Back to Gulls home-page
Gulls Home Page

 Yellow-legged Gull Larus michahellis

 Yellow-legged Gull photo index
Yellow-legged Gull Photo Index


Regular visitor in small numbers in late summer and winter - possibly more than one phenotype involved.


Introduction

Status in the West Midlands Region


Photo Galleries


Introduction

The appearance of Yellow-legged Gulls Larus michahellis in Britain is now a well-established phenomenon. Yellow-legged Gull is a member of the southern group among the 'Herring Gull assemblage', and breeds in the Mediterranean basin, west to the Atlantic seaboard, and east to the southern Black Sea. After breeding it undertakes a moult migration to the north, bringing it in some numbers to western Europe and the SE corner of Britain, and in rather smaller numbers further north and west, including the English West Midlands. For a full discussion of the taxonomy and identification of Yellow-legged Gull and other species or 'forms' within the group, see Yesou (2002) and Garner et al. (1997) but a brief over-view follows.

Historically, Yellow-legged Gull was considered to be a subspecies of the Herring Gull but more recent morphological and behavioural studies indicated that it should be regarded as a good species in its own right. This was further confirmed by DNA studies, which showed that michahellis is genetically differentiated from argentatus. See also comments introducing the Caspian Gull account.

Compared with Herring Gull, Yellow-legged Gull is on average more elegant, with a fuller, squarer, more flat-topped head-shape and relatively longer wings and legs. The orbital ring of Yellow-legged Gull is generally blood-red, creating a darker-eyed appearance compared with Herring Gull. The upperparts are a darker shade of grey than Herring Gull, obviously so compared with British argenteus though approached by darker examples of nominate argentatus. The wing-tip pattern displays more black and less white than Herring Gull. There is generally black on six, sometimes seven, outermost primaries. The black band across the tip of p5 is generally quite thick and unbroken, while the extent of black on the outer primaries as a whole is more extensive than on nominate argentatus. In Scandinavian argentatus there is less black on the outer primaries, often with a full white tip to p10 and no black band on p5 (but see comments below on on yellow-legged Herring Gulls from the Baltic region). The pattern of black is not, however, so distinct from that of argenteus, which also has a black band on p5, but that race has relatively pale grey upperparts so that, generally, confusion should not arise.

In adult Yellow-legged Gulls the legs are a rich yellow, though frequently less intense during the winter months and throughout the year in sub-adult birds. The yellow legs in combination with the relatively dark grey upperparts are often the features which initially attract attention on a standing bird. Note, however, that Herring Gulls sometimes display yellow-legs, particularly individuals of the nominate form from certain areas of Scandinavia and the eastern Baltic region. Individuals from the Baltic region also regularly show a black band on p5, though in most individuals this is narrow compared with michahellis. Also, there is less black in the outer primaries as a whole, with more white at the tip of p10 (sometimes a full white tip) and pale tongues on the inner webs, sometimes approaching a thayeri-pattern on p10. Thus, the pattern is more likely to simulate the wing-pattern of Caspian Gull than Yellow-legged Gull but, as there is black on the outer six primaries, an individual in which the extent of black is most well-developed could be confused with a Yellow-legged Gull. Thus, it is essential to use a full suite of characters when identifying Yellow-legged Gull. The definitive reference for yellow-legged Herring Gulls from the Baltic region is the 1998 paper by Lars Jonsson : 'Yellow-legged Gulls and yellow-legged Herring Gulls in the Baltic' (Alula 4: 74 - 100).

In winter, the head of michahellis remains relatively white and unstreaked compared with the generally well-streaked and messy head-pattern of Herring Gull, though when freshly moulted in autumn michahellis often does show limited head-streaking. This tends to take the form of rather neat streaking around the eye and across the centre of the crown, and this in itself can be quite distinctive. [Conversely, Herring Gulls can become very white-headed by mid to late December, and such individuals can look very different from surrounding Herring Gulls with well-streaked heads. There is a danger of such birds being misidentified as Yellow-legged Gulls by the incautious.]

The timing of the full post-breeding moult in michahellis is somewhat earlier than in Herring Gull, reflecting it's slightly earlier breeding season. Birds appearing in Britain in late June and early July generally show obvious signs of primary moult, with one or more central primaries 'missing' (in fact present but only part grown or just out of 'pin' and so hidden by overlapping 'inner' primaries). When preening, the state of moult and the pattern of the primary tips can often be determined.

