Raptor Population Estimates

by Humphrey Q P Crick

Although the information presented below comes from published sources, it should be noted that the majority derives from surveys undertaken by volunteer ornithologists, particularly those in Raptor Study Groups around the country. Without their high level of enthusiasm for often arduous fieldwork and their support for the objectives and aims of the conservation and conservation science organisations in the UK who need the data, the important information presented below would not be available.

Honey Buzzard Pernis apivorus

Country breakdowns are not available because of confidentiality. This species is under-recorded and under-reported. In 2000, the Rare Breeding Birds Panel (RBBP) coordinated a national survey by calling for more recording and submission of data. The Panel contacted all current workers on the species and provided guidance on how to find Honey Buzzards. However, data were not obtained from all areas and a further 32 pairs or single birds were reported but without nests being confirmed, so the above total is certainly too low. Roberts et al. (1999) made an informal estimate of 50-60 territorial pairs. Non-breeding birds are observed but it is not clear how many there are. RBBP records suggest a large population increase over the past 30 years but some of this increase is likely to be due to better reporting and increased surveying in addition to a real population increase.

Red Kite Milvus milvus

Production of eggs was confirmed by either the presence of a bird sitting on a nest, or a nest seen to contain eggs or young. The figures are reliable: the distribution of this species is well-known and most of the population is monitored each year. In 2000, all occupied areas outside Wales and the Chilterns (the two strongholds) were surveyed; in the Chilterns, a full survey was made of the core area and of sample squares around the edge; in Wales stratified samples were made of both the core and the range edge. The normal age at first breeding is two years, so there must be some non-breeders in the population.

White-tailed Eagle Haliaeetus albicilla

All known territories were recorded and a large area of potential range checked in early spring; it is possible that there were a very few pairs undetected in remote locations. The whole population is closely monitored. The total population is thought to be 80-90 birds and it has increased from two nesting pairs in 1983 (Bainbridge et al., in press).

Marsh Harrier Circus aeruginosus

Females are the appropriate count unit because the species can be polygynous. In 1995 Underhill-Day (1998) organised a full census (based on an exhortation for more complete recording by observers) that provided a higher GB population estimate of 156 females producing eggs was. This probably underestimated the true population because of missed early failures (it only accepted records where breeding was proved) and because it was not based on any formal sampling strategy. Nonetheless, RBBP figures for the years around 1995 were 20-25 lower, showing the even greater extent of under-recording by RBBP for this species. From available information, the total population of full-grown birds present is around 2.4 times greater than the number of breeding females. The core areas are well-monitored, however, so the RBBP data do provide reliable trends and show that the population increase since the 1970s is now tending to plateau off.

Hen Harrier Circus cyaneus

The UK survey in 1998 combined areas covered completely by RSGs with randomly selected 10km squares from the rest of the likely range; the intended coverage was almost complete. It is almost impossible to estimate the number of non-breeders because this species is so widely persecuted (Etheridge et al. 1997): but the number of single birds recorded in the 1998 survey, which is the minimum estimate of the number of non-breeders, averaged 0.28 per territorial pair (Sim et al. in press). Between 1988/9 and 1998 the size of the total population in Great Britain did not change, though there were marked regional changes.

Montagu's Harrier Circus pygargus

Based on records routinely submitted to county recorders; the extent of under-recording and under-reporting is unknown. It is not clear how many non-breeders there are in Britain, but in 1999, there were 8 males reported to RBBP. It is not clear how much of the small increase in reports to the RBBP is a result of better reporting.

Goshawk Accipiter gentilis

The 1995 estimate was based on collation of data from various correspondents, who covered most of the species' British range at the time, a very small proportion of the territories may have been occupied by singletons. The species suffers from serious under-recording and under-reporting. In a population on Gotland, Sweden, only 56% of all males and 33% of all females attempt to nest (Kenward et al. 1999), so there is a total of 4.8 birds for each occupied home range. The non-breeding population in Britain is likely to be slightly higher because of higher male survival, possibly occasioned by greater availability of prey and milder winters than in Gotland (Robert Kenward pers. comm.). The species has increased substantially in numbers in Britain since its re-establishment in 1950 (Petty 1996), but because of widespread under-reporting, the details of the increase are unknown.

