On the morning of Friday 30th March 2001, over 40 dedicated gull-watchers, together as participants of the 5th International Gull Meeting (IGM) at Foz de Douro, Porto, Douro, Portugal, set out for the fishing port of Matosinhos. This was going to be the first field trip of the 5th IGM.
Upon arrival in the fishing port, at c. 8:30 h., many large gulls were observed. Most of these were 'Yellow-legged Gulls' Larus michahellis (?) of the form (forms?) commonly found around the Atlantic Iberian coasts, and of unresolved taxonomic affinity and unsettled nomenclature (hereafter: ‘Galician Gull’). Smaller numbers of Lesser Black-backed Gull L. fuscus of the subspecies graellsii and intermedius were also present. It could be ascertained that at least these three taxa were involved, because individuals colour-ringed as chicks of each of the three were found, with the respective ringers participating in the observations and confirming the presence of 'their' birds (ABe for Galician Gull, PR for graellsii and NHL for intermedius).

At c. 8:45 h., an odd large gull was found perched on a rooftop. Unfortunately, it soon flew off behind the roof, remaining out of sight for the next 30 min. Around 9:15 h the odd gull was relocated at the same rooftop and identified as an American Herring Gull L. smithsonianus in second-winter plumage. During the next 3 hours most gull-watchers got good views of the bird. Photographs and video footage were taken. After 13:00 h, when most gull-watchers had left, it (hereafter: the Friday bird) could not be relocated.

After the Friday bird had not been relocated on the second field trip to the Matosinhos port on the morning of Saturday 31th March, almost all participants of the 5th IGM set out in the early afternoon for the third field trip. Its destination was the Barra do Douro (Cabedelo), an estuarine dune formation stretching c. 700 m northward from the south side of the mouth of the Douro river, and the surrounding mudflats and sandbars. This trip started around 13:30 h and coincided with low tide.
Several 100s Galician Gulls and smaller numbers of Lesser Black-backed Gulls, Mediterranean Gulls L. melanocephalus, Black-headed Gulls Chroicocephalus ridibundus and Little Gulls Hydrocoloeus4 minutus were present. Amidst these, one Common Gull L. canus, and three or four Ring-billed Gulls L. delawarensis were found, one adult, and two, possibly three, first-years.

Around 15:00 h, two observers found a first-year smithsonianus (hereafter: the Saturday bird) on the westernmost accessible sandbank of the estuary, exposed only at low tide, where many Galician and Lesser Black-backed Gulls had congregated. Due to the rapidly incoming tide, 10 hectic min. were spent on the sandbank studying the bird's field marks and relocating it after several short flights, and obtaining video footage. Although the trip was scheduled to be finished at 15:00 h, those participants who had stayed on the safer ground of the northern tip of the Cabedelo eventually also observed the bird, but only distant views could be obtained here.

In the afternoon of Sunday 1st April, after the 5th IGM had been finished, some participants went again to the Douro estuary, in an attempt to obtain better views of the Saturday bird. Around 14:30 h, a first-year smithsonianus was found, and it was assumed that the Saturday bird had been refound less than 1 km away from its original site. It was noticed that the bird showed a number of second generation scapulars, which was at odds with the Saturday bird, but no conclusions were drawn from that dissimilarity. Soon the bird took flight, and was refound on a small riverine beach near the village of Afurada. Here the bird was feeding amongst Galician and Lesser Black-backed Gulls, allowing views down to c. 20 m. Thereafter the bird flew to a fishing hamlet nearby. It was extremely cooperative, readily taking bread thrown out, and at times it was as close as 4 m. Much video footage and 100s of photographs were taken.
It now proved - among others - to have quite a few second generation scapulars, and to be quite worn. The topic of the bird (hereafter: the Sunday bird) being the same individual as the Saturday bird or a different one was not discussed then, nor later that day. It was only later that week, when pictures of both birds were available, that the possibility of the Saturday and Sunday birds being different individuals was considered seriously. Comparison by Martin Elliott of the footage of the two birds confirmed the suspicions that they were actually different individuals: the Saturday bird missing two blocks of median underwing-coverts that were present on the Sunday bird, among others.

