Abstract from BTO Research Report No. 464:

I. Henderson, D. Chamberlain, S. Davies & D. Noble (2007)

Changes in Breeding Bird Populations due to Housing Development Based on Bird Densities and Assemblages along Urban-Rural Gradients
ISBN 978-1-906204-18-1

1. EXECUTIVE SUMMARY

1. To date there has been no quantification of net levels of change in biodiversity due to housing expansion into rural areas. This prevents objective and tiered assessments of risk to landscapes, from those supporting high biodiversity to those holding less significant bird communities or populations.
2. This report combines existing bird data for 2000, from the long-term BTO/RSPB/JNCC Breeding Bird Survey (Noble et al. 2006) with ITE/CEH Land Cover Map data for 2000 to quantify and assesses broad-scale differences between suburban and rural bird communities (species composition and abundance of breeding birds), along urban-rural gradients, on a regional basis.
3. Land cover data were used to define squares into suburban and suburban-adjacent rural squares. The results suggest that those rural squares adjacent to suburban areas have less arable agriculture, and may also be less intensively managed, than rural squares ‘isolated’ in the wider countryside (as suggested by the area of set-aside grass). The finding that broadleaved woodlands cover a greater area in those rural squares that are adjacent to suburban areas is important where the potential for urban expansion exists. Analysis by region showed that these patterns were genuine and not caused by large-scale geographic biases in the data.
4. Bird species richness, diversity and individual species density were analysed in two ways in relation to land cover. First, regression analysis was performed on bird data in relation to continuous land cover variables: suburban, urban, woodland and farmland cover. Second, bird data were compared between suburban and ‘rural’ (as defined above) squares using ANOVA. A key focus was the species used in the England Biodiversity Strategy Indicator on wild birds and species on the individual urban specialists House Sparrow, Swift and House Martin.
5. Species richness, diversity and individual species density showed a significant non-linear response to a suburban land cover gradient in most cases, where there was a peak at intermediate levels of suburban cover. This was not the case, however, for several farmland and woodland indicator species. The explanatory power of the models (as measured by adjusted R2) was very low for all species except Collared Dove, Blackbird, Starling and House Sparrow. When other cover variables, apart from ‘urban’, were included in the model, there were marked improvements in adjusted R2 suggesting that the majority of species respond most strongly to woodland and/or farmland cover.
6. For all bird species combined, bird species diversity and richness were both significantly higher on rural squares than suburban squares. For the 27 urban species there was no significant difference in species diversity between rural squares and suburban squares, and species richness was expectedly higher on suburban squares than on rural squares. For individual woodland species, nine occurred at higher density on rural squares than suburban squares and three showed the opposite pattern. Respective figures for farmland species were eight and two, and for urban species, two and 10.
7. Overall, the results indicate a net loss on species richness and diversity when substituting rural habitats for suburban habitats, particularly in the south and east of England. In the Eastern region, there were over eight more species on average in rural than suburban squares for all species and just under five for indicator species. Differences may be particularly pronounced in Eastern England, with high species richness and significant proportions of several species, such as Turtle Dove, Lesser Spotted Woodpecker, Lesser Whitethroat, Nightingale and Yellow Wagtail. This analysis suggests that such species are likely to be at risk from urban expansion, and is particularly important given that the Eastern region is one that is especially likely to be targeted for further housing development.
8. In terms of possible mitigation, it may be that ‘mature’ greenspace is a critical factor that can increase the value of urban developments, although at the same time, occupy more space. We could provide no assessment of the average ‘quality’ of suburban land cover in this analysis and whether this average could be improved upon on a large scale, in order to reduce the impact of urban expansion into rural areas. It is likely that in some cases, mature suburban and rural gardens and parks are able to support ‘acceptable’ populations of several woodland indicator species. However, developments should probably avoid areas of countryside where relatively high populations of farmland and some woodland specialists exist, or where a high potential exists for such habitats to improve in future (due to their proximity to species–rich areas).
9. Further research: Much greater detail, and finer assessments of the impact of urban expansion on rural bird communities are now possible using existing data along with BBS volunteer support. The BTO, in collaboration with Central Science Laboratories (CSL), can derive densities and areas of buildings, greenspace etc., using Mastermap. In addition, CSL access to ‘Defra aerial photographs’ would allow an assessment of habitat ‘composition’. The two data sources, together with the Breeding Bird Survey data would provide a powerful, spatially-linked analysis of urban impact, at a finer resolution than 1-km squares (e.g. 200 m BBS transect sections), incorporating also measures of habitat condition. BBS volunteer support would be available to ‘ground-truth’ electronic data where necessary, while also collecting new bird data. Such a project would provide a strong evaluation of the suburban ‘habitat’, and of variation in and therefore potential for suburban habitats to support or affect bird assemblages. These data would provide more reliable and quantifiable estimates of impact on rural bird assemblages. At the time of writing, a planned programme of research and timed framework was being developed between BTO and CSL.
10. In addition to these data sources, a further Land cover map survey is scheduled for 2007 and if compatibility with LCM2000 is good, then the BTO proposes to use such data to identify impact, through temporal change, for example, in areas that had become developed between survey years.

Back to Research Reports 455 onwards