Iceland Gulls from the c Braer? disaster^ Shetland 993 D. N. Weir, R. Y. McGowan, A. C. Kitchener, S. McOrist, B. Zonfrillo and M. Heubeck Iceland Gulls Lams glaucoides are uncommon winter visitors to Britain and Ireland, most regularly to Shetland (Lack 986), The conventional view is that occurrences are of Iceland Gulls of the race glaucoides from Greenland (Cramp & Simmons 983). Recently, a few individuals of the Canadian race kumlieni have been recorded, including two specimens in 993, and some sightings have popularly been attributed to 'Thayer's Gull* L. g. thayeri (Millington 993). Two 'invasions' in the northwestern Palearctic were studied, from 4 specimens shot in southwest Iceland in 966 and from 38 in the Faroe Islands in 983. In both cases, the commonest birds were kumlieni adults with very pale wingtips, more or less indistinguishable from glaucoides in the field (Ingolfsson 96; Fjeldsa & Jensen 98; J.K. Jensen in litt.; A. Petersen in litt). An invasion of Iceland Gulls in Shetland in January 993 coincided with the wreck of the oiltanker Braer. About,600 dead seabirds were found in or near Shetland during the following five weeks, including 6 Iceland Gulls, which represents the thirdlargest discrete Palearctic sample. We examined and give full details of of these, and make comparisons with other gull casualties of the Braer. We include details of another dead kumlieni from Shetland in 994. Further studies of Iceland Gulls will be reported elsewhere (Weir el at in prep.). Methods There were 6 Iceland Gulls among the frozen carcasses of some,400 seabirds sent to the Royal (Dick) School of Veterinary Science, Edinburgh University, for postmortem examination. One Iceland Gull from Shetland was discarded, but, including two sent by the bird observatories on Fair Isle, Shetland, and North Ronaldsay, Orkney, the total sample examined was, here given as 'from [Brit. Birds 88: 2, January 99]
6 Iceland Gulls from the 'Braer' disaster Shetland'. We recorded percentage oiling of plumage, external measurements, age, sex, moult, weight, subcutaneous and bodycavity fat scores, profiles of pectoral muscles, gizzard and crop contents and gross pathology. Iceland Gulls were prepared as study skins, or, if too decayed, as skeletons and complete (i.e. feathered) left wings. The gulls were initially aged in winters (st to 3rd and 4th or older), following Cramp & Simmons (983). Firstwinter and secondwinter plumages, however, are variable and sometimes hard to differentiate (Dwight 92). There are no published descriptions from skins of knownage birds ringed as nestlings. We attempted to age the gulls based on the degree of skull ossification from radiographs, so that we could compare these with plumages of skins, but this was unsuccessful. We therefore used two criteria based on feather shapes and measurements to distinguish firstwinter from secondwinter individuals. First, the more rounded tips of tail feathers distinguish firstwinter gulls from older individuals (Dwight 92). Secondly, the wing formulae of suspected secondwinter Iceland Gulls differed consistently from those of all other ages, giving relatively longer, narrower wings (see below). We measured flat, straightened wings and culmens from the base of the bill sheath. These measurements were compared mainly with those of other Iceland Gull specimens from the Palearctic, and either were made by us from three British collections or are from Faroese and Icelandic material measured in a similar way (J.K. Jensen in liu.; A. Petersen in liti.). Iceland Gull wings dry and shrink after specimen preparation (e.g. about 3% in seven months for gulls in the Shetland material), which caused problems when comparing measurements made from specimens ranging from fresh to more than 00 years old. We did not use the main published sets of Iceland Gull measurements. From the texts, or from reexamination of the material, we found that some individuals included under nominate glaucoides would be assigned by us to kumlieni (Dwight 906, 92; Rand 942; Cramp & Simmons 983). In one published set, the data were insufficient for statistical comparisons (Godfrey 986), and in another they had to be treated as unreliable (Smith 966; Snell 989, 99). Adults (fourthwinter or older) in Shetland and comparative material were assigned to one of seven categories of wingtip markings, following Fjeldsa & Jensen (98): see fig.. These span the kumlieniglancoides cline, with Types 6 being treated as kumlieni and Type as nominate glaucoides (Fjeldsa & Jensen 98). By correspondence, we and Icelandic workers ensured that our assignments were as close as possible to those of Fjeldsa & Jensen. Weights of Iceland Gulls in the Shetland sample were approximate, owing partly to lack of time to clean them. Most were damp and sandy, and some had thin films of oil on the plumage, making them heavier, and all were probably dehydrated, which would make them unusually light. The relative importance of these opposite biases could not be determined. Assessments of condition relied more on gross pathology, scores (03) for subcutaneous and bodycavity fat deposits and the transverse concavityconvexity of the pectoral muscles. The crop, proventriculus and gizzard were opened to record food and the presence or absence of oil; presence of oil was usually obvious. In two doubtful cases, gizzard contents were analysed by gas chromatography, but, because of the chemistry of the Gullfaks oil from the Braer, tests were inconclusive.
