- PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON 116(2):293-300. 2003. Diagnoses of hybrid hummingbirds (Aves: Trochihdae). 10. Cyanomyia salvini Brewster, 1893, is an intergeneric hybrid of Amazilia violiceps and Cynanthus latirostris Gary R. Graves Department of Systematic Biology, MRC-1 16, National Museum of Natural History, P.O. Box 37012, Smithsonian Institution, Washington, D.C. 20013-7012, U.S.A. — Abstract. Cyanomyia salvini Brewster, 1893, collected in Sonora, Mexico, is shown to be a hybrid between Amazilia violiceps ellioti and Cynanthus latirostris magicus, whose breeding ranges overlap extensively in northwestern Mexico. This specimen represents the only known intergeneric hybrid between species currently placed in Amazilia and Cynanthus (Sibley and Monroe 1990). & The unique type of Cyanomyia salvini dorsed by Weller Schuchmann (1997) Brewster, 1893, was collected by John C. and Weller (1999), but neither of these ref- Gaboon at Nichosari, Sonora, Mexico, on erences provided corroborating evidence. 31 March 1887. Early references treat C. As a consequence, the taxonomic status of salvini as a valid species (Boucard 1895, Cyanomyia salvini is still in doubt. Here I Ridgway 1911, Gory 1918, Simon 1921), provide a taxonomic assessment of Cyano- although only Ridgway's entry indicated a myia salvini employing the methods and as- personal examination of the specimen. sumptions outlined in Graves (1990) and & Griscom (1934:378) proposed a hybrid or- Graves Zusi (1990), as modified by the C igin for salvini in his revision of Ama- findings of Graves (1998, 1999b). zilia violiceps: Methods "By inference I had always doubted the existence of another species of this genus [Amazilia] in So- The type of Cyanomyia salvini, (J. C. Ga- nora. Geographically and faunally there is no basis boon field number 505), originally part of for one, and the failure to duplicate the type in over the William Brewster Gollection (No. 40 years has further significance, and in part at least strengthens this view. My late esteemed colleague 24,124), was eventually cataloged in the Outram Bangs always supposed that salvini was of Museum of Gomparative Zoology, Harvard A hybrid origin. careful study ofthe colorand struc- University (received in 1918, No. 224,124). tural characters of the type convinces me that Cy- The type was sexed as 6 on the Brewster anomyia salvini Brewster is a hybrid betweenAma- Gollection label and appears to be in defin- zilia violiceps conjuncta [= Amazilia violiceps el- lioti] and Cynanthus latirostris Swainson." itive plumage as judged by the absence of striations on the maxillary ramphotheca, the Griscom's brief description was insufficient absence of distinctive buffy feather tips on to make a convincing case for hybridiza- the dorsal plumage, and the presence of a tion, but Peters (1945) and Phillips (1964) strongly iridescent coronal patch. Descrip- cited Griscom's treatment without substan- tions in this paper refer to definitive male A tive comment. second hypothesis was in- plumage. I compared the type of Cyano- troduced in a succinct footnote by Fried- myia salvini with specimen series of Calyp- mann et al. (1950), who suggested that Cy- te annae, C costa, Selasphorusplatycercus, anomyia salvini is an aberrant example of S. rufus, S. sasin, Stelhila calliope. Archil Amazilia violiceps. This idea was later en- ochus alexandri, Calothorax luciferi. He- 294 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON — Table 1. Ranges (mean ± standard deviation) of measurements (mm) of wing chord, bill length, and rectrix length (Rl—R5) ofadult males ofAmazHia violiceps ellioti, Cynanthus latirostris magicus and a probable hybrid, Amazilia violiceps ellioti X Cynanthus latirostris magicus (= type of Cyanomyia salvini Brewster, 1893; MCZ 224, 124). Amazilia Cynanthus violiceps latirostris Cyanomyia Character N = 