|
|
|
BBS Research Projects Spatial variation in abundance From a conservation viewpoint, knowledge of species abundance at a regional level is crucial for managing biodiversity and protecting declining or endangered species effectively. However, the reliability and accuracy of this knowledge is dependent on the quality of the surveys collecting the data. The BBS is now the main source of continuous monitoring data for common and widespread species in the UK and as such, biologists are increasingly relying on this data to guide management actions. However good the survey, data of this type will always be patchy and incomplete, and it is important for us to understand how this data reflects the real distribution and abundance of the species surveyed. Because many bird species are distributed with respect to habitat (or altitude or climate or space), it is important to examine whether geospatial data at this scale can be used to improve our predictions. In addition, an understanding of the relationships between bird distribution and abundance, with habitat and other environmental factors, would allow us to predict the effect of land-use or climate change on bird species. Two complementary studies examining spatial variation in abundance on BBS squares are discussed below. BBS habitat data is a useful predictor of bird abundance BBS data are regularly used to estimate changes in numbers over time, but this dataset can also be used to explore differences in abundance across different landscapes and regions. In work carried out over the last year, we examined the influence of geographical location and habitat type on counts of birds on BBS squares. As would be expected from differences in breeding ranges, most species varied considerably in abundance in relation to latitude, longitude and altitude. It was therefore important to correct for these effects in models that assess the influence of habitat type. Applying these corrections, we then compared predictions of abundance based on BBS habitat data (collected by the volunteers themselves) to those based on independent landscape classifications from the Centre of Ecology and Hydrology (CEH). Results showed that BBS habitat data collected at the 200 m transect section scale were good predictors of bird abundance and were slightly better than predictions based on broader-scale CEH data. This is a gratifying result, highlighting the value of BBS habitat data, which is not only collected at a finer scale than categories based on mapping or satellite imagery, but which is collected annually and hence very up-to-date. One disadvantage of BBS habitat data is that they are only available for the BBS squares surveyed. It is therefore important to link these data to the best available landscape information for the entire country, in order to make predictions of bird abundance in areas outside surveyed squares. The ability to assess where species occur in greatest numbers and where and in which habitats they are declining has important conservation implications. Freeman, S.N., Noble, D.G., Newson, S.E.
& Baillie, S.R. (2003). Importance of geographical location
and local habitat features for species abundance: analyses using
Breeding Bird Survey data. BTO research Report No. 320. British
Trust for Ornithology, Thetford. Mapping species abundance The last Breeding Bird Atlas of 1988-91 presented maps of species abundance for all widespread bird species in Britain and Ireland (Gibbons et al. 1993). Abundance maps of this type are of huge importance, not only in highlighting the strongholds of particular species or change maps showing areas of significant population change, but they allow information such as this to be made accessible to a much wider audience than would normally be possible. Over the last ten years, there have been considerable advances in the development of software that integrate geostatistics within a Geographic Information System (GIS). In work to examine the potential of these recent software advances, we applied these techniques to produce smoothed distributions of abundance using counts on BBS squares. Geostatistical methods are based on statistical models that model autocorrelation (statistical relationship between counts of birds in this case). Not only do these techniques have the capability of producing a prediction surface, but they can also provide some measure of the accuracy of the predictions. A number of geostatistical interpolation techniques have been developed, of which kriging is the most applicable to this work. Kriging weights the surrounding counts at surveyed sites to derive a prediction for unsurveyed locations. In these, the weights are based on the distance between measured sites and the prediction location, but also on the overall spatial arrangement in the weights (the spatial autocorrelation). For a full discussion of geostatistics and geostatistical methods see Chiles & Delfiner (1999). In the analyses here we have used Simple Kriging and to control for the stratified sampling design of the BBS we use the method of declustering, which preferentially weights the count data, with counts in densely sampled areas receiving less weight and counts in sparsely sampled areas receiving greater weight. This decides how much the data at each site contributes to the calculation of autocorrelation functions across the entire data set. Once a model has been developed, a prediction of mean relative abundance for each 10-km square was made. The data used to produce these maps were the maximum total counts of each species in each BBS square surveyed in 2003. Hence, the map represents the abundance of birds counted during BBS line transect surveys (and hence they are not necessarily all breeding in 2003). The values in the map legends are relative indices related to the mean predicted counts in BBS squares. Results showed that it was possible to produce maps that matched well the expected distribution and abundance for species that are well monitored by the BBS. To take two examples, Meadow Pipit abundance is highest in upland areas of Scotland, Wales and the Peak District, whilst the Collared Dove has its highest abundance in suburban and rural areas around London, in Norfolk from where the species was first recorded, and around Birmingham and Manchester. A quick glance at the last Breeding Bird Atlas will confirm the good match, for these two species at least.
To view relative abundance maps for other species click here Chiles, J. & Delfiner, P. 1999. Geostatistics. Modeling Spatial Uncertainty. John Wiley & Sons, New York. Newson, S.E. & Noble, D.G. (2003) Producing
statistically valid maps of species abundance from UK Breeding Bird
survey counts using Geostatistical Analyst in ArcGIS. BTO Research
Report No. 318. British Trust for Ornithology, Thetford.
|
||||||
|
|
| BTO
Home | JNCC
Home | RSPB
Home |
|