monitoring breeding birds in the united kingdom - European Bird ...
monitoring breeding birds in the united kingdom - European Bird ...
monitoring breeding birds in the united kingdom - European Bird ...
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
R. GREGORY, S. BAILLIE & R. BASHFORD , 2004 - Monitor<strong>in</strong>g <strong>breed<strong>in</strong>g</strong> <strong>birds</strong> <strong>in</strong> <strong>the</strong> United K<strong>in</strong>gdom. In: Ansel<strong>in</strong>, A. (ed.) <strong>Bird</strong><br />
Numbers 1995, Proceed<strong>in</strong>gs of <strong>the</strong> International Conference and 13 th Meet<strong>in</strong>g of <strong>the</strong> <strong>European</strong> <strong>Bird</strong> Census Council,<br />
Pärnu, Estonia. <strong>Bird</strong> Census News 13 (2000):101-112<br />
MONITORING BREEDING BIRDS IN THE UNITED KINGDOM<br />
R. D. Gregory 1,2 , S. R. Baillie 1 & R.I. Bashford 1,2<br />
ABSTRACT. A new annual survey of terrestrial <strong>breed<strong>in</strong>g</strong> <strong>birds</strong> was <strong>in</strong>troduced<br />
<strong>in</strong> <strong>the</strong> United K<strong>in</strong>gdom (UK) <strong>in</strong> 1994. The aim of <strong>the</strong> scheme, known as <strong>the</strong><br />
Breed<strong>in</strong>g <strong>Bird</strong> Survey (BBS), is to monitor population levels of a broad<br />
spectrum of common and widespread species across a representative sample of<br />
sites and habitats <strong>in</strong> <strong>the</strong> UK. Population changes will be <strong>in</strong>terpreted <strong>in</strong> <strong>the</strong> light<br />
of habitat and o<strong>the</strong>r environmental data sets. A primary objective of <strong>the</strong> scheme<br />
is to identify species of conservation concern and, <strong>in</strong> comb<strong>in</strong>ation with o<strong>the</strong>r<br />
data from <strong>the</strong> British Trust for Ornithology's Integrated Population Monitor<strong>in</strong>g<br />
Programme, to provide po<strong>in</strong>ters to <strong>the</strong> causes of population changes. Survey<br />
squares, 1 × 1 km squares of <strong>the</strong> UK National Grid, are chosen as a random<br />
sample stratified by potential observer density. <strong>Bird</strong>s are counted twice dur<strong>in</strong>g<br />
<strong>the</strong> <strong>breed<strong>in</strong>g</strong> season us<strong>in</strong>g standardised l<strong>in</strong>e transects. <strong>Bird</strong>s are recorded <strong>in</strong> three<br />
distance categories from <strong>the</strong> transect l<strong>in</strong>e or as <strong>in</strong> flight. Habitat data are<br />
recorded annually for each 200 m section of transect. In <strong>the</strong> first year of <strong>the</strong><br />
scheme, 1994, over 1 500 squares were surveyed and 190 species recorded. A<br />
total of 76 species were recorded from over 100 separate squares. We aim for<br />
<strong>the</strong> sample size to rise to 2-3 000 squares over <strong>the</strong> next few years. Here we<br />
describe <strong>the</strong> background to <strong>the</strong> <strong>in</strong>troduction of <strong>the</strong> BBS, details of survey and<br />
methods, and prelim<strong>in</strong>ary results from 1994.<br />
1. British Trust for Ornithology, The Nunnery, Thetford, Norfolk<br />
IP24 2PU, United K<strong>in</strong>gdom.<br />
2. Current address: The Royal Society for <strong>the</strong> Protection of <strong>Bird</strong>s, The Lodge,<br />
Sandy, Bedfordshire, SG19 2DL United K<strong>in</strong>gdom.<br />
INTRODUCTION<br />
The need to monitor wildlife populations has arguably never been so great with rapid<br />
development, urbanisation and <strong>in</strong>dustrialisation across <strong>the</strong> globe. Effective bird conservation is not<br />
possible without <strong>monitor<strong>in</strong>g</strong> <strong>in</strong>formation to tell us how population levels are chang<strong>in</strong>g, and ideally,<br />
provid<strong>in</strong>g po<strong>in</strong>ters as to why <strong>the</strong>se changes are tak<strong>in</strong>g place. The <strong>monitor<strong>in</strong>g</strong> of bird populations has<br />
<strong>the</strong> added advantage that <strong>birds</strong> can act as valuable barometers or <strong>in</strong>dicators of <strong>the</strong> general health of<br />
<strong>the</strong> countryside (Furness & Greenwood 1993).<br />
Large-scale surveillance of <strong>birds</strong> <strong>in</strong> <strong>the</strong> UK is made possible by <strong>the</strong> number of skilled and<br />
dedicated ornithologists who are will<strong>in</strong>g to participate <strong>in</strong> survey work. Compared with o<strong>the</strong>r taxa,<br />
<strong>birds</strong> are relatively easy to detect and identify, particularly <strong>in</strong> <strong>the</strong> <strong>breed<strong>in</strong>g</strong> season. The British Trust<br />
for Ornithology (BTO) has a long history of <strong>in</strong>volvement <strong>in</strong> <strong>monitor<strong>in</strong>g</strong> and associated research. It<br />
is <strong>the</strong> unique partnership between skilled volunteers and <strong>the</strong> professional staff that sets <strong>the</strong>ir<br />
research programme apart from o<strong>the</strong>r organisations <strong>in</strong> <strong>the</strong> UK. The BTO works closely with o<strong>the</strong>r<br />
organisations concerned with <strong>the</strong> conservation of <strong>birds</strong>, particularly by <strong>the</strong> statutory conservation<br />
bodies for <strong>the</strong> UK, <strong>the</strong> Jo<strong>in</strong>t Nature Conerservation Committee (on behalf of <strong>the</strong> country agensies),<br />
and <strong>the</strong> Royal Society for <strong>the</strong> Protection of <strong>Bird</strong>s (RSPB). The <strong>in</strong>creas<strong>in</strong>g collaboration among<br />
organisations reflects mutual <strong>in</strong>terest and concern for <strong>the</strong> fate of bird populations <strong>in</strong> <strong>the</strong> countryside.<br />
