Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate...

112
Holocene Vertebrate Fossils from Isla Floreana, Galapagos DAVID W. STEADMAN SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY NUMBER 413

Transcript of Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate...

Page 1: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

Holocene Vertebrate Fossilsfrom Isla Floreana, Galapagos

DAVID W. STEADMAN

SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY • NUMBER 413

Page 2: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

SERIES PUBLICATIONS OF THE SMITHSONIAN INSTITUTION

Emphasis upon publication as a means of "diffusing knowledge" was expressed by the firstSecretary of the Smithsonian. In his formal plan for the Institution, Joseph Henry outlined aprogram that included the following statement: "It is proposed to publish a series of reports,giving an account of the new discoveries in science, and of the changes made from year to yearin all branches of knowledge." This theme of basic research has been adhered to through theyears by thousands of titles issued in series publications under the Smithsonian imprint,commencing with Smithsonian Contributions to Knowledge in 1848 and continuing with thefollowing active series:

Smithsonian Contributions to Anthropology

Smithsonian Contributions to Astrophysics

Smithsonian Contributions to Botany

Smithsonian Contributions to the Earth Sciences

Smithsonian Contributions to the Marine Sciences

Smithsonian Contributions to Paleobiology

Smithsonian Contributions to Zoology

Smithsonian Folklife Studies

Smithsonian Studies in Air and Space

Smithsonian Studies in History and Technology

In these series, the Institution publishes small papers and full-scale monographs that reportthe research and collections of its various museums and bureaux or of professional colleaguesin the world of science and scholarship. The publications are distributed by mailing lists tolibraries, universities, and similar institutions throughout the world.

Papers or monographs submitted for series publication are received by the SmithsonianInstitution Press, subject to its own review for format and style, only through departments of thevarious Smithsonian museums or bureaux, where the manuscripts are given substantive review.Press requirements for manuscript and art preparation are outlined on the inside back cover.

Robert McC. AdamsSecretarySmithsonian Institution

Page 3: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

S M I T H S O N I A N C O N T R I B U T I O N S T O Z O O L O G Y • N U M B E R 4 1 3

Holocene Vertebrate Fossilsfrom Isla Floreana, Galapagos

David W. Steadman

SMITHSONIAN INSTITUTION PRESS

City of Washington

1986

Page 4: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

ABSTRACT

Steadman, David W. Holocene Vertebrate Fossils from Isla Floreana, Galapagos. Smith-sonian Contributions to Zoology, number 413, 103 pages, 25 figures, 4 plates, 12 tables,1986.—This study surveys the late Holocene vertebrate fossil record from Isla Floreana,Galapagos. Over 20,000 fossils from four lava tubes in the arid lowlands near Post OfficeBay are associated with six radiocarbon dates of 2400 years B.P. or younger. The fossils,most of which originated as regurgitated pellets of barn owls, represent more than 1100individual animals of 24 indigenous species. They include six species now extinct onFloreana: Geochelone elephantopus (tortoise), Alsophis biserialis (snake), Tyto punctatissima(Galapagos Barn Owl), Mimus trifasciatus (Floreana Mockingbird), Geospiza nebulosa(Sharp-beaked Ground Finch), and Geospiza magnirostris (Large Ground Finch). Thesespecies are, respectively, 1 st, 7th, 16th, 6th, 15th, and 2nd in abundance among all speciesrecorded as fossils, making up 57% of the individuals in the fossil fauna. In addition, the3rd and 4th most common fossil taxa, Tropidurus grayii (lava lizard) and Zenaida galapa-

f oensis (Galapagos Dove), are extremely rare today on Floreana. Thus, extinction probablyias changed the composition of Floreana's fauna even more than is suggested by the

number of extinct species alone.While the e\idence is circumstantial, I believe that all extinction on Floreana is related

to human impact, directly through predation and habitat alteration, and indirectly throughthe effects or alien animals (rats, mice, cats, dogs, pigs, goats, cattle, and donkeys). Peopleand feral mammals have lived on Floreana since 1832, and most or all vertebrate extinctionoccurred within the succeeding 40-50 years. Direct predation by man and introducedmammals was probably the main cause of extinction only for Geochelone elephantopus,although such predation may have been involved to some extent in each of the otherextinctions. Loss of preferred prey species and human predation probably caused theextinction of Tyto punctatissima. Extinction of Mimus trifasciatus and Geospiza magnirostrismay have resuftea from destruction of Opuntia cactus by feral herbivores. Extinction ofGeospiza nebulosa may be related to habitat changes in the highlands. Buteo galapagoensis(Galapagos Hawk), though not recorded as a fossil, is also apparently extinct on Floreana,with direct human predation likely to be the main cause. All extinction on Floreanaoccurred in historic times; whether this is true elsewhere in the Galapagos awaits moreresearch.

Lasiurus borealis (Red Bat) is the only indigenous mammal recorded as a fossil. Theabsence of fossils of cricetine rodents or Conolophus spp. (Land Iguana) is evidence thatthese forms never occurred on Floreana. The lack of fossils of Coccyzus melacoryphus(Dark-billed Cuckoo) and Dendroica petechia (Yellow Warbler), which occur commonly onFloreana today, is evidence that these two species colonized the Galapagos very recently.

Fossil records depend upon suitable environments of deposition. The Galapagos Islands,being of recent volcanic origin, lack many potentially fossiliferous geological features suchas fine-grained alluvial sediments, indurated sand dunes, or limestone caves and sinkholes.Fortunately, lava tubes, which are very well suited for accumulation and preservation ofvertebrate fossils, are common on certain islands in the Galapagos.

Fossils enable us to reconstruct undisturbed (pre-human) insular faunas more com-pletely than previously possible. Modern biogeographical studies usually do not considernow natural the faunas are; they would benefit by considering changes wrought by humanimpact. The field of island biogeography woula profit from a renewed emphasis on thecollection of original field data rather than the current trend of theoretical studies thatoften are based upon inadequate or poorly understood data.

OFFICIAL PUBLICATION DATE is handstamped in a limited number of initial copies and isrecorded in the Institution's annual report, Smithsonian Year. SERIES COVER DESIGN: The coralMontastrea cavernosa (Linnaeus).

Library of Congress Cataloging in Publication DataSteadman, David W.Holocene vertebrate fossils from Isla Floreana, Galapagos.(Smithsonian contributions to zoology ; no. 413)Bibliography: p.Supt. of Docs, no.: SI 1.27:4131. Vertebrates, Fossil. 2. Paleontology—Recent. 3. Paleontology—Galapagos Islands—

Santa Maria Island. I. Title. II. Series.QL1.S54 no. 413 [QE881] 591 s [566] 85-600002

Page 5: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

Contents

Page

Introduction 1Abbreviations 1Acknowledgments 2

Materials and Methods 3Field Work 3Museum Research 4Modern Specimens Examined 5Place Names 5

Background Studies 6Geology 6Climate 10Paleoclimate 12Vegetation 13

The Fossil Sites 15The Caves 15Stratigraphy 20Chronology 25

The Fossil Fauna 27Systematic Paleontology 37Discussion 60

Human History 60Introduced Mammals 61Extinction 63Theoretical Biogeography and Fossils 76Animals Not Recorded As Fossils 83The Paleontological Potential of the Galapagos 85

Summary and Conclusions 86Literature Cited 89Plates 100

in

Page 6: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have
Page 7: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

Holocene Vertebrate Fossilsfrom Isla Floreana, Galapagos.

David W. Steadman

IntroductionBiologists have long held the Galapagos Islands

in high esteem. Even non-biologists are im-pressed by the living things they see there, espe-cially by the tameness of the native animals.Situated in the equatorial Pacific Ocean, theGalapagos Islands possess a terrestrial flora andfauna that is derived almost entirely from theSouth American mainland, nearly 1000 km tothe east. More than any other locality in theworld, these inhospitable volcanic islands havebeen regarded as a Mecca to which many naturalhistorians sooner or later pay homage. Evolution-ary biologists, biogeographers, and ecologistshave found the Galapagos particularly enticingas an unparalleled "natural laboratory." In at-tempting to reconstruct the evolution of the ver-tebrates that inhabit the Galapagos, biologistshave lacked any paleontological evidence. Dur-ing five trips to the Galapagos since 1978,1 havecollected fossils of reptiles, birds, and mammalsfrom lava tubes on five of the larger islands inthe group—Santa Cruz, San Cristobal, Floreana,Rabida and Isabela. These specimens representthe first serious paleontological effort to docu-ment the evolution, past distribution, and extinc-tion of vertebrates in the Galapagos.

This paper will report comprehensively on thefossils of Floreana, an island with a colorful hu-man history and a tragic biological history. My

David IV. Steadman, Biological Survey, New York State Museum,The State Education Department, Albany, New York 12230.

studies of the fossils from the other four islandsare still incomplete. The fossils described herein,none of which predates with certainty even theHolocene, provide critical new evidence for un-derstanding the extinction and biogeography ofthe vertebrates of Floreana and the entire Gala-pagos Archipelago.

ABBREVIATIONS.—Terminology used through-out this work includes the following:B.P. before present (= before A.D. 1950)CDRS Charles Darwin Research Station, Santa Cruz,

GalapagosITCZ Intertropical convergence zoneMNI minimum number of individuals

The following persons, listed in Table 1, as-sisted in field work on Floreana:

MJCGDDGHH

JRHGMMPDWSENS

Maria Jose CamposGayle DavisDavid GrahamHarvey HelmanJames R. Hill, IIIGodfrey MerlenMiguel PozoDavid W. SteadmanEdward N. Steadman

For species accounts in the "Systemic Paleon-tology" section, the following acronyms are usedfor the four fossil caves. Many fossils were cata-logued as lots, so a single catalogue number mayrepresent many fossils.

CPOI Cueva de Post Office (Inferior)CPOS Cueva de Post Office (Superior)FC Finch CaveBOC Barn Owl Cave

Page 8: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

I examined modern specimens of reptiles,birds, and mammals from the collections thatfollow. These specimens, listed below, are skele-tons (or, for some mammals, skin and skulls)unless stated otherwise.BM(NH) British Museum (Natural History)CAS California Academy of SciencesNMNH National Museum of Natural History, Smith-

sonian InstitutionRIB Robert I. Bowman Collection, San Francisco

State UniversityUCMVZ Museum of Vertebrate Zoology, University of

California, BerkeleyUSNM former United States National Museum, collec-

tions in the National Museum of NaturalHistory, Smithsonian Institution

ACKNOWLEDGMENTS.—My field work in theGalapagos has been funded by Fluid ResearchGrants from the Smithsonian Institution throughS. Dillon Ripley and Storrs L. Olson, with sup-plementary funds from the Graduate StudentDevelopment Fund, University of Arizona. Forpermits and other logistic aid, I thank the staffsof Parque Nacional Galapagos, particularly Mi-guel Cifuentes and Fausto Cepeda, and theCharles Darwin Research Station, especially DonLuis Ramos. For dedicated field assistance in theGalapagos, I am indebted to Maria Jose Campos,Gayle Davis, James R. Hill, III, Godfrey Merlen,Miguel Pozo, Edward N. Steadman, and Lee M.Steadman. Shorter periods of help were kindlyprovided by Jacinto Gordillo, David Graham,Harvey Helman, Paul S. Martin, Mary KayO'Rourke, and Arnaldo Tupiza. The De Roys(Andre, Jacqueline, and Gil), the Devine's (Bud,Doris, and Steve), and the Moore's (Alan andTui De Roy) provided all sorts of advice andassistance during field work. Marsha S. Cox,Minard L. Hall, and Tom Simkin have been veryhelpful in many aspects of my work.

I appreciate the services rendered by the Di-vision of Amphibians and Reptiles, the Divisionof Mammals, and the Division of Vertebrate Pa-leontology at NMNH. For help and companion-ship that is too extensive to detail, I am verygrateful to the people of the Division of Birds,NMNH, and the Laboratory of Paleoenviron-

mental Studies, Department of Geosciences, Uni-versity of Arizona. Museum research was fundedby a Summer Visiting Student Fellowship, a Pre-doctoral Fellowship, and two Scholarly StudiesGrants at the Smithsonian Institution, a NationalScience Foundation Grant (DEB-7923840) toPaul S. Martin, and a National Geographic So-ciety Grant. I thank the following persons whoallowed me to examine specimens: Ian C.J. Gal-braith (BM(NH)), Luis F. Baptista, Barry Roth,and Jacqueline Schonewald (CAS), Robert I.Bowman (RIB), Ned K. Johnson, Victoria M.Dziadosz, and Anne D. Jacobberger (UCMVZ),J. Phillip Angle, Michael D. Carleton, Ronald I.Crombie, Linda K. Gordon, Jeremy J. Jacobs,and Storrs L. Olson (NMNH).

Peter Ballmann kindly translated pertinentportions of Abs et al. (1965) and Steindachner(1876). John E. Cadle, Ronald I. Crombie, andCharles R. Crumly provided assistance with her-petological references. Ellen M. Paige renderedFigures 2, 4, 6, 9, 11, 13, and 15-20. Figure 1is by Lawrence B. Isham. Victor E. Krantz pho-tographed these figures, as well as the plates. Healso printed all other figures. Figures 3, 7, 22,and 23 are based upon photographs by James R.Hill, III, while Figures 8, 10, and 12 were pho-tographed by Edward N. Steadman. I took thephotographs for Figures 5, 14, 24, and 25. Rob-ert J. and Jennifer Emry, Storrs L. Olson andHelen F. James, Jean A. Sammon, and NormanE. and Theresa I. Steadman provided badlyneeded housing during times of maximal anxiety.My knowledge of insular biology has been en-riched by many people, but especially Ronald I.Crombie, Uno Eliasson, Peter R. and B. Rose-mary Grant, Helen F. James, Paul S. Martin,Ernst Mayr, Godfrey Merlen, Storrs L. Olson,Gregory K. Pregill, Clayton E. Ray, Robert P.Reynolds, and Frank J. Sulloway. I particularlyappreciate the help of Gayle Davis with the hu-man history, and of Bruce D. Barnett for hisinformation on introduced animals. Steven M.Chambers and Scott E. Miller graciously volun-teered to identify the land snails and insects,respectively, from the fossil sites. Miller's coor-

Page 9: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413

dination of the insect identifications was donewith assistance and information from L. Burham,T.L. Erwin, R.C. Froeschner, L. Masner, A.F.Newton, L.M. Roth, R.R. Snelling, D.R. White-head, and N.E. Woodley.

This paper represents a revision of my doctoraldissertation in the Department of Geosciences,University of Arizona. Debbie Gaines and Jim I.Mead kindly provided much assistance in dealingwith the academic bureaucracy. I thank the mem-bers of my graduate committee, Paul S. Martin,Owen Davis, C. Vance Haynes, Storrs L. Olson,and Stephen M. Russell, for their helpful criti-cisms of my dissertation. Charles R. Crumly andPeter R. Grant commented on an earlier draftof this manuscript. The final draft was read byW. Ronald Heyer, Paul S. Martin, Storrs L.Olson, and Gregory K. Pregill. This is contribu-tion number 370 of the Charles Darwin Foun-dation for the Galapagos.

Finally, I would like to thank Marsha S. Cox,Paul S. Martin, Storrs L. Olson, and S. DillonRipley for their undying interest in, and supportfor, my paleontological research in the Galapa-gos.

Materials and Methods

Throughout this paper, I will refer to thebones collected in the caves as "fossils." I prefernot to use the term "subfossil," which is oftenassociated with non-mineralized Holocene bones,because many of the specimens from Floreanaare well mineralized and were collected fromwithin sediments. The fossils from Floreana varyin preservation, however, from that of nearlyfresh bone to complete mineralization. It is sim-pler to refer to all of them as fossils than toattempt to distinguish unmineralized frommineralized bones by using the awkward term"subfossil."

FIELD WORK.—I visited Floreana 5 times (25June-9 July 1978, 22 October-1 November1980, 26 December 1980, 10 April 1982, 24, 25May 1983). All fossil collections were made dur-ing the first two trips (Table 1). I also visited the

satellite islands of Champion (4 July 1978, 26October 1980, 24 May 1983) and Caldwell (26December 1980) to observe living plants andanimals. On Floreana, I spent at least 2-3 hoursper day outside of the caves, observing plantsand wildlife, especially the birds.

We (names listed in Table 1) visited the Bahiade las Cuevas region from 22-26 October 1980.The caves of this area, described in Montoriol-Pous and Escola (1975), are not lava tubes butare weathering features in the sides of scoriacones. The sediments in these shallow caves wereessentially unfossiliferous, and seemed to be ofvery little antiquity. Fossils of boney fish, tor-toises (Geochelone elephantopus), and Dark-rumped Petrels (Pterodroma phaeopygta) were col-lected in small numbers from these caves, but theonly organic material commonly found here wasthe trampled dung of feral goats. The limitedfaunal remains from these caves will not be dis-cussed further.

With the maps of Montoriol-Pous and Escola(1975), we searched for the caves they reported,Cueva de Post Office (Inferior and Superior), on25-27 June 1978. Upon finding these lava tubes,we collected fossils for the next 10 days. On 7-9July 1978, we searched for new caves and com-pleted the screening, sorting, and packaging offossils and sediment. We discovered Finch Caveat this time, but collected no fossils there. Wevisited the Post Office Bay region again from 26October to 1 November 1980, collecting fossilsin Cueva de Post Office (Inferior and Superior),Finch Cave, and the newly discovered Barn OwlCave. In each instance, the basic procedure wasas follows:

1. Search for and locate lava tubes.2. If fossiliferous, make maps of the location of

the lava tube and the plan of the floor.3. Collect bones and other organic material on

the surface.4. Look for areas of sediment accumulation

and excavate a test pit, saving sediment sam-ples from all designated levels.

5. Describe the stratigraphy of a wall of thetest pit.

Page 10: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

TABLE 1.—Fossil localities on Floreana (abbreviations of collectors' names explainedin "Introduction").

Locality

Cueva de PostOffice(Inferior)

Edge of inundated zoneRoom 1

Room 2

Room 3

Room 4Excavations 1, 2Excavation 3

Cueva de PostOffice(Superior)

Area 1

Area 2

Area 3Excavations 1, 2Excavations 3, 4

Finch CaveAll areas

Barn Owl CaveRooms 1-3Excavation 1

Collectors

DWS, DGDWS, DG, HH, MPDWS, MJC.JRH, ENSDWS, MPDWS, MJCJRH, ENSDWS, HH, MPDWS, MJCJRH, ENSDWS, HHDWS, MPDWS, MP

DWS, MPDWS, MJCJRH, ENSDWS, MPDWS, MJCJRH, ENSDWS, MPDWS, DG, HH, MPDWS, MJCJRH, ENS

DWS, MJCJRH, ENS

DWS, MJCJRH, ENSDWS, MJCJRH, ENS

Date ofcollection

2Jul 197827Jun 197827Oct 198027Jun 197826Oct 198030Jun 197827Oct 198030Jun 19784Jul 19786Jul 1978

28Jun 197828Oct 198028, 30Jun 197828Oct 198031 Jun 1978ljul 197828Oct 1980

29Oct 1980

30,31 Oct 198031 Oct 1980

6. Screen the sediment from the test pit at aconvenient location outside of the lava tube.

7. If the sediment is highly fossiliferous, thenenlarge the excavation laterally, followingany discernible stratigraphic units.

8. Double-check the original description of thestratigraphic section, taking additional sedi-ment samples if necessary.

9. Line the excavation with a plastic sheet andfill it in with rubble.

10. Screen all remaining excavated sediment ata point outside the cave.

11. Package fossils and sediment samples.12. Transport by boat all collected materials to

CDRS for further sorting and packaging, in

preparation for shipment to NMNH.We used two sizes of screens, XA inch mesh and

'/i6 inch mesh (window screen). All maps weremade by doing a traverse with a Brunton pockettransit and a 20 m line. The vegetation made alonger line impractical outside of the caves.Within the caves, the width of the floor wasmeasured by holding a 25 m steel tape at rightangles to 1 m intervals on the directionally ori-ented 20 m line. These measurements wererounded to the nearest 0.1 m or 0.5 m, depend-ing on limitations of time.

MUSEUM RESEARCH.—At NMNH I unpackedand cleaned the fossils, which then I sorted intobroad taxonomic categories. Precise identifica-

Page 11: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413

tions were made by comparisons with modernskeletons, often aided by a binocular dissectingmicroscope. Measurements were made with dialcalipers of 0.05 mm increments, rounded to thenearest 0.1 mm. The approximately 20,000 fossilspecimens are catalogued in the collection of theDepartment of Paleobiology, NMNH. Repre-sentative samples of the fossils have been pre-sented to the museum of CDRS.

MODERN SPECIMENS EXAMINED.—Rep-tiles: Phyllodactylus galapagensis: USNM223988, 223989. Tropidurus bivittatus: USNM223987. Alsophis barringtonensis: USNM223986.

Birds: Pterodroma phaeopygia: USNM 502214,547924. Puffinus Iherminieri: USNM 488422,498002. Oceanites gracilis: USNM 491398. Pela-godroma marina: USNM 496759. Fregetta tropica:USNM 553239. Nesofregetta albigularis: USNM498012. Halocyptena microsoma: USNM 498395.Oceanodroma tethys: USNM 321612. Oceano-droma castro: USNM 490825, 491205. O. leuco-rhoa: USNM 498389, 498391. O. tristrami:USNM 289202. 0. monorhis: USNM 500261. O.markhami: USNM 497962. O. hornbyi: USNM491400. 0. homochroa: USNM 500220. O. fur-cata: USNM 556268. 0. melania: USNM498407. Nyctanassa violacea: USNM 18028,18501, 318840-318842. Zenaida galapagoensis:USNM 320829. Tyto punctatissima: UCMVZ140963. T. alba: USNM 500619. Pyrocephalusnanus: UCMVZ 130127. Myiarchus magnirostris:UCMVZ 151395. Mimus trifasciatus: BM(NH)1837.2.21.401 (skin), 1899.9.1.3 (skin),1899.9.1.4 (skin); RIB 436SLB, 437SLB. M.macdonaldi: BM(NH) 1899.9.1.6 (skin),1899.9.1.9 (skin); RIB 1424RIB, 1431 RIB;UCMVZ 140968, 140971, 140972. M. parvulus:BM(NH) 1899.9.1.64 (skin), 1899.9.1.68 (skin);RIB 1507RIB, 1509RIB; UCMVZ 140974-140978; USNM 19802, 321068, 321069. Af.melanotis: BM(NH) 1899.9.1.17 (skin). Dendroicapetechia: UCMVZ 130131. Geospiza nebulosa:UCMVZ 93213-93219; USNM 116117. G. fu-liginosa: UCMVZ 93181, 93182, 130277,130278, 130285, 130286, 130296, 130298,130299, 141019-141021; USNM 321070,

345594, 345595. G. fortis: UCMVZ 93140,93141, 93143, 93169-93174, 130137, 130149,130175, 130178, 130181, 130190, 130191,130195, 130198, 130204, 130206, 130247-130249, 141004, 141006; USNM 344819,345593. G. magnirostris: BM(NH)1885.12.14.280 (skin), 1899.9.1.171 (skin);UCMVZ 93084, 130150, 130160, 130164,130170, 140985 (skin), 140993, 140996; USNM291411. G. scandens: UCMVZ 93116-93119,93121-93124; USNM 20734, 345597. G. coni-rostris: UCMVZ 93105. G. crassirostris: UCMVZ93205, 93207, 93208, 130335. G. parvula:UCMVZ 93210, 93211, 130429, 130431,130438, 130454, 130459, 130483, 130490,130508, 141056; USNM 20533. G. pauper:UCMVZ 141054. G. psittacula: UCMVZ130354, 130365, 130367, 130378, 130381,130383, 130386, 130387, 130393, 130397,130402, 130413-130415. G. pallida: UCMVZ93203, 130534, 130544, 130550, 150551,130553, 130554. G. olivacea: UCMVZ 93220,93221, 130563, 130569, 130585; USNM345598.

Mammals: Lasiurus borealis: UCMVZ 65510,101913, 108974, 119904, 126482, 130987,136563, 145006, 145016, 152154. L. ega:UCMVZ 85284, 136048, 140885, 144959,144960. L. dnereus: CAS 13272; UCMVZ145371. L. semotus: UCMVZ 114344, 114345.L. seminolus: UCMVZ 6844. L. floridanus:UCMVZ 70507, 126103. L. intermedius:UCMVZ 84219, 104131. Mus musculus: USNM271417, 361409. Rattus rattus: USNM 664,503767. Felis catus: USNM 253239, 278658. Susscrofa: USNM 651.

PLACE NAMES.—Place names in the Galapagoseither have both Spanish and English versions,or are a combination of the two languages. Thusthe reader must forbear such hybrid combina-tions as "Cueva de Post Office." In this paper, Iwill attempt to use the name that is used mostoften today by residents and scientists in theislands, showing no partiality to Spanish or Eng-lish. For the past 20 years, there has been a trendamong English-speaking peoples to use an in-creased amount of Spanish for place names in

Page 12: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

the Galapagos. I will abide by this trend as muchas possible.

Throughout this paper, I will use the nameFloreana. Like most major islands in the Gala-pagos, Floreana is blessed with several names,reviewed here in an attempt to minimize confu-sion. "Santa Maria," the official Ecuadoreanname for the island, is used only occasionally."Charles," in honor of King Charles II of Eng-land, a Stuart king of the late 17th century, wasthe name used by most English-speaking persons,including buccaneers, explorers, whalers, and sci-entists, until a decade or two ago. The nameCharles is derived from the name "King Charles'sIsland" that was used on the map prepared byWilliam Ambrose Cowley in 1684, and repro-duced in Beebe (1924, fig. 78) and Slevin (1955,map 1; 1959:19). The island named "Santa Mariade l'Aguada" on Cowley's map (this name pre-sumably added after 1684) apparently is alsoFloreana (Markham, 1880; personal observa-tion). The island designated on Cowley's map as"King Charles's Island" is less accurate, in bothlocation and shape, than the one named "SantaMaria de l'Aguada." The name "Isle de Saute"was used for Floreana by Ensign Le Sieur deVillefort of the French frigate Philippeaux in1700. The name "Mercedes" was used by Gen-eral Jose Villamil during his stay on Floreana inthe 1830s and 1840s, according to the accountof Captain Henri Louns, Compte de Gueydon,of the French brig-of-war Le Genie (Slevin,1959:89), although later in the same accountCaptain Gueydon used the name "Mercedes" forIsla San Cristobal. Floreana is the name usedinvariably today by residents and most scientists,regardless of their native tongue. OccasionallyFloreana is spelled "Floriana," especially in worksof the 19th century. The name Floreana is at-tributed to General Villamil, in honor of GeneralFlores, the first president of Ecuador (Wittmer,1961:33). In 1832, Villamil established on Flo-reana the first human colony in the Galapagos.As we shall see, this settlement proved disastrousfor many of the native plants and animals of theisland.

Background Studies

I will introduce the geology, past and presentclimate, and vegetation of Floreana, to providean understanding of the environment in whichthe vertebrates of Floreana once thrived. Asmuch as seems necessary, I will preface the situ-ation on Floreana with more generalized infor-mation on the whole archipelago. The verte-brates themselves have been reviewed in manyscientific and semi-popular publications. Theyneed no introduction here, but certain aspects oftheir biology will be mentioned when appropri-ate.

GEOLOGY.—Perhaps 100 biologists have vis-ited the Galapagos for every geologist who hasmade the same journey. Detailed geological de-scriptions are lacking for most individual islands,and Floreana is no exception. McBirney andWilliams (1969:21-28) provide the only mean-ingful description of the geomorphology of Flo-reana, but their study is based upon only fivedays of field work. General accounts of the ge-ology of the Galapagos are found in Darwin(1869, and earlier editions), Chubb (1933), andWilliams (1966). Many important advances havebeen made in the regional geology of the Gala-pagos during the past twenty years, thanks toplate tectonic theory in combination with tech-niques of age determination. Modern geologicalaccounts that consider plate tectonics, potassium-argon dating, and magnetic polarity determina-tion appear in McBirney and Williams (1969),Hey et al. (1977), Cox (1983), and Simkin (1984).Much of the following is taken from these fourreferences.

The Galapagos Islands are situated on theNazca Plate. They are approximately 30 to 300km south of the Galapagos Spreading Center(Galapagos Island Fracture Zone), which sepa-rates the southward accreting Nazca Plate fromthe northward accreting Cocos Plate (Hey et al.,1977, fig. 1; Cox, 1983, figs. 1, 2). The islandsemerge atop the submarine Galapagos Platform,which is the youngest, highest, and westernmostportion of the Carnegie Ridge. This ridge trends

Page 13: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413

S82°E from the Galapagos nearly to mainlandSouth America, but is obliterated, along with therest of the eastern margin of the Nazca Plate, inthe subduction zone (Chile-Peru Trench) justwest of the South American coast. The CocosRidge trends N43°E from the Galapagos Plat-form toward Costa Rica. There is no evidence,however, that either the Cocos Ridge or theCarnegie Ridge ever served as a land bridge orchain of islands to connect the Galapagos toCocos Island or to the mainland. The GalapagosIslands, as well as the Carnegie Ridge, may haveformed from the Nazca Plate passing over anarea of magma outpouring from the earth's man-tle (a "hotspot"). There is no geological evidencefor subsidence of the Galapagos in general, ashas been suggested by some biologists to explainthe distributions of organisms. Faulting hascaused localized dropping of the land, such as onBaltra, but the regional picture for this young,volcanically active archipelago is not one of sink-ing.

The youngest islands in the group are prob-ably the large western islands of Isabela andFernandina, while the oldest islands are the east-ern ones (San Cristobal, Espanola, Santa Fe,Plaza, Baltra, and northeastern Santa Cruz). Po-tassium-argon dating, magnetic polarity deter-mination, geomorphology, and current volcanicactivity all support the east to west trend ofdecreasing ages of the islands. The oldest islandsprobably have been emergent above the oceanfor 3 to 5 million years, while the youngest islandsare probably less than 1 million years old. Thusthe Galapagos are a very youthful group of is-lands, a fact of utmost importance to those pon-dering the evolution of their flora and fauna.Based on the present rate of eastward movementof the Nazca Plate, the Galapagos probably wereapproximately 200 km west of their present lo-cation when they first began to appear above thesea. They are still moving toward South America,at an estimated rate of 55 mm per year, and thusshould be subducted beneath South America inapproximately 20 million years.

Floreana is located in the south-central portion

of the archipelago (Figure 1). It is the 6th largestisland in the group, having an area of 171 squarekilometers (66 square miles) (Wiggins and Por-ter, 1971). The highest point on Floreana isCerro Paja or Cerro de Pajas (Straw Mountain),which reaches an elevation of 640 meters (2100feet; McBirney and Williams, 1969, pi. 2; Figure2 herein). Although Floreana is roughly circularin shape and attains its highest elevations nearthe center of the island, it is not dominated by asingle large shield volcano with a caldera. Mc-Birney and Williams (1969:21) noted that a cal-dera may have existed at one time on Floreana,but was obliterated by the many younger conesand flows, thus placing Floreana in the "MaunaKea stage" of volcanic development. Floreanahas many parasitic cones, scattered nearly ran-domly, that vary much in size (McBirney andWilliams, 1969, fig. 8). Nearly all of the conesare scoria or lava-scoria cones. In between thesecones are pahoehoe lava flows and many types ofbasaltic ejecta, ranging from ash and cinders upto large boulders. The small satellite islands tothe north and east of Floreana are formed fromtuff or scoria cones that have undergone differ-ing degrees of erosion.

