( Figure 7 to the right) shows a gull that has limited moult of upper-wing coverts, worn greater coverts and tertials and a bill that is quite stubby, with a well-streaked head. Seeing this gull apart from any other first-basics would cause difficulty in identifying it, indeed there are a number of less typical YLG features, aside form the more restricted moult, which might point to it being a LBbG. (Figure 8 below left) shows another first-basic alongside a second-basic that also has little (or no) moult of the tertials and wing coverts, which causes problems with its identity. However, in the field this gull was equal in size to the other YLGs, its body plumage is most like YLG, and the extreme wear and bleaching of the juvenile wing coverts, would suggest that it is YLG.

The second-basic next to it and (Figure 9 above right) show the typical plumage features and bare part colour. The mantle is totally grey (or almost so) and there are some (actually few) third generation, grey coverts, whilst the second generation coverts are very worn and faded (obviously due to their early moult in the first or second calendar year). In comparison, typical 2nd-basic michahellis seen in Britain, show fresher and a lot less worn 2nd-generation coverts. The pattern of the 2nd-generation tertials is interesting in having a ‘saw-toothed’ edge.

In flight (Figures 10 and 11 below) the tail band is fairly broad and at times barred towards the base, the mantle contrasts with faded brown upper-wing coverts and black secondary band. The inner primaries again tend to be darker than on typical michahellis.

The final two images (Figures 10 and 13 below) show a third-basic, with more extensive grey in the plumage and more adult-like bare part colours, and an adult in flight. The adult shows the dark mantle tone (in bright sunlight it is sometimes hard to differentiate from graellsii), the large extent of black in the primaries (though no more than some michahellis), but especially the small square or rectangular, white mirror. This pattern is fairly typical of the adults on Lanzarote and Fuerteventura (with black extending onto p4) and is associated with atlantis, though by no means all show such a pattern. Some show a larger white area on p10 only; some have a small white spot on p9; others can have a totally white tip to p10 and fair sized mirror on p9. By January, there was no sign of the ‘hooded’ appearance quoted for adults of atlantis.

Conclusions
What conclusions can we draw from the slightly superficial treatment I have given such a complex issue? Firstly, I would say that the separation of the Yellow-legged Gulls on the Canary Islands as atlantis is by no means clear. The primary pattern is extremely variable, albeit with a tendency towards smaller mirrors that is all, and the mantle is darker than most but not all michahellis; certainly, I would suggest that racial identification on these features is tenuous (michahellis varies throughout its range), but what else is there? The call seems virtually identical to me; the structure seems subtly larger, with a strong bill and GBbG look, though what they would look like alongside michahellis I don’t know. Perhaps the young birds give us a clue?

Sometimes, certainly in the past, the immature plumages of gulls of the LWHG complex have been ignored when discussing the relationships and positions of various taxa – I like to think, probably naively, that different ‘species’ produce diagnosably different young. Two recent important papers in British Birds - (Baraba Gull by Panov and Monzikov, British Birds 93: 227-241, 2000 and Mongolian Gull by Yesou 2001) ignore this area entirely, which is shame, because I suspect that studies would also provide clues that they are producing young that look different form those seen as being taxonomically ‘close’ – compare first-basic cachinnans and mongolicus, which have often been closely related.

Recently, experience has shown that Caspian Gull has immature plumages that are distinct from Yellow-legged Gull, and at times are easier to separate than adults. The adults of michahellis and the ‘atlantis’ on the Canary Islands are so similar and variable that the extremes of each encompass each other. In reality, the first-basic plumages of the YLGs on the Canary Islands are also only subtly different from michahellis. If you went for any features it would be the even more advanced state of moult, whilst the tail pattern is one of the better features. The darker inner primaries, when present would help, but are they enough to think that they are firstly a different sub-species or secondly, as has been indicated by some, a different species? If you saw one in Britain would you be able to tell from large first-basic graellsii? I leave the answers for further studies, but my opinion is: I doubt it.

Brian Small
April 2001