Numerous winter visitor and passage migrant. Increasing rooftop-breeding population in urban sites

See also:  'Roof-top nesting gulls within the Birmingham boundary' by Jim Winsper

Winter status

In the West Midlands, along with many other regions of England, Lesser Black-backed Gulls became increasingly numerous winter visitors from the mid-1950s. At Blithfield Reservoir, Staffs, for example, there were 900 in October 1960 and up to 3500 during November and December 1967. Harrison et al. (1982) estimated a regional total of around 10000 by the late 1960s and 25000 - 30000 by the mid 1970s. 12000 were estimated in the roost at Draycote Water in December 1974. Numbers declined somewhat towards the end of that decade, with a total population of around 20000. Since then numbers have remained relatively stable, with Harrison & Harrison (2005) estimating the wintering population at the start of the current millennium as still around 20000. Counts at some of the larger roosts are now intermittent but data suggest an increase between 2000 and 2010 but a fall since, coinciding with the closure of some landfill sites. By 2020, although based upon intermittent, non-systematic counts, summed maxima from  principal roosts suggest a regional population of between 15000 and 20000.

The five-yearly mean maximum counts from selected sites during 1986 - 2010, as documented primarily in the Annual Reports of the West Midland Bird Club,  are presented in Table 1.

Table 1. Five-yearly mean maximum counts of Lesser Black-backed Gulls at selected sites in the West Midlands Region, 1986 - 2010.

Roosts of around 2000 have also been reported from Westport Lake in Staffordshire, while as many as 7000 have been estimated feeding at the landfill site at Throckmorton in Worcestershire.

The effects of landfill closures

The closure of landfill sites is generally regarded as a significant factor in reduced numbers of wintering large gulls. The dramatic effects which can follow closure or reduced operations at landfill are vividly illustrated by Throckmorton, where acceptance of domestic refuse ceased in October 2016. At that time a huge decline in the numbers of LWHGs was reported : e.g. 2850 LBBGs in Feb 2016, high numbers through to October 2016 and then just 45 in Nov and 21 in Dec. During 2017 the maximum at Throckmorton was 250 in November. This decline at Throckmorton was mirrored by a substantial reduction in numbers roosting at Westwood Pool, where totals previously in the thousands declined to numbers in the hundreds. Landfill including organic waste recommenced at Throckmorton in 2018, believed to be necessitated by problems with Worcestershire's incinerator plants. Subsequently, there were 1000 LBBGs at Throckmorton in Nov 2018 while during 2020 numbers up to 800 for both LBBGs and HGs were encountered. However, there was no resumption of a substantial and consistent roost at Westwood. Andrew Warr has observed gulls from Throckmorton departing to the east and south-west, the latter in the direction of the Severn Estuary. Presumably, the resumption of organic landfill at Throckmorton will be a temporary measure, so that further declines in the numbers of LWHGs at the site can be anticipated in due course. Thanks are due to Andrew for his input on the Throckmorton / Westwood situation.

In the Tame Valley, the roost at Coton Lakes historically attracted many gulls from landfills at Wilnecote and Packington, both of which have since ceased operations (though, apparently, Wilnecote still has 'capacity' and a continuing option to recommence landfill). The Coton roost reached a maximum of 1200 during 2010 (see table 1 for Tame Valley statistics between 2006 and 2010) but in the following five years the mean annual maximum fell to 520 and between 2016 and 2020 it fell further to just 280. Thus, closure of landfills certainly can lead to profound declines in the numbers of LWHGs at local roosts. However, how this correlates with overall regional numbers remains unclear. It is possible that some / many gulls may simply relocate to other landfills. The level of dependency on landfill may also questioned. Some studies suggest that landfill provides only a relatively small part of the diet of Lesser Black-backed Gulls. Coulson (2019) has emphasised that many gulls make early morning visits to still-damp pasture, where they target invertebrates, which provide an under-estimated component of their food requirements. That said, as and when landfills more-generally disappear from the region, it seems unlikely that so many LWHGs from distant origins will choose to undertake long movements to winter in the area.