Descriptions of the three birds are at the disposal of the CPR. These were derived from field notes, photographs and video footage, including KM's verbal description of the Saturday bird combined with careful interpretation of the video footage for all participants on the Saturday evening in the IGM venue.

Identification

The Saturday and Sunday birds

Identification of first-winter smithsonianus in Europe has been treated by Grant (1986), Mullarney (1990), Moore (1994), Dubois (1997) and Diggin (2001). Their findings serve as a basis for the identification of the Saturday and Sunday birds. Also, Bob Barber's picture collection on the Internet was consulted (http://www.dreamwater.net/bbarber/birds/gulls/HG/1StBasicHG.html).
Although many first-winter smithsonianus have a character of their own, identification is severely hampered by individual and regional variation in both smithsonianus and argentatus/argenteus. In addition, no single character is truly diagnostic in itself, and there are certain areas of overlap between smithsonianus and the two European taxa, rendering some out of range birds impossible to identify. Familiarity with the extensive variation in argenteus and argentatus (see e.g. Grant 1986, Golley 1993) is a prerequisite when attempting to identify a vagrant smithsonianus.

Vagrant first-winter smithsonianus in Europe should be identified on the basis of as large a combination as possible of the following ‘classic’ characters (comparative remarks only with respect to argenteus):

- virtually wholly dark tail
blackish-brown extends much farther towards the tail-base than in the average argenteus, which has a largely pale base rather than paler corners basally; in argenteus, the brown clearly shows as a subterminal band, and the pale fringe is wider than in smithsonianus; the odd argenteus may show a similar pattern
- densely barred uppertail-coverts and rump
in argenteus, dark barring is usually less dense, making the tail-band stand out clearly, rather than reinforcing the impression of a wholly dark tail as in smithsonianus
- uniformly clouded brown underparts, continuing as a shawl onto the hindneck
argenteus can be rather dark underneath, but usually the clouded character typical of classic smithsonianus is lacking, and the pattern is more blotchy or streaked; a shawl is rarely as distinct as in smithsonianus
- heavily barred undertail-coverts
argenteus usually lacks pronounced barring, and often there is no distinct pattern; if barring is present, the brown bars are narrower than the white areas in between
- uniformly dark brown underwing-coverts and axillaries
in argenteus these feathers are paler and mottled rather than uniformly coloured
- uniformly dark brown bases to the greater coverts, most extensively on the outers, and often lacking in the inners
this pattern is rarely seen in argenteus, but the odd individual shows it
- plain dark brown tertials with narrow pale fringes and minor indentations restricted to near the tip
in argenteus, the tertials usually have wider pale edges and tips, and indentations on much of the visible part of the feathers
- inner primary window bleeding onto central primaries (venetian blind effect), and bordered by pronounced broken terminal bar
in argenteus, the window is generally more clearly defined to the inner primaries, with a less pronounced terminal bar
- angular head and parallel-sided bill, with a medium strong gonys-angle, but smallish-headed and shortish-legged compared to body size
this bill shape is consistent, but it is good to realise that overall size and shape are variable, ranging from as large and robust as the largest argentatus to overlap with Thayer's Gull L. thayeri or California Gull L. californicus.

It goes without saying that deciding whether the underparts are dark enough, whether the undertail-coverts show sufficient strong barring, etc., are rather subjective matters. Therefore, photographs and/or video footage will be required in most cases for record assessment by rarities committees.