Status in the West Midlands Region

In recent years, Yellow-legged Gulls have been arriving in Britain in increasing numbers in late summer and early autumn, as part of the species' northward  post-breeding moult migration. The largest numbers are to be found in the south-east coastal counties but smaller numbers regularly reach the West Midlands Region. Intriguingly, comparable numbers are observed in the midlands during winter, whereas on the south coast the majority of Yellow-legged Gulls disperse in late autumn and few are recorded during the winter months. It is possible that individuals recorded in the midlands during winter originate from a different population from those occurring in late summer, and arguably there are phenotypic differences between 'typical' individuals observed during the two seasons.

The first fully-documented Yellow-legged Gull to be detected in the West Midlands area was an adult at Chasewater as long ago as November 1973, though a similar individual had been reported by A. R. M. Blake the previous year (Hume 1978). At that time, the identification and taxonomy of gulls with yellow legs among the 'Herring Gull assemblage' were poorly understood, and, although Hume discussed a number of forms including michahellis, the identity of the Chasewater individual was not attributed explicitly to any particular taxon. In the following years, observations of such individuals became increasingly regular in southern England, and their identification features were extensively clarified. Major advances in understanding of the systematics of the 'large white-headed gulls' also ensued, though universal agreement on the taxonomic status of the various forms - especially which 'forms' warrant recognition as distinct species - has yet to be achieved. Although difficult individuals may still be encountered, and it seems likely that 'michahellis' may itself embrace more than one phenotype, the key identification features of michahellis gulls are now well understood. From the details published by Hume (1978) it is now clear that the Chasewater individual was from the southern michahellis populations and, indeed, by the late 1980s such birds were being observed annually in the region. By the mid 1990's, 40 – 50 individuals were being recorded annually, rising to around 75 per year by the end of the decade. In 1999 a Yellow-legged Gull paired with a Lesser Black-backed Gull L. graellsii and nested at Bredons Hardwick in Worcestershire. Two young were hatched in June but did not survive.

The following analysis describes the temporal and geographical (inter-county) distribution of Yellow-legged Gulls in the West Midlands area and examines how these compare with the national picture.

Methods and limitations

All data published in the West Midland Bird Reports since 1973 have been extracted and examined. Records have been annual since 1986 and up to 2000 it was possible to make reasonably reliable assessments of the number of individuals reaching the region annually. Thus, detailed analysis of seasonal trends etc. is based upon the 15-year period 1986 to 2000. Thereafter numbers reached levels where determination of the numbers of individuals becomes impractical and, therefore, subsequent status trends are based upon annual maxima at well-watched sites.

The period from 1973 to 2000

[ Most of this section is reprinted, with additions and updates, from 'The status of Yellow-legged Gull in the West Midlands region', published in December 2004 in West Midland Bird Report 69: 234-249.]

Even in this earlier period, several factors can affect the assessment and accuracy of such numerical data and the numbers cited should be regarded as 'best estimates' and not as absolute.

Certain issues can lead to over-estimates of numbers, and result from duplication of published data. Adjustments (to some extent subjective) have been applied for the two most obvious of these:

Movement between sites

Observations of Yellow-legged Gulls (and hence published figures in the WMBC Reports) are concentrated on two different types of locality. Firstly are observations at gull-roosts, which assemble in the late afternoon and evening at reservoirs and other sizeable water-bodies. Secondly are observations at daytime feeding and resting sites, such as refuse-tips and certain gravel-pit complexes. Clearly, these two sets of 'data' are not mutually exclusive. Birds feeding during the day at a refuse tip may well be seen later at a lake or reservoir roost. Hence, the published data certainly involve a degree of duplication. Examination of the published data suggests that only in Worcestershire is this likely to be a significant factor. Notable numbers of Yellow-legged Gulls have been reported regularly during the day at Throckmorton tip and similarly significant numbers in the evening at the Westwood Park roost. It is not known what proportion of birds feeding at Throckmorton later roost at Westwood but, as the two sites are only 10 miles apart as the gull flies, it is certain that interchange will occur (identifiable individual gulls of other species, e.g. Glaucous Gull L. hyperboreus, have been identified at both sites on the same day). To allow for this in the following regional and county figures, a somewhat arbitrary 'adjustment for duplication' has been made at the 50% level i.e. it has been assumed that half the numbers involved in sightings at Throckmorton are also involved in sightings at Westwood, and the 'total' figures have been adjusted accordingly.

