Mexican Ducks are not Mallards … probably

2010 January 30
by David J. Ringer
Mexican Duck

Mexican Duck by Pablo Lèautaud

VICKSBURG, MISS. — Version 2.3 of the IOC World Bird List came out last month, and the most interesting change for North American birders is probably the resplitting of Mexican Duck (Anas diazi) from Mallard (Anas platyrhynchos).

The Mexican Duck is resident from southern Texas, New Mexico, and Arizona south through northern and central Mexico. It is essentially non-migratory. Like the Mottled Duck of the Gulf Coast and the American Black Duck of northeastern North America, the Mexican Duck lacks the brightly colored male breeding plumage of the familiar Mallard; both sexes are mottled brown throughout their lives.

Hybrid swarm?

The Mexican Duck has been considered both a full species and a subspecies of Mallard since its formal scientific description in the 1880s. In 1983, the American Ornithologists’ Union lumped it with the Mallard and has left it there ever since, writing in the 1998 seventh edition of its North American checklist that “extensive hybridization in southeastern Arizona, southern New Mexico, and west-central Texas compels merger into a single species (Hubbard 1977).”

The study cited to back up that action was “The Biological and Taxonomic Status of the Mexican Duck” by John P. Hubbard, published by the New Mexico Department of Game and Fish. This study relied on physical characteristics to determine which birds were “hybrids.” Essentially, sets of criteria were established for “pure” Mexican Duck and for Mallard, and birds showing a combination of the features were ruled hybrids. The study concluded that most “Mexican” Ducks occurring in the United States are in fact hybrids and that therefore, the taxon does not deserve recognition as a species under the biological species concept. This analysis became the conventional wisdom and has been parroted ever since.

But it’s bunk. In short, there isn’t much evidence that there is massive, ongoing hybridization between Mallard and Mexican Duck. In a very interesting article, Richard E. Webster writes: “My recent experience in eastern Cochise County [Arizona] and environs is not of the ‘extensive’ hybridization that influenced the decision to lump the two. Rather, it is of no obvious hybridization, because I have not seen an obvious Mallard during the breeding season in the areas I bird, areas in which Mexican Ducks breed at a number of localities in small to moderate numbers.” He goes on to cite studies from the 1980s questioning the “hybrid swarm” theory, breeding bird survey data, and information about breeding biology, all of which seem to contradict the idea that Mexican Duck is being hybridized into oblivion. Apparently no satisfactory genetic studies, which could shed some light on the situation, have ever been done.

What is a Mallard?

The familiar sexually dimorphic Mallard has 10-15 very close relatives around the world, all of which are monomorphic (i.e., lacking a male breeding plumage). Many waterfowl species are capable of interbreeding and producing fertile offspring — even across generic lines — but the situation with the Mallard and its relatives is especially confusing. Interbreeding is rampant under certain conditions; for example, introduced Mallards and indigenous Pacific Black (“Grey”) Ducks (Anas superciliosa superciliosa) have mixed to such an extent in New Zealand that “pure” Grey Ducks are now quite rare.

In North America, Mallards also interbreed with American Black Ducks to a significant degree, but the AOU has always upheld the latter as a valid species.

Kissing cousins?

Two main hypotheses attempt to explain the origins of the Mallard and its closest relatives:

  1. A sexually dimorphic duck, probably rather like the green-headed Mallard of today, was ancestral to the whole clade. Under this scenario, the green-headed Mallard spread out around the world, and various populations began splitting off from it and losing the alternate male plumage, becoming monomorphic. This model can be used to argue that because all the monomorphic forms descended from a dimorphic ancestor, the speciation process did not complete before the populations started coming back into contact again (often due to the activities of humans). Thus, it’s not a stretch to lump them back together.
  2. A monomorphic ancestor gave rise to all or most modern Mallard relatives. Under this scenario, the dimorphic green-headed Mallard is a novelty within the clade. This model can be used to argue that a monomorphic ancestor reached North America and diverged into three populations: American Black Duck, Mexican Duck, and Mottled Duck. Then the green-headed Mallard secondarily invaded North America, coming into contact with the other three populations.

Obviously, if the second hypothesis is closer to the truth, it’s problematic (but not impossible) to treat the Mexican Duck a subspecies of Mallard.

