The recommendations follow the instructions for the London Bird Survey. This London survey was originally co-coordinated by the London Ecology Unit on behalf of the London Biodiversity Partnership.
This method is for counting birds on a walking route. For monitoring purposes, the route is chosen by the recorder, or group of recorders, to be counted monthly over a period of years, so as to track long-term changes in numbers. Observers should be competent at identifying all the species regularly encountered on their route by both sight and sound. Those with less experience of identifying birds will find the method an excellent discipline for learning identification skills, but should register as trainees until they meet the requirements.
Please read this safety information before carrying out a survey.
Some species can only be located at certain times of the year. If you are undertaking a one-off survey of a site you will normally need to make more than one visit to ensure you maximize the proportion of species present at the site that are found during surveying. If you are surveying specific species you should ensure all visits are carried out at the appropriate time of year. If in any doubt consult literature about the species or an acknowledged expert before deciding when to visit.
For ongoing monitoring, the aim is for at least one standard count each month. More frequent counts can be accommodated, and will provide a more precise measure of population changes. It is more important, however, that the counts continue over a long series of years than that they are numerous. Missed months can be ‘imputed’ during analysis, and so gaps are not fatal to the purpose, but the aim should be to count as many months as possible. If the walk is on a site that is subject to the Wetlands Birds Survey, then it is convenient to use the WeBS count dates and include the WeBS data on the standard walk form.
The recording form is carried on a clipboard in the field and the counts are entered directly onto the recording form. Binoculars should be carried to aid identification, although in wooded habitat the majority of registrations will be of birds heard rather than seen.
Standardisation of methods is required so that changes in the counts represent real changes in the birds’ populations, rather than changes in the method. The major requirement for this is that the method, once chosen, is not varied. The recommended methodology for bird surveying is to follow the same walking route on a regular basis.
Choice of site and route
Choice: As the main purpose is to enable regular counting over many years, it is vital that the observer, or team of observers, is happy to continue the survey. For this reason, the chosen sites are likely to be close to home, already under study, or recognized and managed for nature conservation. It is up to the observer to obtain any permission to visit the chosen area.
Length: The method needs a walking route of at least 500 metres. Routes over about four kilometers long will take so long to count that it may be difficult to maintain the counts over the years.
Position: The course of the route should be easily followed on the ground and defined on a large-scale map that is routinely carried in the field. It can be copied onto the back of the recording form. Counting is facilitated by dividing the route into no more than eight sections. It is preferable that these sections correspond with the boundaries of differing bird habitat, existing survey units, or with management ‘parcels’. Most analyses, however, will employ totals from the whole route.
1: The walk
The walk should begin at the same point each time, and cover the sections in the same order. The count should be the sole effort of a single observer (others may accompany the counter for security or social reasons but team counting is not permitted). Walking should be at a comfortably slow pace, with pauses and small diversions as necessary to count and observe the birds, but counting must cease if major departures from the route are undertaken. Variations in the pace should not cause the overall duration of the count to vary more than about 25% either way from the average.
2: Time of day
The recommendation is for the counts to be started at least an hour after dawn and before 11a.m. This avoids great differences in the detectability of birds and the difficulties of sorting out individuals from the dawn chorus. Any other rule is acceptable, if this suits the prime purpose of continued regular counting, but the rule should not be changed once adopted. The times of beginning and ending the count are noted on the recording form. Also circle ‘w’ or ‘s’ on the form to indicate winter or summer time.
Counts should not be undertaken on days with strong winds, thick fog or heavy rain. The recording form has spaces for approximate weather details on the day. Precise details are not required.
The birds are recorded directly onto the recording form (which should be carried on a clip-board in the field). These records should be unambiguous (see specimen example for one method of achieving this). Totals should be calculated shortly after finishing the walk, while the details are still fresh in the memory. There is space on the rear of the form for other issues. It is particularly important to note any changes to the habitat that may influence the counts.
All birds detected are counted, regardless of distance or activity. Some habitats (such as open water or areas of short mown grass) permit full counts, but most records will be of birds detected in places where visibility does not permit full counts. As the count progresses, it is inevitable that there will be doubts over whether some individuals have been counted earlier. Here a simple rule applies: count the birds as new whenever there is reasonable doubt over whether they are the same individuals.
The rationale behind this is twofold. First, it is important that there be no change in the details of method over time that could result in spurious conclusions. Second, is that the method does not seek an absolute number, or census, but rather an index of numbers, for which it is sufficient to hold any bias constant.
If observers are competent and have good faculties, the greatest source of error will be differences in observers’ use of the counting rules. It is desirable, therefore, that different observers counting the same route discuss this issue and compare notes. This may help to ensure that local practice remains the same. The Breeding Bird Survey (BBS) is based on a walk route method using randomly selected 1 kilometre squares across the UK. This includes over a hundred random squares in London allowing trends to be calculated for certain species within London. Anyone interested in taking part should contact the BTO.
Discussion of Other Methodologies
Other methodologies for counting birds should only be used after careful consideration and where it is certain that they meet the aims of the survey. In all cases it should be remembered that accurate long term trends can only be calculated if the methodology can be repeated consistently over a period of time.
The following methods could be considered depending on the circumstances:
1. Site counts: This method aims to count all species at a particular site. This is the method used by the Wetland Bird Survey (WeBS) organised by the BTO. It is appropriate for this survey as wetland birds are usually found in open spaces where they are visible. This method is inappropriate for most other habitats due to the amount of surveyor effort required to achieve a full count and the difficulty in ensuring that effort is consistent from year to year across each of the different habitats at a site. In most instances, a clearly marked walk route as described above is a far more efficient method.
2. Nest counts: This is a site count that concerns either a single species or a small number of species and is an effective method where nests are highly visible and can be easily found (e.g. birds that nest in colonies). Before using this method you should make sure you are aware of any past research on the species that may affect your survey methodology (e.g. House Martin pairs will sometimes build isolated nests away from the main colony so it would be appropriate to widen the survey area.
Use of Collected Data
Raw data can be recorded on a spreadsheet with a column for each species and a row for each day’s count. Analyses are undertaken species by species. ‘Community’ measures like total numbers, richness or diversity are not appropriate with this method. For this the data are linked to another sheet with rows for the months and columns for the years and any missing values are imputed. From this averages can be calculated to illustrate the seasonal changes and trends over the years.
Although statistically sophisticated analyses are possible, the results of these differ little from a simple linear regression of the annual averages against the year. A significant regression coefficient identifies a trend. The confidence limits to the coefficient indicate how much confidence can be put on the estimate of the trend. More sophisticated analyses are justified only when there are results available from many years.
Any trends do not necessarily reflect local changes. Given the mobility of birds, it is essential to show that the local trend differs from that in other parts of London or the country before concluding that the results indicate some local environmental change. This comparison should also correct for the effect of environmental changes that are widespread across England or adjacent regions, such as those resulting from a series of mild winters.
Wider use: This method is designed specifically for surveillance and monitoring of population changes of common bird species over time. In common with other methods developed for this purpose, it is not readily used to provide a ‘complete ‘ list of species using a site, to compare one species with another, nor to provide an estimate of the absolute density or number of any given species.
London Bird Survey Standard Walk Pilot Study Instructions (1999-2000)