A NEW SPECIES OF PALM (: APODIDAE) FROM THE PLEISTOCENE OF

STORRS L. OLSON National Museumof Natural History, SmithsonianInstitution, Washington,D.C. 20560 USA

ABSTRACT.--Anew speciesof palm swift, Tachornisuranoceles, is described from a late Pleistocenecave deposit in central Puerto Rico, the only Greater Antillean island on which swifts of the genusTachornis are not now resident.The fossilspecies differs most conspic- ously from the living speciesT. phoenicobiain being larger. The extinctionof T. uranoceles probablyresulted from the disappearanceof open, dry savannawith scatteredpalm groves. This corroboratesother evidencethat showsdecreasing aridity in the West Indies to have been a major causeof haititat alterationand extinctionat the end of the Pleistocene.Received 24 August1981, accepted4 November1981.

THE discoveryof fossilsof extinct mammals ing, as BlackboneCave was the site that was and in Puerto Rico in the early part of most intensively screenedfor very small ver- this century (Anthony 1918; Wetmore 1920, tebrates. Nevertheless, had fossils of Tachornis 1922) marked the beginning of concertedpa- been presentin reasonablenumbers in any of leontologicalstudies of Antillean vertebrates. the other sites, some of the larger skeletal ele- In 1976 and 1977, expeditionsconducted under ments, such as carpometacarpi and ulnae, the auspicesof the SmithsonianInstitution lo- would almost certainly have been recovered. catedmany additionalfossil deposits in Puerto The depositsfrom BlackboneCave are believed Rico, some of which appear to be older than to be among the oldest yet encountered in any of those previouslyreported. The use of Puerto Rico, and they have yielded other more refined collectingtechniques permitted speciesthat are lackingin the majority of Puer- the recoveryof bonesof very smallvertebrates. to Rican fossil sites (Pregill 1981, Olson and As a consequence,the number of taxa known McKitrick 1982), implying that these species as fossilswas greatlyexpanded. The fossilam- became extinct before the other deposits phibians and reptilesof PuertoRico have been formed. analyzed in detail by Pregill (1981), whose Fossils at Blackbone Cave were originally publicationshould be consultedfor informa- depositedin owl pellets(Pregill 1981), a few of tion on the geology, physiography, and ta- which were found still intact. These were phonomy of the fossil sites. Although the doubtlesscast by the extinctbarn owl Tyto ca- thousands of new specimens of birds have vatica (Wetmore 1920, 1922). This owl was a been only partially identified, it is alreadyev- very proficient and opportunistic predator, as ident that the collections contain a number of may be inferred from the hummingbirdsand taxa not previously known from Puerto Rico. swifts, as well as many other speciesof birds, Many of theseare living speciesthat are found bats, insectivores,reptiles, and amphibians, in elsewhere in the West Indies, but at least three the deposits. represent undescribed, endemic taxa: a new genusand speciesof emberizinefinch (Olson MATERIALS AND METHODS and McKitrick 1982), a very small, delicateform of burrowing owl (Athene)(see Pregill and Ol- Some of the fossils, particuarly ulnae and carpo- metacarpi, were recovered at the fossil site with the son 1981), and a new speciesof palm swift of use of 1/8-inch (0.3-cm) mesh screen, but the smaller the genusTachornis, described herein. specimenswere obtained by transporting135 kg of Although fossilshave been collectedin many screenedmatrix to the laboratory and passing it different caves in Puerto Rico, bones of Tach- through finer mesh (1.5 mm or less). The resulting orniswere found only in one of these--Black- concentratewas pickedwith the aid of a magnifying bone Cave. This may be an artifact of collect- lamp and dissecting microscope. In addition to

