sign in sign up
<<
Home , Accipitridae, Aviceda, Black eagle, Butastur, Buteo, Buteogallus

AMERICAN MUSEUM Norntates PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, N.Y. 10024 Number 2741, pp. 1-20, figs. 1-4, table 1 July 30, 1982 A Revision of the Sub-Buteonine (, Aves)

DEAN AMADON1

ABSTRACT This paper is a taxonomic review of the 25 . Changes from usual treatment include rec- species and approximately 10 genera of chiefly ognition ofthe Asturina, merger ofthe genus Neotropical hawks called sub-buteonines and al- Heterospizias with , and transfer ofthe lied to the more advanced genus . Generic genus Geranospiza to the sub-buteonines. Finally, diagnoses supported by logarithmic ratio diagrams the broad systematics of the chief components of of measurements are presented along with com- the Accipitridae and the place of the sub- ments on intraspecific variation in a few of the buteonine group within it are discussed. INTRODUCTION The term "sub-buteonines" is here used Eight of the 10 sub-buteonine genera here for a group ofhawks and closely allied recognized are Neotropical. Three of the to the large and nearly cosmopolitan genus species in as many genera, cross the United Buteo. As noted later, certain other genera or States border. Two of them, Asturina nitida groups ofgenera may be regarded as sub-bu- and Buteogallus anthracinus, have distribu- teonines in a more general sense, but they are tions that are primarily tropical and subtrop- less closely allied to Buteo and beyond the ical. The third, unicinctus, extends main scope ofthe present paper. As the name a little farther north to Kansas and farther implies and as defined below the sub-bu- south to central . One of the two Old teonines, though often specialized in various World genera, Kaupifalco, is African, the ways are on the whole more primitive than other, , African and Oriental, with Buteo; they represent offshoots or continua- one species ranging north in summer to Ja- tions ofthe stock from which Buteo evolved. pan. By way of contrast, Buteo is well rep- Indeed, the group merges more or less insen- resented in the Holarctic, Neotropical, and sibly with the more primitive species ofButeo Ethiopian regions, but virtually absent from and some authors have included certain sub- the Oriental and entirely so from the Austra- buteonines, notably Asturina and Parabuteo, lian regions. The most primitive species of in that genus. Buteo, such as magnirostris, are Neotropical. ' Lamont Curator of , Emeritus, Department of Omithology, American Museum of Natural History.

Copyright ©) American Museum of Natural History 1982 ISSN 0003-0082 / Price $1.85 2 AMERICAN MUSEUM NOVITATES NO. 2741

