Date: 2/21/25 6:53 pm From: Shaibal Mitra <Shaibal.Mitra...> Subject: [nysbirds-l] Certainty in Bird Identification in Relation to Hybridization: Population Thinking is the Key
Ernst Mayr argued that natural selection was discovered independently by two English people (Darwin and Wallace), rather than by a German, not by chance, but because the tenor of biological thought in mid-19th century Germany was overly typological (focused on the typical, or “perfect” exemplars of structure and function among cells, tissues, organs, or organisms), whereas the English mindset at that time was more influenced by natural history and an appreciation of the variability within biological populations. Whether this historical hypothesis is correct or not, the distinction between typological thinking and population thinking is real, and it offers solutions to many of the most vexing debates in bird identification.
Species are not defined by diagnostic characters; instead, they are recognized via such characters. Similarly, the identification of an individual bird is not simply a matter of discerning the presence of field marks (although it involves this in practice). Identification consists of making an inference regarding the population origins of that individual bird. These distinctions might seem arcane—and it is true that in the vast majority of cases birders will reach the same conclusions regardless of whether they are applying population (better) or typological (not as good) approaches—but again and again, in a small but important subset of identification debates, applying population thinking yields practical and logically consistent solutions that simply can’t be reached via typological thinking. This essay illustrates the utility of population thinking in problems relating to certainty, species boundaries, and hybridization, a using familiar contemporary debate among birders in northeastern North America.
Crabby is the nickname of a Larus gull that has wintered at Old Field Point, in northwestern Suffolk County, Long Island, in recent years, and has been suspected since the earliest days of its discovery as a European Herring Gull. Two recent events have reinvigorated birders’ interest in Crabby: the AOS’s elevation of American and European Herring Gulls to species status; and Hanyang Ye’s superb photo-documentation of Crabby and bold decision to name it a European Herring Gull in eBird. I identified this bird as a European Herring Gull three years ago,* at which time this idea caused quite a stir, as many birders, even the most knowledgeable and highly skilled, weren’t aware that Larus argentatus argentatus included populations in which birds of similar appearance occur frequently.
I knew about yellow-legged Herring Gulls in northern Europe from college days, via Ernst Mayr’s discussion of “Larus omissus” and the Herring/Yellow-legged/Lesser Black-backed Gull complex. Mayr proposed that this complex was possibly a ring species, in which reproductive isolation was weak between many adjacent populations around the northern Northern Hemisphere, but strong at one end of the ring: “where argentatus and fuscus meet along the coasts of Europe they live unmixed side by side, hybridizing only quite rarely.”**
Although some aspects of Mayr’s interpretation of this complex are incorrect or incomplete, his observations regarding European Herring Gull and Lesser Black-backed Gull are not only accurate, but they illustrate beautifully the power of population thinking over typology: populations of European Herring Gull and Lesser Black-backed Gull are large and broadly overlapping. Reproductive isolation is strong and hybridization is rare, as tens of thousands of birds pair assortatively even given the opportunity to choose heterospecific mates. In contrast, the typological perspective fixates on the small number of instances of hybridization, even though these are basically imperceptible at the population level. In terms of identification, if one were to randomly sample a thousand gulls from the area of sympatry, any competent birder would be able to correctly identify all or almost all. Again, what this means is that if you show me a picture of a randomly chosen gull from the Netherlands, I will be able to infer whether it is a Herring Gull or a Lesser Black-backed Gull with extremely high confidence—high enough to satisfy the threshold of certainty, as it relates to identification.
Every now and then (probably less than one in a thousand), I might encounter a hybrid or variant that defies confident assignment to one population or the other. These, I suggest, should be left unidentified, or if intermediate in many ways, provisionally interpreted as hybrids, because hybridization has been known to occur. In the very rare cases where I know that an individual bird is a genuine hybrid, studying and documenting its appearance becomes especially valuable, because this information will be of great utility in inferring the identities of the much larger number of intermediate-looking, provisional hybrids. By “genuine hybrid” I mean something very definite: an individual whose parentage is known confidently and whose parents are confidently attributable to the different species in question. Conversely, an "Olympic Gull" both of whose parents are known, and known to be "Olympic Gulls," is not a hybrid between Western and Glaucous-winged Gull.