230 The Auk 99: 230-235.April 1982 April 1982] FossilPalm Swift 231 bones of Tachornis, this procedure also yielded number of authors (e.g. Lack 1956, Brooke abundant remains of hummingbirds (Trochilidae) 1970) have merged Reinardaand Micropanyp- and many minute specimensof reptiles and am- tila with Tachornis,while keeping Panyptila phibians (Pregil11981).The importanceof usingvery separate.The tarsal morphologyof Reinarda, fine-mesh screensat productive West Indian fossil however, is more specialized than that of sites cannot be overemphasized. The fossil speci- mens of Tachorniswere comparedwith 13 skeletons either Tachornisor Panyptila, which are more similar to each other than either is to Reinarda. of T. phoenicobiain the collectionsof the National Museumof NaturalHistory, SrnithsonianInstitution Thus, it would appear that, if Reinarda and (USNM), American Museum of Natural History Tachornisare merged, Panyptilawould have to (AMNH), and Pierce Brodkorb (PB). Skeletons of be included also. In the present consideration, other generaof swifts in the Srnithsoniancollections the point is moot, as Tachornishas priority were also used in the comparisons.Measurements over the other two names, and the nomencla- were made through a dissectingmicroscope with ture of the Puerto Rican would not be af- dial calipers read to the nearest 0.05 rnm. Specimens fected. to be photographed were first coated with ammo- nium chloride to enhance detail. Tachornisuranoceles, new species

SYSTEMATICS (Figs. 1, 2) Holotype.--Right tarsometatarsus, collec- Family Apodidae Genus Tachornis Gosse 1849 tions of the Department of Paleobiology,Na- tional Museum of Natural History, Smithson- The followingcharacters refer the PuertoRi- ian Institution, USNM 311979 (Fig. ld, e). can fossilsto the genus Tachornis(sensu stricto) Collected28 April 1977 by StorrsL. Olson and and distinguish them from other genera of J. Phillip Angle. swifts: (1) ulna with distinct pointed olecra- Locality.--North-central PuertoRico; "Black- non, unlike Cypseloidinae(see Collins 1976); bone Cave I" (Cueva del Infierno), 1.2 km (2) distal condylesof tibiotarsusnot projecting south of Iglesia Ascension,village of Baraho- far posteriorlyas in Apus and, to a lesserex- na, 2 km northeast of Ciales (18ø20'55"N; tent, in Aeronautes; (3) tarsometatarsus short 66ø26'57'•V). and stout, urnlikethat in Collocalia,Cypsiurus, Chronology.--LatePleistocene. Radiometric or Chaetura; (4) proximal end of tibiotarsus datesindicate a possibleage of between17,000 deflected strongly medially; (5) inner trochlea and 21,000yr B.P. for this deposit;other evi- of tarsometatarsusextending distally well past dencealso supports a Wisconsinanage (Pregill the middle trochlea, and the outer trochlea sit- 1981, Olson and McKitrick 1982). uated well proximad to middle trochlea; (6) Horizon.--Unconsolidated, unstratified cave procoracoidprocess expanded (characters4-6 sedimentsthat probablyformed beneath a for- separate Tachornis,Panyptila, and Reinarda mer roostof the extinctbarn owl, Tytocavatica. from other generaof swifts); (7) shaft of tar- Measurements of holotype.--Length, 7.50 sometatarsusnot as laterally compressed,and mm; proximal width, 2.15 mm; least width of inner and outer trochleae not rotated as far shaft, 1.15 ram; distal width measured diago- posteriorlyas in Reinarda,but similar to Tach- nally acrossthe trochleae,2.25 min. ornisand Panyptila;(8) slitlike roedialproximal Paratypes.--All are topotypes; USNM foramen present, as in Tachornisand Reinarda 311980-312005.Anterior portion of sternum;2 (absent in Panyptila); the fossils agree with left and 1 broken right coracoids;3 left and 2 Tachornisand differ further from Panyptila in right humeri;4 left and 2 right ulnae;4 left and having (9) fenestra in proximal end of tarso- 4 right carpometacarpi;1 broken right femur; metatarsuslarge, (10) distal foramenoval rath- 2 right tibiotarsi. er than more elongate, and (11) postero-prox- Measurementsof paratypes.--SeeTable 1. imal flange of outer trochleanot expanded. Etymology.•reek ouranos,sky, and keles, The highly distinctivetarsal morphology of a racer. The name is proposed as a noun in Panyptila, Tachornis,and Reinarda separates apposition. these genera from all other swifts. (Skeletons Diagnosis.--Differsfrom Tachornisphoenico- of the presumablyrelated genus Micropanyp- bia Gosse1849 as follows: (1) size larger (see tila are not available.) Within this group, a Table 1); (2) posteriorsurface of proximal half 232 STORRSL. ORSON [Auk, Vol. 99