British Birds, vol. 88, no., January 99 Fig.. Colour patterns of the wingtips of adult Iceland Gulls Lams glaucoides. Taken, by permission of the editors of Dansk Om. Foren. Tidsskr., from Fjeldsa & Jensen (98), where the origins are listed for the 22 specimens on which these drawings are based. Results Details of the individuals are given in table. Plumage patterns and morphological criteria were in agreement for presumed firstwinter and secondwinter individuals, giving confidence in our ageing techniques. More than half the birds (3%) were adult and only one (%) was firstwinter. None was in moult. Eleven could be sexed from their gonads, indicating four males and seven females. Measurements suggested that the whole sample contained five males and ten females. Table. Details of dead Iceland Gulls Lams glaucoidesfrom Shetland, Fair Isle and North Ronaldsay, Orkney, JanuaryMarch 993. Sex and age (in winters) 9 >4W? >4W 9 >4W? >4W c?>4w 9>4W 9>4W cj>4w 9 3W 9 2W? 2W cj2w 9 2W cj?w 9 W Subspecies (G = glaucoides, K = kumliem) G G G? G K K?? G Wing 40 40 398 39 40 403 396 440 403 400 42 394 420 4 400 LENGTHS IN MM Tarsus.4 0. 4.6.0 60.0.4.0 64..4.6 2.8 Culmen 44.2 44.2 38.9 40.0 4.8 4.4 44.2 0.0 43.8 4.6 49. 46. 4.9 4.0 42. Gonys 3.2 4..2 4.8.3 4..0.0 4.8 3.6..2 4..0 3. Measurements of other specimens are given in table 2, of which 64 (8%) are from the Palearctic. Measurements in table are within these ranges. There is considerable overlap between the measurements of the adults of nominate glaucoides and kumlieni (tables & 2). Earlier workers suggested that kumlieni have slightly larger bills than glaucoides (Dwight 906, 92; Rand 942). Some of their 'glaucoides" were, on our criteria, kumlieni. On summer specimens of partly sympatric kumlieni and thayeri, bill size and wingtip melanism were not correlated (Snell 989). Winter glaucoides and kumlieni of unknown origin cannot
i? I Table 2. Measurements of mainly Palearctic specimens of Iceland Gulls Lams glaucoides from five European collections, for comparison with 993 Shetland sample. Birds from Baffin Island, Greenland, Icelandj Faroes and Britain & Ireland. * = selected large individual; ** = Newfoundland Measurements in mm: Where taken mean, standard error, range (no.) Race (n) glaucoides () glaucoides kumlieni (6! kumlieni (! () () () (4) Age Ad Ad Ad Ad 3Y 2Y 2Y Y Y Sex S <J 9 9 <J 9 <J 9 BAF * * GRN 4 ICE 4 4 FAR 6 3 4 2 B&l 2 2 * 48.8 40043 403.2 38420 49.4 398439 399.9 3844 WING 408.9 39423 422. 40432 400. 3884 440 406.8 394 ±6.9 () ±3.02 (3) ±3.06 (6) ±2.83 () () () (4) 9.0 66.3 060.6 8.4 4.64 6.9 06 TARSUS. 3.6 60.0.6.6 8 66.6.8 460 ±. () ±0.84 ±.0 (6) ±0.3 (6) () (6) (4) 4.3 40.92. 44.0 39,4.6 4. 432 43.9 394 CULMEN i 44. 424 48. 46.949 43.9 4.648 4.3 43.6 40.4. ±2.04 () ±0.83 (3) ±0. (6) ±0.6 (6) () () (4) 6,.6.8 6.0.0. 6. 48. 4. GONYS 4.8 4.2.4 4,2 4.04.4 8.4 4.9 3.6.3 (2) ±0.33 (8) ±0.4 () () (2) (3) (2) 00 I I!
