Proposal (825) to South American Classification Committee

Treat Sarkidiornis sylvicola as a separate species from Sarkidiornis melanotos

The American Comb-Duck, Sarkidiornis sylvicola has been treated since the 19th century (including Hellmayr and Peters [as Sarkidiornis carunculatus]) as a species apart from the "African" Duck Sarkidiornis melanotos until Delacour & Mayr (1945 Wilson Bull, 57: 3-55) treated as subspecies.

The reasons for this subordination are as follows (Delacour & Mayr 1945: 28):

"The Comb Duck (Sarkidiornis melanotos) includes two well-marked subspecies, one (melanotos) extending from Africa to south-east Asia…, the other (carunculatus) inhabiting South America. We have observed at Clères that the racial hybrids are not intermediate. In such hybrid broods some birds look like pure melanotos and others like pure carunculatus."

This subordination to the Old World taxon was adopted by Meyer de Schauensee (1966. The species of birds of South America), Blake (1977. Manual of Neotropical Birds), and the AOU (1998. Check-list of North American Birds), but not by Wetmore (1965. The Birds of the Republic of Panama), Kear (2005 Ducks, Geese and Swans), or Hoyo & Collar (2014. HBW and BirdLife International Illustrated Checklist of the Birds of the World).

In his phylogenetic classification and a general listing of taxa, Livezey (1997. Annals of Carnegie Museum 66 (4): 457-496) recognizes Sarkidiornis sylvicola as an independent species emphasizing the contrasting coloring of the sides and flanks, gray in S. melanotos and black in S. sylvicola.

The Delacour hybrids were obtained (artificially?) in his particular zoo, in Clères, Normandy, France. Generally, the resulting hybrids have intermediate parental characteristics. I cannot comment on the meaning of non-intermediate hybrids in BSC. However, this curious resilience of characters seems to favor the independence of these phenotypes. Regardless, hybridization in captivity is not a valid basis for considering two taxa to be conspecific under any modern version of the BSC.

There may be other diagnostic differences between sylvicola and melanotos. Apparently, sylvicola has smaller dimensions in both sexes (at least, on average) than melanotos. The calls (the species is basically mute) appear to be lower-pitched, with more bass, in melanotos; however, the sampling (Xeno-canto) is very small.

My vote is in support of the split, considering the well-marked differences between these two. See illustration below:

Fernando Pacheco, May 2019

Comments from Claramunt: “YES. Tough call. However, these birds are much more similar than what those illustrations suggest. Check the illustrations in the HBW instead (see below). Basically, the main difference is black versus grayish flanks. However, sylvicola is also smaller, and del Hoyo & Collar (2014) mentioned the shape of the comb, which seems slightly different, but a more detailed analysis would be desirable.

“That they hybridize in captivity is not evidence of potential free interbreeding in the wild. The statement about hybrids being similar to one or the other parent suggests that the main distinguishing character, the color of the flanks, is produced by a single Mendelian gene. However, the differences between the two taxa are not restricted to a single gene, as there are size differences. In addition, flank color (and maybe comb shape) may be involved in sexual selection and potentially species recognition. Taken together, I think that elevating sylvicola to species is reasonable, pending some falsifying evidence of reproductive compatibility or genomic homogeneity.”

sylvicola:

melanotos:

Comments from Stiles: “YES; as noted by Santiago, the original reason for lumping them was ill-founded.”

Comments from Robbins: “YES, for recognizing Sarkidiornis sylvicola as a species based on the rather dramatic morphological differences. As others have noted, captive hybridity is meaningless for assessing species limits, especially with regard to waterfowl.”

Comments from Zimmer: “YES”. As noted in the Proposal, and by the comments from others on the committee, hybridization in captivity, particularly with a notoriously promiscuous group like waterfowl, is meaningless in establishing species limits. The plumage differences are fairly dramatic, and there are accompanying mensural differences as well as likely differences in comb size and shape, all of which trumps the flimsy basis for lumping these taxa in the first place, in my opinion.”

Comments from Jaramillo: “YES – Particularly as waterfowl are abnormally uniform, not tending to show much geographic variation, other than in species that have culturally mitigated migration routes (geese).”

Comments from Remsen: “YES, but largely because the initial rationale for the lump was based on nearly irrelevant captive breeding. By the way, this one is screaming out for a genetic analysis not for classification but for estimating the age of the split. These two really do not seem to differ very much, phenotypically, thus suggesting a relatively recent split, i.e. transoceanic dispersal. I wish we had comparative information on displays and voice on which to evaluate this one in terms of species rank.”