G I=

I I B I=

Fig. 1. Hindlimb elementsof Tachornis:A, B, C, the living speciesT. phoenicobia;D, anterior view of tarsometatarsusof T. uranoceles,new species(holotype, USNM 311979);E, same, posteriorview; F, anterior view of tibiotarsus of T. uranoceles(USNM 311999). Scale = 5 mm. of shaft of tarsometatarsusmore deeply exca- Table 1. There is no overlap in the measure- vated; (3) distal portion of inner trochleaheavi- ments of the coracoid, humerus, tibiotarsus, or er, more bulbous; (4) anterior surface of shaft tarsometatarsus,and only two of the six fossil of tarsometatarsusmore deeply excavated;(5) ulnae fall within the range of variation of T. anterior distal pit of middle trochlea much phoenicobia.All but two of the eight fossil car- deeper; (6) procoracoidprocess of coracoid pometacarpi, however, fall within the upper largerand projectingfarther medially; (7) ster- limits of the living species.This might suggest no-coracoidalprocess of coracoidwider, not as that the carpometacarpusis proportionately pointed or distally protrudent. No consistent shorter in T. uranoceles, but the intramembral differences, other than size, were found in the ratiosof the meanlengths of the wing elements wing elements, except that in certain speci- are identical for both species. mens of T. uranocelesthe processes(e.g. inter- nal tuberosityof humerus)were heavierthan DISCUSSION in any of the spedmensof T. phoenicobia. Remarks.--Thegreater size of Tachornisur- The , Tachornisphoeni- anocelesis evident from the figures and from cobia, is resident on , , and Ja- April1982] FossilPalm Swift 233

A E B F

I I

C (• D H

Fig.2. Wingelements of Tachornis: A, B, C, D, theliving species T. phoenicobia;E, dorsal view of coracoidofT. uranoceles, newspecies (USNM 312001); F,anconal view of humerus ofT. uranoceles (USNM 311995);G,dorsal view of ulna of T. uranoceles (USNM 311988); H,ventral view of carpometacarpus ofT. uranoceles(USNM 311980).Scale = 5 mm. maicabut is knownonly as a casualvagrant to todayprompted Kepler (1971) to speculatethat PuertoRico (Kepler 1971). Kepler's observation theisland must lack suitable ecological condi- demonstratesthat the species is stillcapable of tionsfor thesebirds. As we shallsee, this is dispersingto theisland, and the discovery of quite probably the case. Tachornisuranoceles shows that palm swifts in- One of the betterimpressions of the habitat deed occurred on Puerto Rico in the Pleisto- of Tachornisphoenicobia is given by Barbour cene. The absenceof Tachornisin Puerto Rico (1943:90), who observedthat in Cuba "the lit- 234 Stoaas L. OLso• [Auk, Vol. 99

TABLE1. Length measurements(mm) of skeletalelements of living and fossil speciesof Tachornis.

T. phoenicobia T. uranoceles Element n Range Mean n Range Mean Coracold 13 7.55--8.10 7.85 2 8.45--8.75 8.60 Humerus 13 6.35-7.10 6.85 5 7.15-7.40 7.30 Ulna 13 8.70-9.70 9.35 6 9.35-10.25 9.85 Carpometacarpus 12 11.75-13.25 12.40 8 12.60-13.45 13.05 Tibiotarsus 12 11.05-11.70 11.35 2 12.10-12.50 12.30 Tarsometatarsu s 13 6.60-7.30 6.90 1 7.50 --

