The Condor92:754-760 0 The Cooper Ornithological Society 1990

AN INTERGENERIC HYBRID (HELIODOXA LEADBEATERI X HELIANGELUS AMETHYSTICOLLIS) FROM NORTHERN COLOMBIA ’

GARY R. GRAVESAND RICHARDL. ZUSI Departmentof VertebrateZoology, National Museum of Natural History, SmithsonianInstitution, Washington, DC 20560

Abstract. We describe a new intergeneric hybrid hummingbird, Heliodoxaleadbeateri x Heliangelusamethysticollis, from the Eastern Cordillera of the Colombian Andes based on external characteristics.A hypothesis of parentage based on external characteristicsis supported by cranial features. Key words: Heliodoxa leadbeateri, Heliangelus amethysticollis; hummingbird;Andes; Colombia;Melbourne A. Carriker,Jr.; intergenerichybrid.

INTRODUCTION of in the National Mu- On 2 1 September 1946, Melbourne A. Caniker, seum of Natural History and the American Mu- Jr. collected an adult male hummingbird that he seum of Natural History. Color descriptionswere tentatively identified as “Heliangelus speciosa?” made under Examolites (Macbeth Corp.). Mea- at Buenos Aires, 6,050 ft (=1,845 m), 8”Ol’N, surements (wing chord, culmen from anterior ex- 72”58’W, Santander de1 Norte, on the eastern tension of feathers) were taken with digital cal- slope of the Eastern Cordillera of the Colombian ipers and rounded to the nearest 0.1 mm. Andes. The specimen (National Museum of Nat- Diagnostic assumptions and methods of hybrid ural History, USNM 392 14 1) was catalogedun- diagnosisbased on plumage charactersand mea- der this name but was never reported in the lit- surements follow Graves (1990). erature, despite the fact that H. speciosa(Salvin The skull of the hybrid was extracted by J. 1892, Graves 1990) was known only from the Phillip Angle using techniques outlined by Olson type specimen in the British Museum. et al. (1987). Zusi compared it with his unpub- During a recent analysis of enigmatic hum- lished data on skulls of speciesrepresenting 104 mingbird taxa from the Andes, Graves examined hummingbird genera,and directly with all species Carriker’s specimen and came to the conclusion, known to occur in the Eastern Cordillera of the on the basis of plumage characters, that it rep- Colombian Andes. The illustrations were drawn resented a previously unreported intergeneric using a dissectingmicroscope and drawing tube. hybrid between the Violet-fronted Brilliant (He- RESULTS liodoxa leadbeateri) and the Amethyst-throated Sunangel (Heliangelus amethysticollis). Because To our knowledge, Carriker’s specimen is the the genus Heliodoxa exhibits derived cranial only hybrid hummingbird from northwestern characters that might be used to test this hy- South America that was sexedinternally and ac- pothesis, we extracted the skull from the speci- companied by field notes (National Museum of men. Zusi’s analysis of the skull supported Natural History). Between 2 September and 3 Graves’ findings. In this paper, we present the October 1946, Carriker and his assistant col- combined analyses of the hybrid. lected 496 specimensin a mixture of second growth and remnant cloud forest separated by MATERIALS AND METHODS pasturesnear the settlement of Buenos Aires and Graves compared the plumage and other exter- on the ridge and mountain rising to the west (Alto nal features of the hybrid with study skins of all de Pozo). This collection (National Museum of Natural History) is composed of speciesthat oc- cur in midelevation (1,700-2,700 m) habitats I Received 19 January 1990. Final acceptance 29 in the Eastern Cordillera (see Hilty and Brown March 1990. 1986).