Most of the surface lavas of Floreana are ofthe Bruhnes Normal Polarity Epoch (0.0-0.79million years old, following Johnson, 1982), al-though lavas of the Matuyama Reversed PolarityEpoch (0.79-2.47 million years old) are exposedalong the northwestern, northeastern, eastern,and southern coasts (Cox and Dalrymple, 1966;McBirney and Williams, 1969:22, 106; Craig S.Bow, pers. comm.; Cox, 1983). Within the Ga-lapagos, Floreana is an island of intermediateage, i.e., one that emerged during the MatuyamaReversed Polarity Epoch (Cox, 1983). Floreanathus is roughly equivalent in age to Pinzon, Ra-bida, and Wenman. It is younger than the oldeastern islands, such as San Cristobal, Espanola,Sante Fe, and Santa Cruz, but older than thewestern and northern islands of Santiago, Isa-bela, Fernandina, Genovesa, Marchena, Pinta,and Culpepper. McBirney and Williams(1969:22) said that "almost surely some [of the

Page 14: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

90°00'

00*00'

01*00

.>CULPEPPER 50 100 KMI I I I i l l

60 MILESVSUP

GENOVESA

SEYMOUR NORTE

FERNANDINA

SAN CRISTOBAL

FLOREANAa

00*00'

oi'oo1

91°00" 90°00"

FIGURE 1.—The Galapagos Islands.

lavas on Floreana] were discharged within thelast thousand years, if not within the last fewcenturies." I disagree with this statement, for Ihave never seen fresh lavas on Floreana. Al-though some of Floreana's lavas are still ratherdevoid of vegetation, they are, nevertheless,much more weathered than are any of the knownhistoric lavas in the Galapagos. McBirney andWilliams (1969:22) also mentioned the reportfrom Captain David Porter, of the U.S. frigateEssex, of seeing a volcanic eruption on Floreanain July 1813. Neither McBirney and Williamsnor any other geologists, however, have found

any lava or other ejecta suggestive of this activity.Simkin et al. (1982:98) listed this eruption as"uncertain (more likely Sierra Negra eruption)."Volcano Sierra Negra, on southern Isabela, isthought to have erupted in 1813 (Simkin et al.,1982:98), and Porter's account of the eruption(in Porter, 1822) did not state specifically that itoccurred on Floreana (T. Simkin, pers. comm.).Until contrary evidence is brought forth, thevolcanoes of Floreana should be regarded asextinct. Probably there has been no volcanicactivity on Floreana for at least several thousandyears.

Page 15: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413

SLA ONSLOW (DEVIL'S CROWN)(II)

•TA CORMORANT

PTA DAYLIGHT

RAOA BLACK BEACH

FIGURE 2.—Floreana and its satellite islands (area enclosed by box shown in greater detail inFigure 4; contour intervals in feet; modified from map prepared by CDRS).

Floreana is composed mainly of alkaline olivinebasalts. Richardson (1933:61) listed the followingrock types from Floreana: "basalt with olivineonly phenocrysts, amygdaloidal basalt, basaltscoriae, olivine bronzite lapilli." McBirney andAoki (1966) and McBirney and Williams(1969:26) reported that plagioclase phenocrystsare rare on Floreana, whereas porphyritic olivineis very common. I have found, however, thatphenocrysts of plagioclase, pyroxene, and olivineare all common in the basalts of the Post OfficeBay region. Further, Chesterman (1963) re-ported phenocrysts of plagioclase in a rock fromFloreana (see below), and Chubb (1933:18)found "felspar" (sic) in large broken crystals near

Post Office Bay. Chemical analyses or detaileddescriptions have been published for rock speci-mens from Floreana as follows: Richardson(1933), 4 specimens (porphyritic olivine basalts);Chesterman (1963), 2 specimens (a crystal-vitric-lithic tuff from Cormorant Bay, and a basalt withphenocrysts of plagioclase and pyroxene fromnear Black Beach); McBirney and Aoki (1966)and McBirney and Williams (1969:26-28, 123,131, 132, 158), 4 specimens (a typical alkalineolivine basalt, a dunite inclusion and a peridotiteinclusion from the same basalt, and a chrome-diopside from the same peridotite).

As one would expect in a young volcanic ar-chipelago, sedimentary deposits and rocks are

Page 16: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

10 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

uncommon in the Galapagos. Non-marine sedi-mentary deposits are especially rare. Alluviumoccurs as small, unindurated, high-energy depos-its that are, as one would expect, unfossiliferous.The very porous volcanic soil of the Galapagospromotes infiltration of rain water rather thanrun-off. The only thorough description of soilsin the Galapagos is in Laruelle (1966), basedupon an altitudinal sequence on Santa Cruz.Fine-grained lacustrine sediments are found incrater lakes on San Cristobal, Genovesa, San-tiago, Isabela, and Fernandina. These sedimentsare described by Colinvaux (1968, 1969), How-miller and Dahnke (1969), and Colinvaux andSchofield (1976a,b), the last authors also report-ing on the fossil pollen and spores from El JuncoLake, on San Cristobal. Bogs are found in thehighlands of Santa Cruz, Isabela, Floreana, andperhaps Santiago (Colinvaux, 1968; Hamann,1975, fig. 2).

Localized marine sedimentary rocks occur onover half of the major islands in the Galapagos.Hickman and Lipps (1985) have divided the ma-rine deposits into 6 categories (tuff cones withmarine fossils; limestone and sandstone interbed-ded with basalt flows; terrace deposits above sealevel; beach rock; supratidal talus debris; recentlyuplifted tidal and subtidal rocks and sand) thatcorrespond to those mentioned by Pitt and James(1983). At Punta Cormorant, Floreana, I havecollected beach rock riddled with marine gastro-pods. Previous workers (Hertlein, 1972, and ref-erences therein) have regarded certain of themarine deposits in the Galapagos to be as old asMiocene or Pliocene, but Hickman and Lipps(1985) report that the ages of these depositsrange from only several hundred years up toapproximately 2 million years, and thus all areprobably either Holocene or Pleistocene. Thesedeposits typically are dominated by mollusks,with coelenterates and echinoids also present.Among the mollusks, gastropods usually out-number pelecypods.

The volcanic rocks that make up most of theGalapagos have been regarded as an extremelypoor environment for preservation of terrestrial

fossils. As mentioned above, there is very littlealluvial sedimentation, and no exposures of fine-grained, stratified terrestrial sediment such asoften contains fossil material on continents. Untilrecently, therefore, vertebrate paleontology inthe Galapagos lagged well behind other branchesof geology and biology. Research on fossil ver-tebrates began in the Galapagos only in the pasttwo decades with the realization that caves (lavatubes) can be rich sources of fossils. Niethammer(1964) reported on mammalian bones from owlpellets on Santa Cruz; only the large extinctrodent "Megalomys" (= Megaoryzomys) curioi oc-curred in what might be termed a paleontologicalcontext. Ray and Whitmore (1973) expressed thegreat need for paleontological research in theGalapagos, and through the interest and encour-agement of them and others at the SmithsonianInstitution, various field biologists collectedbones (mainly of M. curioi) from caves on SantaCruz and Isabela in the 1960's. These fossils, andothers of M. curioi more recently collected, havenow been described by Steadman and Ray(1982).

Birds and reptiles, for which the Galapagos areso well known, remained unknown as Holocenefossils until very recently. Based on field work in1978 and 1980, I reported fossil reptiles, birds,and mammals from Santa Cruz, Floreana, andIsabela (Steadman, 1981). I briefly discussed fos-sils of Darwin's finches from Santa Cruz in re-evaluating the evolution and systematics of thatfamous group (Steadman, 1982).

CLIMATE.—The climate of the Galapagos hasbeen reviewed by Alpert (1963), Palmer and Pyle(1966), and Hamann (1979), but we still lack acomprehensive treatment of this subject. TheGalapagos lie in the "dry zone" of the equatorialPacific. They are not subject to the violent stormsthat can readily occur only several hundred milesto the north. The oceanic currents of the easternequatorial Pacific (reviewed by Wyrtki, 1966,1967) strongly affect climate in the Galapagos.Relatively high islands such as San Cristobal,Floreana, Santa Cruz, and Santiago, have a cli-matic gradient characterized by increasing pre-

Page 17: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 11

cipitation and decreasing temperature as the al-titude increases. (The large, very high islands ofIsabela and Fernandina do not quite fit this gen-eralized pattern.) For Floreana, Hamann (1979)reported mean annual precipitations of 214.1mm at the "En la Playa" station (= Black Beach;elevation 4 m; on the western coast) and 806.0mm at the "Asilo la Paz" station (= WittmerFarm; elevation 300 m; on the south-central partof the island). The number of years upon whichthese means were based was not given, nor wereany temperatures. Cruz and Beach (1983) re-ported two years of data on rainfall and winddirection from five weather stations on Floreana.Their results are discussed below in connectionwith El Nino.

There are two principal seasons in the Gala-pagos—the rainy season lasting approximatelyfrom January through April, and the dry seasonthat makes up the rest of the year. The distinc-tion between these seasons is pronounced, inboth the arid lowlands and moist highlands. Therainy season is a period of contrast, as the mostprecipitation as well as the hottest, sunniest daysoccur at this time. The dry season is character-ized by a fog or mist known as "garua." Thefrequency and intensity of the garua varies fromyear to year. Typically, the garua is more persis-tent in the highlands than in the lowlands. Theweather patterns of the Galapagos may be sum-marized as follows (based mainly on Alpert 1963,and Hamann 1979).

Rainy, Hot Season: January-April; ITCZ (In-tertropical Convergence Zone) moves southwardto just north of the Galapagos (usually 1 °-2°N);air pressure is relatively low; relatively warmtemperatures (at sea level, CDRS weather station,Santa Cruz, mean = 25.9°C, max. = 27.9°C,min. = 24.1°C; at 194 m elevation, Bella Vistaweather station, Santa Cruz, mean = 24.0°C,max. = 24.8°C, min. = 22.6°C); temperatureinversion weakens or dissipates; rain showers atmidday, especially in higher parts of the islands;high levels of sunlight; good visibility, especiallyin the morning; relatively low wind velocities;dominant wind direction east or east-northeast;

low cumulus and altocumulus clouds typicallypresent; cumulus clouds moving from easterlydirection.

Dry, Cool Season: May-December; ITCZ iswell north of the Galapagos (approximately10°N); air pressure relatively high; relativelycool temperatures (at sea level, CDRS weatherstation, Santa Cruz, mean = 22.7°C, max. =27.6°C, min. = 18.8°C; at 194 m elevation, BellaVista weather station, Santa Cruz, mean =20.6°C, max. = 24.5°C, min. = 17.8°C); low-level temperature inversion operative; oftenovercast with poor visibility, especially in themorning; usually good visibility in the afternoon,at least in the lowlands; fog, mist, or drizzle(garua) forms in the morning and usually dissi-pates before noon, but often persists in the high-lands; very little measurable precipitation atlower elevations, but garua may produce meas-urable precipitation in the highlands; dominantwind direction southeast or east-southeast; stra-tocumulus and altocumulus clouds typically pres-ent; stratocumulus clouds moving from southerlyor southeasterly direction.

Changes in temperature, whether seasonal ordaily, are not extreme in the Galapagos. Forexample, on low and barren Baltra, the meandaily maximum and minimum for March (thewarmest month) are 31.1 °C and 23.9°C, whilethe same for September (the coolest month) is26.7°C and 18.9°C (Alpert 1963). (I have con-verted all of the figures in Alpert (1963) andCruz and Beach (1983) from Fahrenheit to Cel-sius, and from inches to millimeters.) The rela-tive consistency in temperature is contrasted,however, by a great variability in precipitation.At Wreck Bay on coastal southwestern San Cris-tobal, Alpert (1963) reported that the annualrainfall for 1950-1958 varied from 37.1 to 1424mm (mean = 503 mm), while mean monthlyrainfall varied from 4.3 mm (June) to 151.4 mm(February). Rainfall also varies greatly from yearto year for any given month in the rainy season.Alpert (1963) reported 1.5 mm for February1950, compared to 487 mm for February 1953,and 0.0 mm in April 1952 and 1954, compared

Page 18: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

12 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

to 458 mm in April 1953. Indeed the "rainy"season in the Galapagos can be extremely arid insome years. For the weather station at WreckBay, mean precipitation during 1950-1958 was417 mm in the rainy season (January throughApril) and only 86 mm in the dry season (theremaining 8 months).

"El Nino" refers to the sudden appearance ofanomalously warm surface water in low latituderegions of the central and eastern Pacific, espe-cially along coastal regions of Peru and southernEcuador, westward through the Galapagos. Thewaters of this region are normally much coolerbecause of upwelling of cool sub-surface waters,as well as the influence of the cool surface watersof the northward flowing Peru current. El Ninocan last from several months to nearly two years.It is not an annual event, but occurs at irregularintervals of 1-12 years (Quinn and Burt 1970).The years 1828, 1845, 1864, 1871, 1877-1878,1891, 1904, 1918, 1925-1926, 1929, 1932,1939-1941, 1943, 1953, 1957-1958, 1965-1966, 1972-1973, and 1982-1983 are generallyregarded as years of well-developed El Nino ac-tivity (Alpert, 1963; Quinn and Burt, 1970; Ra-mage, 1975). The actual cause of El Nino iscontroversial and beyond the purposes of thispaper.

El Nino normally sets in during January toMarch, although it can begin as early as Octoberor as late as May (Wyrtki, 1975; Wooster andGuillen, 1974). Sea surface temperatures in theGalapagos are highest from January throughMarch regardless of the presence or absence ofEl Nino, because of increased solar radiationassociated with the austral summer, as well as asouthward shift in December to February of thewarm tropical surface water that otherwise oc-curs north of the Equator (Wyrtki, 1966). Duringtimes of El Nino, however, sea temperatures areextremely high. This warm water is less salinethan the displaced, cooler water. Many climato-logical and oceanographic phenomena of the1972-1973 El Nino, the best documented severeEl Nino until that of 1982-1983, are describedby Wooster and Guillen (1974) and Ramage

(1975). The extremely severe El Nino of 1982-1983 is described in detail by Cane (1983), Ras-musson and Wallace (1983), and Barber andChavez (1983), and many other papers are ap-pearing on this topic.

Heavy rainfall in the Galapagos and coastalPeru and Ecuador is associated with El Nino.Cruz and Beach (1983) have documented thedrastic increase in precipitation on Floreana dur-ing the 1982-1983 El Nino. At Black Beach theyrecorded only 4 mm per month from Junethrough November 1982, but an astonishing 316mm per month from December 1982 throughMarch 1983. Extremely heavy rainfall occurredon mainland Ecuador and throughout the Gala-pagos during the 1982-1983 El Nino, enoughto form streams and pools in areas of the Gala-pagos that never before experienced surfacewater (personal observation). Marchant (1958,1959) noted that the great majority of nesting interrestrial birds of Ecuador's arid Santa ElenaPeninsula occurred directly after periods of sig-nificant rainfall. Rates of successful avian repro-duction were much lower in 1955-1956 than inthe El Nino interval of 1957-1958. The sameheavy rainfall that means highly successful repro-duction for terrestrial birds spells disaster, how-ever, for the marine birds of the Galapagos andcoastal Peru and Ecuador. The most famousbiological consequence of El Nino is a massivemortality in fish (especially anchovies), whoseplanktonic food resources have been altered bythe lack of adequate upwelling of cool, phos-phate-rich waters. This results in large-scale mor-tality among marine birds that feed on the fish(Murphy, 1936:101-108; Wyrtki, 1966;Boersma, 1978). During the 1982-1983 El Nino,the nesting success of land birds in the Galapagoswas truly phenomenal, as was the lack of repro-duction in the seabirds. Many of these eventswere well monitored by biologists, whose reportsshould be forthcoming in the next year or two.

PALEOCLIMATE.—On the basis of sedimentsfrom El Junco Lake on San Cristobal, Colinvaux(1972) postulated that the Quaternary of theGalapagos was characterized by dry glacial inter-

Page 19: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413

vals and relatively wet interglacials. This wassubstantiated further by the pollen and sporesfrom these sediments (Colinvaux and Schofield,1976a,b). The organic sediments of the past10,000 years in El Junco were underlain by un-datable inorganic sediments that contained a lensof organic sediment dated at >48,000 years B.P.The physical and chemical nature of the inor-ganic sediment suggested alluvial deposition in adessicated lake bed, rather than typical lacustrinedeposition. To account for the apparently dryconditions indicated by the inorganic sediments,Colinvaux (1972) theorized that the ITCZ wasnorth of the equator during the entire year inglacial times, and therefore was unable to pro-duce the rains from January to March or Aprilthat characterize its annual southward movementtoday. The unstable, mixing air of the ITCZ isresponsible for any "normal," heavy rainfall to-day in the Galapagos, as well as in adjacent areasto the north where the ITCZ is present for morethan several months. El Nino may be only a veryintense southward movement of the ITCZ, whichwould involve a southward displacement of theEquatorial Countercurrent and the South Equa-torial Current. If this is so, then El Nino probablydid not affect the Galapagos during Pleistoceneglacial intervals, and Colinvaux's theory is cor-roborated. If El Nino has a less obligatory rela-tionship to the ITCZ, such as in the causal theoryof Wyrtki (1975), then El Nino still could haveinfluenced the climate of the Galapagos duringglacial advances.

The suggestion by Colinvaux (1972) of a morenorthern ITCZ in glacial times touched off sev-eral responses by researchers who did not ques-tion that the Galapagos were probably more aridduring glacial intervals, but who instead pro-posed different mechanisms to account for thisaridity. Newell (1973) was the first to challengeColinvaux's theory, suggesting that Pleistocenearidity could be attained through the ITCZ re-maining south of the equator instead of north.Newell's statement is based upon the relationshipbetween modern seasonal, latitudinal tempera-ture gradients, and the modern seasonal posi-

tions of the ITCZ, with the southward dis-placement of ITCZ increasing with increasingtemperature gradients, especially those of theNorthern Hemisphere. Newell (1973) used 18O/!6O data from ice cores to state that latitudinaltemperature gradients at 20,000 years B.P. weremuch higher that those of any season today, thusthe southward displacement of ITCZ at 20,000years B.P. was greater than even the largest south-ward seasonal displacement today. A major flawthat I see in Newell's hypothesis is that a southernmovement of the ITCZ would result in wetter,not drier, conditions in the Galapagos. (For ex-ample, see the rainfall map in Palmer and Pyle,1966.) Houvenaghel (1974) stated that thehigher temperature gradient of glacial times re-sulted in stronger southeasterly trade winds thatincreased both the rate and duration of upwellingof cool water in the Galapagos. This would resultin drier weather than at present, without involv-ing a southward shift in the ITCZ. Simpson(1975) proposed a model for glacial climates inthe eastern tropical Pacific, and in doing so, shediscounted the models of Colinvaux, Newell, andHouvenaghel as being largely unsubstantiatedand ignorant of relevant data. However, Simp-son's "all-encompassing" model called upon theupwelling of cool water in the Galapagos andcoastal South America, just as the model of Hou-venaghel (1974). Although Simpson's method-ology was somewhat different from Houvenagh-el's, the two models do not seem incompatible,and each deserves testing from other workers.

VEGETATION.—The major botanical surveysof the Galapagos are those of Hooker (1847),B.L. Robinson (1902), Stewart (1911), Svenson(1935), Wiggins and Porter (1971), and Hamann(1981). My description of Floreana's diverse veg-

/ etation is based upon the last two references,supplemented by my own observations. The na-ture of the vegetation in the Galapagos is con-trolled largely by precipitation, which in turn iscontrolled mainly by elevation, directional ex-posure, and size of the island (Alpert, 1963). Onthe higher islands, such as San Cristobal, SantaCruz, Floreana, Santiago, Isabela, Fernandina,

Page 20: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

14 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 3.—Arid zone vegetation on Floreana during dry season, October 1980 lookingnortheast toward Punta Cormorant from entrance to Cueva de Post Office (Inferior) (leaflesstrees are Bur sera graveolens).

Pinzon, and Pinta, clear-cut altitudinal changesin vegetation are evident. Wiggins and Porter(1971) divide the vegetation of the Galapagosinto 6 zones: littoral, arid, transition, Scalesia,Miconia, and fern-sedge. Excellent photographsof these vegetational zones are found in Wigginsand Porter (1971) and Hamann (1979, 1981).Floreana may have each of the 6 zones, althoughthe status of the Miconia and fern-sedge zones is \uncertain because of the severe alteration of thehighland vegetation. Hamann (1981:20) dividesthe Galapagos plant communities into 9 broadcategories, most of which have two or moresubdivisions. They are: forest; closed scrub withscattered trees; scrub; steppe forest (woodland);steppe scrub (scrub woodland); shrub steppe sa-

vanna; desert scrub; broad leaved herb vegeta-tion; and closed bryoid vegetation.

All of the fossil sites are in the arid lowlands,so only that type of vegetation will be described.I will stress woody plants because of their impor-tance in the physiognomy of the vegetation. Spe-cies marked with an asterisk (*) are particularlydominant or conspicuous, at least locally. Thearid zone (Figure 3) covers more area on Flo-reana than any other vegetational zone. It occursfrom near the coast (just in from the littoral zone)upslope to elevations from approximately 80-120 m on the south-facing side of the island to200-300 m on the north-facing side. The highestdiversity of shrubs and small trees is found in thearid zone. Most of these species are deciduous,

Page 21: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 15

resulting in a drastic change in the appearanceof the arid zone in the rainy versus the dryseason. Many of the trees and shrubs are spinyand have relatively small leaves. The trees arespaced rather evenly and far apart, giving thisregion an open appearance. Soil occurs only inisolated, shallow pockets; most of the surface isbarren basaltic rock. Two large arborescent cacti(*Opuntia megasperma and *Jasminocereus thouar-sii) occur in this zone and may be the tallestplants present. These cacti have been reduced innumbers by feral mammals. As a result, the tree*Bursera graveolens is by far the most conspicu-ous plant today in the arid zone of Floreana. Thefollowing other species of trees and shrubs occurin the arid zone: Acacia macracantha, A. rorudi-ana, Alternanthera echinocephala, Borreria dis-persa, B. linearifolia, Castela galapageia, Chamae-syce nummularia, C. punctulata, C. viminea, Cler-odendrum molle, Cordia leucophlyctis, *C. lutea, C.revoluta, *Croton scouleri, *Cryptocarpus pyrifor-mis, Desmanthus virgatus, Geoffroea spinosa, Gos-sypium barbadense, Lantana peduncularis, May-tenus octogona, *Parkinsonia aculeata, *Prosopisjuliflora, *Scalesia affinis, *S. villosa, *Scutia pau-ciflora, Vallesia glabra, and *Waltheria ovata.

The Fossil Sites

THE CAVES.—The four fossiliferous caves(lava tubes) are all very close to one another andcan be reached on foot from the Post OfficeBarrel in under 20 minutes (Figures 4, 5). Atourist trail, constructed in 1980, now leads toCueva de Post Office (Inferior). The elevationsof the caves, at their entrances, range approxi-mately from 20 to 50 m. Blockage from roofcollapses made it impossible to reach the trueorigin or termination of the lava tubes. The flowthat contains the four caves is of relatively youth-ful pahoehoe lava thought to be derived from alarge scoria cone approximately 2 miles (3.2 km)SSE of Post Office Bay (McBirney and Williams,1969:22, 25). This lava is of the Brunhes NormalPolarity Epoch, and thus is no older than 0.79million years (see "Geology"). I am not aware of

any potassium-argon age determinations fromthe Post Office Bay region to complement thepaleomagnetic information. Many pressureridges and mounds also occur in the Post OfficeBay region.

Cueva de Post Office was named, mapped, anddescribed by Montoriol-Pous and Escola (1975).Although it is actually a single lava tube, the twoentrances to this cave lead to separate passage-ways ("Inferior" and "Superior") that are blockedfrom each other by a massive roof collapse, withthe southern end of "Inferior" being separatedfrom the northern end of "Superior" by a taluscone of boulders (Montoriol-Pous and Escola,1975, fig. 2A). The entrances are formed bysmaller roof collapses and are such that escapewould be impossible for any large, non-volantanimal that fell into them. Cueva de Post Office(Inferior) (Figures 6-8) runs downslope to thesea, its lower portion being inundated by saltwater. Large boulders of roof spall are verycommon in the southern portion of Cueva dePost Office (Inferior), especially in Room 2 andin the low-roofed area separating Room 3 fromRoom 4. North of Room 2, most of the floor ofthe cave is barren basalt, free of any sediment.In Room 2 and south thereof, much of the flooris often covered by poorly sorted sediment rang-ing in size from clay to boulders. In Cueva dePost Office (Superior) (Figures 9, 10), thick de-posits of poorly sorted clays to boulders coverthe entire floor.

Finch Cave and Barn Owl Cave have not beennamed, mapped, or described previously. MiguelPozo and I discovered Finch Cave (Figures 11,12) in 1978. The entrance to Finch Cave is thehighest roof collapse that I have found in theGalapagos, being a drop of 18 m to the floor ofthe cave. A large, bouldery talus cone of roofspall dominates the area below the entrance. Thesoutheastern edge of the talus cone is met by atalus slope originating from roof spall concen-trated in the southeastern portion of the cave(only partially mapped). As with most thick talusaccumulations in caves on Floreana, these slopeshave a dip of 30 °-35 °. The original basaltic floor

Page 22: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

16 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

BAHIAPOSTOFFICE

• CUEVA DE POST OFFICE (INFERIOR)

CUEVA OE POST OFFICE (SUPERIOR)

FIGURE 4.—Post Office Bay region, showing location of four fossiliferous caves (contourintervals in feet).

FIGURE 5.—Post Office Bay, looking southeast toward Floreana (all fossil caves located betweenbeach and prominent scoria cone just behind beach).

Page 23: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

MN

FIGURE 6.—Plan view of floor of Cueva de Post Office (Inferior) (note entrance (Figures 7, 8)and location of Excavations 1-3 in Room 4 (Figures 15-17); modified from Montoriol-Pousand Escola, 1975).

Page 24: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

is exposed in much of the northern two-thirds ofFinch Cave. Essentially unfossiliferous, fine-grained sediments (mainly clays and silts) occurin the low-lying sections of the North Room ofFinch Cave. These sediments appear to be de-rived from soil that periodically washed into thecave via the roof collapse, for they are darkerand better sorted than the fossil-bearing sedi-ments excavated from the other three caves.

Barn Owl Cave (Figures 13, 14) was discoveredby James R. Hill on 30 October 1980. This caveis more diverse faunally than any of the othercaves, but because we already had scheduled aboat to take us from Floreana on 1 November1980, our paleontological endeavors in Barn OwlCave were limited to two hurried days. Much ofRoom 1 is exposed to the surface by a huge roofcollapse that did not, however, allow as easyaccess to the cave as the entrance in Room 2.Essentially the entire floor of Barn Owl Cave iscovered by some sort of sediment, ranging from

FIGURE 7.—Looking northeast out of entrance of Cueva dePost Office (Inferior), October 1980, from Room 1 towardMaria Jose Campos above ground.FIGURE 8.—Entrance of Cueva de Post Office (Inferior),October 1980, looking south (Maria Jose Campos and Ed-ward N. Steadman wait to enter).

Page 25: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 19

CUEVA DE POST OFFICE(SUPERIOR)

10 METERS

FIGURE 9.—Plan view of Cueva de Post Office (Superior) floor (note Excavations 1-4(1 and 2detailed in Figures 18, 19); entrance (Figure 10) straddles boundary between areas 2 and 3;modified from Montoriol-Pous and Escola, 1975).

FIGURE 10. —Entrance of Cueva de Post Office (Superior) Maria Jose Campos, October 1980;because of diagonal orientation and smaller size, entrance was ineffective as a natural trap forlarge tortoises.

Page 26: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

20 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FINCH CAVE

O 20 M.

SOUTH TALUS SLOPE

UNMAPPED PORTION OF LAVA TUBE—1-

mainly boulders on the talus slopes to variousfine-grained sediments in low-lying areas. Forexample, the talus slopes in Rooms 1-3 are com-posed mainly of angular basaltic pebbles andboulders up to 1.2 m in diameter, whereas muchof the floor of Room 3 is covered by loose, finesediments (clays through sands). Sediments ofmany particle sizes (clay through pebbles) occurnear the edges of the talus slopes, such as atExcavation 1.

STRATIGRAPHY.—Near the entrance, the floorof each cave was littered with hundreds of bones,as well as pieces of wood. I dug five small testpits in the sediments of Cueva de Post Office,stratigraphic profiles of which are shown in Fig-ures 15-19. Careful sorting of the fine fractionof the matrix yielded a fair number of small,often fragmentary, fossils. With minor excep-

FICURE 11.—Plan view of Finch Cave floor.FIGURE 12.—Looking into Finch Cave, October 1980(James R. Hill stands at bottom (arrow), near edge of maintalus cone).

W:

Page 27: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 21

tions, the fossils within the sediments are of thesame species as occur on the surface of the floor.

The sediments from the different test pits arebroadly similar to each other. They are veryinorganic, being derived from basalts, with somesecondary minerals (gypsum, flourite, apatite,minor calcite). De Paepe (1967) and Montoriol-Pous and Escola (1975) reported gypsum, flour-ite, and apatite as secondary minerals from Cuevade Post Office (Inferior). Gypsum has been re-ported elsewhere in the Galapagos from the saltdeposits of Tagus Crater Lake, Isabela (How-miller and Dahnke, 1969).

The occasional indistinct stratigraphic units inthese poorly sorted clays to pebbles are sugges-tive of very brief water transport, although thelack of well-developed laminations, in combina-tion with the angular nature of the clasts and the

FIGURE 13.—Plan view of Barn Owl Cave floor (noteExcavation 1, shown in Figures 20, 21).

FIGURE 14.—Entrance of Barn Owl Cave, May 1983(Gayle Davis and Godfrey Merlen prepare to enter).

I

. •

Page 28: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

22SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

CUEVA DE POST OFFICE(INFERIOR)

EXCAVATION 2

CUEVA DE POST OFFICE(INFERIOR)

EXCAVATION 3

CUEVA DE POST OFFICE(INFERIOR)

EXCAVATION 1

X X X X X XX X X X X

1 0 -

15X X X X

A x x x

u -

10 -

2 0 -

3 0 -

4 0 -

5 0 -

6 0 -

it Si• ° ; : ' • . • . * • ' • ° ' -

~r.~rZ/r~.—/r

$**$&

••.cb<Ox x~&G~arx

• Q'Zyv' x x•O.7?-C!/x X': '9,Q:tX X X

:§••

f X XX X X

X X XX V

FIGURE 15.—Stratigraphic profile of Excavation 1, Cuevade Post Office (Inferior), Isla Floreana, Galapagos (descrip-tion of sediment: light yellowish brown (10 YR 8/4 dry),light brown (7.5 YR 7/6 wet), unlaminated, poorly sorted;slightly silty, slightly sandy gravel; contact poorly defined(dashed line); clasts of basalt and flourite, very angular tosub-angular, averaging smaller in the top 1 cm; slight reac-tion with 10% HC1; no fossils in upper 6 cm; maximumobserved thickness = 11 cm; lower contact basaltic bedrock).FIGURE 16.—Stratigraphic profile of Excavation 2, Cuevade Post Office (Inferior), Isla Floreana, Galapagos (descrip-tion of sediment: light yellowish brown (10 YR 6/4 dry),dark reddish brown (5 YR 3/3 wet), unlaminated, poorlysorted: pebbly, sandy, silty clay; contact poorly defined

(dashed line); clasts of basalt and gypsum, very angular tosub-angular, averaging smaller in the upper 3 cm; reactionwith 10% HC1; maximum observed thickness = 15 cm; lowercontact basaltic bedrock).FIGURE 17.—Stratigraphic profile of Excavation 3, Cuevade Post Office (Inferior), Isla Floreana, Galapagos (descrip-tion of sediment: yellowish brown (10 YR 5/6 to 10 YR 5/4 dry), reddish brown (5 YR 4/4 wet), poorly laminated,poorly sorted; pebbly, slightly sandy, slightly silty clay, alter-nating with slightly pebbly, slightly sandy, slightly silty clay;poorly defined contacts (dashed lines); basaltic clasts, angularto sub-angular; reaction with 10% HC1; maximum observedthickness = 63 cm; partial lower contact basaltic bedrock).

