Urban gulls and their management in Scotland: review

4 Numbers, trends in numbers and status of gulls in

Scotland

This chapter provides a review of our current knowledge of the numbers, temporal changes in numbers (trends) and status of the gull species that occur commonly and in abundance in Scotland and are known to occur in urban habitats. Following some background information on the general ecology of gulls during and outside the breeding season, we describe the sources of survey information on urban gull numbers that are available and their limitations. For each of the key gull species, we provide details as far as is known of their overall Scottish population size, trends and status, and then specific information on size and trends in numbers in urban areas of Scotland.

4.1 Background ecology and sources of information

Six species of gull occur in Scotland that are either common or abundant (Black-headed Gull, Common Gull, Great Black-backed Gull, Herring Gull, Kittiwake and Lesser Black-backed Gull) and all six have been recorded as occurring within built environments. All tend to breed colonially and to forage and roost communally. Both breeding colonies and roosts are often, though not exclusively, in mixed species groups. All species are opportunist feeders that have apparently been able to capitalise on increased food availability through fishery discards and, with the exception of the Kittiwake, all are generalist feeders that readily take a wide range of edible discards, for example at garbage tips or other waste (Furness & Monaghan 1987).

4.1.1 Breeding season

General ecology

Breeding populations of most species of gull generally increased during the 20th Century in Britain, largely in response to increased protection and food supply (Cramp & Simmons 1983).

Apart from the Kittiwake that feeds mostly at sea, including on fishery discards, all gulls that breed in Britain can forage within a range of habitats including agricultural land, coasts and estuaries including intertidal areas, at sea (including use of fishery discards), garbage tips and other areas where edible refuse is discarded, although preferences and foraging ranges vary between species (e.g. Mudge & Ferns 1982, Furness & Monaghan 1987). As with all birds, their breeding distribution is limited to areas where there are both (i) suitable, secure nesting areas and (ii) suitable available food within a foraging range that does not normally put either the birds or their breeding attempts at risk. A secure nest site for a gull is generally one that is relatively inaccessible to ground predators, or where ground predators are scarce, for example on islands, cliffs, marshes, some moorland, peninsulas and sand dunes, and also on some buildings (e.g. Cramp & Simmons 1983, Mitchell et al. 2004). All gull species normally take at least two years to attain breeding age, with the larger species (Herring, Lesser Black-backed and Great Black-backed Gulls) typically not breeding until at least their fourth calendar year and sometimes until much later (Chabrzyk & Coulson 1976; Coulson et al. 1982; Cramp & Simmons 1983). Even after having recruited into the breeding population, a proportion of adult gulls do not breed every year (Calladine & Harris 1997, O'Connell et al. 1997) and thus, along with the immature cohorts, there will always be individual birds that are not 'attached' to breeding areas even during the breeding season.

Surveys, survey methods and limitations

Three extensive surveys have attempted comprehensive coverage of all coastal breeding seabirds, including gulls, in Britain and Ireland: 'Operation Seafarer' during 1969-70 (Cramp et al. 1974); the 'Seabird Colony Register' in 1985-88 (Lloyd et al. 1991); and 'Seabird 2000' in 1998-2002 (Mitchell et al. 2004); hence, recent population trends for each species within Scotland are reasonably well known. The most recent census, 'Seabird 2000' (1998-2002), was the first to attempt comprehensive coverage of all colonies, including those inland and in urban areas, although some inland and urban colonies were included in earlier surveys. In addition to the three national seabird surveys above, there have been two specific surveys of urban or roof-nesting gulls: one in 1976 (Monaghan & Coulson 1977) and one in 1994 (Raven & Coulson 1997). Operation Seafarer (Cramp 1971, Cramp et al. 1974) and the Seabird Colony Register (Lloyd et al. 1991) included surveys of some roof-nesting and urban gulls, in 1969-70 and 1985-88 respectively, but the two specific surveys in 1976 and 1994 provided information supplementary to that collected during the respective preceding and more extensive surveys. For this review, assessments of trends and current breeding numbers are based largely on the 1974 and 1994 specific surveys and the extensive survey of breeding seabirds undertaken in 1998-2002 ('Seabird 2000', Mitchell et al. 2004). Supporting information has been added from other local surveys (e.g. counts provided in the Scottish Bird Reports or local bird reports) and that provided directly by Local Authority representatives (see Chapter 3).

The 1969-70, 1976 and 1994 surveys were confined to gulls nesting on buildings or other man-made structures frequented by people. They were not comprehensive surveys of gulls nesting in urban areas, in that birds on natural sites (e.g. cliffs or wetlands) within urban areas were excluded. The aim of those surveys was to quantify the habit of roof nesting rather than the total numbers of urban gulls. Seabird 2000 (1998-2002) aimed to survey all colonies of breeding gulls, and obtain habitat details, including whether birds were nesting on buildings frequented by people, were recorded. Seabird 2000 also recorded whether the colonies were in urban or rural areas. The definition of an area as 'urban' is somewhat subjective, however, and nesting on buildings in rural areas may also 'impact' on humans. Hence, for this review, we consider the distribution of all gulls reported as nesting within urban areas and as on buildings, regardless of location. Areas defined as 'urban' by JNCC for Seabird 2000 are shown in Figure 4.1a, while the distribution of all nesting gulls recorded on buildings is shown in Figure 4.1b. In some gull colonies that have been recorded as 'natural' sites, some nesting will occur on derelict or abandoned and unoccupied buildings. For example, many Herring and Lesser Black-backed Gulls nest on, among, and even in the derelict buildings on the island of Inchkeith in the Firth of Forth (pers. obs., see also plate on page 238 in Mitchell et al. 2004). Such colonies have not been included as 'urban' or 'roof-nesting' in the current review.