Origins of wintering Lesser Black-backed Gulls

The majority of Lesser Black-backs wintering in the region are of the race graellsii, which breeds in Britain and also in Iceland, the Faeroes, the Netherlands (where some intergrade with intermedius), Brittany, and Spain. In Britain as a whole, there have been recoveries of birds originating from across this range and also from Belgium, Denmark, Sweden and Finland (Wernham et al. 2002). Most individuals wintering in the West Midlands region are believed to involve British-breeding graellsii originating from Northern England and Scotland (Harrison et al. 1982). However, recoveries of birds ringed during winter include individuals returning to breeding colonies in the Netherlands and Norway (Harrison and Harrison 2005), which indicates the potential for arrivals from disparate regions of the breeding distribution, including intermedius as well as graellsii.  Rock (2002, in Wernham et al. 2002) estimated that almost 60% of foreign-ringed recoveries were 'likely to be intermedius from Norway, Sweden and probably Denmark too'. It is probable that intermedius is under-recorded in the West Midlands region, though the recovery in Norway highlights a further issue : that, conceivably, graellsii may be breeding now in Norway. Paler-mantled Lesser Black-backed Gulls breeding in southern Norway are attributed to intermedius but the population of such paler-mantled birds is increasing and some are now found among colonies of the dark-mantled nominate fuscus in northern Norway. Helbig et al. (2009) wrote:  ' ... the Norwegian intermedius population increased steeply during the 1980s and early 1990s (Lorentsen 2007) and, since the late 1980s, greyish-mantled Lesser Black-backed Gulls with a L. f. intermedius appearance have been found in colonies in northern Norway which were previously occupied solely by the nominate fuscus (Strann & Vader 1992, Bustnes et al. 2006). It is not known whether these new birds originate from expanding intermedius colonies in southern Norway, or from colonies of a third subspecies, L. f. graellsii, which has increased in the North Sea area and in Ireland (Creme et al. 1997, Garthe et al. 1999), as these two subspecies are difficult to distinguish'.

Age distribution among wintering Lesser Black-backed Gulls

The wintering population in Britain involves a high proportion of adults. Prior to the 1940s, adult and younger Lesser Black-backed Gulls in Britain departed between August and October to winter along the Atlantic coasts of southern Europe and Africa (see, for example, Cramp & Simmons 1983 & Coulson 2019). During the 1940s and 1950s, increasing numbers of adults began to remain in Britain during the winter, a phenomenon attributed in large part to the availability of ample food supplies provided by landfill sites. However, the majority of first calendar-year birds continued to migrate to southern climes. Based upon data from 1959 - 1960, Barnes (1961) estimated that 80% - 90% of  LBBGs wintering in Britain were adults, with most of the remainder comprising 2W and 3W. Thus, 1W were very much in the minority. Despite the very substantial national increase in the wintering population since that time, these proportions have remained representative, including in the West Midlands. Entries in WMBC Annual Reports include that 98% of 1000 LBBGs roosting at Draycote on February 28th 1993 were adults. At the Chasewater roost during late-February 2000, Lesser Black-backed Gulls were 'mostly adults' (whereas Herring Gulls were 'mostly immatures'). More-recent confirmation comes from sample counts arranged by Jim Winsper (2020) and discussed in detail in the Ringing Report in WMBC Annual Report No. 85 (covering 2018). Counts supplied from Bartley, Belvide, Chasewater and Draycote revealed between 1% and 16% first-winter individuals, with the higher numbers confined to the earlier post-breeding period and followed by an exodus, with generally lower single-digit percentages during January and February.

The very significant increase in the numbers of LBBGs wintering in Britain and the high proportion of adults in the wintering population can lead to an impression that most adults breeding in the UK now remain here during the winter. However, comparisons between breeding and wintering numbers show that the former are larger by a factor of c. 1.8. This suggests that, although a significantly higher proportion of 1cy gulls depart, there is still considerable migration of adults away from the UK (Banks et al. 2007). From studies centred on the Severn Estuary Region (SER), Rock (2005) estimated that 22% of adults were : ' ... either returning extremely early or not migrating at all and, from preliminary analyses, it appears that overwintering may be more prevalent among urban gulls'.