British Birds, vol. 88, no., January 99 9 Plate. Wingtip markings of races of Iceland Gull Lams glaucoides. Top to bottom: L. g. kumlieni adult, January 993, Orkney (NMSZ 994.60) L g. glaucoides adult, February 993, Shetland (NMSZ 993..3) L. g. kumlieni fourthwinter, January 993, Shetland (NMSZ 993..4) Lg. thayeri adult, July 98, Prince of Wales Is., Canada (Aberdeen Univ. NH Dept 90..22) {Trustees of the National Museums of Scotland)
20 Iceland Gulls from the 'Braer' disaster be separated by measurements. Adult wingtip melanism seems to be the only reliable criterion, and distinctions even in the hand are often difficult. Measurements made by Ingolfsson (96) of another 28 adult male and 26 adult female glaucoides from Iceland are within the ranges in table 2, but wings of males averaged longer, at 430 mm. These birds were measured when fresh, whereas those of male glaucoides in table 2 are partly old specimens, which could account for most of the difference. Partial wing formulae of 30 Iceland Gulls are given by age in table 3, including both Shetland and comparative material. Gulls classed as secondwinter from plumage have wings of a consistently higher aspect ratio than those of all other age classes (i.e. their wings are relatively longer and/or narrower). Table 3. Partial wing formulae of Iceland Gulls Larus glaucoides, from birds readily aged from plumage and with least worn or damaged primaries (National Museums of Scotland and Natural History Museum collections). Primaries from P0 (as 0), others as mm shorter than P0 (means, standard errors and ranges) Age class (no.) Firstyear (4) Secondyear () Thirdyear (3) Fourthyear (3) Older adult (9) 0 0 0 0 0 0 9 3.00 ±0.9 6.36 ±0.8 0 4.33 ±0.6 3 6.00 ±.00 4 2.6 ±0.3 6 8 20.00 ±.8 623 2.82 ±0.82 229 2.33 ±0.6 2022 2.6 ±0.33 222 9.33 ±.33 32 44. ±.66 448 0.9 ±0.9 46 46.00 ±.00 444 46.6 ±.34 4448 4.6 ±.04 422 6 69. ±2.43 64 6. ±.3 688 0.6 ±0.88 692.33+0.88 03 69.6 ±.0 66 99. ±0.9 90 0.3 ±. 962 99.33 ±.8 903 0.00 2. 90 96.33 ±.2 9202 From wingtip markings, four of eight Shetland adults are Type (Greenland glaucoides). Two more are pale, but have damaged tips to primaries 09, which are critical to identification, but they were probably Type or Type 6 (kumlieni). Type 6 breeds mainly in east Baffin Island and possibly in Greenland (Smith 966; Ingolfsson 96; Snell 989; Weir el al. in prep.). The fourthwinter female kumlieni is closest to Type and resembles the type and para types of this race (Dwight 906, plate xv). It is the most widespread form in Baffin Island (Smith 966; Gaston & Decker 98; Snell 989). The adult male kumlieni is large (tables & 2); its wingtip markings are Type 3 in extent although very pale. Remains of an adult kumlieni from Shetland in 994, not in table, are only the left wing; it fits Type 4. Winter weights of Iceland Gulls from the Palearctic are given (table 4). Too many were approximate, or from very small samples, for standard errors to be included usefully. Faroese weights were lowest in the 983 invasion, but Icelandic weights were consistent between an' invasion and all other years. Five of six Icelandic birds examined for fat were classed 'Medium' or 'Fat' (A. Petersen in litt.; Univ. Aberdeen Nat. Hist. Mus. specimen labels). For these reasons, we regarded Icelandic winter weights as 'normal'. (See also Discussion.) Shetland gulls averaged about % of the weight of Icelandic gulls and were even lighter than those from the Faroes in 983, which were described as 'very meagre' (Fjeldsa & Jensen 98).