All the major surveys of gulls undertaken to date have relied predominantly on fieldwork by volunteer surveyors; written instructions were provided and surveys were coordinated by regional organisers with local knowledge of both the birds within their areas and also of the individuals undertaking the survey work. The latter will have assisted in improving the thoroughness of coverage and also the accuracy of counts. Inevitably, however, survey methods for all breeding seabirds have progressed over the 30 years that counts of urban gulls have been made, including the publication of a specific manual to provide guidance on seabird monitoring techniques (Walsh et al. 1995). Any apparent population increases and range expansions therefore need to be considered against a background of improving knowledge of where gull colonies are located and of suitable field survey techniques.

During seabird 2000, five methods were recommended for counting breeding gulls (Mitchell et al. 2004). These were:

Relevant details of these counting methods have been provided in the following sections where they are required to explain the specific limitations of the data collected on urban gulls. Counting breeding gulls can be difficult. Outwith the difficulties of applying any given counting technique in a given context, there are frequently additional issues to address, such as the need to carry out validation studies to allow determination of the proportions of different species present where gulls occur in mixed colonies.

All the national surveys have attempted to provide estimates of the numbers of gulls breeding in each colony. With few exceptions, each colony count provides a best estimate of the number of AONs (based on counts of nests or of individual adult birds), generally with no associated measure of the precision for these estimates. This is because the confidence limits attached to any population estimate must be generated via a sampling approach during counting. Colonial birds, with an extremely heterogeneous distribution ( i.e. large numbers in relatively few places) do not lend themselves easily to such a sampling approach at a national level, although this rigorous statistical approach has been possible within some larger colonies. Current national and regional estimates of gull populations are therefore necessarily the sum of the 'best' estimates for each colony and, as such, contain a sum of the biases associated with all the component individual colony estimates.

Figure 4.1
(a) Gull colonies recorded during 'Seabird 2000' and classed as within urban areas (Source: JNCC).

(b) Gull colonies in Scotland recorded during 'Seabird 2000' or during two roof-nesting gull surveys (1976 and 1994) and classed as nesting on buildings (Source: JNCC)

Missing and negative data

Some urban gull colonies have been reported elsewhere, so that they are known to have been omitted from the 1998-2002 survey ( e.g. Lesser Black-backed Gulls in Edinburgh; Holling 1991, Dott 1994 and previous reports, Calladine 2004). The most recent attempt at a full count of Edinburgh and its surrounds was undertaken in 1994 (Dott 1994); a further attempt to update the results from 1994 is being made during the 2005 breeding season (M. Holling, pers. comm.). The overall extent to which the Seabird 2000 survey missed some extant urban gull colonies is difficult to assess however. Local surveys of roof-nesting gulls report between 250-435 nests (mostly of Lesser Black-backed Gulls but also some Herring Gulls) in Dumfries (Coulson & Coulson 1999-2003), concurrent with, but not reported in the JNCC Seabird 2000 database.

Only two Local Authorities (Aberdeenshire Council and Dundee City Council) raised queries over counts of breeding urban gulls recorded during the Seabird 2000 survey.

In the case of Dundee, the Seabird 2000 survey appeared to have underestimated the numbers that the council believed are now present within the City boundary. Further investigation, including consultation with the observer who provided the Seabird 2000 counts, has suggested that a mixture of factors might be responsible for the apparent discrepancy, including: recent changes to the characteristics of one of the sites (rendering counting somewhat easier in the years following the Seabird 2000 survey); an area containing breeding gulls in the docks possibly missed during the Seabird 2000 counts; differences in the count units used by Seabird 2000 observers and Council staff (i.e. breeding pairs versus total counts of all individuals including non-breeders); and possibly also a real increase in numbers (M.Nicoll, pers. comm.). Unfortunately, the confounding factors make it difficult to assess whether any recent change in numbers has actually occurred. In the case of Aberdeenshire, it is felt that the Seabird 2000 estimate is too low based on the Council's own counts of nests that have been removed from Council-owned properties (C. Campbell, pers. comm.).

JNCC provided us with details of some counts of urban and roof-nesting gulls in addition to those collated for the three national surveys. Any duplicates with the principal databases were removed when we collated this data to provide the population estimates presented in the current report. Of the remaining counts, for purposes of this review, those for roof nesting gulls made between 1990 and 1996 inclusive were added to those for the 1993-95 survey and those from 1997-98 were added to the 1998-2002 survey.