It is worth exploring further the two contrasting perspectives (population vs. typological) regarding individuals that appear intermediate between two species that are known sometimes to hybridize. A large percentage of birders apply the typological perspective and regard many, most, or all intermediate birds as hybrids. But from the population perspective it is important to distinguish, at least conceptually, among at least three different ways in which birds can appear intermediate.
First, the bird might be a genuine hybrid, with one parent from a long line of Herring Gulls and the other from a population of Lesser Black-backed Gulls. The actual frequency of these is important because it is a measure of the strength of reproductive isolation. Everybody agrees that such matings are very rare in European populations and that reproductive isolation between European Herring and Lesser Black-backed Gulls is strong. Second, an intermediate-looking bird might be a recent back-cross between a genuine hybrid and one or the other of the parental species, or between two hybrids. Again, distinguishing these from the first case is important conceptually, even if it might be difficult or impossible in practice, because reproductive isolation might (or might not) be weaker between hybrids and one parental species or the other. Let’s imagine for the sake of argument that Lesser Black-backed and Herring Gulls almost never choose each other as mates (which is true), but that Lesser Black-backed Gulls accept hybrids as mates with zero aversion (and, for fun, let’s say that Herring Gulls avoid mating with hybrids with even greater aversion than they show for “pure” Lesser Black-backed Gulls). If this were true (which it probably isn’t in detail, but it is perfectly plausible), the result would be a slow, steady trickle of Herring Gull genes into populations of Lesser Black-backed Gulls, but not the other way around. If actual hybridization were rare enough (which it is) this would not mean that the two species should be lumped, or that Lesser Black-backed Gull should be regarded as a hybrid swarm. But it would mean that the biological identity of Lesser Back-backed Gulls at the population level includes numerous genes that originated in Herring Gulls. This leads to the third category, in which an intermediate-looking bird might be nothing more than a Lesser Black-backed Gull, all of whose recent ancestors were Lesser Black-backed Gulls, but in whom “Herring Gull genes” have by chance recombined to produce intermediate-looking characteristics. This sort of thing is actually very commonplace, and at a certain point, it isn’t even necessary or important to specify that the genes producing the traits in question had their origin in the population via a remote hybridization event.
Thus, there is even a fourth category. The large amounts of genetic variation observed within many natural populations does not require invoking hybridization at all. Recall that the wolf-like ancestor of domestic dogs harbored within its gene pool all of the genes necessary to produce animals as small as chihuahuas and as large as St. Bernards. Thus, a very small wolf need not be a hybrid with a Coyote, nor a very large Coyote a hybrid with a wolf. The mutations occurring in any population (let’s say, Lesser Black-backed Gulls) will inevitably include some where a gene changes from the fuscus “type” to the argentatus “type.” Thus Herring-like variants ought to be expected among Lesser Black-backed Gulls (and vice versa) even if hybridization were entirely absent, simply because large populations tend to be highly variable. It is here where the weakness of the typological perspective becomes most obvious. Typologists readily acknowledge that hybrids can be surprisingly and unexpectedly variable, and even show characters beyond those of either parent. Although this statement is true to a point, it is at least partly fueled by the uncritical conflation of genuine hybrids with other kinds of variants. The typological perspective overstates the frequency and significance of hybrids and at the same time greatly underestimates the often very high variability of populations as well as their capacity to produce unusual-looking individuals, including individuals that resemble other species in one or more traits, without any need for hybridization at all.
I see a lot of Lesser Black-backed Gulls on Long Island (1,165 observations of 8,382 individuals, according to eBird), and I have noted a small number of birds that I think were likely hybrids with Herring Gull, based on multiple points of intermediacy. But I have also noted innumerable other kinds of variations, including birds that were unusually large, heavy-billed, with narrower or broader than usual tail bands, etc. A particularly interesting category comprises those whose mantle tone was significantly darker than usual. These have raised the question of whether they might represent the subspecies intermedius, or even fuscus. The problem is that although some of them showed other traits of those taxa (e.g., size and shape), others showed more “typical” graellsii-like structure. For now, I don’t feel confident inferring that these originate from populations of intermedius or fuscus.
