Central Canaries (currently classified as atlantis)

Mantle: the upperparts are dark approaching graellsii in some cases but there is always a marked contrast between the black wing-tips and the inner wing, significantly exceeding that shown by graellsii. The demarcation of the black and grey is always sharper in atlantis than in graellsii. The mantle colour is a strong bright blue-grey, more like a very dark argenteus (which are pale blue-grey (Cramp, 1983)) than a pale graellsii (which are a dark ash-grey when fresh with a brown-grey sheen when faded (Grant, 1986)). Bannerman (1963) noted that atlantis in the Canaries has a blue-grey back without a brownish tinge.

Underwing: in December this shows a quite dark trailing edge almost as dark as that of graellsii on the primaries but less obvious on the secondaries. The dark trailing edge is less obvious on moulting birds in August.

Size: always appearing distinctly larger than graellsii in direct comparison; atlantis are often described as small but this is only in comparison with Mediterranean Yellow-legged Gull.

Flight appearance: into strong winds, they exhibit a characteristic jizz, similar to that of argenteus, appearing ponderous, powerful and direct with the carpal pushed well forward, the bulky chest protruding obviously and the wings appearing short and broad, particularly up to the carpal. Also, as in argenteus, the wingtip appears rather blunt. In calmer conditions they frequently soar and employ both thermals and updraughts to manoeuvre effortlessly around the rugged terrain.

Structure: far from dainty. The head appears large for the body and is usually perfectly rounded. The bill appears heavy and is of almost uniform thickness from the base to the tip. There is a pronounced gonydeal angle giving a thick appearance to the terminal third of the bill and a steep angle at the culmen which gives a rather blunt-ended appearance. The neck appears to be thick, not only when hunched but also when walking on the ground or in flight. When alarmed, as in all gulls, the neck can appear thinner. The chest is strong, always appears broad at the base and protrudes slightly. The legs are short, rather like graellsii, being about 70% of the height of the body above the legs rather than being about equal to this height as in Mediterranean Yellow-legged Gull. The feet and toes are large for the size of the bird. The overall effect is rather like a strong argenteus. A minority of birds, about 5%, are noticeably heavy with squarish heads and more protruding chests as in some Mediterranean Yellow-legged Gull. Their legs, however, are still relatively short. In the wingtip, P10 and P9 are often the same length and P8 is not much shorter giving a blunt appearance to the tip. At rest this shows on the folded wing in the position of the primary tips which is best observed in fresh plumage in winter and spring before the tips become too worn. In December it was observed that the tip of P10 is either directly under that of P9 or very slightly outside it, P8 is close to P9 and P5-P7 are widely and evenly spaced. The tips of P8-P10 are bunched as in argenteus: see SW Portugal for relevant literature.

Primaries: the main observations were in July and August. 90% of adults show a clear white mark on P10 with the remaining 10% showing an indistinct or no obvious mark. The white on P10 is in the form of a mirror or a white tip, in approximately equal proportions. At this stage of the moult, some white tips will undoubtedly have been due to wear but some did appear to occur on full-length feathers. About 20% show a small white mirror on P9. From his studies on the Canaries, Volsøe (1951) also found more white on P10, with one almost showing a complete white tip, and a greater frequency of mirrors on P9 than reported for those in the Azores. Further support comes from Vaurie (1965) who noted an increased tendency for white on P9 compared to those in the Azores and from Bannerman (1963) who found a large mirror on P10 and usually no mirror on P9. The black on the wing-tip is usually extensive on the outer primaries with a solid triangle of black extending from the inside of the white mirror or tip to the visible base of P10 on the leading edge and the outer web of P6 on the trailing edge. The black triangle is solid, lacking pale edges to the inner webs as for instance in Caspian Gull. P6 always carries a large black subterminal mark. The markings on P5 are quite variable with 45% showing a broad black band, 30% a thin black band and 25% a broken band, spot or indistinct marking. P4 is often unmarked but sometimes (5% of cases) carries a small black spot on the outer web. Bannerman (1963) found a black band on P5 and a black spot on P4.

Bare parts: in December, these are very bright with the legs being a bright metallic lemon-yellow or bright ochre (earthy yellow - yellow with a fawn tinge) and the bills a bright yellow. During moult in August, the vast majority of adults (90%) have pale yellow bills and dull ochre legs which also look pale yellow in bright light. The other 10% have bright yellow bills and bright yellow-ochre legs which positively gleam in strong light. About 10% of the adults with dull legs have pinky-yellow feet. No adult appears to have flesh legs. The gonydeal spot is bright red in December and is large, often extending just onto the upper mandible. The orbital ring is a bright scarlet red.