15-17 N = 14-15 salvini Wing 53.5-56.1 48.4- 51.9 52.9 (54.6 ± 0.7) (50.4 d: 0.8) Bill 20.1-22.5 18.7-21.6 21.4 (21.4 ± 0.6) (20.3 d: 0.8) Rl 26.3-29.3 22.4-25.3 27.7 (28.2 ± 0.8) (23.8 ^ 0.7) : R2 28.0-30.5 23.9-27.0 28.0 (29.4 ± 0.9) (25.8 d 0.7) : R3 28.3-31.4 26.3-29.4 29.1 (30.1 ± 1.0) (28.6 d 0.8) : R4 28.3-32.5 28.7-32.4 30.6 (30.7 ± 1.1) (30.8 d 1.0) : R5 28.8-32.6 31.0- 33.8 31.2 (30.7 ± 1.2) (32.5 d 0.7) : liomaster constantii, Eugenesfulgens, Lam- tension of feathers); and rectrix length pornis clemenciae, Amazilia beryllina, A. (from point of insertion of the central rec- violiceps (including the type of Amazilia trices to the tip of each rectrix) (Table 1). MCZ violiceps conjuncta Griscom, 1934; Pairs of rectrices are numbered from the in- No. 224,112), Hylocharis leucotis, and Cy- nermost (Rl) to the outermost (R5). Scatter nanthus latirostris magicus, all of which plots of measurements and least squares re- occur in Sonora, Mexico (Friedmann et al. gression lines were used to illustrate size 1950, Howell & Webb 1995), in the collec- differences among specimens. tions of the Museum of Comparative Zo- General color descriptions presented in ology. Because the generic allocation of Appendix 1 were made under natural light. species traditionally placed in Amazilia by I evaluated the color of the medial vane of mm Peters (1945) is in flux, I use the species the dorsal surface of Rl (7 from tip) & taxonomy of Sibley Monroe (1990). De- with a calibrated colorimeter (CR-221 tailed descriptions and photographs of the Chroma Meter, Minolta Corporation) mm type of Cyanomyia salvini were compared equipped with a 3.0 aperture. The mea- with series of the aforementioned species in suring head of the CR-221 uses 45° circum- the National Museum of Natural History, ferential illumination. Light from the pulsed Smithsonian Institution, and with a sup- xenon arc lamp is projected onto the spec- posed immature specimen of Cyanomyia imen surface by optical fibers arranged in a salvini collected at Palmerlee, Cochise circle around the measurement axis to pro- County, Arizona, on 5 July 1905 (Bishop vide diffuse, even lighting over the mea- 1906). The latter specimen (Field Museum suring area. Only light reflected perpendic- = of Natural History 160,998; wing chord ular to the specimen surface is collected for 53.4 mm; bill length = 21.2; R4 = 29.7; color analysis. Colorimetric data from iri- R5 = 29.9) appears to be a female Amazilia descent feathers are acutely dependent on violiceps and will not be further discussed. the angle of measurement, the curvature of Measurements were taken with digital plumage surfaces in museum skins, and the calipers and rounded to the nearest 0.1 mm: degree of pressure applied to the plumage wing chord; bill length (from anterior ex- surface by the Chroma Meter aperture. In VOLUME NUMBER 116, 2 295 Fig. 1. Lateral views of males in definitive plumage: Amazilia violiceps ellioti (top), Cynanthus latirostris magicus (bottom), and a probable hybrid. Amazilia violiceps ellioti X Cynanthus latirostris magicus (= type of MCZ Cyanomyia salvini Brewster, 1893; 224,124). order to reduce measurement variation, I red (fl), and blue-yellow {b). The rationale held the aperture flush with the rectrix sur- is that a color cannot be perceived as red face without depressing it. The default set- and green or yellow and blue at the same ting for the CR-221 Chroma Meter displays time. Therefore "redness" and "green- mean values derived from three sequential, ness" can be expressed as a single value, in situ measurements. I repeated this pro- a, which is coded as positive if the color is cedure twice (five times for the type of Cy- red and negative if the color is green. Like- anomyia salvini), removing the aperture be- wise, "yellowness" or "blueness" is ex- tween trials. Thus, each datum summarized pressed by b for yellows and —b