Unlike some species ofsub-buteonines, none figures of features usually used in generic di- ofthe is characteristic oflowland rain agnoses-, foot, wing, and tail. All the forest. The 10 genera ofsub-buteonines have species of Accipitridae are illustrated in a total of 26 species; Buteo alone has almost Brown and Amadon (1968). Even those fa- the same number-24. miliar with all or most of the taxa involved Of the 10 genera of sub-buteonines no may find these paintings useful in visualizing fewer than six are monotypic and another the characters of the genera and species. two, and Butastur, may with Many of the genera and species of birds, more or less uncertainty, be termed what I especially large conspicuous ones such as (Amadon, 1968) have called "quasi-mono- hawks, were described by Linnaeus and by typic," that is, comprised of a single super- those who shortly followed him. At that time species. While the number of genera seems little was known of relationships and generic excessive, the sub-buteonines are a rather diagnoses were skimpy and often misleading. diverse lot and additional information is Still, the synonymies given in such relatively needed before contemplating further changes. early works as Sharpe (1874) show that the The principal changes made herein are to rec- conclusions of early students as to the affin- ognize Asturina (often regarded as a ities ofvarious genera and species were some- of Buteo), to merge Heterospizias with Bu- times sounder than those expressed decades teogallus, and to transfer Geranospiza to the later. sub-buteonines. The phylogeny of the Accipitridae will in There has been relatively little recent dis- the course of time be increasingly dependent cussion of the status and interrelationships upon the record. Birds ofprey, because of the genera of the family Accipitridae. of their large size and because some were Brown and Amadon (1968) outlined a clas- engulfed at sites such as the Rancho La Brea sification for the Accipitridae and diagnosed tarpits, are not so scarce as . Nonethe- the genera in somewhat general terms. This less, fossils contribute relatively little as yet paper may be regarded as a more formal to our understanding of the sub-buteonines. statement of part of that classification which Occasionally, in fact, faulty assignment of is set forth with original citations and ranges fossils has confused the issue (Amadon, for species and in a checklist of 1963). A complete list of all known fossil the by the late Prof. E. hawks has been provided by Brodkorb Stresemann and the writer which forms part (1964). of volume 1, revised edition, of "Peters Genera of birds are difficult to define; one Check- of the World" (Strese- reason why one finds so few recent ornitho- mann and Amadon, 1979). Original citations logical publications in which there is any at- and other technical details may also be found tempt to do so. This state of affairs is in part in Peters (1931), Friedmann (1950), Hell- the result of the well-known structural ho- mayr and Conover (1949), and Jollie mogeneity of birds. On the other hand, to (1976-1977). make a point that is seldom emphasized, a Dropping back to a somewhat earlier era genus ofbirds may be comparable in geologic one must mention especially Ridgway (1873, age, in degree of ecological uniqueness, and 1876) who in his earlier years was particularly in every other respect except overt morpho- interested in raptorial birds. Since his papers logical characters, to genera of , have been largely forgotten and are scarce, or other groups. The generic diag- I have quoted from them rather freely. Some noses indicate the size and proportions of of Ridgway's conclusions were incorporated each genus; a more detailed analysis of mea- in Friedmann (1950), that volume being a surements follows in a separate section. continuation of Ridgway's "Birds of North A general discussion ofthe sub-buteonines and Middle America" but the geographical and of their place in the family Accipitridae limitations of this work and its general ap- is postponed until the genera have been dealt proach restricted the treatment of genera as with individually. Inasmuch as Buteo is con- such. Nevertheless, it is an important refer- sidered a later, more derived genus it would ence and, for the genera treated, includes line seem logical to define the sub-buteonines and 1982 AMADON: SUB-BUTEONINE HAWKS 3 then compare Buteo with them. This is re- Color phases rare. Outer primaries less versed here, first because the sub-buteonines sharply notched on inner vanes than in Bu- are the main thrust ofthis paper, and second teo; often more than four outer primaries because they are a group of several genera, with emarginations on inner vanes thus sug- as against the single genus Buteo. gesting that they are in general less skilled at I am indebted to Drs. Wesley E. Lanyon soaring than the species of Buteo. Legs usu- and Lester L. Short for reading the entire ally long, coarse, unfeathered; scalation usu- manuscript and for numerous valuable sug- ally as in Buteo but reticulate rather than scu- gestions. The late Dr. Eugene Eisenmann also tellate in Harpyhaliaetus and partially so in read and criticized the entire paper. Many Butastur and Kaupifalco. In general rather others provided useful assistance on one sluggish; several ofthe species feed primarily point or another, among them Dr. Tom J. on cold-blooded or upon Cade, Dr. and Mrs. Frederick N. Hamer- and large centipedes or other . strom, Dr. Ned K. Johnson and Dr. Helmut Most sub-buteonines are tropical or subtrop- Sick. I am grateful to all of them. ical in distribution, where, according to species, they are found in a variety of habi- COMPARISON OF BUTEO WITH tats, including lowland rain forest. SUB-BUTEONINES GENUS BUTEO LACEPEDE, 1799 THE SUB-BUTEONINE GENERA TYPE: Falco buteo Linnaeus. GENUS BUTEOGALLUS LESSON, DIAGNOSIS: Medium-sized to large hawks, 1830 usually of robust form. Bill of moderate size TYPE: Falco aequinoctialis Gmelin. and normal shape. Head fully feathered. SYNONYMs: Urubitinga Lafresnaye, 1842; Wings rather long, longer than tail, broad and type, Falco urubitinga Linnaeus. Hypo- rounded. Outer three or four primaries morphnus Cabanis, 1844, same type. Hetero- notched on inner vane. Tail of moderate spizias Sharpe, 1874, type Falco meridional- length, usually slightly rounded. Tarsus of is Latham. There is dispute as to whether variable length; feathered for at least one- the generic name Urubitinga was properly third of its length on anterior surface. Tibial introduced, so Hypomorphnus has been used usually long, forming "flags." Adults by some authors; now considered a synonym usually barred or monocolored below; im- of Buteogallus, the problem becomes aca- matures usually streaked. Melanistic and er- demic. ythristic phases present in a high proportion RANGE: Southwestern United States (Texas ofspecies. Habitat plains, semi-open or tem- to Utah and Arizona) south to central Ar- perate woodlands, never the interior oftrop- gentina; also Cuba and the Lesser Antilles. ical rain forest. The food of buteos is usually SPECIES: 1- aequinoctialis; 2-subtilis; 3- warm-blooded vertebrates, especially ro- anthracinus; 4- urubitinga; 5- meridionalis. dents, but some species consume reptiles and DIAGNOSIS: Heavily built rather sluggish . For a list ofthe species and their hawks, with long coarse legs and relatively ranges, see Stresemann and Amadon (1979). short, broad rounded wings, the secondaries DIAGNOSIS OF SUB-BUTEONINES: Small to unusually long; tail relatively short (the -sized hawks; form usually robust, oc- species meridionalis of somewhat different casionally slender. Wings typically broad, proportions). Face and even forehead often rounded, with long broad secondary quills. scantily feathered or with bristles only; four Tail variable in length; often rather short, or five primaries weakly emarginate on inner usually with a conspicuous white cross band. vanes. Bill rather long and weak. Adults color and pattern extremely vari- blackish or rufous barred with dusky, always able; young usually different from adult and with a white tail band. Immatures very dif- typically coarsely streaked on breast and ab- ferent; coarsely streaked below; tail with nu- domen. Face and sometimes forehead often merous dusky bars. The species of Buteo- sparsely feathered, bristly, or naked. Flank gallus are partial to wetlands, swampy woods feathers usually short, not forming "flags." or even seacoasts, and feed extensively on 4 AMERICAN MUSEUM NOVITATES NO. 2741 crabs, cold-blooded vertebrates, or occasion- plumage, though lightly barred, is quite un- ally young birds or . like the coarsely streaked immature of all REMARKS: In some early literature (e.g., species of Buteogallus. Sharpe, 1874) Buteogallus was limited to the Turning now to "Heterospizias" meridio- type species, aequinoctialis, whereas the two nalis, the Savanna , here regarded as "black hawks" anthracinus (including sub- belonging to Buteogallus, the species has had tilis) and urubitinga were placed in Urubi- a checkered taxonomic history. First de- tinga. In doing so Sharpe cited as the basis scribed noncommittally as Falco meridio- for such separation not the dissimilar col- nalis by Latham in 1790 it was placed in a oration or denuded forehead and face of the monotypic genus Heterospizias by Sharpe Rufous -hawk (aequinoctialis) but in- (1874, p. 158), who merely separated it off stead insignificant details of proportion. in a key without giving generic characters of Ridgway (1876, p. 141) commented as fol- significance. As a genus it is related to Bu- lows: teogallus, , Asturina, and Para- buteo. It Buteogallus [is] extremely similar to Urubitinga, agrees with Buteogallus in having the main difference being in the coloration; but a coarsely streaked immature plumage, in the differing somewhat also in the relative propor- rufous, finely barred color pattern, and in tion of the lateral toes and the degree of devel- being primarily a of wetlands or low opment of the loral bristles. The latter are al- savanna. The resemblance is so great that I most entirely wanting .... here place it in Buteogallus, even though It is extremely doubtful whether this form meridionalis is a somewhat lankier, relatively should be allowed more than subgeneric rank, longer winged bird than the other species of the resemblance in most respects to Urubitinga Buteogallus and has the face more fully feath- being so exceedingly close. The points of dif- ered. ference pointed out above are indeed trifling, Peters (193 1, p. 226), for reasons unknown, though characteristic; and although the adult plumage is strikingly different from that of any placed Heterospizias in his subfamily Acci- of the species of Urubitinga, the young stage is pitrinae, not . Pl6tnick (1956) quite similar to that of U. anthracina. corrected this error and demonstrated that the genus is a buteonine in the inclusive Peters (1931 , p. 244) did place anthracinus sense, not that and Buteo are them- in Buteogallus, but left the Greater Black selves as distantly related as Peters thought. Hawk urubitinga in Urubitinga (=Hypo- Ridgway (1876, p. 139) was aware ofthe true morphnus), thereby made monotypic. But position ofHeterospizias and wrote: "Similar although the Greater Black Hawk, as the to Urubitinga but wings larger proportion- name implies, is a somewhat larger species ately .... Tibial plumes better developed and and one which has proportionately longer feathers of the pileum longer and more lan- legs than the Lesser Black Hawk, anthraci- ceolate .... Five outer primaries with inner nus, the two are closely allied, so much so webs sinuated .... Bill rather small, exactly that they are often confused both in life and as in anthracina. With a general correspon- as specimens. Thus it was but a natural step dence to Urubitinga in most of the external when Amadon and Eckelberry (1955, p. 68), features ofform, this genus, or more properly reverting to the classification adopted by subgenus, differs sufficiently in the greater some early authors, made Urubitinga a syn- development ofthe remiges and tibial plumes onym of Buteogallus. This has subsequently ... the more lanceolate form of the feathers been followed by Wetmore (1965, p. 229) of the pileum, and the strikingly different and others. coloration." When, however, the comparison Ridgway (1876, p. 167) regarded Asturina is with aequinoctialis it is the other species schistacea Sundevall, now placed in Leucop- ofButeogallus that differ in color. Friedmann ternis, as a "typical" member of the genus (1950, p. 396) noted the similarity in color Buteogallus. In color and perhaps otherwise and pattern of meridionalis and aequinoc- it and its allospecies, plumbea, do to some tialis. extent connect Buteogallus and Leucopternis, The lankier proportions of the Savanna but are closer to the latter. The immature Hawk (meridionalis) permit it to be a some- 1982 AMADON: SUB-BUTEONINE HAWKS 5 what more active