Proposal (229) to South American Classification Committee
Split Coccycua from Coccyzus
Background: The traditional classification of New World Cuculinae consists of four genera, in the following sequence:
I'm not sure if any explicit rationale has ever been published to justify this traditional linear sequence. The species membership in both genera reflects plumage similarities, and has remained fairly stable for a century or so. Ridgway's (1916) classification of North American species is essentially the same as that of AOU (1998) and SACC in terms of generic limits, with the exception that he retained our Piaya minuta in the monotypic genus Coccycua. See Sorenson & Payne (2005) [let me know if you need pdf] for a synopsis of work on the group this century, none of which had sufficient taxon-sampling or resolution to force a change in classification.
New information: Sorenson & Payne (2005) analyzed mitochondrial (ND2) DNA and ribosomal RNA (12S) sequences (1000+ bp each) to produce a phylogeny for the family as a whole, with outstanding taxon-sampling. A summary of the relevant findings for our Coccyzinae (from Fig. 5.5) is as follows:
1. The species of the two genera above, plus Caribbean Hyetornis and Saurothera, comprise a fairly strongly supported group (87% bootstrap).
2. Piaya is not monophyletic if minuta is included; the latter forms a monophyletic group (96% bootstrap) with two South American species ("Coccyzus" pumilus and "C." cinereus). Sorenson & Payne resurrect Coccycua for this group. Coccycua is basal in the group.
3. Extralimital Saurothera and Hyetornis are embedded within the remaining Coccyzus; they form a monophyletic group (95%) but are more closely related to C. erythropthalmus and South American C. lansbergi than the latter are to other Coccyzus, e.g., C. americana. Therefore, to maintain a monophyletic Coccyzus (100% bootstrap; with pumilus and cinereus removed), they merged both genera into Coccyzus.
Sorenson & Payne's (2005) tree (Fig. 5.5) would produce the following linear classification (by convention, with basal taxa first, and with taxa arranged roughly from north to south and west to east within sister taxa). I did this independently from Sorenson & Payne's linear sequence in their table of contents and came up with exactly the same sequence:
Analysis: As noted by Sorenson & Payne, their new groupings also make some "sense" to a degree in terms of plumage and natural history. The species in Coccycua share similarities in bill shape and small body size. Ridgway (1916) placed pumilus and cinereus in their own genus, Micrococcyx, based on wing shape. Hyetornis and Saurothera have vocalizations and plumage similar to true Coccyzus, and their large body size and more rounded wings reflect patterns in morphological evolution of insular birds in general. The only surprise to me was expulsion of minuta from Piaya, but after fiddling around with the skins in our synoptic series, it seems fairly clear that chestnut plumage "comes and goes" throughout this group of cuckoos as a whole, which seem to have repeated plumage themes that reappear among unrelated species and groups.
Recommendation: Sorenson & Payne's data set and analysis are clearly superior to any previous "data" and should be reflected in our classification and linear sequence, and I strongly recommend a YES vote to change our sequence to the one above.
Sorenson, M.D. & R.B. Payne. 2005. Molecular systematics: cuckoo phylogeny inferred from mitochondrial DNA sequences. Pp. 68-94 in R.B. Payne. Bird Families of the World: Cuckoos. Oxford University Press.
Aug. 2006 (in consultation with Bob Payne)
Note: This proposal is a slight modification of the one to AOU/NACC. It was accepted by NACC - see Banks et al. (2006).
Comments from Stiles: "YES. The analyses look solid. Differences in wing shape between minuta and the other two species could simply reflect sedentariness in the former, migratory behavior in the latter."
Comments from Robbins: "YES. Sorenson & Payne (2005) provide solid genetic data for resurrecting Coccycua for minuta and its two close relatives."
Comments from Zimmer: "YES. Evidence seems strong. Despite plumage similarities of minuta to Piaya, it is clearly more divergent vocally from cayana and melanogaster than either of those species is from one another."
