PROC. BIOL. SOC. WASH. 107(3), 1994, pp. 421-428 DESCRIPTION OF A NEW OF SUBFOSSIL EAGLE FROM MADAGASCAR: STEPHANOAETUS (AVES: FALCONIFORMES) FROM THE DEPOSITS OF AMPASAMBAZIMBA

Steven M. Goodman

Abstract.— StwcvdiX bones recovered from subfossil deposits at Ampasam- bazimba, Madagascar, are described as a new species of eagle, Stephanoaetus mahery. These bones presumably date from the Holocene. S. coronatus, oc- curring on the African mainland, is the only extant member of this , and throughout much of its range feeds extensively on primates weighing up to about 12 kg. Bones of a remarkable assortment of lemurs have been recovered

from Ampasambazimba and it is suggested that S. mahery may have preyed upon primates. Further, the legendary Rokh, a giant reputed to have occurred on Madagascar, from the tales of Sinbad and Marco Polo, may have been derived from S. mahery. Resume.— Une nouvelle espece d'aigle, Stephanoaetus mahery, fait I'objet d'une description realisee a partir de divers ossements provenant des depots subfossiles d'Ampasambazimba, Madagascar. L'age estime de ces ossements est de I'holocene. Stephanoaetus coronatus, qui se rencontre sur le continent africain, est actuellement I'unique espece connue dans ce genre. Cette espece

se nourrit communement de primates pesant environ 1 2 kg. De remarquables

ossements de lemuriens ont ete retrouves a Ampasambazimba et il a est suggere que S. mahery pourrait etre un predateur de primates. Par ailleurs, la legende de I'oiseau rock, celebre oiseau-geant ayant ete vecu a Madagascar dont on trouve des details dans I'Histoire de Sinbad le marin et de Marco Polo pourait etre celle de S. mahery. Famintinana. -TsLolambiby maro hita avy tao amin'ny farita misy taratsie- fan-taolambiby ao Ampasambazimba, Madagasikara, no noheverina fa kara- zamboromahery vaovao, Stephanoaetus mahery. Araka ny tombatombana mi- kasika ireo taolana ireo, dia efa hatramin'ny vanimpotoana jeolozika "Holocene" no nisian'izy ireo tao ambanin'ny tany. S. coronatus, hita any amin'ny kaoti- nanta afrikana, no hany karazana mitovy amin'io mbola velona ankehitriny izay mivelona ara-dalana amin'ny fihinanana karazana varika milanja sahabo ho 12 kg eo ho eo izay hitany ao amin'ny faritra iainany. Mbola nahitana karazan-taolambiby voafaritra ho varika ihany koa tao Ampasambazimba, noho izany azo heverina fa io S. mahery io dia nihaza sy nihinana karazam- barika. Ankoatr'izany, ilay angano mikasika an'ilay hoe "Rokh" na ilay vo- rombe malaza niana tany Madagasikara, nalaina avy tao amin'ny Tantaran'i Sinbad sy Marco Polo dia mety tsy ho hafa fa'ity atao hoe S. mahery ity ihany.

The modem bird fauna of Madagascar is course of the last few millennia the island well-known for its high level of endemism, has undergone drastic ecological change, a of the 201 extant resident species 105 (52%) portion of which clearly is human induced are endemic (Langrand 1990). During the (Perrier de la Bathie 1921, Battistini& Ver- 422 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON

Fig. 1 . Tarsometatarsus of Stephanoaetus mahery, new species, holotype, MAD 549 1 , Museum National d'Histoire Naturelle, Service de Paleontologie, Paris, left dorsal and right ventral views.

in 1972, Bumey 1987). In general, little is and successionally infilled to form a marsh known about what effects these changes have (MacPhee et al. 1985). During excavations had on the avifauna. The best known group at Ampasambazimba a large number ofbird of extinct on the island is the elephant bones were recovered, which until now re- birds (Family Aepyomithiformes) which mained largely unstudied. The majority of consisted of at least seven species in two this material consists of waterbirds, but different genera (Brodkorb 1963). The exact bones of raptorial birds were also recovered. sequence of events that led to their extinc- Within this material are elements of a large tion is unknown, but at least one species eagle previously unknown to science and may have been extant until the turn of the here it is proposed to call it: 17th-century, about the same time Euro- peans arrived on Madagascar (Flacourt Stephanoaetus mahery, new species 1658, p. 165). Other smaller bird species Figs. 1, 2 are known to have gone extinct in the past few millennia (Milne Edwards & Grandidier Holotype. —heft tarsometatarsus, collec- 1895, Andrews 1897, Goodman & Ra- tions of the Laboratoire de Paleontologie, voavy 1993). Museum National d'Histoire Naturelle,