The differences between first-winter smithsonianus and the European large white-headed gull taxa other than argenteus, excluding the white-winged taxa, are more pronounced:

- argentatus is similarly-sized and dark individuals can be problematic; usually they have more checkered barring on the wing-coverts, with more white showing in between the brown, the tertials are more extensively notched white and with broader white tips, and the tail is more broadly fringed pale with sparser barring on the upper- and undertail-coverts

S graellsii and intermedius and Baltic Lesser Black-backed Gull L. f. fuscus (hereafter fuscus) tend to be smaller and comparatively narrower- and longer-winged; odd individuals with largely dark tails are a pitfall, but the underparts tend to be coarsely streaked and mottled dark brown, rather than uniformly clouded; usually they lack an obvious pale mid-wing window and dense barring on the upper- and undertail-coverts, and develop less pale on the bill-base

S cachinnans and Mediterranean michahellis, although of similar size as smithsonianus, can be easily eliminated by their largely whitish heads, underparts, and upper- and undertail-coverts; both retain fully black bills throughout their first year of life; the bill of Mediterranean michahellis is comparatively heavier, with a more pronounced gonydeal angle and a blunter tip; apart from their smaller size, Galician Gull and atlantis share most of the Mediterranean michahellis plumage characters, while Galician Gull has a rounder head, a less heavy and more gently downcurved bill and comparatively shorter legs, much like argenteus

S Great Black-backed Gull L. marinus is larger and stronger-billed; it shows more white in between the brown of the upperparts and wing-coverts, is coarsely blotched brown on the underparts, and has sparse, irregular blotching on the tail-base and the uppertail-coverts.

The Sunday bird showed a complete set of the classic first-winter smithsonianus characters mentioned here, making its identification rather straightforward. The bird showed a large body with a comparatively small head, appeared more thickset but of similar size as the surrounding Galician Gulls, with rather shortish-looking legs. Contrary to the Saturday bird, it had moulted many scapulars, and looked quite worn.
Due to the circumstances, the Saturday bird could not be studied in the field and documented as thoroughly as the Sunday bird. Nevertheless, this was more than sufficient for assessing its identity as smithsonianus, since the bird fitted all the classic first-winter smithsonianus characters. It showed hardly any noticeable wear, with perhaps only one juvenile scapular replaced by a second-generation (first-winter) feather. It was not particularly dark, being considerably paler than first-year Lesser Black-backed Gulls present. Particularly the juvenile tertials looked remarkably fresh.

As for first-years, another species that might cause confusion, is Kelp Gull L. dominicanus. This species is normally restricted to the southern hemisphere. Since the early 1980s, there are several records of Kelp Gull from Africa north of the equator, as far north as Mauritania (Pineau et al. 2001). Others may be found still farther north, and the Iberian peninsula could be a good candidate for the first European records. How to distinguish between first-year smithsonianus and Kelp Gull is beyond the scope of this paper, and we refer to Jiguet et al. (2001) for identification of Kelp Gull.

The Friday bird

Identification of second-winter smithsonianus in Europe is far from straightforward. Even more caution needs to be exercised in differentiating them from same-age Herring Gulls, than for first-winters.
Most cautionary remarks made for identification of first-winters, also hold for second-winters.
It is particularly important to realize that quite a few second-winter argenteus show an all-dark or mostly dark tail (darker than in first-winter plumage) and heavily patterned head, neck, and/or underparts. Ahmad & Elliott (2000) highlighted the subtleties involved in a range of features that tell a 'black-tailed' and otherwise odd second-winter argenteus apart from a classic second-winter smithsonianus.
Most second-winter smithsonianus will not be as eye-catching amidst same-age argenteus/argentatus as first-winters. Therefore, the identification process can not do without a similarly long list of ‘classic’ characters, of which as large a combination as possible should fit the relevant birds (comparative remarks only with respect to argenteus):