predator and "pirate," in tus. It is a "megasubspecies" in the usage of the sense of Meinertzhagen (1959) than are Amadon and Short (1976). It is of interest the other buteogallines.,I watched one as it that B. anthracinus, unlike urubitinga, ex- pursued a stork (Euxenura galeata) in a long tends only to the northern coastal regions of circling flight until the latter dropped its prey, . To the south its ecological probably a , which the hawk seized and niche may be filled by Leucopternis schista- ate. A pair of Savanna Hawks circling over cea and plumbea. a little valley in the uplands of western Ar- gentina looked very buteonine; one of them GENUS PARABUTEO RIDGWAY, 1874 had the tail of a dangling from its bill. TYPE: Buteo harrisi Audubon = Parabuteo The is also closely related to unicinctus harrisi. the Bay-winged Hawk, Parabuteo unicinctus. SYNONYM: Antenor, Ridgway, 1874. Type, The question would seem to be whether Par- same as Parabuteo, name preoccupied. abuteo should follow Heterospizias into Bu- RANGE: Southwestern United States, south teogallus. I am tentatively retaining it for rea- to Chile and . sons given later. SPECIES: 1- unicinctus. COMMENTS ON SPECIES: I now follow Mon- DIAGNOSIS: Medium-sized, robust hawk; roe (1963, 1968) in treating subtilis as a wings and tail long, tail slightly rounded. species and not as a subspecies of anthraci- Outer four primaries emarginate on inner nus. Restricted to the coastal Pacific zone vane. Lores with bristles only. Bill relatively from southernmost to northwestern longer than in Buteo. Legs long, robust; talons (including the Pearl Islands, ) large, strongly curved, acute. Color pattern subtilis differs somewhat from anthracinus of adult, especially of the North American in bodily configuration and in habits (Dickey subspecies, harrisi, distinctive and contrast- and van Rossem, 1938; Slud, 1964), and the ing. Immature very different: mottled dor- two seem to be in contact without intergrad- sally, coarsely streaked ventrally as in Buteo- ing. I have elsewhere (Amadon, 1961) dis- gallus. "The cranium is very similar to that cussed the curious examples of subtilis, lack- of Buteo Uamaicensis] borealis, the details of ing melanin in the plumage, which occur in structure being exactly the same .... Its up- the southern part of the species' range. per portion is more depressed, presenting in The black hawks of the Atlantic coast of this respect, a closer resemblance to that of and the off-lying islands (utilensis) Asturina" (Ridgway, 1876, p. 161). and of Cuba (gundlachii) differ from conti- REMARKS: Elsewhere Ridgway (1874, p. nental anthracinus in somewhat the same 249) stated that his new genus Parabuteo way as does subtilis, but the divergence is less finds its closest ally in Buteogallus. I would and these taxa are probably still subspecies modify this only slightly to say that its very ofanthracinus. On the other hand, the species closest ally is the Savanna Hawk (meridio- aequinoctialis, which like subtilis is chiefly a nalis), then placed in a monotypic genus Het- coastal form of small size, ranging from the erospizias, but here in Buteogallus. Mr. Wil- Orinoco Delta east and south well down the liam Mader (in litt.), who has studied both coast ofBrazil, might be closer to subtilis than species intensively in the field, concurs with to anthracinus, though its dissimilar color this opinion. Both species are somewhat and plumage pattern suggests that it is not lankier and longer winged than most species particularly close to either. Alternatively, all of Buteogallus. But meridionalis is close three species may comprise a superspecies, enough to be placed in that genus, whereas and indeed the range of anthracinus termi- Parabuteo, in my opinion, may be retained. nates to the eastward on the north coast of It is a bird of open chaco, savanna, and even South America about where that of aequi- desert rather than wetlands and is a more noctialis commences. active powerful species than the buteogal- Buteogallus (anthracinus) gundlachii of lines, preying primarily upon warm-blooded Cuba, long considered a full species, may be vertebrates: , wood rats, and the like, written with the species name in parentheses and also birds. Falconers find it a good sub- to indicate some lingering doubt as to its sta- ject: docile and easy to handle, but fearless 6 AMERICAN MUSEUM NOVITATES NO. 2741 and rugged in hunting. Mader (1975) discov- REMARKS: Rupornis is now considered a ered a peculiarity in its mating: many nests synonym of Buteo, properly in my opinion. have a trio ofadults, two males and a female, Ofthe subgroups ofthat genus, it is probably and these trios produce more young than do the most primitive and hence nearest the sub- pairs. A pair or trio has been known to start buteonines. Early authors, including even another brood while still feeding the young Ridgway, were impressed by the resemblance of an earlier one in the nest. of nitida to the (Astur To summarize, Parabuteo is close to Bu- [=Accipiter] gentilis), and hence the name teogallus, especially to meridionalis. Its re- "Asturina" and the former vernacular "Mex- semblances to Buteo, in which genus some ican Goshawk." Everyone is now agreed that would place it, are partly convergence, not the resemblance is superficial. that any of the sub-buteonines are far from Recent works treat Asturina as a synonym that genus. ofButeo. Its one species, nitida, is placed near the beginning of that genus, along with leu- corrhous, ridgwayi, lineatus, and sometimes GENUS ASTURINA VIEILLOT, 1816 brachyurus. Johnson and Peeters (1963) set TYPE: Asturina cinerea Vieillot = Falco up an exclusively New World group they nitidus Latham. called "woodland buteos" for these species RANGE: Southern Texas to southern Ari- (except leucorrhous, which would presum- zona, thence south to Argentina, in lightly ably have to be added, though it does not wooded or savanna country. have the color characteristics of the others). SPECIES: 1- nitida (includes plagiata), but Asturina itself, however, is, I am sure, a sub- see Johnson and Peeters (1963). buteonine. Ridgway was correct in saying DIAGNOSIS: Rather small hawk. Wings rel- that its closest relative is Leucopternis and atively short, only four outer primaries sin- ifthe immature plumage were like that ofthe uate on inner web. Tail long, almost square adult, it could be placed there. But it is not; or slightly emarginate. "Feet large and robust rather it is coarsely streaked like that of Bu- ... reaching almost to the end of the tail teogallus. Asturina thus stands between these ... tarsi very robust compared to the toes two genera. This may seem difficult to accept ... the frontal and posterior rows ofscutellae if it is compared with B. urubitinga, or even very distinct and regular ... claws strong, B. anthracinus, but ifthe comparison is with well curved, but not very acute" (Ridgway, the small, extensively barred B. aequinoc- 1874, p. 244). Color pattern ofadult distinc- tialis, the relationship is more evident. As- tive: pale gray throughout more or less finely turina is the oldest of the three names, and barred, according to subspecies, with darker if any further reduction of genera were con- gray, especially ventrally. Immature quite templated it would be necessary to combine different: streaked or blotched ventrally. An Leucopternis, Buteogallus (and Parabuteo) open, often dry country species that prefers with it. But one is hardly prepared to bring as food. the smaller species of Leucopternis into the Ridgway (op. cit.) wrote further: "The same genus with Buteogallus anthracinus and species of Rupornis [magnirostris and leu- B. urubitinga. Asturina does have some char- corrhous] have been associated with the acters ofits own, as outlined above. The fact species of the present genus [e.g., by Sharpe, that only four outer primaries are emargin- 1874] but are very distinct. The genera (or ated suggests Buteo, and is probably corre- more properly subgenera) most nearly allied lated with the fact that in general it inhabits to Asturina are Leucopternis Kaup oftropical more open country than the species of Bu- America, and probably Kaupifalco Bona- teogallus and Leucopternis and is perhaps a parte of Africa. The former differs mainly in more skilled flier and predator. more or less rounded instead ofemarginated Ridgway noted the similarity of Asturina tail and in having the old and young to the (Kaupifalco mono- similar; the latter in having the posterior face grammicus) of Africa. It is of interest that of the tarsus without a well defined row of Thiollay (1978) in a comparison of the rap- transverse scutellae." tors of the lowlands of southeastern Mexico 1982 AMADON: SUB-BUTEONINE HAWKS 7 and of the Ivory Coast found a striking con- feathered but never as denuded as in some vergence in and habits between species ofButeogallus. Wings relatively short the two genera (see also Niles, 1979). It may and broad. eventually seem possible to combine them, Ridgway (1876, p. 174) added: Leucopter- though Leucopternis will have to be consid- nis is ered also; it is more like Kaupifalco than is Asturina in one respect-the similarity of similar to Urubitinga [=Buteogallus] but tarsus not more than twice as long as the middle toe; young to adult. inner webs of only four or five outer primaries Johnson and Peeters (op. cit.) did not con- cut (the cutting being a distinct emargination); sider the number of emarginate primaries in and- nostrils more nearly circular, and rather defining their group of woodland buteos; in- vertical than horizontal. The genus is most sim- deed the group contains species with three ilar to Urubitinga, but presents the above well (lineatus), four (platypterus), and even five marked differences, with the addition ofanother (magnirostris) incised quills. The character and equally important one, viz., the similarity was used extensively in the earlier literature of the young and adult stages in plumage .... to subdivide Buteo. It is one of those nu- There is also a very close relationship to Astur- merous characters which are sometimes use- ina, but in the latter the old and young stages not. For of plumage are exceedingly dissimilar, the nos- ful, sometimes example the subgroup tril is very decidedly horizontal, and the emar- ofAmerican buteos swainsoni, albicaudatus, ginations ofthe primaries end abruptly with the polyosoma, and galapagoensis seems to be fourth, on which it is very distinct. natural and all ofthem have three emarginate The sternum ofL. semiplumbeus most resem- primaries. Yet one would not be inclined to bles that of Asturina. . ., from which it differs associate Buteo lineatus with this group, even mainly in smaller size. That of Rupornis (rufi- though it also is New World and has three cauda) [=Buteo magnirostris ruficauda] differs notched primaries. It is possible that the sim- conspicuously in much greater posterior breadth ilarities of the woodland buteos to one and larger foramina, which in the other two are another are in part superficial and ecological, sometimes nearly or quite obsolete, and always just as is the resemblance to them in plum- small. ages of Accipiter striatus and A. cooperii. At the time Ridgway wrote he considered schistacea and plumbea, now placed in Leu- GENUS LEUCOPTERNIS KAUP, 1847 copternis partly because their immature plumage is similar to that of the adult, as TYPE: Falco melanops Latham. members of Buteogallus. This further em- SYNONYM: Morphnarchus Ridgway, 1920, phasizes how closely interrelated these genera type Leucopternis princeps P. L. Sclater. and Asturina are. RANGE: Mexico to Argentina, usually in The species of Leucopternis are for the moist forest. most part sluggish inhabitants of forest, SPECIES: 1- melanops; 2- kuhli; 3-schis- where they catch cold-blooded , tacea; 4-plumbea; 5-semiplumbea; 6-la- large , and the like and only now and cernulata; 7- albicollis (including ghies- then a or bird. breghti and occidentalis); 8-polionota; 9- COMMENTS ON SPECIES: The genus Leucop- princeps. ternis offers several problems at the species Leucopternis has the following superspe- level. Plumbea and schistacea, one west and cies: (melanops + kuhlii); (schistacea + one east ofthe Andes in lowland forest, form plumbea), (albicollis + polionota). a as do and DIAGNosIs: Small to medium-sized sub- superspecies; melanops kuhlii, buteonines. Immature plumage like that of one north and one south of the Amazon. adult in color and pattern; tail with promi- Whether the last two or semiplumbea of nent white band. Color variable from species Middle America have any particular tie with to species; ranging from dull black through- the little known lacernulata of southeastern out to almost entirely white. One species, L. is more doubtful. Both lacernulata and princeps, finely barred on breast, recalling polionota, discussed below, have been greatly Asturina and . Face sparsely reduced by the deforestation of southeastern 8 AMERICAN MUSEUM NOVITATES NO. 2741

