Nick Rossiter email: nick.rossiter1 at btinternet.com

http://nickrossiter.org.uk/ylgweb/

West Barn, Ordley, Hexham, Northumberland, England NE46 1SX

Paper presented at 6th International Gull Meeting, Lammi, Finland, 8th-11th August 2002.

The Yellow-legged Gulls found on the Atlantic coasts are considered to comprise four forms: dark, Cantabrican, a cline from dark-Cantabrican and Mediterranean Larus michahellis. The area where the Atlantic and Mediterranean forms meet is of particular significance for looking at the boundaries of the Atlantic forms. Data for the gulls from this intersection area are presented and contrasted with that from other Atlantic areas. It is thought that the intersection zone is fairly narrow stretching for no more than 40-50km to the west and 60-170km on the less researched eastern side. For the Atlantic forms as a whole, the results from these morphological studies and recent DNA studies converge to support the differentiation of dark Atlantic Yellow-legged Gull (southern atlantis) from both michahellis and northern atlantis.

Knowledge of the Yellow-legged Gull complex has substantially increased over the last decade with Klein (1994), Garner & Quinn (1997) and Jonsson (1998) describing the Mediterranean form, provisionally classified as Larus michahellis, and the form from further east L. cachinnans. Recent DNA evidence (de Knijff et al 2001) suggests that michahellis and cachinnans are not closely related. In this paper L. michahellis is sometimes called Mediterranean Yellow-legged Gull.

The Atlantic complex of Yellow-legged Gulls has been relatively neglected. This may be because it appears that its members are more sedentary and less likely to appear in northern Europe. Other factors include variability, at least at a fine level of detail from locality to locality, so that it has been difficult to define precisely its characteristics. The traditional classification (Dwight, 1922; 1925) of Yellow-legged Gulls is to include those found on the Atlantic islands of the Azores, Canaries and Madeira as atlantis and those on the Atlantic continental coasts of Morocco, Portugal and Spain as michahellis. However, some authors have acknowledged that the birds on the Atlantic continental coasts have some resemblance to atlantis (Stegmann 1934; Dwight 1925; Bannerman 1963; Etchécopar & Hüe 1967). Jonsson (1998) thought that birds breeding on the Atlantic coast of the Iberian peninsula and Morocco stand morphologically closer to atlantis and are better treated as such.

To obtain objective information on their appearance, field studies have been made of Yellow-legged Gulls by Rossiter (2002) since 1990 in the Canary Islands, Madeira, Morocco (Agadir-Tanger), Spain (Andalucia, Minho, Santander), Portugal (Algarve, Setúbal-Porto) and France (Gironde). An obvious omission is the Azores. Southwards the range also extends through Western Sahara to the very north of Mauritania.

The full results of this work, including statistics for wingtip patterns in adults, calls, moult, first-year, second-year and third-year characteristics are available at Rossiter (2002). It was considered that four forms of Yellow-legged Gulls are found on the Atlantic:

1. A dark form in the Canary Islands, Madeira, Morocco (Atlantic coast Agadir northwards) and south Portugal (Algarve). Other studies as above suggest the range extends to the Azores, Western Sahara and the very north of Mauritania.
2. A pale form on the northern Atlantic coast of Spain, sometimes called Cantabrican Yellow-legged Gull (Teyssèdre 1984).
3. An intermediate form, a cline between the dark and pale forms, in western Portugal and north west Spain (Bermejo 1999).
4. Mediterranean Yellow-legged Gulls breeding in south west and western France (Yésou, 1991) and western Andalucia in Spain.

Within each range there is significant variation so populations are not as homogeneous as in say Mediterranean Yellow-legged Gulls. However, in terms of size, structure and calls, populations 1)-3) are clearly different from Mediterranean Yellow-legged Gulls. The Atlantic form in general is smaller (by 5-10%); has a less attenuated appearance at rest; shorter legs, typically 70% of height of body above legs compared to about 100%; rounded heads, lacking angular appearance; a blunter wingtip giving bunched primary tips on the folded wing; slightly smaller white primary tips on the folded wing; slightly darker mantle shade with a marked blue-grey sheen; shriller long calls lacking a guttural component, less deep mew calls; long calls given at 60° rather than 90°; ochre tint to the yellow legs and less white in the wingtip with only 30% carrying a mirror on P9. The dark form is characterised by a further 17 characteristics including aspects such as a relatively dark mantle, even less (22%) with a mirror on P9 and a heavy bill with steep gonydeal angle and a steep angle at the culmen (Rossiter 2002).