- heavily and densely barred undertail-coverts
the dark bars are often broader than the pale bars in between, particularly on the longer feathers; unmoulted uppertail-coverts show a similar pattern; this dense barring on the tail-coverts is rarely, if ever, shown in second-winter argenteus, and is near-diagnostic
- a complete row of uniformly and contrastingly dark greater coverts, also clearly visible in flight (nearly as dark as the secondaries)
second-winter argenteus usually shows thin and intricate barring on the greater coverts ('spaghetti-pattern'); quite a few show very dark outer greater coverts, but a complete row of entirely dark greater coverts seems rare
- a completely blackish tail or nearly so, except for some pale on the inner webs of the outer feathers
some second-winter argenteus are identical in this respect
- uniformly dark lower breast and belly, which may continue onto the upper breast and hindneck, creating a uniformly dark shawl
head and upper neck may be heavily streaked, but can be whitish; some second-winter argenteus may have an identical pattern on the underparts though
- solidly dark tertials (on which paler fringes may be due to wear)
many second-winter argenteus show some delicate barring here, but sometimes the tertials are all-dark, like in smithsonianus
- lack of regular barring on the median and lesser coverts
these may be as uniformly dark as the greater coverts; on argenteus usually paler and checkered
- very dark underwing-coverts, contrasting with silvery-grey remiges
in argenteus, the underwing-coverts are less dark and often mottled, contrasting less with the primaries
- head and bill shape, and leg length, see Saturday and Sunday birds
- somewhat subdued inner primary window, bordered by a row of broad blackish subterminal patches
the basic colour of the window is often quite brownish, not as pale greyish as in argenteus, adding to the uniform appearance of the upperwing
- the pattern of the upperparts (though not a really distinctive character in second-years, this may be helpful too)
in both taxa, the upperparts range from entirely grey to entirely brownish with dark anchor markings; in the former case, the upperparts are distinctly pale grey in smithsonianus and contrast markedly with the uniformly dark underparts (particularly when they continue into a dark shawl on the hindneck) and wing-coverts; in the latter case, smithsonianus shows a rather irregular and mottled pattern on the mantle and scapulars, unlike the neatly barred upperparts of many argenteus.

As in first-winters, the differences with other European large gull taxa are more pronounced:

S argentatus upperparts, if of the grey type, are decidedly darker than in smithsonianus, and if of the brownish type, are barred rather than mottled; there is distinctive pale fringing on the primary tips, and the tail is broadly fringed pale

S graellsii, intermedius, fuscus and marinus have much darker upperparts and considerably whiter underparts and tail with often a comparatively narrow subterminal black band and largely white uppertail-coverts

S cachinnans, Mediterranean michahellis, Galician Gull and atlantis have darker upperparts, and virtually white head and underparts.

The Friday bird exhibited a solid suite of the 10 characters listed above, that undisputedly made it a classic smithsonianus in second-winter plumage. This combination of characters does not occur in second-years of any of the European large white-headed gull taxa. As explained by Ahmad & Elliott (2000), second-winter argenteus from the dark end of the scale as for tail pattern can be troublesome, but generally would lack the extent of uniformity in darkish underparts colouration, darkness of tertials and greater coverts, and the density in upper- and undertail-covert barring shown by classic second-winter smithsonianus.
Individual smithsonianus may show additional identification features, which are useful when scanning gull flocks in Europe. In the case of the Friday bird, the odd combination of an argentatus-like size and shape (though smithsonianus need not be this big), and an argenteus-like mantle colour, contrasting strongly with a strikingly dark inner wing, supplied a good supplementary field character.