for blues. in Table 2 represents the mean of 6 (paren- The third coordinate, L, ranging from to tal species) or 15 (type of C. salvini) in- 100, describes the "lightness" of color; low dependent colorimetric measurements. values are dark, high values are light. The Colorimetric characters were described in more light reflected from the plumage, the terms of opponent-color coordinates (L, a, higher the L value will be. Visual systems & & b) (Hunter Harold 1987). This system is in hummingbirds (e.g.. Goldsmith Gold- based on the hypothesis that signals from smith 1979) differ significantly from those the cone receptors in the human eye are of humans and the relevance of opponent coded by the brain as dark-light (L), green- color coordinates to colors perceived by — 296 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON hummingbirds is unknown. In any case, the L, a, b color system permits plumage color to be unambiguously characterized for tax- onomic purposes. Results and Discussion I considered three hypotheses proposed by previous authors Cyanomyia salvini represents (1) a hybrid, Amazilia violiceps X Cynanthus latirostris, (2) an aberrant ex- ample of Amazilia violiceps, or (3) a valid species. For brevity I use the epithet, sal- vini, in the remainder of the paper. I found no evidence that salvini repre- sents a subdefinitive plumage or geographic variant of any known taxon. The possibility that salvini represents an aberrant plumage of Amazilia violiceps can be rejected be- cause salvini has substantially shorter wings. All evidence is consistent with the hypothesis that salvini represents an inter- generic hybrid, Amazilia violiceps X Cy- nanthus latirostris. Several characters of salvini facilitate the identification of its pa- rental species (Appendix 1): (a) brilliant bluish-purple crown; (b) white chin, throat, midline of breast and abdomen white; (c) greenish-blue subterminal spots or bars on white feathers at the lateral margins of the throat and upper breast; and (d) absence of rufous or buff pigmentation on the second- aries or rectrices (Fig. 1). Here I present a synopsis of the critical steps of the hybrid diagnosis. The pool of potential parental species may be quickly narrowed by focusing on the white ventral plumage ofsalvini. Among the potential pa- rental species that occur in Sonora, Mexico, only Amazilia violiceps possesses white ventral plumage from chin to undertail co- 48 50 52 54 56 58 verts. Bluish-green spotting on the lateral WING margins of the chin, throat, and breast feathers of salvini were inherited from the Fig. 2. Bivariate plots of measurements (see Table other parental species. Both Amazilia ber- 1): Amazilia violiceps ellioti (hollow triangle), Cynan- yllina and Cynanthus latirostris have green thus latirostris magicus (A), and a probable hybrid, or bluish-green plumage from chin to upper (i^) Amazilia violiceps ellioti X Cynanthus latirostris magicus (= type ofCyanomyia salvini Brewster, 1893; breast. Because brown or reddish-brown MCZ 224,124). pigments appear to exhibit consistent pen- VOLUME NUMBER 116, 2 297 etrance in hummingbird hybrids (Banks & ^ ON & Johnson 1961, Graves Newfield 1996), Q 0^0 ^00 in A. beryllina can be eliminated from further +1 +1 +1 p consideration because the secondaries and a s d -^ rectrices of salvini lack buff or rufous pig- ment. Only one pair of species (A. violiceps ^ ^ I^J X C. latirostris) could have contributed the -j:5 u ^ _= >< to unique combination of characters exhibited QQ 1 by salvini (Appendix 1). 