Brazil. The late E. Kaempfer, who collected albicollis a superspecies. He states that it is in that part of Brazil during the 1 920s never usually found in the hills, whereas L. lacer- secured polionota and only two specimens of nulata, prefers the lowlands. lacernulata. The former has long been a rare The sequence in which the remaining gen- bird; the American Museum lacked it until era of New World sub-buteonines are given we secured one from the Brazilian national is oflittle significance; all are offshoots ofthe collection. preceding core group. In the Leucopternis albicollis complex there are three problems as to species limits. First, the whitest of the forms of the "White GENUS BUSARELLUS LAFRESNAYE, 1842 Hawk," ghiesbreghti of Mexico to Costa TYPE: Falco nigricollis Latham. Rica, was at one time considered specifically RANGE: Mexico to central Argentina in distinct from albicollis. When Meyer de lowlands near ponds, sluggish streams and Schauensee described an intermediate form, lagoons. williaminae, from , the question SPECIES: 1- nigricollis. seemed answered; Wetmore (1965, p. 253), DIAGNOSIS: Medium-sized, stocky sub-bu- however, reverted to treatment of ghies- teonine, with very broad wings and short tail. breghti as a species. I think this unjustified In general like Buteogallus but adapted in and see no reason even to express doubt by various ways for catching fish and with a placing the name albicollis in parentheses. somewhat different color pattern. Soles of The second doubtful member of the albi- toes and feet covered by "rugose and thorny collis group was described as a species, Leu- spicules" (Sharpe, 1874, p. 159); talons long, copternis occidentalis, by Salvin in 1876 from strongly curved, and acute. In both these re- the Andean slopes of western . Its spects it is like other fish-catching hawks and entire back is uniform gray in contrast to the . black markings ofthe other taxa in the group. Ridgway (1876, p. 142) wrote of Busarel- It may be subtropical rather than tropical in lus: zonal preference and is separated by the An- des from its nearest relatives. For this reason General form and appearance of Buteogallus Brown and Amadon (1968) and others have aequinoctialis. Bill, feet, and claws very strong; listed it as a wings large (but primaries short), and out- species, though subspecies status stretched feet reaching beyond the end of the has been suggested by others. Stresemann in rather short, nearly even tail .... Inferior sur- the manuscript, upon which Stresemann and face of the toes with acute papillae .... Claws Amadon (1979) is based listed it as Leucop- very strong ... deeply grooved beneath. Head ternis (albicollis) occidentalis. I am now will- normally feathered .... Secondaries much de- ing to follow this usage, which is an educated veloped and very broad. Plumes of the outer guess that the form has not yet crossed the face of the tibiae short and close .... This species threshold. strongly characterized and very remarkable ge- This brings us to the third , the one nus exhibits a striking analogy to [Os- named from Sao Paulo, Brazil by Kaup in prey] in the very strong and slightly graduated 1847 as Asturina claws, the close feathering ofthe tibiae, the gen- (Leucopternis) polionota. eral form of the bill, and the sharp spicules of Peters (1931, p. 242) wrote it L. (albicollis?) the toe pads. polionota; perhaps he was the first to suggest that it may be a subspecies ofalbicollis. Hell- REMARKS: This species often soars high in mayr and Conover (1949, p. 173, fn.) also the blue, when its broad wings and short tail pointed this out, but thought it best to leave impart a distinct contour and make it appear polionota as a species until its relationships larger than it is; it returns to earth in an im- with lacernulata were clarified. The latter, pressive plunge, sometimes interrupted. It however, is a considerably smaller bird, pos- does not indulge in such dives when fishing, sibly allied to L. melanops and kuhli. Dr. but seizes its prey in shallow stoops. Olson Helmut Sick, one of the few who has had (in press) has recently come to a very different field experience withpolionota, wrote me that conclusion as to the affinities of Busarellus. he prefers to keep it as a species, forming with He found that in six of seven specimens the 1982 AMADON: SUB-BUTEONINE HAWKS 9 phalanges of the second toe are anky- soaring. As befits its aquiline size, it nests on losed, a feature found elsewhere only in the ledges of lofty cliffs when they are available, milvine kites (, , and a few oth- but in low country builds in trees. The food ers) and in the true sea-eagles (Haliaeetus, includes medium-sized rodents but it is said Ichthyophaga, but not Gypohierax). On this to come to carrion. The slight resemblance basis and some general resemblances, Olson of Geranoaetus to the sea-eagles (Haliaeetus) would place Busarellus with this group of is, as will be evident from the above diag- kites. This I find difficult to believe, unless, nosis, superficial. perchance, such kites which I (Brown and In various respects Geranoaetus suggests Amadon, 1968, chap. 1) consider to be rel- Buteogallus and Harpyhaliaetus, whereas the atively primitive members of the Accipitri- barred breast recalls that of Leucopternis dae, or their antecedents, gave rise to the sub- princeps. It also has some characters of its buteonines, one ofwhich, Busarellus, retains own, as listed above and certainly warrants in most individuals this feature. generic status. In Hellmayr and Conover's (1949, p. 144) volume on the Falconiformes of the , we find it well disguised as GENUS GERANOAETUS KAUP, 1844 Buteofuscescens, even the species name hav- TYPE: Falco aguai Temminck = ing been unnecessarily changed. I have else- melanoleucus Vieillot. where (Amadon, 1963) outlined the peculiar RANGE: Southern South America, extend- sequence ofevents whereby: (1) certain more ing north at higher elevations in the Andes or less fragmentary bones of fossils from to . were assigned to Geranoaetus SPECIES: 1- melanoleucus = fuscescens of because they were Buteo-like, but large; (2) some authors. other fossil or subfossil bones were assigned DIAGNOSIS: Size very large, female much to the living G. melanoleucus; (3) Wetmore larger than male; wings long and broad; tail (1933), finding no difference between the fos- short, wedge-shaped. Breast finely barred; sil bones and those ofButeo, called them that; feathers pointed and overlapping the white (4) thus the Recent Geranoaetus melanoleu- upper abdomen; feathers of nape and crown cus found itself in Buteo along with the fos- also lanceolate. Immature very different, sils. streaked and splotched below as in Buteo- So far as the living species is concerned, gallus or Harpyhaliaetus. Bill and head large. it was all a mistake. Legs long and coarse. Ridgway (1876, p. 131) added: "Form intermediate between that of GERANOSPIZA KAUP, 1847 Buteo and Haliaetus. Wings very long, the GENUS primaries rigid ... five ... with their inner TYPE: Falco gracilis Temminck = Gera- webs deeply emarginated. Tail less than half nospiza nigra gracilis. the wing, graduated, the feathers very stiff. SYNoNYM: Ischnosceles Strickland, 1844, Feathers of the head and neck cuneate, ap- same type species. Considered preoccupied proaching lanceolate, as are those ofthe lower by Ischnoscelis Burmeister, 1842 (Insecta) breast; tibial plumes well developed; second- and for that reason renamed Geranospiza by aries 17. Somewhat intermediate between Kaup. Wetmore (1965, p. 253) concluded Buteo and Haliaetus, this genus is neverthe- that Ischnosceles is not preoccupied and used less very distinct from either .... The feet it, though it had never been in general usage ... are much more like those of Buteo, there and not at all for about 100 years. Thus the being a well developed web between the outer name Geranospiza may be retained on the and middle toes, of which not a trace is seen grounds ofusage, iffor no other reason (Mon- in Haliaetus, while the scutellation is also roe, 1968, p. 85). that of the typical Buteones." RANGE: Mexico to Argentina. REMARKS: The Gray Eagle-buzzard inhab- SPECIES: 1- nigra. The species varies in its open or lightly wooded country, both in color from black in the north (Mexico) to plains and mountains. Like Busarellus it ap- finely barred with pale grayish in the south pears broad-winged and short-tailed when (Argentina) but there is a cline from the one 10 AMERICAN MUSEUM NOVITATES NO. 2741 to the other and it is now evident that there the Asian Ictinaetus malayensis has is but one species. a short outer toe as in and Gera- DIAGNOSIS: "Form very slender, the wings nospiza, and like them it often feeds on nest- and tail very long, the head small, bill weak lings. Ictinaetus is not closely related to these and tarsi extremely elongated and slender. genera, so in its case at least, some degree of Outer toe very much shorter and weaker than convergence is indicated .... If similarities of the hind limb are discounted, the inner ... its claw disproportionately the evidence for close relationship of Polybo- small and weak. Tibio-tarsal joint flexible roides and Geranospiza is much weakened. both backward and forward! Secondaries Other similarities are such features as general much developed, reaching nearly to the end coloration, broad secondaries, weak bill, and of the primaries, and very broad. Bill much some cranial resemblances. On the other hand, as in Nisus [=Accipiter]; nostril obliquely hor- differences include facial feathering (a much izontal, oval. Tarsus ... with frontal and pos- larger bare zone in Polyboroides), tarsal cover- terior series of broad transverse scutellae, ing (reticulate in Polyboroides, scutellate in these often fused into continuous plates; Geranospiza), and totally different immature claws normal. Tibial feathers short and close; plumages. On balance I am inclined to doubt not plume-like ... fourth to the sixth quills that the two genera are closely related. longest, the first shortest and much bowed; Cooper (1980), after further comparisons, outer six with inner webs sinuated. Tail long, came to the same conclusion. nearly equal to the wing, rounded, the feath- Ictinaetus has relatively short legs and does ers very broad" (Ridgway, 1873, p. 82). not share the modified tibiotarsal joint ofthe REMARKS: This small hawk probes with its other two genera. The peculiar shortening of long legs for and the like in crevices, the outer toe, which is found in all three gen- knotholes, the bases of bromeliads and era probably narrows the foot and makes it among rocks (Brown and Amadon, 1968, p. easier to withdraw it from narrow nests or 378). Called "" presumably be- crevices, especially when prey is