Bird populations and environmental change Module 5 final report

by
A.M. Wilson & R.M. Fuller
Contact: Andy Wilson (andy.wilson@bto.org). British Trust for Ornithology, The Nunnery, Thetford, Norfolk, IP24 2PU, UK

Contents

• Executive Summary - HTML
• Part 1: Main report (text) - PDF, 165 Kb
• Part 2: Main report (figures) - PDF, 343 Kb
• Appendix 1: survey instructions - PDF, 455 Kb
• Appendix 2: population density estimates - PDF, 42 Kb
• Appendix 3: inter set correlations of environmental variables with axes - PDF, 33 Kb
• Appendix 4: CCA ordination diagrams - whole survey - PDF, 29 Kb
• Appendix 5: CCA ordination diagrams - English squares - PDF, 24 Kb
• Appendix 6: CCA ordination diagrams - Scottish squares - PDF, 25 Kb
• Appendix 7: CCA ordination diagrams - Welsh squares - PDF, 23 Kb
• Appendix 8: CCA ordination diagrams - Environmental Zone 1 - PDF, 24 Kb
• Appendix 9: CCA ordination diagrams - Environmental Zone 2 - PDF, 23 Kb
• Appendix 10: CCA ordination diagrams - Environmental Zone 3 - PDF, 22 Kb
• Appendix 11: CCA ordination diagrams - Environmental Zone 4 - PDF, 20 Kb
• Appendix 12: CCA ordination diagrams - Environmental Zone 5 - PDF, 22 Kb
• Appendix 13: CCA ordination diagrams - Environmental Zone 6 - PDF, 20 Kb
• Appendix 14: CCA ordination diagrams - 20 farmland indicator species - PDF, 8 Kb
• Appendix 15: CCA ordination diagrams - 38 woodland indicator species - PDF, 9 Kb

Note: Because of it's size, the report is split into 17 separate PDF files. The Executive Summary is also provided separately, as HTML text.

Note: you will need Adobe® Acrobat Reader®, if you don't have this installed on your computer, click here. 

For viewers with visual difficulties, information on the services provided to improve the accessibility of Acrobat documents is available from the following web site: http://access.adobe.com

• Go to the Module 5 (Bird populations and countryside change) information page

   Click here to visit the BTO web site

Executive Summary

1. Birds are potentially valuable indicators of some aspects of the quality of the countryside for wildlife.  Furthermore, trends in bird populations form one of the 15 Quality of Life headline indicators published annually by the UK Government.  It is important to understand how bird populations are affected by spatial differences across the countryside in habitat availability and to changes across time in the quantity and quality of these habitats.  This project, which was Module 5 of Countryside Survey 2000 (CS2000), sets a new baseline for achieving this.  The bird counts were undertaken during the spring and summer of 2000.

2. The main aim of the CS2000 bird counts was to estimate the abundance of breeding birds in a large sample of the 1-km squares for which detailed information on land use, vegetation and habitat features were collected during the CS2000 Field Survey.  These data were used to provide quantitative descriptions of bird assemblages at three different scales: Great Britain, individual countries of Great Britain and Environmental Zones.  The six Environmental Zones are aggregations of ITE Land Classes chosen to reflect major environmental variation in the UK.  These Environmental Zones were those used as a regional framework for reporting the main results of CS2000.  The CS2000 bird counts were also used to generate Avifaunal Zones, each of which contained a distinctive breeding assemblage of birds.  Preliminary analyses were undertaken of gradients and patterns in bird populations in 1-km squares in relation to summary data on landscape and habitat composition of individual squares.

3. Bird surveys were conducted using up to 4 km of line transect counts within each 1-km square, using the same methods to those adopted by the BTO/JNCC/RSPB Breeding Bird Survey (BBS).  Numbers of birds were recorded separately for 200 m sections of the transects according to three distance bands either side of the observer.  Densities of most bird species were estimated for Great Britain, individual countries and Environmental Zones using Distance Sampling techniques.  An index of the abundance of each species within individual 1-km squares was derived from the frequency with which it was recorded in the 200 m transect sections i.e. the proportion of sections occupied.  A comparison of density estimates and frequency indices for all species showed a close correspondence between these two measures with only colonial species (e.g. House Martin and several seabirds) or very widespread but low density species  (e.g. Cuckoo) emerging as outliers.

