Recovery of acidified waters in the UK

WP 1: Targets for recovery

Reference sites

Analogue matching is a palaeolimnological technique that aims to find matches for fossil sediment samples from a set of modern surface sediment samples. Modern analogues were identified that closely matched the pre-disturbance conditions of eight of the UK AWMN lakes using diatom- and cladoceran-based analogue matching.

These analogue sites were assessed in terms of hydrochemistry, aquatic macrophytes and macro-invertebrates as to their suitability for defining wider hydrochemical and biological reference conditions for acidified sites within the AWMN. The analogues identified for individual AWMN sites show a close degree of similarity in terms of their hydrochemical characteristics, aquatic macrophytes and, to a lesser extent, in the macro-invertebrate fauna.

The reference conditions of acidified AWMN sites are inferred to be less acidic than today and to support a wider range of acid-sensitive aquatic macrophyte and macro-invertebrate taxa than that recorded in the AWMN lakes over the period of monitoring since 1988.

Chemical-biological database and interactive web page

Determination of chemical-biological response curves for national applications

To improve models of the relationships between the presence of key biological indicator groups and water chemistry, development of the chemical-biological database has continued under this programme. Diatoms and invertebrates have been used because they are key structural components of freshwater ecosystems and are extremely sensitive to changes in acidity.

These models are essential to predict the probability of occurrence of key taxa under pre-acidified conditions where hindcast water chemistry data are provided by static or dynamic models (i.e. to provide knowledge of likely biological targets for recovery under "reference" conditions) and the probable biological deviation from target conditions (i.e. damage) in future under different emissions reduction scenarios.

Note that in either case the chemical-biological models can only predict the suitability of chemical conditions for the presence of key taxa and do not account for other important factors such as habitat suitability or presence of biological populations from which dispersion into chemically recovered areas may occur.

Predictive model development focused on the "alkalinity + xDOC" formulation of ANC using both logistic regression and generalized additive modeling techniques. Response models were fitted to all widespread taxa (38 diatoms, 29 invertebrates) in the database, plus aggregated groups of acid-sensitive diatoms and invertebrates.

A total of 36 diatoms and 22 invertebrates showed a significant response to ANC, varying from linear/monotonic, Gaussian symmetric unimodal and complex linear or non-symmetric unimodal responses. For all these taxa, it is therefore possible to predict probability of occurrence relative to chemical conditions under past or future conditions suggested by hydrochemical models.

Re-evaluation of critical loads/limits criteria

Data collated on three themes under the current programme were presented at the DEFRA workshop on 27th February 2004; reconstructions of pre-industrial ANC, evidence for biological responses to ANC values of 0 and 20 µ eq L-1 in current data and current status of UK freshwaters in terms of ANC and critical load exceedance with these two values of ANCcrit.

It was found that ANC values below 20 µ eq L-1 are rare in unimpacted systems, with very few sites showing evidence for pre-industrial values below this limit in dynamic modelling or palaeolimnological reconstruction exercises. Monitoring data suggest that ANC = 20 µeq L-1 may be an important biological recovery threshold, while zero ANC is associated with a high probability of damage. Critical loads exceedance maps using both 0 and 20 µ eq L-1 are very consistent in pattern and the higher value of ANCcrit does not greatly extend the exceeded areas or numbers of exceeded sites.

Several options for revision of the critical chemical limit for the UK were considered. The consensus of the workshop was that a general ANCcrit value of 20 µ eq L-1 should be applied in the UK, except where site specific modeling data suggest that pre-industrial values may have been lower, in which case 0 µ eq L-1 should be used. These revised criteria were used in the April 2004 freshwater critical load submission to CCE.