Rooftop nesting

The breeding distribution of Lesser Black-backed Gulls also underwent a radical expansion during the latter part of the 20th century, with the greatest change being the increase in inland nesting, mainly on man-made structures such as rooftops (see Balmer et al. (2013) and citations therein). Rooftop breeding did not extend to the West Midlands until the mid-1980s and has been attributed to expansion from colonies in the SER (see Harrison & Harrison (2005) and Winsper (2014) for further discussion). The first instances of breeding documented in WMBC Annual Reports were at Worcester in 1986 and in central Birmingham in 1987 (though Rock (2004) notes that a single pair is considered to have bred in Worcester in 1982). Detailed assessments of breeding numbers at Worcester and Birmingham were limited until the late 1990s and early 2000s but the indications are that numbers began to increase substantially during that period. Surveys in Worcester by Peter Rock indicated 230 pairs by 2003, 540 pairs by 2006 and over 900 pairs by 2020. In Birmingham, during 2008 to 2011, a survey conducted by Jim Winsper of  'Rooftop nesting gulls within the Birmingham boundary' found at least 555  'apparently occupied nests' (AONs) of Lesser Black-backed Gulls. A pilot survey in 2018, organised in Birmingham by the BTO and DEFRA, found 1921 AONs. WMBC Reports have also documented rooftop breeding in other urban areas of the West Midlands, including Kidderminster, Solihull (including Elmdon), Stoke, Stourport, West Bromwich and Wolverhampton. Smaller numbers of Herring Gulls have also been involved. These data, would suggest that the breeding population of Lesser Black-backed Gulls in the region had risen to over 3000 pairs (6000 individuals) by 2020. This is of the order of 25% to 30% of the currently estimated wintering population and, hence, numbers during the breeding season will now contribute significantly to the overall annual profile of numbers. 

Rooftop roosting

Most nesting activity has ceased by the end of July or early August, with adults and their offspring dispersing and declining numbers of gulls then to be seen at rooftop colonies during the day. However, gulls may return to roost in the initial weeks of the post-breeding season and there is evidence that they may be joined by other individuals, producing rooftop roosts of substantial size. Other rooftop roosts occur at sites not used for breeding (while Black-headed Gulls, which are not confirmed rooftop breeders in England, may also establish rooftop roosts at this time). Some rooftop roosts disperse by the end of September e.g. at Elmdon in this region and at Thetford in Norfolk (per Dawn Balmer) but, according to Deacon (2019), others in England may persist and even peak in mid-winter. Banks et al. (2007) noted that studies by Shedden (1983) and Austin et al. (2003), based on traditional water-based roosts, suggested that there are only limited movements to and from roosts after dusk. However, according to Deacon (2019), gulls may arrive at rooftop roosts in the middle of the night. Nocturnal movements may vary according to season and differ between rural and urban sites. Discussing local movements of urban breeding gulls tracked by GPS, Spelt et al. (2019) noted : 'Street lighting enables them [urban gulls] to forage at night. Indeed, whereas urban gulls spend approximately 50% of their time flying during the day, they spend 35% of their time flying at night'. While gulls foraging in urban areas at night, utilising street lighting, are quite likely to be noticed (though not necessarily seen or reported by bird-watchers), gulls arriving at roosts well after nightfall are less likely to be detected. Rooftop roosts as a whole are very probably under-recorded while known sites have been little studied. Are gulls arriving at rooftop roosts well after dark relocating from traditional water-based roosts, from nearby land-based or other rooftop pre-roosts, arriving from nocturnal feeding excursions or some combination of these and other sources? At rooftop roosts which disperse at the end of September, it seems likely that locally breeding gulls which have delayed southbound migration until this time, then trigger dispersal of the entire roost, including gulls from elsewhere which have joined the roost in the post-breeding season from July to September. However, if some relocate to traditional roosts, then any significant upturn in numbers there during late-September may not be due exclusively to recently arriving winter visitors from outside the region. Thus, potentially, migration patterns of gulls inferred from numbers at traditional roosts could be somewhat  misleading.