The statutory requirement to monitor <strong>birds</strong> arises <strong>in</strong> law from <strong>the</strong> Wildlife and Countryside Act<br />
1981, <strong>the</strong> <strong>European</strong> Union <strong>Bird</strong>s Directive, and <strong>the</strong> Ramsar, Bonn and Berne Conventions.<br />
- 101 -
The scope of <strong>monitor<strong>in</strong>g</strong><br />
A total of 215 bird species breed <strong>in</strong> <strong>the</strong> UK on a regular basis (Gibbons et al. 1993). Over 80 %<br />
of those species are censused or surveyed on an annual basis through a variety of surveys supported<br />
by government and non-government conservation organisations (Fig. 1: Greenwood et al. 1995).<br />
Specific surveys cover population changes of <strong>the</strong> Heron, wetland <strong>birds</strong> (10 species), sea<strong>birds</strong> (22<br />
species), and rare <strong>breed<strong>in</strong>g</strong> <strong>birds</strong> (56 species: be<strong>in</strong>g def<strong>in</strong>ed as hav<strong>in</strong>g populations of fewer than 300<br />
pairs). We envisage that a small number of <strong>birds</strong> (27 species) can be covered by improved collation<br />
of data collected by bird clubs, reserve wardens and raptor <strong>monitor<strong>in</strong>g</strong> groups (Fig. 1). There is a<br />
group of ten species for which <strong>the</strong>re is no specific survey at present.<br />
Figure 1. Monitor<strong>in</strong>g of terrestrial <strong>breed<strong>in</strong>g</strong> <strong>birds</strong> <strong>in</strong> <strong>the</strong> United K<strong>in</strong>gdom.<br />
The graph shows <strong>the</strong> approximate number of species monitored by various annual surveys<br />
and record<strong>in</strong>g groups. The proposed <strong>monitor<strong>in</strong>g</strong> of species covered by raptor and owls<br />
<strong>monitor<strong>in</strong>g</strong>, and improved data collation, have yet to be implemented. Data from autumn<br />
Wetland <strong>Bird</strong> Survey counts would be used to monitor a small number of <strong>breed<strong>in</strong>g</strong><br />
waterfowl which are not covered by o<strong>the</strong>r schemes.<br />
The rema<strong>in</strong><strong>in</strong>g <strong>breed<strong>in</strong>g</strong> <strong>birds</strong> (89 species) represent <strong>the</strong> common and widerspread species of<br />
<strong>the</strong> countryside and <strong>the</strong>se form <strong>the</strong> focus of BTO's annual <strong>monitor<strong>in</strong>g</strong> programme. At this scale, it is<br />
not feasible nor efficient to count or estimate total population sizes as it is for <strong>the</strong> rarer species, and<br />
<strong>monitor<strong>in</strong>g</strong> must be based on sample surveys. For <strong>the</strong> <strong>in</strong>formation to be of greatest use, it is crucial<br />
for <strong>the</strong> sample of sites to be representative of <strong>the</strong> countryside as a whole. The primary objective of<br />
bird <strong>monitor<strong>in</strong>g</strong>, or more strictly surveillance, is to provide an <strong>in</strong>dex of population changes from<br />
year to year. Although <strong>the</strong> terms <strong>monitor<strong>in</strong>g</strong> and surveillance tend to be <strong>in</strong>terchanged <strong>in</strong> <strong>the</strong><br />
literature, <strong>monitor<strong>in</strong>g</strong> implies (1) <strong>the</strong> measurement of changes aga<strong>in</strong>st predef<strong>in</strong>ed standards of<br />
targets, (2) <strong>the</strong> collection of supplementary data to help <strong>in</strong>tepret and understand population changes,<br />
and (3) that <strong>the</strong>re are clear objectives (Baillie 1990, Furness & Greenwood 1993).<br />
- 102 -
COMMON BIRD CENSUS<br />
The Common <strong>Bird</strong> Census (CBC) has been <strong>the</strong> ma<strong>in</strong> <strong>monitor<strong>in</strong>g</strong> tool for common <strong>birds</strong> <strong>in</strong> <strong>the</strong><br />
UK over <strong>the</strong> last 30 years. The CBC is based on territory mapp<strong>in</strong>g and is one of <strong>the</strong> longest runn<strong>in</strong>g<br />
schemes of its k<strong>in</strong>d <strong>in</strong> Europe, hav<strong>in</strong>g been established <strong>in</strong> <strong>the</strong> early 1960s (Marchant et al. 1990). It<br />
has proved highly valuable <strong>in</strong> advanc<strong>in</strong>g our knowledge and understand<strong>in</strong>g of population<br />
fluctuations among British <strong>birds</strong> (e.g. O'Connor & Shrubb 1986, Marchant et al. 1990, Baillie 1990,<br />
Greenwood & Baillie 1991, Peach et al. 1991, Fuller et al. 1995, Greenwood et al. 1995). There is<br />
general consensus that territory mapp<strong>in</strong>g provides a reasonable estimate of most territorial<br />
populations (Dawson 1981, Verner 1985, Verner & Ritter 1988, Bibby et al. 1992) and <strong>the</strong> aims<br />
and methods of <strong>the</strong> CBC have been reviewed on a periodic basis (O'Connor & Fuller 1984,<br />
Marchant et al. 1990, Baillie & Marchant 1992).<br />
Despite <strong>the</strong> considerable achievement of <strong>the</strong> CBC, <strong>the</strong> scheme has a number of limitations and<br />
<strong>the</strong> BTO has been explor<strong>in</strong>g alternative methods for population <strong>monitor<strong>in</strong>g</strong> for some time (Baillie &<br />
Marchant 1992, Gregory et al. 1994). The limitations of <strong>the</strong> CBC are as follows:<br />
1. Survey sites are not chosen at random. In consequence, <strong>the</strong>re is <strong>the</strong> potential for species,<br />
geographical and habitat bias. In fact, <strong>the</strong> CBC has been restricted to farmland and<br />
woodland, and o<strong>the</strong>r habitats have been excluded. The distribution of survey plots mirrors<br />
