3. BIRD CENSUSES

3.1 Common Birds Census (CBC) and Waterways Bird Survey (WBS)

The CBC and WBS, established in 1962 and 1974 respectively, monitor annually population levels of common birds (c.35% of the c.200 species breeding annually in the UK). They are based on the territory mapping method, which provides good indices of populations from the c.200 CBC plots and c.100 WBS plots each year (Marchant et al., 1999). Both the spatial distribution of birds and habitat details of the plots are recorded, so habitat relations of the birds can be analysed (e.g. Donald & Forrest, 1995).

The schemes focus on habitats - lowland farmland, lowland woodland, and linear waterways - where birds were considered to be most under threat at the time the schemes were set up. Farmland CBCs in southern and eastern Britain have been shown to be typical of farmland in that area. We have previously shown that the habitat and geographical distribution of plots has remained constant, that trained analysts are consistent in their interpretation of the data, and that there is no tendency for plots that are losing birds at more than the average rate to be more likely to be dropped from the sample.

One of the key findings from the CBC has been the discovery of dramatic population declines among common, mostly farmland, birds over the last 25 years (Marchant & Gregory, 1994; Fuller et al,. 1995; Baillie et al., 1997; Siriwardena et al.,1998). The CBC data have played a major role in recent revisions to the conservation priorities of birds in the UK, including the lists of Birds of Conservation Importance, Birds of Conservation Concern and the Biodiversity Steering Group List. Its most high profile use has been as the core terrestrial data for one of the government’s Quality of Life Indicators, published by DETR in September 1998. Mountford indices for 73 species monitored by CBC or WBS are included in the calculation of the headline wild bird indicator, as well as the supplementary woodland and farmland birds indicators that showed the steepest declines (section 2.3).

BTO has developed and implemented a modification of the Mountford method for indexing population trends (with confidence limits) more precisely than traditional methods. Recent work has explored the use of other statistical methods, such as log linear poisson regression, which uses the data in an efficient manner and has the potential to model covariate effects. Further work has developed statistical methods to identify, objectively, (i) similarities between the trends of different species or taxanomic groups and (ii) periods in which species have been increasing or declining (Siriwardena et al.,1998).

3.2 Breeding Bird Survey (BBS)

Despite the considerable achievements of the CBC and WBS, these schemes have limitations, the most serious being the lack of a predefined sampling framework. CBC plot distribution is biased towards south-east Britain and only farmland and woodland habitats are covered. Territory mapping is very time-consuming for both fieldworkers and analysts and this has limited its expansion. The BTO explored alternative methods of monitoring common and widespread species over several years, culminating in the introduction of the BTO/JNCC/RSPB Breeding Bird Survey in 1994 (Gregory et al.,1996; Noble et al., 1999). The BBS is a simpler survey that is less time-consuming for volunteers and can be processed more efficiently by BTO HQ. The main aims of the new monitoring scheme were to:

1. Improve geographic representativeness.
2. Improve habitat coverage.
3. Increase species coverage.

BBS was introduced in 1994, as a joint partnership between BTO, JNCC and RSPB. It is based on a formal randomised sampling design, with stratification by observer density. Volunteer birdwatchers count all birds seen and heard along two standardised line transect counts per year in 1x1km squares. The same sample of squares is surveyed each year. The number of squares covered in the UK has increased considerably since its inception, from 1,565 in 1994 to 2,307 in 1998. The long-term goal is to survey 2,000-3,000 squares each year, with at least minimum coverage across all UK regions. The BBS has been run in parallel with CBC for a number of years, and research is being initiated to analyse the period of overlap to develop methods for calibrating the data from these two schemes so that long-term trends can be linked.