Non-breeding gulls in summer

Large numbers of predominantly immature gulls have occurred during the summer at some landfill sites, a phenomenon described as 'recent' in 1982 by Harrison et al., who mentioned 2000 LBBGs at Packington Landfill and 3000 at Ladywalk in June 1974. Subsequently, at the (hitherto) huge landfill site at Throckmorton in Worcestershire (described above), numbers reached 1000 during August 1999, 3000 during August 2009 and 2500 during June and July 2010. There is little indication of such mid-summer numbers from the data published from traditional roosts but mid-summer counts at traditional roosts have always been relatively limited. While confined to a small number of sites, the gulls in these large summer assemblies will provide another significant contribution to the regional monthly distribution of Lesser Black-backed Gulls.

Changes in the seasonal profile of numbers

The changes in status and behaviour outlined above clearly have repercussions for the seasonal profile of numbers. Figure 1 shows the historical pattern from personal counts at two reservoir roosts : Blithfield during 1980 - 1985 and Bartley during 1998 - 2003. The former time-span, from Blithfield, will involve primarily winter-visiting gulls, as the species was not breeding in the region at that time and there was no significant roosting of immature gulls at the site during the summer months. The profile at Bartley dates from a time when the numbers of Lesser Black-backs breeding in the region were beginning to increase from a low base. At both roosts, Lesser Black-backed Gull numbers were low during the summer months but then increased rapidly from July to October, reaching a peak in November at Blithfield and in December at Bartley. Harrison et al. (1982) also detected a November peak based on region-wide data from 1955 to 1978. Numbers at both Blithfield and Bartley remained relatively constant through January and February. However, wintering birds starting to disperse as February advanced were no doubt offset by passage migrants beginning to move through to breeding sites elsewhere. By late March or early April numbers became relatively small. Overall, the two sets of data show a comparable profile through much of the year ('bell-shaped' when plotted from July through winter to June, which captures the natural profile of numbers at that time). However, the November peak of the earlier (Blithfield) period declined and the annual peak shifted towards mid-winter (Bartley).

Comparing Fig. 1 with Fig. 2, the 12-month cycle still exhibits increasing numbers in the post-breeding season and declining numbers through the early spring. However, during the late autumn and winter, there is a more 'bi-modal' and less 'bell-shaped' distribution. There are proportionately much-more significant numbers through July to October; a subsequent slight decline in numbers through November and December; an increased prominence of numbers during January and February.

A change in the relative balance between winter and summer numbers is to be expected, given the significant numbers now breeding on rooftops in urban sites.  Changes in the migration strategies of local gulls and an increase in passage migrants are also likely to be involved. As early as 1983, Cramp and Simmons (BWP3) noted evidence that adults were beginning their southwards migration in August to October but starting to return north from November, with gulls from their 2W onwards reappearing in British roosts in rising numbers through December to February. This was written in 1983 but, with numbers breeding inland nationally increasing rapidly since then and breeding not starting in the West Midlands until the mid-1980s, changes in the monthly profile in the West Midlands might be expected to have accelerated more-recently.

Intriguingly, as noted above, there are certain sites which contribute significantly to the total in particular months but little in other months; they do not adhere closely to the overall annual profile. They emphasise further the complexity of the issues contributing to the overall pattern. A particular case is Upton Warren, in Worcestershire. Here, in recent years, there has been a large influx of Lesser Black-backed Gulls during September and October, reaching 2200 in October 2019, but subsequent numbers in the winter months have been relatively small. In Fig. 2b, Upton contributes 29% to the September total and 27% to the October total. In stark contrast, the percentages from November to February are between 0.7% and 2.9%. The profile for all other sites combined shows a higher total for November than for October. The reasons for such an anomalous profile are not clear.

Figure 2a. Recent monthly distribution of Lesser Black-backed Gull as indicated in Fig. 2a but with contributions from Upton Warren shown separately.
Note the significant contributions to the regional Sept and Oct totals provided by this single site

Other races

Very small numbers (generally single figures) of the race intermedius are reported in the West Midlands region each winter. This race breeds in southwest Scandinavia, Denmark and the Netherlands and migrates south and south-west to wintering grounds in western Europe and west Africa. Thus, it is to be expected in Britain and it is likely that it is under-recorded in the region. There are a number of ringing recoveries of this form. However, field identification of birds away from the breeding range is not straightforward, as there is some variation in the shade of grey of the upperparts in graellsii and very dark individuals can be found breeding in LBBG colonies in Britain. In Denmark and the Netherlands many of the breeding birds are considered to be intermediate between graellsii and intermedius. There is some debate over the validity of intermedius as a subspecies rather than simply a cline among graellsii.