British Birds, vol. 88, no. I, January 99 2 Table 4. Winter weights of Iceland Gulls Ixinis glaucoides in invasions and other years from Iceland, the Faroes and Shetland. Data: Iceland (a) all adult glaucoides; source all Ingolfsson (96); Iceland (b) 6 of adult; sources (A. Petersen in tor.), 2 (BWP 983); Faroes 29 of 44 adult; sources 24 (Fjeldsi & Jensen 98), 20 (J.K. Jensen, unpubl.); Shetland of adult; sources (this study), c = circa (approximately) Country Iceland 966 (a) all years (b) Faroes invasion other years all years Shetland 993 No. of 99 26 28 4 32 6 Weight (g) MEAN c0 69 2 66 c600 RANGE 640039 480820 2080 48080 4900 No. of 36 28 0 2 Weight (g) MEAN c90 90 84 80 90 c60 RANGE 8090 00890 00890 60000 Detailed internal examination of the Shetland Iceland Gulls* indicated the likelihood of their deaths even if there had been no oil; proximate and ultimate causes of death could not be separated. Other aspects of their condition oiling, amounts of food in the stomachs and numbers of casualties are compared with those of three other gull species in the Shedand material (table ). Most comparisons were only suggestive, but Iceland Gulls tended to have less fat than others, more often had wasted pectorals, less often contained food, less often had heavily oiled plumage, but more often had ingested oil. The relative abundance of the four species in Shetland in winter is indicated broadly by Shetland Bird Reports (Anon. 96992). Although Iceland Gulls were most numerous in 993, and dead birds of a rare species clearly would attract attention, casualties were still a disproportionately large number of all gulls found dead. Table. Features of condition and oiling for different species of gull (Laridae) in samples of dead birds after the Braer wreck in Shetland, January 993. GBBG = Great Blackbacked Gull Larus murium; HG = Herring Gull L. argetiiatiis; Kittiwake Rissa iridactyla; Iceland Gull L. glaucoides. = not recorded, c = circa (approximately) Percentage positive for feature (no. examined) Feature of condition or oiling GBBG + HG Kittiwake Iceland Gull Moderatehigh, (score 23} fat scores for subcutaneous fat bodycavity fat deposits Pectoral muscles convex in transverse profile Much food, crop to gizzard 0% or more of plumage oiled Oil in gizzard (22) (22) c0% (22) 3% () 63% (32) 4% () 8% (6) 0% (64) 30% (64) 42% (43) % (3) 0% (3) 0% (2) % (3) 2% () % (3) *The Shedand Iceland Gulls, on detailed internal examination, exhibited dry, sticky subcutaneous tissues, diffusely pale tan kidneys and urates visible in the cloacae. Their lungs were oedematous (the tissues swollen with fluid) and most had apparent enlargement of adrenals (small glands close to the kidneys). The kidneys showed diffuse swelling owing to fluid and some degeneration of the epithelial cells (in the thin lining of internal cavities). Livers, spleens, bursae (sacs that serve to reduce friction in tendons) and hearts appeared relatively normal. Lesions found were consistent with severe dehydration and starvation, which may well have resulted in death, in conjunction with the direct effects of ingesting oil.