The specific surveys of roof nesting gulls in 1976 and 1994 both collated past records and circulated requests for information on the locations of colonies, including negative counts. The latter provide a sounder basis for determining whether colonies detected only in later surveys were newly established or simply were not counted previously, but unfortunately the reporting of such negative counts was not comprehensive (Monaghan & Coulson 1977); in addition, negative counts that were reported ( i.e. no breeding gulls present in a given geographical area) may have been unreliable in some cases ( i.e. the area in question was not checked by the respondent). Observed increases in breeding gull numbers between these two specific surveys may be overestimated if colonies were unknown or unreported during the earlier one.

Timing of counts

The timing of survey work for the 1976 survey is not reported. For the 1994 and 1999-2002 surveys, observers were asked to undertake counts between mid-May and early June (Raven & Coulson 1997, Mitchell et al. 2004) to coincide with the period when the maximum number of nests is expected to be occupied (Wanless & Harris 1984). The actual dates when surveyors counted breeding gulls are not available for either the 1976 or 1994 surveys. For the 1998-2002 survey, however, the count date is known for 73% of the counted AONs of all gulls in urban or man-made areas. Of these, just 29% were within, or close to the optimum period ( JNCC data). Based on what is known about the influence of the timing of counts on estimated breeding numbers at natural colonies, the most recent estimates of urban gull breeding numbers would be expected to be underestimates. The extent to which this is the case is unknown, however. In fact, counts that fall later than the optimum window for counting might cause less bias to estimates in urban than in natural gull colonies because of the generally higher breeding success at urban sites (such that few pairs might fail before late counts were made, see Section 5.2). However, an additional problem related specifically to urban-nesting gulls is the extent to which some nests may have been removed by residents in urban areas before counts were made starting in May: for example, in some areas where specific counts of urban-nesting gulls have been made, 10-15% of nests had been removed by the end of May, when many counts would generally be made (J. Coulson, pers. comm.). In the absence of studies of the breeding phenology and success of urban gulls in Scotland, it is not possible to make further suggestion as to the extent, if any, to which current numbers are underestimated.

Visibility of nests

Gulls nest on a variety of substrates and even where nesting on roofs, the extent to which nesting areas are visible can be variable. Nests on the roofs of buildings are rarely visible from ground level and suitable vantage points, for example taller buildings, are not always available ( e.g. Monaghan & Coulson 1977). The roofs and buildings on which gulls nest can be complex structures and some nests (including their attending birds) may be undercover or otherwise concealed. Once again, such problems will lead to some counts being underestimates. Of all the AONs of gulls counted in urban and man-made sites in Scotland during 1998-2002, 78% were made from vantage points and hence could be underestimates, again by an unknown amount (see below). A further 17% of all the AONs counted were from aerial surveys, for which no ground-truthing to check for nests that were not visible from the planes is documented.

Even where rooftops, or other sites where gulls are nesting are accessible directly and visibility is good, accurate counting can be difficult. Even experienced counters miss some nests (Ferns & Mudge 1981, Wanless & Harris 1984), and appropriate correction factors will not always have been determined and applied, to account for the missed proportions. Only 3% of the AONs counted in urban and man made sites were counted from the ground rather than from vantage points or planes however, most of those involving scanning generally vertical parts of buildings and other structures ( JNCC data).

Outside of Scotland, in the Bristol area, one worker on urban gulls used a combination of counts of adult birds (assessed by plumage characteristics) and AONs to estimate the size of local breeding colonies, using additional correction factors (of between 5% and 25% depending on roof structure) to account for missed nests (Rock in prep.). Rock's study considers the data reported by Seabird 2000 for that area to underestimate breeding gull numbers by around 60%, through a combination of undercounting at individual colonies and also missing sites. The full methods used in this study were not available for us to assess directly here, however, and it is likely that no systematic validation of the correction factors used has been undertaken. Assuming that counts of birds and of nests can be correctly associated using validated correction factors, then such an approach can lead to a more accurate estimate of colony size. It follows that at least some counts of urban gull colonies in Scotland may be serious underestimates but there are no appropriate validation studies to allow further assessment of the extent of this problem. A comparable discrepancy is also found between Seabird 2000 data for roof-nesting gulls in Lerwick, Shetland and a more intensive census (Okill 2004), although the latter survey was undertaken one year later than the former so that a direct comparison may not be valid. If counts that involve correction factors to account for less than full visibility of nests are to be compared through time, then it is important that the criteria on which the correction factors are based are carefully documented. Assessment should also be made of whether the correction factors change through time ( e.g. as could occur if numbers increase and areas of the colony that differ in character from the initial area become occupied).

Counts based on individual birds

At some gull colonies, counts have been made of attending adults, rather than of nests or nest sites ( e.g. where nests are not visible but birds are). For Seabird 2000, such counts were systematically divided by two to give a count of AONs (Mitchell et al. 2004). Such correction approaches are biased according to variation in the actual attendance rates of adults at nests; less than 1% of all urban and roof nesting AONs in Scotland during Seabird 2000 were derived in this manner, however, so that the influence of such bias on overall estimates is negligible.