4. A total of 336 squares was surveyed, with samples of between 29 and 93 in each of the Environmental Zones.  The data were contributed by a combination of volunteer and contract workers.  A total of 171 species was recorded.  The birds detected in these squares were a good representation of terrestrial bird communities in Great Britain.  There was a strong relationship between the number of each species recorded on the CS2000 bird counts and the latest British population estimates.  The only species that were markedly under-recorded were certain colonial nesting seabirds.  The dominant species in each of the Environmental Zones are listed.   Chaffinch and Wren are the only species to feature in the top 10 for each of the six Zones while Meadow Pipit was the most abundant species in the three Zones covering the uplands and marginal uplands.

5. Cluster analysis suggested that there were similarities in bird assemblages between Environmental Zones 1 and 2 (English and Welsh lowlands), zones 3 and 4 (the English and Welsh uplands and the Scottish lowlands) and zones 5 and 6 (the Scottish uplands).  There was a marked difference between Zones 1 and 2 and the remaining four Zones, indicating a fundamental divergence in the bird assemblages between the English and Welsh lowlands and the rest of Britain.  However, Environmental Zone 6 was also found to differ strongly from the other zones in that the bird community there was dominated by relatively few species.

6. Avifaunal Zones were developed using cluster analysis to aggregate Land Classes into zones where bird assemblages were similar.  This analysis suggests that between five and eight Avifaunal Zones are identifiable, which are broadly similar to the Environmental Zones in that they show strong north-south and altitudinal differentiation.  It is suggested that the concept of avifaunal zones is explored further, possibly using independent datasets, such as the New Atlas of Breeding Birds and the BTO/JNCC/RSPB Breeding Bird Survey.

7. The relationships between bird assemblages and environmental variables were explored in a preliminary way using Canonical Correspondence Analysis applied to the frequency indices.  All these analyses were undertaken at the scale of individual 1-km squares.  These analyses effectively identified major gradients in birds assemblages, in terms of species composition and the relative abundance of species, and identified those habitats and environmental attributes most closely related to these gradients.  The results are presented mainly in the form of separate ordination diagrams for Great Britain, individual countries and Environmental Zones.

8. A great deal of detailed information for each individual species concerning their broad relationships with environmental variables can be derived from these diagrams.  The patterns highlighted by these analyses were different at the three spatial scales and between the six Environmental Zones.  At the scale of Great Britain, invariants, such as climate, altitude, easting and northing, were found to be very important determinants of bird assemblages.  However, at the country and Environmental Zone scales, the effects of other factors, especially broad habitats could be seen more clearly.  Dominant land-uses and habitats, notably the areas of arable, human settlement and broad-leaved woodland were frequently among the most important features affecting gradients in bird assemblages.  The relationships between bird assemblages and environmental variables varied considerably between Environmental Zones.

9. Analyses were conducted for the 20 farmland and 40 woodland species that contribute to the Government’s Quality of Life headline indicator.  These analyses explored relationships between these species and a large suite of environmental variables, but excluding the invariants such as climate and topography.  Farmland species that have not decreased since the 1970s tend to be ones associated with improved grassland or areas with substantial quantities of woodland or human settlements.  Those farmland species that have declined strongly are often ones associated with arable-dominated landscapes.  The woodland species show a more complex pattern with declining species being associated with a diversity of landscapes and woodland types.

10. Most of the environmental variables found to be significant in determining bird communities are already recorded systematically for the BTO/JNCC/RSPB Breeding Bird Survey.  One possible exception is the length of linear features, such as hedgerows, dykes, walls and roads, which may be inadequately recorded by the BBS.

11. Our work could be readily repeated as part of any future Countryside Survey, particularly as all the transect routes were mapped.  Repeating the bird counts alongside the Field Survey in future Countryside Surveys would enable changes in bird numbers to be related to changes in land cover and habitat quality as measured by the Field Survey.  It is also recommended that careful consideration be given to how the Countryside Survey data could be more effectively integrated with the nationwide atlases of breeding and winter bird distribution that are periodically compiled by the British Trust for Ornithology.

12. The strength of the CS2000 bird data lies in understanding how large-scale pattern in the landscape affects bird communities rather than in understanding the micro-habitat relationships of birds or the effects of habitat management.  The data can be used to model relationships between land-use and bird populations in ways that allow predictions of the broad consequences for bird populations of large-scale shifts in future land-use.  Much previous work of this type has focused on lowland farmland habitats but the CS2000 data could be used to address a range of broad strategic land-use and habitat creation issues in both the lowlands and uplands.  Examples of possible applications are given and recommendations are made about how this might be taken forward.  Models need to be scale- and region-specific and they should draw on data from both the 1-km squares and 200 m transect sections.  The models should incorporate CS2000 Land Cover Map data as well as the CS2000 Field Survey data.  The possibility of incorporating data from the CS2000 Vegetation Plots within 1-km squares into the models should also be explored.