tle palm swift is gregarious,and the colonies scribedand illustratedby Sick (1948).The only are scatteredwidely over vast areas of sterile, nest yet reportedfor Micropanyptilawas stated semi-arid grasslandsin which grow scattered to be similar to that of Tachornis(Bond 1956). clumpsof variouspalmetto-like palms. Among Puerto Rico now has a rather depauperate the dry, pendent dead fans of these trees the palm flora. Most of the specieseither grow in swifts stick their watchpocketnests .... col- wet forestor probably do not presentthe right onies do not occur in all of the localities which growth form to be attractiveto Tachornis.Kep- strike one as being most suitable." In , ler (1971:310) mentionsthat the endemicroyal Gosse (1849: 62) described these swifts as oc- palm Roystoneaborinquena appears "to offer curring "over the grass-piecesand savannasof similar ecological conditions" to those of the lowlands,the marshy flats at the seaward species of Roystoneaelsewhere. The Puerto mouths of the valleys, as well as the pens of Rican species, however, occurs in "hillsides the mountain slopes." He described nests as and forest... in moist or wet districts" (Brit- being found in coconut palms (Cocos)and ton and Wilson 1923: 112), which does not con- "palmetto (Chaemerops)"[probably = Sabal]. form with the preferenceof Tachornisfor drier, It is now somewhat difficult to determine the open areas. Furthermore, the speciesof Roy- original habitat of T. phoenicobiain Hispaniola stonea do not retain pendent dead fronds and Jamaicabecause the most readily ob- alongside the trunk. served colonies are found in exotic palms in It is possible that one or more species of botanical gardens and parks or even in the palm in which T. uranocelesnested became en- thatched roofs of dwellings. Wetmore and tirely extinct in Puerto Rico or became so re- Swales(1931: 265) mention observingthe birds duced that there were no longer sufficient "alighting among the dead hanging fronds of numbers to support viable populations of the royal palms" (probably not Roystonea, Tachornis.It is worth noting that two species however; see below), and Lack (1976: 276) of palms, Gaussiaattenuata and Sabal causi- statesthat in Jamaicathey are found especially arum,are either nearly restrictedto, or are most in "the thatchpalm Sabaljamaicensis." Orlando abundant in, the arid southwestern part of A. Garrido (pers. comm.) informs me that in Puerto Rico (Britton and Wilson 1923), where Cuba Tachornisnests in jata palms,a nameap- relict dry forest has been able to persist. The plied to a number of speciesof Copernicia, dead fronds are not retained in Gaussia, how- many of which grow in isolatedgroves in open ever, so it is unlikely that palms of this genus country. were ever important in the economy of Tach- I suspectthat the original habitat of Tach- ornis.In PuertoRico today, the lower fronds of ornisphoenicobia is much as portrayedby Bar- Sabal are so consistentlystripped for use in bour (1943)---open,rather arid grasslandor sa- mats and baskets(Robert W. Read, Dept. Bot- vanna with isolatedclumps of palms of a type any, Smithsonian Institution, pers. comm.) that retain their deadfronds hanging alongside that almostno suitablenesting sites remain for the trunk. In Brazil, the closelyrelated swift any individuals of Tachornisphoenicobia that Reinardasquamata nests in exactly similar sit- might potentiallycolonize the island. uations in palms of the genusMauritia, as de- There is considerable evidence to show that April 1982] FossilPalm Swift 235 xeric habitats were more prevalent in Puerto LITERATURE CITED Rico, and in the West Indies generally,during ANTHONY,H. E. 1918. The indigenous land mam- the last glacialadvance (Pregill 1981, Pregill mals of Porto Rico, living and extinct. Mem. and Olson 1981, Olson 1982). At the end of the Amer. Mus. Nat. Hist. n.s. 1: 329-435, pl. 55-74. Pleistocenethe West Indies evidently became BARBOUR,T. 1943. Cuban ornithology.Mem. Nuttall more mesic, with the result that open, arid Ornithol. Club 9: 1-144. habitats contracted or disappeared. This BOND,J. 1956. Nesting of the Pygmy Palm Swift. Auk 73: 457. caused the extinction or reduction in range of BRITTON, N. L., & P. WILSON. 