Length of residence

It should be noted that 'totals of individuals' and 'monthly distribution of records' are numerically distinct. Individual gulls may remain in the area for more than one month and thus contribute several 'units' to the 'monthly distribution' of records. Thus, the analysis of the monthly distribution of numbers reveals the number of birds present in a given month and not birds necessarily arriving in that month nor the total number of individuals involved. Rarely are the published data precise enough to distinguish recent arrivals from individuals remaining from a previous month. To arrive at an estimate of the minimum number of individuals, monthly increments have been used. For example, if - at a given site - there were up to four birds in October and up to six birds in November, it is assumed that only two of these individuals actually arrived in November, with four remaining from October. This 'minimalist' approach will inevitably produce an under-estimate of the true numbers involved, and this may be significant at some sites. However, there is no alternative method of producing consistent, comparative data between sites and between years.

Occasionally, estimated totals of the number of birds occurring at a given site during a calendar year have been published in the regional Report. Such figures will often include a considerable number of individuals observed during January. Many of these birds will have arrived, in fact, during the later months of the previous year and are not new arrivals in the current year. Their occurrence will have already been published in the Report for the previous year and their contribution to the statistics incorporated into the previous year's totals.

As a result of these factors, it should be noted that the estimates for 'total numbers' in the following analysis will not always accord with some of the comments in the corresponding WMBC Report for a given year.

The above adjustments will help to reduce inflated figures resulting from duplicated data in the published figures.

Conversely, if there is in fact a relatively rapid 'turnover' of individuals at certain sites or in certain years, numbers taken from the published figures may under-estimate the true situation. Most of the published monthly figures probably relate to 'raw' day maxima (the highest number of individuals seen on any one day during the month) rather than the number of different individuals identified during that month. In these circumstances, the number of different individuals may significantly exceed the published monthly figures (see WMBC Reports for 1994 and 1995). In the absence of structured data, no correction or adjustment can be applied for this situation.

The final factor to take into consideration is that of mis-identifications, which are probably seasonally biased towards the early months of the year. This will affect observations at roosts rather than at tips and daytime 'loafing sites'. At roosts, the light is inevitably failing, the range of observation is usually quite distant, and certain characters such as leg-colour are unlikely to be observed. Despite these impairments to observation, there is evidence that the approach to identification at such roosts is often over-simplistic. Frequently, identification appears to be based solely on a combination of a relatively dark grey mantle and a white, unstreaked head. While these features are a good starting point in autumn and early winter, they are quite inadequate from late-December onwards, as by that date some Herring Gulls L. argentatus have acquired a very white and unstreaked head. As some Herring Gulls may also show a relatively dark grey mantle, identification of Yellow-legged Gulls in the period from late December onwards requires this proviso to be kept very clearly in mind. At roosts it will usually be impossible to determine leg-colour, but confident identification will involve structure, wing-tip pattern and the appearance of the eye, as well as mantle colour and head pattern.

Even at tips and gravel pits, where the leg colour of a large white-headed gull may be among the first things to attract attention, the possibility of a Herring Gull with yellow legs has to be considered. Such birds have been claimed on a number of occasions and a certain individual was trapped and ringed at Throckmorton in December 1990. For discussion of the appearance of Herring Gulls with yellow legs, and photographs illustrating how confusing some individuals can be, see Jonsson (1998).

A further possible source of misidentification has emerged, with an apparently increasing number of hybrids. These can involve not only hybrids involving michahellis as one parent and e.g. Herring Gull as the other, but also hybrids between Herring Gull and Lesser Black-backed Gull, with the resulting offspring sometimes resembling Yellow-legged Gull. Such birds have been recorded in recent years from the Bristol area (Peter Rock in litt.).

However, although care should always be applied in identifying Yellow-legged Gulls, the influence of identification errors should not be exaggerated. As Yellow-legged Gulls are becoming more numerous visitors, occasionally misidentified individuals will, hopefully, become proportionately less and less significant and will not unduly affect the prevailing patterns of occurrence.

The analysis below must take into account the limitations of the numerical data. The graphics are based upon statistics for a 15-year period and will reveal seasonal patterns and numerical trends, but the numbers involved should not be regarded as absolute.

Annual Distribution

Following the first record in November 1973, there were just 18 further Yellow-legged Gulls in the following eleven years and there were no records at all in 1985. The first record for Warwickshire came from Draycote in June and July 1976 and the first record for the 'West Midlands County' at Walsall in November 1984. Regionally, records were then annual from 1986, including the first for Worcestershire, at Throckmorton, in November of that year.