<strong>the</strong> distribution of <strong>the</strong> human population and is baised towards <strong>the</strong> south and east of <strong>the</strong><br />
UK.<br />
2. Fieldwork, data compilation and mapp<strong>in</strong>g analysis are extremely time consum<strong>in</strong>g (Gregory<br />
et al. 1994). Fieldwork <strong>in</strong>volves n<strong>in</strong>e or ten visits to a census plot each <strong>breed<strong>in</strong>g</strong> season and<br />
many hours <strong>in</strong> data collation. Note that <strong>in</strong> Brita<strong>in</strong> territory analysis is carried out by BTO<br />
staff ra<strong>the</strong>r than by volunteers, as is <strong>the</strong> case <strong>in</strong> o<strong>the</strong>r countries (van Dijk 1992). The time<br />
required for territory mapp<strong>in</strong>g fieldwork and analysis is roughly seven times that of an<br />
equivalent technique, such as po<strong>in</strong>t counts (Gregory et al. 1994).<br />
Taken toge<strong>the</strong>r, <strong>the</strong>se factors have limited <strong>the</strong> expansion of CBC coverage <strong>in</strong> <strong>the</strong> UK and to<br />
some extent limited <strong>the</strong> use of <strong>the</strong> data, although much can be done at <strong>the</strong> analytical stage to<br />
alleviate problems (Greenwood et al. 1995). Over a number of years <strong>the</strong> BTO has tested a number<br />
of new approaches to population <strong>monitor<strong>in</strong>g</strong> <strong>in</strong> a UK sett<strong>in</strong>g (Gregory & Baillie 2004). The aim is<br />
to develop a <strong>monitor<strong>in</strong>g</strong> scheme based on a formal sampl<strong>in</strong>g strategy for volunteer birdwatchers<br />
which uses simple and efficient methods and thus allows large numbers of people to participate<br />
over a wide geographical area.<br />
ALTERNATIVES TO TERRITORY MAPPING<br />
L<strong>in</strong>e and po<strong>in</strong>t-count transects<br />
The obvious alternatives to territory mapp<strong>in</strong>g are l<strong>in</strong>e or po<strong>in</strong>t count transects, both of which<br />
are relatively efficient and have been used widely <strong>in</strong> population <strong>monitor<strong>in</strong>g</strong> <strong>in</strong> Europe and North<br />
America (Bibby et al. 1992). A number of field trials have tested <strong>the</strong> suitability of <strong>the</strong>se methods<br />
(Gregory et al. 1994, Gregory & Baillie 2004). They showed <strong>the</strong> comparability of <strong>monitor<strong>in</strong>g</strong><br />
results derived from different count<strong>in</strong>g methods, <strong>the</strong> efficiency of <strong>the</strong> methods, and <strong>the</strong><br />
practicability of us<strong>in</strong>g randomly chosen survey sites <strong>in</strong> <strong>the</strong> UK. A l<strong>in</strong>e transect approach was<br />
recommended because it was preferred by volunteers and was slightly more precise. Fur<strong>the</strong>rmore,<br />
l<strong>in</strong>e transects can be applied effetively <strong>in</strong> a wider range of habitats than po<strong>in</strong>t counts (Bibby<br />
et al. 1992).<br />
- 103 -
The f<strong>in</strong>al, and arguably most important, aspect of <strong>the</strong> new scheme was <strong>the</strong> formal sampl<strong>in</strong>g<br />
strategy. Our trials, and similar experience with national surveys of s<strong>in</strong>gle-species (Shrubb & Lack<br />
1991, Donald & Evans 1995), have demonstrated <strong>the</strong> feasibility of a random sampl<strong>in</strong>g design <strong>in</strong> <strong>the</strong><br />
UK. The advantages of this are considerable as it provides <strong>the</strong> only mechanism for produc<strong>in</strong>g what<br />
is demonstably unbiased and thus representative <strong>in</strong>formation (Bibby et al. 1992). A desk-based<br />
study assessed a number of different strategies for <strong>the</strong> selection of survey squares (Gregory &<br />
Baillie 1994, 2004). These simulations showed that simple random sampl<strong>in</strong>g compared very well<br />
with any of <strong>the</strong> more complicated strategies <strong>in</strong>volv<strong>in</strong>g landscape type. It was concluded that a<br />
random selection, stratified by potential observer density, should be used for <strong>the</strong> new survey.<br />
BREEDING BIRD SURVEY<br />
Aga<strong>in</strong>st this background, an annual survey of terrestrial <strong>breed<strong>in</strong>g</strong> <strong>birds</strong> was <strong>in</strong>troduced <strong>in</strong> <strong>the</strong><br />
UK <strong>in</strong> 1994 (http://www.bto.org/survey/bbs.htm). The strategic aims of <strong>the</strong> BBS are:<br />
1) to be based on a formal randomised sampl<strong>in</strong>g design,<br />
2) to improve <strong>the</strong> geographical representation of <strong>monitor<strong>in</strong>g</strong>,<br />
3) to improve <strong>the</strong> habitat representation of <strong>monitor<strong>in</strong>g</strong>, and<br />
4) to <strong>in</strong>crease species coverage of <strong>monitor<strong>in</strong>g</strong> (as a consequence of <strong>the</strong> o<strong>the</strong>r po<strong>in</strong>ts).<br />
The BBS thus aims to provide precise <strong>in</strong>formation on year-to-year and longer term changes <strong>in</strong><br />
population levels for a broad spectrum of common species across a range of regions and habitats.<br />
The parallel record<strong>in</strong>g of land use and habitat change will facilitate a better understand<strong>in</strong>g of <strong>the</strong><br />
factors responsible for population changes, and this will be particularly important for populations <strong>in</strong><br />
decl<strong>in</strong>e. In a wider context, <strong>the</strong> BBS aims to promote a greater understand<strong>in</strong>g of <strong>the</strong> population<br />
biology of British <strong>birds</strong> through a partnership between large numbers of voluteer and a small<br />