j7541 HYBRID HUMMINGBIRD 755

Once we concluded that the specimen was not pectoral bands. Coeligena torquata may be elim- a typical example of any describedhummingbird inated as a parental species because it has a taxon, we consideredthree options: the specimen strongly patterned tail, whereas the rectrices of represented(1) a rare genetic plumage variant of the hybrid are unmarked. Thus, H. amethysti- some Andean species;(2) a hybrid, or (3) a new collis is indicated as one of the parental species. species.Our analysesfocussed on the 50 + species However, this species has only a small green of hummingbirds that occur in the northern half frontlet at the base of the bill, not the large front- of the Eastern Cordillera above 1,000 m eleva- let of the Carriker specimen. Only two species tion, and especially on the 13 species of hum- with unpatterned tails could have contributed a mingbirds that Carriker collected near Buenos large frontlet and greenish gorget to the hybrid: Aires (Colibri thalassinus, C. coruscans, Chlo- H. leadbeateri and H. jacula. Heliodoxa jacula rostilbon stenura, Adelomyia melanogenys,He- can be eliminated tentatively because it occurs liodoxa leadbeateri, Coeligena coeligena, C. he- at lower elevations (SOO-1,500 m) and has not lianthea, Boissonneauaflavescens, Heliangelus been recorded in the Eastern Cordillera north of amethysticollis, Ocreatus underwoodii, Metal- Cundinamarca (Hilty and Brown 1986). Also, lura tyrianthina, Aglaiocercuskingi, and Acestru- males of H. jacula have a violet spot in the center ra heliodor). of the throat, and a H. jacula x H. amethysti- Although the specimen resembles H. ame- collis hybrid would probably have a small throat thysticollisin its pattern of white pectoral band spot of some shade of purple or pink. All other and glittering gorget,it differs significantly in pro- Eastern Cordilleran speciesare eliminated by two portions and color from that speciesand all oth- or more well-defined external charactersthat are ers. These charactersdemonstrate that the spec- not found in the Carriker specimen (data avail- imen is not a plumage variant or an undescribed able from the authors). race of any known speciesof hummingbird. Thus, plumage pattern and color indicate that Determining whether a unique specimen rep- the Carriker specimen is a hybrid between H. resentsa hybrid or a valid speciescan be difficult. leadbeateri and H. amethysticollis (Fig. 1). The Nevertheless, a number of Andean specieshave wing chord and culmen measurements (mm) fall been described from only one or two specimens within the range of Colombian samples of males (e.g., Graves 1988). In such cases, rejection of of the two parental speciesbut are closer to H. alternatives, especially hybridism, has been con- leadbeateri: leadbeateri (n = 15) wing chord, strued as proof of specific status. Our initial in- 67.3-72.2, culmen, 17.3-20.7; amethysticollis(n vestigations of potential hybridism were based = 15),wingchord, 67.5-71.8,culmen 14.8-16.1; on characters of the male plumage and on an hybrid, wing chord, 71.8, culmen, 18.2. assumption of polygenic inheritance of most A complete description of the external char- plumage characters(Graves 1990). acters of the hybrid and comparisons with its The specimen exhibits a number of plumage parental species follows. Descriptions of struc- characters,such as a brilliant gorgetand frontlet, tural colors are unusually subjective and actual that are restricted to the subfamily Trochilinae. color varies with the angle of inspection and di- The body plumage, remiges, rectrices, bill, and rection of light. For these reasonswe use general feet of the hybrid are relatively unmodified, lack- color terms. ing such elaborations as the tail rackets, tibia1 plumes, and emarginated primaries found in CAPITAL TRACT some trochiline hummingbirds. We assumedthat, In leadbeateri,an oblong, brilliant violet frontlet if the specimenwere a hybrid, the parental species extends from the posterior edge of the crown to would be similarly unmodified, and that they the base of the bill. The hindcrown and the feath- would possesscollectively the distinctive plum- ers bordering the frontlet are dark bronzy green. age charactersexhibited by the hybrid: (1) white The crown of amethysticollisis dark green with pectoral band; (2) large brilliant frontlet; and (3) a small brilliant green frontlet at the base of the a large, well-defined, brilliant gorget. No poten- bill. The crown of the hybrid is dark green with tial parental speciesexhibits all three characters. bronze highlights; a brilliant golden-green front- Among the Eastern Cordilleran species,only let extends from the center of the crown above two morphologically unspecialized species,Coe- the orbits to the base of the bill. Thus, the hy- ligena torquataand H. amethysticollis,have white brid’s frontlet is intermediate in size between 156 GARY R. GRAVES ANDRICHARD L. ZUSI