One of the best known subfossil localities Paris, MAD 5491 (Figs. 1, 2). The element on Madagascar is Ampasambazimba, lo- is complete except a portion of the hypo- cated about 85 km west of the capital city tarsus is broken. of Antananarivo. The site is at the edge of Loca//Yv- —Ampasambazimba, Madagas- a former lake created by the damming of a car, and accessioned in 1925 as part of a river by a lava flow. Subsequently, the lava collection of over 750 specimens donated barrier was downcut, and the lake drained by "Gouvemment general de Madagascar Fig. 2. Comparison oi Stephanoaetus coronatus (MNHN 1901.209) tarsometatarsus (top) and S. mahery (MNHN MAD 5491) tarsometatarsus (bottom). 424 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON par rintermediaire de M. Samat" and con- proximalia, particularly in reference to the sisting of "ossements d'oiseaux sub-fossiles crista plantaris mediana; configuration of provenant d'Ampasambazimba, Antsirabe, the sulcus extensorius and associated mus- et Morarano pres de Betafo (W. d'Antsir- cle attachments; and proportion and mor- abe)." The locality listed in the general cat- phology of the trochlea metatarsi I, fossa alog for the holotype is Ampasambazimba. metatarsi, foramen vasculare distale, sulcus Before 1925 several different excavations hypotarsi, and crista hypotarsi (medialis and were conducted at this site and it is impos- lateralis). The holotype tarsometatarsus of sible to determine during which season the S. mahery differs from comparative mate- holotype was collected. rial of S. coronatus by being longer {tio = Chronology. —No radiometric date is 2.15, P = 0.028, Table 1). Further, the cris- available for the Stephanoaetus mahery re- tae plantares are more prominent and sulcus mains. They are presumed to be Quater- flexorius deeper in coronatus than mahery. nary, probably Holocene. This is supported Paratypes.— All in collections of the La- by material of Megaladapis grandidieri, an boratoire de Paleontologie, Museum Na- extinct lemur excavated from Ampasam- tional d'Histoire Naturelle, Paris, and con- bazimba, that has yielded a radiocarbon date sisting of distal pedal phalanges (claws)— of 1035 ± 50 years B.P. (Tattersall 1973). first digit of right side (MAD 5428) and left However, since no stratigraphic informa- side (MAD 5423), distal one-quarter of left tion is available for the Ampasambazimba ulna (MAD 5587), and pelvic fragment material reported on herein, it is unknown (MAD 4944). Except for MAD 5428, which if the Megaladapis and Stephanoaetus re- is from Ampasambazimba and is part of the mains were associated. same accession as the holotype, exact col- Measurements of holotype. —Greatest lection localities of the other bones are un- length 108.0 mm; proximal breadth 26.1 known, but are presumably from Quater- mm; and distal breadth 27.9 mm. (See von nary deposits on Madagascar. den Driesch 1 976, fig. 62a-c for illustrations Discussion. —The genus Stephanoaetus and descriptions of the measurements.) was previously unrecorded on Madagascar, Etymology.— The. name mahery is from and S. mahery represents the largest known the Malagasy adjective meaning powerful. to inhabit the island during the This is in reference to the presumed strength Quaternary. The extant raptor community this eagle would have possessed. contains distinctly smaller birds, except for Diagnosis.— The. falconiform tarsometa- the Madagascar Fish Eagle (Haliaeetus vo- tarsus is one of the most distinctive osteo- ciferoides), which in modem times is lim- logical elements of raptors and has impor- ited to the northern and northwestern por- tant systematic characters for diagnosing tion of Madagascar, is one ofthe rarest eagles genera (Jolly 1977). The type specimen of in the world (Langrand 1990). S. mahery is exceptionally robust in both The majority of raptors show sexual size- overall size and proportions. Articular sur- dimorphism with females being larger than faces and muscle attachments are excep- males, as is the case with S. coronatus

tionally prominent. The specimen was com- (Brown et al. 1982). Since comparative ma- pared to all genera of African eagles, and terial of 5. coronatus used in this study is with the exception of Stephanoaetus, these comprised of both sexes, and the holotype are morphologically different. The Ampa- subfossil tarsometatarsus is larger than all

sambazimba tarsometatarsus is referred to of these modern specimens, it is concluded the genus Stephanoaetus by the following that S. mahery does not fall within size vari-

characters (terminology follows Baumel ation of S. coronatus (Table 1). Six of the [1979]): position of foramina vascularia eleven specimens of S. coronatus used in VOLUME 107, NUMBER 3 425

Table 1.— Tarsometatarsus measurements (mm) of the holotype of Stephanoaetus mahery and a series of extant S. cownatus.