Between 1986 and 2000, the number of Yellow-legged Gulls visiting the region was probably between 700 (based on monthly 'increments') and 1200 (based on summated 'monthly maxima') but, allowing for 'rapid turnover' (see 'Methods and Limitations') and over-looked non-adult gulls (see 'Age Distribution') the true figure could be significantly higher.

By the late 1990's, 70 or more individuals were occurring annually, with up to 16 at a time at Draycote Water in Warwickshire and up to 11 at Westwood in Worcestershire.

The annual pattern of records based upon estimated minima (see 'Methods and Limitations') is presented in Figure 1.

  Annual totals of Yellow-legged Gulls in the West Midlands

Fig 1. Estimated Annual Totals of Yellow-legged Gulls in WMBC Region, 1973 - 2000

Annual numbers rose steadily through the late 1980’s and then remained relatively constant at around 50 individuals per year for the first eight years of the 1990’s. A second upsurge then occurred, with around 75 individuals per year during the three years up to 2000.

An examination of the figures by county (Figure 2) shows that, within the regional pattern, there are small but significant differences between the four counties.

 

  Annual totals of Yellow-legged Gulls in the West Midlands by county

Fig. 2.  Estimated Annual Totals of Yellow-legged Gulls by County, 1986 - 2000

 

Warwickshire.

Numbers reached a peak in 1993 and then declined somewhat in the ensuing four years. Numbers increased from 1998 but only by 2000 did numbers exceed the 1993 level.

Worcestershire.

Records in Worcestershire reached a plateau between 1989 and 1996, and then displayed a significant increase in the subsequent four years.

Staffordshire

Numbers reported from Staffordshire were relatively erratic but rather higher between 1988 and 1992 than between 1993 and 1997. As in all three ‘shire’ counties, numbers increased in the three-year period up to 2000.

West Midlands

In the West Midlands County, only Bartley Reservoir consistently produced records of Yellow-legged Gulls prior to 2000 and the total number of individuals is too low to determine trends of statistical significance. There were marginally more records in 1995 and 1996 than in other years but the pattern has generally been relatively constant, at around 7 individuals per year. (In more recent years, the Stubbers Green area has produced more significant numbers - see below.)


The annual pattern of records for the region as a whole is much as might be expected: a small number of records initially, during the years when Yellow-legged Gulls established themselves as visitors to the area (and to Britain as a whole); a fairly steady stream of observations through the late 1980s and 1990s; and finally a second upturn in the number of records, probably indicative of a genuine increase but also partly the result of increasing observer awareness and effort.

The development of records in Worcestershire closely mirrors this pattern. In Warwickshire, a polynomial trend line plotted through the data again portrays a similar underlying trend (Figure 3). However, there is an interesting irregularity in the pattern, with a peak in 1993 followed by a decline in records, before a recovery in the late 1990's.

 Numbers and trend-line for YL Gulls in Warks

Fig. 3.  Annual numbers and trend line for Warwickshire

This 'irregularity' reflects a significant change over the years in the distribution of records within Warwickshire, with the Tame Valley providing the majority of earlier records but Draycote Water the majority of observations in more recent years.

In the ten years 1986 – 1995, the Tame Valley was the principal area, and numbers reported there exceeded those from Draycote by a factor of nearly two to one. In the following five years this situation was completely reversed and numbers at Draycote exceeded those in the Tame Valley by a factor of six to one. Changes at both locations were involved. There was a significant increase at Draycote but also a precipitous decline in the Tame Valley, where only three were reported in the four years 1996 – 1999. A partial return to form occurred in 2000, when at least seven individuals were recorded.

The decline in the Tame Valley in the second half of the 1990’s may well be related to the distribution and operational practices of refuse tips in northern Warwickshire, allied to the disappearance of suitable  'loafing areas' at nearby gravel pits and kindred sites (e.g. flooding of previously 'dry' pits in the Kingsbury area, and abandonment of the fly-ash lagoons at Ladywalk following the closure of Hams Hall power-station). 

(The excavation of extensive new sand and gravel pits in the Middleton, Drayton Bassett and Dosthill areas, on the Warwickshire/Staffordshire border north of Kingsbury, and an extension of landfill operations in the Dosthill/Wilnecote area produced conditions that once again attracted Yellow-legged Gulls on a regular basis from 2002 but the running down of the landfill operations in 2011 has again lead to much-reduced numbers of LWHGs in the middle Tame Valley.)

