glaucoides & kumlieni

(last update: October 12, 2011)

Iceland & Kumlien's Gull adult summer

Current status of hybridisation

(Peter Adriaens, Dec 2011)

Despite many writings in magazines and various internet fora, nobody actually knows the exact taxonomic position of Kumlien's Gull. The literature on the subject is confusing; research is hampered by the inaccessibility of both the (high Arctic) breeding areas and breeding habitat, i.e. tall and steep cliffs. Some authors treat the taxon as a subspecies of Iceland Gull (e.g. Godfrey 1986; AOU 1998; Olsen & Larsson 2003), a few as a valid species in its own right (such as Taverner 1937; Sutton 1968; Chu 1998; Browning 2002), and according to others it is a population of hybrids (a so-called hybrid swarm) produced by interbreeding between Thayer's and Iceland Gulls (Dwight 1925; Weir et al 2000). Although the latter theory indeed seems to offer a valid explanation for the enormous plumage variation seen in Kumlien's Gull, it is misleading nonetheless, as more seems to be going on than merely hybridisation:

• The breeding areas of Thayer's and Iceland Gull have been separated for well over a century now. The breeding area of Iceland Gull shrunk significantly from 1860 and by 1900 the species was confined to Greenland (Weir et al 2000). Thayer's Gull, on the other hand, breeds in the high Arctic region of Canada and has disappeared as a breeding bird from Greenland since 1930 (Boertmann 2001). Moreover, its distribution in Greenland had always been limited to the extreme northwest (above 77°N), while Iceland Gull mainly breeds in the southern part of the island, reaching its northern limit at 74°N along the west coast (Boertmann 2001).
• Preliminary DNA results have shown that Thayer's Gull is more closely related to Glaucous-winged Gull than to either Iceland or Kumlien's Gull (AERC TAC 2003; Gay et al 2005). Chu (1998) had reached the same conclusion before, but on the basis of morphological and skeletal data.
• The current breeding population of Kumlien's Gull (c. 10,000 pairs) is about twice the size of that of Thayer's Gull (4,000 to 6,000 pairs) (Weir et al 2000) - a strange discrepancy for any "hybrid population".
• Kumlien's Gull has a separate range, both in summer and in winter. Its breeding area is mostly limited to the southern half of Baffin Island, Canada. Thayer's Gull breeds in the northern half of Baffin Island and further to the north and west (Howell & Dunn 2007). Therefore, it is only on Baffin Island that the breeding areas of both taxa meet or, perhaps, overlap to a certain extent. In Greenland, no more than two breeding records of Kumlien's Gull are known so far, so its range does not really overlap with that of Iceland Gull either (Boertmann 2001). Kumlien's Gull mainly winters along the east coast of Canada and the United States, while Thayer's Gull migrates to the west coast and Iceland Gull mainly winters in Greenland and Iceland (Howell & Dunn 2007; Olsen & Larsson 2003).
• There are also small ecological differences. In winter, Kumlien's and Iceland Gull are more of a seabird than Thayer's Gull, which appears more strongly dependent on coastal areas (Howell & Mactavish 2003).
• A general rule used by many authors to identify a certain taxon as a valid subspecies is that about 70-75% of the population has to differ from at least 99% of all other populations in one or more characters. The colour of the wing tips as well as the iris colour separate about 98% of adult Kumlien's Gulls from all adult Iceland Gulls, while body size, structure, colour of the upperparts, and primary pattern separate them from Thayer's Gull too. Thus, there seems to be no reason in this respect not to consider Kumlien's Gull a valid subspecies (Amadon 1949; Howell & Mactavish 2003).
• It is true that some authors have reported extensive interbreeding between Kumlien's and Thayer's Gull on Baffin Island, but at least some appear to have underestimated the separation of both taxa in the field. The "evidence" for extensive interbreeding actually seems to be quite poor. One particular study from 1986, for instance, which reported many mixed pairs (Gaston et al 1986), simply states that all breeding pairs were censused from a small aeroplane and that only those birds with "wing tips as pale as in Glaucous Gull" were written down as Kumlien's Gulls! The chapter on hybrdization in the Weir et al paper (2000) relies mainly on Knudsen 1976, Gaston & Decker 1985, Snell 1989, and Renaud et al 1981 (the latter misspelled as 'Reynaud'). Several of these references are freely available online nowadays (see References below). The Knudsen (1976) reference is mainly an abstract (12 lines) of a lecture given at the 1975 Pacific Seabird Group annual meeting (12 December 1975). The actual lecture itself was not published. The abstract mentions four birds collected at Home Bay, Baffin Island, that were identified as hybrids of Larus thayeri and L. glaucoides on the basis of measurements, iris colour, and plumage patterns. It goes on to say that "mixed pairs of these two species were also seen". It does not use the name kumlieni, but presumably that is what L. glaucoides means in this case. From this short abstract alone, it is impossible to know what criteria the author used to distinguish between Thayer's Gull and dark Kumlien's Gull, nor is it clear how many mixed pairs were seen. The Gaston & Decker paper (1985) is short (2 pages); it reports on a colony of 90 - 100 birds on eastern Southhampton Island. Of twelve pairs studied, five were mixed (Thayer's x Kumlien's Gull). However, the criteria used to distinguish between Thayer's and Kumlien's Gull were not very sophisticated; the 'darkest' bird they considered to be Kumlien's Gull had only "medium brown markings on the primaries - distributed in exactly the same fashion as typical Thayer's". The latter taxon they define as having "extensive black or brown pigmentation at the tips of the five longest primaries". However, Howell & Elliott (2001) have shown that most Thayer's Gulls in their core wintering range have black(ish) markings at the tips of six longest primaries, and Howell & Mactavish (2003) have shown that some Kumlien's Gulls in their core wintering range have blackish wingtips (with the markings on the five longest primaries). The Renaud et al paper (1981) does not say anything about hybridization. It simply reports on birds seen just north of Baffin Island in 1978-1979. The authors mention "Iceland Gull (L. glaucoides)" as a "rare/uncommon summer visitant" (no nesting observed), and Thayer's Gull as a "very common visitant; probably breeds". It is not clear why the Weir et al paper (2000) includes this reference as "evidence" of hybridization. Snell (1989) visited 14 gull colonies at Home Bay, Baffin Island, in 1985 - the same colonies that another researcher named Smith had visited twenty years earlier. He did not find any Thayer's Gulls at these sites, or perhaps he included them all in "melanistic glaucoides" in his paper - that is not entirely clear to me. In any case, he only mentions glaucoides for the colonies visited, and does not explicitly mention thayeri for these visits. He describes non-assortative mating between "melanistic glaucoides and non-melanistic glaucoides" (contra Smith's results), but not mixed breeding between Thayer's and Kumlien's Gull. Essentially, I strongly get the feeling from reading the paper that Snell could not be bothered distinguishing between Thayer's Gull and dark-winged Kumlien's Gull. Again, it seems doubtful to use this paper as "evidence" for extensive interbreeding between these two taxa.
  Whatever the case, mixed breeding limited to a rather small contact zone (such as Home Bay, Baffin Island) would not necessarily affect the taxonomic position of Kumlien's Gull (cf. the situation of Herring and Glaucous Gulls in Iceland - two taxa whose status as distinct species is generally accepted).