Assigning proportions to species in mixed colonies

In mixed colonies, the determination of the proportion of a count to assign to a particular species can provide a further potential source of error if counts are based on nests that cannot readily be identified as belonging to a particular species. As the majority of counts of urban gulls have been made from vantage points, most AONs will have been assigned to species by identification of the bird(s) in attendance. For the 17% that were counted by aerial survey during Seabird 2000, specific identification during the surveys, or from photographs, would have been difficult or impossible however. Recommended survey methods suggest that representative sample head counts are made to determine the proportion of each species present. The two species most likely to be involved are Herring Gull and Lesser Black-backed Gull. There are potential biases associated with differences in attendance rates at colonies between the two species, but at sites where these have been determined, the direction of bias ( i.e. for Herring or for Lesser Black-backed) is confounding (Calladine 2004), such that the effect on survey estimates is unknown.

Derivation of trends

Although the methods of counting during earlier surveys of urban gulls are less precisely documented than for Seabird 2000, the majority of counts will have been from vantage points (Monaghan & Coulson 1977, Raven & Coulson 1997). Although the actual counts are likely to have been underestimates in many cases (for reasons given above), where there have been subsequent counts for the same areas, comparisons to establish general trends should still be valid. Where different methods have been used, for example counts from vantage points in one survey and aerial counts in another, however, in the absence of any calibration of the two methods comparison should be made with caution. Aerial surveys were used only in the 1998-2002 survey and for only 17% of the total AONs counted, so that any effect on national trends should be relatively minor but more local trend information may be biased and should be treated with caution. For urban nesting gulls, counts from vantage points have been the principal method used during all three comprehensive seabird surveys and differences in counting methodologies are unlikely to be a major factor contributing to apparent changes in the numbers of breeding gulls in urban areas.

4.1.2 Outside the breeding season

General ecology

Outside of the breeding season, the requirement for secure nest sites is absent and gulls generally become more widespread. They still require secure roosting sites that are within range of suitable foraging areas. Gulls typically roost communally and often in mixed species groups. Roosting sites are normally on open water, including the sea in calm conditions, although some roosts and 'pre-roost gatherings' can be located on buildings. As birds are not tied to attending eggs or young at this time, wintering gulls are able to range more widely and will not necessarily use the same roosts and foraging areas throughout the non-breeding season. Hence, some individual gulls may make more transitory use of urban areas in winter.

Surveys, survey methods and limitations

A systematic survey due to be completed during winter 2005/06 aims to provide the first comprehensive estimate of gull numbers in Britain during winter, including a measure of the precision of that estimate (Burton et al. 2004). The general distribution of wintering gulls in Britain, including measures of relative abundance, was mapped for the three winters 1981/82 to 1983/84 (Lack 1986), and coordinated mid-winter counts at some larger roosts in Scotland have been made in 1983, 1993 and 2003 (Bowes et al. 1984, Burton et al. 2003, Burton et al. 2004). These latter surveys targeted major gull roosts, where over 1000 gulls were known or expected to congregate (Figure 4.2). Based on counts from 116 roosts in 1993, minimum mid-winter population estimates for Scotland have been produced (Burton et al. 2003) but the numbers of birds roosting away from those large roosts identified for the 1993 survey will not be known until completion of a survey of randomly selected tetrads (2km by 2km squares) in 2005/06 (Burton et al. 2004). The 2005/06 winter roost survey has not been designed to provide estimates of the numbers of gulls using urban areas in winter.

Figure 4.2
Key roosts targeted for the mid-winter surveys of roosting gulls in 1983, 1993 and 2004 (Source: BTO data).

Derivation of trends

Indices of change in the numbers of gulls wintering in Scotland have been produced already, based on counts from between 41 (for Lesser Black-backed Gull) and 112 (for Herring Gull) roosts that have been surveyed in more than one of the three coordinated mid-winter counts (Burton et al. 2004). Trends assessed from mid-winter counts of large roosts do not necessarily reflect trends in abundance within urban areas nor their abundance at other times outside of the breeding season. They should reflect broad trends in the numbers of gulls in Scotland at that time of year however. The distribution of roosts for which the trends have been derived to date are concentrated on the east coast, in the central belt and south-west Scotland (Figure 1). The winter trend information currently excludes some important areas for wintering gulls (all species except the Lesser Black-backed Gull), notably on the west coast, the Western Isles and in Shetland (Lack 1986) and, therefore, may not be representative of changes in the numbers of wintering gulls throughout Scotland.

4.1.3 Questionnaire survey

The questionnaire survey that was sent to every Local Authority in Scotland (Section 2.3 and Appendix 1) asked respondents for any information that they had on the size and trends in numbers in their areas, together with how any gulls were distributed spatially and details of the methods used to obtain any counts that they provided. The resulting information was used to compare with that obtained from the Joint Nature Conservation Committee from the 'Seabird 2000' survey and, alongside other more local information sources, was used to investigate any possible gaps in Seabird 2000 coverage. None of the counts provided as a result of the questionnaire survey provided detail of the numbers of gulls using urban areas in winter.

4.2 Species accounts

4.2.1 Black-headed Gull (Larus ridibundus)

The majority of Black-headed Gulls breeding in Scotland do so at inland sites. An estimated 43,191 apparently occupied nests ( AONs) were counted in 1998-2002, most in the eastern half of the country (Dunn 2004). Earlier comparable counts are only available for coastal colonies, which represented 16% (6,888 AONs) of the total breeding population in 1998-2002; in 1998-2002, they showed a marked decline from 18,226 AONs in 1969-70 and 9,554 AONs in 1985-88 (Dunn 2004). The Black-headed Gull is currently on the Amber List of Birds of Conservation Concern in the UK, based on the moderate decline (25-49%) in its breeding population over the past 25 years (Gregory et al. 2002).