Page 29: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 23

CUEVA DE POST OFFICE(SUPERIOR)

EXCAVATION 1

3 0 -

CUEVA DE POST OFFICE(SUPERIOR)

EXCAVATION 2

1 0 -

o

o .

> • % •

FIGURE 18.—Stratigraphic profile of Excavation 1, Cuevade Post Office (Superior), Isla Floreana, Galapagos (descrip-tion of sediment: light yellowish brown (10 YR 6/4 dry)dark reddish brown (5 YR 3/4 wet), poorly laminated, poorlysorted; pebbly, sandy, silty clay alternating with sandy, siltyclay; contacts poorly defined (dashed lines); clasts basaltic,very angular to sub-rounded; no reaction with 10% HCI;maximum observed thickness = 36 cm; basaltic bedrock notreached).

20-

25-

3 0 -

lFIGURE 19.—Stratigraphic profile of Excavation 2, Cuevade Post Office (Superior), Isla Floreana, Galapagos (descrip-tion of sediment: dark yellowish brown (10 YR 6/4 dry),reddish brown (5 YR 4/4 wet), unlaminated, very poorlysorted, pebbly, sandy, silty clay; no discernible stratigraphiccontacts; clasts basaltic, angular to sub-rounded, more com-mon in upper 2 cm; very slight reaction with 10% HCI;maximum observed thickness = 30 cm; basaltic bedrock notleached).

Page 30: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

24 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

BARN OWL CAVEEXCAVATION 1

1 £°/#"20-

o

Q-UJO

3 0 -

40-

50-

X X X *X X X X

X X X XX ^ X

UNIT A

UNIT B

UNIT C

UNIT C

FIGURE 20.—Stratigraphic profile of Excavation 1, BarnOwl Cave, Isla Floreana, Galapagos (description of sediment:Unit A: brown (7.5 YR 4/4 dry), reddish brown (5 YR 4/4wet), moderately well laminated, poorly sorted, clayey,sandy, pebbly silt, alternating with well-laminated, poorlysorted, clayey, silty, sandy gravel; basaltic clasts very angularto sub-rounded; no reaction with 10% HCI; maximum ob-served thickness = 10 cm; indistinct lower contact (dashedline). Unit B: brown (7.5 YR 4/4 dry), reddish brown (5 YR4/4 wet), unlaminated, moderately well sorted; silty, sandy,pebbly clay; basaltic clasts sub-angular; overall prismatictexture; strong reaction with 10% HCI; maximum observedthickness = 6 cm; sharp lower contact (solid line). Unit C\:Yellowish brown (10 YR 5/6 dry), dark brown (7.5 YR 4/6wet), unlaminated, moderately well sorted, silty, sandy clayto clayey, sandy, gravely silt; basaltic clasts, rarely up to 20cm in diameter, sub-angular to sub-rounded; strong reactionwith 10% HCI; maximum observed thickness = 22 cm;indistinct lower contact (dashed line). Unit Cy. light yellowishbrown (10 YR 6/6 dry), dark brown (7.5 YR 4/6 wet),unlaminated, very poorly sorted, clayey, sandy, gravely silt;basaltic clasts rarely up to 20 cm in diameter, smaller clastsof flourite and calcite, very angular to sub-rounded; strongreaction with 10% HCI; maximum observed thickness = 6cm; lower contact basaltic bedrock).)

poor sorting, suggests that these sediments werenot transported very far. Probably both waterand gravity sliding were involved, but only forshort distances. When one considers that nearlyall of the sediment in the caves is derived eitherfrom weathering products of roof spall or fromloose soils washed into the cave from the surface,it appears that this sediment has seldom beentransported for more than 20 m. There is noground water in the caves, and occasional heavyrains are the only source of water. The only timethat I saw sediment entering the caves was during

the torrential rains of May 1983, when at leastseveral cm of new sediment had been washedinto Barn Owl Cave, and a lesser amount intoCueva de Post Office (Inferior). The sedimentsin the caves of Floreana seem to be unaffectedby rains of ordinary strength and duration. In-stead, the fossils became incorporated into thesediments only during very rare intervals whensheetwash from exceptional rains enters the cavethrough the roof collapse. At least in Cueva dePost Office (especially Inferior), trampling of thesediment by tortoises and humans was also im-

Page 31: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 25

FIGURE 21.—Excavation 1, Barn Owl Cave, October 1980 (test pit on left; excavation enlargedtoward wall, following stratigraphic units).

portant in mixing the bones into the loose sedi-ment.

Excavation 1 of Barn Owl Cave (Figures 20,21) was by far the richest of any excavation onFloreana. The sediment of Excavation 1 wasbasically similar to that of the other caves in itspoor development of lamination, rounding, andsorting. The major difference was that Excava-tion 1 was divisible into several distinct strati-graphic units that were highly fossiliferous. Thesediments of Excavation 1 must have formeddirectly below a roost of Tyto punctatissima.

CHRONOLOGY.—There is little potential fordeveloping a Quaternary alluvial chronology inthe Galapagos because of the scarcity of alluvium.The very porous volcanic soil of the Galapagosis conducive to infiltration of rain water ratherthan run-off. Another shortcoming is the appar-

ent absence of cultural remains from the periodbefore the discovery of the islands by the Spanishin the 16th century. This means that culturalchronologies based on lithics, ceramics, and as-sociated radiocarbon dates are not possible in theGalapagos, except that the presence of any hu-man artifacts or introduced animals is almostsurely indicative of the past 400 years. Severalmore recent faunal datum points can be used toprovide minimum ages for certain deposits. Forexample, historical records show that tortoises(Geochelone elephantopus) became extinct on Flo-reana around 1850. Therefore, tortoise bones inany deposit on Floreana would indicate that thedeposit is at least 130 years old. Snakes (Alsophisbiserialis), hawks (Buteo galapagoensis), barn owls(Tyto punctatissima), mockingbirds (Mimus trifas-ciatus), and two species of Darwin's finches (Geo-

Page 32: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

26 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

spiza nebulosa and G. magnirostris) also becameextinct on Floreana in the mid-1800s, althoughthe timing of these extinctions is less well docu-mented than that of Geochelone. Nevertheless,these species are also useful indicators of theminimum age of a fossil site.

Three of the six radiocarbon dates in Table 2were first reported in Steadman (1981). The onlyother published radiocarbon dates from the Ga-lapagos are associated with palynological studiesof late Quaternary lacustrine sediments in ElJunco Lake, San Cristobal (Colinvaux, 1972; Col-invaux and Schofield, 1976a,b). Unfortunately,no excavations on Floreana have yielded charcoalor plant remains suitable for radiocarbon dating.Each of the dates in Table 2 is from surfacematerial. All ages are corrected for 13C as re-ported. The delta 13C values for these samplesare about as expected; those for the wood sam-ples correspond to values characteristic of thosemodern woody plants that are generally C3 plants(Stuiverand Polach, 1977; DeNiro and Epstein,1978). The delta I3C value for the horny scuteof Geochelone is reasonable for an animal that eatsa wide variety of plants, but dominantly cactus (aCAM plant), as well as a fair amount of tropicalgrasses, many of which may be C4 plants (Stuiverand Polach, 1977; DeNiro and Epstein, 1978).

Potassium-argon age determinations on basaltsfrom the walls of the lava tubes would give amaximum age for the sediments and fossils. Com-bined with the radiocarbon determinations onsurface material, this could bracket the age ofthe sediments and sub-surface fossils. I have notsampled these basalts to determine their suitabil-ity for dating, but recent advances in potassium-argon dating can permit determinations as youngas 30,000 years B.P. or less under favorable con-ditions (Gramlich et al., 1971). Future paleonto-logical endeavors in the Galapagos or any othervolcanic areas should keep this technique inmind. Another technique with promise for fu-ture studies in the Galapagos is the radiocarbondating of land snail shells as described by Good-friend and Hood (1983) and Goodfriend andStipp (1983). Land snails occur commonly in

most vertebrate fossil sites in the Galapagos,where the generally non-carbonate environmentof the basaltic substrate should eliminate theproblem of contamination of the shells by "dead"carbonate.

The radiocarbon data provide no evidencethat any of the fossil sites is older than the Ho-locene. The presence of numerous fossils of his-torically extinct species on the surface means thatthe sub-surface fossils in each of the caves aremore than approximately 130 years old. By su-perposition alone, the fossils in Excavation 1 ofBarn Owl Cave are more than 640±50 years old,while those of Excavation 4 of Cueva de PostOffice (Superior) must be older than 1030±160years B.P. (Table 2). That organic material con-tinued to accumulate in caves during historictimes is indicated by the modern radiocarbondate on Prosopis and the presence of introducedmammals. With the extinction of tortoises andbarn owls, however, the present rate of accu-mulation of vertebrates in the caves of Floreanacan be no more than a trickle compared with therate before human colonization of the island.

The main chronological shortcoming in theFloreana caves is not being able to establish bothstratigraphic and chronological control for thefossils within a single deposit. The sub-surfacelevels consist only of bones and inorganic sedi-ment; they lack plant material suited for radi-ocarbon dating. The surface of the sediments,on the other hand, contains wood or epidermalscutes of tortoises that are excellent for radiocar-bon dating, but lacks the stratigraphic frameworkneeded to relate the dated material unequivo-cally to the fossils or to other datable material onthe surface of the same cave. This uncertainty ofchronology is well demonstrated by the dis-cordant ages of the three radiocarbon samplesfrom the surface of Cueva de Post Office (Infe-rior) and the two samples from Barn Owl Cave.There simply is no way to demonstrate that adated piece of wood is contemporaneous withthe undated bones that lie next to it. Neverthe-less, the chronological discrepancies here aremeasured only in hundreds or at most a few

Page 33: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 27

TABLE 2.—Radiocarbon age determinations from Floreana (all done at the Universityof Arizona Radiocarbon Laboratory).

Number

A-2088

A-2089

A-2090

A-3283

A-2512

A-3280

Material

Burseragraveolens(wood; ~9gm)

Prosopisjuliflora(wood; ~ 1 Ogm)

Geocheloneelepkantopus(horny epidermalscute of carapace; ~7gm)

Burseragraveolens(wood; ~5gm)

Acacia sp.(wood; ~17gm)

cf. Cordia sp.(wood; ~5gm)

Locality

Cueva de PostOffice (Inferior)Room 3: surface

Cueva de PostOffice (Inferior)Room 1: surface

Cueva de PostOffice (Inferior)Room 1: surface

Cueva de PostOffice (Superior)Excavation 4

Barn Owl CaveRoom 3: surface

Barn Owl CaveExcavation 1:surface

Age

990± 120yrs. B. p.delta I5C =-23.7% PDB

80±110yrs. B. p.(modern)delta ISC =-25.4% PDB

310±80yrs. B. p.delta 1SC =-13.9% PDB

1030±160yrs. B. P.delta I3C =-26 .01% PDB

2420±25 yrs. B. p.delta 1SC =-25 .5% PDB

640±50 yrs. B. p.delta ISC =-26.9% PDB

thousands of years; most or all of the fossil faunacan safely be assumed to be late Holocene in age.

The Fossil Fauna

All of the fossil sites on Floreana are in lavatubes with roof collapses that permit the entry ofanimals for potential fossilization. Tables 3-6 listsystematically the fossils from each locality withineach cave, and Table 7 summarizes the faunafrom all of the caves. The fossils originate mainlyor entirely in two different ways—from preyitems of barn owls (Tyto punctatissima), or fromthe roof collapse acting as a natural trap intowhich an animal falls but cannot escape.

Natural trap localities are dominated by tor-toises (Geochelone elephantopus), especially largeindividuals, that fell into the roof collapses andwere unable to escape from the cave. Boneyaccumulations resulting solely from natural traps

are less rich, both in number of species and innumber of individuals, than those derived fromthe regurgitations of barn owls. Fossils tend tobe concentrated near the entrance (roof collapse)of the lava tube in either situation, but in naturaltraps the fossils cluster rather randomly aboutthe entrance, thinning as one goes farther fromthe entrance. In former barn owl sites, the fossilsusually are concentrated below ledges on thewalls that served as the owls' roosts.

Townsend (1928:157, 159) made this graphicdescription of the tortoises that had been trappedin Cueva de Post Office (Inferior).It would be difficult to imagine a more effective trap fortortoises than the well-like entrance to the cave from whichthe skeletons were taken. It must have operated automati-cally as a death trap for centuries. The brushy half-concealedentrance is merely a hole in the ground a dozen feet indiameter and twenty feet deep. With a steep slope at oneside, the unlucky tortoise that tumbled in did not necessarilystrike bottom with a fatal crash, but rather rolled down anincline it could not ascend.

Page 34: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

TABLE 3.—Faunal summary of Cueva de Post Office (Inferior) (E = extinct; I = introduced byman; for each taxon, first number = number of specimens; number in parentheses = minimumnumber of individuals represented by specimens (MNI); introduced species not considered incalculation of totals and percentages).

Species

Osteichthyes, sp. indet.Geochelone elephantopus (E)Phyllodactylus bauriiTropidurus grayiiAlsophis biserialis (E)Puffinus IherminieriNycta nassa violaceaZenaida galapagoensisMimus trifasciatus (E)Geospiza nebulosa (E)G. fuliginosaG. magnirostris (E)G. olivaceaGeospiza, sp. indet.Passeriformes, sp. indet.Aves, sp. indet.

AIus musculus (I)Rattus rattus (I)Equus asinus (I)Mammalia, sp. indet. (I)

TotalPercentage of total

Edge ofinundated

zone:surface

637(2)

37(1)

1(1)120(2)

674(3)12.4(1.2)

Room 1:surface

1096(33)

8(1)

2(1)58(7)

7(3)

1(1)35(8)

2(0)

1(1)1(1)

1209(54)22.2(21.3)

Room 2:surface

286(21)

2(1)1(1)1(1)

11(2)37(3)

13(2)

22(0)

373(31)6.8(12.2)

Room 3:surface

512(16)

31(7)8(3)

42(5)

1(1)

594(32)10.9(12.6)

Room 4: Excavationsurface 1

6(1)

1(1)

141(1)

28(5)6(2)

1(1)5(1)

cf.l(l)

47(11) 141(1)0.9(4.3) 2.6(0.4)

TABLE 4.—Faunal summary of Cueva de Post Office (Superior) (E = extinct; I = introducedby man; for each taxon, first number = number of specimens; number in parentheses =minimum number of individuals represented by specimens (MNI); introduced species notconsidered in calculation of totals and percentages).

Species

Geochelone elephantopus (E)Phyllodactylus bauriiTropidurus grayiiAlsophis biserialis (E)Zenaida galapagoensisMimus trifasciatus (E)Geospiza fuliginosaG. fortisG. magnirostris (E)Geospiza, sp. indet.Passeriformes, sp. indet.Aves, sp. indet.

Mus musculus (I)Rattus rattus (I)

TotalPercentage of total

Area 1:surface

3(2)

3(2)5(2)

51(4)

62(10)2.2(4.1)

Area 2:surface

191(21)

1(1)13(5)2(1)

22(4)36(5)

414(42)

13(3)33(0)

2(1)

1(1)

725(82)26.4(33.7)

Area 3:surface

8(3)

1(1)18(2)

1(1)30(3)

4(1)

6(0)

68(11)2.5(3.0)

Excavation1

198(11)17(5)11(4)44(4)

1(1)1(1)

4(3)2(2)1(0)

3(1)

279(31)10.2(12.8)

Excavation2

17(6)2(1)1(1)8(4)

1(D

3(2)

32(15)1.2(6.2)

Page 35: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 29

Excavation2

204(7)3(2)

1(1)9(2)

3(2)

1(1)1(1)

1(0)

223(16)4.1(6.3)

Excavation3

1(1)1770(25)

44(16)143(20)109(11)

31(11)8(6)

cf.2(2)

23(8)2(2)

1(1)23(2)32(0)

1(1)

2189(105)40.2(41.5)

Total

1(1)3874(103)

47(18)147(23)764(17)142(2)

2(1)162(34)66(17)

2(2)2(2)

156(26)3(3)

1(1)27(3)54(0)

2(2)121(3)

1(1)1(1)

5450(253)

Percentageof

total

0.0(0.4)71.1(40.7)

0.9(7.1)2.7(9.1)

14.0(6.7)2.6(0.8)0.0(0.4)3.0(13.4)1.2(6.7)0.0(0.8)0.0(0.8)2.9(10.3)0.0(1.2)0.0(0.4)0.5(1.2)' .0(0.0)

Excavation3

494(15)11(3)15(3)26(1)

5(2)7(2)

159(8)

12(2)102(0)

cf.l(l)

831(36)30.2(14.8)

Excavation4

345(25)

8(1)

1(1)5(3)4(2)

1(1)1(1)

372(23)

13(1)

750(58)27.3(23.9)

Total

1256(83)32(11)66(16)81(11)37(13)84(16)

1(1)1(1)

1004(81)2(2)

42(8)141(0)

5(2)

2(2)

2747(243)

Percentageof

total

45.7(34.2)1.2(4.5)2.4(6.6)2.9(4.5)1.3(5.3)3.0(6.6)0.0(0.4)0.0(0.4)

36.5(33.3)0.1(0.8)1.5(3.3)5.1(0.0)

The rocky floor of the cave is not wide but leads into afew low passages under the lava, all strewn with dry bonesof tortoises that had crept everywhere in search of an outlet.The brittle remains of the earlier victims had been crawledover repeatedly and gradually broken up by those that wereentrapped subsequently from time to time.

The bleached and bony remains of those not too anti-quated and fragile to be removed, had long lost their darkhorny plates which lay curled and twisted beside them. In adozen of these, both carapace and plastron were practicallyintact, while skulls and leg bones had usually been disturbedand scattered. A considerable amount of broken tortoiseremains had long since become mixed with the soil of thecave floor. The later arrivals lay where they died, their largewhite carapaces showing conspicuously as our flashlightswere turned in their direction.

On 16 January 1929, K.P. Schmidt, S.N. Shur-cliff, and an unnamed Ecuadorean also visitedCueva de Post Office (Inferior) during the CranePacific Expedition of the Field Museum of Nat-ural History. Shurcliff (1930:104) describedtheir activities as follows.

We let ourselves down a dark narrow shaft with a long ropeand came into a large underground cavern. There were nogood tortoise shells in this cavern for Schmidty says that Dr.C.H. Townsend had been there ahead of us but we foundanother adjoining cavern which Dr. Townsend must haveoverlooked for therein we discovered several dozen shells ofthe extinct Charles Island tortoise, three in perfect condi-tion. Schmidty is delighted to secure these shells for themuseum, for this type of tortoise has been extinct for morethan a hundred years.

The "adjoining cavern" described by Shurcliff isprobably Room 3 or Room 4 of Cueva de PostOffice (Inferior) and not Cueva de Post Office(Superior), as the latter has a completely separateentrance from the former and contained veryfew surface remains of large tortoises. Townsendprobably obtained most or all of his tortoisespecimens from Room 2 and the northern partof Room 3 in Cueva de Post Office (Inferior).

I believe that Tyto punctatissima was responsiblefor the deposition of nearly all non-tortoise fossilsin the Floreana caves. This belief is substantiatedby its occurrence in the fossil fauna of Barn OwlCave, these fossils being the first specimens of T.punctatissima ever taken on Floreana. Except fornon-hatchling tortoises, the fossil fauna of each

Page 36: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

30 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

TABLE 5.—Faunal summary of Finch Cave (E = extinct; I = introduced by man; for eachtaxon, first number = number of specimens; number in parentheses = minimum number ofindividuals represented by specimens (MNI); introduced species not considered in calculationof totals and percentages).

Species

Geochelone elephantopus (E)Phyllodactylus bauriiTropidurus grayiiAlsophis biserialis (E)Zenaida galapagoensisMyiarchus magnirostrisMimus trifasciatus (E)Geospiza fuliginosaG. fortisG. magnirostris (E)G. pauperPasseriformes, sp. indet.Aves, sp. indet.Lasiurus borealis

Rattus rattus (I)Mammalia, sp. indet. (I)

TotalPercentage of total

South talusslope:

surface

11(3)

12(2)

1(1)

1(1)

87(5)

1(0)

113(12)13.5(13.2)

Main taluscone,

south slope:surface

11(2)

7(3)

1(1)11(2)

23(1)223(9)

276(18)33.0(19.8)

Main taluscone.

east slope:surface

35(5)

4(1)

29(6)

4(2)

66(7)

4(1)6(0)

1(1)

148(22)17.7(24.2)

Main taluscone.

west slope:surface

9(3)2(1)

21(3)39(1)26(5)

1(1)7(2)2(1)

86(7)

1(1)5(0)

14(0)24(2)

10(2)

237(27)28.4(29.7)

NorthRoom:surface

11(4)

1(1)

4(1)

1(1)20(3)

1(0)

23(2)

61(12)7.3(13.2)

Total

77(17)2(1)

45(10)40(2)71(15)

1(1)11(4)

3(2)24(2)

482(31)

1(1)11(1)20(0)47(4)

10(2)

1(1)

835(91)

Percentage oftotal

9.2(18.7)0.2(1.1)5.4(11.0)4.8(2.2)8.5(16.5)0.1(1.1)1.3(4.4)0.4(2.2)2.9(2.2)

57.7(34.1)0.1(1.1)1.3(1.1)2.4(0.0)5.6(4.4)

cave consists almost entirely of animals that arepotential prey items for barn owls. Although T.punctatissima is extinct now on Floreana, it stilloccurs commonly in lava tubes on several otherislands in the Galapagos, where it is responsiblefor rich deposits of both modern and fossil bones(Niethammer, 1964; Abs et al., 1965; Steadman,1981; Steadman and Ray, 1982; Groot, 1983).The only other owl in the Galapagos is AsioJlammeus, the Short-eared Owl, which is notknown to roost in lava tubes. Owls, and barnowls in particular, are known elsewhere to havebeen important accumulators of bones in caves,occasionally resulting in fossil deposits (for ex-ample, see Davis, 1959; Trost and Hutchison,1963; Guilday et al., 1977; Olson, 1978; Pregill,1981, 1982; Levinson, 1982; Olson and Hilgart-ner, 1982).

Excavation 1 of Barn Owl Cave clearly dem-onstrates the importance of rich, ancient, barnowl roosts in determining the past fauna of an

island. This site yielded many more fossils thanthe excavations in any of the other caves (Tables3, 4, 6, 7), and also provided the first unequivocalevidence that T. punctatissima once occurred onFloreana. The barn owls that deposited the bonesof Excavation 1 sampled the local fauna verythoroughly; except the seabird Puffinus Ihermi-nieri, each native taxon that was recovered as afossil from Floreana occurred in this deposit. Inaddition to T. punctatissima, the following taxaof birds were found in Barn Owl Cave but not inthe other nearby caves: Pterodroma phaeopygia,Oceanodroma castro, Pyrocephalus nanus, Geo-spiza crassirostris, Geospiza scandens, and Geospizaparvula. Of these, only Pterodroma phaeopygiamay not have been brought into the cave by T.punctatissima.

Table 8 depicts the past feeding habits of T.punctatissima on Floreana, considering only na-tive animals that are small enough to be preyitems. Note the large difference in relative abun-

Page 37: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413

dance of reptiles and birds in surface versus sub-surface levels. Reptiles make up only 19.4% ofthe MNI recovered from surface collections, butthey represent 57.0% of the MNI from excava-tions. The same values for birds are 79.0% and42.0%, respectively. These values are remarka-bly similar to each other for all four of the surfacecollections. Those of the excavations from Cuevade Post Office (Inferior) and Barn Owl Caveresemble each other closely, but differ signifi-cantly from those of Cueva de Post Office (Su-perior). This difference is due to the method ofcollection used in Excavation 4 of Cueva de PostOffice (Superior), explained at the end of thissection. Bats are a minor component (less than2%) in either case. Tortoises average much largerin size in the surface collections than from sub-surface levels. Therefore, a lower percentage oftortoises from the surface were potential preyitems of T. punctatissima. I have allowed for thisdifference by including only a small percentageof their MNI in calculating Table 8 (see caption).Even if one were to omit tortoises from consid-eration altogether, the remaining reptiles (= liz-ards and snakes) are still relatively much morecommon in the sub-surface levels (Table 7).Much or all of this difference probably is due toa collecting bias, because persons making thesurface collections had a better sight image forskulls, rostra, and mandibles of Geospiza magni-rostris than for any of the bones of lizards andsnakes. Further, these elements of G. magniros-tris, as well as many of the post-cranial elementsof Zenaida galapagoensis, are larger and moreconspicuous than most bones of lizards or snakes.For these reasons I believe that the excavationsare more truly representative of the actual rela-tive abundances of nearly all the species in thefossil fauna. This idea is discussed further below.A testimony to the importance of one's sightimage is shown by the fact that experienced fieldscientists such as Charles Haskins Townsend andKarl Patterson Schmidt could collect numeroustortoise bones from Cueva de Post Office (Infe-rior) while overlooking entirely the abundantbones of many smaller species.

Barn owls prefer to eat rodents when they areavailable. Rodents and other small mammals usu-ally make up over 90% of the diet in species ofTyto on continents, as seen in innumerable stud-ies, such as Wallace (1948) and Trost and Hutch-ison (1963) in North America, Vernon (1972) inAfrica, De Bruijn (1979) in Europe, Morton andMartin (1979) in Australia, and Herrera andJaksic (1980), Jaksic and Yanez (1980), and Jaksicet al. (1982) in South America, Europe, andNorth America. Birds and reptiles seldom com-prise more than a few percent of the prey itemsof Tyto, although Otteni et al. (1972) found thatthe frequency of mammals in pellets of T. albain southern Texas varied year to year from65.5% to 98.9%, while the frequency of birdsvaried from 1.1% to 34.3%. Davis (1959), Ver-non (1972), and Ruprecht (1979) also reportedindividual cases of birds making up more than10% of the prey items of continental barn owls.

The food habits of T. punctatissima in the Ga-lapagos are better known than for most insularbarn owls. Salvin (1876:494) and Gifford(1919:194) reported grasshoppers in the stom-achs of barn owls collected on Santa Cruz andperhaps other islands. Snodgrass and Heller(1904:266) reported rats (Mus = Rattus) as themain food of barn owls on Isabela. Niethammer(1964) reported only the mammals in pellets ofT. punctatissima and Asioflammeus from the Cas-cajo Mountain region of Santa Cruz, althoughhe did not distinguish which of these two owlswas responsible for which pellets, if indeed heknew. The mammalian prey items were as fol-lows: 2 bats (1 Lasiurus borealis, 1 Lasiurus ciner-eus) and 382 rodents (145 Rattus rattus, 230 Musmusculus, 1 Megaoryzomys curioi, 5 Nesoryzomysindefessus, 1 N. darwini). Absetal. (1965, includ-ing the authors' subsequent manuscript correc-tions in reprints) reported on pellets of T. punc-tatissima from two localities on Santa Cruz (3 kmfrom Academy Bay, and Cascajo Mountain).Most of these pellets appear to be the same asthose reported by Niethammer (1964). Abs et al.(1965) recorded from these pellets 119 Rattusrattus, 224 Mus musculus, 1 Megaoryzomys curioi,

Page 38: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

32 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

TABLE 6.—Faunal summary of Barn Owl Cave (E = extinct; I = introduced by man; for eachtaxon, first number = number of specimens; number in parentheses = minimum number ofindividuals represented by specimens (MNI); introduced species not considered in calculationof totals and percentages).

Species

Osteichthves, sp. indet.Geochelone elephantopus (E)Phyllodactylus bauriiTropidurus grayiiAlsophis biserialis (E)Pterodroma phaeopygiaOceanodroma castroNyctanassa violaceaArdeidae, sp. indet.Non-passerine Aves, sp. indet.Zenaida galapagoensisTyto punctatissima (E)Pyrocephalus nanusMyiarchus magnirostrisMimus trifasciatus (E)Geospiza nebulosa (E)G. fuliginosaG. fortisG. magnirostris (E)G. crassirostrisG. scandensG. parvulaG. pauperG. olivaceaGeospiza, sp. indet.Passeriformes, sp. indet.Aves, sp. indet.Lasiurus borealis

Mus musculus (I)Rattus rattus (I)Felis domesticus (I)Sus scrofa (I)

TotalPercentage of total

Room 1:surface

319(20)2(2)

102(16)144(3)

1(1)

129(9)2(1)

1(1)65(8)

1(1)5(4)3(2)

186(16)2(1)

1(1)2(0)

64(0)19(0)

2(1)

1(1)2(1)1(1)

16(1)

1050(87)9.4(16.7)

Room 2,floor:

surface

61(7)2(1)

14(2)

29(4)

11(2)

1(1)1(1)

149(12)

1(1)

14(0)11(0)2(1)

2(2)cf.3(2)

296(32)2.7(6.2)

Room 2,talus slope:

surface

558(10)

5(2)

1(1)

26(5)

23(4)

2(1)276(26)

1(1)1(1)

17(1)11(0)

921(52)8.3(10.0)

Room 3:surface

23(5)

6(2)2(1)

32(5)

1(1)

11(3)

1(1)1(1)

69(7)

1(1)1(1)

1(0)9(0)

158(28)1.4(5.4)

Excavation1: surface

15(2)

4(1)

KD

9(2)

3(1)

1(1)75(7)

8(3)9(0)

cf.l(l)

125(18)1.1(3.5)

6 Oryzomys species (= Nesoryzomys indefessus and/or N. darwini), 1 Lasiurus borealis, 1 Lasiuruscinereus, 2 Zenaida galapagoensis, 1 Myiarchusmagnirostris, 2 Mimus parvulus, 8 Dendroica pete-chia, 4 Geospiza fuliginosa, 4 G. fortis, 1 G. scan-dens, 3 G. pallida, 12 G. olivacea, and 1 undeter-mined bird. Thus birds made up 9.7% of theprey remains listed by Abs et al. (1965). Harris

(1974:121; 1982:121) summarized the food ofT. punctatissima as "mainly rats and mice, somesmall birds. Eats more crickets, grasshoppers andscorpions than the Short-eared Owl." The veryrich fossil site at Cueva de Kubler, Santa Cruz, isderived from prey remains of T. punctatissima(Steadman, 1981). This fossil fauna is dominatedby native rodents, but also includes many species

Page 39: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 33

Excavation1: test pit

102(6)114(18)559(17)254(5)

1(1)

64(11)

31(10)

1(1)4(3)

1(1)66(7)

1(1)4(0)

75(8)239(0)

1(1)32(5)

cf.l(l)

1517(90)13.7(17.3)

Excavation1: Unit A

2(1)175(3)242(39)778(17)446(9)

60(5)2(1)

3(1)17(2)2(1)5(3)

86(6)

1(1)1(1)1(1)6(0)

51(0)256(0)

2(1)

cf.l(l)

2136(92)19.2(17.7)

Excavation1: Unit B

40(4)79(10)

1372(16)90(2)

12(4)3(1)

68(5)

1(1)46(3)

6(1)

230(0)

1(1)

1948(48)17.5(9.2)

Excavation1: UnitC,

49(1)73(8)

1428(20)109(3)

1(1)

cf.l(l)29(0)34(4)2(1)

1(1)12(4)56(3)

2(1)

1(1)52(5)

cf.l(l)

1(1)11(2)

173(0)263(0)

2(1)

2301(59)20.7(11.3)

Excavation1: Unit C2

1(1)

5(2)13(1)

560(1)7(1)

1(1)

3(1)22(2)

5(3)14(0)28(0)

2(1)

661(14)6.0(2.7)

Total

2(1)1343(59)512(78)

4273(95)1060(26)

1(1)1(1)1(1)1(1)

589(1)402(50)

10(5)

1(1)17(7)

285(38)5(4)

18(13)13(9)

1027(91)5(4)2(2)

1(1)2(2)4(4)

35(6)425(12)

1066(0)12(7)35(8)

8(6)

1(1)16(1)

11113(520)

Percentageof

total

0.0(0.2)12.1(11.3)4.6(15.0)

38.5(18.3)9.5(5.0)0.0(0.2)0.0(0.2)0.0(0.2)0.0(0.2)5.3(0.2)3.5(9.6)0.1(1.0)0.0(0.2)0.2(1.3)2.6(7.3)0.0(0.8)0.2(2.5)0.1(1.7)9.2(17.5)0.0(0.8)0.0(0.4)0.0(0.2)0.0(0.4)0.0(0.8)0.3(1.2)3.8(2.3)9.6(0.0)0.1(1.3)

of birds and reptiles. On islands such as Floreanathat lacked rodents, however, T. punctatissimawas obliged to eat small reptiles, birds, and in-sects, at least until Rattus and Mus were intro-duced. Using "weight %" and not "% individ-uals," Groot (1983) reported that barn owls onSanta Cruz ate 12.03% insects, 0.01% spiders,3.42% birds, and 84.54% mammals, the last cat-

egory represented almost entirely by introducedRattus and Mus. Unfortunately, the species-levelidentifications of birds in Groot (1983) are notreliable, based upon my own examination of thespecimens involved.

