Every year around this time, birders anxiously await the decisions of the American Ornithological Society (AOS) Classification Committee. This committee makes decisions on taxonomic issues for North American birds, including splitting (dividing one species into two or more species) and lumping (grouping two or more species into one species), renaming or reclassifying species, listing established exotics, etc. Any changes are adopted by the American Birding Association (ABA), the “official” listing authority for the North American birding community.
This year, the Mexican Duck is up for reconsideration. In the ABA area, this duck is found in Texas, New Mexico, and Arizona. It is currently recognized by the AOS and ABA as a Mallard subspecies. Indeed, Mexican Ducks look nearly identical to Mallards, with one important exception: adult males don’t have the glossy green head, white neck ring, chestnut chest, gray flanks, and black rump that we see in male Mallards; all Mexican Ducks look like female Mallards. Given this prominent difference, the taxonomy of this bird has proven controversial, and it is still recognized as a separate species by some taxonomic authorities.
History of Mexican Duck Classification
The Mexican Duck was first described as a distinct species in 1886 by Robert Ridgway, the first full-time curator of the United States National Museum (now known as the Smithsonian Institute). This designation lasted for nearly a century. However, in the sixth edition of Clements Checklist of Birds of the World, in 1983, the Mexican Duck was lumped with Mallard and treated as a subspecies. This reclassification was largely based on work by John P. Hubbard, New Mexico Department of Fish and Game, published in 1977. Hubbard showed that in the southern part of their range (central Mexico), the plumage of Mexican Ducks was distinct from Mallards, but in the northern part (southern United States), their plumage was more similar to Mallards. These data were interpreted as evidence for “extensive hybridization in southeastern Arizona, southern New Mexico and west-central Texas” (American Ornithologists Union 1983), which warranted lumping the Mexican Duck with the Mallard. Notably, however, full species status was retained by some taxonomic authorities, such as the International Ornithological Congress.
This conflicting taxonomy, and the corresponding controversy, has persisted. In 2010, a proposal to reinstate species status for the Mexican Duck was submitted to the Classification Committee. This proposal reflected a general shift in the scientific world regarding what constitutes a species, which included the recognition of regular and natural occurrence of hybridization among many species (recent genomic data suggest that hybridization is the norm not the exception). The proposal rightly argued that Mallards hybridized extensively with several Mallard-like species, including the American Black Duck, Mottled Duck, Hawaiian Duck, and (Russian) Spot-billed Duck. Hybridization between Mallards and Mexican Ducks, the extent of which remained unknown, was not a compelling reason to treat them as the same species. Furthermore, my friend, colleague and mentor, Kevin McCracken, published evidence from mitochondrial DNA (a “special” type of DNA inherited only from mothers) suggesting that Mexican Ducks are more closely related to American Black Ducks and Mottled Ducks rather than Mallards. Despite this evidence, the Classification Committee rejected the reclassification with a vote of 5 (yes) to 6 (no), retaining the Mexican Duck as a subspecies of the Mallard.
In 2018, the controversy grew. A new proposal was submitted to the Classification Committee arguing that the entire Mallard Complex experienced high rates of hybridization and that Mexican Ducks were no less different from Mallards than other members of the complex, and for consistency, Mexican Duck should be elevated to full species status. The proposal to split Mexican Duck from Mallard was again rejected with a vote of 3 (yes) to 8 (no). Further complicating the taxonomy, Clements Checklist went ahead and elevated Mexican Duck to full species, breaking from the tradition of adhering to the AOS classification. Because eBird taxonomy follows Clements, this decision resulted in an addition to the life lists of many eBird users, despite the species not being recognized by the AOS or the ABA.
Given the controversy surrounding the species status of Mexican Ducks, the proposal to split them from Mallards has resurfaced again. New data has come to light, which confirms the fact that the Mexican Duck situation is no different than the situation in Black Ducks and Mottled Ducks. In addition, recent DNA evidence suggests that hybridization is probably not as rampant as we once thought and that there are likely mechanisms that maintain the Mexican Duck as a distinct biological entity despite hybridization. AOS’s decision to split these species will ultimately depend on the interpretation of available DNA evidence by the Classification Committee.
Recent DNA Evidence
For the past decade or so, the genome of Mexican Ducks and other Mallard-like ducks has been investigated by Philip Lavretsky (University of Texas El Paso). His research began at Wright State University as a doctoral student in my lab. Here, I will highlight a few of our most important findings that will be considered by the AOS Classification Committee.
In 2014, we published a paper based on 18 DNA markers showing weak differences among all Mallard-like ducks. Notably, however, Mexican Ducks were less similar to Mallards than were American Black Ducks (see phylogenetic tree below). Yet, Black Ducks are recognized as distinct species whereas Mexican Ducks are not.
As part of his PhD research, Dr. Lavretsky examined over 3,000 DNA markers for Mexican Ducks and Mallards. These data were published in 2015 in the journal Molecular Ecology. Consistent with the plumage work by John Hubbard (see above), we found that Mexican Ducks from Texas and New Mexico were genetically more similar to Mallards than were Mexican Ducks from Mexico. Rather than finding evidence for rampant hybridization, however, we surprisingly found, three distinct populations of Mexican Ducks – southern U.S., Sonora (west coast of Mexico) and interior Mexico. We found only a single individual out of 105 samples that had compelling evidence of hybridization — one Mexican Duck from the southern U.S. appeared to have about 25% Mallard genes. This frequency of hybridization is much lower than what was found in other studies for Black Ducks, Mottled Ducks, and Hawaiian Ducks.
In 2019 we published a paper in ‘Molecular Ecology’ that compared Mexican Ducks, Mallards, Black Ducks, and Mottled Ducks at more than 3000 DNA markers. Like our earlier work, we again found that Mallards and Black Ducks were genetically more similar to each other than were Mallards and Mexican Ducks. We also found that the sex-chromosomes (similar to the XX vs. XY chromosomes in humans) were quite different between Mallards and the other Mallard-like ducks. Finally, we provided additional evidence that hybridization was not as rampant as previously believed.
These data support some level of separation between Mallards and Mexican Ducks, but is it enough to warrant splitting them into different species? It is important to recognize that speciation is a process, not an event, and it can take millions of years for complete reproductive isolation to evolve. There exists a continuum between populations of a single species and fully differentiated species, and many taxonomic pairs fall within this continuum (the gray zone of speciation). There is no rule as to how long this process will take or how different the pairs need to be to warrant the recognition of different species. Indeed, any hard “line” would be subjective. So, in my opinion, the best we can do is apply some sort of consistency for these “gray zone” cases.
The Mexican Duck is firmly within this “gray zone.” It can be distinguished from the Mallard in both plumage and DNA, and there is limited evidence of “extensive hybridization”. Although recognizable, the genetic differences between Mallards and Mexican Ducks are weak, at best, suggesting that they may have only recently split from a common ancestor. Importantly, however, these weak differences are at the same level as the differences between Mallards, Black Ducks and Mottled Ducks, each of which is recognized as different species. If we want to apply consistency to our designation of species, then we really only have two options: (1) Treat the Mexican Duck as a full species separate from Mallards, or (2) lump Mallards, Mexican Ducks, Black Ducks, and Mottled Ducks (and perhaps Hawaiian Duck, Spot-billed Duck, and Philippine Duck) as a single species. I, for one, anxiously look forward to the decision of the AOS Classification Committee on this treatment.