Details of survey design, species, habitat and regional coverage can be found in the BBS annual reports (e.g. Gregory et al., 1996; Noble et al., 1999). The most recent report contains population changes between 1994 and 1998 in the UK, as well as separate tables of indices for trends in Scotland and Wales. Trends are modelled using log linear poisson regression with weighting to account for the stratified sampling design. Reliable indices for a total of c.100 species are generated using BBS data. Because this is a relatively new scheme, we were particularly interested in evaluating the design of the survey, from data collection to analytical methods. A study undertaken in 1998 to examine the effects of observer deviations from the ideal transect to the actual transect found that there was a slight habitat bias away from farmland towards boundaries such as roads (Field & Gregory, 1998). In another study, we examined our procedures for data extraction and some of the assumptions of the log linear regression model that we use to calculate annual population changes (Field & Gregory, 1999). Timing of visits was found to be highly consistent between years, and confirmed that a maximum count for each species over the two visits is the most effective way of counting both residents and migrants. A linear trend model was found to indicate broadly similar population changes to the standard year-by-site index model, but more sophisticated models will be developed as more years of data become available.

Annual habitat recording is an important part of the design of the scheme, because it measures environmental change with which to interpret changes in bird populations. Associated research examining bird-habitat relations using BBS data has made considerable progress in quantifying habitat use. One paper on broad-scale habitat use of some declining British birds has been published (Gregory & Baillie, 1998) and another paper on habitat use by granivorous birds is in press.

The recording of a number of mammal species was introduced in 1995 and continues to be well-supported by volunteers. About 40 species of mammals are currently detected on BBS squares, and recent analyses suggest that BBS mammal recording has the potential to detect a 25% decline over 25 years, for about 15 species.

3.3 Single Species Surveys

Species living at low densities, those with highly clumped distributions, and those with special nesting requirements are often poorly covered by general census schemes. We monitor such species through special surveys - either one-off surveys or repeated surveys at fixed intervals. Preliminary results from the just completed survey of Nightingales have yielded an estimate of between 3,500 and 4,000 singing males. Recent evaluations of wader populations include a survey of Lapwings undertaken in spring 1998, and a survey of breeding waders in the Outer Hebrides. Other surveys conducted in the last five years in the breeding season include, Skylark (1997), Woodlark (1997), Rook (1996), Red-breasted Merganser and Goosander (1987 & 1997) and Barn Owl (1994-97). The Skylark survey of 1997/98 is the most recent of the non-breeding surveys.

3.4 Heronries Census

Begun in 1928, a sample survey is conducted annually in England and Wales, with periodic full censuses (extended to Scotland in 1985). This is the longest-running breeding bird population monitoring scheme in the world. A statistical model for indexing the population, initially developed with the University of Dundee, has recently been applied to the entire dataset. A NERC-funded project to update and computerise historical Heronries data and produce a definitive set of national and regional trends has been completed and a draft report is being circulated.

3.5 Raptor Censuses

Raptor population monitoring requires methods adapted to the secrecy and low population density of raptors, to their vulnerability to persecution, and to the number of specialist organisations and individuals involved in their study. BTO has led national surveys of a number of raptor species, often in collaboration with local Raptor Study Groups. Perhaps the most famous series of surveys has been the 10-yearly censuses of Peregrines, that established the depth of the species’ decline due to the effects of organochlorine pesticides and charted the species spectacular recovery.

The most recent survey has been a three-year national survey of Barn Owls, in collaboration with the Hawk and Owl Trust, based on a random stratified sample of 1,100 tetrads. This has provided the first replicable survey for the Barn Owl, which is on the Biodiversity Steering Group Conservation Concern List. The survey also provided population estimates for Little Owl.

3.6 Waterways Breeding Bird Survey (WBBS)

Pilot work on a new survey of birds using linear waterways was carried out in 1998 (Marchant & Gregory, 1999). Using a modified BBS method with transects along randomly-selected stretches of waterway, as well as a sample of canals, observers surveyed 169 waterways in 1998. Analyses revealed that WBBS yielded proportionally more data on riparian species (including Dipper, Kingfisher, Goosander, Little Grebe, Mute Swan & Grey Wagtail) than the BBS, and covered considerably more species than WBS. A comparison of bird abundance on canals with a close season of coarse fishing, to those without, revealed no significant impact (Marchant et al., 1999). The WBBS will be continued in 1999 and 2000, and estimates of bird abundance will be related to the habitat data from the River Habitat Survey data collected by the Environment Agency.