Few studies have been made of the modernfeeding habits of barn owls on islands outside ofthe Galapagos. The West Indian species of Tyto

Page 40: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

34 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

TABLE 7.—Faunal summary of all localities (E = extinct; I = introduced by man; for eachtaxon, first number = number of specimens; number in parentheses = minimum number ofindividuals represented by by specimens (MNI); introduced species not considered in calculationof totals and percentages).

Species

Osteichthyes, sp. indet.Geochelone elephantopus (E)Phyllodactylus bauriiTropidurus grayiiMsophis biserialis (E)Pterodroma phaeopygiaPuffinus IherminieriOceanodroma castroNyctanassa violaceaArdeidae, sp. indet.Non-passerine Aves, sp. indet.Zenaida galapagoensisTyto punctatissima (E)Pyrocephalus nanusMyiarchus magnirostrisMimus trifasciatus (E)Geospiza nebulosa (E)G. fuliginosaG. fortisG. magnirostris (E)G. crassirostrisG. scandensG. parvulaG. pauperG. olivaceaGeospiza, sp. indet.Passeriformes, sp. indet.Aves, sp. indet.Lasiurus borealis

\lus musculus (I)Rattus rattus (I)Felis domesticus (I)Sus scrofa (I)Equus asinus (I)Mammalia, sp. indet. (I)

Total

Percentage of total

Cueva de Post Office(Inferior)

Surface

1900(71)

3(2)646(4)

1(1)

2(1)

128(21)

58(11)

2(2)

132(17)

3(1)22(0)

1(1)121(3)

1(1)1(1)

2897(131)14.4(11.8)

Excavations

1(1)1974(32)

47(18)144(21)118(13)

141(1)

34(13)

8(6)2(2)

24(9)

3(3)

1(1)24(2)32(0)

1(1)

2553(122)12.7(11.0)

Cueva de Post Office(Superior)

Surface

202(26)2(2)

31(7)2(1)

26(7)

71(10)

469(47)

13(3)39(0)

2(1)

1(1)

855(103)4.2(9.3)

Excavations

1054(57)30(9)35(9)79(10)

11(6)

13(6)

1(1)1(1)

535(34)

2(2)29(5)

102(0)

3(1)

1(1)

1892(140)9.4(12.6)

Finch CaveSurface

77(17)2(1)

45(10)40(2)

71(15)

1(1)11(4)

3(2)24(2)

482(31)

1(1)

11(1)20(0)47(4)

10(2)

1(1)

835(91)4.1(8.2)

Barn OwlSurface

976(44)4(3)

131(23)148(6)

225(25)3(2)

1(1)113(18)

2(2)7(6)7(5)

755(68)4(3)2(2)

1(1)Kl)3(0)

112(4)50(0)

4(2)

3(3)6(4)

1(1)16(1)

2550(217)12.6(19.6)

also eat a higher percentage of birds and reptilesthan their continental counterparts. Wetmoreand Swales (1931:234-236) reported a variety ofsmall birds, lizards, and bats in pellets of T.glaucops from Hispaniola. Elsewhere in the West

Indies, Buden (1974) and Johnston (1974;1975:299) reported a fair diversity of birds, liz-ards, and bats in pellets of T. alba from theBahamas and Grand Cayman, respectively. Ineach of these Antillean samples, the percentage

Page 41: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413

Barn OwlExcavation

2(1)367( 15)508(75)

4142(72)912(20)

1(1)

1(1)

1(1)589(1)177(25)

7(3)

1(1)16(6)

172(20)3(2)

11(7)6(4)

272(23)

1(1)

1(1)1(1)3(3)

32(6)313(8)

1016(0)8(5)

32(5)3(2)

8563(303)42.5(27.4)

Total

Surface

3155(158)8(6)

210(42)836(13)

1(1)

3(2)

450(68)3(2)

2(2)253(43)

2(2)12(10)31(7)

1838(163)4(3)2(2)

2(2)

1(1)3(0)

139(9)131(0)51(6)

6(5)138(10)

1(1)16(1)

1(1)2(2)

7137(542)35.4(49.0)

Excavations

3(2)3395(104)

585(102)4321(102)1109(43)

1(1)141(1)

1(1)

1(1)589(1)222(44)

7(3)

1(1)16(6)

193(32)5(4)

12(8)7(5)

831(66)

1(1)

1(1)1(1)6(6)

35(9)366(15)

1150(0)8(5)

36(7)3(3)

13008(565)64.6(51.0)

Grand total

3(2)6550(262)

593(108)4531(144)1945(56)

1(1)142(2)

1(1)3(2)

1(1)589(1)672(112)

10(5)

1(1)18(8)

446(75)7(6)

24(18)38(12)

2669(229)5(4)2(2)

1(1)3(3)7(7)

38(9)505(24)

1281(0)59(11)

42(12)142(13)

1(1)16(1)

1(1)2(2)

20145(1107)

Percentage of total

Surface

0.0(0.0)44.3(29.2)

0.1(1.1)2.9(7.7)

11.7(2.4)0.0(0.0)0.0(0.2)0.0(0.0)0.0(0.4)0.0(0.0)0.0(0.0)6.2(12.5)0.0(0.4)0.0(0.0)0.0(0.4)3.5(7.9)0.0(0.4)0.2(1.8)0.4(1.3)

25.8(30.1)0.0(0.6)0.0(0.4)0.0(0.0)0.0(0.4)0.0(0.2)0.0(0.0)2.0(1.7)1.8(0)0.7(1.1)

Excavations

0.0(0.4)26.1(18.4)

4.5(18.0)33.2(18.0)

8.5(7.6)0.0(0.2)1.1(0.2)0.0(0.2)0.0(0.0)0.0(0.2)4.5(0.2)1.7(7.8)0.0(0.5)0.0(0.2)0.1(1.1)1.5(5.7)0.0(0.7)0.1(1.4)0.0(0.9)6.4(11.7)0.0(0.2)0.0(0.0)0.0(0.2)0.0(0.2)0.0(0.1)0.3(1.6)2.8(2.6)8.8(0.0)0.1(0.9)

Percentage ofgrand total

0.0(0.2)32.5(23.7)

2.9(9.8)22.5(13.0)9.6(5.0)0.0(0.1)0.7(0.2)0.0(0.1)0.0(0.2)0.0(0.1)2.9(0.1)3.3(10.1)0.0(0.4)0.0(0.1)0.1(0.7)2.2(6.8)0.0(0.5)0.1(1.6)0.2(1.1)

13.2(20.7)0.0(0.4)0.0(0.2)0.0(0.1)0.0(0.3)0.0(0.6)0.2(0.8)2.5(2.2)6.4(0.0)0.3(1.0)

Rank in

abundance(based on MNI)

191537

201920192020

41620136

159

102

1719201814128-

11

------

of reptiles and birds was lower than in the Flo-reana fossil sites because of a dominance of man-introduced rodents (Rattus, Mus). The relativescarcity of Rattus and Mus in the deposits fromFloreana suggests that T. punctatissima became

extinct there very shortly after the introductionof these rodents, for T. punctatissima preys heav-ily on Rattus and Mus elsewhere in the Galapagos(Niethammer, 1964; Abs et al., 1965; Groot,1983; personal observation). Had barn owls be-

Page 42: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

36 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

TABLE8.—Percentages of MNI of native vertebrates in fossilsites on Floreana, derived from data in Table 7, to attemptto determine past food habits of Tyto punctatissima. ForGeochelone elephantopus, the numbers in this table includeonly 10% of the MNI from surface deposits, and only 25%of the MNI from excavations. For birds, the numbers in thistable do not include Pterodroma phaeopygia, Nyctanassa vio-lacea, "Ardeidae, species indeterminate," "Non-passerineAves, species indeterminate," and Tyto punctatissima, becausethese taxa are unlikely to be food items of T. punctatissima.See text for additional explanation.

Location

Cueva de PostOffice (Inferior)

SurfaceExcavation

Cueva de PostOffice (Superior)

SurfaceExcavation

Finch CaveSurface

Barn Owl CaveSurfaceExcavation

TotalSurfaceExcavation

Reptiles

19.761.8

16.243.3

19.7

20.760.0

19.457.0

Birds

80.338.1

83.856.7

75.0

78.238.2

79.042.0

Bats

-

-

5.3

1.11.8

1.51.0

come extinct on Floreana in this century, forexample, one would expect to find more remainsof Rattus and Mus in the caves.

Volcanic activity is one last remote possibilityfor accumulation of fossils in the lava tubes.Mearns (1903) found 16 species of passerinebirds dead in caves in Yellowstone National Park,Wyoming, which he attributed to poisonous gasand heat associated with geothermal activity.Similarly, the heat and gas of volcanic eruptionscould kill animals in the Galapagos, perhaps es-pecially birds that entered the lava tubes to es-cape the danger on the surface, only to die whenconditions became deadly in the cave as well.

Excavations 3 and 4 of Cueva de Post Office(Superior) represent an attempt to test the im-portance of screening cave sediments. These ex-cavations were adjacent to each other in a gravely

talus slope that lacked any natural stratigraphicunits (Figure 9). The sediment of Excavation 3was screened outside of the cave in standardfashion, whereas that of Excavation 4 was notscreened, but was very meticulously hand-pickedin situ for several hours by slowly churning thetop 0.2 m of sediment, and collecting any bonesthat appeared. By digging to a depth of 0.3 m inExcavation 3, we determined that essentially allof the fossils in this talus slope occurred in theuppermost 0.1 m of sediment. Because both Ex-cavations 3 and 4 were deeper than 0.1 m, all ofthe fossils contained in their sediments were po-tentially obtainable during our operations. Ex-cavation 3 had an area (in m) of 1.2 X 0.6, or0.72 m2, while Excavation 4 had an area of 3.5X 0.7, or 2.45 m2. Thus their effective volumesof fossiliferous sediment were 0.072 m3 and0.245 m\ respectively. Therefore, Excavation 4would be expected to yield approximately 3 timesas many fossils as Excavation 3, if both methodsof collection were equally efficient in recoveringavailable fossils. Table 4 shows that this was notthe case; the screened sediments of Excavation 3yielded more specimens than the larger volumeof unscreened sediments from Excavation 4. Fur-thermore, all common taxa except Geospiza mag-nirostris were underrepresented in Excavation 4,both in numbers of specimens and in MNI. Thebones of Geospiza magnirostris, and to a lesserextent, the immature Geochelone elephantopus,were relatively conspicuous and easily collectedby hand (no large individuals of tortoises werepresent in these excavations). The small reptiles(snakes and lizards) were very much underrepre-sented, with a total of 52 specimens (7 MNI) inExcavation 3 versus only 9 specimens (2 MNI) inExcavation 4. Extremely small bones, such asthose of Phyllodactylus baurii, are collected onlyvery rarely without the use of screens. There-fore, the relative abundances of animals from thesurface collections are not truly representative ofthe entire fossil fauna. Compared to hand-col-lecting, screening also yields more fossils perMNI, and more unidentifiable fragmentary spec-imens.

Page 43: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 37

Systematic Paleontology

Mollusks, insects, and vertebrates were re-covered from the Floreana caves. Each groupwill be discussed separately.

Phylum MOLLUSCA

Class GASTROPODA

A detailed account of the land snails from theFloreana caves has been prepared (Chambers andSteadman, in press), from which the following isabstracted. Helicina cf. nesiotica was found onlyin Excavation 1 of Barn Owl Cave. Naesiotusgalapaganus was by far the most abundant speciesin all of the sites, but this species has been col-lected alive only very rarely.

Mollusks from the Floreana Caves

Order ARCHAEOGASTROPODAFamily HELICINIDAE

Helicina cf. nesiotica DallOrder STYI.OMMATOPHORA

Family PUPILLIDAEGastroprocta clausa (Reibisch)

Family BULIMULIDAENaesiotus galapaganus (Pfieffer)

Phylum ARTHROPODAClass INSECT A

Insects are represented by at least 143 individ-uals of some 21 species. Scott E. Miller, of theMuseum of Comparative Zoology, Harvard Uni-versity, has coordinated the study of these in-sects. His report, which I have edited, follows.

Although most of the insect specimens havebeen identified at least to genus, only limitedconclusions can be drawn for the following rea-sons.

1. Modern insects of Floreana (and the Gala-pagos in general) are poorly known. Althoughthe insect fauna was listed by Linsley and Usinger(1966) and Linsley (1977), many species arepoorly known and not recognizable from litera-

ture alone, and many taxa have not been sampledadequately.

2. While most of the insects that were identi-fied are endemic to the Galapagos Islands, mostare widespread within the Galapagos, and notenough is known of their modern ecology tocomment on the fossils. Most species are presentin a small modern collection made on Floreanain 1970 by R.S. Silberglied (now in the Museumof Comparative Zoology).

3. Most of the specimens are from surfacelayers. The few specimens from sub-surface lay-ers are generally poorly preserved.

The dominant group of insects here, as in mostQuaternary deposits, is the Order Coleoptera(beetles). The Family Carabidae (ground beetles)is most abundant, especially Calosoma granatense,a species widespread within, but confined to, theGalapagos (Basilewsky, 1968). Other beetles rep-resented are Bostrichidae (wood borers), Tene-brionidae (darkling ground beetles), and Curcu-lionidae (weevils).

Oothecae (egg-cases) that are apparently fromthe cockroach Periplaneta americana occur in sev-eral deposits, as do leg fragments probably fromthe same species. This cosmopolitan cockroach isgenerally thought to have originated in Africa(Roth, 1982), in spite of Durden's (1978) sugges-tion that it may have originated in Central Amer-ica. As indicated by specimens in shipwrecks, P.americana was being spread by man in the NewWorld at least as early as the sixteenth century(Durden, 1978; Roth, 1982). Man probablybrought P. americana to the Galapagos as well,since it occurs only in surface layers of Floreanacaves (and also in fossil sites on Isla Santa Cruz;Miller, unpublished data). A few fragments ofthis cockroach are found in sub-surface layers oftest pits, but these specimens are probably theresult of contamination. In Cueva de Post Office(Superior), a specimen of Evaniidae (an ensignwasp) occurs along with the oothecae of Peripla-neta, some of which show parasitic holes. Eva-niids, which are parasitic on cockroach oothecae,have not been reported previously from the Ga-lapagos.

Page 44: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

38 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Insects from the Floreana Caves

Order ORTHOPTERAFamily BLATTIDAE (cockroaches)

Periplaneta sp., near P. americana (Linnaeus)—cosmo-politan

Family unknown (not Blattidae)species unknown (mandible and miscellaneous parts)

Order HEMIPTERA

Family CYDNIDAE (burrower bugs)Dallasiellus murinus (Van Duzee)—Galapagos and

EcuadorOrder COLEOPTERA

Family CARABIDAE (ground beetles)Calosoma granatense Gehin—GalapagosPterostkhus [=Feronia of Van Dyke 1953] sp., maybe P.

waterhousei (Van Dyke)—Galapagosspecies unknown (not same as above species?)

Family BOSTRICHIDAE (wood borers)Amphicercus galapaganus Lesne—Galapagos

Family TENEBRIONIDAE (darkling ground beetles)Ammophorus sp. A, probably A. cooksoni C. Water-

house—GalapagosAmmophorus sp. BAmmophorus sp. CStomion sp., probably S. galapagoense C. Waterhouse—

Galapagostwo unknown species (not same as above species)

Family CURCULIONIDAE (weevils)Gerstaeckeria galapagoensis Van DykePantomorus sp., probably P. caroli Van Dyke

Order DIPTERAFamily CALLIPHORIDAE (blow flies)

unidentified species (probably this family)Order HYMENOPTERA

Family EVANHDAE (ensign wasps)Evania sp.

Family FORMICIDAE (ants)Camponotus sp.Odontomachus bauri (Emery)—NeotropicalSolenopsis geminate (Fabricius)—Neotropical and south-

ern Nearctic

Phylum VERTEBRATA

Class OSTEICHTHYES

Order through species indeterminate

(Unknown Bony Fish)

MATERIAL.—3 specimens, representing atleast 2 individuals (Tables 3, 6, 7). CPOI: USNM338446 (vertebra). BOC: USNM 338115 (2 ver-tebrae).

These three small, undiagnostic vertebrae arefrom one or more species of bony fish. Tytopunctatissima is not known to eat fish, so theseremains probably entered the caves as undigestedfood of another bird. Of the species recorded asfossils from the caves, Pterodroma phaeopygia, Puf-finus Iherminieri, Oceanodroma castro, and Nycta-nassa violacea eat fish, and perhaps Mimus trifas-ciatus once scavenged fish from the nearby shoresof Floreana.

Class REPTILIA

Order CHELONIA

Family TESTUDINIDAE

Geochelone elephantopus (Harlan)

(Galapagos Tortoise)

MATERIAL.—6550 specimens, representing atleast 262 individuals (Tables 3-7). All skeletalelements, as well as eggshell fragments, are rep-resented. CPOI: USNM 284647, 284648,284671, 284673, 284682-284695, 330570,330577, 330581, 330583, 330586, 330589,330594, 330595, 330599, 338247, 338253,338259, 338264, 338264, 338268, 338275,338279, 338286, 338291, 338294, 338298,338464-338475. CPOS: USNM 284473,284477, 284500, 284595, 284641, 284645,338193, 338207, 338211, 338212, 338218,338219, 338222, 338223, 338228, 338231-338233, 338238, 338239, 338244, 338245,338441. FC: USNM 331472, 331502, 331557,331561, 331564, 331579, 331582. BOC:USNM 331186, 331187, 331213, 331214,331281, 331285, 331313, 331352, 331414,331431, 331441, 331608, 331610, 338118,338152, 338166, 338171, 338476.

Geochelone elephantopus is the most commonfossil taxon, yet tortoises have been extinct onFloreana since approximately 1850 (see "Extinc-tion"), resulting in a paucity of well-documentedspecimens from this island. I regard all Galapagostortoises as conspecific under the name elephan-topus, based on Testudo elephantopus Harlan

Page 45: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413

1827. This treatment is supported by the veryhigh levels of genetic similarity in all extant formsof Galapagos tortoises, based on starch gel elec-trophoresis (Marlow and Patton, 1981) and theirrelatively uniform cranial morphometry(Crumly, 1984). Gunther (1877b) and Van Den-burgh (1914:244-259) wrote detailed summar-ies of the troubled history of nomenclature andsystematics of tortoises throughout the Galapa-gos, and much of my account of the Floreanatortoise is extracted from these sources. Harlan(1827) stated no specific island locality for theholotype (an immature animal) of T. elephanto-pus, but Van Denburgh (1914:245-249) rea-soned that it probably was from Floreana. If so,the Floreana tortoise would be known as Geoche-lone elephantopus elephantopus. Testudo ephippiumGunther 1875, was based on a specimen of un-known locality or date of collection, but wasregarded by Gunther (1875) to be from Floreanabecause it resembled a description of the Flo-reana tortoise by Porter (1822). Gunther(1877b:ll, 62, 82) later regarded T. ephippiumto be from Santa Cruz as well as Floreana, butBaur (1889) believed that it was from Pinta.Finally, Gunther (1896) and Van Denburgh(1914:252) believed that T. ephippium was fromPinzon.

The first adequate descriptions of the Floreanatortoise were those of Gunther (1902:185-192,pis. 16-21) and Van Denburgh (1914:317, pis.55, 56), but all specimens involved were onlyprobably, but not certainly, from Floreana. Tes-tudo galapagoensis Baur 1889, is the first namebased on a specimen regarded by the describeras being certainly from Floreana. The holotypeof T. galapagoensis is a skeleton taken on Floreanain 1833 by Commodore John Downes of the U.S.frigate Potomac, and presented to the BostonSociety of Natural History. Under the name T.galapagoensis, Gunther (1902) described andcompared this and two other specimens suppos-edly from Floreana. One of the two "other"specimens, taken by a whaling ship in the 1830's,may actually be from Floreana, whereas theother was purchased on Floreana in 1871 by A.Agassiz, and thus probably was transported by

local residents to Floreana from another island.Van Denburgh (1914:255, 256, 316, 317) re-garded T. galapagoensis as a synonym of T. ele-phantopus, whereas Broom (1929) disagreed,stating that T. elephantopus was not from Flo-reana. If Harlan's holotype is not actually fromFloreana, the Floreana tortoise would be knownas Geochelone elephantopus galapagoensis, a treat-ment followed by Pritchard (1967:168). How-ever, only leg bones remain of Harlan's holotypeaccording to Van Denburgh (1914:247), whileRothschild (1902) reported the specimen to belost. To summarize, the history of nomenclatureof Galapagos tortoises, particularly that of Flo-reana, is confused by the uncertain collectionlocalities for various type specimens. This no-menclatorial mess may never be untangled satis-factorily. The Floreana tortoise may be best re-garded merely as Geochelone elephantopus, con-specific with all other tortoises in the Galapagos.

The morphology of the Floreana tortoise hasnot been thoroughly documented, and this factis largely responsible for the problems of nomen-clature just discussed. The first specimens oftortoises unquestionably from Floreana wereskeletons (including essentially complete cara-paces, plastrons, and skulls) from Cueva de PostOffice (Inferior) reported and figured by Town-send (1928), and described by Broom (1929).These authors did not name the cave from whichthe tortoise bones were collected, but the descrip-tion of the region, the cave itself, and the tortoisebones (Townsend, 1928; see quote in "The FossilFauna"), leave little doubt that this cave was theone named Cueva de Post Office (Inferior) byMontoriol-Pous and Escola (1975), who also re-ported the presence of tortoise bones. The CraneExpedition of 1929 (Shurcliff, 1930:104) and theHancock Expedition of 1933 (Banning, 1933:4,5) also found bones of tortoises near Post OfficeBay, probably in the same cave. Strauch(1936:84) mentioned that Dr. Friedrich Ritterentered a cave that contained tortoise bones,near Post Office Bay, in 1930. Again, this cavewas probably Cueva de Post Office (Inferior), themost easily accessible of the fossiliferous caves onFloreana.

Page 46: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

40 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

I do not know how many tortoise bones wereremoved by these and other unknown partiesbefore my own visits to Cueva de Post Office(Inferior), but the number probably was consid-erable. There is no evidence that bones of speciesother than tortoises were removed by any ofthese explorers. I have stored in the CDRS mu-seum additional tortoise bones from this cavethat could not be transported to NMNH becauseof space limitations. Because of these specimensand those removed in the 1920's and 1930's (andperhaps thereafter), the MNI's and overall rela-tive abundances of Geochelone in Tables 3 and 7are too low by an unknown quantity. Most or allof the tortoises taken from Cueva de Post Office(Inferior) prior to our visits in 1978 and 1980were large individuals, so their removal has littleor no effect on the calculated past feeding habitsof Tyto punctatissima (Table 8). No intact cara-paces or plastrons remained when we first visitedthe cave in 1978, whereas the fossils in the threeother caves sampled seem to have been undis-turbed prior to our collections. As mentionedpreviously, it seems remarkable that none of the20th century explorers ever reported (or no-ticed?) that the floor of Cueva de Post Office(Inferior) was strewn with hundreds of bones ofanimals other than tortoises.

Most of the tortoises that I collected from thecaves had been trapped after falling through aroof collapse. Cueva de Post Office (Inferior) wasan especially effective natural trap for large tor-toises (Figures 7, 8, 22), many of which survivedthe fall through the roof collapse and wanderedto the far corners of the cave before succumbing.The other three caves contained a much higherpercentage of immature individuals. A completedistribution of size classes is represented by thetortoise fossils from all of the caves, ranging fromhatchlings up to adults that weighed perhaps 150kg or more. This distribution suggests that thecaves were sampling a healthy population of tor-toises. In Barn Owl Cave and probably in othercaves as well, many of the smallest tortoises mayrepresent prey items of Tyto punctatissima, whichapparently can prey upon tortoises aged several

FIGURE 22.—Fossils of Geochelone elephantopus on floor ofRoom 2, Cueva de Post Office (Inferior), October 1980(recently constructed tourist trail allowed easy access to cave,resulting in extensive damage to tortoise bones).

months or younger. I am not aware of any reportof T. punctatissima preying upon small tortoisestoday, but there are no published studies of thefood habits of barn owls in areas where tortoisesreproduce at natural levels. That barn owlswould have fed regularly on very young tortoisesseems reasonable, however, especially on a ro-dent-free island such as Floreana. On Santa Cruz,for example, very small tortoises occur in theclearly 7)tfo-derived fossil site at Cueva de Kubler.This cave would have been ineffective as a nat-ural trap, and thus most of the tiny tortoisesthere probably were prey items of T. punctatis-sima.

Page 47: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 41

Order SQUAMATA

Suborder SAURIA

Family GEKKONIDAE

Phyllodactylus baurii Garman

(Gecko)

MATERIAL.—593 specimens, representing atleast 108 individuals (Tables 3-7). Most majorskeletal elements are represented. CPOI: USNM284674, 330592, 330596, 338248, 338255,338261, 338266, 338270, 338277, 338281,338296, 338300, 338457. CPOS: USNM284472, 284593, 338206, 338215, 338226,338235, 338243, 338439. FC: USNM 331473.BOC: USNM 331280, 331432, 331438,331606, 331612, 338109, 338119, 338154,338168, 338173, 338433, 338434.

Osteological distinctions, if they exist, neverhave been reported for the various species ofPhyllodactylus that occur in the Galapagos. Thefossils from Floreana agree with a skeletal speci-men of Phyllodactylus from the Galapagos, and Irefer them to the species baurii simply because itis the only species of Phyllodactylus recorded onFloreana. I follow the species-level systematics ofVan Denburgh (1912b), who used the name P.baurii for the geckos of Floreana, Gardner-near-Floreana, Champion, Enderby, Espanola, andGardner-near-Espanola. Based upon allozymeelectrophoresis, Wright (1983) has reported thatPhyllodactylus from Floreana is most similar tothose populations from Santiago, Bartolome, andMarchena, and to be least similar to those fromSan Cristobal, Santa Fe, and Fernandina. Thesystematic conclusions to be drawn from thesefindings are unclear.

Geckos were first reported from Floreana byGunther (1877a) as P. galapagensis Peters, basedupon specimen(s) taken in June 1875, by Com-mander W.E. Cookson of HMS Peterel. Garman(1892) regarded the geckos of Floreana, Espan-ola, and Gardner-near-Espanola as distinct from

those of other islands, and named them as a newspecies, P. baurii, based upon a specimen col-lected by George Baur at Las Cuevas, Floreana,in 1891. Heller (1903) and Van Denburgh(1912b) provided descriptions and measure-ments of P. baurii and other geckos of the Gala-pagos. I use Garman's original spelling of thespecific epithet baurii, although most authorssince Garman (1892) have dropped the penulti-mate "i."

Unlike the other terrestrial reptiles of Flo-reana, Phyllodactylus has been reported to becommon, even near inhabited areas such as BlackBeach (Heller, 1903; Slevin, 1931). J.R. Slevin'sfield notes of 1905-1906 on Phyllodactylus (inVan Denburgh, 1912b, and Fritts and Fritts,1982) reported these geckos to be plentiful inthe lowlands under loose lava blocks, bark, anddried wood. For example, Slevin collected 125geckos on Floreana on 11 October 1905, and 69geckos on 23 May 1906. He also encounteredeggs of geckos frequently, thus indicating appar-ently healthy rates of reproduction. I do notknow the status of Phyllodactylus on Floreanatoday, never having looked for it. It may still bedoing well, for its inconspicuous habits wouldmake predation difficult for introduced preda-tors such as rats, cats, and dogs.

I believe that most or all of the geckos re-covered from the four caves represent prey itemsof Tyto punctatissima. Both Phyllodactylus and Tytoare active nocturnally. Like all geckos, Phyllodac-tylus is so sure-footed that it is unlikely to fallaccidentally into a roof collapse and becometrapped in a cave. Phyllodactylus is a prey item ofTyto on Santa Cruz as well, in both a modern andfossil context. In northern coastal Peru, Thomasand Thomas (1977) also reported remains ofPhyllodactylus from a modern bone deposit at-tributed to owls, possibly to Tyto alba.

Had not the sediments from the caves on Flo-reana been sifted through a fine-mesh (V\a inch)screen, the relative abundance of Phyllodactyluswould have been grossly misjudged in the fossilfaunas. The tiny bones of Phyllodactylus are neverretained in coarse-mesh (V4 inch) screens.

Page 48: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

42 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Family IGUANIDAE

Tropidurus grayii (Bell)

(Lava Lizard)

MATERIAL.—4531 specimens, representing atleast 144 individuals (Tables 3-7). All skeletalelements are represented. CPOI: USNM284675, 330578, 330596, 338254, 338260,338265, 338269, 338276, 338280, 338287,338292, 338295, 338299, 338447. CPOS:USNM 284478, 284498, 284592, 338192,338208, 338214, 338227, 338234, 338240,338440. FC: USNM 331474, 331503, 331504,331560, 331563, 331580, 338326, 338436.BOC: USNM 331184, 331193, 331194,331209, 331216, 331279, 331286, 331312,331353, 331416, 331433, 331439, 331605,331611, 338120, 338153, 338167, 338172.