FIGURE 1. Lateral (A) and ventral (B) view of adult male specimens(from left to right) of Heliungelus amethysticollis,H. leadbeaterix H. amethysticollishybrid (USNM 39214 l), and Heliodoxaleadbeateri. leadbeateri and amethysticollis. The feathering on the dorsal base of the bill in amethysticollis VENTRAL TRACT extends rostrally to the level of the anterior limit The auriculars and the side of the neck of lead- of the external nares or slightly beyond (0.4 mm); beateri are dull green and contrast slightly with in leadbeateri it extends I 1.5 mm beyond; the the brilliant throat in indirect light. In amethys- hybrid is intermediate (1.1 mm). When viewed ticollis, the auriculars are dull blackish-green, be- head-on in direct light, the lores and the green coming greener on the side of the neck, and con- plumage adjacent to the frontlet of leadbeateri, trast sharply with the gorget.The auricular region amethysticollis, and the hybrid appear sooty of the hybrid is intermediate in color but resem- black. Both parental speciesand the hybrid have bles amethysticollis in contrasting sharply with a small white postocular spot. the gorget. The large brilliant green gorget of leadbeateri extends from the gular area between SPINAL TRACT the rami of the bill posteriorly to include the The back, scapulars,and rump of both parental upper breast. Contrast between the gorget and speciesand the hybrid are dark green. The dark surrounding plumage is indistinct except when bronze-green crown and hindneck of leadbeateri the former is seen to glitter (viewed head-on in contrastswith the back, whereascontrast in ame- direct light). The chin of amethysticollisis sooty thysticollisis variable but less pronounced than black, bordered posteriorly by a well-defined, in leadbeateri. The hybrid is intermediate. brilliant amethyst gorgetthat extends only to the HYBRID HUMMINGBIRD 157 lower throat. The gorgetof the hybrid resembles than in amethysticollis; those of the hybrid are amethysticollis in general configuration and intermediate in shape. feather shape, but resembles leadbeateri in ex- CAUDAL TRACT tending anteriorly to the gular area; the color is golden-green (same as the frontlet) in direct light The tail of leadbeateriis moderately forked. The but exhibits bluish-green reflections at some an- central rectricesare green with bronzy highlights; glesin indirect light, and thus is intermediate but the outer pairs of rectrices are progressively much closer to leadbeateri. blacker (the outermost is often entirely black). The lower breast and belly of leadbeateri are The color is similar in amethysticollisbut the tail green; the undertail coverts are dull green to is unforked or slightly forked. The tail of the bronze-green tipped with grayish-white or huffy- hybrid is colored like those of the parents. Its white. The gorget of amethysticollisis bordered shapeis uninformative becauseall rectriceshave posteriorly by a white band on the upper breast, basal sheaths. The central pair is almost fully and then a green band acrossthe breast (brilliant developed; the others are about half-grown. when viewed head-on). In indirect light, the low- er breast is dull green and doesnot contrast great- BILL ly with the sides. The feathers along the midline The bills of leadbeateri and amethysticollis are of the lower breast and belly are broadly edged essentially straight with a gradually decurving with buE, the undertail coverts are white or pale culmen and a correspondingly recurving gonys. gray with broad white margins. The hybrid is In amethysticollis the tip is slightly swollen be- intermediate; the gorget is bordered posteriorly cause the ventral outline of the mandibular ra- by a white pectoral band spotted with green mus in lateral view is slightly concave in contrast feather tips. When viewed head-on, the lower to the convex gonys; the culmen begins to de- breast is crossedby an indistinct band of golden- curve somewhat distal to the point where the green iridescence; in indirect light it is interme- gonys beginsto recurve. The tomium of the man- diate in color and pattern between the two pa- dible is also recurved near the tip. The bill tip rental species.The undertail coverts of the hy- in leadbeateri is more evenly attenuate, without brid are intermediate - bronzy-gray with white the appearance of a swollen tip. The lateral pro- margins. Downy white feathersare present in the file of the hybrid is more like leadbeateri in the vent area of the hybrid and both parental species. even attenuation of the bill tip. The tomia of The tibia1 feathers of all three forms have a both speciesand the hybrid are unserrated. In tufted appearance caused by long barbs with ra- the hybrid the ventral bars of the upper jaw and diating hairlike barbules. In amethysticollisthese their rhamphothecae were broken by shot an- feathers are gray with short white tips. In lead- terior to the nasal cavities, causing the bill to beateri and the hybrid, they are more extensively collapse inward and become much narrower in white, making the tufts more prominent. The ventral view than it was in life (Fig. 1). feathers that cover the distal end of the tibiotar- The rhamphotheca of the mandible of ame- sus and proximal end of the tarsometatarsusare thysticollisis two-toned in its proximal one-half short and structurally unmodified. These are to two-thirds, the ventral portion being blackish brownish-gray in amethysticollis;brownish-gray and sharply demarked from the dorsal yellowish with prominent white tips in leadbeateri, and portion. Most leadbeateri differ in being uni- intermediate-brownish-gray with inconspic- formly blackish, or dark brown and somewhat uous tips-in the hybrid. paler in the dorsal portion. The hybrid is inter- mediate, blackish below and whitish and some- ALAR TRACT what blotched with pale brownish on the dorsal The flight feathers of the parental species and portion. hybrid are similar in color. The wing coverts of The lack of rhamphothecal corrugationson the leadbeateri are greener and more iridescent than hybrid indicates that it is an adult. in amethysticollis;those of the hybrid are inter- mediate. The flight feathersof the parental species OSTEOLOGY are roughly similar in shape, but the inner pri- In an osteological test of Heliodoxa x Helian- maries of leadbeateri are slightly more pointed gelusparentage, the hybrid might be expectedto 758 GARY R. GRAVES ANDRICHARD L. ZUSI