22 Iceland Gulls from the 'Braer' disaster The crop, proventriculus and gizzard of 3 Iceland Gulls were opened. Only eight contained any food, six of them very little. We noted a crustacean claw (possibly Cancer) in one, mussel Mytilus (in one), beaks of squid Gonatus, probably including G. steenstrufrii (two), worn fish bones, large and small (two), a small Herring Clupea harengus (one) and unidentified tissue (two). One contained a piece of plastic. Food items were similarly few in gizzards of gulls in the 983 Faroes invasion (Fjeldsa & Jensen 98). By contrast, the main study of winter feeding ecology in the Palearctic was from gizzard contents of 3 adults, in Iceland in 966 (Ingolfsson 96). Discussion That a rare Palearctic invasion of Iceland Gulls should coincide with a major oil spill was extraordinary. This coincidence, given constraints on collecting, was the only way in which a substantial sample could be obtained and examined in the UK. While the sample was much smaller than those from the Faroes in 983 or Iceland in 966, it was similar to one or both in the high proportion of adults, some of them Canadian; in secondwinter birds outnumbering firstwinter birds; and in the high proportion of females. The last two features are also typical of specimens from noninvasion years in the Faroe Islands and in Britain and Ireland (Ingolfsson 96; Fjeldsa & Jensen 98; J.K. Jensen in litt.; A. Petersen in lite; Weir ct al. in prep.). Invasions of Britain and Ireland have been recorded since the one in winter 82/3 (J. A. HarvieBrown, mss. 823), which extended from at least Orkney to North Yorkshire, with many hundreds of mainly adult Iceland Gulls in the inner Firth of Forth, where fields were being manured with locally caught fish; one was noted as having wingtip markings. Another invasion occurred in southwest England in 84 (Saunders 884). During 90, several contributors to British Birds recorded unusually many and widespread Iceland Gulls with relatively many adults (Kay 90; Tucker 90). The wellknown, widespread 982/83 invasion involved very many; the high proportion of adults included at least four wellmarked kumueni (Hogg 984). The 993 invasion was similarly large and included at least one and perhaps as many as five kumlieni besides those recorded here (Millington 993; Rogers et al. 994). After 983, a very few kumlieni repeatedly wintered in the same Scottish or Irish localities; one in Peterhead Harbour was annual up to 994 (S. Alexander in litt.). Palearctic nesting was reported on Jan Mayen Island in the 880s (Cramp & Simmons 983). This followed two British invasions in trie preceding decade (above). Almost all the Palearctic invasions referred to have peaked in early January. The Faroese invasion followed severe westerly storms and very low sea temperatures off southwest and south Greenland (Fjeldsa & Jensen 98). The 993 invasion of Shetland and elsewhere in Britain and Ireland was similarly during very severe and prolonged westerly storms (Murray 993; Nightingale & McGeehan 993; Ecological Steering Group on the Oil Spill in Shetland 993). Apart from Iceland, where invasions of adults may be the result of increased food supply (Faber 820, as cited in Saunders 884; Ingolfsson 96), adults reach the Palearctic mainly when displaced from their normal winter ranges by such storms. Weights indicate that, the farther south and east they are driven, the more