It is more conservative, I think, and definitely more logical, to allow for variability within and among the rapidly growing populations in Greenland, where most of our birds come from. From the population perspective, these rapidly growing populations were founded by relatively small numbers of individuals drawn from a genetically diverse source. This means that some genetic variants that occur at low frequency in Europe would be over-represented in Greenland, and therefore on Long Island. It also means that bulky bodies, pale mantles, pinkish legs, etc. cannot be ascribed to hybridization with Herring Gull with any kind of certainty, and that even provisional confidence in hybrid origin should be reserved, until we know more, for birds showing intermediacy in multiple ways. Failing to distinguish between genuine hybrids and birds regarded as likely hybrids makes it impossible to improve our understanding of how genuine hybrids actually look.
And of course, Crabby’s intermediacy between Lesser Black-backed Gull and American Herring Gull is limited to just two traits, mantle color and gonys pattern. In important traits like size and shape (which, by the way, generally show intermediacy in hybrid vertebrates), Crabby resembles a Herring Gull. The red orbital ring resembles the condition in many Larus taxa, including northern populations of European Herring Gull, as well as Lesser Black-backed Gull. And the wingtip pattern is completely unlike that of Lesser Black-backed, unusual for American Herring, but again similar to those found in northern populations of European Herring Gull. There is no basis to infer that Crabby originated in a colony of North American Herring Gulls with a Lesser Black-backed Gull as the other parent, nor among Greenland Lesser Black Backs, with an American Herring Gull—or even a vagrant European Herring Gull—as the other parent.
Crabby's identification as a European Herring Gull is based on the observation that in terms of all these important Larus field marks its appearance best matches those found regularly in that taxon. Is it possible, as Mayr and others have hypothesized, that some of the genes that characterize the northern populations of argentatus might have originated via past hybridization with lesser Black-backed and/or Yellow-legged Gulls? Of course!—but this is purely speculative and it is not relevant to the rationale for identification. Although the gene pools of natural populations have diverse and sometimes highly complex histories, sometimes involving hybridization, it is very often possible to infer which sort of population an individual bird comes from.
Last points: it has been observed that Crabby is not a “typical” European Herring Gull. First, note that European Herring Gulls consist of many, genetically connected populations that vary appreciably in overall appearance. Crabby can’t be expected to conform to the range-wide norms of the species. But knowledgeable observers have noted that even among the northern, yellow-legged populations, most individuals have drabber legs in winter and lose their head streaking earlier than this. Aren’t these obstacles? The first answer is that, yes, Crabby is not perfectly typical. But the next is that neither deviation is positively more suggestive of any alternative interpretation, and certainly not suggestive of a Lesser Black-backed x American Herring Gull hybrid. The second answer is more granular. It has been observed that yellow leg color is unexpectedly prevalent among typically pink-legged American Herring Gulls and Great Black-backed Gulls at this site, and even more generally along the North Shore of Long Island (and beyond) at this season (Mitra, unpublished data). This very specific observation suggests that diet might play a role here, mitigating concern over Crabby’s deviance. Similarly, retention of winter plumage late into the winter/spring does not point to any particular alternative interpretation, but it might be expected in a vagrant. For those who believe that the split of European Herring Gull and American Herring Gull was warranted (I am skeptical), Crabby might something of a Rosetta Stone; if it turns out that European Herring Gulls really are reproductively isolated from American, then it would make sense that perennially mateless Crabby’s hormone cycles might be slowed depressingly.
The question ultimately is the degree of certainty that each person might require. European Herring Gull is the best single interpretation, but what can we say about the next most likely contenders, such as a hybrid, which, even if unlikely, might nevertheless cast shade on accepting a first state record? This is a legitimate question. I offer the discussion above to explain why the hybrid hypothesis is weaker than some would suggest, but it is up to each to decide where the bar is set.
*eBird checklist from the day after its discovery by Patrice Domeischel and others (unedited, even to soften the hyperbolic language regarding over-reporting of hybrids):
https://ebird.org/checklist/S103599677