This, in essence, is the argument the IOC uses in resplitting Mexican Duck, citing a 2001 study of Mottled Duck and its relatives:

McCracken et al. (2001) challenged the subspecies status of the Mexican Duck. They found that the Mexican Duck is the southwestern sister “species” of the Mottled Duck and in turn the American Black Duck, all members of a set of original and monomorphic “mallards” that speciated in North America before dimorphic “greenhead” Mallards expanded their range from Europe to North America. They are all closely related members of a recent allopatric radiation with no postzygotic barriers to gene exchange between them. However, they mate assortatively, and do not interbreed freely. McCracken et al (2001) therefore recommended that “Mottled ducks and Mexican ducks be designated as species so that the nomenclature is consistent with phylogeny.” (The Mexican Duck)

Looking closer

The 2001 study is fascinating and well worth a read (link below). Each Mallard carries one of two different genetic signatures in a certain region of its mitochondrial DNA. (Remember that mitochondrial DNA is passed directly from a female to her offspring without recombination.) The trick is that though the Mallard population carries both signatures, the Mottled, Black, and Mexican ducks sampled have only one: type B of the haplotype in question.

McCracken et al. suggest that this pattern was best explained by an invasion of green-headed Mallards into North America. Still able to interbreed with their New World cousins, they would have picked up the mitochondrial genes of the three monochromatic species by crossing with local females, and so the Mallard population would end up with both genetic signatures.

However, the authors note that this hypothesis could be called into question if the type B haplotype were discovered in Mallards in Asia.

So they studied the Mallards in Alaska and the Old World and in 2005 published a paper showing high percentages of type Bs in Alaska but virtually none in Asia and Europe (with one caveat), supporting the idea that Mallards picked up the type B haplotype when they reached the New World. But the high percentage of type Bs in Alaska (far away from the current breeding ranges of any North American monochromatic species) and a confusing result centered around a population of Eastern Spot-billed Ducks in northeast Asia still leave questions open:

In summary, existing data do not support a single explanation for paraphyly of mtDNA haplotypes in Mallards. Instead, a combination of evolutionary processes, including historical and contemporary range expansion, hybridization, and incomplete lineage sorting have likely shaped the current distribution mtDNA haplotype lineages in Mallards and closely related species inhabiting both Asia and North America. Kulikova et al. 2005

To split or not to split?

Despite unanswered questions, the best available evidence indicates that North America’s three monochromatic populations (Mexican, Mottled, Black) did not split off from green-headed Mallards but instead came into contact with that taxon later, and that gene flow then ensued.

This means that treating Mexican Duck as a subspecies of Mallard while elevating American Black Duck and Mottled Duck is probably inconsistent. I say “probably” because paraphyly — reticulation — does happen, so it’s conceivable for populations A and B to diverge, for C to then diverge from B, and then for A and B to remain compatible and even come back together, leaving C a distinct population.

However, it’s not at all clear that this is what has happened or is happening in this case.

But another potential problem is that significant genetic work has never been done on the Mexican Duck; the 2001 McCracken study sampled only four Mexican Ducks from one location in San Luis Potosí.

I can understand why bodies like the AOU might want further clarity in what’s obviously a very complex situation before making a change, but it seems pretty clear that leaving Mexican Duck lumped with Mallard does not reflect the best available information on the complex. Treating the Mexican Duck as a valid species — for now — is a good move by the IOC.


American Ornithologists’ Union (AOU) (1998): Check-list of North American Birds (7th edition). American Ornithologists’ Union, Washington, DC.

IOC (2009): The Mexican Duck.

Kulikova, Irina V.; Drovetski, S.V.; Gibson, D.D.; Harrigan, R.J.; Rohwer, S.; Sorenson, Michael D.; Winker, K.; Zhuravlev, Yury N. & McCracken, Kevin G. (2005): Phylogeography of the Mallard (Anas platyrhynchos): Hybridization, dispersal, and lineage sorting contribute to complex geographic structure. Auk 122 (3): 949-965. PDF

McCracken, Kevin G.; Johnson, William P. & Sheldon, Frederick H. (2001): Molecular population genetics, phylogeography, and conservation biology of the mottled duck (Anas fulvigula). Conservation Genetics 2 (2): 87–102. PDF

Webster, Richard E.: The Status of Mottled Duck (Anas fulvigula) in Arizona. Arizona Field Ornithologists.

Related posts:

  1. Ducks, ducks, ducks
10 Responses
  1. January 31, 2010

    Fascinating. As you did when we discussed this issue, you manage to make a complex distinction at least somewhat intelligible!