If we are to postulate that the Atlantic and Mediterranean Yellow-legged Gulls are different taxa, it is obviously important to determine what the populations of Yellow-legged Gulls look like in the intersection area between the Mediterranean and the Atlantic. The purpose of this paper is to report on the preliminary findings here.

Visits were made to the meeting points of the Atlantic and the Mediterranean at Tanger in Morocco (August 1993, 1994 and 2001), Asilah to the south west of Tanger in Morocco (August 1994), Nerja and Estepona in east Andalucia (May 1996; August 2001), Cádiz, Conil and Barbate in west Andalucia (August 2001) and Tarifa, Gibraltar and La Linea in the Straits of Gibraltar (August 2001). The detailed results of a more recent visit in June 2002 to west and east Andalucia are not available yet but some observations are made from it. Gulls were observed closely at each of these locations and many photographs taken for subsequent analysis of aspects such as wingtip pattern.

From a study of all features, there appear to be three forms in the Atlantic/Mediterranean intersection area:

• Large size being considerably bigger than graellsii.
• Relatively pale mantle, lacking obvious blue tones.
• Boxy heads, long heavy bill with blunt tip, sturdy long legs equal in length to height of body above.
• Head moult: most (60%) showed brown speckles over the forehead and the rest of the head (not neck); the remainder were often fairly white still.
• Leg colour was a pale yellow with a slight orange tint; feet were yellow.
• Bill was a rather washed out yellow with a not very strong orange-red gonydeal spot.
• Calls: the mew was like a deep graellsii; the long deeper than graellsii and hoarser (more guttural).

In the visit in June 2002, breeding colonies of the dark form were located at Cádiz on the islands to the north west of the main port and more interestingly on the cliffs at Barbate some 40km from the Tarifa area. Adult Mediterranean Yellow-legged Gulls were seen over a number of marismas from Barbate to Sanlúcar suggesting breeding in these areas. There appeared to be no Mediterranean Yellow-legged Gulls breeding in the large colony of the dark form on the Barbate cliffs. The intersection form was noted breeding on islands at Tarifa and in the port area at Estepona. The forms present at each locality appeared to be similar to those seen in August 2001.

A statistical analysis of adult wingtip pattern, calls and first-year features was made. The full results are available at (Rossiter 2002). Here two tables are given showing for comparative purposes the data for all Atlantic localities. Table 1 shows the wingtip patterns for adults derived from an analysis of the photographs for all the sites visited. Data for the individual localities are shown first, followed by summary data for each postulated form. The rows, in bold italics, are the data for the Atlantic-Mediterranean intersection area. The total sample size for all the Atlantic forms varies from 347-451 for the various features which seems to be adequate for analytical purposes. The samples for the individual areas are much less of course and caution should be applied in comparing one locality's results with another. The bottom line shows the data for the Mediterranean Yellow-legged Gulls in south west Spain.

Table 1 shows that the wingtip patterns for the intersection and the Mediterranean form are similar with P9 showing a mirror in 32% of the former and 35% of the latter. Both these forms also show a high incidence of a broad band on P5 (62% and 67% respectively) and of a spot on P4 (44% and 38% respectively). Compared to the dark form, they both show more white and more black in the wingtip.

Table 2 shows the results for the calls observed in the field. Those for the intersection form are reasonably close to those of the dark form with 77% of long calls like shrill graellsii compared to 58% of the dark form. The mew calls are closer still with 85% like graellsii compared to 75% of the dark form. The calls of the intersection form are quite different from those of the Mediterranean Yellow-legged Gulls which were deeper throughout and also hoarser in the case of the long call. Therefore it appears that the intersection form is like Mediterranean Yellow-legged Gull in wingtip and the dark form in calls. As described earlier, in structure, size, mantle shade and colour, it is intermediate between the two forms.