Taxa Greatest length Proximal breadth Distal breadth

mahery, sp. nov. MAD 5491 108.0 26.1 27.9

coronatus

Male (n = 2) 99.7, 102.4 22.2, 25.8 23.5, 27.2 Female (n = 2) 100.5, 102.9 25.2, 25.3 26.2, 27.2 Combined (n = 11)> 102.0 ± 2.7 25.1 ± 1.6 27.0 ± 1.4 98.2-105.9 22.0-26.7 23.5-28.9

Combined descriptive statistics presented as mean ± standard deviation and followed by range. this Study were captive birds; no significant ofS. coronatus are important to subdue large difference in size or osteological characters prey. The hind talons ofS. mahery resemble was found between wild and captive indi- those of S. coronatus in absolute size and viduals. The paratype material of 5. mahery massiveness (Fig. 3). Thus, it is assumed is difficult to distinguish from modem S. that S. mahery would have been as formi- coronatus, at least in part due to the frag- dable a predator as modem S. coronatus. mentary nature of some of the bones. Fourteen species oflemurs, seven ofwhich One intriguing question associated with are extinct, have been identified from the the discovery of S. mahery is what did it Ampasambazimba material (Tattersall eat? Food habits of S. coronatus, the only 1973, Dewar 1984). If indeed these species extant member of this genus and which oc- were temporally sympatric, and there is ev- curs throughout much of sub-Saharan Af- idence that they were, this assemblage would rica (Brown et al. 1982), have been studied represent the highest primate species diver- in the Kibale Forest of Uganda (Skorupa sity known for any site, fossil or modem,

1989, Struhsaker & Leakey 1990), where on the island (NicoU & Langrand 1 989), and this immensely powerful eagle feeds on an perhaps anywhere else in the world. Body assortment of relatively large prey. This in- weights of lemurs known from the deposits cludes blue duiker (Cephalophus monticola) of Ampasambazimba and that might have and Peter's duiker (C. callipygus), adults of been preyed upon by S. mahery include the latter species weigh up to 20.5 kg (King- (from L. Godfrey, pers. comm. for extinct don 1982); red colobus (Colobus badius), taxa [*] and Richard & Dewar [1991] for adult males weigh between 9 and 12.5 kg extant taxa): *Archaeolemur edwardsi (24.5 (Kingdon 1974); black and white colobus kg), *Hadropithecus stenognathus (16.7 kg), (C. guereza), adult males weigh up to 10.5 *Mesopropithecus pithecoides (9.7 kg), *Pa- kg (Struhsaker 1975); and a variety of small- chylemur insignis (10.0 kg), Indri indri (6 er primates, carnivores, and birds. By num- kg), and Propithecus spp. (3.6-5.8 kg). ber, over 80% of this eagle's diet at Kibale Primatologists studying behavior ofmod- is composed of primates (Struhsaker & Lea- em lemurs have observed that several large key 1990). diurnal and social species have a strong ste- On the basis of tarsometatarsus length, reotypic response to birds of prey flying the general size of S. mahery is estimated overhead (Sauther 1989, Macedonia 1990). to be comparable to that of S. coronatus, Evolution of this behavior is difficult to ex- and it is assumed that the former species plain in light of little evidence of the extant would have been capable of taking equiv- raptors predating on this group of lemurs, alent size prey. Further, the long hind talons young or old (Goodman et al. 1993). Recent 426 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON

Fig. 3. Comparison of Stephanoaetus coronatus (MNHN I90L209) first digit from left and right side (top) and S. mahery first digit left side (MNHN MAD 5423) and first digit right side (MNHN MAD 5428) (bottom). discovery of another extinct eagle on Mad- determine if some of the bones recovered agascar in the genus Aquila, and the possi- from the site are prey remains of 5'. mahery. biUty that it ate diurnal lemurs, has been The legendary Rokh of numerous Middle interpreted as a potential evolutionary force Age travelers, including Marco Polo, and in the development of this anti-predator from the tales of Sinbad and A Thousand behavior (Goodman 1994). Now that Qua- and One Nights, may have been a bird found ternary deposits in Madagascar have yield- on Madagascar. Several authors have equat- ed remains of Stephanoaetus, an eagle larger ed the Rokh with elephant birds (e.g., La- than Aquila and a presumed primate spe- vauden 1931), while others have rejected cialist, this explanation can be further sup- this idea and identified it as a bird of prey ported. Hunting techniques used by S. co- (Decary 1937, Allibert 1992). If indeed the ronatus such as puncturing the prey's heart Rokh is not totally imaginary, inhabited with its massive and powerful hind talons Madagascar, and was a large raptor, it may or by flight blows (Brown et al. 1982, Leland have been derived from S. mahery. & Struhsaker 1993) may leave no or subtle Comparative material examined.— Few tell-tale marks in the victim's bone remains. skeletons of large eagles exist in the world's However, a detailed study of food remains museum collections, and a substantial por- recovered from 5". coronatus nests and feed- tion of these are those of birds that died in ing sites might reveal some characters that captivity (*). The holotype of S. mahery was would help to diagnose how this eagle dis- compared to modem skeletal material ofthe patches and dismantles prey. These char- following eagles (See Acknowledgments for acters could then be compared to lemur ma- definitions of acronyms.): Haliaeetus vocifer- terial excavated at Ampasambazimba to oides (MNHN not cataloged; UA 77), H. vo- VOLUME 107, NUMBER 3 427

cifer (BMNH 1954.38.51; MNHN 1908.60), figure 2. L. M. A. Rakotozafy drafted the Ichthyophaga ichthyaetus (BMNH not cata- Malagasy summary. L. Godfrey and T. loged), Terathopius ecaudatus (UMMZ Struhsaker graciously provided unpub- 158569), Spilornis cheela (MNHN A4025), lished information. For comments on an Morphnus guianensis (FMNH 106705), earlier draft of this paper I am grateful to Harpia harpyja (BMNH 1862.3.19.14, H. James and D. Steadman. 1862.3.14.19; MNHN *1878.577), Pithe- cophaga jefferyi (BMNH 1910.2.11.1a, Literature Cited *1961.23.1; FMNH *106496), all extant AUibert, C. 1992. Le monde austronesien et la civ- Aquila spp. (excluding A. gurneyi), Hieraae- ilisation du bambou: une plume qui pese lourd: *1 *1 9 tusfasciatus (MNHN 870.326, 10.424), I'oiseau Rokh des auteurs arabes.— Taloha 11: Lophaetus occipitalis (FMNH 313295; 167-181. MNHN 1880.32; USNM 431671), Spizae- Andrews, C. W. 1897. On some fossil remains of carinate birds from central Madagascar. ?M5a>Mafw5 (BMNH 1850.8. 15. 13; FMNH — Ibis 3 (7th series):343-359. *289556), S. ornatus (BMNH 1952.1.177; Battistini, R., & P. Verin. 1972. Man and the envi- * 1889. Ste- FMNH *288164; MNHN 150), ronment in Madagascar: past problems and phanoaetus coronatus (AMNH *4256; problems of today. Pp. 31 1-337 /« R. Battistini BMNH 1952.1.178, 1954.30.42; MNHN and G. Richard-Vindard, eds., Biogeography and ecology in Madagascar. Junk, The Hague. * 190 1.209; NMK OB 125, OB849; USNM Baumel, J. J. (ed.) 1979. Nomina avium. *345669, *346652, *346654, *346655, anatomica Academic Press, London, xxv + 637 pp. 432 1 and Polemaetus bellicosus (BMNH 80), Brodkorb, P. 1963. Catalogue of fossil birds. Part 1 1853.10.21.1, 1952.1.179, 1954.30.43, (Archaeopterygiformes through Ardei- 1957.9.1, 1984.101.1; USNM 430533). formes).— Bulletin of the Florida State Museum 7:179-293. Brown, L. H., E. K. Urban, & K. Newman. 1982. Acknowledgments The birds of Africa. Vol. 1. Academic Press, London, xiii + 521 pp. The following colleagues allowed access Bumey, D. 1987. Late Holocene vegetation changes to or lent material under their care: G. Bar- in central Madagascar.— Quaternary Research rowclough, American Museum of Natural 28:130-143. Decary, R. 1937. La legende Rokh et YAepyor- History (AMNH), New York; R. Pry-Jones, du nis. —Bulletin de 1' Academic Malgache 20 (nou- British Museum of Natural History velleserie): 107-1 13. (BMNH), Tring; C. Lefevre, Laboratoire Dewar, R. E. 1984. in Madagascar: the d'Anatomic Comparee, J.-F. Voisin & J. loss of the subfossil fauna. Pp. 574—593 in P. S. Cuisin, Laboratoire de Zoologie, and D. Martin and R. G. Klein, eds.. Quaternary ex- Goujet, Laboratoire de Paleontologie, Mu- tinctions: a prehistoric revolution. University of Arizona Press, Tucson. seum National d'Histoire Naturelle Flacourt, E. de. 1658. Histoire de la grande ile Mad- (MNHN), Paris; L. Bennun, National Mu- agascar. Paris. seums of Kenya (NMK), Nairobi; B. Rak- Goodman, S. M. 1 994. The enigma of antipredator otosamimanana. Service de Paleontologie, behavior in lemurs: Evidence of a large extinct eagle on Madagascar. — International Journal of Universite d'Antananarivo (UA), Antana- Primatology 15:97-102. narivo; R. Payne and R. Storer, University , & F. Ravoavy. 1993. Identification of bird of Michigan Museum of Zoology (UMMZ), subfossils from cave surface deposits at Anjoh- Ann Arbor; S. Olson, National Museum of ibe, Madagascar, with a description of a new Natural History (USNM), Washington, D.C. giant Coua (Cuculidae: Couinae). — Proceedings ofthe Biological Society ofWashington 106:24- The study was supported by grants from 33. Conservation International, the Field Mu- , S. O'Connor, & O. Langrand. 1993. A review seum of Natural History (FMNH), and of predation on lemurs: implications for the World Wide Fund for Nature. K. Telfer drew evolution of social behavior in small, nocturnal 428 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON

primates. Pp. 51-66 in P. M. Kappeler & J. U. World Wide Fund for Nature, Gland, xvii + Ganzhom, eds., Lemur social systems and their 374 pp. ecological basis. Plenum Press, New York. Perrier de la Bathie, H. 1921. La vegetation mal-

Kingdon, J. 1974. East African mammals. An atlas gache.— Annales du Musee Colonial de Mar-

of evolution in Africa. Vol. I. University of Chi- seille 9:1-266. cago Press, Chicago, x + 446 pp. Richard, A. F., & R. E. Dewar. 1991. Lemur ecol-

. 1982. East African mammals. An atlas of ogy.— Annual Review of Ecology and System- evolution in Africa. Vol. IIIc. University of Chi- atics 22:145-175. cago Press, Chicago, v + 393 pp. Sauther, M. L. 1989. Antipredator behavior in troops Jolly, M. 1977. A contribution to the morphology of free-ranging Lemur catta at Beza Mahafaly and phylogeny of the Falconiformes (part III). — Special Reserve, Madagascar. — International Evolutionary Theory 2 (5):209-300. Journal of Primatology 10:595-606.

Langrand, O. 1 990. Guide to the birds of Madagas- Skorupa, J. P. 1989. Crowned Eagles iSfrep/zawoaeruy

car. Yale University Press, New Haven, xi + [sic] cownatus in rainforest: observations on 364 pp. breeding chronology and diet at a nest in Ugan- Lavauden, L. 1931. Animaux disparus et legendaires da. -Ibis 131:294-298. de Madagascar.— Revue Scientifique, Paris 69: Struhsaker, T. T. 1975. The red colobus monkey. 297-308. University ofChicago Press, Chicago, xiv -1-311 Leland, L., & T. T. Struhsaker. 1993. Teamwork tac- pp.

tics.— Natural History 102:43-48. , & M. Leakey. 1990. Prey selectivity by

Macedonia, J. M. 1990. What is communicated in crowned hawk-eagles on monkeys in the Kibale the antipredator calls of lemurs: evidence from Forest, Uganda.— Behavioral Ecology and So- playback experiments with ringtailed and ruffed ciobiology 26:435-443.

lemurs. -Ethology 86:177-190. Tattersall, I. 1973. A note on the age of the subfossil

MacPhee, R. D. E., D. A. Bumey, &N. A. Wells. 1 985. site of Ampasambazimba, Miarinarivo Prov- Early Holocene chronology and environment of ince, Malagasy Republic— American Museum Ampasambazimba, a Malagasy subfossil lemur of Natural History Novitates 2520:1-6. site. — International Journal of Primatology von den Driesch, A. 1976. A guide to the measure- 6:463-489. ment of bones from archaeological

Milne Edwards, A., & A. Grandidier. 1895. Sur des sites. — Peabody Museum Bulletin, no. 1. ossements d'oiseaux provenant des terrains re- cents de Madagascar.— Bulletin Museum du Field Museum of Natural History, Roo- d'Histoire Naturelle, Paris 1:9-11. sevelt Road at Lake Shore Drive, Chicago, Nicoll, M. E., & O. Langrand. 1989. Madagascar: revue de la conservation et des aires protegees. Illinois 60605, U.S.A.