3 Hybridization of Amazilia violiceps with -a u any of the small gorgeted species (i.e., Ca- ^ o s lypte annae, C. costa, Selasphorus platy- 1 cercus, S. rufus, S. sasin, Stellula calliope, Archilochus alexandri, Calothorax lucifer) would likely produce offspring with one to 9. ^ Q —^ —'t many iridescent gorget feathers (Graves & O 0<3 1+^/11 +1 +1 s-^o Zusi 1990; Graves 1996, 1999a). In a sim- Sjj dr~- O—N '1 ilar fashion, Lampornis clemenciae (bril- TO3 liant gorget and large white tail spots), Eu- a C genes fulgens (brilliant gorget and black a SX:3 r(nN ^'^^ breast and abdomen), Hylocharis leucotis (white postocular stripe and black chin and 1 § 8^ auriculars), and Heliomaster constantii —r- d<T\ (brilliant gorget and semi-concealed white S 2 I 1 1 rump patch) are unlikely to be a parental species because they each possess a suite of plumage characters not observed, even as Q ON r-; C/5 1—( 1—1 traces, in salvini. +1 +|+'S As a second step, I tested the restrictive c5 00 ON Cvl hypothesis with an examination of size and O K? 1 O(Nn —vd external proportions (Fig. 2). Measurements .>2 '^S^ xXjcj of trochiline hybrids fall within the men- ~~! ^ X (N (N 'J d sural ranges exhibited by their parental spe- 1 (mN (N c cies as a consequence of a polygenic mode & of inheritance (Banks Johnson 1961). +1 O d (N ^ Measurements of Amazilia violiceps ellioti S-i S in r—^j (N and Cynanthus latirostris overlap in four of the seven characters and the percent differ- ence in character means is modest (larger O'-n^ '-^^ ^ species divided by smaller): wing chord Cd H II ^ ^ (8.3%), bill length (5.4%), Rl (18.5%), R2 (14.0%), R3 (5.2%), R4 (0.3%), and R5 (5.9%). Measurements of salvini fall at or cd bleestswteheann tthhee mcheaarnactvearluemefaonrsR4(oorf1A..0 vmioml- •5 -^5 ^ ^;^> 'g!'"^o^ "-K^2 iceps) of the postulated parental species, I ^'2 -C5^ .^NJ uRaneedcs,trpiirnxedscieocvlteoerrdalvbacylauselesesa,isntapssapqlruvoaixrniiemsfaatrlleegbrteehsteswiveoaenln.- 3^ IU§ y<NN •^.~"s^^V^^.;^J U-'C?^^fs*;«:-!^'U>"*~^vs'>^j:i< — 298 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON 4.5 sey, Alison Pirie, .and Jeremiah Trimble A A (Museum of Comparative Zoology, Har- CwO A A c 3.0 vard University) for permission to study the -^o 1.5 A type of Cyanomyia salvini, David Hafner (Occidental College) for specimen loans, ^A^ ^ Tom Schulenberg for tracking down the 0.0 A w Bishop Collection specimen, and John CO A^ Bates (Field Museum of Natural History) i-1.5 1^ A CD for lending it. Travel was supported by the 60)-3.0 Alexander Wetmore Fund, Smithsonian In- 10 15 20 25 30 35 stitution. Literature Cited CO (O c Banks, R. C, & N. K. Johnson. 1961. A review of — North American hybrid hummingbirds. Con- dor 63:3-28. ^ Bishop, L. B. 1906. Uranomitra salvini in Arizona.^ Auk 23:337-338. Boucard, A. 1893. Genera of humming birds, part 2. ((00 Published by the author, London. <cD Brewster, W. 1893. Description ofa new hummingbird — 0) from northern Mexico. Auk 10:214-215. m Cory, C. B. 1918. Catalogue of birds ofthe Americas, — 15 20 25 30 35 part 2, No.l. Field Museum of Natural His- tory Zoological Series 13:1-315. Lightness (L) Friedmann, H., L. Griscom, & R. T. Moore. 1950. Dis- tributional check-list of the birds of Mexico, Fig. 3. Bivariate relationships of L, a, b color co- part 1. Pacific Coast Avifauna No. 29, 202 pp. ordinates: Amazilia violiceps ellioti (hollow triangle), Goldsmith, T. H., & K. M. Goldsmith. 1979. Discrim- Cynanthus latirostris magicus (A), and a probable hy- ination of colors by the black-chinned hum- brid {ik), Amazilia violiceps ellioti X Cynanthus lati- — mingbird, Archilochus alexandri. Journal of rostris magicus (= type of Cyanomyia salvini Brew- Comparative Physiology A 130:209-220. MCZ ster, 1893; 224,124). Graves, G. R. 1990. Systematics of the "green-throat- ed sunangels" (Aves: Trochilidae): valid taxaor — the character means of the postulated pa- hybrids? Proceedings of the Biological Soci- ety of Washington 103:6-25. rental species (Table 2, Fig. 3). 1996. Diagnoses of hybrid hummingbirds In summary, evidence obtained from . (Aves: Trochilidae). 