These fossils compare favorably with a skeletalspecimen of Tropidurus from the Galapagos.There are no known osteological distinctionsamong the various species of Tropidurus in theGalapagos, except perhaps for size. My referal ofthe fossil material to the species grayii is basedpurely upon geography, as T. grayii is the onlyspecies of Tropidurus known from Floreana. Bell(1843:24) originally described T. grayii as Leioce-phalus grayii, from "numerous specimens" col-lected on Floreana and San Cristobal by CharlesDarwin in September 1835. It is not known howmany of the specimens came from Floreana, orhow many there were altogether. A synonymy ofgeneric names applied to T. grayii is listed in Baur(1892). I follow the species-level systematics ofVan Denburgh and Slevin (1913), who restrictedT. grayii to Floreana and its satelite islets ofGardner-near-Floreana, Champion, and En-derby. In addition, I observed two males and onefemale of 7. grayii on Isla Caldwell on 26 Decem-ber 1980. Surprisingly, Wright (1983) found theTropidurus from Floreana to be most similar tothose from far-away Pinta, and least similar tothose from Marchena and San Cristobal, basedon allozyme electrophoresis. These findings werenot interpreted in a strict systematic manner.

On most islands in the Galapagos, Tropidurusthrives in the face of natural predation by hawks(Buteo), owls (Tyto, Asio), mockingbirds (Mimus),and snakes (Alsophis). There is no reason todoubt that the same was once true on Floreana.However, Tropidurus on Floreana apparently hassuffered greatly from predation by feral mam-mals, especially rats, cats, and dogs. Baur(1895:70) reported Tropidurus as "exceedinglyrare" on Floreana in 1891. Heller (1903) blamedthis rarity on predation by domestic animals,especially cats. The Webster-Harris Expeditionfound no Tropidurus on Floreana during a weekin 1897 (Rothschild and Hartert, 1899:99, 100).In May 1899, Heller found no Tropidurus duringthree days of collecting in the western and centralparts of Floreana. Van Denburgh and Slevin(1913) regarded Tropidurus as almost extinct onFloreana, based upon the California Academy ofSciences Expedition taking only 16 specimens ofthis lizard during 28 days on this island in 1905-1906. This compares to their capture of 15 spec-imens in only two hours on Gardner-near-Flo-reana, 16 on Champion in 1 xh hours, and 31 onEnderby in one hour. Slevin (1931:40, 1935:20;and in Fritts and Fritts, 1982:17, 18) stated thatcats have nearly exterminated Tropidurus on Flo-reana, with the last population occurring in lownumbers near the northeastern coast. My ownfield work on Floreana has been mostly on thenorthern coast near Post Office Bay, Punta Cor-morant, and Bahia de las Cuevas. Tropidurus isextremely rare in the first two regions. It is notexcessively rare at Bahia de las Cuevas (i.e., a fewof them can be found with little difficulty), butstill is much less common here than in compara-ble dry lowlands elsewhere in the Galapagos.

The small body size of Tropidurus on Floreanais as striking as its rarity. To assess whether ornot this small body size may also be an artifact ofthe unnatural conditions that exist now on Flo-reana, I visited Isla Champion, a tiny islet 0.7 kmnorth of Floreana, with pristine native vegetationand no introduced mammals. On 26 October1980 and 24 May 1983, I was very impressed byhow much larger (especially males) and more

Page 49: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 43

abundant Tropidurus grayii is on Champion thanon nearby Floreana. My conservative field esti-mates of snout-vent length of T. grayii on Cham-pion range from 100 to 120 mm for males, and90 to 110 mm for females, compared to 50 to60 mm for individuals of either sex on Floreana.Some published snout-vent lengths for T. grayiifrom Floreana are as follows: Boulenger (1891),59-65 mm for Darwin's five specimens (fourmales, one female); and Van Denburgh andSlevin (1913:165), 65 mm for an adult maletaken in 1905-1906. The status of Tropiduruson Champion versus Floreana is paralleled bythat reported by P.R. Grant (1975) on HermanosIII, a small island off the southeast coast ofIsabela. He found the Tropidurus on HermanosIII to be much larger than those from Isabela.Likewise, P.R. Grant (1975) reported that Dag-mar Werner found the Tropidurus on the smallisland of Elizabeth to be larger than those fromthe adjacent islands of Fernandina and Isabela.Certain lizards in the West Indies show similartrends of size and survival. Baskin and Williams(1966) reported that the teiid lizard Ameivapolops of St. Croix was exterminated by the in-troduced mongoose (Herpestes auropunctatus),but survived on two small islets near St. Croixthat were mongoose-free. Likewise, we observedonly one medium-sized individual of Ameiva gris-woldi during four weeks on mongoose-riddenAntigua in 1980 and 1983, but found very largeindividuals of A. griswoldi to be common on thenearby small islet of Great Bird Island (Steadmanetal., 1984).

The fossil material of Tropidurus includesmany individuals that are larger than any modernspecimens from Floreana, and are approximatelythe size of modern individuals from Champion.Again, similar situations have been reportedfrom the West Indies. Among Puerto Rican liz-ards, Pregill (1981) found a larger size in lateQuaternary fossils of the iguanids Anolis cuvieriand A. occultus and the anguid Diploglossus pleeithan in any modern specimens. In the Bahamas,late Quaternary fossils of Anolis sagrei are alsolarger than modern specimens (Pregill, 1982).

The larger size of fossil Tropidurus, along withthe examples of large individuals surviving onsmall, nearby, predator-free islets, suggests thatthe small size of today's Tropidurus on Floreanais the result of excessive predation by introducedcats, dogs, and rats.

Among the diagnostic characters reported forTropidurus on Floreana are its small size (espe-cially in males), its low amount of sexual di-morphism in color and mid-dorsal crest devel-opment, and its reversed sexual dimorphism insize, with males slightly smaller than females (VanDenburgh and Slevin, 1913; Carpenter, 1966,1970). I believe that these characters do notrepresent the natural condition in T. baurii. Thesmall size arises because individuals are nevergiven the chance to reach full size. In any orga-nism with indeterminate growth, a sharp increasein rate of predation will reduce mean body sizeof the population, particularly in a lizard such asTropidurus that becomes more conspicuous topredators as its size increases. The average sizeof Ameiva polops in St. Croix decreased this cen-tury from 6-7 inches to 4-5 inches as theirpopulation was being reduced by mongoose pre-dation (Baskin and Williams, 1966). The reducedlevels of sexual dimorphism in size, color, andmid-dorsal crest development in T. grayii of Flo-reana can also be explained by the artificiallyhigh rate of predation they now experience,which does not allow males to attain fully any ofthe normal adult male characteristics. The dif-ferences in display movements between malesand females of Tropidurus in the Galapagos (de-scribed by Carpenter, 1966) may make malessomewhat more conspicuous than females. Thiswould result in males being preyed upon at anearlier average age and thus accentuate the arti-ficially low levels of sexual dimorphism. Thatthere may be an abnormal sex ratio in Tropiduruson Floreana is suggested by the collection of 14females versus only 1 male by the CaliforniaAcademy of Sciences Expedition in 1905-1906.

In conclusion, I would suggest that size can beused as taxonomic character for insular reptilesonly with great caution. Inter-island comparisons

Page 50: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

44 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

of size in closely related reptilian taxa shouldconsider historical information on predation,such as the introduction of alien predators or theextinction of natural predators.

Suborder SERPENTES

Family COLUBRIDAE

Alsophis biserialis (Gunther)

(Floreana Snake)

MATERIAL.—1945 specimens, representing atleast 56 individuals (Tables 3-7). All skeletalelements are represented. CPOI: USNM284649, 284676, 330574, 330575, 330591,330593, 330597, 338249, 338256, 338262,338267, 338271, 338278, 338282, 338288,338301, 338450, 338458, 338460. CPOS:USNM 284479, 284499, 284594, 338213,338220, 338221, 338225, 338229, 338236,338241, 338242, 338438. FC: USNM 331475,338327. BOC: USNM 331185, 331192,331210, 331217, 331434, 331440, 331607,331613, 338121, 338169, 338174, 338431,338432.

These fossils compare favorably with a skeletalspecimen of Alsophis from the Galapagos. Noosteological descriptions exist for the snakes ofGalapagos; I assume that the fossils are conspe-cific with A. biserialis, the only snake known fromFloreana. The snakes of the Galapagos are xe-nodontine colubrids, a complex assemblage ofneotropical snakes in which generic level system-atics is very unsettled (John E. Cadle, pers.comm.). Traditionally, snakes of the Galapagosare placed in the genus Dromicus, but I followMaglio (1970) in placing them in Alsophis, agenus that otherwise occurs on mainland SouthAmerica and in the West Indies. For species-levelsystematics, I follow Van Denburgh (1912a), inrealization that a paucity of specimens leaves thisquestion unsettled. Until the description of A.biserialis eibili from San Cristobal (Mertens,1960), A. biserialis was known only from Flo-reana, Gardner-near-Floreana, and Champion.Thus the snake of Floreana and its satellites is

now known as A. biserialis biserialis.Snakes are apparently extinct on Floreana (see

"Extinction"). The only snake ever collectedfrom life on Floreana was a young specimentaken by Charles Darwin in September 1835,which Gunther (1860) described as Herpetodryasbiserialis. Dr. A. Habel secured a snake in 1868that supposedly came from Floreana, but VanDenburgh (1912a) stated that this specimen camefrom either Santa Cruz or Rabida. Heller (1903)mentioned, without any details, that the HasslerExpedition of 1872 took a snake on Floreana.However, Steindachner (1876:305) reportedthat the Hassler Expedition found (but did notcollect) only 1 snake in all of the Galapagos, thisbeing on Rabida. The only other specimen of A.b. biserialis is an adult female collected on Gard-ner-near-Floreana on 3 October 1905, by theCalifornia Academy of Sciences Expedition (VanDenburgh, 1912a; Slevin, 1931). Snakes arefound commonly today, but not collected, onChampion and Gardner-near-Floreana (De Rid-der, 1976; B. Voigt, 1977a,b; P.R. Grant, 1980;B.R. Grant, 1981).

Snakes seem unlikely to fall into natural trapswith any frequency. Instead I believe that mostof the snakes recovered from the caves of Flo-reana were prey items of Tyto punctatissima. Judg-ing from their occurrence as fossils in Cueva deKubler, Santa Cruz, snakes were eaten regularlybut in rather low numbers by T. punctatissima.Floreana lacks native rodents, so barn owls mayhave eaten snakes there more frequently than onSanta Cruz, although confirmation of this sug-gestion awaits completion of the study of thefossil fauna from Cueva de Kubler. The smallsize of T. punctatissima would make a fully adultsnake difficult to procure. Therefore, snakes inthe Galapagos fossil deposits may not include thelargest individuals in the population.

Class AVES

The standard references for the distributionand/or systematics of birds within the Galapagosare Salvin (1876), Ridgway (1897), Rothschildand Hartert (1899, 1902), Snodgrass and Heller

Page 51: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 45

(1904), Gifford (1913), Loomis (1918), Swarth(1931), and Harris (1973, 1974, 1982). Herein Iwill follow my own systematic judgements, whichoften agree with those of one or more of theauthors cited above. I have used trinomials onlywhen subspecies can be distinguished osteologi-cally.

Order PROCELLARIIFORMES

Family PROCELLARIIDAE

Pterodroma phaeopygia (Salvin)

(Dark-rumped Petrel)

MATERIAL.— 1 specimen, representing 1 indi-vidual (Tables 6, 7). BOC: USNM 331599 (alarphalanx).

Pterodroma phaeopygia is larger than any otherprocellariid that occurs regularly in the Galapa-gos. This petrel may be too large to be a regularprey item of Tyto punctatissima, although Groot(1983) reported it from barn owl pellets on SantaCruz. Pterodroma phaeopygia nests and roosts to-day in caves, crevices, and burrows in the high-lands of larger islands in the group, includingFloreana. Coulter (1982) reviewed the currentstatus of P. phaeopygia on Floreana, finding it tobe declining rapidly in numbers, with most indi-viduals confined to Cerro Pajas. Barn Owl Cave,although in the lowlands, may have served as aformer roosting or nesting area for P. phaeopygia.In October 1980, we found bones of P. phaeopy-gia of little apparent antiquity in the shallowcaves of the Bahia de la Cuevas region on thenortheast coast of Floreana. These bones provideadditional evidence that P. phaeopygia once wasmuch more widespread on Floreana than today.Pterodroma phaeopygia is also known from lateQuaternary fossils in the Hawaiian Islands (Olsonand James, 1982b:32).

Puffinus Iherminieri Lesson

(Audubon's Shearwater)

MATERIAL.—142 specimens, representing atleast 2 individuals (Tables 3, 7). CPOI: USNM

284678 (carpometacarpus), 330587 (partial as-sociated skeleton).

This species is distinguished from all otherprocellariids in the Galapagos by its small size. Itis, however, larger than any species of oceanitid.Puffinus Iherminieri is a very common residentbird in the Galapagos, nesting in holes and crev-ices on shoreline cliffs on Floreana and manyother islands. It may have been incorporated intothe fossil fauna as prey of Tyto punctatissima.Quaternary fossils of P. Iherminieri have beenreported from the Hawaiian Islands (Olson andJames, 1982b:33), many islands in the West In-dies (Brodkorb, 1963:246; Olson and Hilgart-ner, 1982), and St. Helena Island (Olson,1975:20).

Family OCEANITIDAE

Oceanodroma castro (Harcourt)

(Harcourt's (Madeiran) Storm Petrel)

MATERIAL.—1 specimen, representing 1 indi-vidual (Tables 6, 7). BOC: USNM 338323 (man-dibular articulation).

I compared this specimen to all species ofoceanitids that occur in the Galapagos exceptFregetta grallaria, which is a very rare non-breed-ing visitor. Most species that occur anywhere inthe eastern Pacific were considered as well. Thefossil is larger than in Oceanites gracilis, O. ocean-icus, Oceanodroma tethys, O. leucorhoa, or Halocyp-tena microsoma, and smaller than in Oceanodromamarkhami or O. tristrami. It is approximately thesize of Fregetta tropica, Pelagodroma marina,Oceanodroma castro, O. furcata, O. homochroa, O.hornbyi, O. melania, or 0. monorhis, but resemblesthe mandible of O. castro and differs from allothers of similar size in having a shorter andstouter Proc. mandibulae medialis.

Oceanodroma castro nests in many localities inthe Galapagos, the nearest to Post Office Baybeing Isla Champion. This storm-petrel is noc-turnal on its breeding grounds, and its occur-rence in Barn Owl Cave is undoubtedly due topredation by Tyto punctatissima. Predation onstorm-petrels by barn owls is not without prece-

Page 52: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

46 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

dent, for Bonnott (1928) reported that T. albafed heavily on O. leucorhoa at Castle Rock, coastalCalifornia. Late Quaternary fossils of 0. castroare known from the Hawaiian Islands (Olson andJames, 1982b:33), Ascension Island (Ashmole,1963; Olson, 1977), and St. Helena Island (Ol-son, 1975).

Order Incertae Sedis(following Olson, 1979)

Family ARDEIDAE

Nyctanassa violacea (Linnaeus)

(Yellow-crowned Night Heron)

MATERIAL.—3 specimens, representing atleast 2 individuals (Tables 3, 6, 7). CPOI: USNM330568 (radius, tarsometatarsus). BOC: USNM331208 (vertebra).

Nyctanassa violacea is unique in size amongherons in the Galapagos, all other taxa beingeither larger or smaller. It is resident on Floreanaand many other islands in the archipelago. Night-herons are too large to be preyed upon by Tytopunctatissima, but they may have died in Cuevade Post Office (Inferior) and Barn Owl Cavewhile searching for large insects. Nyctanassa vio-lacea commonly occurs well inland in arid low-land areas of the Galapagos (Gifford, 1913:59;Harris, 1974:82, 1982:83; personal observation).It is recorded only from surface levels, so anotherpossibility is that the remains of a dead individualwere carried into the cave by a feral scavenger,such as a cat, dog, or pig. Quaternary fossils ofN. violacea have been reported from Florida andthe West Indies (Brodkorb, 1963:285; Olson andHilgartner, 1982).

Ardeidae, species indeterminate

(Unknown Heron)

MATERIAL.—1 specimen, representing 1 indi-vidual (Tables 6, 7). BOC: USNM 331372 (un-gual phalanx).

This specimen is from an immature, medium-sized heron. It is likely to represent either Ardeaalba (Great Egret) or Nyctanassa violacea, but isnot sufficiently diagnostic to allow specific iden-tification.

Non-passerine Aves, order indeterminate

(Unknown Non-Passerine Bird)

MATERIAL.—589 specimens, representing atleast 1 individual (Tables 6, 7). BOC: USNM331366, 338170 (elements uncertain).

These shattered, very fragmentary fossils dom-inated the lowest unit (C2) of Excavation 1, BarnOwl Cave. They are of an indeterminate largebird, but are too incomplete to reveal any diag-nostic features. Because these fossils are associ-ated stratigraphically and are all preserved simi-larly, they probably belong to a single individual.

Order COLUMBIFORMES

Family COLUMBIDAE

Zenaida galapagoensis Gould

(Galapagos Dove)

MATERIAL.—664 specimens, representing atleast 112 individuals (Tables 3-7). All skeletalelements are represented. CPOI: U S N M284646, 284677, 284696, 330569, 330576,330579, 330580, 330584, 330585, 330590,330600, 330759, 338258, 338263, 338272,338283, 338293, 338297, 338302, 338449,338454. CPOS: U S N M 284474, 284576,284597, 284644, 338194, 338209, 338216. FC:USNM 331468, 331501, 331556, 331578,331585. BOC: U S N M 330690, 330701,330715-330717, 330728, 330735, 330736,330741, 331183, 331195, 331206, 331219,331277, 331287, 331311, 331318, 331351,331355, 331365, 331412, 331428, 331590,331600, 338103, 338110, 338124, 338155,338175,338306.

These fossils agree in detail with a skeleton of

Page 53: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 47

Z. galapagoensis, the only columbid that occursin the Galapagos. Males of this dove are signifi-cantly larger than females in external measure-ments (Ridgway, 1897:617, 618; Gifford,1913:11, 111) and likewise the fossils fall intotwo distinct size categories. Sundevall (1871) pro-posed the genus Nesopelia for Z. galapagoensis,but neither osteology nor plumage support ge-neric separation of this dove.

Zenaida galapagoensis is a ground-dwellingdove that occurs throughout the Galapagos, al-though it usually avoids the moist highlands onlarger islands except Floreana. The concentra-tion of dove bones in the Floreana caves is un-doubtedly due to predation by Tyto punctatissima.This statement is corroborated by the large num-ber of immature individuals represented. Zen-aida galapagoensis is common also in the Tyto-derived fossil site at Cueva de Rubier, SantaCruz.

In pristine times, Z. galapagoensis was abun-dant and conspicuous throughout all dry regionsof the Galapagos. Early accounts of the islandsmention the tameness and great numbers ofthese doves, which were slaughtered and eatenregularly by the crews of passing ships. For ex-ample, Darwin (1871:173, 174) stated:

Cowley (in the year 1684) says, that the "turtle-doves wereso tame, that they would often alight upon our hats andarms, so as that we could take them alive: they are notfearing man, until such time as some of our company didfire at them, whereby they were rendered more shy." Dam-pier also, in the same year, says that a man in a morning'swalk might kill six or seven dozen of these doves. At present,although certainly very tame, they do not alight on people'sarms, nor do they suffer themselves to be killed in such largenumbers. It is surprising that they have not become wilder;for these islands during the last hundred and fifty years havebeen frequently visited by buncaniers [sic] and whalers; andthe sailors, wandering through the woods in search of tor-toises, always take cruel delight in knocking down the littlebirds.

These birds, although now still more persecuted, do notreadily become wild: in Charles Island, which had then beencolonized about six years, I saw a boy sitting by a well witha switch in his hand, with which he killed the doves andfinches as they came to drink. He had already procured alittle heap of them for dinner; and he said that he had

constantly been in the habit of waiting by this well for thesame purpose.

Galapagos doves are still remarkably tame to-day on islands such as Genovesa, Santa Fe, Ra-bida, and Fernandina that are not populated byhumans or feral predators. They are reluctant tofly and can be baited with seeds to feed from aperson's hand. As much as any species in theGalapagos, doves represent the extreme tame-ness characteristic of insular birds.

The rarity of doves on Floreana in 1905-1906was attributed to predation by feral cats, dogs,and humans (Gifford, 1913:6, 8). I agree thatferal predation is probably a major cause of therarity of doves on Floreana, but other factorsmay be involved as well. For example, P.R. Grantand K.T. Grant (1979) showed that approxi-mately 50% of the dove nests they studied onGenovesa were located in old mockingbird nests,and these nests had a higher rate of fledglingsuccess than those not using old mockingbirdnests. Thus the extinction of Mimus trifasciatuson Floreana may have removed many primenesting sites for the doves. Further, P.R. Grantand K.T. Grant (1979) noted that doves on Gen-ovesa fed on the flowers, seeds, and pulp ofOpuntia. The unnatural scarcity of Opuntia onFloreana has depleted an important source offood, as well as nesting sites.

The future of doves on Floreana is not prom-ising; I doubt if they will survive this century.The only doves that I saw during 4 weeks onFloreana were a pair that frequented ledges onthe deep, vertical walls of the entrance to FinchCave in 1978, where they were apparently pro-tected from predators. A similar situation existson Santa Cruz, where Gifford (1913:6) founddoves to be common or abundant in 1905-1906.Since that time, cats, dogs, and people have pop-ulated Santa Cruz, and doves are rare there today(personal observation). The survival of doves inthe entire archipelago depends upon islands likeGenovesa, Santa Fe, Rabida and Fernandinabeing maintained free of human residents andferal predators.

Page 54: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

48 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Order STRIGIFORMES

Family TYTONIDAE

Tyto punctatissima (Gray)

(Galapagos Barn Owl)

MATERIAL.—10 specimens, representing atleast 5 individuals (Tables 6, 7). BOC: USNM331189 (vertebra), 331288 (coracoid), 331316(manus phalanx), 331371 (2 pedal phalanges),338114 (2 vertebrae), 338156 (3 pedal pha-langes).

The coracoid (Plate 1) and vertebrae agreewith a partial skeleton of Tyto punctatissima fromSanta Cruz, while the phalanges (not available inthe partial skeleton) agree qualitatively withthose of the nearly cosmopolitan T. alba, anddiffer from those of the only other owl in theGalapagos, Asio flammeus (Family Strigidae). Tytopunctatissima has been regarded as a subspeciesof T. alba by most authors of the past two dec-ades, but I recognize T. punctatissima as a fullspecies, very distinct from the races of T. alba inmainland North, Central, and South America.Tyto punctatissima is smaller in all external meas-urements than T. alba, and every skeletal elementof T. punctatissima (best known from numerousfossils from Cueva de Kubler, Santa Cruz) iseasily distinguished from that of T. alba, beingonly approximately 2/3 as large. Furthermore, theplumage in T. punctatissima is much darker thanin T. alba.

Tyto punctatissima is known from San Cristobal,Santa Cruz, Baltra, Santiago, Isabela, and Fer-nandina, but is extinct on Floreana (see "Extinc-tion"). Harris (1973) regarded barn owls on Flo-reana as "probably extinct" (p. 270), and "prob-ably once resident, now not present" (p. 274). Iam aware of no evidence for the former occur-rence of T. punctatissima on Floreana, however,other than the fossils I found in Barn Owl Cave.Groot (1983:169) reported "only one doubtablerecord of Barn Owl pellets found on Floreana byJ. Hatch (Brosset 1963) came to my knowledge."Brosset (1963), in fact mentioned no such record.

The fossils from Barn Owl Cave include at leastone immature individual, thus establishing thatbarn owls nested as well as roosted on Floreana.

Order PASSERIFORMES

Various passerine birds dominate the modernterrestrial avifauna of the Galapagos. Each spe-cies is small enough to be preyed upon by Tytopunctatissima, and I believe that by this methodthey were incorporated into the fossil deposits.

Family TYRANNIDAE

Pyrocephalus nanus Gould

(Galapagos Vermilion Flycatcher)

MATERIAL.— 1 specimen, representing 1 indi-vidual (Tables 6, 7). BOC: USNM 338382 (hu-merus).

This specimen agrees qualitatively with thehumerus of Pyrocephalus rubinus of mainlandNorth, Central, and South America. It is distin-guished from the humerus of other Galapagospasserines except Myiarchus magnirostris in lack-ing the medial Fossa pneumotricipitalis. It issmaller than the humerus of M. magnirostris.

Ridgway (1894) described a new species,Pyrocephalus carolensis, on the basis of specimensfrom Floreana. I agree with Swarth (1931:88,93) in regarding P. carolensis as a synonym of P.nanus. Most authors since Swarth (1931) haveregarded P. nanus as a subspecies of P. rubinus,but I believe that the differences in osteologyand adult plumage between these two taxa war-rant species-level recognition. The wings and tailof P. nanus are much shorter than in P. rubinus.Bones of the wing are much smaller in P. nanusthan in P. rubinus, while the reverse is true forbones of the leg. Reduction of the pectoral girdleand enlargement of the pelvic girdle are commonphenomena in insular birds. It seems unlikelythat P. nanus could undergo any further reduc-tion of the pectoral assemblage, for such wouldrender it flightless or very nearly so. I cannotthink of any selective advantage for flightlessness

Page 55: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

N U M B E R 4 1 3 49

in an arboreal flycatcher such as Pyrocephalus,which catches insects on the wing. The maleplumage of P. nanus is a lighter and duller redthan in P. rubinus. The female plumage of P.nanus is very unlike that of P. rubinus, beingyellow below with slight or no streaking, while P.rubinus is more or less cream-colored below,suffused w ith varying amounts of pink, and withprominent brown streaks. DeBenedictis (1966)noted differences between the songs and songflights of P. nanus compared to those of P. rubi-nus of the mainland (California, Arizona, Col-ombia).

Two species of Pyrocephalus have evolved inthe Galapagos (P. dubius from San Cristobal, andP. nanus from nearly all other islands), thus pro-viding additional evidence that the Galapagosforms have been distinct genetically from theancestral P. rubinus for quite some time. Pyroce-phalus dubius differs from P. nanus mainly in sizeand female plumage. Like P. nanus, it is verydistinct from P. rubinus of the mainland.

Pyrocephalus nanus is most common on islandssuch as Floreana that have moist highlands. Onthese islands it seldom if ever nests in the aridlowlands, occurring there mainly in the non-breeding season. Thus its rarity in the fossil faunais easily understood.

Myiarchus magnirostris (Gould)

(Large-billed Flycatcher)

MATERIAL.—18 specimens, representing atleast 8 individuals (Tables 5-7). FC: USNM331477 (humerus). BOC: USNM 330718 (quad-rate), 330761 (quadrate), 338113 (humerus),338305 (mandible), 338311-338314 (4 mandi-bles), 338377-338381 (5 humeri), 338420 (4quadrates).

Each of these fossils agrees with a skeleton ofM. magnirostris, and differs from all other Gala-pagos passerines, except P. nanus, as follows:quadrate, in ventral aspect, with entire mandi-bular articulation more rectangular (less trian-gular) in outline, and with a relatively shallow

depression or notch in the central portion; man-dible, w ith relatively small retroarticular process,the internal articular process nearly horizontal inposterior aspect, the ramus relatively straight andweak, and the symphysis broad and flat. Myiar-chus magnirostris is distinguished from Pyroce-phalus nanus by its larger size.

Ridgway (1893) proposed a new genus Eribatesfor M. magnirostris, but few subsequent authorshave recognized this taxon. The placement of M.magnirostris in Myiarchus seems well substanti-ated by morphology and behavior (Lanyon,1978:603-609).

This species is found throughout the Galapa-gos, at nearly all elevations. It is common todayin the lowlands of Floreana.

Family MIMIDAE

Mimus trifasciatus (Gould)

(Floreana Mockingbird)

MATERIAL.—446 specimens, representing atleast 75 individuals (Tables 3-7). All skeletalelements are represented. CPOI: USNM284358-284366, 284651, 284651, 284681,330670, 330671, 338290, 338456. CPOS:USNM 284367-284373, 284475, 284575,284586, 284639, 284643. FC: USNM 331467,331476, 331499. BOC: USNM 330686,330729, 330744, 330748, 331177-331180,331196, 331205, 331220, 331276, 331289,331310, 331317, 331354, 331369, 331411,331427, 331593, 331601, 338104, 338111,338125,338157, 338419.

Mimus trifasciatus is easily recognized osteolog-ically as the largest passerine bird that occurs inthe fossil sites. The fossils agree qualitatively andquantitatively (Table 9) with skeletons of M. tri-fasciatus from Champion. They are consistentlymuch larger than all skeletal elements of M.parvulus (Plate 2), while they are slightly smallerthan M. macdonaldi in the rostrum and mandible,but similar in size to M. macdonaldi in most post-cranial elements. Skeletons of M. melanotis were

Page 56: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

50 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

TABLE 9.skeletonssizes).

—Osteological measurements of adult Mimus (in mm), comparing Floreana fossils with modernfrom other Galapagos islands (all means rounded to nearest 0.1 mm because of small sample

Specimens

Mimus trifasciatus(Floreana)

meanrangenumber

Mimus trifasciatus(Champion)

meanrangenumber

M. macdonaldi(Kspanola, Gardner-near-Espaiiola)

meanrangenumber

M. parvulus barringtoni(Santa Fe)

meannumber

M. p. bauri(Genovesa)

meanrangenumber

M. p. pa rvulus(Santa Cruz)

meannumber

M. p. personatus(Rabida, Pinta)

meanrangenumber

Length ofrostrum

28.3

1

29.029.0-29.1

2

32.231.2-34.1

5

25.41

28.227.6-28.8

4

21.51

24.522.3-26.8

3

Length ofmandible

42.8+42.8±42.9

2

42.842.0-43.6

2

46.144.7-47.1

5

39.01

42.240.3-43.6

4

34.81

38.435.8-41.8

3

Maximumdepth ofmandible

3.23.1-3.4

2

3.13.0-3.2

2

3.12.9-3 3

5

2.81

2.82.6-2.9

4

2.41

2.62.5-2.7

3

Width ofmandibular

articu-lation

5.24.9-5.4

5

5.04.9-5.2

2

5.45.1-5.5

5

4.2I

4.64.4-4.8

4

4.11

4.64.3-4.7

3

Length ofhumerus

29.428.7-30.2

9

29.028.4-29.6

2

28.427.2-29.4

5

24.21

27.025.9-27.8

3

24.91

25.123.7-26.4

3

Proximalwidth ofhumerus

7.57.3-7.7

5

7.36.9-7.7

2

7.57.1-7.9

5

6.31

7.06.7-7.2

3

6.01

6.66.4-6.9

3

Distalwidth of

humerous

6.46.0-6.8

15

6.25.9-6.6

2

6.55.9-6.9

5

5.31

(i.O

5.8-6.22

5.01

5.75.3-6.0

3

Length ofulna

33.431.4-34.7

9

32.831.8-33.9

2

32.630.9-33.8

5

26.91

31.129.4-32.2

3

27.91

28.726.9-30.1

3

not available, but this species, confined to SanCristobal, is similar in all external measurementsto M. parvulus and thus its skeleton would besmaller than the fossils from Floreana. X-raysprovide the only means of comparing the fossilswith the holotype of M. trifasciatus, taken onFloreana in October 1835 by Captain Robert

Fitzroy of the Beagle. When compared directlyto the x-rays, the Floreana fossils (especially therostrum and mandible) are more similar in sizeto the three specimens of M. trifasciatus than toany other species of Mimus in the Galapagos.