clenched in cause ofits long legs; they are not long enough the foot. In Ictinaetus, which feeds somewhat to make the name more than fanciful. differently-slowly soaring above the forest Geranospiza has been thought to be related and snatching nests or their contents from to the African Banded Hawk (Poly- the treetops-there is a further modification: boroides = Gymnogenys) because both have the talons are narrow, thin, and somewhat "doubled-jointed" tarsi, and feed similarly straightened, thus lessening the chance that by grappling in cavities for prey. There are the foot will become snagged when the bird also resemblances in color and proportions. snatches chicks from a nest. In Geranospiza Burton (1978) compared the limb bones of and Polyboroides, which do not feed in flight the two genera and has published an illus- and which frequently drag prey from crev- trated report. He found similar modifications ices, the talons are normally curved. in the two genera, but they do not involve The position of Polyboroides is doubtful. fundamental changes. He summarized as fol- Brown and Amadon (1968, p. 368) suggested lows: relationship to the serpent-eagles, noting the reticulated tarsi and the resemblance of the The detailed resemblance between the two gen- immature plumage to that of . I am era in the modifications of their tibiotarsi and still of that opinion, but Brown (1972) stud- tarsometatarsi appears to strengthen the as- ied a nesting pair of Polyboroides and con- sumption that they are related, particularly since cluded that the genus is more closely related the shared characters are derived ones. Never- to the buteonines. Further studies are needed. theless, convergence can by no means be ruled I here follow an earlier suggestion (Brown out. If such features as the orientation of the supratendinal bridge and the form and siting of and Amadon, 1968, p. 21), placing Gera- the calcaneal ridges are simple consequences of nospiza with the sub-buteonines. Jollie the overall narrowing of the joint, they could (1976-1977, pt. 3, p. 119) also placed Gera- perfectly well have arisen independently. In this nospiza with the buteonine and sub-buteo- connection, it is of some interest to note, as nine hawks. Apparently he does not discuss Brown and Amadon (1968) pointed out, that the genus in any detail in his compendious 1982 AMADON: SUB-BUTEONINE HAWKS I1I work; it is not included in a list of genera REMARKS: Friedmann and before him Pe- which he dissected or which had been dis- ters (1931) recognized the genus Urubitornis; sected by others; he mentions it in a short list that is, they placed each of the two species ofgenera whose position in the family, even under discussion in a monotypic genus. I as he wrote, was still uncertain. I think he (Amadon, 1949) set forth reasons why this assigned it properly. seems to be unnecessary. Coronatus has a So far as general coloration and plumage long pointed crest, solitarius does not; oth- are concerned, the southern (Argentina) gray, erwise the two are similar. There are many barred subspecies of Geranospiza is some- birds of prey in which the crest varies enor- what like Polyboroides typus, but equally like, mously among closely related species or sub- for example, Asturina nitida. The black, species: e.g., Spizaetus c. cirrhatus vs. S. (cir- northern race of Geranospiza nigra, from rhatus) limnaeetus or ptilorhynchus which the species takes its name, resembles orientalis vs. P. ptilorhyncus torquatus. Hell- some of the species of Buteogallus and Leu- mayr and Conover (1949, p. 197) swung to copternis; like them it has a white tail bar. the other extreme and made solitarius a sub- Immatures of Geranospiza nigra or at least species ofcoronatus. In this I am certain they those of some of the subspecies, are similar were wrong, especially since solitarius is a to the adults; this may be a tie with Leucop- bird offorest, often at subtropical elevations, ternis. The same is not true of Polyboroides. whereas coronatus frequents semi-open sa- The assumption is then that Geranospiza vanna country. belongs to the Neotropical sub-buteonine as- Wetmore (1965, p. 239), who again rec- semblage and is only superficially like Poly- ommended that Urubitornis be recognized, boroides. pointed out the extraordinary resemblance of solitarius to the (Bu- teogallus urubitinga). But if coronatus were GENUS HARPYHALIAETUS LAFRESNAYE, 1842 black instead of dark brown and lacked a TYPE: Harpyia coronata Vieillot. crest, it too would resemble B. urubitinga. SYNoNYM: Urubitornis J. Verreaux, 1856. He stated that the tarsal covering of Buteo- Type, solitarius Tschudi. gallus urubitinga, "especially at its lower RANGE: Mexico to Argentina. end" suggests that of solitarius, but I cannot SPECIES: 1-solitarius; 2-coronatus. see that it departs from the usual buteonine DIAGNOSIS: Eagle-sized sub-buteonines. type. As Friedmann (1950, p. 389 and fig. Primaries relatively short, secondaries long 25) noted, the former has the "acrotarsium and broad, almost reaching tips of primaries and planta tarsi with a continuous series of in the folded wing. Adults monocolored broad transverse scutella." blackish or dark brown according to species, The Black was placed in the with white tail bar. Immatures unlike adults: genus Circaetus by its describer, Tschudi, in paler above, somewhat mottled; below dull 1844. Both species of Harpyhaliaetus do white or buffy, coarsely streaked and blotched have some resemblance to this African genus with dark or blackish brown. "Tarsi ... with- of serpent-eagles and especially to C. cine- out a continuous series of transverse scutella reus. This is true not only in bodily size and on either acrotarsium or planta tarsi ... in- configuration but in the reticulate covering stead acrotarsium covered for the greater part ofthe tarsi. Nonetheless, it is virtually certain with small, rough, hexagonal scales; about that this resemblance is superficial. Very little 6-8 large, broad transverse scutella on the is known of the habits of either species of middle portion" (Friedmann, 1950, pp. 67, Harpyhaliaetus but at two nests of solitarius 415-416; fig. 28, p. 418). Legs and feet large, in Mexico, the only food observed was large extending to or beyond the tip of the rather (Harrison and Kiff, 1977). The rough, short tail. Talons relatively small, not strongly reticulated tarsi may serve the same function curved. Friedmann's description ofthe tarsal in Harpyhaliaetus that they are assumed to scutellation of Harpyhaliaetus is based upon in Circaetus and related genera, namely, to solitarius, but that of coronatus is similar, or provide protection from the fangs of ven- if anything even more reticulated. omous reptiles. Hence the reticulation is con- 12 AMERICAN MUSEUM NOVITATES NO. 2741 cluded to be a derived character in Harpy- Many large insects and sometimes small ver- haliaetus, not a primitive one. tebrates are consumed, most of them seized I continue to regard the species solitarius from the ground or picked from foliage. and coronatus as forming a natural genus Harpyhaliaetus, derived from buteogalline GENUS KAUPIFALCO BONAPARTE, 1854 stock, but not quite as close to Buteogallus TYPE: Falco monogrammicus Temminck. urubitinga as might be thought at first glance. SYNONYM: Asturinula Finsch and Hart- Its two species may form a superspecies, laub, 1870, same type. though probably not. RANGE: Sub-Saharan Africa. SPECIES: 1- monogrammicus. GENUS BUTASTUR HODGSON, 1843 DIAGNOSIS: Small, trim hawk, barred be- TYPE: Circus teesa Franklin. low, gray above, with prominent median RANGE: Africa; from east to the throat stripe as in Butastur. Immature plum- larger islands ofthe East Indies, north in east- age very similar to that of adult. Wings of ern Asia to southeast Siberia and Japan. medium length; tail slightly rounded and SPECIES: 1- rufipennis; 2- teesa; 3-liven- with white bar. Bill small; nostrils rounded, ter, 4-indicus. Rujipennis is African, teesa with a bony tubercle. Posterior row of tarsal Indian, liventer southern Oriental and East scutes tending to be somewhat subdivided Indian and indicus northern Oriental. The (fig., Jackson and Sclater, 1938, p. 187). four may constitute a superspecies. REMARKS: The is similar DIAGNOSIS: Rather small hawks; wings and to some of the smaller species of Leucopter- tail long; body slight; flight buoyant. Color nis, not least, as Ridgway noticed long ago, pattern brown to rufous with fine ventral in that the immature plumage is like that of streaking or wavy barring; immatures tend- the adult. Yet Kaupifalco may be more ing to be more streaked but much like adult. closely allied to its Old World compatriot, No white tail bar, but with a dark central Butastur, as indicated by the slightly unusual throat streak. The typical buteonine row of tarsal scutellation and other resemblances. large tarsal scutes, fore and aft, tend to be One hesitates to place the two Old World broken up and subdivided, especially the genera Kaupifalco and Butastur between posterior one. Perhaps for this reason, or be- some of the closely allied Neotropical ones, cause Sharpe (1874) had Butastur and Kau- and since they are relatively unspecialized it pifalco associated with such genera as Cir- is best to place them before that group, as for caetus, Hartert (1912-1921, p. 186 and fig. example in Stresemann and Amadon (1979, 196) stated that Butastur finds its closest al- p. 349). In this paper, since the Neotropical lies with the serpent-eagles, not with the bu- group is larger and more diversified, it has teonines. Everything else, however, points to been considered first. the reverse. The Lizard Buzzard inhabits semi-open REMARKS: Butastur rufipennis of Africa is and dry country, sometimes around villages. quickly attracted to grass fires, where it flaps Rather sluggish, it spends much time perch- along without soaring but with buoyant ing, occasionally uttering a loud, melodious flight, consuming insects wafted skyward by call. It drops to the ground to seize lizards, the heat and flames. Butastur indicus is a rodents and the like. Small birds do not seem species of deciduous woodland, comparable to fear it. perhaps with such a hawk as Buteo platyp- terus in eastern North America. It is the most migratory of all sub-buteonines, reaching RELATIONSHIPS OF THE Taiwan, the , and the East Indies. GENERA OF SUB-BUTEONINES For the most part the species of this genus The interrelationships ofthe genera ofsub- prefer savannas and semi-open woodland. buteonine hawks are self-evident from the Though rather sluggish, they have an accip- above diagnoses but may be summarized as iter or even -like profile when perched. follows. Four of them-Buteogallus, Para- 1982 AMADON: SUB-BUTEONINE HAWKS 13