number of professional ornithologists.<br />
Selection of survey squares<br />
Survey squares were selected as a random sample, stratified by potential observer density,<br />
from with<strong>in</strong> regions across <strong>the</strong> UK (Gregory & Baillie 1994). The 128 BTO regions were reduced<br />
to 83 regions by <strong>the</strong> amalgamation of <strong>the</strong> smallest ones. Sampl<strong>in</strong>g from fewer, larger regions avoids<br />
<strong>the</strong> problem of extremely small adm<strong>in</strong>istrative regions. The number of squares to be surveyed <strong>in</strong><br />
each region was calculated by multiply<strong>in</strong>g observer desity by a constant and sett<strong>in</strong>g a m<strong>in</strong>imum<br />
level of coverage. We anticipate that professional will be necessary to ensure m<strong>in</strong>imum coverage <strong>in</strong><br />
remote areas.<br />
At a local level squares are allocated to volunteers through a network of voluntary regional<br />
organisers. Each organiser receives a list of target squares for <strong>the</strong>ir region with <strong>the</strong> <strong>in</strong>struction that<br />
squares should be allocated <strong>in</strong> strict order from <strong>the</strong> top downwards. This is essential <strong>in</strong> ma<strong>in</strong>ta<strong>in</strong><strong>in</strong>g<br />
<strong>the</strong> random design of <strong>the</strong> survey. The same squares are surveyed year after year and new volunteers<br />
are found if <strong>the</strong> orig<strong>in</strong>al volunteer drops out. The way <strong>in</strong> which observer changes are handled <strong>in</strong> <strong>the</strong><br />
analysis will be addressed once sufficient data have been ga<strong>the</strong>red. Regional organisers play a vital<br />
role <strong>in</strong> coord<strong>in</strong>at<strong>in</strong>g fieldwork and oversee<strong>in</strong>g <strong>the</strong> return of data. In limited circumstances, <strong>the</strong>y can<br />
report squares as be<strong>in</strong>g uncoverable, for example, because a landowner refuses access permission or<br />
<strong>the</strong> square is complelely <strong>in</strong>accessible. In such cases, <strong>the</strong> square can effectively be ignored, although<br />
organisers submit <strong>in</strong>formation on land use and habitat type to check for any bias <strong>in</strong>troduced by <strong>the</strong>ir<br />
exclusion.<br />
- 104 -
Survey methods<br />
BBS fieldwork <strong>in</strong>volves three visits to a survey square each year. The first visit is to record<br />
details of <strong>the</strong> habitat and to mark <strong>the</strong> survey route (Fig. 2). <strong>Bird</strong> counts are carried out on <strong>the</strong> second<br />
and third visits. The survey route comprises two parallel l<strong>in</strong>es, each 1 km <strong>in</strong> length (although for<br />
practical reasons <strong>the</strong>re is often substantial deviation about <strong>the</strong> ideal). Each of <strong>the</strong>se l<strong>in</strong>es is divided<br />
<strong>in</strong>to five sections, mak<strong>in</strong>g a total of ten 200 m sections, and <strong>birds</strong> and habitats are recorded with<strong>in</strong><br />
<strong>the</strong>se units. Habitat type and land use are recorded annually on a habitat form. This form describes<br />
both <strong>the</strong> habitat surveyed along <strong>the</strong> actual and <strong>the</strong> 'ideal' transect if <strong>the</strong>re is a deviation. By record<strong>in</strong>g<br />
<strong>the</strong> 'ideal' transect route we are able to access whe<strong>the</strong>r observers avoid or prefer particular habitats<br />
with<strong>in</strong> <strong>the</strong>ir squares.<br />
Figure 2. An idealised BBS route through a 1 × 1 km square.<br />
The lower draw<strong>in</strong>gs show proposed survey routes through 1 × 1 km squares.<br />
- 105 -
All <strong>the</strong> survey forms were designed so that <strong>the</strong> data could be readily computerised. For<br />
example, coded <strong>in</strong>formation on species names, county codes, wea<strong>the</strong>r conditions, and habitat type<br />
allow detailed <strong>in</strong>formation to be summarised and <strong>in</strong>put efficiently. Habitat <strong>in</strong>formation is recorded<br />
on <strong>the</strong> first reconnaissance visit to <strong>the</strong> square. The appropriate habitat codes are chosen from an<br />
established hierarchical system (Crick 1992), which is common to a range of BTO schemes.<br />
Observers record <strong>the</strong> primary and secondary habitat for each transect section with up to four levels<br />
of detail (Fig. 3). The primary habitat is <strong>the</strong> dom<strong>in</strong>ant habitat type with<strong>in</strong> a particular transect<br />
section. In <strong>the</strong> example given, transect 1 comprises tilled land with a hedgerow without trees, an<br />
active farmyard, and <strong>the</strong> dom<strong>in</strong>ant crop be<strong>in</strong>g an autumn cereal. There is no secondary habitat<br />
with<strong>in</strong> <strong>the</strong> section and so this space is left blank. Note that <strong>the</strong> ideal transect lies with<strong>in</strong> 20 m of <strong>the</strong><br />
actual route and <strong>the</strong> habitats do not differ (Fig. 2). Transect 2 is a similar area of farmland which<br />
conta<strong>in</strong>s a small woodland. The first habitat codes are <strong>the</strong> same as those above. The second codes<br />
are for a young coniferous plantation with low human disturbance, a moderate shrub layer, and<br />
sparse field layer. Note that <strong>the</strong> ideal route is 50 m from <strong>the</strong> actual survey route and <strong>the</strong> ideal goes<br />
through <strong>the</strong> young plantation.<br />
Figure 3.<br />
An example of a BBS habitat record<strong>in</strong>g form (see text for details).<br />
