Mantle: slightly paler and less obviously blue-grey but most still show a blue tinge particularly on the scapulars at rest. The shade is obviously paler than graellsii but generally darker than Mediterranean Yellow-legged Gull. Cramp (1983) also considered that the mantle shade of western Iberian forms was slightly darker than that of Mediterranean Yellow-legged Gull. Garner (1998) examined some skins from Berlengas in south-west Portugal and confirmed that the mantle shade was between that of Mediterranean Yellow-legged Gull and atlantis. Joiris (1978) assessed lusitanius as paler than gulls he considered to be argentatus. In view of the variability of argentatus, this comparison is not very critical as pointed out by Garner & Quinn (1997).

Underwing: not such an obviously dark area on the trailing edge; there was a dusky area on the primaries but very little shading on the secondaries.

Size: smaller; slightly larger than graellsii in general but there is some overlap with 5% actually being smaller than adjacent graellsii in direct comparison. Joiris (1978) thought lusitanius was smaller than gulls he assessed as argentatus. The present study suggests that they are indeed much smaller than argentatus, being smaller than atlantis in the central Canaries and this view is supported by the measurements given in Tables 1-2..

Structure: on balance slighter; the wings are generally narrower up to the carpal and appear long as in graellsii although the wingtip remains blunt; most (80%) appear sturdy but about 20% are quite slightly built and more like graellsii with a small head and a less bold chest. The bills of the minority are, however, still quite heavy and this can give a rather unbalanced look to some birds. In Cramp (1983, p.835), a less-pointed wingtip was reported for Atlantic Iberian michahellis with P9 and P10 given as equal longest primaries and P8 17-23mm shorter. The wingtip shape implied by Cramp for Atlantic michahellis is actually quite close to that for argenteus. Joiris (1978) does not describe structural features.

Primaries: tendency for less black and more white; in about 50% of adults, the wing-tip shows an extensive solid triangular black area as in those in the central Canaries but quite frequently (35%) the black area is reduced with extra grey on P7 intruding into the black area to make it slightly V-shaped. In a small minority (15%), usually involving the slighter birds, the black is heavily indented giving a wing-tip more like argenteus. 50% of birds showed a mirror, sometimes very small, on P9 and P10 is always well-marked with a white mirror or tip. 30% have an indistinct, or even no, marking on P5. Cramp (1983) also thought that the west Iberian forms had slightly less black in the primaries than Mediterranean Yellow-legged Gull.

Bare parts: legs are mainly a rather clean bright lemon-yellow colour with no hint of orange but a few show a richer ochre colour.

Calls: as in the Central Canaries. Bermejo (1999) agrees, saying that the voices of all the birds found in the Iberian Atlantic are sharper than those of the Mediterranean Yellow-legged Gull,

First-years: often rustier-brown with 15% ruddy brown and 60% grey-brown with a ruddy tinge. Joiris (1978) described the first-years in January as being finely-barred on the mantle and with a pale head with strongly patterned dark markings in the loral area. The present study found these features in 45% and 75% of birds respectively.

Moult in first-summers: in late March, no first-summer seen had started primary moult; the situation was quite variable in first-summers in mid-May with 40% still to start moult, 30% only moulting P1 and 10% at the most advanced state of having shed P4.

Second-years: in March, as in the eastern Canaries.

Information needed: moult in adults.