The numbers of Black-headed Gulls present in Scotland are boosted in the winter months with immigrants of mostly northern European breeding origins (Mackinnon & Coulson 1987; Fletcher 2002). In winter, the Scottish distribution of the species is concentrated along the east coast, in the central belt and around the Solway Firth and Firth of Clyde (Flegg 1986). The minimum number of Black-headed Gulls roosting in Scotland in January 1993 was 155,539 (Burton et al. 2003). Based on counts from 101 roosts for which more than one year of data were available, this represented an increase of 26% since 1983, but there has been a subsequent decline of 66% between 1993 and 2004 (Burton et al. 2004).

Black-headed Gulls were excluded from the specific roof-nesting surveys of 1996 and 1994 and were not recorded during 1969 (Cramp 1971). During Seabird 2000 (1998-2002), nesting Black-headed Gulls were only reported in four areas defined as urban, City of Aberdeen (68 AONs), industrial site at Dyce (1050 AONs), Montrose (1 AON) and Hunterston (50 AONs). Other reported sites are defined as rural, although some of the 100 AONs were on buildings at Carstairs Junction, Clydesdale (defined as rural) (Table 4.1, Figure 4.3). Although Black-headed Gulls were excluded from the specific surveys of roof-nesting 'large gulls', as no mention is included in earlier reports (Monaghan & Coulson 1977, Raven & Coulson 1997), it is likely that the occurrence of this species nesting on roofs is truly relatively recent.

4.2.2 Common Gull (Larus canus)

The majority of breeding Common Gulls in Britain are found in Scotland, with an estimated 48,113 AONs counted during 1998-2002, about half of which breed inland and half on the coast and this species is possibly the most widespread gull in Scotland (Tasker 2004). The coastal portion of the breeding population was given as 20,467 AONs in 1998-2002, representing increases from 12,229 in 1969-70 and 15,134 in 1985-88 (Tasker 2004). The distribution of inland breeding Common Gulls has contracted over the same period, with marked declines at some large colonies in north-east Scotland (Tasker 2004). Based on the decline of the overall population, and also an 'unfavourable status' elsewhere in Europe, the Common Gull is currently on the Amber List of birds of Conservation Concern in the UK (Gregory et al. 2002).

As with Black-headed Gulls, numbers of Common Gulls are augmented in winter by birds of largely northern European breeding origins (Douse 2002). Although the species is still widespread in Scotland in winter, many areas in the Highlands are vacated (Vernon 1986). In January 1993, the minimum number of Common Gulls wintering in Scotland was 79,742 (Burton et al. 2003). Although there is some evidence of a decline between 1983 and 2004, the change is not statistically significant based on counts from 103 roosts with more than one year of count data (Burton et al. 2004).

A single nest of Common Gulls was reported on a building in 1969, at Dalcross Airport, Inverness (Cramp 1971) and none was reported during 1974-76 (Monaghan & Coulson 1977). Roof-nesting has occurred in Aberdeen since at least 1984 (Sullivan 1985) where, by 1993-95, 182 AONs were present (Table 4.2, Raven & Coulson 1997). The incidence of roof nesting has continued to increase and become more widespread (Table 4.2, Figure 4.4). By 1998-2002, the largest reported concentration of urban and roof-nesting Common Gulls was in Aberdeen, with 410 AONs (Table 4.2), including 280 AONs reported from the City centre and the rest on an industrial estate at Dyce. Over 50% of Common Gulls nesting on buildings were outside of areas classed as urban (Table 4.2), with some concentrations in industrial complexes on the periphery of urban areas (Tasker 2004). Roof nesting Common Gulls have not been reported in Britain and Ireland outside of Scotland (Tasker 2004).

4.2.3 Lesser Black-backed Gull (Larus fuscus)

During 1998-2002, 25,057 AONs of breeding Lesser Black-backed Gulls were estimated in Scotland, with concentrations in the Firths of Forth and Clyde and across the central belt (Calladine 2004). The majority of Lesser Black-backed Gulls in Scotland breed on the coast (86%, 21,565 AONs). The 1998-2002 count of coastal breeding Lesser Black-backed Gulls was the highest of the three comprehensive surveys: 12,031 in 1969-70 and 19,524 in 1985-88. For some colonies that have been monitored annually, peaks were reached in the 1990s and there is some evidence of recent declines however (Calladine 2004). The Lesser Black-backed Gull is currently on the Amber List of Birds of Conservation Concern, principally because the UK holds the majority of the global breeding population of the race L. f. graelsii (Gregory et al. 2002).

The majority of Lesser Black-backed Gulls leave Scotland after breeding, spending the winter at locations extending between southern Britain and north-west Africa (Rock 2002). There has been an increasing tendency for more birds to remain within their breeding range throughout the year over recent decades, however, although numbers in Scotland during the winter months are still considerably less than in the summer. In winter, they tend to be concentrated around the Firth of Clyde and the Solway Firth (Baker 1980, Hickling 1986a, Rock 2002), with a minimum of 452 counted at roosts in January 1993 (Burton et al. 2003). In January 2004, based on counts at 41 winter roosts, there has been a subsequent 112% increase in the number of Lesser Black-backed Gulls in Scotland in mid-winter (Burton et al. 2004).