Primary moult in mid-August of adults was similar for all forms ranging from P5/P6 to P8/P9 with P7/P8 usually the most frequent state (46% in Mediterranean Yellow-legged Gulls). There were some differences in the juveniles. Those of the Mediterranean Yellow-legged Gull, compared to the dark form, showed less heavy markings on the breast and belly, whiter heads (except for the eye spot) and were less grey-brown with many having a ruddy tinge. The tertials showed bold white tips rather than the narrow pale fringe usually shown by the dark form. The intersection form showed varied tertial patterns including pale fringe (46%), white tip (33%) and notches (17%) compared to the uniform pattern of the dark form where 88% show a pale fringe. The intersection form was also intermediate between the dark form and Mediterranean Yellow-legged Gull in duskiness of head, breast and belly although significantly closer to the dark form.

Notes: percentages are rounded to nearest 5 for individual localities.

The intersection area appears to be narrow. On the western side it extends no more than 40-50km to Barbate cliffs in west Andalucia and Asilah in Morocco. The eastern side is not so well researched. Certainly Estepona, 60km to the east, is in the intersection zone but by time Nerja is reached, some 40km to the east of Malaga and 170km east of the Straits, most conform to Mediterranean Yellow-legged Gull although about 25% are distinctly on the small size (Rossiter, 2002). The best that can be said at present is that the intersection zone appears to be at least 80km wide and is unlikely to be wider than 220km. Jonsson’s studies (1998) at Marbella were on the eastern fringe of the intersection zone and any specimens taken from Gibraltar, for instance by de Knijff et al (2001), are from the middle of the intersection zone.

It is of considerable relevance to compare the recent DNA work by Liebers, Helbig and de Knijff (2001) with the morphological results obtained in the present study. Liebers et al consider atlantis to comprise both the dark and clinal dark-Cantabrican forms defined in the present study. They find that:

• The form michahellis has a more shallow haplotype and lower nucleotide diversity than atlantis and its haplotypes are a subset of atlantis, suggesting that michahellis is derived from atlantis as the older lineage.
• The form atlantis has held a large population over a long period, residing in an area of relative climatic stability.
• The michahellis population is less diverse than atlantis with no significant geographical variation while there is significant differentiation within all five atlantis colonies analysed.
• An unexpected differentiation was found between northern atlantis (Azores, west Portugal and Galicia) and southern atlantis (Madeira, Morocco). Since northern atlantis is similar to michahellis in many respects, this suggests that the southern atlantis contributed little to michahellis.
• Gene flow between atlantis and michahellis is probably continuing because of the connection of the Atlantic to the Mediterranean at Gibraltar. Also michahellis is breeding now on the Atlantic coast of France further increasing the likelihood of genetic exchange with atlantis.

There does appear to be considerable convergence between the genetic studies and the present morphological studies. In particular:

• The differentiation by Liebers et al between southern atlantis (corresponding to the dark Atlantic Yellow-legged Gull form of the present study) and northern atlantis (corresponding in part to the dark-Cantabrican cline of the present study) gives independent corroboration of the forms found in the Atlantic.
• The significant differentiation by Liebers et al between one atlantis population and another gives independent corroboration of the variety of forms found in the present morphological studies.
• The identification of the Atlantic Ocean-Mediterranean Sea intersection as a gene mixing zone is also supported by the present morphological studies which find a population with characteristics between atlantis and michahellis in the area.

Morphological and genetic studies suggest that the southern form of atlantis (Canaries, Madeira, Morocco and perhaps south Portugal and south west Spain), called dark Atlantic Yellow-legged Gull in this study, is differentiated from both Mediterranean Yellow-legged Gull and northern atlantis (Azores, west Portugal, north west Spain). The intersection area between the dark Atlantic form and Mediterranean Yellow-legged Gull appears to be narrow, perhaps as little as 100km and probably no more than 220km.

http://nickrossiter.org.uk/ylgweb/characteristics atlantic yellow-legged gulls.htm (June).

Bakker, T, Atlantic Yellow-legged Gulls, at http://home.wanadoo.nl/cursorius/yellowl_gull.htm

Lewis, B, Yellow Legged Gulls of Portugal, at http://www.bway.net/~lewis/birds/Portugal/port.html

Newell, D, Dick Newell's Birds, at http://magikcircle.com/birds/

Newell, D, & Reid, M, Cantabrican Yellow-legged Gulls, at http://www.martinreid.com/mcaninx.html

Regan, K, & Fisher, I, (2001), Madeiran atlantis, at http://www.wheatear.pwp.blueyonder.co.uk/gull_page.htm

Home Page Yellow-legged Gulls

© Copyright Nick Rossiter 2002.