Mimus trifasciatus lives today on the small off-shore islands of Champion and Gardner-near-

Page 57: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 51

TABLE 9.—Continued

Proximal Proximal DistalLength of depth of Distal Distal Length of width of width of

carpometa- carpometa- Length of Length of width Length of width of tarsomet- tarso- tarso-carpus carpus coracoid femur of femur tibiotarsus tibiotarsus a tarsus metatarsus metatarsus

18.9 4.6 22.5 27.3 5.4 52.8+ 4.3 41.6 4.8 3.617.8-19.9 4.0-5.0 21.5-23.2 26.7-28.0 5.0-5.8 51.2±54.2+ 4.0-4.6 40.1-42.7 4.4-5.1 3.5-3.7

9 8 7 7 13 9 19 7 14 14

18.6 4.5 22.3 27.2 5.2 53.4 4.3 40.6 4.8 3.418.1-19.1 4.4-4.6 21.5-23.1 26.8-27.7 5.0-5.4 52.7-54.1 4.2-4.4 39.9-41.3 4.6-4.9 3.2-3.5

2 2 2 2 2 2 2 2 2 2

19.2 4.5 23.2 26.8 5.2 50.1 4.3 38.6 4.7 3.718.3-20.0 4.2-4.7 22.0-23.9 25.5-27.7 4.6-5.4 47.6-52.0 4.2-4.5 - 36.3-40.5 4.5-4.9 3.4-3.9

5 5 4 5 5 5 5 5 5 5

15.41

18.61

22.71

4.41

43.41

3.51

33.01

4.01

3.01

17.6 4.3 20.9 25.1 4.9 46.8 4.0 35.7 4.6 3.317.0-18.1 4.2-4.3 20.2-21.5 24.3-25.8 4.8-4.9 43.9-49.1 3.9-4.1 33.5-37.0 4.4-4.7 3.2-3.4

3 3 3 3 3 3 3 3 3 3

15.91

3.71

19.31

24.01

4.21

45.21

3.51

34.91

4.01

3.21

16.3 4.0 19.8 24.0 4.5 45.6 3.7 35.8 4.3 3.115 2-17 1 3 8-4 2 18.7-20.8 22.8-25.6 4.2-4.8 43.3-48.4 3.6-3.8 33.5-37.8 4.2-4.3 2.9-3.2

3 3 3 3 3 3 3 3 3 3

Floreana, but is extinct on Floreana. The ecologyand behavior of this species are covered in thesection on extinction. Unfortunately, Gould(1837b) named no specific island as the type-locality of M. trifasciatus, although the data onthe two Beagle specimens seem to indicate clearlythat they were taken on Floreana. The holotype

was collected by Darwin, and the other was takenby Fitzroy. Sulloway (1982b) gives additionalinformation on these specimens.

Mimus trifasciatus is one of 4 species of mock-ingbirds in the Galapagos according to Swarth(1931:104-131) and most subsequent authors.Davis and Miller (1960) united all Galapagos

Page 58: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

52 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

mockingbirds under the name trifasciatus (notmelanotis, as stated by I. Abbott and L.K. Abbott,1978), restricting the nominate subspecies to Flo-reana and its satelites. Even if one were to followDavis and Miller (1960) in lumping all of theGalapagos mockingbirds into a single species,none of which is sympatric, the name trifasciatuswould still apply to the Floreana mockingbirdbecause of priority. Sundevall (1871) listed themockingbirds from San Cristobal, Santa Cruz,Santiago, and Floreana all as "Mimus melanotis."On this basis, Salvin (1876:466) and Ridgway(1890:119, 122) erroneously reported that twospecies of mockingbirds, Af. trifasciatus and Af.melanotis, occurred on Floreana.

Recently, I. Abbott and L.K. Abbott (1978)analyzed variation in length of the bill, wing, andtarsus in all populations of mockingbirds in theGalapagos, but reached no firm systematic con-clusions other than recommending that Af. me-lanotis of San Cristobal be lumped with the wide-spread M. parvulus. Mimus trifasciatus is distin-guished from other species of mockingbirds inthe Galapagos as follows (based upon my ownexamination and measurements of skins and skel-etons, supplemented by the data in Ridgway,1907:245-248; Swarth, 1931:111-117; Bow-man and Carter, 1971:248; and I. Abbott andL.K. Abbott, 1978): larger size (much larger inall respects than M. melanotis or M. parvulus,slightly larger than or equal to Af. macdonaldi ofEspanola in all external and skeletal measure-ments except those of the bill); entire dorsumdarker with little or no streaking; less white inthe tail; auricular region lighter in color thancrown (other species have a dark facial patch);brown breast band present (breast white withvarying amounts of streaking in other species).Gifford (1919:207) noted that the song of Af.trifasciatus is different from that of other Gala-pagos mockingbirds. Bowman and Carter (1971)were unable to hybridize captive Af. parvuluswith either Af. trifasciatus or M. macdonaldi. Fur-ther, a canonical analysis of three external meas-urements (I. Abbott and L.K. Abbott, 1978)clustered Af. trifasciatus as far from other Gala-

pagos populations as it was from Af. longicaudatusof the mainland.

The Floreana mockingbird was first describedby Gould (1837b) as Orpheus trifasciatus, butthereafter most authors reported it as Mimustrifasciatus (see Sharpe, 1881:346, and referencestherein) until Ridgway (1890) named the newgenus Nesomimus for all Galapagos mockingbirds.Rothschild and Hartert (1899:142; 1902:381)questioned the validity of the genus Nesomimus,but still recognized it, as have most other authorssince. I agree with I. Abbott and L.K. Abbott(1978) that the distinctions between mocking-birds of the mainland and the Galapagos do notwarrant generic recognition. Designating thespecies melanotis as the type of his new genus,Ridgway (1890:102) diagnosed Nesomimus as fol-lows: "Similar to Mimus BOIE, but bill longerand much more compressed basally, and tarsusmuch longer (nearly twice as long as middle toeinstead of only about one-third longer)." Thesecharacters do not hold. Although Af. trifasciatusand Af. macdonaldi do have longer bills than anyof their non-Galapagos congeners, this is not thecase for Af. parvulus or Af. melanotis, where billsof equal length may be found in M. longicaudatusand Af. saturinus of mainland South America.Even when all Galapagos mockingbirds are in-cluded in the genus Mimus, the variation inlength and curvature of the bill in Mimus is nogreater than in Toxostoma, another genus of Mim-idae. Also, the bill may be as "compressed bas-ally" in M. polyglottos, Af. gundlachii, or M. gilvusas in the Galapagos birds. The tarsus of Galapa-gos mockingbirds is not relatively longer thanthat of Af. gundlachii, Af. thenca, or M. longicau-datus. In fact, mockingbirds of the Galapagos aresimilar enough to Af. longicaudatus to suggestvery strongly that the latter is their direct ances-tor. To separate these species in different generamasks their relationships. The relatively shortertail in the Galapagos birds presents no problemsystematically, for reduction in relative taillength has occurred in other passerines of theGalapagos (flycatchers, finches) when comparedwith their nearest relatives on the mainland.

Page 59: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

N U M B E R 4 1 3 53

Family FRINGILLIDAE

Subfamily EMBERIZINAE

Genus Geospiza Gould

(Darwin's Finches)

Darwin's finches consist of 13 living species inthe Galapagos and 1 species on Cocos Island,Costa Rica. Among the major islands in the ar-chipelago (Figure 1), 3 to 8 species of Darwin'sfinches are recorded from each of the relativelysmall, dry islands, and 9 to 11 species from eachof the larger, more environmentally diverse is-lands. Floreana had 11 species of Geospiza,among which only G. pallida has not been shownto be resident. The evolution of Darwin's finchesis an interesting topic that lies beyond the scopeof this paper. P.R. Grant (1981) and Steadman(1982) review the origin and adaptive radiationof this fascinating group of small, tame birds.

I follow the generic classification of Steadman(1982) and the species-level classification of Lack(1945, 1947). For distributional data, I followLack (1969) and Harris (1973, 1974, 1982), ex-cept where stated otherwise. Identifications offossil Darwin's finches are based on my ownunpublished osteological diagnoses and interspe-cific comparisons. Identification of post-cranialspecimens usually was possible only for G. mag-nirostris, which is larger than any of the otherspecies of Darwin's finches. All other identifica-tions are based on cranial elements, especiallythe rostrum, mandible, and quadrate.

Geospiza nebulosa Gould

(Sharp-beaked Ground Finch)

MATERIAL.—7 specimens, representing atleast 6 individuals (Tables 3, 6, 7). CPOI: USNM330675 (mandible), 338459 (pterygoid). BOC:USNM 330694 (mandible), 330708 (rostrum),330712 (mandible), 330721 (quadrate), 330743(rostrum).

This finch has been recorded historically fromSan Cristobal, Floreana, Santa Cruz, Isabela, Fer-

nandina, Santiago, Pinta, Genovesa, Culpepper,and Wenman. Today, G. nebulosa is either ex-tinct, very rare, or of uncertain status on the firstfour or five of these islands (see "Extinction").

Lack (1945, 1947) united all populations ofthe Sharp-beaked Ground Finch under the nameG. difficilis, a treatment followed by all subse-quent authors until Sulloway (1982a,b) showedthat the earlier name G. nebulosa Gould (1837a)is based on an extinct population from Floreana,and therefore G. difficilis Sharpe (1888) is a syn-onym of G. nebulosa for reasons of priority. Sul-loway (1982a,b) has also determined the specificisland localities for most of the Beagle specimensof Darwin's finches. His measurements of thesespecimens reveal slightly larger bill dimensionsfor the Floreana birds than for any other popu-lations of G. nebulosa. This large size is corrobo-rated by the fossils reported herein, which arealso larger than the cranial elements of otherforms of G. nebulosa, but differ both qualitativelyand quantitatively from those of G. nebulosa'sclosest relatives (G. fuliginosa, G. fortis, and G.scandens).

As is also the case with G. magnirostris, thefossils from Floreana of G. nebulosa representmore individuals than are known from skins.Geospiza nebulosa is known from Floreana by fourskins whose history has been studied by Sulloway(1982b). Two of these specimens, collected byFitzroy in 1835 and Kinberg in 1852, are knownto be from Floreana because of their associatedfield data. The other two, collected by Darwinin 1835, most probably came from Floreana be-cause of circumstantial evidence of the chronol-ogy and geography of Darwin's collecting, andbecause these specimens, one of which is now-lost, agree closely with the other two specimens.

Geospiza nebulosa occurs as a fossil with aboutthe same frequency as G. olivacea. On large is-lands such as Santa Cruz and presumably Flo-reana, both of these species tend to be morecommon, and do most or all of their nesting,above the arid coastal region. They occur regu-larly in the lowlands only in the non-breedingseason (Gifford, 1919:238). Geospiza nebulosa oc-

Page 60: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

54 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

curs more commonly as a fossil than otherfinches, such as G. crassirostris, G. parvula, andG. pauper, that are regarded as characteristic ofareas more humid than the region surroundingthe caves. Therefore, G. nebulosa probably wasonce rather common in the middle and higherelevations of Floreana. It may even have nestedin the lowlands, where it was at least a regularvisitor.

Geospiza fuliginosa Gould

(Small Ground Finch)

MATERIAL.—24 specimens, representing atleast 18 individuals (Tables 3-7). CPOI: USNM330669 (mandible), 330672 (mandible). CPOS:USNM 330679 (mandible). FC: USNM 330681(skull with rostrum), 330683 (mandible), 330684(mandible). BOC: USNM 330689 (mandible),330693 (mandible), 330697 (mandible), 330704(mandible), 330705 (mandible), 330706 (mandi-ble), 330709 (rostrum), 330713 (mandible),330714 (mandible), 330725 (mandible), 330731(mandible), 330737 (rostrum), 330739 (ros-trum), 330742 (mandible), 331188 (mandible),338303 (rostrum), 338304 (rostrum), 338309(mandible).

Geospiza fuliginosa occurs nearly throughoutthe Galapagos. It is the second most commonfossil finch on Floreana, and almost certainly itsnumbers would be augmented if some of thespecimens relegated to "Geospiza, species indeter-minate" were less fragmentary. Now that G. mag-nirostris no longer dominates the finch fauna ofFloreana, G. fuliginosa has become by default themost common finch of the lowlands surroundingPost Office Bay, vying with Dendroica petechia asthe most common land bird in this region today,at least during my observations in June and July1978, and October 1980.

Geospiza fortis Gould

(Medium Ground Finch)

MATERIALS.—38 specimens, representing atleast 12 individuals (Tables 4-7). CPOS: USNM

330678 (rostrum). FC: USNM 330685 (ros-trum), 330680 (associated partial skeleton; jugal,palatine, quadrate, rostrum, mandible, and ma-jor elements of the wing and leg; 23 total speci-mens). BOC: USNM 330688 (mandible), 330692(mandible), 330699 (mandible), 330700 (mandi-ble), 330702 (jugal), 330711 (mandible) 330745(quadrate), 330752 (mandible), 330753 (mandi-ble), 330757 (mandible), 331199 (jugal), 33810(mandible), 338370 (pterygoid).

The partial associated skeleton of G. fortis fromFinch Cave represents the only positively identi-fied post-cranial fossils of any Darwin's finchexcept G. magnirostris. Because of this associatedskeleton, G. fortis has a higher number of speci-mens, relative to MNI, than in any other finchesexcept G. magnirostris. Geospiza fortis is very wide-spread in the Galapagos. It is the third mostcommon fossil finch, and only G. fuliginosa ismore common today in the lowlands of Floreana.Geospiza fortis is extremely variable in size onFloreana (Lack, 1947, PI. IV, tables XXIV,XXIX), as corroborated by the fossils.

Geospiza magnirostris Gould

(Large Ground Finch)

MATERIAL.—2669 specimens, representing atleast 229 individuals (Tables 3-7). All skeletalelements are represented. CPOI: USNM284459-284471, 284652-284670, 284672,284679, 284680, 338462, 338250, 338251,338273, 338284, 338289, 338462. CPOS:USNM 284374-284450, 284452-284458,284480-284497, 284501-284574, 284579-284591, 284600-284638, 284640, 284642,338182-338189, 338195-338204, 338246,338442-338444. FC: USNM 331442-331466,331478-331498, 331505-331555, 331558,331559, 331562, 331565-331571, 331573-331577, 331586-331588. BOC: USNM330722, 330724, 330762-330767, 331114-331176, 331198, 331202-331204, 331221-331275, 331290-331309, 331319-331350,331356-331361, 331373-331410, 331419-331426, 331589, 331591, 331592, 331596-

Page 61: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 55

331598, 338105, 338106, 338126-338149,338158-338163, 338176-338179, 338308,338369, 338372-338375, 338410-338418,338424, 338435.

Geospiza magnirostris is recorded from eachsite. It is the most common avian taxon in thefossil fauna. The fossils include several nearlycomplete associated skeletons (e.g., Figure 23).Both cranially and post-cranially, its large sizedistinguishes G. magnirostris from all other Dar-win's finches (Figure 24; Plates 3, 4).

Geospiza magnirostris, like G. nebulosa, is nowextinct on Floreana. These two finches resembleeach other further in having been major sourcesof controversy among ornithologists of the pastcentury. The basic questions have been: (1) didG. magnirostris ever occur on Floreana?; and (2)if so, why did it become extinct there? Sulloway

(1982a,b) has answered the first of these ques-tions in the affirmative, just as he did for G.nebulosa. In addition, the paleontological evi-dence (2669 fossils) for the occurrence of G.magnirostris on Floreana is unequivocal. By com-bining paleontological evidence with data frommodern ecological studies, the second questionnow can also be addressed (see "Extinction").

Wherever it occurs, G. magnirostris is restrictedmainly to the arid lowlands. This finch occurs,or did occur, on most of the major islands in theGalapagos, but it is seldom as abundant as at least2 or 3 smaller sympatric finches. Based on MNI,G. m. magnirostris is more than 12 times as com-mon in the Floreana fossil deposits as the 2ndmost abundant finch, G. fuliginosa. Barring theunlikely possibility that Tyto punctatissima on Flo-reana was a highly specialized feeder on G. m.

FIGURE 23.—Associated skeleton of Geospiza magnirostris from Finch Cave (USNM 331573;approximately life size).

Page 62: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

56 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 24.—Adult male skins of Geospiza magnirostris (inlateral aspect); top, G. m. magnirostris, BM(NH)1885.12.14.280, Floreana; bottom, G. m. strenua, BM(NH)1899.9.1.171, Genovesa (individuals from Genovesa arelargest of any surviving populations of G. magnirostris; spec-imens X 0.4 life size).

magnirostris, there is little doubt that this largefinch was the most common lowland bird onFloreana before human contact.

Geospiza crassirostris (Gould)

(Vegetarian Finch)

MATERIAL.—4 specimens, representing 4 in-dividuals (Tables 6, 7). BOC: USNM 330691(mandible), 330698 (quadrate), 330707 (mandi-ble), 338387 (tarsometatarsus), 338426 (quad-rate).

Geospiza crassirostris has been recorded fromall of the large, biologically diverse islands in theGalapagos. It nests no lower than the transitionzone on islands such as Floreana, but descendsinto the lowlands in the non-breeding season. InFebruary and March 1906, Gifford (1919:243)found G. crassirostris on Floreana only above1000 feet elevation. I have never seen this speciesin the arid zone of Floreana. It is surprising thatas many as four individuals of G. crassirostris werepreserved as fossils, but if Darwin's three speci-mens of G. crassirostris actually came from Flo-reana, as Sulloway (1982b) has suggested, thenthis finch may have been more common on thisisland in pristine times than today.

Geospiza scandens (Gould)

(Cactus Finch)

MATERIAL.—2 specimens, representing 2 in-dividuals (Tables 6, 7). BOC: USNM 330696(rostrum), 330703 (jugal).

Geospiza scandens occurs on most islands in thearchipelago. It is generally characteristic of aridlowlands, although Gifford (1919:239) reportedG. scandens to occur at higher elevations onFloreana, particularly when introduced orangeswere ripe. In the lowlands of Floreana, Gifford(1919:239) found G. scandens to be most com-mon near the localized stands of cactus. Today,G. scandens seem to be fairly common in the PostOffice Bay region, but it is decidedly less com-mon than G. fuliginosa or G. fortis. The rarity ofG. scandens as a fossil is puzzling, especially whenone considers the former abundance of cactuson Floreana. The overwhelming prehistoricabundance of G. magnirostris may have affectedthe numbers of G. scandens.

Geospiza parvula (Gould)

(Small Tree Finch)

MATERIAL.—1 specimen, representing 1 indi-vidual (Tables 6, 7). BOC: USNM 330730 (man-dible).

Geospiza parvula is the smallest and most wide-spread of the tree finches. It is common on thelarger, higher islands, including Floreana, andnests mainly in the transition and Scalesia zones;it occurs regularly, sometimes commonly, in thearid zone, especially outside of the breeding sea-son. Tentatively, I would rate G. parvula as thefourth most common finch in the lowlands ofFloreana today, behind G. fuliginosa, G. fortis,and G. scandens. I have seen up to 5 or 6 individ-uals in a single day (26 June 1978) in the PostOffice Bay region. Based on its modern abun-dance, one might expect G. parvula to haveoccurred as a fossil in slightly greater numbers.The actual relative abundance of G. parvula inthe fossil record may be masked, however, be-

Page 63: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 57

cause this species is difficult to distinguish osteo-logically from G. fuliginosa. Certain of the frag-mentary fossils reported as "Geospiza, species in-determinate" may pertain to G. parvula.

Geospiza pauper (Ridgway)

(Medium Tree Finch)

MATERIALS.—3 specimens, representing 3 in-dividuals (Tables 5-7). FC: USNM 330682 (man-dible). BOC: USNM 330695 (rostrum), 330738(rostrum).

Geospiza pauper is endemic to Floreana, nest-ing in and above the transition zone. In the non-breeding season, G. pauper regularly occurs inthe arid lowlands. Gifford (1919:249) reportedthat it was seldom encountered below 1000 feetelevation in 1906, although a few specimens weretaken in the lowlands. I found G. pauper to becommon at low elevations near Bahia de las Cue-vas on 23 October 1980, and observed a singleindividual near Post Office Bay on 26 June 1978.The relative abundance of G. pauper in the fossilfauna seems to be more or less as expected.

Geospiza olivacea (Gould)

(Warbler Finch)

MATERIAL.—7 specimens, representing 7 in-dividuals (Tables 3, 6, 7). CPOI: USNM 330673(mandible), 330676 (quadrate), 338285 (frontal).BOC: USNM 330726 (mandible), 331200 (cran-ium), 338423 (quadrate), 338425 (quadrate).

Surprisingly, G. olivacea is the fourth mostcommon fossil finch on Floreana. The relativeabundance of G. olivacea may be biased somewhatby its very small size, which greatly facilitatesidentification. Nevertheless, G. olivacea seems tobe decidedly uncommon in the northern low-lands of Floreana today, where I have neverobserved it in spite of looking for it outside ofthe nesting season (June, July, October). Gifford(1919:220, 223) found G. olivacea to occur un-commonly in the "wooded interior and on thesouth slope" of Floreana in 1906, with breeding

records in February and May. The fossils suggestthat G. olivacea occurred in the arid lowlands ofFloreana more commonly in the past than today,although my observations of its present statuswere limited.

Geospiza, species indeterminate

(Unidentifiable Darwin's Finches)

MATERIAL.—38 specimens, representing atleast 9 individuals (Tables 3, 4, 6, 7). CPOI:USNM 330674. CPOS: USNM 330677,338210. BOC: USNM 330687, 330710,330719, 330720, 330732-330734, 330740,330746, 330749-330751, 330754-330756,330758, 330760, 331201, 338150, 338307,338315-338322, 333871, 338376, 338427-338430.

Indeterminate species of Geospiza are repre-sented by 3 cranial fragments, 2 rostra, 2 jugals,1 pterygoid, 11 quadrates, and 19 mandibles.Each of these fragmentary fossils is smaller thanG. magnirostris and larger than G. olivacea, butcannot be distinguished from two or more otherspecies of Darwin's finches. Most of these speci-mens agree in size with G. fuliginosa or G. par-vula, and very likely belong to one of thesespecies.

Passeriformes, family through speciesindeterminate

(Unknown Passerine Birds)

MATERIAL.—505 specimens, representing atleast 24 individuals (Tables 3-7). CPOI: USNM284650, 330582, 330598, 338252, 338257,338274, 338406-338409, 338451, 338461.CPOS: USNM 284577, 284599, 338190,338224, 338230, 338237, 338404, 338405. FC:USNM 331572, 331584, 338402, 338403.BOC: USNM 331181, 331367, 331368,338112,338383-338401.

This category consists of passerine post-cranialelements that clearly are not from Mimus trifas-ciatus or Geospiza magnirostris because of their

Page 64: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

58 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

small size, or are too fragmentary to determineaccurately even their size. Most of the specimensfall into the first category, and of these, the greatmajority undoubtedly represent Darwin'sfinches. As stated previously, the smaller speciesof Geospiza seem to be impossible to distinguishon the basis of post-cranial elements.

AVES, order through species indeterminate

(Unknown Birds)

MATERIAL.—1281 specimens, representing nonew individuals (MNI) because sufficient num-bers of diagnostic avian fossils were recoveredfrom each of the same sites and levels (Tables 3-7). CPOI: USNM 338445, 338452, 338453,338455. CPOS: USNM 284578, 338181,338191, 338205. FC: USNM 331469, 331500,338437. BOC: USNM 330727, 331182,331197, 331207, 331218, 331278, 331362,331363, 331370, 331413, 331429, 331430,331594, 331595, 351602, 331603, 338107,338108, 338122, 338123, 338164, 338165,338180,338325.

This material consists of avian post-cranial ele-ments that are too fragmentary for identificationeven to order. From their size, however, all ornearly all of these specimens represent birds nolarger that Zenaida galapagoensis. Very likelythese specimens pertain mainly to the three mostcommon fossil species—Geospiza magnirostris,Zenaida galapagoensis, and Mimus trifasciatus.

Class MAMMALIA

Order CHIROPTERA

Family VESPERTILIONIDAE

Lasiurus borealis (Muller)

(Red Bat)

MATERIAL.—59 specimens, representing atleast 11 individuals (Tables 5-7). Nearly all skel-etal elements are represented. FC: USNM331470, 331581, 331583. BOC: USNM

331282, 331315, 331364, 331604, 338117,338151,338324.

Lasiurus borealis is a very widespread speciesthat occurs in much of North, Central, and SouthAmerica. The form of L. borealis that occurs inthe Galapagos was originally described as a newspecies, Atalapha brachyotis, by Allen (1892:47),who noted that "this insular form closely resem-bles A. varia [= L. borealis] in coloration, size andproportions, except that it has much smallerears." Allen's holotype and only specimen wasfrom San Cristobal. Most authors since Allen(1892) have regarded L. brachyotis as a validspecies, although Niethammer (1964) noted thatbrachyotis may be only a race of borealis. While Ihave not made a critical study of the externalmorphology of L. "brachyotis" versus L. borealisof the mainland, I have examined the skulls,dentaries, and other bones of these forms, find-ing no justification for recognition of brachyotisas a species distinct from borealis. In the absenceof a series of skins from the Galapagos, Allen'scharacter of smaller ears in "brachyotis" cannotbe evaluated. The Galapagos form may deserverecognition as a subspecies of borealis, but thepaucity of specimens prevents any decision in thisregard. Lasiurus borealis from the Galapagos islarger than most races of L. borealis from themainland, although it is the same size as certaincontinental races.

To my knowledge, no specimens of L. borealishave ever been taken before on Floreana. With-out providing the date or exact locality, Brosset(1963) noted a sight record of Lasiurus (speciesundetermined) from Floreana, this being theonly evidence of the occurrence of bats on Flo-reana until the fossils reported herein. I neversaw bats during my field work on Floreana. Theonly other bat in the Galapagos is L. cinereus,which is significantly larger than L. borealis. Spec-imens of L. cinereus from the Galapagos havebeen reported only from Santa Cruz.

What little is known of the biology of bats inthe Galapagos has been reviewed by Brosset(1963), Orr (1966), Vanalek (1982), and Clark(1984). Brosset (1963) noted that Tyto punctatis-

Page 65: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 59

sitna will prey upon L. borealis, and probably thisis how the fossils from Floreana were depositedin the caves, for L. borealis roosts in vegetation,not in caves. I have also identified fossils of L.borealis from Cueva de Rubier, Santa Cruz, thesebones certainly being derived from prey items ofT. punctatissima.

Introduced Mammals

The remaining mammals have been intro-duced on Floreana by man. Because they are nota part of Floreana's natural fauna, I have notincluded these species in any of the totals inTables 3-7. None of these specimens is likely tobe older than A.D. 1832, the year that GeneralJose Villamil established the first large humansettlement on Floreana (see "Human History").For nomenclature, I follow Corbet (1978).

Order RODENTIA

Family MURIDAE

Subfamily MURINAE

Mus musculus Linnaeus

(House Mouse)

MATERIAL.—42 specimens, representing atleast 12 individuals (Tables 3, 4, 6, 7). Mostmajor skeletal elements are represented, includ-ing maxillae and dentaries. CPOI: USNM330573, 338448. CPOS: USNM 284598,338217. BOC: USNM 331190, 331283,331435,331437,331609.

Although we collected this introduced rodentfrom excavations as well as surface deposits, acritical look at its occurrence in the excavationsreveals no evidence for great antiquity of Musmusculus on Floreana. From Cueva de Post Of-fice (Inferior), it is represented by 1 pelvis in the0-3 cm level. From Cueva de Post Office (Su-perior), it occurs again only in the 0-3 cm level

(1 premaxilla, 2 dentaries). From Excavation 1in Barn Owl Cave, M. musculus was found onlyin the test pit, in the form of 26 specimens (3MNI) from the 0-10 cm level, 5 specimens (1MNI) from the 10-20 cm level, and 1 specimenfrom the 30-40 cm level. The stratigraphy oftest pits is never fool-proof, and the six specimensof Af. musculus from below 10 cm are undoubt-edly contaminants that dribbled into lower levelsduring excavation of the test pit. After comple-tion of the test pit, we enlarged Excavation 1horizontally in a much more controlled manner,according to its stratigraphic units. Mus musculuswas not present in the stratigraphically controlledexcavation. This little rodent is discussed furtherin "Introduced Mammals."

Tyto punctatissima probably was the main orsole agent for deposition of Mus musculus in thecaves. The rarity of both Mus musculus and Rat-tus rattus, however, suggests that these two intro-duced rodents co-existed on Floreana with T.punctatissima for only a very short time.

Rattus rattus (Linnaeus)

(Black Rat)

MATERIAL.—142 specimens, representing atleast 13 individuals (Tables 3-7). All skeletalelements are represented. CPOI: USNM330571, 330572, 330588. CPOS: USNM284451, 284476. FC: USNM 331471. BOC:USNM 331211, 331284, 331415, 331436,338116.

As with Mus musculus, the specimens of Rattusrattus from excavations cannot be considered asevidence of its prehistoric occurrence on Flo-reana. From Excavation 3 of Cueva de PostOffice (Superior), this introduced rodent is rep-resented by a single humerus from the upper-most 10 cm of sediment, which therefore couldhave been on the surface originally. From Exca-vation 1 of Barn Owl Cave, R. rattus occurs onlyas one upper incisor and one dentary from the0-10 cm level of the test pit, and one vertebrafrom Unit A, the highest unit. Rattus rattus isdiscussed further in Introduced Mammals.

Page 66: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

60 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Order CARNIVORA

Family FELIDAE

Felis catus Linnaeus

(Domestic Cat)

MATERIAL.— 1 specimen, representing 1 indi-vidual (Tables 6, 7). BOC: USNM 331191 (ra-dius).

Felis catus is recorded only from the surface ofRoom 1 in Barn Owl Cave. Cats are too large tobe prey items of Tyto punctatissima. (Instead, catsmay have preyed upon T. punctatissima duringthe barn owl's last days on Floreana.) Felis catusthrives today in the lowlands of Floreana.

Order ARTIODACTYLA

Family SUIDAE

Sus scrofa Linnaeus

(Domestic Pig)

MATERIAL.—16 specimens, representing atleast 1 individual (Tables 6, 7). BOC: USNM331212 (13 post-cranial elements), 331215(skull, mandible), 331314 (radius).

These bones were scattered over an area ap-proximately 4 m in diameter on the surface ofRoom 1 of Barn Owl Cave. They likely belongto one individual of 5. scrofa because of theirsimilarity in preservation, and because they allrepresent a very young pig. Until very recently,feral pigs were common on Floreana.

Order PERISSODACTYLA

Family EQUIDAE

Equus asinus Linnaeus

(Donkey)

MATERIAL.— 1 specimen, representing 1 indi-vidual (Tables 3, 7). CPOI: USNM 338477 (ker-atinous hoof sheath).

The lack of any other donkey remains makesme question that the entire donkey was ever inthe cave. Instead, a dog, cat, pig, or rat may havebrought the hoof sheath of E. asinus into thecave. Donkeys are common today in the lowlandsof Floreana.

Mammalia, order indeterminate

(Unknown Mammal)

MATERIAL.—2 specimens, representing 2 in-dividuals (Tables 3, 5, 7). CPOI: USNM 338478(fused thoracic vertebrae). FC: USNM 338463(cranial fragment).