2. Hy—brid origin of Erioc- plumage color and pattern, as well as from nemis soderstromi Butler. Proceedings of the external size and shape, is consistent with Biological Society ofWashington 109:764-769. the hypothesis that Cyanomyia salvini is an . 1998. Diagnoses of hybrid hummingbirds intrageneric hybrid between Amazilia viol- (Aves: Trochilidae). 6. An intergeneric hybrid, Aglaiocercus kingi X Metallura tyrianthina, iceps ellioti and Cynanthus latirostris mag- from Venezuela.—Proceedings of the Biologi- icus, whose breeding ranges overlap exten- cal Society of Washington 111:511-520. sively in northwestern Mexico. Cyanomyia . 1999a. Diagnoses of hybrid hummingbirds salvini Brewster, 1893, is thus available in (Aves: Trochilidae). 7. Probable pa—rentage of taxonomy only for the purposes of hom- Calliphlox iridescens Gould 1860. Proceed- ings of the Biological Society of Washington onymy. 112:443-450. 1999b. Diagnoses of hybrid hummingbirds . Acknowledgments (Aves: Trochilidae). 8. A provisional hypothesis for the hybrid origin of Zodalia glyceria — I thank Richard Banks and Richard Zusi (Gould, 1858). Proceedings of the Biological for reviewing the manuscript, Doug Cau- Society of Washington 112:491-502. VOLUME NUMBER 116, 2 299 -, & N. L. Newfield. 1996. Diagnoses of hybrid in direct light. The remainder ofthe dorsum from hind- hummingbirds (Aves: Trochilidae). 1. Charac- neck to the rectrices is grayish-olive; the transition be- terization of Calypte anna X Stellula calliope tween the purple crown and olive hindneck is abrupt. and the possible effe—cts of egg volume on hy- The back, shoulders, upper tail coverts, and central bridization potential. Proceedings of the Bio- rectrices are faintly glossed with silvery-green when logical Society of Washington 109:755-763. viewed head-on. The rectrices are unmarked. -, & R. L. Zusi. 1990. An intergeneric hybrid Feathers of the forecrown and crown of latirostris hummingbird {Heliodoxa leadbeateri X Helian- (to a point immediately posterior of the eyes) are dark geliis amethysticollis) from northern Colom- green, margined with bronze giving the forehead and bia.—Condor 92:754-760. frontal part of the crown a dull appearance. The re- Griscom, L—. 1934. The ornithology of Guerrero, Mex- mainder of the dorsum from hindcrown to rump is ico. Bulletin of the Museum of Comparative glossy green (showing scattered bluish-green, brilliant- Zoology 75:368-422. ly iridescent feather barbs when viewed head-on). Up- Howell, S. N. G. & S. Webb. 1995. The birds ofMex- pertail coverts are darker, contrasting slightly with the ico and northern Central America. Oxford Uni- green rump and the bluish-black rectrices. The outer versity Press, Oxford, UK, 851 pp. two or three pairs of rectrices (R3, R4, R5) are nar- Hunter, R. S., & R. W. Harold. 1987. The measurement rowly tipped with gray. of appearance, 2nd edition. Wiley, New York, The crown of salvini is intermediate in appearance 411 pp. between that of violiceps and latirostris, showing a Peters, J. 1945. Check-list of birds of the world, vol. mixture of deep blue, purplish-blue iridescence, be- 5. Museum of Comparative Zoology, Cam- coming greener toward the posterior of the coronal bridge, Massachusetts, 306 pp. area. The intensity of iridescence increases posteriorly Phillips, A. R. 1964.—Notas sistematicas sobre aves from the forecrown to center of the crown (viewed Mexicanas, III. Revista de la Sociedad Mexi- head-on). This bluish-green iridescence at the posterior cana de Historia Natural 25:217-242. edge of the coronal area blends into deep bluish-green Ridgway, R. 1911. Birds of North and Middle Amer- on the hindneck and back, changing to dull green on ica. Bulletin of the United States National Mu- the lower back and rump. The back and scapulars of seum 50, part 5. salvini appear bluer than those of either parental spe- Sibley, C. G., & B. L. Monroe, Jr. 1990. Distribution cies. This seems to be another example of the "blu- and taxonomy of birds of the world. Yale Uni- ing" phenomenon observed in some hummingbird hy- versity Press, New Haven, Connecticut, 1111 brids (Graves 1998, 1999b). The intensity of irides- pp. cence on the lower back of salvini is intermediate be- Simon, E. 1921. Histoire naturelle des Trochilidae tween that observed in the postulated parental species. (synopsis et catalogue). Encyclopedia Roret, L. There is no appreciable contrast between uppertail co- Mulo, Paris. verts and rectrices in salvini. The rectrices are inter- Weller, A.