Annual numbers reported in Staffordshire between 1986 and 2000 varied somewhat erratically. The majority of records through the period as a whole came from the three large reservoir roosts, at Belvide, Blithfield and Chasewater. Very few were recorded at any of these sites in 1993, 1996 and 1997. As with the other ‘shire’ counties, numbers increased in the three years 1998 –2000, with Belvide in particular reporting higher numbers than hitherto. A key change between the early and later years of the period was the decline in latter years of reports from Westport Lake, near Stoke. Reports were annual between 1986 and 1993, with around six individuals each year. Since 1994, however, there have been relatively few reports and none at all in 1994, 1999 and 2000. This reflects a general decline in reports from Westport of gulls of all species, which may to some extent reflect changes in observer activity.

Monthly Distribution

Figure 4 shows the monthly distribution of records of Yellow-legged Gulls in the region between 1986 and 2000. The data are plotted from summer through winter (rather than January to December) to match the natural profile of arrivals and departures. Each column shows the summated totals for that month ('the number of birds present') and also the estimated total of arrivals in that month, calculated as described in 'Methods and Limitations'.

 Monthly distribution of YL Gulls in the West Midlands

Fig. 4.  Monthly distribution of Yellow-legged Gulls in WMBC region,
1986 – 2000, plotted from summer through winter

The corresponding monthly distributions for the individual counties are shown in Figure 5 (note the different 'Numbers' scale in each diagram).

 Monthly distribution of YL Gulls in the West Midlands, by county

Fig. 5.  Monthly distribution of Yellow-legged Gulls by county, 1986 – 2000

The monthly profile of numbers across the region as a whole (Figure 4) shows a bi-modal pattern, with an initial summer influx in July, a slight decline in numbers through the autumn, and then a second and more substantial peak during the winter months between October and January. Examining the monthly distribution by county, however, reveals a wide disparity between the four counties in the proportion of summer records (Figure 5). The regional profile - with a small but distinct summer peak followed by a more pronounced winter peak - is reflected closely in the records from Worcestershire and the West Midlands County.  In Warwickshire, numbers in summer have been significantly higher as a proportion of the total, with July the single most productive month. In Staffordshire, by contrast, there have been very few reports between July and September, with records almost exclusively during the winter months.

Age distribution

Data on the age of Yellow-legged Gulls in the region is available for about 75 percent of records. Since 1994, many of the records published in the WMBC Reports have been tabulated, facilitating ready reference to numbers but unfortunately losing data on the age of the individuals involved.

Many observers do not actively search for non-adult Yellow-legged Gulls, particularly first and second-year individuals, so the proportion of non-adults recorded will certainly be less than the true number occurring. The early arrivals in June and July might be expected to consist largely of adults but the true age-distribution later in the winter remains to be fully clarified. Nevertheless, of the records published with age-data, 68 percent were adults. Of the 32 percent non-adults, 4% were first-years (1W/1S), 8% second-years, 11% third-years, and 9% fourth-years.

Comparisons with other areas

The largest concentrations of Yellow-legged Gulls in the UK are to be found in the south-east coastal counties, between Essex and Dorset. At Pagham Harbour in Sussex, for example, over 300 may occur at a time during July and August (per Richard Fairbank and the Sussex Ornithological Society) while over 300 have been recorded at a single site in Dorset during September (per Shaun Robson and the Dorset Bird Club). Over 380 were at East Tilbury in Essex during September 2001 (per Howard Vaughan, Paul Wood and the Essex Birdwatching Society).

In these south-east coastal counties, there is a large influx between July and September, with numbers peaking in August or early September. Numbers then decline during the autumn. There is no evidence of a second influx of 'wintering' birds and, indeed, relative to the numbers present in late summer, there is an almost complete exodus during the winter months.

This mirrors the situation in northern continental Europe, where numbers also peak in August and early September, and decrease after October (Jonsson, op. cit.).

Figure 6 shows the monthly profile of numbers during 1986 – 2000 in Sussex, based on ‘monthly maxima’ extracted from the Sussex Bird Report by Richard Fairbank, and ‘estimated arrivals’ calculated by the author as per ‘Methods and Limitations’.