Lesser Black-backed Gulls have been reported nesting on buildings in the Bristol Channel area since the 1960s (Cramp 1971). By 1976, small numbers were found in north-east England (Monaghan & Coulson 1977). Roof nesting had become widespread across the Central Belt and occurred elsewhere on the east coast of Scotland by 1993-95 (Table 4.3, Raven & Coulson 1997), with numbers generally increasing by the 1999-2002 survey (Table 4.3, Calladine 2004). The vast majority (>98%) of the reported 4,309 AONs in 1998-2002 were in urban areas, the remainder being largely within industrial complexes ( JNCC data). Most roof-nesting Lesser Black-backed Gulls in Scotland are found currently in the Central Belt, where they now breed in many urban areas (Figure 4.5). Concentrations are reported in the western parts of the Central Belt ( e.g. Glasgow, Renfrew and Cumbernauld: Table 4.3). Recent counts from the east (Edinburgh and its surrounds) are lacking, although significant numbers are thought to be present: at the last attempt at a complete count in 1994, a minimum of c.80 pairs was estimated for Edinburgh and its surrounds, with a further c.30 pairs on roofs at Granton harbour (Dott 1994). Outside of the Central Belt, breeding Lesser Black-backed Gulls are found in numbers in the City of Aberdeen, Dumfries and in Ayrshire (Figure 4.5). The first instance of roof-nesting by Lesser Black-backed Gulls in Shetland was reported in 2003 (Okill 2004).

4.2.4 Herring Gull (Larus argentatus)

A predominantly, though not exclusively, coastal breeding species in Scotland, 72,130 AONs of Herring Gulls were estimated during 1998-2002 (Madden & Newton 2004). The coastal proportion of that count (71,659 AONs) compares to 159,237 in 1969-70 and 92,950 in 1985-88 (Madden & Newton 2004). Based on similar declines across the UK, the species is currently placed on the Amber List of Birds of Conservation Concern (Gregory et al. 2002), and recent trends appear to qualify it for the Red List ('high concern', Mitchell 2004).

Many Herring Gulls from Scotland disperse south outside of the breeding season, with substantial numbers moving to the Irish and southern North Sea coasts. Wintering numbers in Scotland are augmented, particularly on the east coast, by birds from of Fenno-Scandinavia breeding origins (Calladine 2002). The winter distribution of Herring Gulls is broadly similar to that during the breeding season, although there are greater concentrations inland during the winter, notably in the Central Belt (Hickling 1986b). In January 1993, a minimum of 90,972 Herring Gulls roosted in Scotland (Burton et al. 2003). Although the derived index based on counts from 112 roosts suggests a subsequent decline by 2004, the trend is not statistically significant (Burton et al. 2004).

Herring Gulls have nested on some roofs in southern England since at least the 1920s and in north-east Scotland since at least the 1950s (Cramp 1971). Relatively small numbers were recorded nesting in Scottish east coast towns by 1974-75, and these had increased substantially, including occupation of rooftops in the Central Belt, by 1993-95 (Monaghan & Coulson 1977, Raven & Coulson 1997, Table 4.4). With the notable exception of the City of Aberdeen, numbers reported at many sites already holding roof-nesting Herring Gulls by 1993-95 were similar in 1998-2002. Many additional sites had either become occupied, or were first reported, in the intervening period, however, and the Herring Gull remains the most abundant roof- and urban-nesting gull in Scotland (Table 4.4). As with the Lesser Black-backed Gull, the majority (>95%) of Herring Gulls reported as nesting on buildings (6,202 AONs) in 1998-2002, did so in urban areas (Table 4.4). The largest concentrations of roof-nesting Herring Gulls in Scotland are in the east coast towns between Dundee and Inverness, with smaller colonies across the Central Belt, in Berwickshire, Ayrshire and in Dumfries (Figure 4.6). As with the Lesser Black-backed Gull, recent data from the Edinburgh area are lacking: the most recent attempt at a full survey in 1994 recorded a minimum of c.40 pairs in Edinburgh and its surrounds and a further 12 pairs on roofs at Granton Harbour (Dott 1994). The biased distribution of roof-nesting Herring Gulls towards the east coast of Scotland contrasts markedly with the concentrations of roof-nesting Lesser Black-backed Gulls in the Central Belt and broadly reflects the differences in the breeding distributions of the two species in the wider countryside (Gibbons et al. 1993, Mitchell et al. 2004).

4.2.5 Great Black-backed Gull (Larus marinus)

With an almost exclusively coastal breeding distribution, 14,776 AONs of Great Black-backed Gulls were estimated during 1998-2002, concentrated in the Northern Isles and on the north and west coasts (Reid 2004). In 1969-70, 15,950 and in 1985-88, 15,315 AONs were counted in Scotland. The species is currently on the Green List of Birds of Conservation Concern in the UK (Gregory et al. 2002).