British Birds, vol. 88, no., January 99 23 starved they become. For adults, near the eastern end of the winter range, the 993 Shetland occurrences effectively were a seabird 'wreck'. On top of this, Iceland Gull feeding habits may have made them exceptionally vulnerable to the widely dispersed Braer oil. In the northwestern insular Palearctic, the normal winter feeding habits of Iceland Gull and Kittiwake Rissa tridactyla are wholly marine, whereas those of Great Blackbacked Gull L. marinus and Herring Gull L. argentatus are largely littoral (Ingolfsson 96; Cramp & Simmons 983). This could explain differences between species in oiling and in the relative frequency of casualties. Iceland Gulls may have been especially likely to dip in flight for oiled food items on the surface in a contaminated area. Our studies of Iceland Gulls have implications for field identification. It was thought that wingtip measurement, melanism (the extent of dark primary patterning) and other characters were continuously clinal from Nearctic Herring Gulls L. a. smithsonianus through thayeri and kumlieni to glaucoides (Dwight 92). The smilhsonianusthayeri dine was doubted by MacPherson (96). This portion of the cline remains uncertain, but thayerikumlieniglaucoides intergrades are evident in a review of historical records and museum specimens (Weir et al. in prep.). Winter adult specimens collected from Greenland to Britain and Ireland include increasingly high proportions of pale kumlieni, which are indistinguishable from glaucoides in the field (Weir et al. in prep.). It follows that sight records are at best of 'kumlienilike' or 'glaucoidesuke' individuals. Even the taxa used here are only convenient definitions. The probability of Palearctic occurrences of thayeri is low. This race winters mainly on the CanadianUS Pacific coast and, during summer, migration and winter, is the least pelagic Iceland Gull race (Weir et al. in prep.). In a 4year study near Montreal, only 3% of,392 migrating Iceland Gulls were thought to be thayeri (Steeves et al. 989). Most dark young individuals wintering on the east North American coast were thought to be in the kumlienithayeri cline, in a good article on identification (Zimmer 99). The extreme darkness of young kumlieni, however, darker even than young smithsonianus, was noted when these birds were first found breeding (Kumlien 89). The Shetland and North Ronaldsay gulls were said to be the first British kumlieni specimens (Millington 993). There is at least one earlier: a thirdwinter female in the Natural History Museum, Tring (89..0.), taken in Shetland on 24th November 869. It was registered as glaucoides, but the tips of primaries 08 had brown immature markings on the outer webs and pale adult kumlieni Type 2 markings on the inner webs (DNW pers. obs.). This seems to be the first Palearctic record. The recent National Museums of Scotland series of adults from Scotland and the Faroes includes most of the wingtipmelanism types in the kumlieniglaucoides cline. Acknowledgments The authors are most grateful to the Shetland Wildlife Response Coordinating Committee for organising the collection and delivery of corpses, and to Scottish Natural Heritage and the National Museums of Scotland for providing grants to cover the costs of processing the dead birds killed by oil from the Braer. We also thank Christian Salvesen pic, P&O, A&J Beveridge Ltd, Lever Bros, and Baroid (UK) Ltd for providing additional resources and materials. We further thank Dr A. Petersen and J.K. Jensen for access to data, Dr R. PrysJones for
24 Iceland Gulls from the 'Braer' disaster access to specimens al the Natural History Museum, Dr N. P. Ashmole, Dr J. Uttley and P. Harvey for commenting on an earlier draft, and Dr W. R. P. Bourne for further comments and for arranging access to specimens from Aberdeen University Natural History Museum. We are grateful to the editors of Dansk Orn. Foren. Tidsskr. for permission to reproducefig.. Summary In January 993, an invasion of emaciated, mainly adult, Iceland Gulls Lams glaticoides was stormdriven to Shedand, where it coincided wifh the Braer oiltanker disaster. Of