Harpia gp.

FIG. 1. Suggested phylogeny ofsub-buteonines and close allies. The Asturina group includes Asturina, Buteogallus, Leucopternis and Parabuteo. For composition ofAquila group and Harpia group see caption for fig. 2. buteo, Asturina, and Leucopternis form a relative Parabuteo the same treatment. Yet closely knit group. They could all be sub- Parabuteo is ecologically and morphologi- sumed under Asturina, the oldest name, with- cally a step away from Buteogallus (including out major distortion ofthe facts, yet to bring Heterospizias) and I prefer to maintain it. the smaller, more slender species of Leucop- The other New World genera of sub-bu- ternis in the same genus with the large, coarse teonines-Busarellus, Geranospiza, Gera- black hawks (Buteogallus) would scarcely be noaetus, and Harpyhaliaetus -are all to vary- acceptable. Furthermore Asturina is usually ing extents specialized offshoots ofthe above placed in Buteo and it is advisable to await nuclear group. Busarellus is very close to agreement that it is a sub-buteonine before Buteogallus but specialized for catching fish consideration ofcombining other genera with (but see Olson, in press). Harpyhaliaetus is it. Short of this, the only step that might be Buteogallus converted to eagle-size, with ap- taken now that Heterospizias has been placed propriate changes in proportions. Gerano- in Buteogallus would be to accord its close spiza resembles various species of Buteogal- 14 AMERICAN MUSEUM NOVITATES NO. 2741 lus and Leucopternis in many ways; the is in other respects intermediate between the validity of the genus would not appear to be latter group and the true Buteones; the resem- open to question. Geranoaetus as already blance is especially great in the wings, the noted has a mixture of characters of several secondaries having a similar excessive de- of the other genera, plus some of its own. velopment, while the shallow sinuation ofan As to the Old World genera Butastur and indefinite number of primaries is another Kaupifalco the latter, if it were Neotropical, point ofagreement." Whether the two groups might be regarded as congeneric with Leu- are close enough to render such superficial copternis or rather the other way around since similarities ofsignificance is a moot question. Kaupifalco is the older name. In such spa- Aside from the long bare tarsi, some ofthese tially separated taxa, however, one suspects eagles resemble the larger, shorter winged a degree of convergence and it is quite pos- species ofthe "booted" eagles (with feathered sible, as already noted, that Kaupifalco is tarsi) such as coronatus of closer to its Old World relative Butastur than Africa. It must be admitted that the whitish to Leucopternis. immatures of Morphnus and Harpia do re- The genera of sub-buteonines will natu- semble the similar plumage stages of Ste- rally be listed with the specialized offshoots phanoaetus, Spizaetus, and Oroaetus. None- last followed by Buteo. This means that the theless, the coarsely scaled, long legs and unspecialized sub-buteonines that are nearest other features of the Morphni suggest that Buteo, such as Parabuteo and Leucopternis, they are an earlier offshoot of sub-buteonine will be separated from it by genera less closely stock, whereas the are a more allied, but such dislocations are inevitable in recent development from something very any linear list. I prefer to place Kaupifalco close to Buteo. The skeletons of Buteo and and Butastur before the New World sub-bu- are extremely similar. teonines as noted and hence near certain Old Finally a few words, which may be consid- World genera such as to which they ered as supplementary to the discussion in may conceivably be allied. In a linear clas- Brown and Amadon (1968), about the clas- sification one can only make such compro- sification of the family Accipitridae as a mises as seem best and resort to diagrams or whole. I still regard the kites as the most other devices to set forth details of the ap- primitive members of the family, though parent phylogeny (figs. 1, 2). Stresemann (Stresemann and Amadon, 1979, Four genera of tropical eagles, all mono- p. 273) had them listed last, and hence pre- typic, may be thought of as descendants of sumably considered them advanced. The an earlier offshoot of sub-buteonine stock. kites require study, in part to see whether They are Morphnus and Harpia in the Neo- their subdivision by Ridgway (1876) and tropics; Pithecophaga in the Philippines, and later by Friedmann (1950) into three sections Harpyopsis in New Guinea. The first two are should be maintained and at what taxonomic closely related; Morphnus might be united level. The subgroup centered in Milvus, with Harpia, but is much more slightly built. seems to lead naturally into the sea and fish Pithecophaga and Harpyopsis are allied to eagles (Haliaeetus and Ichthyophaga). The each other also, but less obviously. These last latter in turn may be perhaps linked with the two are insular relicts; there is nothing similar accipitrid ("Old World") through to them elsewhere in the Old World. Shufeldt the genus Gypohierax. Two problem genera, (1919) mentioned some osteological similar- Ictinaetus and Gypaetus, may belong in this ities between Harpia and Pithecophaga. vicinity, though it is customary to leave the Ridgway (1876, p. 167) set up a group former with the booted eagles of the Aquila "Morphni" for Harpia (his Thrasaetus) and group. Morphnus; to this could presumably be The serpent-eagles (Circaetus and related added Harpyopsis and Pithecophaga. He genera) are probably derived from -like wrote: "The pterylosis of Urubitinga [=Bu- stock also, though no direct links remain. teogallus, in particular B. urubitinga] pre- Two genera, Polyboroides and Melierax, may sents many points of resemblance to that of belong here. The advanced and widely dis- the Morphni ... and it is likely that the genus tributed genera Circus and even Accipiter, if 1982 AMADON: SUB-BUTEONINE HAWKS 15

