01 January 2017

Thayer's the Iceland Gull - One Species

Respected reader,

After much consideration, I recently announced a drastic change in my position regarding the Thayer's-Kumlien's-Iceland gull complex. I have given this taxonomic conundrum countless hours of thought, both in the field and at my desk. The subject is not one that I take lightly as it has troubled me from almost the start of my interest in gulls. It wasn't until early last year, however, when I was preparing an article to be published on separating Thayer's and Kumlien's Gulls, that I candidly began to question my own pedagogy on this topic. I effectively abandoned the manuscript as I could not faithfully reconcile the identification of these forms without more clarity on their taxonomy.

Based on my experience here on the wintering grounds, and careful review of all the literature, I have resolved to following Snell, Godfrey et al (1986) who classified Thayer's as the darkest subspecies of Iceland Gull. Given all the information that is currently available, I believe that this is the most sound and practical approach.

Going forward, I now regard all 3 taxa as conspecific:
  • Iceland Gull (L. glaucoides glaucoides)
  • Kumlien's Iceland Gull (L.g. kumlieni
  • Thayer's Iceland Gull (L.g. thayeri).

My usage of "Kumlien's" and "Thayer's" is no more than an abbreviation of the common names above. Iceland is a broad name that can refer to all 3 subspecies collectively, or specifically nominate depending on the context. Note that this classification is by no means novel, having been employed by various workers, ornithological unions, and records committees around the world for some time.

The premise of this grouping is that the subspecies are clinal, with upperparts and primary patterns from palest to darkestBetter presented, perhaps, is that no evidence is sufficiently convincing to the contrary -- this is at the crux of the matter.

Gaps, Gaps and More Gaps

Consequently, until we have proper knowledge of what a Kumlien's Gull is, the argument for a two-species solution will remain flawed.

For those who champion the current two-species classification held by the AOU, what justification can be given for treating Kumlien's exclusively as a subspecies of Iceland Gull? That is, why not classify it as the paler subspecies of Thayer's Gull? I reiterate - verbatim - the question asked by Howell and Elliott: "At what point do Thayer's become reliably separable from Kumlien's?" Are we aimlessly trying to separate a paler form of a species from its darker form (or vice versa)? Or are we trying to distinguish hybrid populations from their parent species? These are critical questions that should not be underplayed via constructs that we've developed on the wintering grounds.

It is not very difficult to summarize what typical adult thayeri look like in California, or what many winter glaucoides should look like in Iceland. I don't deny these forms are real and identifiable. In fact, we've become so adept at field identification that we've even formulated an "operational" definition for wintering Kumlien's Gulls in Newfoundland. But spend a few winters on Lake Michigan or Lake Superior and the pieces soon begin to crumble.

A continuum of phenotypes is on display in the Great Lakes region where conjuring a line to separate dark-end Kumlien's and pale Thayer's can't be done with much satisfaction. Arbitrary decisions are often made when these "problematic" birds are found by cognizant observers. The need to assign a name to what we see is desirable, but to pretend we have genuine knowledge of what labels these individuals should be given, is futile (I myself have been guilty of engaging in this mangled labeling). To try to remedy this confusion we've readily come up with names to band-aid this problem: Tweener, Thiceland, Lake Michigan Gull, and so on and so forth. I ask, rhetorically, if Thayer's and Kumlien's have the potential to overlap in every observable field character - and they supposedly do have the potential to overlap in every observable field character - then what is to prevent someone from labeling the majority of "Thayer's" seen on the Great Lakes as dark-end Kumlien's? This is the big elephant in the room and it is pivotal to my argument.