Count visits are timed so that <strong>the</strong> first is <strong>in</strong> <strong>the</strong> early part of <strong>the</strong> <strong>breed<strong>in</strong>g</strong> season (April to mid-<br />
May) and <strong>the</strong> second <strong>in</strong> <strong>the</strong> late part (mid-May to <strong>the</strong> end of June). It is recommended that visits<br />
should be at least four weeks apart. The average visit time is around 90 m<strong>in</strong>utes. Volunteer counters<br />
are asked to beg<strong>in</strong> <strong>the</strong>ir counts between 6 am and 7 am so that <strong>the</strong>y co<strong>in</strong>cide with maximum bird<br />
activity, but avoid concentrated song activity at dawn (Bibby et al. 1992). Volunteers record all <strong>the</strong><br />
<strong>birds</strong> <strong>the</strong>y see or hear as <strong>the</strong>y walk methodically along <strong>the</strong>ir transect routes. They are encouraged to<br />
- 106 -
pause, listen and scan for <strong>birds</strong> as <strong>the</strong>y walk along <strong>the</strong>ir transects. <strong>Bird</strong>s are noted <strong>in</strong> three distance<br />
categories (with<strong>in</strong> 25m, between 25-100 m, or over 100 m to ei<strong>the</strong>r side of <strong>the</strong> l<strong>in</strong>e) measured at<br />
right angles to <strong>the</strong> transect l<strong>in</strong>e, or as <strong>in</strong> flight. Record<strong>in</strong>g <strong>birds</strong> <strong>in</strong> distance bands is important<br />
because it gives a measure of bird detectability <strong>in</strong> different habitats and allows relative population<br />
density to be estimated us<strong>in</strong>g DISTANCE sampl<strong>in</strong>g (Bibby et al. 1992, Buckland et al. 1993). Note<br />
that <strong>the</strong> estimates of absolute densities derived us<strong>in</strong>g this method are based on a number of key<br />
assumptions.<br />
<strong>Bird</strong>s are noted <strong>in</strong> <strong>the</strong> field on specially designed forms us<strong>in</strong>g two-letter species-codes, which<br />
were developed with<strong>in</strong> <strong>the</strong> CBC (Fig. 4). Observers are encouraged to summarise this <strong>in</strong>formation<br />
on count summary forms as soon after <strong>the</strong> fieldwork as is convenient. Header <strong>in</strong>formation on <strong>the</strong><br />
forms <strong>in</strong>cludes <strong>the</strong> observers' name, address, telephone number, <strong>the</strong> square reference, county code,<br />
date, whe<strong>the</strong>r <strong>the</strong> bird count was early or late <strong>in</strong> <strong>the</strong> season, wea<strong>the</strong>r conditions, and <strong>the</strong> start<strong>in</strong>g and<br />
f<strong>in</strong>ish<strong>in</strong>g times of <strong>the</strong> two halves of <strong>the</strong> transect (Fig. 4). Counties are recorded us<strong>in</strong>g 4-letter codes<br />
which are used across BTO schemes. Wea<strong>the</strong>r codes describe cloud cover, ra<strong>in</strong>, w<strong>in</strong>d, and visibility,<br />
on arbitrary three-po<strong>in</strong>t scale. We discourage bird counts <strong>in</strong> heavy ra<strong>in</strong>, poor visibility, or strong<br />
w<strong>in</strong>ds.<br />
Figure 4. An example of a BBS field record<strong>in</strong>g form and summary sheet (see text for details).<br />
- 107 -
The majority of forms are returned to BTO Headquarters through our network of regional<br />
organisers. Thus organisers are <strong>in</strong> a position to chase up outstand<strong>in</strong>g forms, answer queries and note<br />
<strong>in</strong>terest<strong>in</strong>g observations. BTO acknowledges <strong>the</strong> receipt of each form directly with each observer<br />
and we provide <strong>the</strong>m with <strong>in</strong>formation on census work and related projects on a regular basis. On<br />
receipt, count summary and habitat forms are double-checked by staff for clarity and obvious errors,<br />
before be<strong>in</strong>g sent to be <strong>in</strong>put by an outside agency. Field forms are kept for reference and to allow<br />
any fur<strong>the</strong>r check<strong>in</strong>g. A series of computer progams are used to check <strong>the</strong> computerised data set.<br />
These programs look for <strong>in</strong>consistency or errors <strong>in</strong> <strong>the</strong> different codes (a number of similar species<br />
codes are commonly confused) and check for unlikely bird counts or occurences.<br />
Results from <strong>the</strong> first year<br />
Coord<strong>in</strong>ates for a total of 2 434 1 × 1 km squares were distributed to regional organisers <strong>in</strong><br />
1994. 1 553 (64 %) of <strong>the</strong>se squares were surveyed (Fig. 5), 734 (30 %) were not surveyed and 147<br />
(6 %) were reported as uncoverable. A small number of squares 68 (4 %) were surveyed by<br />
professional teams <strong>in</strong> north and west Scotland funded by RSPB. This is an area with very few<br />
volunteer counters. The map shows an encourag<strong>in</strong>g spread of survey squares although <strong>the</strong>re are a<br />
number of gaps which we <strong>in</strong>tend to fill <strong>in</strong> <strong>the</strong> com<strong>in</strong>g years. The priority <strong>in</strong> 1995 has been to<br />
improve <strong>the</strong> proportion of squares that are surveyed and only those regions achiev<strong>in</strong>g a high level of<br />
coverage <strong>in</strong> 1994 received any new squares for 1995. The eventual sample size of BBS is to be<br />
2 000-3 000 1 km² squares.<br />
Figure 5. Distribution of Breed<strong>in</strong>g <strong>Bird</strong> Survey squares <strong>in</strong> 1994.<br />
The sampl<strong>in</strong>g strategy is designed so that <strong>the</strong>re are larger numbers of squares where <strong>the</strong>re<br />