Comments from Jaramillo: "Yes and NO - Yes, split Coccycua from Coccyzus, but No on the actual membership of Coccycua. The issue is more than just wing shape, or chestnut in the plumage, but I need more convincing that minuta is not a Piaya. The bill morphology of minuta to me is much more like Piaya, than Coccycua. Small size may link these three species, but minuta is small for a Piaya, as a Coccyzus it actually would be rather a large one. So really small size links pumila and cinerea, but not minuta. The green bill of minuta is unlike all other cuckoos in this group except Piaya cayana. The highly graduated tail of minuta is like Piaya or Coccyzus, while pumila and cinerea have much less strongly graduated tails. Vocally, pumila and cinerea outwardly sound quite different, but the repetitive nature of the songs and the tempo is similar. The explosive nature of some Piaya calls (the chip-churr call of cayana for example) seems to be matched by a call of minuta on the Hardy tapes (assuming correct identification). There is also a purring call attributed to minuta on Xeno-canto that sounds similar to that of Piaya melanogaster. The common 'chwip" call to me sounds allied to calls of Piaya cayana rather than Coccycua (I know cinerea pretty well from Argentina, but pumila only from recordings). This is more from the gut than anything else, but I do want to raise the question of whether the analysis could have mis-represented the relationship of minuta? Maybe others more familiar with the methodology could comment. I just think that pumila and cinerea "make sense" to me, but not minuta, not at all." [Note from Remsen -- I'll record this as a NO vote.]
Comments from Pacheco: [NO] "Yes" to accept pumilus and cinereus apart from Coccyzus, but "no" considering strictly the title of this present proposal. The type-species of genus Coccycua is Cuculus monachus = Coccyzus minutus [Little Cuckoo]. More recently, conclusions of Hughes (2006. Systematics and Biodiversity 4: 483-488) support the resurrection of Micrococcyx to incorporate the Ash-coloured and Dwarf cuckoos, and the monotypic Coccycua for the little cuckoo."
Comments from Bob Payne: "Here are comparisons of the trees in Hughes (2006) and Payne (2005). Our tree is a complete species taxa sample of cuckoos, it is not limited by choice of a single (questionably appropriate) outgroup taxon to root the tree, and it has much higher bootstrap values, especially for the clade/non-clade in question, Coccycua. I ran the Hughes 2006 paper past Mike Sorenson, and he agrees the outgroup problem is the major problem with that paper."
"Hughes (2006) presented a tree of molecular phylogeny of certain cuckoos. The tree did not recover a clade that included Coccycua with minuta and cinerea, distinct from the other New World cuculine cuckoos. Otherwise, the tree is consistent with the estimate of cuckoo phylogeny in Sorenson & Payne (in Payne 2005). The following compares the taxon sampling and bootstrap support for the clade in question, and indicates stronger support for the tree in Sorenson & Payne (2005).
"Taxa: 18 species are recognized in New World Cuculinae (Payne 2005). All 18 species were sequenced in Sorenson & Payne (2005). Basal to the other species was a clade composed of minuta, pumila, and cinerea; the clade had 96% bootstrap value, and the genus Coccycua Vieillot 1816 was recognized for the clade. Within this clade, minuta was basal, and the sister clade with pumila and cinerea had 100% bootstrap support, this sister clade corresponds to "Micrococcyx" Ridgway 1912. The taxa most closely related to these New World cuckoos in the mitochondrial phylogeny were the Old World malkohas (Phaenicophaeus and 5 other genera) and crested cuckoos (Clamator).
"Hughes 2006 included only 11 of the 18 species taxa of New World cuckoos, with a single outgroup taxon, Cuculus, to root the phylogeny. Hughes 2006 tree showed 10 of these New World cuckoos (including minuta, excluding only cinerea) to be a clade. This clade (excluding cinerea) had a bootstrap support of only 69%. The consensus tree did not show cinerea and the 10 New World cuckoos to be more closely related to each other than cinerea was to Cuculus - the 10 species, C. cinerea, and Cuculus formed an unresolved three-branch polytomy.