1 cm

FIGURE 2. Diagnosticfeatures of the skull: (A) Heliodoxu Zeadbeuteri;(B) hybrid (USNM 392141); (C) Heliungelusamethysticollis. Left, lateral view; right, dorsalview; middle, vomer in dorsalview. Seetext for explanation.Features a-f areprobably derived in Heliodoxu (sensulato). Vomers enlarged 20% relativeto skulls.

exhibit a mosaic of the derived charactersof its example, the nasal cavities and conchae of parent genera, intermediacy of characters, and Schistes are largely unossified but those of He- the absenceof derived charactersof other genera. liangelus, Heliodoxa, and the hybrid are ossified. Several derived characters of the skull distin- The Carriker specimen exhibits a straight lateral guish Heliodoxa (sensu lato, including Clytolae- profile of the cerebellar prominence. Although ma, Sternoclyta, Polyplancta; Hylonympha not the braincase is damaged and somewhat dis- available). No charactersof the skull were found torted, this feature is not an artifact because its to be derived for Heliangelus. shape is maintained by the unbroken left half of A derived feature of Heliodoxa is the modified the braincase. lateral profile of the cerebellar prominence of the Several other charactersare probably derived braincase-relatively straight as opposed to for Heliodoxa. One is a marked anterior projec- evenly curved in other hummingbirds (Fig. 2- tion of the nasal surface of the ectethmoid (Fig. a). A profile like that of Heliodoxa is found also 2-b). A similar feature occurs in the unrelated in Patagona, whoserelationships are unresolved, genus Eutoxeres (Phaethornithinae) and is man- and in Schistes, which is not phylogenetically ifested in a less extreme form in many other close to Heliodoxa. Both genera can be elimi- genera including Heliangelus. The size of the an- nated as possible parents on other grounds. For terior projection of the ectethmoid in the hybrid HYBRID HUMMINGBIRD 759