A similar conundrum exists with presumed pale Kumlien's and nominate Icelands wintering in the North Atlantic Ocean (particularly between eastern Greenland and the western UK). Kumlien's is increasing annually in this region and there is no shortage of birds that appear to be nominate glaucoides, less than a small amount of faint peppering on the outer 1-2 primaries (see this example from southwest Iceland in late winter). Unsettling is presumed kumlieni with entirely white wingtips found on both continents. Some of these individuals have pale gray upperparts similar to Glaucous Gull, while others have upperparts that appear to be similar to American Herring Gull or slightly paler. Circular reasoning abounds, yet we are told to either chalk it up to a highly variable taxon that overlaps with both thayeri and glaucoides, or to embrace the uncertainties inherent in this complex.

I beg the question, why can't we admit that at this time there's an existing phenomenon that is beyond our reach? How is it acceptable that we not have a clear understanding of what a species looks and sounds like throughout its breeding range, and what relation - if any - exists among the taxa it purportedly interbreeds with?? What harm is done, then, in accepting a white-winged subspecies on one end, a black-winged subspecies on the other, and an intermediate form that suitably bridges the two? If there are no lines that can be objectively drawn, then those who insist on the current undeveloped split take on a great burden of proof. Prove away. 

At any rate, as information continues to be shared - easier and faster than ever - we're learning that these confusing birds appear to be increasing, and they're inconveniently turning up in places that weren't known to harbor them with any frequency just two decades ago (and these are just the individuals being reported!). I've indicated to a number of people whom I've discussed this topic with, we're either becoming better at recognizing problematic birds, or the "lines" on both ends are becoming blurrier. In due time, it behooves us to recognize and appreciate this remarkable cline. 

West Coast Thayer's - East Coast Kumlien's

Problem birds regularly have the potential of taking on two different labels depending on the regional observer casting a vote. Conservative records committees have only added to the bewilderment. As one records committee member from Washington state tried to explain to me, "The range maps don't lie, Kumlien's doesn't winter on the Pacific Coast". Ultimately, these committees have the job of making "yes" or "no" decisions, and to that end, I don't lay much blame on any one organization. After all, most committees admit that they await clarification on this taxonomy from the AOU.

The AOU's decision to remove Thayer's Gull from the Herring complex in 1973 was well supported. The two do not breed in sympatry and isolating mechanisms are discernible (see Macpherson 1961). But an obvious void in our knowledge persists. Thayer's, as a taxonomic unit, was never defined with rigor or consistency. Regrettably, the AOU was ill-informed about the relatedness of thayeri and kumlieni, relying on Neal Smith's erroneous study as its bedrock of information (Smith 1966). Sutton and others have unmasked Smith's study which shrouded the taxonomy of Thayer's in much confusion.

Accordingly, qualifying Thayer's Gull as distinct and separate from Kumlien's Gull was never substantiated on the breeding grounds. We now know that two forms - darker-winged and paler-winged birds - appear to interbreed randomly (nonassortative mating) to an unknown degree (Gaston & Deceker 1985). Admittedly, whether these darker birds are Thayer's or dark Kumlien's is uncertain. It seems the authors who have visited these sites where reported interbreeding occurs (South Hampton Island & Baffin Island) didn't systematically define the breeding adults they observed. In fact therein lies the problem: Very little work has been invested in thoroughly studying these taxa on the breeding grounds (and understandably so). Conflicting accounts and obscure reports of what the general phenotypes look like throughout the ranges of the Iceland complex keep us in want. Ambiguity is an understatement here.

Hybrid Swarm?

Naturally, a simple fix for Kumlien's Gull is to peg it a hybrid form, using its vast array of primary patterns as evidence. The chief paper supporting the 'hybrid swarm' theory deserves close scrutiny and I encourage you to read this thought-provoking piece (Weir et al. 2000). The arbitrary definitions we know from the wintering grounds are precisely what the authors use as their starting point, perpetuating more circular reasoning. Table 1 in Weir presents us with morphological overlap among the three taxa and "marginal separation" of the museum specimens used in their analysis. This is not surprising. The authors don't rely on any data that they collected from the breeding colonies, and the paper does not address any genetic relationships. In many ways, the authors make just as much the case for the "one species" theory as they do for the "hybrid swarm" theory.