Comparative monthly distribution of YL GUlls in Sussex

Fig. 6.  Comparative monthly distribution of Yellow-legged Gulls in Sussex,
1986 – 2000, plotted summer through winter

Just 40 miles inland, numbers in the vice-county of Surrey are much lower and more comparable with those in the West Midlands (data per Jeffery Wheatley and the Surrey Bird Club). Numbers increase rapidly through July, peak in August and then decline slowly through the autumn and winter months. Apart from a slight increase in the total number recorded during January, there is little evidence of a 'bi-modal' distribution. Winter records are regular but totals during 1994 to 2000 never exceeded 60% of the July/August peak and could well involve birds remaining from the summer.

The number of Yellow-legged Gulls congregating on the south coast of England is very large compared with the West Midlands. Over 300 individuals may be present at one time in Essex , Sussex and Dorset during the peak months of  August and September. In comparison, numbers at any one time in the West Midlands region rarely exceed 20 to 30 individuals during the winter peak in December. It may be dangerous to jump to conclusions about differences in the seasonal distribution between southern England and central England when the total 'populations' are of such different magnitudes. However, the increasing numbers in winter in the West Midlands, and the fact that numbers reach a peak at that season, is very much at odds with the pattern in the English heartlands for Yellow-legged Gulls on the south coast and also in northern continental Europe.

How does the pattern in the West Midlands region compare with adjoining midlands counties?

During 1997, Jon Baker carried out detailed observations of Yellow-legged Gulls at Dix Pit in Oxfordshire (Baker, 1998). Numbers in the early months of the year were low, with none from late April through May. The first returning birds were in early June, followed by ‘a small passage of birds for the rest of June and July’. There were seven birds on July 21st and August 1st, after which ‘the species became scarce again’. Numbers then increased significantly through October , reached a day maximum of 26 on November 26th , and declined only slightly during December. Thus, the seasonal pattern at this site in Oxfordshire mirrors that for the West Midlands region.

In Derbyshire, the pattern also matches that in the West Midlands area. 'The first birds are usually noted in June and July followed by a big influx in August. Numbers then drop off slightly in September and October before rising again to a peak in December. Slightly lower numbers are noted in January with numbers then falling away until the last birds are noted in April' (Keys, 2002).

In Northamptonshire, annual numbers have increased since 1995 and reached around 100 in 2001 (per Mike Alibone and the Northamptonshire Bird Club). 'There is one clear peak emerging, in August/September. Birds begin to arrive in July and are evident until October. There are a few winter records but there is no discernible second peak.' Thus, the seasonal pattern in Northamptonshire does not mirror that for the West Midlands region and is closer to that evident in southern England.

Discussion

There are disconcerting differences between the apparent seasonal patterns in various midland counties.

Within the West Midlands region, there is a significant winter peak in Staffordshire, Worcestershire and the West Midlands County. In Warwickshire, although there are significant numbers in winter, July has been the single most productive month.

In a study at an Oxfordshire site and also in Derbyshire, peak numbers are again during the winter but in Northamptonshire numbers peak in August/September and there is no recorded winter influx.

Such differences are highly unexpected in contiguous inland counties.

The winter peak evident in a majority of midland counties is also at variance with the seasonal profile of Yellow-legged Gulls in their 'heartland' areas on the south coast.

A high proportion of summer records in Warwickshire comes from the Tame Valley and involves 'daytime' observations at 'resting' sites such as the gravel workings at Dosthill and, in earlier years, the fly-ash lagoons at Hams Hall Power Station. By comparison, records from Draycote much more often involve roosting birds and, significantly, are more concentrated in the winter period, September to March.

In Staffordshire, a very high proportion of records come from the gull roosts at the three major reservoirs, Belvide, Blithfield and Chasewater, and, in this county, records have been almost exclusively in winter.

Observations at gull-roosts are much less frequent during the summer, before the winter influx of the majority of gulls of other species attracts observers' attention.

How genuine is the very different pattern in the West Midlands area compared with the south coast? Are the differences in seasonal patterns between counties in central England substantive? Is the bi-modal distribution and a winter peak perhaps an artefact, arising primarily from the on-set of concentrated 'winter roost watching' in late autumn?

The observations of Baker in Oxfordshire (op. cit.) encompassed all months of the year and can be assumed to be real and free from seasonal bias.

In order to test for seasonal bias in the West Midlands regional data,  I have analysed subsets of data from sites where there has been at least a degree of year-round observation at the appropriate time of day, and over a period of several years.