Sparrowhawk Accipiter nisus

The GB estimates were based on extrapolation from intensive study areas, which provided information on the relationship between Sparrowhawk numbers, altitude and forest area (Newton 1986), using Forestry Commission data on the extent of forests at different altitudes. Pairs with new nests provide a good index because the nest remains recognizable as such for the rest of the season, and in subsequent years, giving tangible evidence of a breeding attempt long after the builders have given up (Ian Newton pers. comm.). Recent demographic evidence is that there are 0.28 non-nesting females for each nesting female; because the birds are monogamous, there are probably similar numbers of males, so the total population is c.1.28 times greater than the number nesting (Newton & Rothery in press). CBC and BBS data indicate that Sparrowhawk numbers have stabilized or even declined since the mid 1990s.

Buzzard Buteo buteo

Soaring birds were counted in a stratified random sample of tetrads and the number of breeding pairs taken to be half the number of soaring birds recorded (24500-33800), though with the caveat that this might underestimate numbers because some females may have been incubating during the later stages of the survey. However, such counts cannot be considered to provide more than a rough estimate of the number of territorial pairs, although they are a reasonable basis for monitoring changes over time. Coverage was only 74%, with evidence of bias towards occupied squares, so the true figure was probably in the lower end of the range given (Taylor et al. 1988). Overall, the total population, including non-breeders, is probably more than twice the number of territorial pairs (Kenward et al. 2000). CBC, BBS and local studies show that populations are increasing and spreading their range

Golden Eagle Aquila chrysaetos

All but 13 home ranges and all potential nesting areas within the species British range were systematically surveyed. The 13 home ranges not visited might have held 3 pairs. A home range was considered to be occupied by a pair if two adult birds were seen together or if there was evidence of eggs having been laid. An additional 69 home ranges held single birds (Green 1996). It is likely, however, that singletons will have been under-recorded, so Watson's (1997) estimate (derived from intensive sample surveys) that 15% of the population may be non-breeders is probably a better figure. The size of the British population did not change between 1982 and 1992, though there was some regional variation.

Osprey Pandion haliaetus

There is an annual census of all known sites (which are consistently re-used), plus searching of likely areas and checking of sightings by the public, so only a few sites may be missed each year. It should be noted that two pairs nested in England in 2001. The demography of this population has not been sufficiently worked out to attempt estimates of how many non-territory-holders there are in Britain. Ospreys were at very low numbers or extinct in Scotland during 1917-53 but have increased rapidly following their recolonization in 1954 (Dennis 1995).

Kestrel Falco tinnunculus

The figures are based on densities observed in intensive study areas and the proportion of the country occupied by the species. Village (1990) estimated that at least 30% of a tagged population in lowland English farmland was non-breeding (i.e. did not belong to a pair that produced eggs). In the same area, direct counts showed mean densities of breeders and non-breeders to be 47 and 32 birds/100 km2, a proportion of 32% non-breeders. Kestrels had recovered from the deleterious effects of organochlorine pesticides by the mid-1970s but subsequently declined rapidly and are now no more common than in the early 1960s. Furthermore, although the CBC indicates that abundance has been stable for the last 15 years, the BBS suggests that a further decline has occurred since 1994.

Merlin Falco columbarius

The GB total of c1300 was thought to be accurate within +/-200 (Rebecca & Bainbridge 1998). In areas where long-term intensive studies guaranteed 100% coverage, independent checks during the 1993-4 census showed that they provided unbiased estimates of population size. In the rest of the country, a random selection from a regular grid of survey squares was surveyed using standardized techniques, and population size was extrapolated from these data. The NI estimate was based on unpublished data from RSGs and RSPB fieldworkers. Greenwood et al. (in press) calculated that there might be as many as 800 non-breeding birds in addition to those with nests. Between the British surveys of 1983-4 and 1993-4, Merlin numbers increased in most areas and were stable elsewhere (Rebecca & Bainbridge 1998).

