006) diatom-inferred pH values from the most recent sediment traps, when compared with sediment core tops, provide evidence of recovery at most sites but its extent, so far, is slight when compared with the reference period. The substantial difference between sites in their reference conditions indicates clearly that recovery success can not be measured against a single target pH or, by inference, ANC value and must be assessed on a site-specific basis.