One or two birds of the nominate race fuscus ('Baltic Gull') have also been reported in most years. This race breeds in the Baltic area and northern Fennoscandia. Generally, it migrates south-east to winter predominantly in the Middle East and east Africa, though 1cy individuals from Northern Norway are more likely to move west, at least initially (Helberg et al. 2009). The breeding population of nominate fuscus is declining.

The identification of fuscus away from the breeding grounds is fraught with difficulties. The old perception that nominate fuscus could be identified on the basis of very black upperparts is now known to be false, as some individuals from the range of intermedius can appear just as dark in the field. It is structurally smaller and more elongated than intermedius, on average, but individual variation makes these characters unreliable. The status, movements and identification of 'Baltic Gull' were discussed in a key paper by Lars Jonsson  (Jonsson, L. 1998. Baltic Lesser Black-backed Gull Larus fuscus fuscus - moult, ageing and identification. Birding World 11: 295 - 317). However, a proviso re suspended moult in the other races has tempered some conclusions, even in this reference. The timing of the moult cycle seemed likely to provide a more reliable route to confident identification. The nominate race moults later than intermedius and graellsii and, prior to migration, adults normally moult only the two inner-most primaries. Thus, the outer primaries are still in place and all of uniform age during August and September, when the other two races are in active wing-moult. Unfortunately, it has recently been demonstrated that occasional individuals of intermedius (and graellsii) undergo 'suspended' moult, so that they too may show a moult cycle comparable to that of fuscus. Thus, positive identification of nominate fuscus remains problematic. It remains essential, however, to determine the precise state of moult with any suspected fuscus, as an individual in active wing-moult during the late-summer / early-autumn is certainly not fuscus.

The highest probability of success now appears to reside with 2cy individuals. For discussion of the more-advanced wing and body moult of fuscus in its 2cy and the possibility of confident identification of individuals of this age, see the relevant pages on the Gull Research Organisation webpages at: http://www.israel.gull-research.org/02cyapril.html and the references there to the relevant paper on 'Primary moult of Baltic Gull during the first 15 months (in Dutch Birding 28: 158 - 161) by Hannu Koskinen & Visa Rauste. Even further limitations emerged subsequently. The following appears on the website of BBRC, which now adjudicates claims of nominate fuscus (see http://www.bbrc.org.uk/subspecies-information/skuas-to-terns):

'Winters (2006) went on to propose a tighter definition of which second and third calendar year birds might be acceptable but this position was subsequently reviewed by Altenburg et al. (2011) who concluded that only second calendar year birds in April to June (and perhaps  into the summer) can be safely identified.'

Thus, the identification of fuscus away from the breeding grounds remains a considerable challenge. Ringing recoveries provide the surest route and offer some support for fuscus being a rare visitor to Britain, as several colour-ringed  fuscus have now been observed, as have individuals meeting the 2cy criteria mentioned above.

Given the difficulties in identification, it is considered that most historical claims of 'Baltic Gull' from the West Midlands Region will certainly prove inadequate and it is probable that none is sufficiently rigorous to prove acceptable. It is highly unlikely, for example, that many of the documentations submitted to county recorders hitherto have included details of state-of-moult. In consequence, historical reports claimed from the West Midlands Region are not itemised here.

Acknowledgments

References and acknowledgments for the material on this site can be found <here>. However, the text on Lesser Black-backed Gull has proved much more complex than those dealing with most other species and I would like to acknowledge here the particular contributions of Dawn Balmer, Nick Barlow, Lee Johnson, John Oates, John Sirrett, Andrew Warr and Jim Winsper, who each provided valuable correspondence and responded to my questions (and speculations). Data on age-distribution was supplied to Jim Winsper by Terry Hinett (Bartley), Steve Nuttall (Belvide), Ian Ward (Chasewater) and Tim Marlow (Draycote).