Accipiter gp.

Igp-

Aegypius gp.

Non-Milvine KitesN~ FIG. 2. Suggested phylogeny of family Accipitridae. This is a slightly amended version of the phy- logenetic tree presented in Brown and Amadon (1968, p. 20). The contents of the groups (abbreviated "gp." on this figure and on fig. 1) are: Non-Milvine Kites- 13 genera- through Ictinia in Stresemann and Amadon (1979, pp. xiii-xiv). Milvus gp.-Milvus, , Lophoictinia, Hamirostra. Haliaeetus gp. -Haliaeetus, Ichthyophaga. gp. -All the accipitrid (=Old World) vultures, including doubtfully Gypaetus and very doubt- fully Gypohierax. Cincaetus gp. -The five genera of serpent eagles and hawks, including doubtfully Polyboroides. Accipiter gp.-Includes, after recent reduction (Amadon, 1978) only Accipiter and Urotriorchis. Sub-buteonines-The 10 genera treated in detail in this paper; see figure 3. Harpia gp.-Harpia, Morphnus, Harpyopsis, Pithecophaga. Aquila gp. -All ofthe "booted" eagles, including very doubtfully Ictinaetus. Genera Ictinaetus through Polemaetus in Stresemann and Amadon (1979, p. xv). the latter is as close to the subgenus Micro- buteonines in the narrower sense used in this nisus of Melierax as it appears to be, may be paper, and finally Buteo and the Aquila offshoots of similar stock. Next are the sub- group. Aquila comprises eight genera and 16 AMERICAN MUSEUM NOVITATES NO. 2741

TABLE 1 Mean Measurements of Total Length, Wing, Tail, and Tarsus (in millimeters) and Weight (in grams) for Selected Species of Sub-buteonine and Buteonine Hawksa Total Weight Length Wing Tail Tarsus Kaupifalco monogrammicus meridionalis 9 311 180 235 148 54 Butastur indicus 8 404 235 320 190 56 Butastur rufipennis 2 360 205 317 177 56 Buteogallus aequinoctialis 725 295 312 160 78 Buteogallus urubitinga ridgwayi 6 1160 342 376 243 116 Buteogallus meridionalis 935 315 407 203 101 Asturina nitida plagiata 485 265 251 160 71 Leucopternis melanops 6 307 210 212 140 61 Leucopternis albicollis ghiesbreghti 8 652 288 354 225 83 Parabuteo unicinctus harrisi 2 998 300 361 233 87 Busarellus nigricollis 9 705 365 392 180 80 Geranospiza nigra nigra 6 349 232 302 239 88 Geranoaetus melanoleucus 2420 414 506 228 109 Harpyhaliaetus coronatus 2950 450 542 290 130 Harpyhaliaetus solitarius - 445 512 244 127 Buteojamaicensis borealis 1126 315 380 223 85 a Sometimes, as indicated, the data are for one subspecies or one sex only, others represent the averages for the two sexes. The logarithms of these measurements, or of the cube root of the measurement, for weight, were used in plotting the ratio diagrams of figure 3 and figure 4. a total of 30 species, all so close that only in millimeters or for weight, in grams, may moderate havoc would be wreaked by placing be found in various publications, e.g., Brown them all in Aquila. and Amadon (1968) and Friedmann (1950). The cosmopolitan or nearly so genera Cir- Wing, tail, and tarsus sometimes vary in- cus, Accipiter, Buteo, and Aquila appear on dependently ofgeneral size, for example, Bu- the one hand to be more advanced or derived teogallus urubitinga has relatively longer tarsi than many related or putatively related trop- than the other species ofthat genus. It is use- ical genera but at the same time to have as ful to have a measure of general size when a rule simpler color patterns, and a lesser evaluating such variation. Total weight, de- development of crests and similar features. spite its variability, is one such index (Ama- This may reflect the fact that such genera, all don, 1943). In Brown and Amadon (1968) of which have temperate or even subarctic an effort was made to search the literature zone nesting species, successfully coped with for weights, but even so none at all were the glacial period and thus acquired certain found for some taxa and very few for many characters and adaptations, such as migra- of the others. For present purposes I have tion, which contributed both to their dis- been able to find at least one weight of a persal and overall success. Parallels may be species from each ofthe 10 genera. For Har- found in several other families, e.g., compare pyhaliaetus exactly one weight was available, crows (Corvus) with some of the tropical that of an individual of H. coronatus in the magpies (Kitta, etc.) or Sturnus with some of zoo at Belem, Brazil, kindly sent to me by the tropical starlings. Dr. H. Sick. Only for Parabuteo unicinctus were the mean weights of long series of spec- MEASUREMENTS imens available (Hamerstrom and Hamer- strom, 1978: birds trapped, processed, and The generic diagnoses presented above released), and from others kindly sent to me contain qualitative comparisons ofmensural by Prof. T. Cade. Weights for several species characters-general size and lengths of ap- are from the contributions of Dr. F. Haver- pendages, in particular wing, tail, and tarsus. schmidt to the of . Af- The actual measurements of these features ter figures 1 and 2 were completed, I found 1 982 AMADON: SUB-BUTEONINE HAWKS 17

Cube root weight

Wing

Tail

Tarsus 0.75 0.80 0.90 1.00 1.10 120 1.30 1.50 1.80 2.00 2.20 Ratio Scale FIG. 3. Comparison by logarithmic ratio diagram of the proportions of wing, tail, tarsus and (cube root of) weight in the 10 genera of sub-buteonines. Usually the type species of the genera are presented, but see text. Buteogallus meridionalis is included because it is usually placed in a separate genus Het- erospizias. The names correspond with those of table 1, which gives the mean measurements used in constructing figures 3 and 4. The ratios for the various measurements as compared with those of Asturina nitida, here selected as a typical sub-buteonine, may be read off directly from the ratio scale on the graph. Cube root of weight may be taken as indicative of general size. Thus from the graph one may conclude that Leucopternis melanops does not differ significantly in proportions from Asturina; that Geranospiza has long wings, tarsi and especially tail in proportion to its weight; that Buteogallus meridionalis has relatively longer wings than B. aequinoctialis (diverges more widely from cube root weight), etc. Only the initial letters of species names are shown on this figure and on figure 4; these enable the species to be determined from table 1 in the instances where more than one species ofa genus have been included.

that K. H. Voous (1969) had published in the flesh or from skeletons would better weights ofSuriname raptors based in part on reflect some aspects of general size. Unfor- Haverschmidt's work. Taxa by taxa compar- tunately such data are even scarcer than are ison indicated that his figures did not differ weights. sufficiently from those used here to warrant Another measurement of general size of reworking the graphs. birds may be obtained from museum study When weights are compared with linear skins. Although such specimens are empty measurements such as wing length it is best skins stuffed with cotton or a similar sub- to use the cube root ofthe weight (easily ob- stance, the preparator does endeavor to pro- tained from logarithms) to reduce it to a lin- duce a skin that resembles as closely as pos- ear equivalent of the other measurements sible a bird in the flesh as lying in a supine (Romer and Price, 1940; Amadon, 1943). position. By selecting well made skins, "total Weight to be sure does not reveal that, for length," as measured from the tip of the bill example, Geranospiza nigra is a much slen- to the tip of the tail gives an approximation derer bird than Buteogallus aequinoctialis. of the general size of the bird. When similar Perhaps some index based on length, breadth, taxa are being compared, for example sub- and depth of the body as taken from birds species ofthe same species, such total lengths 18 AMERICAN MUSEUM NOVITATES NO. 2741

Kaupifalco m.

Butastur r.

Buteogallus a.

Buteogallus u.

Buteogallus m.

Asturina n.

Leucopternis m.

Leucopternis a.

Parabuteo u.

Busarellus n.

Geranospiza n.

Geranoaetus m.

Harpyhaliaetus c.

Harpyhaliaetus s.

Buteo j.

0.20 0.25 0.27 0.30 0.40 0.50 0.60 0.70 0.75 0.90 1.00 1.10 1.20 1.30 1.40 1.60 Log ratio scale FIG. 4. Comparison by logarithmic ratio diagram of the proportions of wing, tail and tarsus in 14 species of sub-buteonines and of Buteojamaicensis. In this figure body length, as defined in the text, has been used as the standard of comparison; hence differences in proportions as indicated on the graph should be independent of variation in general size. Major differences in proportions agree with those in figure 3; minor fluctuations are not statistically significant owing to inadequacies of material (see text). Additional species, as compared with figure 3, were included to show intrageneric variation in propor- tions, for example the relatively long tarsi of Buteogallus urubitinga. Buteo jamaicensis is similar or identical in proportions to B. buteo, the type species of the genus.