Key References

Baillie, S.R., Gregory, R.D. & Siriwardena, G.M. (1997). Farmland bird declines: patterns, processes and prospects. In: BCPC Symposium Proceedings No. 69: Biodiversity and Conservation in Agriculture (ed. R.C. Kirkwood), pp. 65-87. British Crop Protection Council, Farnham.

BTO, HOT, NT, NTS, RSPB, SRSG, WRSG, WTs., SOC, WWT, WWF-UK. (1997). Birds of Prey in the UK: Back From the Brink. BTO et al., Thetford.

Carter, S.P., Gregory, R.D. & Baillie, S.R. (1997). Abundance, distribution and habitat use of breeding Goosanders Mergus merganser and Red-breasted Mergansers Mergus serrator on British Rivers. Bird Study, 44(1): 1-12.

Donald, P.F. & Evans, A.D. (1995). Habitat selection and population size of Corn Buntings Miliaria calandra breeding in Britain in 1993. Bird Study, 42: 190-204.

Donald, P.F. & Forrest, C. (1995). The effects of agricultural change on population size of Corn Buntings Miliaria calandra on individual farms. Bird Study, 42: 205-215.

Field, R.H. & Gregory, R.D. (1998). Habitat Bias in Actual and Ideal Transect Lines in Breeding Bird Surveys, 1994-1997. BTO Research Report No. 214. BTO, Thetford.

Field, R.H. & Gregory, R.D. (1999). Measuring Population Changes from the Breeding Bird Survey. BTO Research Report No. 217. BTO, Thetford.

Fuller, R.J., Gregory, R.D., Gibbons, D.W., Marchant, J.H., Wilson, J.D., Baillie, S.R. & Carter, N. (1995). Population declines and range contractions among farmland birds in Britain. Conservation Biology, 9: 1425-1441.

Gregory, R.D. & Baillie, S.R. (1998). Large-scale habitat use of some declining British birds. Journal of Applied Ecology, 35: 785-799.

Gregory, R.D., Bashford, R.I., Balmer, D.E., Marchant, J.H., Wilson, A.M. & Baillie, S.R. (1996). Breeding Bird Survey 1994-1995. Annual Report No. 1. BTO, Thetford.

Marchant, J.H. & Gregory, R.D. (1994). Recent population changes among seed-eating passerines in the United Kingdom. In: Bird Numbers 1992: Distribution, Monitoring and Ecological Aspects (eds. E.J.M. Hagemeijer & T.J. Verstrael), pp. 87-95. Statistics Netherlands, Voorburg/Heerlen & SOVON, Beek-Ubbergen.

Marchant, J.H. & Gregory, R.D. (1999). Waterways Breeding Bird Survey Pilot Study 1998: Adaptation of BBS Census Methods to Rivers and Canals. BTO Research Report No. 218.

Marchant, J.H., Gregory, R.D., Balmer, D.E. Gough, S.J. & Wilson, A.M. (1999). A Comparison of Breeding Bird Numbers Along Canals With and Without a Close Season for Fishing. R&D Technical Report W193. Environment Agency, Bristol.

Marchant, J.H., Sanderson, F. & Glue, D. (1999). Changes in breeding bird populations, 1997-98. BTO News, 222: 10-14.

Noble, D.G., Bashford, R.I., Marchant, J.H., Baillie, S.R. & Gregory, R.D. (1999). The Breeding Bird Survey 1998. Annual Report No. 4. BTO Research Report No. 225. BTO, Thetford.

Siriwardena, G.M., Baillie, S.R., Buckland, S.T., Fewster, R.M., Marchant, J.H. & Wilson, J.D. (1998). Trends in the abundance of farmland birds: a quantitative comparison of smoothed Common Birds Census indices. Journal of Applied Ecology, 35: 24-43.

Sitters, H.P., Fuller, R.J., Hoblyn, R.A., Wright, M.T., Cowie, N. & Bowden, C.G.R. (1996). The Woodlark Lullula arborea in Britain: population trends, distribution and habitat occupancy. Bird Study, 43: 172-187.

Toms, M. (1998). Box 1: Project Barn Owl. BTO News, 223: 11.

Back to Research Summary Contents