Although some Great Black-backed Gulls disperse south after breeding, principally to the coasts of the Irish and North Seas, birds of Fenno-Scandinavia breeding origins winter in Scotland (Reid 2002). The winter distribution remains similar to that of breeding birds, although birds also occur in inland, lowland areas during the winter (Monaghan 1986). A minimum of 2,970 Great Black-backed Gulls roosted in Scotland in January 1993 (Burton et al. 2003), with a suggested decline, based on counts from 96 roosts, of 23% between the 1993 count and January 2004 (Burton et al. 2004).

The only records of roof nesting by Great Black-backed Gulls in Britain up to 1974 were in Cornwall (Cramp 1971, Monaghan & Coulson 1977). By 1994, a total of 6 pairs was reported nesting on buildings in Scotland (Raven & Coulson 1997). The number reported in urban areas was 21 AONs by 1998-2000, with sizeable colonies also established at the Nigg Oil Terminal in Easter Ross (136 AONs, principally on industrial structures) and 57 AONs at Quedal on Rousay, Orkney (Figure 4.7); at the latter site, man-made structures were not recorded as the principal habitat at the colony however ( JNCC data). Great Black-backed Gulls have also been recorded nesting in Dumfries in 2003 (one pair) and 2004 (two pairs; J. Coulson pers. comm.).

4.2.6 Kittiwake (Rissa tridactyla)

Breeding exclusively on the coasts, 282,213 AONs of Kittiwake were estimated in Scotland during 1998-2002, with most colonies on the North Sea coast and in the Northern Isles (Heubeck 2004). The two earlier surveys found greater numbers, with 359,425 AONs estimated in 1985-88 and 346,097 AONs in 1969-70. Due to the recent decline, the Kittiwake is currently on the Amber List of Birds of Conservation Concern (Gregory et al. 2002).

Unlike the other five species of breeding gulls considered in this report, Kittiwakes feed mainly on pelagic shoaling fish, although they also take fishing discards. They do not normally feed inland or utilise garbage (Heubeck 2004). Outside of the breeding season, Kittiwakes are found almost exclusively at sea, and although they are present around all Scottish coasts throughout the winter, Scottish-bred birds can range widely, including crossing the Atlantic (Coulson 1986, 2002).

Although Kittiwakes do nest on some buildings and other man-made structures at the coast, and have done so in Scotland since at least the 1930s (Cramp 1971), and also on some natural cliffs in urban areas, this species has not been identified with problems associated with urban gulls in Scotland (Chapter 3).

Figure 4.3
The distribution of urban and roof-nesting Black-headed Gulls in Scotland reported during the 'Seabird 2000' survey (1998-2002). Numbers in the legend refer to AONs. Crosses show additional from which urban and roof-nesting Black-headed Gulls were reported during 1968-1998 (Source: JNCC data).

Figure 4.4
The distribution of urban and roof-nesting Common Gulls in Scotland reported during the 'Seabird 2000' survey (1998-2002). Numbers in the legend refer to AONs. Crosses show additional sites from which urban and roof-nesting Common Gulls were reported during 1968-1998 (Source: JNCC data).

Figure 4.5
The distribution of urban and roof-nesting Lesser Black-backed Gulls in Scotland reported during the 'Seabird 2000' survey (1998-2002). Numbers in the legend refer to AONs. Crosses show sites from which urban and roof-nesting Lesser Black-backed Gulls were reported during 1968-1998 (Source: JNCC data, Coulson & Coulson 1999-2002).

Figure 4.6
The distribution of urban and roof-nesting Herring Gulls in Scotland reported during the 'Seabird 2000' survey (1998-2002). Numbers in the legend refer to AONs. Crosses show additional sites from which urban and roof-nesting Herring Gulls were reported during 1968-1998 (Source: JNCC data, Coulson & Coulson 1999-2002).

Figure 4.7
The distribution of urban and roof-nesting Great Black-backed Gulls in Scotland reported during the 'Seabird 2000' survey (1998-2002). Numbers in the legend refer to AONs. Crosses show additional sites from which urban and roof-nesting Great Black-backed Gulls were reported during 1968-1998 (Source: JNCC data).

Table 4.1

Numbers of Apparently Occupied Nests ( AONs) of urban and roof-nesting Black-headed Gulls reported during three survey periods, by administrative area of Scotland (Source: JNCC data).

Administrative area	1974-76	1993-95	1998-2002 (urban only)	1998-2000 (urban and other roof-nesting)
Angus	.	.	1	1
Argyll and Bute	.	.	0	19
Banff and Buchan	.	65	0	22
City of Aberdeen	.	175	1118	1118
Clydesdale	.	.	0	100
Cumnock and Doon Valley	.	.	0	0
Cunninghame	.	210	50	50
Eastwood	.	.	0	12
Inverness	.	.	0	67
Kincardine and Deeside	.	.	0	1
Ross and Cromarty	.	.	0	1
TOTAL	0	450	1169	1391

Table 4.2

Numbers of Apparently Occupied Nests ( AONs) of urban and roof-nesting Common Gulls reported during three survey periods, by administrative area of Scotland (Source: JNCC data).