may be compared directly to show size dif- tios, calculated for six taxa at random, are ferences as was done by Chapman (1940, p. sufficiently alike to indicate that such is the 422) for Zonotrichia capensis. When the tail case, especially in view of the fact that total or bill vary independently of general size, length is here based on measurements ofonly these may be subtracted from the total length one to three specimens for each taxon, and to give "body length" (Amadon, 1943, p. the weights, also few in number, are not from 165); the length ofthe neck as well should be the same specimens: Kaupifalco monogram- subtracted from total length in long-necked micus .368; Butastur ruJipennis .369; Leu- species such as herons. In hawks it is suffi- copternis melanops .357; Geranoaetus me- cient to subtract the length of tail, which in lanoleucus .440; Parabuteo unicinctus .400; birds is comprised offeathers only, from total and Geranospiza nigra .360. length to give body length. This was done Figures 3 and 4 are logarithmic ratio dia- here (table 1). grams as devised by Simpson (1941) and later In figure 3 (cube root of) weight is used as used by Amadon (1950) and others. In such an index of general size and in figure 4 body diagrams a standard of comparison is first length, taken from skins as just defined. If selected. In figure 3 it is Asturina nitida, cho- both are valid indexes to general size, the sen because it is a sub-buteonine of average ratio ofone to the other should be fairly con- size and proportions and the first named of stant from taxon to taxon. The following ra- a cluster of very closely allied genera, the 1982 AMADON: SUB-BUTEONINE HAWKS 19 others being Buteogallus, Parabuteo, and enough prepared for taking that measure- Leucopternis. From the ratio scale on the ment. graph one may read off directly the ratio of LITERATURE CITED the other taxa on the graph to Asturina for each listed measurement. The figures for cube Amadon, D. 1943. Bird weights as an aid in . root ofweight provide a basis point for com- Wilson Bull., vol. 55, pp. 164-177. paring the differences in proportions per se. 1949. Notes on Harpyhaliaetus. Auk, vol. 66, By copying offthe scale and placing the read- pp. 53-56. ing 1.00 at any data point, the ratios of the 1950. The Hawaiian honeycreepers (Aves, measurements of all the others on that hor- Drepaniidae). Bull. Am. Mus. Nat. izontal line to the point (species) selected may Hist., vol. 95, art. 4, 262 pp. be read directly from the scale; a property of 1961. Remarks on the genus Buteogallus. No- such logarithmic scales. vedades Colombianas, vol. 1, no. 6, pp. In figure 4 body length as defined above is 358-360. used as the standard of comparison, thus 1963. Comparison offossil and Recent species: some difficulties. , vol. 65, no. eliminating size as a factor to be considered 5, pp. 407-409. in comparing the proportions of the various 1964. Taxonomic notes on birds of prey. Am. taxa. Because of the scanty data and the fact Mus. Novitates, no. 2166, 24 pp. that not all measurements are from the same 1968. Further remarks on the superspecies specimens, only the more pronounced dif- concept. Syst. Zool., vol. 17, no. 3, pp. ferences in proportions should be accepted 345-346. as significant. For example, Geranospiza is Amadon, D., and D. Eckelberry characterized by relatively very long wing, 1955. Observations on Mexican birds. Con- tarsus, and especially tail. Supporting what dor, vol. 57, no. 2, pp. 65-80. was said above, the more significant inter- Amadon, D., and L. Short 1976. Treatment of subspecies approaching generic variation in proportions is in agree- species status. Syst. Zool., vol. 25, no. ment on the two graphs even though the 2, pp. 161-167. measurement of general size employed was, Brodkorb, P. as noted, not the same. Many ofthe less pro- 1964. Catalogue of fossil birds, part 2. Bull. nounced variations in proportions, especially Florida State Mus. Biol. Sci., vol. 8, no. when they are in agreement on the two graphs 3, pp. 195-335. will probably be demonstrated to be signifi- Brown, L. cant later when more adequate series ofmea- 1972. The breeding behavior of the African surements are available. Harrier Hawk Polyboroides typus in The type species of each genus is included Kenya. Ostrich, vol. 43, pp. 169-175. Brown, L., and D. Amadon on the graphs except for Butastur where very 1968. Eagles, hawks and of the world. poor material of the type, B. teesa, was at Hamlyn, London; McGraw-Hill, New hand. When more than one species ofa genus York. 2 vols., 945 pp. is included it is to illustrate some point, for Burton, P. J. K. example the relatively long legs of Buteogal- 1978. The intertarsal joint of the Harrier- lus urubitinga. The actual measurements Hawks Polyboroides spp. and the Crane used are given in table 1. Usually an average Hawk Geranospiza caerulescens. Ibis, are vol. 120, pp. 171-177. of the two sexes (females invariably Chapman, F. M. larger) was used, but in a few instances where 1940. Post-glacial history of Zonotrichia ca- weights were of one sex only all the data are pensis. Bull. Am. Mus. Nat. Hist., vol. for that sex as indicated. Only means are 77, art. 8, pp. 381-438. given, complete measurements are in the Cooper, J. E. works cited above supplemented by various 1980. Additional observations on the intertar- The sal joint of the African Harrier-Hawk sources, as noted, especially for weights. Polyboroides typus. Ibis, vol. 122, pp. body length measurements were taken from 94-98. specimens in the American Museum of Nat- Dickey, D. R., and A. J. van Rossem ural History. Often only one or two speci- 1938. The birds of . Field Mus. mens in a suite ofa dozen or so were carefully Nat. Hist. Zool. Ser., vol. 23, 609 pp. 20 AMERICAN MUSEUM NOVITATES NO. 2741

Friedmann, H. Plotnick, R. 1950. Birds of North and Middle America, 1956. Posicion sistematica del genero Heter- Falconiformes. U.S. Natl. Mus., Bull. ospizias. El Hornero, vol. 10, pp. 50, pt. 11, 793 pp. 136-139. Hamerstrom, F., and F. Hamerstrom Ridgway, R. 1978. External sex characters ofHarris' Hawks 1873. Catalogue of the ornithological collec- in winter. Raptor Res., vol. 12(1/2), pp. tion in the Museum ofthe Society. Proc. 1-14. Boston Soc. Nat. Hist., vol. 16, pp. Harrison, E., and L. Kiff 43-106. 1977. The nest and egg of the Black Solitary 1874. History of North American birds, In Eagle. Condor, vol. 79, pp. 132-133. Baird, S. F., T. M. Brewer and R. Ridg- Hartert, E. way (eds.), Land birds, vol. 3, 560 pp. 1912-1921. VOgel palaarktischen fauna, vol. 2, Boston, Little Brown. part 9, pp. 1089-1216. 1876. Studies of the American Falconidae. Hellmayr, C., and H. Conover Bull. U.S. Geol. and Geog. Surv. Terr., 1949. Catalogue ofbirds ofthe Americas. Field vol. 2, pp. 91-182. Mus. Nat. Hist. Zool. Ser., vol. 13, pt. Romer, A. S., and L. W. Price. 1, no. 4, 358 pp. 1940. Review of the Pelycosauria. Geol. Soc. Jackson, F. J. (completed by W. L. Sclater) of Amer., Special Paper no. 28, 538 pp. 1938. The birds of Kenya Colony and the Sharpe, B. Uganda Protectorate, 1. Gurney and 1874. Catalogue of birds in the British Mu- Jackson, London, 542 pp. seum, vol. 1. British Museum (Nat. Johnson, N. K., and H. J. Peeters Hist.), London, 480 pp. 1963. The systematic position ofcertain hawks Shufeldt, R. in the genus Buteo. Auk, vol. 80, no. 4, 1919. Osteological and other notes on the pp. 417-446. Monkey-eating Eagle ... Pithecophaga Jollie, M. jefferyi Grant. Phil. Jour. Sci., vol. 15, 1976-1977. A contribution to the morphology pp. 3 1-58. and phylogeny of the Falconiformes. Sick, H., and D. M. Teixeira Evol. Theory, vol. 1(1976) pp. 285-298 1977. The egg of the Crowned Solitary Eagle. (pt. 1); vol. 2 (1977) pp. 115-300 (pts. Condor, vol. 79, p. 133. 2 and 3); vol. 3 (1977) pp. 1-141 (pt. 4). Simpson, G. G. Mader, W. J. 1941. Large pleistocene felines ofNorth Amer- 1975. Extra adults at Harris' Hawk nests. Con- ica. Am. Mus. Novitates, no. 1136, 27 dor, vol. 77, pp. 482-485. pp- Meinertzhagen, R. Slud, P. 1959. Pirates and predators. Edinburgh, Oliver 1964. The birds of . Bull. Amer. and Boyd, 230 pp. Monroe, B. L., Jr. Mus. Nat. Hist., vol. 128, 430 pp. 1963. Three new subspecies of birds from Stresemann, E., and D. Amadon Honduras. Occas. Papers Mus. Zool., 1979. Falconiformes, In Mayr, E. and G. W. Louisiana State Univ., vol. 136, 7 pp. Cottrell (eds.), Check-list ofbirds of the 1968. Distributional survey of the birds of world, vol. 1., revised, Cambridge, Honduras. Am. Ornith. Union Ornith. Mus. Comp. Zool., pp. 270-425. Monographs, no. 7, 458 pp. Thiollay, J. M. Niles, D. M. 1978. Comparaisons entre les peuplements de 1979. [Review ofl Comparaisons entre les peu- Falconiformes des plaines cutieres du plements de Falconiformes des plaines Mexique et de Cote-d'Ivoire. Gerfaut, cutieres du Mexique et de Cote-d'Ivoire vol. 58, pp. 139-162. by J. M. Thiollay. Auk, suppl. 96, pp. Voous, K. H. 18b-19b. 1969. potential in birds ofprey from Olson, S. Surinam. Ardea, vol. 57, pp. 117-148. [In press] The distribution of fused phalanges Wetmore, A. ofthe inner toe in the Accipitridae. Bull. 1933. Status of the genus Geranoaetus. Auk, British Ornith. Cl. vol. 50, p. 212. Peters, J. L. 1965. The birds of the Republic of Panama, 1931. Check-list ofbirds of the world, 1. Har- pt. 1. Smithsonian Misc. Coll., vol. 150, vard Univ. Press, Cambridge, 345 pp. 483 pp.