1923. Descriptive diversespecies of vertebrates(Pregill and Ol- flora•Spermatophyta(Part). New YorkAcad. Sci. son 1981).I interpretthe presenceof Tachornis Sci. Surv. PortoRico and Virgin Islands5: 7-158. uranoceles in the Pleistocene of Puerto Rico as BROOKE,R. K. 1970. Taxonomic and evolutionary indicating that open prairie or savanna,with notes on the subfamilies, tribes, genera and isolatedgroves of large palms, occurredin the subgeneraof the swifts (Aves: Apodidae). Dur- area of the caves where the fossils were de- ban Mus. Novitates 9: 13-24. posited.This habitatwas replacedby the Sub- COLLINS,C. T. 1976. A review of the Lower Miocene tropicalMoist Forestthat characterizesthe re- swifts (Aves: Apodidae). Smithsonian Contr. gion today(Pregill 1981, Fig. 3), with the result Paleobiol. 27: 129-132. that the palmsand their attendantpopulations GossE, P. H. 1849. The birds of Jamaica. London, John van Voorst. of Tachorniscould no longer survive. The re- KEPLER,C. B. 1971. First Puerto Rican record of the duction or lossof areas of open savannasuit- Antillean Palm Swift. Wilson Bull. 83: 309-310. able asforaging sites for T. uranocelesprobably LACK, D. 1956. A review of the genera and nesting played as significanta role in the extinctionof habits of swifts. Auk 73: 1-32. that speciesas the lossof nesting sites. --. 1976. Islandbiology, illustrated by the land birds of Jamaica.Oxford, BlackwellSci. Publ. ACKNOWLEDGMENTS OLSON,S. L. (Ed.). 1982. Fossilvertebrates from the Bahamas. Smithsonian Contr. Paleobiol. 48. I am indebtedto the many peoplewho participat- --, & M. C. McKitrick. 1982. A new genus and ed in the Smithsonian fossil collecting expeditions species of emberizine finch from Pleistocene (see Pregill 1981) and must particularlysingle out cave deposits in Puerto Rico (Aves: Passeri- Noel Snyder,who providedthe impetusfor the en- formes). J. Vert. Paleontol. 1. tire projectas well asboundless enthusiasm for find- PREGILL,G. K. 1981. Late Pleistoceneherpetofaunas ing new fossildeposits. Frederick V. Grady, in ad- from Puerto Rico. Misc. Publ. Univ. Kansas dition to taking part in the field work, spent Mus. Nat. Hist. 71: 1-72. countlesshours picking fine concentrate,and the --, & S. L. ORSON.1981. Zoogeographyof West fossilsdescribed here are a testimonyto the keenness Indian vertebrates in relation to Pleistocene cli- of his eye. J. Phillip Angle assistedme on a return matic cycles.Ann. Rev. Ecol. Syst. 12: 75-98. trip to PuertoRico to collectmatrix for fine screening and to obtain skeletal material of recent birds for SICK, H. 1948. The nesting of Reinardasquamata. Auk 65: 169-174. comparativepurposes. This trip was made possible WETmORE,A. 1920. Five new speciesof birds from througha grant from the NationalGeographic So- cave deposits in Porto Rico. Proc. Biol. Soc. ciety. Comparativematerial of Tachorniswas kindly Washington33: 76-81. lent by PierceBrodkorb and CharlesT. Collins.The --. 1922. Bird remains from the caves of Porto photographsof the minute bones are the exacting Rico. Bull. Amer. Mus. Nat. Hist. 46: 297-333. work of Victor E. Krantz. I am also grateful to K. --, & B. H. SWALES. 1931. The birds of JeffreyBickart, Charles T. Collins,Gregory K. Pregill, and the . Bull. U.S. Natl. Robert W. Read, and David W. Steadman for their Mus. 155: 1-483. commentson the manuscript.