specimens, eight were adult and only one was firstwinter. Two of six adults which could be assigned to a race were Canadian kumlieni. Compared with other gull casualties, Iceland Gulls appeared to be especially likely to ingest oil. The specimen sample was like diosc from invasions in southwest Iceland (966) and in the Faroes (983). References ANON. 96992. Shetland Bird Reports. Lerwick. CRAMP, S & SIMMONS, K. E. L. (eds.) 983. The Birds of the Western Pakarctic. vol 3. Oxford. Dwic.HT, J. 906. Status and plumages of the whitewinged gulls of the genus Lams. Auk 23: 2643. 92. The gulls (Laridae) of the world; their plumages, moults, variations, relationships and distribution. Bull. Arner. Mus. Nat. Hist. 2: 3402. ECOLOGICAL STITRING GROUP ON THH OH. SPILL IN SHETLAND. 993. An Interim Report on Survey and Monitoring, May 993. Edinburgh. FJKI.DSA, J., &JHNSKN, J.K. 98. 'Invasion' af Hvidvingede og Kumlien's Mager Lamsglaucoides glaucoides og kumlieni pa Nolso pa Faeroerne. Dansk Orn. Foren. Tidsskr. 9: 0306. GASTON, A. J., & DHCKHR, R. 98. Interbreeding of Thayer's Gull Lams thayeri and Kumlien's Gull Lams glaucoides kumlieni on Southampton Island, Northwest Territories. Can. FieldNat. 99: 229, GODFREY, W. E. 986. K Birds of Canada. Revised edn. Ottawa. HARVIHBROWN, J. A. 823. Unpublished ms. journals. National Museums of Scodand Library. HOGG, A'. 984. Scottish Bird Report 983. Scot. Birds 3: Suppl. INGOI.FSSON, A. 96. The Feeding Ecology of Five Species of Large Gulls (Lams) in Iceland. Unpublished PhD dissertation. University of Michigan, Ann Arbor. KAY, G. T. 90. The Iceland Gull in winter. Brit Birds 43: 399402. KUMI.IHN, L. 89. Contributions to the natural history of Arctic America. US Nat. Mus. Bull. : 9. LACK, P. 986. The Atlas of Wintering Birds in Britain and Ireland. Calton. MACPIIHRSON, A. H. 96. Observations on Arctic Ijjrus gulls and on the taxonomy of L. thayeri Brooks. Arctic Inst. North America Tech. Paper : 40. MILLINGTON, R. 993. Identification and status of Kumlien's Gull. Birding World 6: 006. MURRAY, A. 993. Recent reports. Scottish Bird News 30. NIGHTINGAI.H, B., & McGlHHAN, A. 993. Recent reports. Brit. Birds 86: 02, 0, 200. RAND, A. L. 942. Lams kumlieni mid its allies. Can. FieldNat. 6: 2326. ROGHRS, M. J., & THH RARI'I IHS COMMI'ITKE. 994. Report on rare birds in Great Britain in 993. Brit. Birds?,!: 3. SAUNDHRS, H. 884. A Histoty of British Birds by Wm. Yarrell. 4di edn. vol. 3. London. SMITH, N. G. 966. Evolution of some arctic gulls (Ijims): an experimental study of isolating mechanisms. Ornith. Monogr. 4. SNF.LL, R. R. 989. Status of Lams gulls at Home Bay, Baffin Island. Colonial Waterbirds 2: 2 23. 99. Conflation of the observed and the hypothesized: Smith's 96 research in Home Bay, Baffin Island. Colonial Waterbirds 4: 96202. STKKVKS, J., Hoi.0 IAN, S., & BATHURST, R. 989. Migration of Larids at Beauharnois, Quebec, 96980. Can. FieldNat. 03: 2338. TuGKiiR, B. W. (ed.) 90. Iceland and Glaucous Gulls in Britain in 90. Brit. Birds 43: 409 43.
British Birds, vol, 88, no., January 99 2 WKIR, D. N., KITCHENER, A. C, & MCGOWAN, R. Y. In prep. Distribution changes in operationally defined taxa of the Iceland Gull IMTUS glaucoides Meyer 822. ZIMMKR, K. J. 99. The impossible identification zone. Plumage variation in 'KumlienY Iceland Gull. Birding (99): 24269. D. N. Weir, R, Y. McGowan and A. C. Kitchener, Department of Natural History, Royal Museum of Scotland, Chambers Street, Edinburgh EHl JF S. McOrist, Department of Veterinary Pathology, Royal (Dick) School of Veterinary Science, University of Edinburgh, Veterinary Field Station, Easter Bush, Midlothian EH2 9RG B. Zonfrillo, Department of Zoology, University of Glasgow, Glasgow G2 8QQ M. Heubeck, AURISISOTEAG Ltd, Broonies Taing, Sandwich, Shetland ZE2 9HH