Calls: Generally rather taciturn but prolonged observations in August 1998 on the large flock at Campo de Golf produced considerable information. The most frequent long call, heard in about 80% of cases, is close to that of graellsii in its chuckling nature but slightly shriller and somewhat less resonant, rather like a shrill 'tinny' graellsii. However, a minority, about 20%, give calls closer to the shrill clamour of argenteus but perhaps slightly deeper. About half of such calls were given in a rather emphatic weary manner. In Garachico in northern Tenerife, the long calls are closer to argentatus than to graellsii. The generally shriller nature suggests an influence of argenteus so perhaps it is not surprising that long calls are usually delivered in a similar way to argenteus with the neck at a 60° angle, slightly steeper than the typical 45° of argenteus. Mew calls are very similar to graellsii, sharing their low-pitched richness but occasionally slightly shriller. Other calls include the fundamental (single deep note) which is deep and again close to graellsii and the rallying call which is also like graellsii in tempo and delivery but slightly shriller. In December, calls are often heard from the gulls as they wheel around the tops of the rugged hills near Playa de Las Americas. The long, mew and fundamental calls at this time of year are the same as the predominant forms in August. Moreno (1988) confirms the relatively shrill nature of the long calls of atlantis in the Canary Islands by indicating that it emits an asperant gua, gua, gua, less deep than the ga, ga, ga of the Lesser Black-backed Gull. Teyssèdre (1983) provides indirect support. In February 1982 she visited Funchal in Madeira and Las Palmas (Gran Canaria) and Santa Cruz (Tenerife) in the Canaries. Although no formal sonogram measurements were taken of atlantis, she thought that the calls of the Canary Islands birds were shriller than those of the Madeiran ones.

Moult in adults: in mid-August all the adults were in full moult with typical flight pattern being P1-P6 newly grown, P7 growing, P8 missing and P9-P10 still intact. Head moult was assessed on 17th August 1998: 28% showed quite heavy speckling on the front, sides and top of the head and a brown wash over the back of the head giving a hooded effect; 72% showed quite definite brown on the head, in the form of speckles, particularly on the eyes and forehead and none were all white-headed. In each category the necks were white. Heavy head streaking, as reported for Azores atlantis (Garner, 1998; Jonsson, 1998), was not observed on any adults. The speckling observed is very obvious in dull light but is completely 'burnt out' in the bright sunshine of midday. In December the heads are a gleaming white and all primaries are present indicating that moult is complete.

Direct Comparison with adult Lesser-black Backed Gull: in direct comparison in flight against strong winds, graellsii look much more streamlined with long thin wings, small body and head and a less deep amplitude to the wing beat. They also make less strenuous progress, tacking frequently to make progress into the wind. The most obvious flight distinctions are provided, from below, by the more extensive black on the primaries in the Lesser Black-backed Gull, extending obviously to P4/P5, and, from above, by the greater contrast between the outer primaries and the slightly paler back and upperwing in the Yellow-legged Gull. Both species show a pronounced dark trailing edge to the underwing but this feature is more striking in graellsii particularly on the secondaries. On the ground intermedius stand out more obviously from atlantis than graellsii because of their darker mantles. Close examination showed that the following differences held very well in August. The graellsii have orange-yellow legs and bill while those of the atlantis are generally a dull ochre and pale yellow respectively, and the graellsii have white heads while those of the atlantis are speckled. The graellsii are typically shedding P4 at this time (Cramp, 1983) while the atlantis are shedding P8. Structurally the graellsii show a slighter chest, a smaller head, slenderer legs of a similar length, narrower wings with the white tips of P6-P10 more evenly spaced on the folded wing, a less massive head and neck and a slighter, more tapering and more pointed bill. Dwight's measurements (1925, p.209) for Azores atlantis and graellsii show that atlantis is 1-10% larger in every measurement made with atlantis having particularly large toes, bill and tail. Bannerman (1963) noted that atlantis in the Canaries is 'rather larger' than graellsii. In December the reverse moult condition applies with graellsii still in the final stages of moult and the atlantis having completed their moult.