is intermediate between those of Heliodoxa and gelus (Fig. 2-g). The hybrid also resemblesHe- Heliangelus. liangelus in that the lateral wall of the rostra1 Most species of Heliodoxa exhibit bilateral nasal concha is broadly concave between a pro- asymmetry of the dorsal bar of the upper jaw jecting roof and broad floor in lateral view (Fig. (Fig. 2-c) and of the adjacent bony roof of the 2-h). In leadbeaterithis lateral concavity is nar- vestibular and respiratory nasal cavities (Fig. 2- rower and reduced in length-a derived feature d). The asymmetry is associated with elongate in hummingbirds. The roof of the nasal cavities hyoid horns that pass on one side (right or left) is depressedbelow the ridge of the dorsal bar in of the dorsal bar and over the right or left nasal Heliodoxa and lies almost at the level of the ridge chambers to enter a sheath in the concavity of in Heliangelus (Fig. 2-i); in this the hybrid is the upper jaw (Zusi, unpubl.). The hyoids vary intermediate. It is also intermediate in the ori- in length among species such that asymmetry entation of the nasal surface of the ectethmoid ranges from none to well marked within the ge- (swept back in Heliodoxa and transverse in He- nus. Heliodoxa leadbeaterishows marked asym- liangelus; Fig. 2-j). The median spine of the metry. Asymmetry outside Heliodoxa is found vomer is long in Heliodoxa, short in Heliangelus, in Eugenes, Topaza, Chalybura, Heliomaster, and intermediate in the hybrid (Fig. 2-k). Ensifera, and Coeligena. Asymmetry of the dor- In summary, osteologyof the hybrid skull ex- sal bar and nasal roof is pronounced in leadbeat- hibits derived charactersof Heliodoxa, similar- eri and absent in Heliangelus. The hybrid is sym- ities to Heliangelus, and intermediacy between metrical (a slight swelling on the left side of the the two genera. It lacks derived characters of base of the dorsal bar may or may not represent other genera. These features support the hypoth- asymmetry related to displacement of elongate esis that a speciesof Heliodoxa is one parent of hyoid horns to the right); the tongue and hyoid the hybrid, and that a speciesof Heliangeluscould horns are, of course, missing. The anterior tips be the other. of the hyoid horns of amethysticollislie well short of the craniofacial hinge when the tongue is re- DISCUSSION tracted into the bill; those of the hybrid could The analysis of independent data sets is a pow- have extended forward somewhat beyond the erful but under-utilized tool in hybrid diagnoses. hinge without causingasymmetry. Lack of asym- In this case,the parental hypothesissuggested by metry is thus not an argument against leadbeateri plumage characterscould have been falsified by as a parent; it probably reflects intermediacy in skeletal characters.Had this occurred, we would length of the hyoid horns. have re-examined the diagnostic assumptions of Probably derived independently in Heliodoxa the analyses and considered alternate parentage is a truncate form of the body of the vomer from or other hypothesessuch as atavism (seeBuckley dorsal view in contrast to a smoothly tapered 1982, Graves 1990). form (Fig. 2-e). A truncate vomer is found also Problematic specimens of hummingbirds are in Amazilia, Chlorostilbon,their relatives, and a relatively common in museum collections, and few other genera. In Heliodoxa (and some Ama- many of those reported in the literature are re- zilia) the anterolateral angles of the vomer are garded as intergeneric hybrids (see examples in accentuated by anteriorly directed spikes. The Meyer de Schauensee1966). The high frequency vomer of the hybrid is intermediate in shapeand of “intergeneric” hybridization in humming- lacks the anterior spikes of Heliodoxa. Another was interpreted by Sibley (1957) and Short feature of Heliodoxa is an anteriorly forked bony and Phillips (1966) as evidence that the genera roof of the vestibular nasal cavity from dorsal are oversplit. Banks and Johnson (1961, p. 26) view (Fig. 2-l). A similar feature appears else- however, suggestedthat “lack of intrageneric hy- where in a variety of genera. In the hybrid, the bridization in . . . North American humming- anterior profile of the roof of the vestibular nasal birds may result from the perfection of intrage- cavity is forked as in Heliodoxa. None of the neric isolating mechanisms without the above charactersis unique to Heliodoxa, but in concurrent development of intergeneric isolating combination they are diagnostic of the genus. mechanisms.” However, the significance of hy- Among other features, the roof of the