Discussion

There is only one record of smithsonianus for Portugal (Moore 1994, de Juana A. et al. 1995, CM), of a first-year observed at Faro, Algarve, 30-31 December 1992. Spain also has one record, of a first-year at Xixón, Asturies, on 3 January 1991 (de Juana A. et al. 1995).
The first record for the Western Palearctic was of a bird ringed as chick on Kent Island, New Brunswick, Canada, on 13 August 1936, that came aboard a ship at sea 480 km NW of Cape Finisterre, Galicia, in November 1937 (Gross 1940, Cramp & Simmons 1983).
Most records of smithsonianus in the Western Palearctic come from Ireland. After the first in 1986 (O'Sullivan & Smiddy 1990), no fewer than 10 were recorded in 1990 (O'Sullivan & Smiddy 1991), all first-winters. The identification of three of these was treated in detail by Mullarney (1990). Up to 1999 the total number of accepted records for Britain and Ireland is 27. Five of these are British, all first-winters, and of the Irish records only two were second-winters, with no older birds recorded thus far (O'Sullivan & Smiddy 1990-1992, Rogers et al. 1992-1994, 1996-2000, Smiddy & O'Sullivan 1993). The only remaining Western Palearctic country that has smithsonianus on the list is France, with four individuals recorded from 1993 through 1999 (Dubois et al. 1995a, 1995b, Frémont et al. 2000). Surprisingly, only one is a first-winter, in January in suburban Paris, the others are a second-winter in January and February on the SW coast, and also recorded as a third-winter in December of the same year, a second-summer in June at sea off W Brittany, and a juvenile on 3 August 1997 at Le Portel, Pas-de-Calais. How many did we miss there during the 4th IGM in October 1999?
With three smithsonianus in Porto in 3 days, an no previous records in the area, it is tempting to speculate that they have been overlooked. It is to be hoped that local observers will be successful in finding them in the future.

Acknowledgements

The authors, participants of the 5th IGM who either found or identified the three smithsonianus, or contributed to documenting the records through their field notes, sketches, photographs or video footage, wish to express their sincere gratitude to all other participants for their stimulating company in the field and in the IGM venue at Foz de Douro: Kenneth Bengtsson, Cecilia Johansson, Risto Juvaste, Vegard Ankarstand Larsen, Radka Lezalova, Pedro Cardia Lopez, Mike Marsh, Kjeld Pedersen, Johnny Pedersen, Steve Piotrowski, Jaroslav Simek, Ian Smith, Roy Smith, David Sowter, David Sturm, Alberto Velando, Clive Watson and José Maria Zapata.
We are particularly grateful to Martin Elliott for scrutinizing the footage of the two first-year smithsonianus and finding irrefutable differences between them.
Steve Howell, Bruce Mactavish and Kevin McLaughlin commented on various smithsonianus identification topics. Peter van Scheepen provided considerable help in taking care of large chunks of the mountain of e-mail communication.

References

Ahmad, M. & Elliot [sic], M. (2000). Red Herrings? Birdwatch 99: 22-24.

Beaubrun, P.-C. (1993). Status of Yellow-legged Gull (Larus cachinnans) in Morocco and in the western Mediterranean. In: Aguilar, J.S., Monbailliu, X. & Paterson, A.M. (Eds.), Estatus y conservación de aves marinas. Actas del II Simposio MEDMARAVIS, Madrid, pp. 47-55.

Carrera, E., Trias, J., Bermejo, A., de Juana, E. & Varela, J. (1987). Étude biométrique des populations ibériques et nord-africaine du Goéland leucophée Larus cachinnans. Oiseau Rev. Fr. Ornithol. 57: 32-38.

Cramp, S. & Simmons, K.E.L. (Eds.) (1983). The Birds of the Western Palearctic, Vol. 3. Oxford University Press, Oxford.

Crochet, P.-A., Bonhomme, F. & Lebreton, J.-D. (2000). Molecular phylogeny and plumage evolution in gulls (Larini). J. Evol. Biol. 13: 47-57.

Crochet, P.-A., Lebreton, J.-D. & Bonhomme, F. (2002). Systematics of large white-headed gulls: patterns of mitochondrial DNA variation in western European taxa. Auk 119: 603-620.

Diggin, J. (2001). The American Herring Gulls in County Cork. Birding World 14: 62-65.

Dubois, P.J. (1987). Notes on the 'Cantabrican' Herring Gull. In: Grant, P.J., Sharrock, J.T.R., Taggar, S. & Shirihai, H. (Editors), International Bird Identification, Proc. 4th Int. Identification Meeting, Eilat, 1st-8th November 1986, Eilat, pp. 41-42.

