PROPOSAL: that the broadly conceived species Chlorostilbon mellisugus (as per Zimmer 1950 and Schuchmann 1999 in HBW ­ the accounts are credited to Bündgen, but were written to Schuchmann's orders) be considered a superspecies, which (in South America) includes four allospecies, as discussed in Stiles (1996). Details of morphology, coloration and distribution are given in that paper, as well as a more thorough explanation of the reasoning. The allospecies are:

a) C. melanorhynchus (subspecies melanorhynchus and pumilus, though the latter might not be recognizable) of W Colombia and NW Ecuador: distinguishable from adjacent C. gibsoni by its shorter, all-black bill, relatively longer wing, shorter tail that is less forked, sex for sex. Males differ additionally in having a bright crown, females in having much green in the rectrices with some dark grey at the bases of the lateral rectrices.

b) C. gibsoni (subspecies gibsoni, nitens and chrysogaster ­ although the distinctness of nitens has been questioned) of N and C Colombia (Magdalena valley) and NW Venezuela (L. Maracaibo basin): tail longer and more forked than other SA forms in both sexes; differs from other SA forms (except olivaresi) in having at least basal half of mandible pink or red. Differs from melanorhynchus in having dull crown in male; females have extensive pale grey in bases of lateral rectrices; chrysogaster, the form that most closely approaches mellisugus, differs most from it in size (larger); tail much more deeply forked to (females) double-rounded.

c) C. mellisugus (subspecies mellisugus, caribaeus, duidae, subfurcatus, phaeopygos, napensis ­ if distinct from the preceding) of E Colombia S to N Bolivia, E through Venezuela and adjacent Brasil to the Guyanas: differs from the preceding forms in much less forked tail, sex for sex, lateral rectrices of females all blue except for narrow pale greyish tips (no green or basal grey); from gibsoni in all-black bill and (males) brighter crown more blue in breast.

d) C. olivaresi (monotypic), restricted to the Sierra de Chiribiquete of SE Colombia: much larger than all other forms; resembles gibsoni in bill color, pattern of rectrices of female, dull crown of male; resembles mellisugus in shallowly forked tail, sex for sex, and strongly bluish breast of males.

The two treatments of this complex mentioned above advocate subsuming all forms into mellisugus, and both share certain flaws. Zimmer (1950) starts out dividing the complex into "eastern" and "western" groups, ascribing chrysogaster to the former and nitens to the latter, then noting their great similarities and considering them (and hence both "groups", conspecific. Bündgen (1999) does much the same, starting out with "northern" and "southern" groups. My data suggest a rather more complex picture, in which variation simply does not break down along such neat geographic lines. The variation in color and morphology is more of a mosaic, rather than the smooth cline suggeested by Schuchmann: in particular, melanorhynchus and gibsoni break up the trend, and within gibsoni the trend reverses with chrysogaster having the most deeply forked tail; olivaresi fits no cisandean trend. Moreover, where the ranges of two forms abut, there is no hint of intermediacy or hybridization in any specimen I have seen to date (and I have examined a considerable number of additional specimens since 1996). No substantive new evidence has been published supporting a contrary view, hence I propose that we follow the arrangement I put forward in 1996.

References:
Bündgen, R. 1999. Species account of Chlorostilbon mellisugus. HBW vol. 5, p. 574.

Schuchmann, K.-L. 1999. Family Trochilidae. HBW vol. 5, pp.

Stiles, F. G. 1996. A new species of emerald hummingbird (Trochilidae: Chlorostilbon) from the Sierra de Chiribiquete, southeastern Colombia, with a review of the C. mellisugus complex. Wilson Bulletin 108:1-27.

Zimmer, J. T. 1950. Studies of Peruvian birds, no. 58: the genera Chlorostilbon, Thalurania, Hylocharis and Chrysuronia. American Museum Novitates no. 1474, pp. 1-12.

============

Comments from Remsen: I vote YES on this proposal, for reasons outlined here and in Stiles (1996). The only reason that the baseline list (from Dickinson 2003) started with a broad C. mellisugus is that I had unfortunately forgotten that Stiles (1996) had more in it than the new species description.

Comments from Nores: "Si estoy de acuerdo en separar Chlorostilbon melisugus en tres especies. Las razones dadas por Stiles son bastante convincentes, especialmente las diferencias morfológicas (como la cola mas furcada o menos furcada), y la coloración como el caso del pico o de la corona."