vestibu- bridization at different taxonomic levels can be lar nasal cavity of the hybrid is narrower and determined only after parentage has been iden- longer than in Heliodoxa, and similar to Helian- tified critically in each case, and after the phy- 160 GARY R. GRAVES AND RICHARD L. ZUSI logeny of hummingbirds has been determined Memorial Fund of the American Museum of Natural accurately. Examples of well-documented hybrid History (GRG) and the ResearchOpportunities Fund, Smithsonian Institution (GRG, RLZ). analyses are rare, and investigations of phylog- eny are still in the formative stages(e.g., Zusi LITERATURE CITED and Bentz 1982; Zusi, unpubl.). Among carefully BANKS,R. C., AND N. K. JOHNSON.1961. A review analyzed hybrids, we think those described by of North American hybrid hummingbirds. Con- Banks and Johnson (196 1) and Short and Phillips dor 63:3-28. (1966) among the “genera” Archilochus, Calypte, BERLIOZ, J. 1932. Notes critiques sur quelquesTro- Stellula, and Se1asphoru.s represent hybridiza- chilides du British Museum. Oiseau 2:530-534. BERLIOZ, J. 1938. Notes critiques sur des Trochilidts. tion between very closely related species.By con- Oiseau 8:3-19. trast, the hybrids Eugenes x Cynanthus (Short BUCKLEY,P. A. 1982. Avian genetics,p. 21-110. In and Phillips 1966) and Heliodoxa x Heliangelus M. Petrak [ed.], Diseasesof cageand aviary birds. involve hybridization between species in distinct 2nd ed. Lea and Febiger, Philadelphia, PA. GRAVES,G. R. 1988. Phylloscarteslanyoni, a new genera that are not sister taxa. No hybrids have species of bristle-tyrant (Tyrannidae) from the been reported between species from different lower CaucaValley of Colombia. Wilson Bull. 100: subfamilies (Phaethomithinae, Trochilinae). Only 529-534. one (Ruschi 1944) has been reported between GRAVES,G. R. 1990. Systematics of the “green- “primitive” and “advanced” trochilines (see Zusi throated sunangels”(Aves: Trochilidae):valid taxa or hybrids? Proc. Biol. Sot. Wash. 103:6-25. and Bentz 1982), and one (Berlioz 1932, 1938) GRAY,A. P. 1958. Bird hybrids. Commonwealth Ag- between the small, gorgeted hummingbirds (last ricultural Bureaux, Bucks, England. 19 genera of Peters [ 19451) and other groups (data HILTY, S. L., AND W. L. BROWN. 1986. A guide to from Gray 1958, Meyer de Schauensee 1966). It the birds of Colombia. Princeton Univ. Press, Princeton, NJ. is premature to estimate the frequency of hy- MEYER DE SCHAUENSEE,R. M. 1966. The speciesof bridization at different taxonomic levels or the birds of South America. Livingstone Press, Nar- implications of such patterns. beth, PA. BISON, S. L., J. P. ANGLE, F. v. GRADY, AND H. F. ACKNOWLEDGMENTS JAMES. 1987. A technique for salvaginganatom- ical material from study skins of rare or extinct We thank R. C. Banks, L. C. Binford, R. Bleiweiss, birds. Auk 104:510-512. and K. C. Parkes for helpful comments on the manu- Pamas, J. 1945. Check-list of birds of the world. Vol. script and the curators and staff of the American Mu- 5. Museum of Comparative Zoology, Cambridge, seum of Natural History, Pierce Brodkorb Collection, MA. California Academy of Sciences,Carnegie Museum of Ruscm, A. 1944. Novo case de hibrido entre OStro- Natural History, Delaware Museum of Natural His- chilideos: Thalurania glaucopis x Melanotrochi- tory, Field Museum of Natural History, Florida Mu- lusfuscus (Trochilidae, Aves). Bol. Mus. Nat. Rio seum of Natural History (University of Florida), In- de J. Zool. 241-4. stitut Royal de SciencesNaturelles de Belgique, Los SALVIN,0. 1892. Catalogue of birds in the British Angeles County Museum, Moore Laboratory of Zo- Museum. Vol. 16. ology (Occidental College), Museum of Comparative SIBLEY,C. G. 1957. The evolutionary and taxonomic Zoology (Harvard University), Museum of Natural significanceof sexualdimorphism and hybridiza- History (University of Kansas), Museum of Natural tion in birds. Condor 59: 166-l 9 1. Sciences(Louisiana State University), Museum of Ver- SHORT, L. L., ANDA. R. PHILLIPS. 1966. More hybrid tebrate Zoology (University of California, Berkeley), hummingbirds from the United States. Auk 83: Museum of Zooloav (Universitv of Michigan). Pea- 253-265. body Museum of Natural History (Yale U&&ity), ZUSI, R. L., AND G. D. BENTZ. 1982. Variation of a Royal Ontario Museum, and San Diego Museum of muscle in hummingbirds and swifts and its sys- Natural History for the loan of specimens.Final figures tematic imolications. Proc. Biol. Sot. Wash. 95: were inked by Deborah Roney. Financial support for 412-420. - museumwork was provided by the Frank M. Chapman