Dubois, P.J. (1997). Identification of North American Herring Gull. Br. Birds 90: 314-324.

Dubois, P.J. & CHN. (1995a-b). Les oiseaux rares en France en 1993-1994. Rapport du Comité d'Homologation National. Ornithos 2: 1-19, 145-167.

Frémont, J.-Y. & CHN. (2000). Les oiseaux rares en France en 1999. Rapport du Comité d'Homologation National. Ornithos 7: 146-173.

Golley, M. (1993). Identification of argentatus Herring Gull. Birding World 6: 32-38.

Grant, P.J. (1986). Gulls: a guide to identification. Second edition. Buteo Books, Vermillion.

Gross, A.O. (1940). The migration of Kent Island Herring Gulls. Bird-Banding 11: 129-155.

Jiguet, F., Jaramillo, A. & Sinclair, I. (2001). Identification of Kelp Gull. Birding World 14: 112-125.

Joiris, C. (1978). Le Goéland argenté portugais (Larus argentatus lusitanius), nouvelle forme de Goéland argenté à pattes jaunes. Aves 15: 17-18.

de Juana A., E. & Comité Ibérico de Rarezas de la SEO. (1995). Observaciones homologadas de aves raras en España y Portugal. Informe de 1993. Ardeola 42: 97-113.

Klein, R. & Buchheim, A. (1997). Die westliche Schwarzmeerküste als Kontaktgebiet zweier Großmöwenformen der Larus cachinnans-Gruppe. Vogelwelt 118: 61-70.

Liebers, D. & Helbig, A.J. (1999). Phänotypische Charakterisierung und systematische Stellung der Armenienmöwe Larus armenicus. Limicola 13: 281-321.

Mayr, E. & Ashlock, P.D. (1991). Principles of systematic zoology. Second edition. McGraw-Hill, New York.

Moore, C.C. (1994). The first record of American Herring Gull for Portugal. Airo 5: 32-34.

Mullarney, K. (1990). American Herring Gulls in Ireland. Birding World 3: 96-100.

Munilla, I. (1997). Desplazamientos de la Gaviota Patiamarilla Larus cachinnans en poblaciones del norte de la peninsula Ibérica. Ardeola 44: 19-26.

O'Sullivan, O. & Smiddy, P. (1990-1992). Thirty-seventh to thirty-ninth Irish Bird Report, 1989 to 1991. Irish Birds 4: 231-257, 423-462, 571-610.

Pineau, O., Kayser, Y., Sall, M., Gueye, A. & Hafner, H. (2001). The Kelp Gull at Banc d'Arguin - a new Western Palearctic bird. Birding World 14: 110-111.

Rogers, M.J. & Rarities Committee. (1992-1994, 1996-2000). Report on rare birds in Great Britain in 1991-1993, 1995-1999. Br. Birds 85: 507-554, 86: 447-540, 87: 503-571, 89: 481-531, 90: 453-522, 91: 455-517, 92: 554-609, 93: 512-567.

Rus, A. (2000). Gaviota Patiamarilla: una población controlada. Tierra 28: 36-41, 44-47.

Sangster, G., Hazevoet, C.J., van den Berg, A.B., Roselaar, C.S. & Sluys, R. (1999). Dutch avifaunal list: species concepts, taxonomic instability, and taxonomic changes in 1977-1998. Ardea 87: 139-165.

Smiddy, P. & O'Sullivan, O. (1993). Fortieth Irish Bird Report. Irish Birds 5: 79-102.

Teyssèdre, A. (1984). Comparaison acoustique de Larus argentatus argenteus, L. fuscus graellsii, L. cachinnans (?) michahellis et du Goéland argenté à pattes jaunes cantabrique. Behaviour 88: 13-33.
------------------------------------------------------------------------------------------------------