are more potential observers and <strong>the</strong> map shows a concentration of po<strong>in</strong>ts around large<br />
cities.<br />
- 108 -
The range of species recorded <strong>in</strong> 1994 was extremely broad, <strong>the</strong> total species count was 190,<br />
with 76 species recorded from over 100 squares, and 21 species from between 50-100 squares<br />
(Table 1). This suggests that we would be able to monitor <strong>the</strong>se species with at least a moderate<br />
degree of precision. The most widespread species were Chaff<strong>in</strong>ch Fr<strong>in</strong>gilla coelebs, Woodpigeon<br />
Columba palumbus, Blackbird Turdus merula, and Wren Troglodytes troglodytes. A wide range of<br />
habitats was covered (Table 2). Farmland predom<strong>in</strong>ates with<strong>in</strong> <strong>the</strong> sample, as it dom<strong>in</strong>ates land use<br />
<strong>in</strong> <strong>the</strong> UK (Barr et al. 1993), <strong>the</strong> next most common land uses were human sites (<strong>in</strong>clud<strong>in</strong>g urban,<br />
suburban and rural sites) and woodland, followed a range of rarer land types (Table 2). We expect<br />
<strong>the</strong> coverage of species and habitats to <strong>in</strong>crease modestly as <strong>the</strong> scheme grows.<br />
DISCUSSION<br />
The launch of <strong>the</strong> BBS as an annual bird <strong>monitor<strong>in</strong>g</strong> scheme marks an important and<br />
significant advance <strong>in</strong> British ornithology. In <strong>the</strong> medium- to long-term <strong>the</strong> BBS will take over <strong>the</strong><br />
<strong>monitor<strong>in</strong>g</strong> role of <strong>the</strong> CBC by produc<strong>in</strong>g population <strong>in</strong>dices for a range of common <strong>breed<strong>in</strong>g</strong> <strong>birds</strong>.<br />
Calibration of <strong>the</strong> trends from <strong>the</strong> two schemes will be essential to ma<strong>in</strong>ta<strong>in</strong> <strong>the</strong> long-term time<br />
series <strong>in</strong>itiated by <strong>the</strong> CBC <strong>in</strong> <strong>the</strong> 1960s. We plan to have a considerable period of overlap between<br />
<strong>the</strong> two surveys. It will be some time before <strong>the</strong>re are sufficient BBS data to produce mean<strong>in</strong>gful<br />
time trends and <strong>in</strong> <strong>the</strong> meantime data from <strong>the</strong> CBC will be <strong>in</strong>valuable source of <strong>in</strong>formation.<br />
Fur<strong>the</strong>r work is required <strong>in</strong> <strong>the</strong> development of <strong>the</strong> analysis for year-to-year <strong>monitor<strong>in</strong>g</strong> data<br />
from stratified BBS samples, and longer-term population changes. The national statistics will be<br />
derived from a region-by-region analysis. The fact that different regions have different sampl<strong>in</strong>g<br />
<strong>in</strong>tensities can be <strong>in</strong>corporated <strong>in</strong> <strong>the</strong> analysis and will not bias estimates of change. Population<br />
changes are likely to be weighted by <strong>the</strong> size of regions, sampl<strong>in</strong>g <strong>in</strong>tensity with<strong>in</strong> regions, and <strong>the</strong><br />
number of bird registrations with<strong>in</strong> occupied squares. We will explore <strong>the</strong> use of bird counts <strong>in</strong><br />
different distance categories and <strong>the</strong> use of early and late counts on a species-by-species basis. For<br />
example, it might be sensible to treat residents, early-arriv<strong>in</strong>g migrants and late-arriv<strong>in</strong>g migrants <strong>in</strong><br />
different ways. In this way, it should be possible to maximise <strong>the</strong> precision of between-year changes<br />
for groups of species with particular lifestyles. Recent advances <strong>in</strong> <strong>the</strong> analysis of <strong>monitor<strong>in</strong>g</strong> data,<br />
particularly <strong>the</strong> implementation of methods based on <strong>the</strong> Logl<strong>in</strong>ear Poisson regression (ter Braak et<br />
al. 1994, van Stre<strong>in</strong> et al. 2004) provides great potential for use with BBS data. It is <strong>the</strong> nature of<br />
such time-series that <strong>the</strong> data conta<strong>in</strong> miss<strong>in</strong>g values, are spatially auto-correlated, and that <strong>the</strong> bird<br />
counts will be non-normally distributed. All <strong>the</strong>se factors, however, can be assessed and<br />
<strong>in</strong>corporated <strong>in</strong> <strong>the</strong> regression models developed by van Stre<strong>in</strong> et al. (2004) to create robust <strong>in</strong>dicies.<br />
BBS data will form an <strong>in</strong>tegral part of <strong>the</strong> BTO's Integrated Population Monitor<strong>in</strong>g (IPM)<br />
programme by provid<strong>in</strong>g <strong>in</strong>formation on populat<strong>in</strong> changes among common <strong>birds</strong>. The IPM aims to<br />
develop a better understand<strong>in</strong>g of bird populations by <strong>in</strong>tegrat<strong>in</strong>g <strong>in</strong>formation on bird numbers with<br />
that on survival and <strong>breed<strong>in</strong>g</strong> success (Baillie 1990, Greenwood et al. 1993). It is <strong>the</strong>n possible to<br />
build population models to explore <strong>the</strong> dynamical behaviour of a particular species and build <strong>in</strong><br />
environmental variables such as wea<strong>the</strong>r, <strong>in</strong> order to dist<strong>in</strong>guish natural population fluctuations<br />
from those brought about by <strong>the</strong> actions of man.<br />
- 109 -
Table 1. Species recorded from <strong>the</strong> Breed<strong>in</strong>g <strong>Bird</strong> Survey 1994.<br />
Species are listed <strong>in</strong> descend<strong>in</strong>g order of occurence for A) species recorded from over 100<br />