The first of these surface remains is from animmature individual, dog-sized or larger. Thecranial fragment is approximately the size ofCapra or Sus, but lacks any diagnostic features.

Discussion

HUMAN HISTORY.—An understanding of Flo-reana's human history is crucial to this study, forall vertebrate extinction on Floreana is relatedtemporally to human arrival. Many accounts ofthe settlements on Floreana are vague and con-flicting, and that which follows is my best attemptto piece together briefly the major historicalevents on this island.

Except for the dubious possibility of limited,temporary Amerindian encampments (Heyer-dahl and Skjolsvold 1956; Heyerdahl 1963), thediscovery of the Galapagos Islands is credited toFray Tomas de Berlanga, the Bishop of Panama,in 1535. The particular islands that he visitedcannot be determined with certainty from hiswritings. Privateers and pirates frequented theGalapagos, including Floreana, in the late 17thand early 18th centuries. Here these sailors couldmake repairs and obtain food (especially tor-toises) and water in between raids on Spanishships and coastal towns of Latin America. In thelate 1700s, Galapagos waters became the focusof whaling ships from the United States andBritain. Initially, Floreana was visited often be-cause it offered both tortoises and fresh water,

Page 67: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 61

as well as good anchorage. When tortoises be-came very rare in approximately 1840 (see "Ex-tinction"), Floreana became a little less popularwith the whalers, who were drawn there never-theless by the fresh water and vegetables.

In 1807, Patrick Watkins became Floreana'sfirst long-term resident. This Irish sailor wasprobably the first settler anywhere in the entirearchipelago. Slevin (1959:105) and others havesummarized Watkins' harsh life on Floreana,based upon the diary, written in 1813, of Mid-shipman William W. Feltus of the U.S. frigateEssex, and upon the journal of Captain DavidPorter of the Essex (Porter, 1822). Watkins livedon Floreana for approximately 2 years beforehijacking a boat to mainland South America.

In 1832, General Jose Villamil established asizeable colony on Floreana with the permissionof the newly formed Ecuadorean government.This colony was the first significant human set-tlement anywhere in the Galapagos. Villamil'scolony is described by Darwin (1871:141, 142),who noted the presence of introduced plants andanimals there in 1835. Villamil's settlementlasted only until 1845 or 1846, after which Flo-reana was sporadically inhabited (although ratherfrequently visited) until 1929, when two Ger-mans, Dr. Karl Friedrich Ritter and Frau DoreStrauch, settled in the highlands, remainingthere until Ritter's death in 1934, when Strauchreturned to Germany (Strauch, 1936). The Witt-mer family, also of Germany, moved to Floreanain 1932. The Wittmer's initially consisted ofHeinz, his wife Margret, and a son. Another sonand a daughter were born subsequently. FrauWittmer, who now lives at Black Beach (thefamily originally settled in the highlands), hasrecounted her family's experiences on Floreana(Wittmer, 1961). Later in 1932, BaronessAntoinette (Eloise) von Wagner-Bousquet movedto Floreana with three men. Controversy andmistrust filled the lives of the Baroness and hercompanions, as well as Ritter, Strauch, and theWittmer's, as one can easily discern from theconflicting accounts in Strauch (1936), W.A. Ro-binson (1936), Conway and Conway (1947), and

Wittmer (1961). The ensuing tragic events thatoccurred on Floreana in 1934 still rate as themost popular of unsolved mysteries in the Gala-pagos (Treherne, 1983).

Various persons have moved to Floreana sincethe 1930's, the most prolific being the Cruzfamily, some of whose 11 children still live onFloreana (Gayle Davis, pers. comm.). Wittmer(1961:221, 234) reported that approximately 50persons lived on Floreana during the period of1956-1959. Floreana's population today is stillaround 50 (Gayle Davis, pers. comm.), livingboth in the highlands and at Black Beach.

INTRODUCED MAMMALS.—"We may inferfrom these facts what havoc the introduction ofany new beast of prey must cause in a countrybefore the instincts of the indigenous inhabitantshave become adapted to the stranger's craft orpower" (Darwin 1871:176).

Perhaps more than any other island in theGalapagos, Floreana has been adversely affectedby mammals introduced by man. These feralmammals, which may be more harmful to nativeinsular animals than man himself, are probablyinvolved in the extinctions of as least six speciesof vertebrates on Floreana, although evidencefor this accusation is no better than circumstan-tial. Only the extinction of the hawk Buteo gala-pagoensis is difficult to relate to the impact offeral mammals. Large animals are more vulner-able to direct human predation than smaller an-imals because they are more conspicuous. OnFloreana, predation by man may be the chiefcause of extinction for Geochelone and Buteo, butmay be involved as well in the loss of Alsophisand Tyto.

Slevin (1959:7), Leveque (1963), and Thorn-ton (1971:268) noted that the following speciesof feral mammals occurred on Floreana: blackrat (Rattus rattus), house mouse (Mus musculus),dog {Canis familiaris), cat {Felis catus), pig {Susscrofa), goat (Capra hircus), cow {Bos taurus), andburro or donkey (Equus asinus). Eckhardt (1972)also listed all of these species as currently inhab-iting Floreana, with cats and donkeys designatedeither as especially abundant or particularly de-

Page 68: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

62 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

structive. Eckhardt (1972) is an excellent refer-ence for the ecological effects of introducedplants and animals in the Galapagos, although hemakes little reference to Floreana in particular.During my field work on Floreana in 1978 and1980, I noted cats, goats, and donkeys to becommon in the arid northern coastal region. Idid not observe pigs, cattle, or dogs on Floreana,but I have spent only part of a single day in thehighlands. I have not seen Rattus on Floreana,but I have identified bones of both Rattus andMus from cave deposits on Floreana as well asfrom modern pellets of Short-eared Owls (Asioflammeus) that roost on Champion but huntmainly on nearby Floreana. The islands of SanCristobal, Santa Cruz, Santiago, and Isabela havemore or less the same species of feral mammalsas Floreana (Leveque, 1963; Eckhardt, 1972),but on these four larger islands feral mammalsseem to have done somewhat less damage tonative vertebrates. Table 10 is a very currentassessment by Bruce D. Barnett of the introducedanimals on Floreana, where the low figures fordogs, pigs, and cattle are due to an effectivehunting and poisoning campaign for the pastseveral years.

Once established, most populations of feralmammals thrived on Floreana up to the present,or at least up until the past decade. Sheep, men-

TABLE 10.—Numbers of individuals of introducedvertebrates on Floreana, February 1984 (data from BruceD. Barnett, as related to Marsha S. Cox).

Animal

MiceRatsDogsCatsPigsGoatsCattleBurrosHorsesDucksChickensPigeons

Domestic

00

252044

3265

551712

4272

Feral

"Infinite""Infinite"

0

<100010-205000

50-60>1000

10

-1500

tioned only by Tanner (1888) and A. Agassiz(1892:68), would be an exception. Most speciesof feral mammals escaped into the wilds of Flo-reana within 10 years after 1832, the year thatVillamil established the settlement there. It ispossible that buccaneers or whalers releasedgoats on Floreana prior to 1832, but the sugges-tion of Wittmer (1961:32) that goats and cattlewere released on Floreana by Fray Tomas deBerlanga in 1535 seems highly unlikely. CaptainDavid Porter of the Essex released goats on San-tiago as early as 1814 (Porter, 1822), but I knowof no earlier report of feral mammals anywherein the Galapagos. Darwin (1871:142) noted feralgoats and pigs on Floreana in September 1835.By 1846, Berthold Seemann of HMS Heraldstated (in Van Denburgh, 1914:226) that "wilddogs, pigs, goats, and cattle had increased won-derfully" on Floreana, and Villamil's settlementalso owned approximately 2000 head of cattle.

Captain A.H. Markham visited Floreana in1880 in HMS Triumph, finding the island to be(1880:744) "in undisturbed possession of the so-called wild cattle. . . . . donkeys, dogs, pigs, andother animals that had been left to run wild onthe abandonment of the island by the formerinhabitants." In 1887, Midshipman M. Estienneof the French corvette Decres reported (inSlevin, 1959:103) abundant donkeys on Flo-reana, as well as wild cattle and pigs. CaptainTanner (1888) of the Albatross noted large num-bers of cattle, sheep, hogs, horses, and donkeysrunning wild on Floreana in 1888. A. Agassiz ofthe Albatross found cattle, donkeys, sheep, goats,hogs, cats, dogs, and "common fowl" in a feralstate on Floreana in 1891 (A. Agassiz, 1892:68).Slevin (1931:39-43) reported cats, dogs, goats,cattle, pigs, and donkeys seen on Floreana by theCalifornia Academy of Sciences Expedition inOctober 1905. Strauch (1936), Conway and Con-way (1947), and Wittmer (1961) collectivelymentioned rats, mice, cats, dogs, pigs, goats,cattle, donkeys, and horses on Floreana between1930 and 1960, with pigs and cattle being verydestructive to their crops.

Patton et al. (1975) stated that rats occurred

Page 69: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 63

on Floreana by sometime in the 1800s, notingthat Darwin made no mention of rats being therein 1835. The first specimen of Rattus from Flo-reana was taken by the Hopkins Stanford Gala-pagos Expedition in 1898-1899 (Heller 1904).Rats may have been living on Floreana in 1835or very shortly thereafter, but were simply over-looked. On Santa Cruz, where rats seem to beabundant today (personal observation), they caneasily go unnoticed to someone who is not spe-cifically looking for them.

While I am convinced that alien introductionstriggered the extinction of native forms, the di-rect evidence of such is elusive. Rats may beinvolved in the extinction of Geochelone (preda-tion on eggs and hatchlings) and Alsophis (pre-dation on young). Atkinson (1977) hypothesizedthe potentially detrimental effects of Rattus rattuson birds in Hawaii, through direct predation andtransmission of disease. There is, however, verylittle unequivocal evidence of actual predationon birds by Rattus (Norman 1970, 1975), or oftheir role in transmission of disease. Atkinson(1977) noted that avian extinction coincided withthe introduction of Rattus on Hawaii, Lord HoweIsland, and Big Smith Cape Island in New Zea-land. The role of Rattus in the extinction of birdson Floreana is purely speculative. I am not awareof studies of any sort on Mus in Galapagos, nordo I know that Mus has had any damaging effecton the native vertebrates of Floreana or anyother island in the Galapagos or elsewhere.

Certain vertebrates on Floreana probably havesuffered heavily from predation by feral cats anddogs. For example, Strauch (1936:93) and Witt-mer (1961:136) both owned cats on Floreanathat regularly ate finches and doves. Cats anddogs may be involved in the extinctions of Geo-chelone, Alsophis, and perhaps Buteo, Tyto, Mimus,Geospiza nebulosa, Geospiza tnagnirostris, and per-haps in the present rarity of Tropidurus andZenaida. Fortunately, feral dogs have been elim-inated on Floreana in the past decade by poison-ing and shooting (Barnett, 1982), thus ending aperiod of 140 years as the largest carnivores onthe island.

Among the introduced ungulates of Floreana,only the pig may prey directly on vertebrates.While rooting around in the soil and groundcover, pigs probably kill small individuals andeggs of Geochelone and Alsophis. Koford (1966)noted that residents of Santa Cruz claim that pigsdestroy the nests and eggs of tortoises.

Goats, cattle, and burros destroy the nativevegetation by their relentless browsing and graz-ing. Goats are especially notorious for habitatalteration on islands, both in the Galapagos andelsewhere (Coblentz, 1978). In arid regions suchas the Galapagos, succulents are preferred be-cause of their high water content. Hamann(1975) has reviewed the damage done to vege-tation by introduced herbivores, especially goats,on Santa Cruz, Santa Fe, and Pinta; the samegeneral principles apply to Floreana. Koford(1966) mentioned that Floreana had, at thattime, the densest population of donkeys in theGalapagos.

EXTINCTION.—Seven species of vertebratesfrom Floreana have become extinct, six of whichare recorded as fossils. The fossils themselvesprovide no evidence concerning the cause ofextinction, but the main importance of the fossilsis in determining the presence or absence of aspecies at a given time, in this case the lateHolocene. Having established that a certain spe-cies once lived on Floreana, then various kindsof independent, circumstantial evidence can besought to suggest causes of extinction.

The chronology of extinction on Floreana sug-gests man's involvement; all or nearly all extinc-tion has occurred since man's arrival. Except forthe barn owl, all extinct vertebrates from Flo-reana are known to have survived into historictimes, i.e., to A.D. 1835 or later. While the barnowl is not recorded definitely from Floreanaother than from fossils, I believe that its extinc-tion was also in historic times (see below). A fairlyextensive fossil record from Floreana has dis-closed no dear cases of prehistoric extinction. Ifman's impact is a major cause of extinction onislands, then a paucity or lack of prehistoricextinction is not unexpected on an archipelago

Page 70: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

64 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

such as the Galapagos where prehistoric manprobably never occurred, or at most visited on avery limited basis. New, dated fossil faunas fromother islands in the Galapagos will provide thecrucial test. I predict that most or all extinctionin the Galapagos will be shown to have occurredin historic times. Radiocarbon dating is essential.For example, Steadman and Ray (1982) couldsay little about the extinction of the giant rat ofSanta Cruz, Megaoryzomys curioi, because of thelack of radiocarbon dates from any of the locali-ties from which it was collected.

Geochelone elephantopus: Presumably, extinc-tion of tortoises on Floreana was due mainly todirect human predation, although predation oneggs and young by introduced mammals such asrats, cats, dogs, and pigs may have also contrib-uted. Various accounts of visits to the Galapagosin the 17th, 18th, and early 19th centuries men-tion the tremendous numbers of tortoises thatlived there, and Floreana was no exception. Thebuccaneers, whalers, and explorers of this timeoften took on board hundreds of tortoises duringa single landing. As a result, tortoises becamevery rare or extinct on islands that initiallyseemed to have an inexhaustible supply. My ac-count of the history of man-tortoise relationshipsin the Galapagos is taken mostly from Baur(1889), Van Denburgh (1914), Townsend(1925a), and Slevin (1935, 1959).

Floreana has not been active volcanically inhistoric times, so volcanic catastrophies can beeliminated as a possible cause of extinction oftortoises there. Fernandina is the only island inthe Galapagos where volcanism may be involvedin the extinction or rarity of tortoises. Only asingle tortoise has ever been collected on Fernan-dina (on 6 April 1906, by Rollo Beck). Fernan-dina is the most active volcano in the Galapagos,and most of its area is covered with fresh, barrenlava flows. If tortoises ever occurred naturallyand are now extinct on Fernandina, then thisextinction is undoubtedly due to natural means,i.e., volcanic activity.

Tortoises can survive for months without foodor water, so they are an ideal source of fresh

meat for sailors. According to Captain BenjaminMorrell (1832; in Baur, 1889:1055, 1056):

They are an excellent food, and have no doubt saved thelives of thousands of seamen employed in the whale-fishingin those seas, both American and Englishmen. I have knownwhale-ships to take from six to nine hundred of the smallestsize of these tortoises on board when about leaving theislands for their cruising grounds; thus providing themselveswith provisions for six to eight months, and securing themen against the scurvy. I have had these animals on boardmy own vessels from five to six months without their oncetaking food or water; and on killing them I have found morethan a quart of sweet fresh water in the resceptacle [sic]which nature has furnished them for that purpose, whiletheir flesh was in as good condition as when I first took themon board. They have been known to live on board some ofour whale-ships for fourteen months under similar circum-stances, without any apparent diminution of health orweight.

On large islands, female tortoises occur morefrequently in the arid coastal regions than males,which tend to concentrate in the humid high-lands (Hendrickson, 1966). Thus human preda-tion initially took many more females than males,simply because the females were more accessible.Medium-sized tortoises (50-100 lbs.) were pref-erentially sought because they were not tooheavy to transport, yet they could yield signifi-cant quantities of meat. Tortoises were takenmainly for their meat, but their oil was also anattractive commodity for many tortoise hunters.Residents of various islands in the Galapagos,including Floreana, killed tortoises regularly andin large numbers for their oil alone. The tortoisesfrom Floreana and Hood Island (= Espafiola)may have been favored above those of otherislands, for Porter (1822:233) stated that thosefrom Espafiola "were of a quality far superior tothose found on James Island. They were similarin appearance to those of Charles Island, very fatand delicious."

Rose (1924:354) reasoned that Floreana wasthe island where buccaneer Edward Da vies of theBatchelor's Delight obtained many tortoises in1687, which would be the first record of humanpredation on the Floreana tortoise. Throughmuch of the 18th century, however, removal of

Page 71: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 65

tortoises from Floreana probably occurred at alow rate because of the infrequency of shipspassing through the Galapagos. American andBritish whaling ships became numerous in Gala-pagos waters from the late 1700's until the late1800's. These ships were very destructive totortoises.

By the year 1800, tortoises were still abundanton Floreana, although the whaling ships un-doubtedly had already reduced their numbers.The largest decline of Floreana tortoises oc-curred in the last 50 years of their existence, theperiod from 1800 to 1850. Amasa Delano (1817)reported that tortoises were plentiful on Flo-reana in 1801 and several years thereafter.Known removals of tortoises from Floreana from1812 through 1837 are compiled in Table 11,although this list is very incomplete. For exam-ple, no fewer than 31 whaling ships called atFloreana between 13 October 1832 and 30 Au-gust 1833 (Reynolds, 1835; in Baur, 1889). Rey-nolds (in Townsend, 1925a) estimated that eachof these ships took on board 200 tortoises,whereas Townsend (1925a) reported an averageof 138 tortoises taken by each of 9 whaling shipsthat called at Floreana in early and mid-1830s.Thirty-one ships taking 138 tortoises each overa period of IOV2 months would result in theremoval of 4890 tortoises per year from Floreanaby ships alone, and this figure may be an under-estimate. Baur (1889) stated that about 100,000tortoises were removed from the Galapagos sincetheir discovery, but subsequently he revised thisfigure (in Townsend, 1925a,b) to 10,000,000.The truth may lie somewhere in between.

E.C. Cornell (in Townsend, 1925a:95) notedthe presence of hatchlings in 1816, suggestingthat tortoises were still reproducing at a signifi-cant rate, which suggests further that most or allof the feral predators were absent at that time.Visits by whaling ships to Floreana increased infrequency after 1832, for General Villamil's set-tlement could provide passing ships with freshfruit and vegetables. The availability of freshwater, produce, and tortoises made Floreana themost beneficial stop in the entire archipelago.

TABLE 11.—Tortoises taken from Floreana by ships (mostlyAmerican whalers) from 1812 to 1837 (data mainly fromTownsend (1925a, 1928), supplemented by Porter(1822:160) and Slevin (1935; 1959:74, 128)).

Year/Ship

1812 Sukey1813 Essex

Briton1814 u

1816 FJiza«

Apollo1820 Essex1824 Wasp

" Loan1828 India1831 Magnolia

Frances1832 Hector1833 Octavia1834 Bengal

MossBenezetL. C. Richmond

1835 BarclayBenezetPioneer

1836 OhioPioneer

1837 Eliza AdamsTotal

Number oftortoises

25030

400-500"a few"

24"boatloads"

74300

60100394100155179226+235100350120?

5040pp?24

3311+

Number of daysof tortoise

hunting

p

1

11127??p?

257?2

1551368424

77+

Van Denburgh (1914:220) reported that thepeople who colonized Floreana in 1832 (see "Hu-man History") had, with the help of their feralmammals, "reduced the number of tortoisesupon Charles Island so rapidly and to such anextent that within three years [= 1835] the peo-ple were obliged to send hunting parties to otherislands to procure a supply of food." In 1835,Darwin (1871:144) encountered on James Island(= Santiago) a group of men from Floreana whowere hunting tortoises. Yet tortoises were stillbeing hunted successfully on Floreana at thattime, for Darwin (1871:142) noted that, "thestaple article of animal food [of the people living

Page 72: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

66 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

on Floreana in 1835) is supplied by the tortoise.Their numbers have of course been greatly re-duced in this island, but the people yet count ontwo days' hunting giving them food for the restof the week."

Very shortly after 1835, however, the Flo-reana tortoise declined to the point of no return.The combined hunting pressure from humanresidents, whalers, and feral mammals was simplyexhaustive. The whaler's logbooks examined byTownsend (1925a, 1928) record no tortoisestaken from Floreana after 1837, except for a fewobtained in 1847 by the ship Congaree and in1848 by the ship Aurora. Townsend reasoned,however, that these tortoises probably were notnative to Floreana, but were imported fromother islands.

Reasonable estimates of the date of extinctionof the Floreana tortoise range from 1840 (Heller,1903) to 1850 (Broom, 1929). The French frig-ate La Venus visited Floreana in 1838, just threeyears after the Beagle's visit, and made collectionsof birds and plants, but made no mention oftortoises (Slevin, 1959:56). Berthold Seemann ofHMS Herald visited Floreana in 1846 and re-ported the tortoise to be extinct, whereas wilddogs, pigs, goats, and cattle were abundant (VanDenburgh, 1914:226). Dr. Kinbergof the Swed-ish ship Eugenie found no tortoises on Floreanain 1852. The U.S. whaling ship Fabius visitedFloreana for one day in 1858, sending "boatsashore after turtle" (Townsend, 1925a:82).Their results are not known. Professor LouisAgassiz, on board the U.S. steamer Hassler, pur-chased a tortoise on Floreana in 1872, and Town-send (1928) listed three tortoises taken on Flo-reana in 1882 by the U.S. whaler Atlantic. Broom(1929) rightly cautioned, however, that any tor-toises collected on Floreana after 1850 wereprobably brought there from another island bythe settlers of Floreana. The following statementof Heller (1903:45) is misleading in that it impliesthat Commander W.E. Cookson of HMS Peterelcollected tortoises on Floreana in 1875: "Com-mander Cookson . . . collected some reptiles,chiefly tortoises, at Abingdon, Albemarle, andCharles." In fact, Cookson (1876) reported that

the Floreana tortoise had become extinct 20-30years before 1875. (Steindachner (1876) andGunther (1877a,b) reported on the reptiles col-lected by Cookson.) All subsequent explorers ofFloreana have also failed to find living tortoises,including the thorough search made by the Cal-ifornia Academy of Sciences Expedition in 1905-1906 (Slevin, 1931:5; see J.R. Slevin's herpeto-logical field notes in Van Denburgh, 1914:317,318, and in Fritts and Fritts, 1982). All thingsconsidered, the year 1850 is a very reasonableestimate of the date of extinction of the Floreanatortoise, although it must have been extremelyrare during the last decade of its existence.

Tortoises on Floreana may have become ex-tinct or nearly so even if direct human predationhad never occurred; the 150 years of combinedefforts by alien mammals may have been suffi-cient by itself to reduce tortoise populations be-yond recovery. Rats, cats, dogs, and pigs eatyoung tortoises or tortoise eggs, while feralbrowsers and grazers (donkeys, cattle, and espe-cially goats) are detrimental to tortoise popula-tions through competition for food (MacFarlandet al., 1974a,b). Even if the adult population werehealthy, the feral predators may be capable ofpreventing successful recruitment. Cookson(1876) stated that dogs kill not only very youngtortoises, but also those weighing up to 60 lbs.Townsend (1928) mentioned that rats and catsfeed on newly hatched tortoises. Thus it may bethat Floreana's feral mammals played a signifi-cant role in wiping out the tortoises; predationby seafarers and residents probably was not theunique cause of their extinction.

Alsophis biserialis: This snake is known fromFloreana by only one specimen taken in 1835 byCharles Darwin. Presumably it is extinct there,but extinction is difficult to prove because of theinconspicuous habits of snakes. Alsophis may pos-sibly survive on Floreana in extremely reducednumbers, simply having eluded collectors of thepast 150 years. Snakes had never been collectedon San Cristobal, for example, until three speci-mens were taken in 1957 (Mertens, 1960). TheCalifornia Academy of Sciences Expedition col-lected one snake on Gardner-near-Floreana Is-

Page 73: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 67

land on October 1905, this specimen apparentlybeing conspecific with the snake of Floreana itself(Van Denburgh, 1912a:337). As mentioned pre-viously, snakes are common today on Championand Gardner-near-Floreana. The CaliforniaAcademy of Sciences Expedition collected 98snakes on other islands in the Galapagos from 24September 1905 to 25 September 1906 (VanDenburgh, 1912a:330; Slevin, 1931:34, 150), anaverage of one snake for every 3.7 days in thefield. They spent 28 days on Floreana (see theherpetological field notes of J.R. Slevin in VanDenburgh, 1914:317, 318, and in Fritts andFritts, 1982), which therefore could have beenexpected to yield approximately 7 or 8 snakes.Instead they found none.

Scarcity or absence of snakes on Floreana canbe attributed to man and his introduced mam-mals. Van Denburgh (1912a:338) stated:

Snakes must be very rare on Charles Island, for none wereseen there by any member of our expedition [CAS], althoughcareful search was made for them. It is probable that theravages of the smaller kinds of mammals that have beenintroduced there—particularly rats and cats—have pushedthem to the verge of extinction, as they have the Tropidurus.It is probable that a longer search would show that snakesare still to be found on Champion and Enderby as well as onGardner, for Tropiduri [sic] still are fairly abundant on allthese islets.

Presumably the predatory actions of rats, cats,dogs, and pigs, as well as humans, have beendevastating to the Floreana snake. Humans areparticularly fond of killing snakes of any sort, outof simple fear and loathing.

The vegetational damage wrought by goats,cattle, pigs, and donkeys on Floreana may alsohave been detrimental to snakes. Goats alonemay not be very damaging to snakes in the shortrun because snakes are still common today onSantiago, Espanola, and Sante Fe, three islandsthat are, or have been until recently, heavilypopulated by goats. Discussing modern tropicalsituations, Janzen (1976:372) stated:

Through intensive grazing, browsing, and trampling, espe-cially near watercourses during severe dry seasons, largeherbivores should greatly reduce the cover available forreptiles, the small vertebrate prey of snakes, and the insectsavailable to reptiles.

Similar damages by large herbivores on Floreanamay have been very detrimental to Alsophisbiserialis as well as the lizard Tropidurus grayii.

Judging from the size of collections from in-dividual islands of the California Academy ofSciences Expedition (Van Denburgh, 1912a:330), snakes were common in 1905-1906 onEspanola, Santa Cruz, and Santa Fe, and werereasonably common on Santiago, Fernandina,Baltra, and Rabida. At that time, only Santiagoof the above islands had rats (Patton et al., 1975)and pigs, and probably none of these islands hadferal cats or dogs. Isabela, San Cristobal, andFloreana were the large islands where snakeswere rare or non-existent in 1905-1906; each ofthese islands already had been settled by peopleand had obtained a variety of feral mammals,including rats. The effect of introduced mam-mals on snakes can be inferred further from thecase of Santa Cruz. As noted above, snakes werecommon on Santa Cruz when this island lackedintroduced mammals. Since the 1920's, SantaCruz has been colonized intensively by people,and now it boasts a full complement of rats, cats,dogs, pigs, goats, cattle, and donkeys, not tomention several thousand people. Snakes arevery rare on Santa Cruz today; I have never seena snake there in over four months of field work.Snakes are common today in the Galapagos onlyon Espanola, Santa Fe, Seymour Norte, Pinzon,Rabida, Santiago, and Fernandina (Robert P.Reynolds, pers. comm.; personal observation).Each of these islands except Santiago lacks intro-duced mammals. Santiago is heavily populatedby goats, pigs, and rats, so the abundance ofsnakes there is difficult to explain.

To summarize, the decline and probable ex-tinction of snakes on Floreana is more speculativethan that of the Floreana tortoise. It is mucheasier to say that an island has no tortoises thanto say that it has no snakes. Most or all of theintroduced mammals were well established andundoubtedly affecting the snake population ofFloreana by the 1830's. Snakes were probably invery low numbers by 1850, and may not havesurvived the close of the 19th century.

Buteo galapagoensis: Specimens of B. galapa-

Page 74: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

68 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

goensis have never been reported from Floreana,nor is this hawk mentioned specifically to occurthere in early accounts, such as those of Porter(1822), Darwin (1871), and Markham (1880).Nevertheless, this passage from Gould (1841:25),based upon Darwin's observations on hawks in1835, provides fairly convincing evidence that B.galapagoensis did indeed once live on Floreana:" . . . on all islands, it [the hawk] is excessivelynumerous. . . . It is extremely tame, and fre-quents the neighborhood of any building inhab-ited by man. . . . These birds will eat all kinds ofoffal thrown from houses. . . . They are said tokill young doves, and even chickens. . . . " Thesestatements could pertain only to Floreana, forthe other three islands visited by the Beagle wereuninhabited by people in 1835. Because it wentunrecorded on Floreana by all subsequent collec-tors, B. galapagoensis must have died out or beenseverely depleted soon after 1835.

Providing no details, Thornton (1971:150),Harris (1973; 1974:37, 86; 1982:37, 87), Vries(1975), and Groot (1983) have noted that hawksare extinct on Floreana. This extinction mayhave been due to human predation because B.galapagoensis is very tame. On Isabela and Fer-nandina I have approached these hawks to within0.5 m. This hawk can be killed easily with fire-arms, and in fact, is still hunted in the Galapagos(Duffy, 1981; Groot, 1983). Hawks are also pre-sumably extinct on the settled island of San Cris-tobal (Harris, 1973; 1974:37,86; 1982:37, 87),where the last specimen was taken in 1905-1906by the California Academy of Sciences Expedi-tion (Swarth, 1931:50). On Santa Cruz, hawkswere abundant in 1905-1906 (Gifford,1919:190), but are very rare today (Harris,1973). Santa Cruz was first settled in the 1920'sand 1930's, and now supports more people thanany other island in the Galapagos.

Snodgrass and Heller (1904:265) noted thatthe absence of hawks on Floreana may be due tothe scarcity of Tropidurus, but it seems unlikelythat a change in availability of prey, such as thedrastic reduction in numbers of snakes and lavalizards, could have caused the hawk's extinction.Buteo galapagoensis is an extremely opportunistic

and versatile predator and scavenger (Vries,1976), and probably could have thrived on therats, mice, and large mammal carrion that be-came available on Floreana with the arrival ofman. Direct human predation appears to be themost reasonable explanation for the loss of hawkson Floreana.

Tyto punctatissima: Barn owls were not re-corded definitely from Floreana until I foundtheir fossils in Barn Owl Cave. Nevertheless, Ibelieve that barn owls died out in historic times,as with all other extinct vertebrates from Flo-reana. Otherwise, the remains of black rats (Rat-tus rattus) and house mice (Mus musculus) in thecaves would be difficult to explain. The otherowl in the Galapagos, Asio flammeus, does notroost in lava tubes.