"Hughes 2006 tree with 11 taxa was otherwise consistent with Sorenson & Payne 2005 tree for the 18 New World taxa. Both indicated the "Coccyzus" of Peters 1940 was polyphyletic.
"Numerical comparison of the trees: the bootstrap support value for a 3-species Coccycua clade was 96% in Sorenson & Payne 2005. The bootstrap support for a clade that included Coccycua minuta and excluded cinerea was much lower, 69% (Hughes 2006), and the latter tree did not sample the species pumila.
"Hughes 2006 tree had fewer taxa, it did not include the closest relatives of the New World Coccyzus group: the malkohas and the crested cuckoos. The choice of a single, distantly related cuckoo (Cuculus) probably affected the rooting of Hughes 2006 tree; another outgroup might have recovered the three-species Coccycua clade. The lower bootstrap results indicate lower support for the Hughes 2006 tree than for Sorenson & Payne 2005 tree.
"Hughes (2006) proposed recognition of another genus, Micrococcyx Ridgway 1912, based on the lack of recovered clade that includes minuta and cinerea. Hughes did not compare the other pumila, which is generally thought to be closely related to cinerea.
"Hughes & Baker 1999 ("Phylogenetic relationships of the enigmatic hoatzin (Opisthocomus)" Molecular Biology & Evolution, 1999, 16:1300-1307) reported a genetic phylogeny of several cuckoo taxa, including "Chalcites," Cacomantis, Crotophaga, Guira, and Neomorphus, using the same mitochondrial genes as Hughes 2000. Why were these not included in Hughes 2006, as outgroups? And, Hughes 2003, Naturwissenschaften 90:231-233 reported molecular phylogenies in other New World cuckoos, using the same genes, but these were not compared in Hughes 2006 to the Coccyzus group. The choice of outgroup taxa (a single taxon) is of greatest concern in interpreting Hughes 2006.
"Also, Hughes & Baker (1999) was based on certain erroneous nucleotide sequences (as evaluated in Sorenson et al., 2003, "More taxa, more characters: the hoatzin problem is still unresolved." Mol. Biol. Evol. 20:1484-1499; not cited in Hughes 2006, where Hughes & Baker 1999 was cited).
"Elsewhere, Hughes 2000 had a tree based on morphological characters that showed Coccyzus most closely related to Clamator (as in Sorenson & Payne 2005). Coccyzus was less closely related by numbers of branches to Cuculus (or to Tapera, as in an earlier ms version of Hughes 2006). Hughes 2000 (a morphological comparison) listed Clamator as sister genus to Coccyzus, but Hughes 2006 did not compare genes of Clamator.
"The morphological characterization for the three-species Coccycua genus in Payne (2005) was: "Small New World cuckoos with a rounded to graduated tail. The type, by monotypy, is Coccyzus minutus Vieillot (Peters 1940), which has rufous plumage and short rounded wings. Wing shape is rounded in Coccycua minuta, intermediate in C. pumila and pointed in C. cinerea, although not as pointed as in most Coccyzus cuckoos." Partly this follows the descriptions and comparisons in Ridgway (1912). Hughes 2000 did not report on the morphology of pumila or cinerea.
"In summary, the phylogenetic estimate and the systematic treatment of cuckoos in Payne (2005) have better support than the estimate and systematics in Hughes (2006). Both agree that the broad-sense genus Coccyzus of Peters is polyphyletic, and both agree that the large lizard-cuckoos "Saurothera" of the West Indies are within the monophyletic narrow-sense genus Coccyzus."
Comments from Schulenberg: "YES, although I don't think our work is done. I agree with Fernando that what we probably are looking at is a monotypic Coccycua (minuta), and another genus for pumilo + cinerea. I haven't seen Hughes 2006, and if Payne and Sorenson are right, then maybe I needn't bother. But I don't see that the Sorenson and Payne phylogeny is inconsistent with a two-genus "solution" for these three species either."