Others have used the widely known Western x Glaucous-winged Gull hybrid phenomenon to argue Kumlien's is a hybrid form. This comparison, in my opinion, is poorly devised. Western and Glaucous-winged Gull have settled differences that are well known. They differ genetically (Pons et al 2005), they differ vocally, and they have yellow versus pink colored orbital rings. One is somewhat migratory while the other is mostly sedentary. There's also contrasting preferences for nesting habitat; Glaucous-winged Gull prefers estuaries and fjords, whereas Western Gull does best along California's coastal upwelling. Hence, there is some ecogeographic isolation that prevents either parent species from being swamped (Hoffman 1978, Bell 1997, Moncrieff 2012). The hybrid swarm in the Pacific Northwest is a fascinating event, but nonetheless it's a contained hybrid swarm in the breeding season.

On the other hand with thayeri, kumlieni and glaucoides, unique taxonomic units have not been adequately delineated -- this is undeniable. All are cliff nesters, all wear the same color orbitals and there is no compelling evidence from the breeding colonies that they have different voices (although some preliminary information from the wintering grounds might prove they do have distinct voices). There's a complete absence of molecular data from birds retrieved on the breeding grounds - this is most troubling. The genetic work that has been done consists of small sample sizes - 3 thayeri and 1 glaucoides - all of birds taken on the wintering grounds. We're not told how or why those specimens were selected, and the experimental design aimed to place these "species" in a phylogeny, makes no obvious effort to prevent false negatives. Add to this the uncertainty regarding how far north thayeri and glaucoides breed, and whether or not there is currently any contact between the two, and it becomes clear that our understanding of this complex is still emerging, at best. Although kumlieni may very well prove to be a large hybrid population, to say it is one without any evidence is bad science.

Concluding Remarks

The puzzle that surrounds this complex is presently beyond all of us. To declare that "the cline" is unequivocally true, is untenable. With all due respect, to insist unique biological entities exist is simply premature. Too many unanswered questions loom, and if we are honest with ourselves the null hypothesis is the fresh start that's needed in order to make progress.

Dissent is healthy and I do lament relinquishing one species from my checklist. I will gladly reverse my view in the future if convincing research becomes available. For now, I maintain that Iceland Gull is a highly variable polymorphic species with what appears to be 3 subspecies.


I would be remiss if I did not mention Ron Pittaway's excellent article on the history and taxonomy of "Thayer's Gull". Pittaway's paper has summarized this conundrum better than any other existing source that I know of.  

Gaston, A.J. and R. Decker. 1985. Interbreeding of Thayer’s Gull, Larus thayeri, and Kumlien’s Gull, Larus glaucoides kumlieni, on Southampton Island, Northwest Territories. Canadian Field-Naturalist 99.

Godfrey, W .E. 1986. The Birds of Canada. Revised Edition. National Museums of Canada, Ottawa.

Macpherson, A.H. 1961. Observations on Canadian Arctic Larus gulls, and on the taxonomy of L. thayeri Brooks. Arctic Institute of North America.

Pittaway, R. 1999. "Taxonomic History of Thayer's Gull". Ontario Birds 17(1):1-13.

Pons, J.M., P.A. Crochet, A Hassanin. 2005. Phylogenetic relationships within the Laridae (Charadriiformes: Aves) inferred from mitochondrial markers. Molecular Phylogenetics and Evolution 37.

Smith, N.G. 1966. Evolution of some arctic gulls (Larus): an experimental study of isolating mechanisms. Ornithological Monographs 4. American Ornithologists’ Union, Washington, D.C.

Snell, R.R. 1989. Status of Larus Gulls at Home Bay, Baffin Island. Colonial Waterbirds 12(1).

Weir, D.N., R.Y. McGown, A.C. Kitchener, S. McOrist, B. Zonfrillo and M. Heubeck. 1995. Iceland Gulls from the Braerdisaster, Shetland 1993. British Birds 88.