survey squares and B) species recorded from between 50 and 100 squares.<br />
Species<br />
Proportion<br />
of survey<br />
squares<br />
occupied<br />
(%)<br />
Number<br />
of survey<br />
squares<br />
occupied<br />
A. Over 100 squares occupied<br />
Chaff<strong>in</strong>ch 87 1349<br />
Woodpigeon 87 1342<br />
Blackbird 86 1340<br />
Wren 86 1334<br />
Carrion Crow 83 1290<br />
Rob<strong>in</strong> 82 1265<br />
Blue Tit 79 1220<br />
Starl<strong>in</strong>g 73 1124<br />
Skylark 72 1113<br />
Great Tit 71 1102<br />
Dunnock 69 1076<br />
Swallow 68 1048<br />
Song Thrush 66 1023<br />
Magpie 65 1014<br />
Willow Warbler 62 964<br />
House Sparrow 60 924<br />
Greenf<strong>in</strong>ch 59 909<br />
Pheasant 58 894<br />
Jackdaw 54 835<br />
Yellowhammer 52 806<br />
L<strong>in</strong>net 51 785<br />
Goldf<strong>in</strong>ch 50 769<br />
Rook 49 767<br />
Collared Dove 46 716<br />
Swift 46 710<br />
Blackcap 46 707<br />
Mistle Thrush 46 706<br />
Whitethroat 43 667<br />
Pied Wagtail 43 659<br />
Mallard 42 656<br />
Cuckoo 42 652<br />
Chiffchaff 41 640<br />
House Mart<strong>in</strong> 34 524<br />
Meadow Pipit 33 515<br />
Lapw<strong>in</strong>g 31 487<br />
Long-tailed Tit 30 471<br />
Kestrel 29 453<br />
Stock Dove 29 451<br />
Jay 27 419<br />
Coal Tit 27 411<br />
Black-headed Gull 24 369<br />
Moorhen 24 368<br />
Bullf<strong>in</strong>ch 24 366<br />
Great Spotted Woodpecker 23 362<br />
Green Woodpecker 23 358<br />
Feral Pigeon 23 358<br />
Curlew 23 350<br />
Goldcrest 22 346<br />
Grey Heron 22 337<br />
Species<br />
Proportion<br />
of survey<br />
squares<br />
occupied<br />
(%)<br />
Number<br />
of survey<br />
squares<br />
occupied<br />
contd.<br />
Herr<strong>in</strong>g Gull 20 317<br />
Lesser Black-backed Gull 19 292<br />
Buzzard 18 284<br />
Reed Bunt<strong>in</strong>g 17 271<br />
Garden Warbler 17 265<br />
Red-legged Partridge 17 259<br />
Sparrowhawk 14 220<br />
Treecreeper 14 214<br />
Nuthatch 13 204<br />
Lesser Whitethroat 13 200<br />
Wheatear 12 187<br />
Grey Partridge 12 183<br />
Canada Goose 12 183<br />
Oystercatcher 11 175<br />
Spotted Flycatcher 11 172<br />
Sedge Warbler 10 160<br />
Corn Bunt<strong>in</strong>g 10 151<br />
Turtle Dove 9 139<br />
Yellow Wagtail 8 126<br />
Coot 8 126<br />
Raven 8 118<br />
Tree Sparrow 7 112<br />
Mute Swan 7 112<br />
Grey Wagtail 7 112<br />
Tree Pipit 7 108<br />
Common Gull 7 104<br />
Snipe 7 103<br />
B. Between 50 and 100 squares occupied<br />
Sisk<strong>in</strong> 6 96<br />
Redstart 6 95<br />
Tufted Duck 6 94<br />
Lesser Redpoll 6 90<br />
Marsh Tit 6 88<br />
Shelduck 5 85<br />
Red Grouse 5 81<br />
Golden Plover 5 81<br />
Cormorant 5 80<br />
Hooded Crow 5 72<br />
Little Owl 4 66<br />
Wh<strong>in</strong>chat 4 65<br />
Willow Tit 4 63<br />
Wood Warbler 4 63<br />
Great Black-backed Gull 4 63<br />
Redshank 4 57<br />
Tawny Owl 4 56<br />
Reed Warbler 4 56<br />
Sand Mart<strong>in</strong> 3 54<br />
Common Sandpiper 3 54<br />
Stonechat 3 53<br />
- 110 -
Table 2. Habitat composition of 200 metre transect sections with<strong>in</strong> <strong>the</strong> Breed<strong>in</strong>g <strong>Bird</strong> Survey 1994.<br />
These figures do not reflect <strong>the</strong> habitat composition of <strong>the</strong> UK as a whole as <strong>the</strong>y are not<br />
corrected for regional variation <strong>in</strong> sampl<strong>in</strong>g effort.<br />
Major Habitat<br />
Percentage of transect<br />
sections<br />
Number of transect<br />
sections<br />
Farmland 54.2 8 175<br />
Human Sites 15.7 2 372<br />
Woodland 12.1 1 821<br />
Heathland & Bogs 8.0 1 208<br />
Grassland 5.0 756<br />
Scrubland 2.3 347<br />
Water Bodies 1.7 260<br />
Inland Rock 0.6 83<br />
Coastal 0.4 53<br />
Miscellaneous 0.1 11<br />
TOTAL 100.0 15 086<br />
ACKNOWLEDGEMENTS<br />
We thank <strong>the</strong> many BTO volunteers, both fieldworkers and regional organisers who have<br />
contributed to <strong>the</strong> <strong>monitor<strong>in</strong>g</strong> programmes discussed <strong>in</strong> this paper. The CBC and PCP were<br />
supported under a contract from <strong>the</strong> JNCC (on behalf of English Nature, Countryside Council for<br />
Wales, Scottish Natural Heritage, and under a contract from <strong>the</strong> Department of <strong>the</strong> Environment<br />
(Nor<strong>the</strong>rn Ireland). The project to evaluate sampl<strong>in</strong>g strategies was funded by RSPB. The BBS is<br />
jo<strong>in</strong>tly funded by <strong>the</strong> BTO, <strong>the</strong> JNCC and <strong>the</strong> RSPB. We are grateful to <strong>the</strong> members of <strong>the</strong> BTO's<br />
Integrated Population Monitor<strong>in</strong>g Work<strong>in</strong>g Group, Prof. Steve Buckland, Drs. Rhys Green,<br />
Nicholas Aebischer, John Goss-Custard and Dorian Moss for helpful advice. We also thank Drs.<br />
David Stroud, Ken Smith, David Gibbons, Jeremy Greenwood, Will Peach, and Humphrey Crick<br />
for helpful discussion. John Marchant, Dawn Balmer, Andy Wilson, and <strong>the</strong> late Dr Steve Carter<br />
provided assistance.<br />
REFERENCES<br />
Baillie, S.R. (1990): Integrated population <strong>monitor<strong>in</strong>g</strong> of <strong>breed<strong>in</strong>g</strong> <strong>birds</strong> <strong>in</strong> Brita<strong>in</strong> & Ireland.<br />
Ibis 132:151-166.<br />
Baillie, S.R. & J.H. Marchant (1992): The use of <strong>breed<strong>in</strong>g</strong> bird censuses to monitor common <strong>birds</strong> <strong>in</strong> Brita<strong>in</strong><br />
and Ireland - current practice and future prospects. Vogelwelt 113:172-182.<br />
Barr, C.J., R.G.H. Bunce. R.T. Clarke, R.M. Fuller, M.T. Furse, M.K. Gillespie, G.B. Groom, C.J. Hallam,<br />
M. Hornung, D.C. Howard & M.J. Ness (1993): Countryside Survey 1990. Ma<strong>in</strong> Report. London:<br />
Department of <strong>the</strong> Environment.<br />
Bibby, C.J., N.D. Burgess & D.A. Hill (1992): <strong>Bird</strong> Census Techniques. London: Academic Press.<br />
