Home > Research > Services > Report Abstracts > 236

Abstract from BTO Research Report No 236:

Chamberlain, D.E., Warren, R.W., Crick, H.Q.P., Hall, J., Metcalf, S., Ormerod, S., Whyatt, D. & Vickery, J.A.

Acidification and terrestrial birds.

Executive summary

1. The UK Government has recently signed up to the 1999 UN ECE Protocol to Abate Acidification, Eutrophication and Ground-level Ozone under the Convention on Long- Range Transboundary Air Pollution. A more stringent strategy for EU countries, the National Emissions Ceilings Directive, is under development. While much is known about impacts of acidification on aquatic ecosystems in the UK, there is a real need for robust indicators that integrate the effects on terrestrial ecosystems. They should be able to measure the impact of abatement strategies and be readily understood by the general public.

2. Acidification reduces calcium availability in aquatic and terrestrial ecosystems, resulting in decreased productivity and reduced population sizes of some bird species. This project assesses the potential use of birds as cost-effective bio-indicators that integrate the effects of acidification within terrestrial ecosystems. It combines information from uniquely detailed historical data sets for bird distribution, abundance and productivity with those for acid deposition and critical loads. It investigates the extent of past and present effects of acidification on terrestrial birds and investigates whether the data sets can form the basis of a cost-effective bio-monitor of terrestrial acidification. It concentrates on those relationships that show adverse effects of acidification because these will provide the best potential candidates as bio-monitors of acidification abatement.

3. Acidification data used in the analysis:
(a) UK Critical Loads. These were supplied by the Centre for Ecology and Hydrology. The critical loads define the sensitivity of the land to acid deposition of both Sulphur and Nitrogen species. Critical loads for all ecosystems combined for each 10-km square were used in the analysis to incorporate a measure of acid sensitivity in conjunction with accumulated exceedance (see below).
(b) UK Deposition of Sulphur and Nitrogen. Deposition fields, including backcasts where required, were produced by the HARM model (University of Edinburgh) on a 10-km scale for 1955, 1970, 1983 and 1990. The lack of a reliable spatially disaggregated historical emissions inventory for oxidised or reduced Nitrogen in the UK limited the search for correlations between bird data and deposition, constraining it to the use of Sulphur deposition only.
(c) UK ‘Exceedance Ratio’. Accumulated exceedances (AE) of critical loads for all ecosystems combined were calculated on a 10-km scale by the Centre for Ecology and Hydrology for 1955, 1970, 1983 and 1990 using the HARM modelled deposition fields. It was considered reasonable to use the combination of the historical N deposition together with the more reliable historical S deposition to calculate historical changes in AE. A unit of AE in a sensitive area might be expected to be more damaging than a unit of AE in a less sensitive area. Thus after exploring a number of methods of using the exceedance data in the analysis, the ratio of the AE: 5% CLSMAX was calculated and used in the final analysis.

4. Bird data used in the analysis:
Distribution data was taken from the two Breeding Bird Atlases, the first in 1968-72 and the second in 1988-91. The second atlas contains more detailed information concerning frequency of occurrence. Data used consist of:
(a) the frequency of occurrence of species in 10-km squares in the 1988-91 atlas; and
(b) the presence/absence of species in 10-km squares in both atlas periods. This latter data set was used to determine the number of 10-km squares experiencing losses (local extinction) and gains (local colonisation) between the two periods.
Data on reproductive performance was analysed with respect to acidification and other environmental variables. Six measures of reproductive performance were considered: lay date, clutch size, brood size, partial brood loss, clutch failure rate and brood failure rate.

5. In the analyses of bird distribution and breeding performance, regression models were initially constructed that contained just environmental variables (e.g. geographical and habitat variables). We then tested whether Sulphur deposition and exceedance ratio made any significant additional contributions to the models, thereby attempting to control for any confounding environmental factors.

6. The hypothesis that anthropogenic acidification has had a detrimental effect on bird species distributions in Britain was tested by analysing three ornithological data sets in relation to (1) Sulphur deposition and (2) exceedance ratio:
(a) frequency of species occurrence in 1988-91 (a measure of relative abundance in 10 km squares);
(b) species loss in 10-km squares between 1968-72 and 1988-91 (change in presence/absence in 10-km squares);
(c) species gain in 10-km squares between 1968-72 and 1988-91.
Nine species were selected for analysis that were invertebrate feeders and that had shown a change in range of at least 10% between the two atlas surveys. The species represented a range of ecological requirements.

7. Species distributions were significantly affected by a range of variables including latitude, altitude and habitat type. Significant additional contributions of either Sulphur deposition (SDep)or exceedance ratio (ER) were identified for all of the species selected for analysis, thus supporting the hypothesis of a detrimental effect of acidification on distribution:
(a) Occurrence in 1988-91 was less likely in 10-km squares with greater acidification for Redshank (SDep), Dipper (SDep), Stonechat (ER), and Lesser Whitethroat (ER).
(b) Species extinction between 1968-72 and 1988-91 was less likely for Lapwing (SDep & ER), Redshank (SDep), Dipper (SDep), Redstart (SDep) and Ring Ouzel (SDep) in 10-km squares showing a decrease in acidification.
(c) Species colonisation over the same period was more likely for Little Ringed Plover (SDep) and Lesser Whitethroat (SDep) in 10-km squares showing a decrease in acidification.