First-years: freshly-fledged juveniles in July and August are a dark grey-brown colour and only a small minority (10%) show a rusty tinge as in some, particularly worn, Mediterranean Yellow-legged Gull. The heads always show a large dark patch around the eye and the majority (85%) also show dense speckling on the crown and more pronounced streaks on the upper neck. Some show speckling on the forehead as well but in others this is strikingly white as in Mediterranean Yellow-legged Gull (Jonsson, 1998). The remaining 15% appear to have all-white heads except for the eye patch though close inspection often reveals indistinct speckling. The breasts are heavily streaked in 85% of birds with under 5% very lightly marked. The bellies have a tendency to be slightly paler, being heavily streaked in 65% of birds and sparsely streaked in the remainder. None have virtually unmarked bellies. The bills are generally (95%) all-black though a few (5%) show a pale base and some a small white tip. In fresh plumage, about a third have a scaley mantle and two-thirds a less contrasting mantle. Jonsson (1998) also noted that first-winter atlantis can be very strongly patterned. The underwing usually (80%) shows some contrast between the grey-brown underwing coverts and the paler rest of the wing and, in 35% of birds, the contrast is very obvious. Jonsson (1998) also noted that the underwing coverts of atlantis were frequently solidly dark. The legs are flesh-coloured, sometimes quite a bright pink.

The primaries are the darkest part of the plumage being an obvious dark brown-black on most individuals. The pale window on the middle primaries is usually neither obvious (5% of cases) nor non-existent (15%). Instead 80% of the juveniles show an indistinct pale window which needs to be looked for carefully. Moreno (1988) emphasised the importance of the white tips (presumably including the pale inner webs) on the inner primaries of atlantis for distinguishing them in the Canary Islands from juvenile graellsii. As indicated above 80% of atlantis juveniles did show at least some paleness in the primary window.

The inner wing always shows a distinct secondary bar but the presence of an additional greater coverts bar is more optional. About 55% of juveniles show two bars giving a striped effect on the inner wing. A much more constant feature (90% of cases) is the gleaming white rump, carrying only a few dark flecks and contrasting with a broad dark solid brown-black subterminal tail band. The band is usually broader than in Mediterranean Yellow-legged Gull, being closer to graellsii in this respect. The tertials are like those in graellsii (Dubois & Yésou, 1984) showing dark grey-brown centres with a thin complete outer pale edge.

The analyses of the samples obtained in visits in 1998 and 2000 are shown separately in Table 6. The results are remarkably similar for July 2000 and August 1998, the main differences being that the birds were more grey-brown in July (90% against 80% in August) and also showed a greater tendency to not show a pale primary window (20% against 15%) and to have an all black bill (100% against 90%). The colony at Los Cristianos on 28th July 2000 had almost finished for the season but residual activity included one juvenile hunger-crying and about ten pairs apparently still in territory protecting juveniles hidden under rocks. Three very weak-flying juveniles were seen in this area in the next ten days. At Garachico on 8th August 2000, one unfledged juvenile was present and about six pairs still appeared to be ongoing. The juveniles have whining calls like argenteus in their first year.

By December the mantle colour has faded slightly but is still predominantly grey-brown and the heads are paler but still retain many dusky marks as well as the obvious eye patch.

Moult in first-summers: by late July of their second calendar year, most have completed primary moult but a few are retarded enough to show P8 growing and P9 missing.

Second-years: in late July and August, when the birds are about one year old, their appearance is very variable. Compared to juveniles the body is generally a much cleaner white with much reduced breast and belly streaking. However, some retain quite extensive streaking on the sides of the breast and the belly. The heads retain a conspicuous eye spot but the head as a whole is usually much whiter. Some show extensive streaking on the lower neck and around the back of the neck. The strikingly pale examples of first-summers/second-winters reported in July for an undisclosed location in the Canary Islands (Dubois 2001) are at the extreme end of the spectrum for paleness and advanced maturity in the central Canaries. They are discussed later (see eastern Canaries). The outer primaries are a conspicuous dark brown-black, showing a much greater contrast with the scapulars and back which are beginning to show the adult grey colour. The inner wing shows a pronounced secondary bar but the greater coverts bar is rarely obvious. The pale window on the inner primaries is obvious and the tail and upper rump are a bright white, virtually devoid of mottling, with a neat dark subterminal tail band. At rest the mantle looks a rather uniform mixture of brown and grey with few obvious spots on the greater coverts. The underwing also shows a tendency to a greater uniformity with the underwing coverts rarely being obvious. The bill varies from still almost completely black to a pale pink-yellow base extending one-quarter of the bill length. The legs are rarely bright pink, often being a darkish straw colour or a creamy colour.

Moult in second-summers: some show dusky heads in August with streaking also extended onto the necks. Others show paler heads with a dark eye spot.

Third-years: the heads were dusky in August with streaking also extending onto the sides and back of the necks. A few third-summers do have quite pale legs in August with a fleshy tinge not seen in adults.

Comment: the total of 800 at Campo de Golf is an exceptional gathering for the island in view of the estimated total population for Tenerife of 412-438 pairs in 1987 (Paterson 1997) but 700 were found in the nearby Médano area in November 1991 (Guttiérez & Dominguez 1993).

Copyright © Nick Rossiter 1999-2001. All rights reserved.