Two such subsets of data from the West Midlands region have been used. Firstly, records from Throckmorton, which is the only site for which the published data regularly encompassed most months of the year and was available throughout the period 1986 –2000. To provide a second, independent 'control set', I have observed Yellow-legged Gulls on a year-round basis at Bartley and Frankley Reservoirs during the four years 1999 – 2002 (and thus slightly outside the principal period of analysis). Numbers of Yellow-legged Gulls at this site are very small but the site has the advantage (in this context) that there is both a diurnal 'loafing' area and an over-night roost, and as a consequence Yellow-legged Gulls frequent the site in both the summer and winter months.

The monthly distributions resulting from these two subsets of data are presented in Figure 7.

Monthly distribution of YLGs at Throckmorton and Bartley/Frankley 

Fig. 7.  Monthly distribution of Yellow-legged Gulls at two sites with year-round observations

The results from each site are very similar. There is a more-even balance between summer and winter numbers compared with data from the region as a whole but the bi-modal distribution and winter peak are still evident.

It can be concluded that 'seasonal bias', resulting from the on-set of  'winter gull-roost watching', does inflate the relative prominence of the winter peak in numbers but is not wholly responsible for the bi-modal pattern in the monthly distribution.

Hence, there is evidence that the winter peak in numbers in the West Midlands area is genuine. Rather than a single peak of birds in summer, as on the south coast, it would seem that there is a summer influx, a brief hiatus, and then a second influx of winter visitors.

This pattern is echoed in data from Derbyshire and Oxfordshire but not in Northamptonshire. The apparently disjointed distribution of wintering birds within adjoining midland counties is highly unexpected  but analysis of selected 'year-round' subsets of data re-affirms the bi-modal pattern and a winter peak in numbers across most of the West Midlands area.

Where do the late arriving wintering birds come from? The simplest explanation is that they are birds that arrived in summer in southern England or continental Europe as part of the principal pre-moult migration and then subsequently relocated within (or to) Britain as part of a secondary post-moult migration.

The seasonal pattern on the south coast of England, and in northern continental Europe, shows that most of the Yellow-legged Gulls moving north in late summer disperse during the winter. Clearly, given the large numbers involved, most must return south at this time. However, perhaps a small number instead moves north and west into the more land-locked areas of Britain, to areas where reservoirs and refuse tips attract such large numbers of Herring Gulls and Lesser Black-backed Gulls at this season.

A more intriguing explanation is that they involve a second and relatively small influx from southern Europe, perhaps from another part of the species' breeding range. At least a proportion of the individuals found in England in winter appear to average smaller, darker and rounder-headed than is typical for Yellow-legged Gulls originating from the Mediterranean basin. Such characteristics have been associated with Yellow-legged Gulls from the Atlantic seaboard of southern Europe and NW Africa (Martin Garner and Dick Newell in litt.). If a second population is involved, such gulls may perhaps favour different sites from those arriving from the Mediterranean basin, and this could go some way to explaining the uneven distribution of Yellow-legged Gulls across the midlands during the winter.

The period since 2000.

Since 2000, the number of Yellow-legged Gulls visiting the region has reached levels where only summarised data are published in the regional Reports and detailed analysis is no longer possible. Generally, the species is now recorded in all months but the broad picture of a bi-modal pattern of occurrence (late-summer and winter influxes, respectively) is firmly established. It is likely that the number of individuals annually is now well into three figures.

In 2004, a count of 18 on August 31st at Chasewater (Staffs) constituted a regional record at that time for a single locality. During 2007, a new regional locality-high of 20 occurred at Draycote (Warks) on January 23rd, while at least 22 individuals were identified there during the last week of the month. In 2010, a new record was established at Draycote, with 25 on July 28th and 26 on July 30th.  At least 36 individuals were identified on these two dates, comprising 14 adults, one fourth-year, 7 second-years and 14 juveniles. In Worcestershire, 16 were counted at Throckmorton landfill in October 2009. In the West Midlands 'county', 14 were at Stubbers Green on July 21st 2010, with at least 31 different individuals during the month.

The five-yearly mean maximum counts from selected sites during 2001 - 2010 are presented in Table 1.

5-year Mean Draycote Tame Valley Throckmorton Westwood Belvide Blithfield Chasewater Bartley Stubbers Green
2001 - 2005 9 3   8 7 2 8 4  
2006 - 2010 19 6 9 6 9 4 8 8 8

Table 1. Five-yearly mean maximum counts of Yellow-legged Gulls at selected sites in the West Midlands Region, 2001 - 2010.