  2. Lisa Berger permalink
    January 31, 2010

    Using Hubbard’s analytical approach, I’m poised to publish a treatise on historical paraphyly of the starling and new world meadowlarks. A brief summary follows:

    Both starlings and meadowlarks originally derive from a common ancestor. Mitrochondrial DNA studies indicate the new world founder population resided in what is now Alaska and the northwest United States, and was probably introduced approximately 18,000 years ago by hominids who rapidly populated the west coast from Alaska to South America.

    My interest was first piqued by observation of phenotypical similarities, including but not limited to stocky build, long stout bill, waddling style of foraging, remige position that produces airplane-like gliding, short neck, and short retrices. This led to my current genetic research.

    Originally homozygous, multiple nucleotide polymorphism in a polynucleotide sequence of closely associated genes that affect bill and plumage color, produced dimorphic phenotypes resulting in plumage similar to each modern day species. The probably shift in genetic translation may have been spontaneous or the result of a wide variety of pressures on the species, including epigenetic gene methylamine as a result of soot contaminants from volcanic activity or coercion-encompassing fires set by Paleolithic hominids to drive herds of game over cliffs, by cosmic radiation, or by glaciations and climatic shift.

    Allopathic speciation began to occur when the northern-most part of the polyphonic population was fairly rapidly selected against due to dark coloration as the glaciers advanced and to reproductive stress from shorter breeding season and lowered food resources. The dimorphic phenotype better succeeded with its mottled plumage in a habitat not fully covered in ice and snow.

    For approximately 5,000 years, the populations grew more divergent. The starling probably became extinct in the new world during this time, although recessively homozygous remnant populations of starlings may still exist today. The European Starling was reintroduced to the eastern U.S. in the late 1870s and again in 1890, and has since spread across the continent.

    It will be interesting to observe how the two species interact and if speciation will be maintained or will merge. On the basis of my research, I propose that more study be devoted to these closely related species, which will answer the question: Are all birds that appear to be European Starlings, in fact, European Starlings, or are some individuals, especially those resident populations in the southwest U.S. and Mexico, black meadowlarks?

    I just couldn’t resist ;)

    • David J. Ringer permalink*
      January 31, 2010

      Great Scott! You might be on to something!

  3. Lisa Berger permalink
    January 31, 2010


    Your article is really a fine distillation of research about speciation of the Mallard clade. Keep up your great work. Articles like yours challenge folks to find out more, and to expand their knowledge and understanding of nature.

    Good Stuff!

  4. Steven Mlodinow permalink
    February 4, 2010

    Most interesting. Thanks for putting the information “out there” for us to read.

    When I was living in southern CA and birding much in southeastern AZ, I had made friends with a local hunter/rancher/researcher. He had done a bit of everything. Anyway, I’d not necessarily take his words, recounted by me 20+ years later, as gospel, but he said that at some point not long (a decade??? or two???) before the species were lumped, a number of Mallards were released into the breeding range of Mexican Duck in Arizona. Per what my friend said, this did lead to some hybridization, and that led to the “Mallards and Mexican Ducks interbreed freely” statement.

    My bottom line is that if such is true, an unnatural occurrence as described might lead to a temporary situation (in this case, hybridization) that does not reflect any long term reality.

    Steven Mlodinow

    • David J. Ringer permalink*
      February 4, 2010

      Hi Steven,

      You’re welcome, and thanks for the comment.

      Yes, the papers mention the fact that deliberate release of Mallards has probably clouded the picture in various ways. And it’s also clear that Mallards now breed in areas they didn’t before human intervention (e.g., Florida, and parks all over the continent), which has all sort of implications.

      I hadn’t heard that particular anecdote — interesting.

  5. February 6, 2010

    Great post! In 27 years of birding on the border I have seen very few Mallards and never one during the breeding season. A more interesting question is whether Mottled Ducks and Mexican Ducks interbreed. They are sympatric here in the Rio Grande Valley and not many birders(myself included) identify them properly. During the bird festival this fall, Jon Dunn identified both species in the same flock.

    • David J. Ringer permalink*
      February 6, 2010

      Excellent point, Dan. I remember studying a pair of ducks at Bentsen and wondering how in the world to tell whether they were Mottleds or Mexicans…. Are there distinguishing characteristics?

Trackbacks and Pingbacks

  1. AOU to leave Mexican Duck in limbo
  2. Mexican (Mallard) Duck | Birdspix

Comments are closed.