-A. 1999. Violet-crowned Hummingbird, mediate in color between those of latirostris and viol- AElglyirotttr,ia&viJo.licSeaprsg.ataPl.,5e9d9s..inHJa.nddeblooHkoyoof, tAh.e iarceepwso(rTnabalned2a).ppTehaertooutbeerrreetcatirniceedsf(rRo4m,aRs5u)bdoeffisnailtviivnei Birds of the World, vol. 5. Barn-owls to Hum- plumage. Rl and R2 are fresh and unworn, whereas m,in&gbKi.r-dLs.. SLcyhnuxchEdmiacnino.ns,19B9a7r.ceTlhoenah,yb7r5i9dpopr.i- ftheeatthieprsofofRs3alivsinsiliagrhetliyntweorrmne.diWaitenginccoovleorrtsanadnddefglirgehet gin of a Venezuelan trochilid, Amazilia distans Wetmore & Phelps 1956.—Ornithologia Neo- of melanism to those of violiceps and latirostris. The ventral plumage of violiceps is snowy white tropical 8:107-112. from chin to undertail coverts. Feathers on the chin, throat, and upper breast are pure white to the base, Appendix with only a few pale gray barbs at the base of the 1 rachises. Comparative description of selected characters of The chin and upper throat of latirostris are deep adult male Amazilia violiceps ellioti, Cynanthus lati- purple (extending laterally to the auriculars and eye- rostris magicus, and a probable hybrid, A. violiceps ring), blending into dark bluish-green on the lower ellioti X C. latirostris magicus (= Cyanomyia salvini throat; this latter color continuing posteriorly to the Brewster; MCZ 224,125). Descriptions of structural vent. A few feathers in the chin are fringed with white. colors are unusually subjective, as color seen by the Feather disks from chin to vent exhibit iridescent high- observer varies according to the angle of inspection lights when viewed head-on. Vent feathers are white. and direction of light. For this reason I use general Undertail coverts oflatirostris are gray, paling to white color descriptions. at the distal margins (most coverts are gray along the The forecrown and crown of violiceps exhibit bril- rachis to the tip). Some shorter coverts have an oval liant bluish-purple iridescence when viewed head-on gray subterminal spot. 300 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON The chin and throat of salvini are white but basal present in violiceps in the form of grayish-olive sides, feather barbs are much grayer than in violiceps. The with scattered weakly iridescent, pale bluish-green auriculars are iridescent purple and greenish-purple feather disks. However, the medial extension ofdarker (posteriorly). Feathers at the sides of the throat have feather tips observed in violiceps never reaches the ex- purple (anterior) or purplish-green (posteriorly) subter- tent observed in salvini. The few undertail coverts re- minal bars, imparting a spotted appearance to the sides maining on the specimen of salvini are pure white. of the throat. This spotting coalesces on the sides of The bill of violiceps is dull grayish-yellow (red in the breast to form an incomplete pectoral band formed life) tipped with dark brown or blackish-brown (<15% of white feathers with large subterminal greenish-blue of the bill length). Bill color in latirostris is similar or greenish-purple disks. The sides of salvini are dark but the dark tip is more extensive. The pattern of pig- bluish-green (about the same color as in latirostris). mentation in salvini is intermediate ofthat observed in The tendency toward an incomplete pectoral band is adult male specimens of violiceps and latirostris.