8. In addition to relationships showing an apparently detrimental effect of acidification, there were a number that showed apparently beneficial relationships, e.g. where species occurrence was linked to areas where exceedance ratio was high. We tabulated the numbers of relationships showing significant detrimental and apparently beneficial effects on species distributions from all analyses involving combined habitat and acidification variables. For exceedance ratio, 4 relationships showed a detrimental effect, 9 a beneficial effect and 13 were not significant or inconclusive due to significant interaction terms. For Sulphur deposition, the respective figures were 13, 5 and 8. Thus there is little general evidence that exceedance ratio affects species in the expected direction, but Sulphur deposition appears to be more consistently detrimental with respect to species distribution.

9. There was no evidence that species richness in 1988-91 was lower in acidified areas, but richness of resident insectivorous, migrant insectivorous and resident omnivorous passerines increased significantly in 10-km squares that had experienced the greatest decreases in acidification.

10. Forecast reductions in acidification over the next decade were available: (1) the REFERENCE scenario, and (2) the WGS31B scenario, upon which the Gothenburg Protocol is based. These forecasts were used to predict changes in species range by 2010 from the models derived for Sulphur deposition, the most consistent predictor. Lapwing, Little Ringed Plover and Lesser Whitethroat showed the greatest predicted changes, with increases in range of 13%, 13% and 9% respectively, compared to 1990, under the WGS31b scenario. These species were most sensitive and therefore likely to be the best candidates for bio-monitors of acidification. Species richness was predicted to change by only an average of 1 species per 10-km square under the same scenario. Although the REFERENCE scenario produced results 20-40% smaller than the WGS31B scenario, the standard deviations were relatively large and the absolute sizes of the differences in predicted changes were relatively small, the most extreme being 5% (for Lapwing).

11. There were a number of species that showed distinct geographic regions of high sensitivity to acidification. However, there were no areas that were consistently the most sensitive across species. Given the specific nature of the models for each species or species group and the lack of consistent spatial trends, it is difficult to identify regions where there is high sensitivity to changes in acidification across a range of species. Any consideration of areas in which birds are particularly sensitive to acidification within a monitoring framework would have to be undertaken on a species-by-species basis.

12. The effects of acidification, as measured by critical load exceedance and acid deposition, were analysed with respect to reproductive performance for four invertebrate-feeding passerines, Dipper Cinclus cinclus, Song Thrush Turdus philomelos, Ring Ouzel T. torquartus and Great Tit Parus major.

13. Measures of breeding performance were significantly affected by a range of variables including latitude, longitude and altitude. These significant predictors were included in the models before the addition of acidification variables. The majority of relationships between aspects of breeding performance and acidification were curvilinear in nature.

14. These regression models were used to predict effects of reductions in acidification forecast for 2010 under the WGS31B scenario on breeding performance. Predicted decreases in breeding performance were more common than increases (7 versus 3) but the latter were greater in magnitude. Brood survival rates in Dipper and Ring Ouzel were predicted to increase by 23% and 13% respectively compared with 1990 rates and Song Thrush clutch survival was predicted to increase by 7% under a scenario of reduced acid deposition by the end of the decade. Although the predictions under the REFERENCE scenario were smaller in magnitude than for the WGS31B scenario, the associated standard deviations were relatively large compared to the absolute size of the differences.

15. Overall, in both the investigation of bird distributions and breeding performance, detrimental impacts were more frequently detected using Sulphur deposition data than those based on exceedance ratio. Furthermore, species distribution was more sensitive to acidification than breeding performance. Thus the effects of Sulphur deposition on bird distributions are likely to form the most useful potential bio-monitors for acidification abatement strategies in the UK.

16. While breeding performance measured by the British Trust for Ornithology's Nest Record Scheme may prove to be a useful bio-monitor of acidification in the future, the species investigated here appeared not to be especially sensitive and further investigations are required to identify more sensitive species. It would be important to investigate the breeding performance of species showing significant relationships between distributions and acid deposition at a finer scale (i.e. intensive field studies).

17. The atlas data sets showed a number of apparently detrimental associations between Sulphur deposition and bird species distributions. This is despite the lack of sensitivity inherent in this presence/absence data set. It would be especially valuable to follow up these investigations by analysis of the BTO/JNCC/RSPB Breeding Bird Survey (BBS) and the Waterways Breeding Bird Survey (WBBS) data sets. This provides annually recorded, spatially referenced relative abundance data. Such data could be used to measure changes in presence/absence in 1-km survey squares, or they could be used to measure changes in relative abundance, which may be more sensitive to changes in acid deposition and might correlate with exceedance ratio. The BBS could provide a useful, frequently updated, system for tracking the responses of British bird populations identified in this study to reductions in Sulphur and Nitrogen emissions in the future.

18. Due to the slow recovery time of ecosystems to reductions in deposition, future monitoring of bird species sensitive to acid deposition may be more likely to indicate ecosystem recovery in the period starting in 1990, as opposed to the period from 1970 to 1990. The time delay between reductions in critical load exceedances and ecosystem recovery may be considerable and it will be important to monitor ecosystem recovery in the future. A future study could link to a dynamic modelling exercise for the UK using a model such as MAGIC which takes into account such temporal aspects. This may be greatly facilitated by the recent interest shown in dynamic modelling across Europe (for example at the recent Task Force on Integrated Assessment Modelling in Stockholm) which may result in the generation of detailed additional data for the UK.

Back to Research Reports 159-247