Administrative area	1974-76	1993-95	1998-2002 (urban only)	1998-2002 (urban and other roof-nesting)
Angus	.	.	19	19
Argyll and Bute	.	1	40	97
Banff and Buchan	.	.	0	1
Caithness	.	12	86	86
City of Aberdeen	.	182	410	410
Clydesdale	.	.	0	12
Cumbernauld and Kilsyth	.	.	3	3
Cunninghame	.	.	4	4
Dumbarton	.	.	40	77
Eastwood	.	.	0	38
Gordon	.	.	0	77
Inverclyde	.	2	19	22
Inverness	.	20	96	107
Kincardine and Deeside	.	.	0	210
Lochaber	.	25	20	20
Moray	.	.	0	5
Nairn	.	.	0	7
Orkney	.	.	0	67
Perth and Kinross	.	.	0	1
Renfrew	.	.	0	8
Ross and Cromarty	.	34	0	366
Shetland	.	.	1	1
Sutherland	.	.	13	17
Western Isles - Comhairle nan eilean			0	1
TOTAL	0	276	751	1656

Table 4.3

Numbers of Apparently Occupied Nests ( AONs) of urban and roof-nesting Lesser Black-backed Gulls reported during three survey periods, by administrative area of Scotland (Sources: JNCC data, Coulson & Coulson 1999-2002).

Administrative area	1974-76	1993-95	1998-2002 (urban only)	1998-2002 (urban and other roof-nesting)
Angus	.	0	7	7
Banff and Buchan	.	5	4	27
Bearsden and Milngavie	.	10	118	118
City of Aberdeen	.	50	159	159
City of Dundee	.	7	65	65
City of Edinburgh	.	113	.	.
City of Glasgow	.	140	640	640
Clackmannan	.	.	88	88
Clydebank	.	.	30	30
Clydesdale	.	.	40	40
Cumbernauld and Kilsyth	.	350	449	449
Cunninghame	.	32	97	97
Dumbarton	.	175	134	143
Dunfermline	.	.	38	38
East Kilbride	.	.	250	250
East Lothian	.	1	.	.
Eastwood	.	.	39	39
Falkirk	.	4	179	179
Inverclyde	.	141	189	189
Inverness	.	.	6	6
Kilmarnock and Loudon	.	1	134	134
Kirkcaldy	.	.	6	6
Kyle and Carrick	.	143	311	311
Moray	.	1	28	28
Motherwell	.	.	34	34
Nithsdale	.	.	370	370
Renfrew	.	211	518	518
Ross and Cromarty	.	.	0	6
Strathkelvin	.	12	329	329
Sutherland	.	.	1	1
West Lothian	.	15	.	.
Western Isles - Comhairle nan eilean		.	0	8
TOTAL	0	1411	4263	4309

Table 4.4

Numbers of Apparently Occupied Nests ( AONs) of urban and roof-nesting Herring Gulls reported during three survey periods, by administrative area of Scotland (Sources: JNCC data, Coulson & Coulson 1999-2002).

Administrative area	1974-76	1993-95		1998-2002 (urban only)	1998-2002 (urban and other roof-nesting)
Angus	0		448	398	398
Banff and Buchan	.		322	444	473
Bearsden and Milngavie	.		.	6	6
Berwickshire	22		125	194	194
Caithness	.		131	50	50
City of Aberdeen	1		2020	3370	3370
City of Dundee	9		.	296	296
City of Edinburgh	.		43	.	.
City of Glasgow	.		6	19	19
Clackmannan	.		.	35	35
Clydebank	.		.	0	0
Cumbernauld and Kilsyth	.		13	17	17
Cunninghame	.		4	19	19
Dumbarton	.		5	10	30
Dunfermline	.		.	18	18
East Kilbride	.		.	20	20
East Lothian	.		23	.	.
Eastwood	.		.	0	1
Falkirk	.		.	24	24
Inverclyde	.		35	40	40
Inverness	3		150	356	356
Kilmarnock and Loudon	3		.	15	15
Kirkcaldy	.		.	10	10
Kyle and Carrick	.		159	144	144
Moray	1		32	153	164
Motherwell	.		.	1	1
Nairn	9		.	80	80
Nithsdale	.		3	65	65
North East Fife	.		2	.	.
Orkney	.		.	0	1
Perth and Kinross	.		.	1	1
Renfrew	.		.	0	1
Ross and Cromarty	1		35	83	295
Shetland	6		59	21	21
Strathkelvin	.		.	5	5
Sutherland	.		.	33	33
West Lothian	.		5	.	.
Wigtown	0		.	.	.
TOTAL	55		3620	5927	6202

Table 4.5

Numbers of Apparently Occupied Nests ( AONs) of urban and roof-nesting Great Black-backed Gulls reported during three survey periods, by administrative area of Scotland (Source: JNCC data).

Administrative area	1974-76	1993-95	1998-2000 (urban only)	1998-2000 (urban and other roof-nesting)
Angus	.	.	2	2
Banff and Buchan	.	.	.	.
City of Aberdeen	.	2	9	9
City of Edinburgh	.	1	.	.
Cunninghame	.	1	1	1
Dumbarton	.	.	0	3
Inverclyde	.	1	2	2
Inverness	.	.	5	5
Kyle and Carrick	.	1	.	.
Moray	.	.	1	1
Orkney	.	.	0	58
Ross and Cromarty	.	.	0	255
Sutherland	.	.	1	1
TOTAL	0	6	21	337