ter Braak, C.F.F., A.J. van Strien, R. Meijer & T.J. Verstrael (1994): Analysis of <strong>monitor<strong>in</strong>g</strong> data with many<br />
miss<strong>in</strong>g values: which method In Hagemeijer, E.J.M. & T.J. Verstrael (eds.): <strong>Bird</strong> Numbers 1992:<br />
Distribution, <strong>monitor<strong>in</strong>g</strong> and ecological aspects. Proc. 12th Int. Conf. IBCC and EOAC,<br />
Noordwijkerhout, The Ne<strong>the</strong>rlands: 663-673. Beek-Ubbergen: SOVON.<br />
Buckland, S.T., D.R. Anderson, K.P. Burnham & J.L. Laake (1993): Distance sampl<strong>in</strong>g: estimat<strong>in</strong>g<br />
abundance of biological populations. London: Chapman & Hall.<br />
Crick, H.Q.P. (1992): A bird-habitat cod<strong>in</strong>g system for use <strong>in</strong> Brita<strong>in</strong> and Ireland <strong>in</strong>corporat<strong>in</strong>g aspects of<br />
land-management and human activity. <strong>Bird</strong> Study 39:1-12.<br />
Dawson, D.G. (1981): The usefulness of absolute ("census") and relative ("sampl<strong>in</strong>g" or "<strong>in</strong>dex") measures of<br />
abundance. Studies <strong>in</strong> Avian Biology 6:554-558.<br />
- 111 -
van Dijk, A.J. (1992): The Breed<strong>in</strong>g bird <strong>monitor<strong>in</strong>g</strong> programme of SOVON <strong>in</strong> <strong>the</strong> Ne<strong>the</strong>rlands. Vogelwelt<br />
113:197-209.<br />
Donald, P.F. & A.D. Evans (1995): Poplation size and habitat selection of Corn Bunt<strong>in</strong>g Miliaria calandra<br />
<strong>breed<strong>in</strong>g</strong> <strong>in</strong> Brita<strong>in</strong> <strong>in</strong> 1993. <strong>Bird</strong> Study 42: 190-204.<br />
Fuller, R.J., R.D. Gregory, D.W. Gibbons, J.H. Marchant, J.D. Wilson, S.R. Baillie & N. Carter (1995):<br />
Population decl<strong>in</strong>es and range contractions among lowland farmland <strong>birds</strong> <strong>in</strong> Brita<strong>in</strong>. Conservation<br />
Biology 9, 1425-1441.<br />
Furness, R.W. & J.J.D. Greenwood (eds., 1993): <strong>Bird</strong>s as monitors of environmental change. London:<br />
Chapman & Hall.<br />
Gibbons, D.W., J.B. Reid & R.A. Chapman (1993): The New Atlas of Breed<strong>in</strong>g <strong>Bird</strong>s <strong>in</strong> Brita<strong>in</strong> and Ireland:<br />
1988-1991. London: Poyser.<br />
Greenwood, J.J.D. & S.R. Baillie (1991): Effects of density- dependence and wea<strong>the</strong>r on population changes<br />
of English passer<strong>in</strong>es us<strong>in</strong>g a non-experimental paradigm. Ibis 133 (suppl.): 121-133.<br />
Greenwood, J.J.D., S.R. Baillie, R.D. Gregory & W.J. Peach (1995): Some new approaches to conservation<br />
monitorng of British <strong>birds</strong>. Ibis 137 (suppl.): 16-28.<br />
Greenwood, J.J.D., S.R. Baillie, H.Q.P. Crick, J.H. Marchant & W.J. Peach (1993): Integrated population<br />
<strong>monitor<strong>in</strong>g</strong>: detect<strong>in</strong>g <strong>the</strong> effects of diverse changes. In Furness, R.W. & J.J.D. Greenwood (eds.):<br />
<strong>Bird</strong>s as Monitors of Environmental Change: 267-342. London: Chapman & Hall.<br />
Gregory, R.D. & S.R. Baillie (1994): Evaluation of sampl<strong>in</strong>g strategies for 1-km squares for <strong>in</strong>clusion <strong>in</strong> <strong>the</strong><br />
Breed<strong>in</strong>g <strong>Bird</strong> Survey. Thetford: BTO Research Report 139.<br />
Gregory, R.D. & S.R. Baillie (2004): Survey design and sampl<strong>in</strong>g strategies for <strong>monitor<strong>in</strong>g</strong> <strong>breed<strong>in</strong>g</strong> <strong>birds</strong>.<br />
<strong>Bird</strong> Census News, Vol. 13, 2000 (2004): 19-31.<br />
Gregory, R.D., J.H. Marchant, S.R. Baillie & J.J.D. Greenwood (1994): A comparison of population changes<br />
among British <strong>breed<strong>in</strong>g</strong> <strong>birds</strong> us<strong>in</strong>g territory mapp<strong>in</strong>g and po<strong>in</strong>t-count data. In Hagemeijer, E.J.M. &<br />
T.J. Verstrael (eds.): <strong>Bird</strong> Numbers 1992: Distribution, <strong>monitor<strong>in</strong>g</strong> and ecological aspects. Proc.<br />
12th Int. Conf. IBCC and EOAC, Noordwijkerhout, The Ne<strong>the</strong>rlands: 503-512. Beek-Ubbergen:<br />
SOVON.<br />
Marchant, J.H., R. Hudson, S.P. Carter & P. Whitt<strong>in</strong>gton (1990): Population trends <strong>in</strong> British <strong>breed<strong>in</strong>g</strong> <strong>birds</strong>.<br />
Tr<strong>in</strong>g: British Trust for Ornithology.<br />
O'Connor R.J. & R.J. Fuller (1984): A re-evaluation of <strong>the</strong> aims and methods of <strong>the</strong> Common <strong>Bird</strong> Census.<br />
Tr<strong>in</strong>g: BTO Research Report 15.<br />
O'Connor R.J. & M. Shrubb (1986): Farm<strong>in</strong>g and <strong>birds</strong>. Cambridge: Cambridge University Press.<br />
Peach, W.J., S.R. Baillie, L.G. Underhill (1991): Survival of British Sedge Warblers Acrocephalus<br />
schoenobaenus <strong>in</strong> relation to west African ra<strong>in</strong>fall. Ibis 133 (suppl.): 300-305.<br />
Shrubb, M. & P.C. Lack (1991): The numbers and distribution of Lapw<strong>in</strong>gs Vanellus vanellus nest<strong>in</strong>g <strong>in</strong><br />
England and Wales <strong>in</strong> 1987. <strong>Bird</strong> Study 38: 20-37.<br />
van Strien, A.J., E.J.M. Hagemeijer & T.J. Verstrael: Assess<strong>in</strong>g trends <strong>in</strong> <strong>birds</strong> <strong>monitor<strong>in</strong>g</strong> data.<br />
<strong>Bird</strong> Census News, Vol. 13, 2000 (2004)<br />
Verner, J. (1985): Assessment of count<strong>in</strong>g techniques. In Johnston, R.F. (ed.): Current Ornithology 2:247-<br />
302. New York: Plenum Press.<br />
Verner, J. & L.V. Ritter (1988): A comparison of transects and spot mapp<strong>in</strong>g <strong>in</strong> Oak-P<strong>in</strong>e woodlands of<br />
California. Condor 90: 401-419.<br />
- 112 -