Hobby Falco subbuteo

The RBBP figures are almost certainly under-estimates: Prince & Clarke (1993) estimated (by density extrapolation) that the population in Cambridgeshire was twice that reported by RBBP. It is likely that a range of 500-1000 occupied territories is reasonable, although this upper limit might not be great enough. Please note that 1990 was the first year since 1887 that Hobbies appeared to hold territory in Scotland, and 1994 had the only confirmed egg-laying yet; between 1990-1999, up to 6 occupied territories have been recorded in Scotland (RBBP figures). Greenwood et al. (in press) calculate that there may be between 250-500 non-breeding birds in addition to territory holders. Both CBC and RBBP suggest a sustained population increase over the past 30 years.

Peregrine Falco peregrinus

The UK survey achieved almost full coverage except in north and west Scotland. It attempted to visit all known sites and also to check potentially suitable areas, emphasis being given to checking potential sites early in the season to assess site occupancy before any pairs departed their sites. Extrapolations of occupancy in unsurveyed areas were based on occupancy rates in the surrounding region. The non-breeding population of singletons may be equal to the number of territorial pairs (Ratcliffe 1995). Decadal surveys show that the pesticide-associated declines of the 1950s were largely reversed by 1981 and by 1991 most populations were higher than ever before recorded, though not in north and west Scotland (Crick & Ratcliffe 1995).

Barn Owl Tyto alba

The British figures represent the first replicable baseline population estimate, based on a stratified randomized sample survey with validation of the results by independent observers. The total of 4000 has approximate 95% confidence limits of +/-1200. In many areas, this species relies on prey whose numbers fluctuate markedly. In poor years many may not breed but in good years most do so (Taylor 1994). If we take this into account, then the total number of individuals (breeding and non-breeding) in the population may be up to 2.6 times the number of pairs with eggs. The proportion of CBC plots on which Barn Owls occur has declined from the mid-1960s and shown some recovery since the late 1980s.

Little Owl Athene noctua

Likely nest sites were examined in a stratified random sample of tetrads, aimed at surveying Barn Owls. Observers were not asked to provide evidence of breeding and it is possible that some cavities used only for roosting or for preparing food (Hibbert-Ware 1938) were included in the count, which may therefore overestimate breeding numbers. Coverage was incomplete but did not vary with stratum and there was no evidence of serious bias. The non-breeding population is unknown. CBC data show some variation in numbers over time but no significant long-term trends, but crepuscular species such as this are poorly monitored by general schemes such as CBC.

Tawny Owl Strix aluco

A national survey in 1989 suggested an average GB density of 10 territories per 10-km square and was used to derive estimates in Gibbons et al. (1993). Reasonable margins of error on the estimate are 15,000-40,000 (S.M. Percival, pers. comm.). Most Tawny Owls that fail to secure a territory in their first autumn or winter starve to death (Southern 1970, Hirons et al. 1979), so the number of territories is a good measure of the total population at the start of the breeding season. But note that on average (with very marked variations), one-third of territory-holders do not lay eggs (Southern 1970), so the number of territories is much greater than the number of pairs producing eggs. The CBC shows some variation in numbers but no significant long-term trends, but poorly monitors nocturnal species.

Short-eared Owl Asio otus

These are probably the most unreliable estimates of any except those for the next species. Data on this species are so scant that the distinction between territory-holding, nest-building and egg-laying pairs cannot usually be made. The British estimate is based on the number of 10-km squares with evidence of occupancy and breeding, multiplied by general knowledge of typical densities in occupied squares. The Welsh figure is consistent with more direct information (Lovegrove et al. 1994). Information on non-breeders and trends are unreliable.

Long-eared Owl Asio flammeus

The numbers of this species are probably less well known than those of any other British owl or raptor. The British estimate is based on the number of 10-km squares with evidence of occupancy and breeding, multiplied by general knowledge of typical densities in occupied squares. The estimate from Wales is probably too great, as only four confirmed breeding pairs were found during atlas-fieldwork (Lovegrove et al. 1994); that for the Isle of Man fits with the numbers routinely reported (Craine & Moore 1997; Cullen 2000). Information on non-breeders and trends are unreliable.

Go to Page 3 - Raptor Population References