It is considered that there is significant 'turn over' of Yellow-legged Gulls visiting the region and that, during the course of a year, the number of individual Yellow-legged Gulls visiting the region may well be two or three times the maximum day/site counts, as presented in the Table.

See also a selection of  images in the Photo Gallery, to illustrate the appearance of Yellow-legged Gulls and to facilitate considerations of origins and phenotypic variation. The item Studies of an individual Yellow-legged Gull also includes a series of images showing the progression of a single individual from 2W (3cy) to adult plumages.

Summary

Following the first fully-documented record in 1973, Yellow-legged Gull has become a regular visitor to the West Midlands region and has been annual since 1986. By the late 1990s, records numbered 75 or more individuals annually and by 2010 annual totals were in three figures. Up to 26 have been present at a time at Draycote (Warwickshire) and up to 18 at Chasewater (Staffordshire). Nearly 70% of birds identified have been adults, though a proportion of immatures, particularly first and second years, is probably over-looked. A detailed analysis of data from 1986 to 2000 indicates some notable differences between the four counties in the annual and monthly distributions of records. Although natural patterns of movement in the counties may differ to some extent, it is also likely that patterns are influenced by  a) sudden changes in habitats at key sites which make them either more or less attractive to gulls, and  b) the relative extent to which observations are focussed on roosts, which are watched principally in winter. The latter factor, in particular, could influence the apparent seasonal patterns of arrivals and departures. However, published observations as a whole and analysis of carefully chosen subsets of data both suggest arrivals in two phases: a summer influx around July, associated with the species' northerly post-breeding moult migration from the Mediterranean basin, and a larger winter influx between October and January. Such a bi-modal pattern and a relatively significant influx in winter are at odds with the profile of numbers at key sites for Yellow-legged Gulls on the south coast of England and also in northern continental Europe. In these areas, where relatively large numbers arrive in the summer, there is a major exodus during the mid-winter months. The winter influx in parts of the Midlands involves very small numbers compared with the peak populations found on the south coast and perhaps a small proportion of individuals departing from southern England undertakes a secondary post-moult migration further north and west rather than returning immediately to southern Europe. A more intriguing possibility is that such winter arrivals involve gulls from another part of the species' range in southern Europe, perhaps including the Atlantic seaboard. Such gulls may perhaps favour different sites from those arriving from the Mediterranean basin, and this could go some way to explaining the uneven distribution of Yellow-legged Gulls across the midlands during the winter.  

Acknowledgements

Principal thanks go to the small but enthusiastic group of gull-watchers in the West Midlands region who have produced the data upon which this article is based. I am very grateful to the following for providing comparative data or publications on Yellow-legged Gull for their respective areas: Mike Alibone (Northamptonshire Bird Club), Richard Fairbank (Sussex Ornithological Society), David Hawkins (Oxford Ornithological Society), Rodney Keys (Derbyshire Ornithological Society), Shaun Robson (Dorset Bird Club), Howard Vaughan and Paul Wood (Essex Birdwatching Society), and Jeffery Wheatley (Surrey Bird Club). David Emley clarified specific points in respect of Staffordshire. Martin Garner and Dick Newell provided valuable comments on current debate and speculation regarding the appearance, variations and possible origins of Yellow-legged Gulls observed in England during the winter.

References

Baker, J. 1998. The Yellow-legged Gull in Oxfordshire. Birds of Oxfordshire 1997: 72-76.

Dean, A.R. 2004. The status of Yellow-legged Gull Larus michahellis in the West Midlands region. West Midland Bird Report 69: 234-249.

Garner, M., Quinn, D. and Glover, B. 1997.  Identification of Yellow-legged Gulls in Britain. British Birds 90: 25-62, 369-383.

Hume, R.A. 1978. Variations in Herring Gulls at a Midland roost. British Birds 71: 189-191.

Jonsson, L. 1998. Yellow-legged Gulls and yellow-legged Herring Gulls in the Baltic. Alula 4: 74-100.

Keys, R. 2002. The status and occurrences of Yellow-legged Gull and Caspian Gull in Derbyshire. Derbyshire Bird Report 2001: 133-134.

Yesou, P. 2002. Systematics of Larus argentatus cachinnans fuscus complex revisited. Dutch Birding 24: 271-298.


  Top of Page