The extirpation of T. punctatissima may haveoccurred through severe population declines inits preferred prey species, and probably wasaided by direct predation on the owls by people,cats, or dogs. Rodents dominate the diet of T.punctatissima on other islands. Floreana has al-ways lacked native rodents, which would haveappeared in the fossil sites had they been present.Thus T. punctatissima never may have been ascommon on Floreana as it is (or was) on therodent-bearing islands of San Cristobal, SantaCruz, Santiago, Isabela, and Fernandina. Groot(1983) attributed the absence of barn owls onFloreana to the lack of rodents, but the fossilrecord shows that barn owls did exist and breedon Floreana in the absence of native rodents.Based upon fossils, the 7 most common verte-brate prey items for barn owls on Floreana were,in descending order, Geospiza magnirostris, Tro-pidurus grayii, Zenaida galapagoensis, Phyllodac-tylus baurii, Mimus trifasciatus, Alsophis biserialis,and hatchling Geochelone elephantopus (Tables 7,8). Collectively, these 7 species made up approx-imately 86% of the barn owl's diet. Geospizamagnirostris, Mimus trifasciatus, Alsophis biserialis,and Geochelone elephantopus are now extinct onFloreana, whereas Tropidurus grayii and Zenaidagalapagoensis survive only in extremely reducednumbers. Phyllodactylus baurii is the only one ofthe 7 most common prey species that still prob-

Page 75: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 69

ably occurs on Floreana in "normal" numbers.But this tiny gecko is the smallest vertebrate onFloreana and could never provide the bulk of anowl's diet. Otteni et al. (1972) noted a drasticdecline in yearly reproductive rates of T. alba inTexas when their preferred food (in this case,small mammals) declined in numbers. Therefore,a rapid but long-term decline in preferred prey,such as occurred on Floreana, could have provedfatal to an insular population of barn owls. Themajor decline in prey species occurred approxi-mately from 1830 to 1860, so T. punctatissimamay have died out on Floreana by the 1860s or1870s.

Killing of barn owls by people and feral mam-mals may also have been involved in their disap-pearance. Such predation may have providedonly the "coup de grace," or it may have beenjust as important as loss of prey in wiping out T.punctatissima. In the absence of predation fromman and feral mammals, T. punctatissima mayhave been able to adjust to the change in preyavailability (i.e., fewer native animals, more in-troduced rats and mice) that occurred with thepeopling of Floreana. Harris (1973) stated thatman probably was directly involved in the loss ofT. punctatissima from Floreana, presumablythrough predation. I believe that most 19th cen-tury encounters between human residents andbarn owls were fatal to the latter, for this owl isamazingly tame, and can be killed very easily. Aswith hawks, residents still kill barn owls today onother islands in the Galapagos (Duffy, 1981;Groot, 1983). In 1906, the California Academyof Sciences Expedition collected one barn owl onIsabela with a stick, and three on Santa Cruz thatwere perched only 6-7 feet away (Gifford,1919:194). On Santa Cruz, I have approachedbarn owls to within arm's reach without makingthem fly. That T. punctatissima could have ex-isted on Floreana in the 19th century but eludedcollectors is not unrealistic, for few ornithologistsvisit caves or do much collecting at night.

Mimus trifasciatus: This mockingbird hasbeen extinct on Floreana since approximately the1870s. It survives today only on Champion andGardner-near-Floreana, with respective total

populations (in 1980) of 48 and 150-200 individ-uals (P.R. Grant, 1980). Other than the fossilsreported herein, only 3 specimens of Af. trifascia-tus are known from Floreana. The Beagle Expe-dition collected two specimens in September1835 (Gould, 1837b), and Dr. Kinberg of theSwedish frigate Eugenie collected a single speci-men in May 1852 (Sundevall, 1871). As far as Ican determine, no specimens of Af. trifasciatushave been taken on Floreana since 1852. Thelast record of mockingbirds on Floreana is fromDr. A. Habel, whose field notes of 1868 (inSalvin, 1876:472) reported that "the MockingThrushes there [Floreana] differed in their live-lier and more intelligent habits, and in theirsuperior powers of song [compared to Af. mac-donaldi of Espanola.]" Habel took no specimensof mockingbird from Floreana.

Baur (1895) was the first to suggest that Af.trifasciatus was extinct on Floreana, followed byRidgway (1897:482, 483). Ridgway was unawareof Kinberg's specimen taken in 1852, and thusstated that nobody had collected Af. trifasciatussince Darwin's visit in 1835. Rothschild and Har-tert (1899:142, 143) were next to mention theextinction of mockingbirds on Floreana, basedon the failure of collectors such as Habel (in1868), the naturalists of the Albatross (in 1888,1891), Baur and Adams (in 1891), and Websterand Harris (in 1897), as well as others, to procureany specimens. (As noted above, however, Habeldid observe mockingbirds on Floreana.) Since1899, Af. trifasciatus has been reported to beextinct or probably extinct on Floreana by Snod-grass and Heller (1904:358, 359), Rothschild(1907:xi, xii), Gifford (1919:207), Swarth(1931:114, 117), Hellmayr (1934:334), Lack(1947:23), Harris (1968, 1973, 1974:36, 37,128; 1982:36, 128), Bowman and Carter (1971),Thornton (1971:79,80,157,160), and I. Abbottand L.K. Abbott (1978). Swarth (1931:117) sug-gested that the mockingbirds collected by theBeagle Expedition may have been taken on Gard-ner-near-Floreana instead of Floreana itself. Thiscaution was reiterated by Thornton (1971:160)and Bowman and Carter (1971).

As mentioned previously, Af. trifasciatus still

Page 76: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

70 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

occurs on Floreana's satellite islands of Cham-pion and Gardner-near-Floreana. I have not vis-ited Garner-near-Floreana, so I will confine mydiscussion to the mockingbirds living on Cham-pion. The flora of Champion seems to be pristine(Figure 25; also see plant lists in B. Voigt and A.Voigt, 1977; P.R. Grant, 1980). We are fortun-ate that this small offshore island has preserveda flora that probably resembles that which ex-isted on the adjacent lowlands of northern Flo-reana before the arrival of man. During visits toChampion on 4 July 1978, 26 October 1980,and 24 May 1983,1 observed at least 12 differentmockingbirds, each of which was either on theground or in an arborescent prickly pear cactus(Opuntia megasperma var. megasperma). Althoughother trees and shrubs were common, such asBursera graveolens, Croton scouleri, Jasminocereus

thouarsii, Parkinsonia aculeata, and Prosopis juli-flora, I did not observe the mockingbirds to usethese plants in any way. The only time that I sawa mockingbird on Champion use any plant otherthan Opuntia was a single bird that landed brieflyin a shrub (Cordia luted) on 24 May 1983. Myobservations are corroborated by those of severalother authors. Gifford (1919:207) reported find-ing mockingbirds on Champion near Opuntia,with "a nest or two . . . in nearly every good-sized cactus tree." Harris (1974:128) reportedthe nest of M. trifasciatus to be "usually a substan-tial mass of twigs placed in a cactus." B. Voigt(1977b) noted: "Old nests found only in cacti,about 2 m above ground." Mockingbirds fromother islands in the Galapagos also use Opuntiafrequently as a nesting site (Rothschild and Har-tert, 1902:383; Gifford, 1919:209-214; Harris,

FIGURE 25.—Gayle Davis among the abundant cactus, Opuntia megasperma var. megasperma(Champion Island, May 1983).

Page 77: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 71

1974:126, 128; P.R. Grant and N. Grant, 1979),but they are not as completely dependent uponOpuntia as is Af. trifasciatus.

Opuntia may also be important to Af. trifascia-tus as food, although no quantified food studiesexist. Mimus trifasciatus feeds on flowers andfruits of Opuntia, as well as grasshoppers, ants,moths, spiders, and centipedes from the groundand rotting Opuntia pads, eggs of Sula nebouxii,and flowers of the prostrate vine Convulvulus(Hatch, 1965; Bowman and Carter, 1971; DeRidder, 1976; B. Voigt, 1977a,b; B. Voigt andA. Voight, 1977; B.R. Grant, 1981; Harcourt,1982).

To summarize, Af. trifasciatus feeds, nests, androosts exclusively in or near Opuntia, a plant thatis essential to the survival of Af. trifasciatus onChampion. A major vegetational difference be-tween Champion and Floreana is Champion'smuch greater density of Opuntia. One wouldexpect to find Opuntia commonly in the aridlowlands of Floreana just as on other islands inthe Galapagos, but in fact both species of colum-nar cactus (Jasminocereus thouarsii var. thouarsii,Fig. 46A of Dawson, 1962; and Opuntia mega-sperma var. megasperma, Figure 25) survive onFloreana mainly on steep, inaccessible, rocky out-crops, such as on the tops of dikes or verticallyeroded tuff cones. Floreana's small, offshore is-lets are the only places left in the Floreana regionwhere cacti are still more or less as they existedbefore human contact (Figure 25). Porter(1822:162) found "prickly pears in great abun-dance" on Floreana in 1813, so there is no reasonto doubt that cactus was not abundant on Flo-reana before 1832, the initial year of major hu-man occupation. The botanist Hugh Cumingvisited Floreana in 1829 (Howell, 1941), threeyears before Villamil settled the island, but I amunaware of any notes describing the cactus hepresumably found there.

The destruction of cactus on Floreana wascaused by feral ungulates. The arid lowlands ofFloreana lack fresh surface water, so the succu-lent stems and pads of cactus provide a very goodsource of moisture for large introduced herbi-

vores. Townsend (1930) and Dawson (1962)blamed feral donkeys and cattle for the scarcityof Opuntia on Floreana. Koford (1966) and Ha-mann (1975) discussed damage to Opuntia in theGalapagos by feral goats, an extreme example ofwhich is found by comparing figures 1 and 6 ofHamann (1975). According to Dawson (1962)and Andre De Roy (pers. comm.), donkeys arestronger and therefore more destructive to cac-tus than are goats, so that goats alone are lesseffective in destroying cactus than is the combi-nation of goats and donkeys. Beginning in the1830's, I believe that goats and donkeys heavilydamaged the cactus of Floreana. As the Opuntiadisappeared, so did the mockingbirds.

The extinction of Af. trifasciatus on Floreanawas not as rapid as that of Geospiza magnirostris,another bird whose loss I attribute to the rarityof Opuntia in the lowlands of Floreana. I doubtthat the longer survival of Af. trifasciatus was dueto its being any less dependent on Opuntia thanwas G. magnirostris. Instead, Af. trifasciatus mayhave been slightly better at avoiding predationfrom feral cats and dogs. Such predation mayhave added the final touches to wiping out bothG. magnirostris and Af. trifasciatus, once they werelocalized because of loss of habitat.

Predation by dogs and especially cats has beensuggested as the sole cause of extinction of Af.trifasciatus, although Swarth (1931:117), whoquestioned whether or not Af. trifasciatus everoccurred on Floreana, doubted that such preda-tion was an important factor. Swarth pointed outthat the northern coast of Floreana, adjacent toChampion, lacked cats and dogs early in the 20thcentury according to J.R. Slevin. Swarth's sug-gestion seems unlikely, however, for cats anddogs were certainly established elsewhere on Flo-reana at this time, as indicated by Slevin himself(1959:7; see "Introduced Mammals"). Harris(1974; 1982:36, 37) also doubted that cats wereresponsible for the loss of Af. trifasciatus on Flo-reana. On Santa Cruz and southern Isabela, Af.parvulus appears to thrive today in the presenceof cats and dogs, and the same is true for Af.melanotis on San Cristobal. Behavioral differ-

Page 78: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

72 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

ences, however, may have made M. trifasciatusmore vulnerable to predation than M. parvulusor M. melanotis. By leading a rather specializedlife that involves spending essentially all of itstime either in cactus or on the ground, M. trifas-ciatus may have been much easier to prey uponthan M. parvulus or M. melanotis, which are notso dependent upon Opuntia. Lastly, Duffy (1981)suggested that pathogens, perhaps introduced bydomestic fowl, may have been involved in theloss of M. trifasciatus.

If feral ungulates were removed from Flo-reana, the vegetation of the arid lowlands mayregenerate toward what it was like before the19th century. If Opuntia increased in abundance,M. trifasciatus might be able to live once againon Floreana, if feral carnivores were eradicated.Re-colonization of Floreana would probably re-quire transplanting individuals from Championor Gardner-near-Floreana, for I. Abbott andL.K. Abbott (1978) noted that no stray mocking-birds from these nearby islands have ever beenrecorded on Floreana. Captain David Porter(1822:163) found mockingbirds in great num-bers on Floreana in 1813. To restore their for-mer abundance would seem to be a worthy goalof future conservation efforts in the Galapagos.

Geospiza nebulosa nebulosa: Lack (1945:9, 10,14, 15) was the first author to suggest that G.nebulosa may be extinct on Floreana. Until Lack'sdiscovery that G. nebulosa was a large form of G."difficilis," ornithologists since Salvin (1876) hadregarded G. nebulosa as a synonym of G. fortis(see account of G. nebulosa in "Systematic Paleon-tology"). This synonymy not only masked thetrue relationships of G. nebulosa, but also coveredup the fact that nobody had collected G. nebulosaon Floreana since 1852. Geospiza fortis, on theother hand, had been collected regularly on Flo-reana since the Beagle's visit in 1835, and is stillcommon there today. By regarding G. nebulosaas synonymous with G. fortis, ornithologists hadno clue that they were ignoring an extinct, mor-phologically recognizable population. Subse-quent to Lack (1945), the following authors havenoted that an extinct form of G. "difficilis" may

have occurred on Floreana: Lack (1947:23, 120;1969), Bowman (1961:270), Paynter (1970:162),Harris (1973; 1974:36, 144; 1982:36, 145), Sul-loway (1982a,b, who resurrected the name ne-bulosa), and Schluter and P.R. Grant (1982).

Geospiza nebulosa was either extinct or veryrare on Floreana by the 1860s or 1870s. It waslast collected in 1852 by Dr. Kinberg of theSwedish frigate Eugenie. Floreana was visitedrather frequently from the 1860s onward, yet nospecimens of G. nebulosa were procured. Almostcertainly it was extinct no later than the turn ofthe century, for the California Academy of Sci-ences Expedition found none during their inten-sive collecting effort on Floreana in 1905-1906.That the Beagle crew took 3 specimens of G.nebulosa from Floreana in 1835 suggests that itwas not a rare bird at that time. This was justthree years after the initial settlement of Flo-reana, so G. nebulosa may not yet have beenaffected greatly by human habitation. Neverthe-less, total extinction of G. nebulosa may haveoccurred rapidly once its population began todecline. On Santa Cruz, G. nebulosa was commonin 1905-1906 (Gifford, 1919:238), but appar-ently had vanished by the late 1930's (Lack,1945:13), within a decade after the highlands ofSanta Cruz began to be cleared extensively forcattle grazing.

Referring to the history of Santa Cruz, Lack(1945:9, 10, 13), Bowman (1961:270), Harris(1974, 1982:36), and Sulloway (1982b) postu-lated that habitat destruction in the highlandsmay have been responsible for the extinction ofG. nebulosa on Floreana as well. Bowman(1961:270) noted that predation from feral catsmay have been another factor. I. Abbott et al.(1977:170) cited "lack of preferred habitat orfood" and "competition" as reasons for the ab-sence of G. nebulosa from Floreana. If "lack ofpreferred habitat" can be taken to mean habitatdestruction, then the reasons of I. Abbott et al.(1977) seem compatible with those of Sulloway(1982b), who discussed the extinction of G. ne-bulosa more thoroughly than any previous au-thor. Sulloway stated (1982b:89):

Page 79: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 73

Although once found on all of the larger islands in thearchipelago, G. difficilis has definitely become extinct onseveral of them, probably owing to ground clearing andcultivation in the humid zone. It is hardly surprising, then,that G. difficilis may have encountered this same fate onCharles Island, which was the first island to be settled, in1832. Within just a few years, ecological disturbances asso-ciated with the settlement were already manifesting them-selves. Darwin [see Barlow, 1963:264] specifically noted thatthe larger species of ground finches, which normally preferthe arid lowlands, were extremely common on Charles Islandnear the cleared tracts as the highlands settlement. Indeed,these ground finches had become quite troublesome to thesettlers, eating seeds buried up to six inches in the cultivatedfields. Thus, by the mid-1830s any endemic Charles Islandpopulation of G. difficilis would have been faced with twothreats to its continued existence: a diminishing habitat andincreased competition from other species of Darwin's finchesthat are normally restricted to the lower altitudes.

To summarize, I believe that the extinction ofG. nebulosa on Floreana, as well as on Santa Cruzand perhaps San Cristobal and Isabela, can beattributed to a combination of habitat destruc-tion (probably affecting both nesting and feed-ing), predation by feral cats and possibly dogs,and perhaps competition from newly invadingfinches such as G.fortis and G.fuliginosa. Schluterand P.R. Grant (1982) concluded that interspe-cific competition is the best available explanationfor the nearly mutually exclusive distributions ofG. nebulosa and G. fuliginosa. This would seemto support competition as an important factor inthe extinction of G. nebulosa on Floreana andelsewhere.

On Genovesa and Pinta, P.R. Grant and B.R.Grant (1980) and Schluter and P.R. Grant (1982)studied in detail the food habits of G. nebulosaacutirostris and G. n. difficilis, respectively, but toextrapolate these results to Floreana would berisky because the bill in G. n. nebulosa of Floreanais so much larger than in G. n. difficilis andespecially G. n. acutirostris. Our information onthe habits of G. nebulosa in the highlands of largeislands is from Gifford (1919:238), who foundthem commonly on Santa Cruz in 1905— 1906

in the thickly vegetated region of the lower humid belt,usually feeding on the ground under bushes, often in flocks.More than once we shot at one, mistaking it for a rail, so

skulking were its habits. . . . in the arid region below 75 feetelevation . . . they were found, as formerly in the humidbelt, under bushes, digging vigorously in the grass and dryleaves.

Gifford found G. nebulosa to feed on theground also on Santiago, and summarized theirhabits by stating that (p. 239) "it is strictly terres-trial and does not feed in the trees as do theother species." Its terrestrial habits may havemade G. nebulosa more vulnerable to predationby feral carnivores than other small finches.

Geospiza magnirostris magnirostris: This finchis almost certainly extinct on Floreana. Claims toits existence after the Beagle collections are re-futed in Steadman (1984). Based on its abun-dance as a fossil, G. m. magnirostris must havebeen an extremely common bird prior to thearrival of humans. It makes up 75% of the totalfossil finch fauna, being over 12 times morenumerous than any other species of Darwin'sfinch. Geospiza m. magnirostris constitutes 86% ofthe finches from surface deposits, and 65% ofthe finches from excavations. Probably the lastfigure more closely approximates the actual rel-ative abundance of this large finch as a fossil, for,as stated previously, the surface collections werebiased by the size of G. magnirostris. Regardless,the fossil evidence suggests that G. m. magnirostriswas once very abundant in Floreana's lowlands.

In 1813, Porter (1822:163) found on Floreanagreat numbers of "a small black bird, with aremarkably short and strong bill, and a shrillnote." Porter's bird was probably G. m. magniros-tris. This species was still common on Floreanain 1835, three years after the establishment ofVillamil's colony. The Beagle Expedition col-lected 15 specimens of Darwin's finches on Flo-reana in 1835, of which five, or 33%, were G. m.magnirostris. If we assume the likelihood that theBeagle specimens of G. m. magnirostris were col-lected in the lowlands of Floreana, whereas atleast some of the Beagle specimens of G. nebulosaand G. crassirostris (which total five specimens)were from the highlands of Floreana, then G. m.magnirostris would constitute up to 50% of thelowland finch fauna collected by the Beagle crew.

Page 80: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

74 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

This figure is not too far from the 65% derivedfrom paleontological data, although the relativeabundances of the Beagle specimens are at best acrude approximation of relative density becauseof small sample size, collectors' biases, and thepossibility of lost specimens.

Geospiza m. magnirostris probably disappearedfrom Floreana within a decade or two after theonly historic specimens were collected. That noadditional specimens were taken after 1835 issignificant negative evidence, for G. m. magniros-tris was probably rather conspicuous in life, andby necessity all trips to Floreana had to passthrough its habitat in the arid lowlands. Theearliest post-Beagle collecting efforts on Floreanawere those of Adolphe-Simon Neboux andCharles-Rene-Augustin Leclancher of theFrench frigate Venus in 1838, Thomas Edmon-ston of HMS Herald in 1846, and Dr. Kinbergof the Swedish frigate Eugenie in 1852. Collec-tively, these three expeditions spent more than13 days (perhaps as many as 20 days) on Floreanaand obtained specimens of Geospiza nebulosa, G.fuliginosa, G. fortis, G. scandens, and G. parvula,but not G. magnirostris (Sundevall, 1871; Sharpe,1888; Sulloway, 1982b). It is particularly note-worthy that the Venus collectors obtained nospecimens of G. m. magnirostris during their 11days on Floreana in 1838, only three years afterthe Beagle's visit. The Beagle crew may havecollected G. m. magnirostris during its final sev-eral years of abundance.

Salvin (1876:479) specifically noted that Dr.A. Habel did not encounter G. magnirostris dur-ing his visit to Floreana from 27 July to 12August 1868. Gifford (1919:225) reported thatthe California Academy of Sciences Expeditiontook specimens of G. magnirostris on Floreana,but Swarth (1931:147) pointed out that Giffordwas actually referring to G. fortis and not G.magnirostris. Bowman (1961:20, 271) cited aspecimen he collected as evidence of the modernand past occurrence of G. magnirostris on Flo-reana. This specimen is much too small, however,to represent the large Floreana race of G. mag-nirostris (Steadman, 1984).

As on Floreana, G. m. magnirostris has not beenrecorded from San Cristobal since the Beagle'svisit in 1835 (Sulloway, 1982b). Ridgway(1890:121) noted that the Albatross Expeditionrecorded it there in 1888, but he cited no speci-mens or other details of this record. I havesearched in vain for such a specimen in theUSNM collection, where the Albatross specimensare housed. San Cristobal was settled approxi-mately a decade after Floreana, so perhaps G. m.magnirostris survived there a little longer than onFloreana.

Rothschild and Hartert (1899) were the firstto suggest that G. m. magnirostris may be extincton Floreana.

On Charles Island . . . probably at least one or two thick-billed finches have become extinct. As the earliest settlementof men has been on Charles Island, and as we know thatthey had no regard for the birds—sailors, finding the tame-ness of the birds strange and novel, used to take a cruelpleasure in knocking them down with sticks—we are prob-ably right in ascribing these disappearances merely to humaninfluence [p. 142].

It is probable that G. magnirostris [on Floreana] is exter-minated or extremely scarce. This is quite possible when weconsider that Nesomimus trifasciatus has disappeared fromCharles Island, and that these finches, according to Darwin[see Gould, 1841:100], did "much injury by digging up rootsand seeds from a depth of even six inches." It is thereforeto be supposed that they were killed by the colonists, whocomplained of their injuries, and who first settled on CharlesIsland about 1830 [p. 154].

Townsend (1928:168; 1930:154) suggestedthat the loss of G. m. magnirostris was due topredation by cats and rats. Lack (1947:23) saidthat its extinction probably was due indirectly tothe human settlement on Floreana, which maymean habitat destruction or predation by feralmammals. Duffy (1981) mentioned introducedpathogens as a possibility. The following authorshave regarded G. m. magnirostris as possibly,probably, or certainly extinct on Floreana, with-out proposing any specific cause: Ridgway(1901:493, 496), Lack (1946, 1969:253, 254,261), Paynter (1970:161), Harris (1973,1974:143; 1982:144), and Sulloway (1982a,b).

To summarize the story up to this point, G. m.

Page 81: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 75

magnirostris was once a very common bird in thelowlands of Floreana, but its extinction occurredrapidly, perhaps within a decade after the Bea-gles visit to Floreana in 1835. This extinctionhas been attributed to human agency throughpredation by introduced cats and rats, as well ashumans themselves. The fossil record sheds nonew light on the extinction problem, but doesdemonstrate the former abundance of G. m. mag-nirostris.

By itself, predation from alien animals doesnot seem to be sufficient to cause the rapiddecline and extinction of G. m. magnirostris onFloreana. Other ground finches, such as G. fuli-ginosa and G. fortis, were probably less commoninitially than G. m. magnirostris. Why should thesespecies survive such predation while G. m. mag-nirostris did not? If predation, especially by catsand humans, could act this rapidly by itself, thenone might expect that Zenaida galapagoensiswould have become extinct first. Yet this ex-tremely tame and tasty, ground-dwelling dovehas survived (although just barely) 150 years ofpredation on Floreana from man and his feralmammals.

As with Mimus trifasciatus, I believe that thesevere depletion of Opuntia on Floreana by feralungulates is related to the extinction of G. m.magnirostris. This idea has recently been putforth by B.R. Grant and P.R. Grant (1982:653),but requires additional evidence and explana-tion. Both the bill of G. m. magnirostris and theseed of Opuntia megasperma var. megasperma areextremely large. In fact the seed of O. m. var.megasperma is much larger and harder than inany other Opuntia in the Galapagos, including O.m. var. orientalis of San Cristobal and Espanola(Howell, 1933; Dawson, 1962, fig. 65A-C, 1966;Wiggins and Porter, 1971:545; B.R. Grant andP.R. Grant, 1982). Howell (1934:516) describedthe seed of O. m. var. megasperma as "like small,somewhat compressed marbles about a half inchin diameter." No finch of ordinary strength couldcrack such a seed.

I. Abbott et al. (1977) formulated a compositeindex of size (depth) and hardness (kg of force)

of seeds to evaluate the ability of Darwin's finchesto feed on different types of seed. B.R. Grantand P.R. Grant (1982:653) report that the size-hardness index for seeds of O. m. var. megaspermafrom Champion is greater than 20, compared to10.9 for O. m. var. orientalis from Espanola anda range of 2 to 7 for other forms of Opuntia inthe Galapagos. The size-hardness index of Opun-tia seeds appears to be correlated positively withthe bill size of G. magnirostris from the sameisland; Floreana simply represents the highestvalues for each of these characters: These largesizes evolved together on Floreana, resulting ina dependence of G. m. magnirostris on Opuntiafor much of its food. In addition, Racine andDownhower (1974, table 4) reported that O. m.var. megasperma produces a much greater volumeof seeds than any other Opuntia in the Galapagos.Opuntia m. var. orientalis from San Cristobal hasthe second greatest seed volume, but has thegreatest fruit volume of all. B.R. Grant and P.R.Grant (1981, 1982) noted that the fruit sur-rounding the seed is fibrous in 0. megasperma,whereas it is fleshy in O. echios, O. helleri, and O.galapageia. Thus both the seed and fruit of O.megasperma are tougher than in other species ofOpuntia.

Geospiza magnirostris feeds not only on theseeds of Opuntia, but also the flowers, pollen,fruit, and pad fibers (P.R. Grant and B.R. Grant,1980; B.R. Grant and P.R. Grant, 1981). Opuntiaoccupied 22.5% of the foraging time of G. mag-nirostris on Rabida in June (I. Abbott et al.,1977:181). On Genovesa in November, G. mag-nirostris spent 64.6% of its foraging time onOpuntia seeds on the ground, as well as 4.5% onOpuntia pad fiber (P.R. Grant and B.R. Grant,1980, table 18). Geospiza m. magnirostris may alsohave depended heavily upon Opuntia for nestingsites, for 28 of 46 nests of G. m. strenua onGenovesa were in Opuntia (P.R. Grant and B.R.Grant, 1980, table 12).

If G. m. magnirostris was dependent uponOpuntia for feeding and nesting, then a cause ofits extinction on Floreana can be proposed. Be-ginning in the 1830's, feral goats and donkeys

Page 82: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

76 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

severely reduced the Opuntia in the lowlands ofFloreana, thereby destroying the major sourceof food and nesting sites for G. m. magnirostris.Feral herbivores may also have competed withG. m. magnirostris for the relatively few remain-ing fruits of Opuntia. By itself, reduction of Opun-tia may not have completely wiped out G. m.magnirostris, but the small, localized groups thatsurvived may have succumbed to predation byrats, cats, dogs, or humans. Thus I suggest thatthe extinction of both Mimus trifasciatus and G.m. magnirostris was due mainly to the loss ofOpuntia. The apparently more rapid extinctionof the latter species of bird may be attributedeither to a need for larger territories (i.e., agreater inability to survive in small, localizedpatches of Opuntia), or to a greater vulnerabilityto feral predators, or both.

The extinction of G. m. magnirostris on SanCristobal may have occurred in the same manneras on Floreana because, as mentioned above, theOpuntia on San Cristobal also has large, hardseeds and tough fruit. B.R. Grant and P.R. Grant(1982) have recently suggested that G. magniros-tris (subspecies unknown) may once have oc-curred on Espafiola, but became extinct beforeit was recorded by scientific collectors. This in-teresting idea awaits verification by the discoveryof fossils on Espafiola, which might be difficultbecause no lava tubes have ever been reportedthere.

In discussing extinction on Floreana I havetaken the position that all faunal losses are dueto the impact of man or his introduced mammals.My position is supported by the fossil record,which discloses no clear-cut losses in the last fewthousand years before human arrival. This raisesseveral questions. For example, to what degreeare insular extinctions random events? By docu-menting extinctions, can the Holocene fossil rec-ord be used to illuminate studies of island bio-geography? Can human impact or other causesof extinction be related to the observed differ-ences in intensity and chronology of extinctionon different groups of islands?

THEORETICAL BIOGEOGRAPHY AND FOSSILS.—

The past 20 years have witnessed an explosionof publications on the diversity and distributionof insular biotas. Stimulated mainly by the theo-retical models of Preston (1962a,b) and Mac-Arthur and Wilson (1963, 1967), researchershave analyzed the floras and faunas of manyislands and archipelagos, attempting to support,modify, or refute these models, which are partof an all-encompassing "equilibrium theory ofisland biogeography." My main intention here isto demonstrate the importance of paleontologi-cal evidence in understanding patterns of insularbiogeography. I will preface this, however, withsome non-paleontological criticisms of theoreti-cal island biogeography.

The species-area equation, S = CAZ, is onemodel of MacArthur and Wilson (1967:8, 9) thathas received an enormous amount of attention."S" is the number of species of a particular tax-onomic group on an island, "A" is the area of theisland, "C" is a constant "that depends on thetaxon and biogeographic region, and in particu-lar most strongly on the population density de-termined by these two parameters," and "z" is aconstant "that changes very little among taxa orwithin a given taxon in different parts of theworld." Both C and z are fitted to the dataavailable for S and A. For a related set of insularfloras or faunas, a log-log plot of S (Y-axis) versusA (X-axis) will yield a fitted regression line ofsupposed predictive abilities whose slope is z andwhose Y-intercept is C. Preston (1962a) ex-pressed the species-area equation as N = KAZ,but since most workers follow the terminologyof MacArthur and Wilson (1967), I will do thesame.

The species-area equation, as well as otheraspects of the equilibrium theory of island bio-geography, has enjoyed great popularity; the fewserious attempts to point out its shortcomingshave generally been ignored. For example, Haasexposed several major flaws in the species-areaequation, stating (1975:371):

In fitting data of any sort there is bound to be some error,which may produce uncertainty about the fitted constants[C, z] and thus the conclusions based on them. The reader

Page 83: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 77

should not be forced to program the data anew himself inorder to judge how much to trust these fitted constants.. . . wide confidence intervals are not untypical and shouldbe considered more fully before any discussion of the differ-ences between particular z values, which may have beeninfluenced very strongly by a few data points. . . . the confi-dence intervals have a tendency to flare as they move awayfrom the mean, the most stable point on the regression line.Errors or poor fits about the data points will tend to "rock"the regression line about the mean, affecting the end pointsof the curve most. For this reason, the confidence intervalsflare. This means that if one has a species-area curve andwishes to use it to predict the number of species on a nearbyisland, he must be ready to accept almost any number asfitting his prediction (making it meaningless) if the island isvery much smaller or larger than those already considered.

Haas (1975:372) also noted that the theorybehind the interpretation of the constants C andZ is poorly understood (and therefore ignored),but that this

really should come as no surprise, since many factors affect-ing species abundance influence and, in turn, are absorbedby these two constants. This is especially so with C, which isaffected variously by the density of the organisms, numberof species in the taxa, degree of isolation (MacArthur andWilson, 1967), and the scale with which the area is measured.In fact, so much variation is sopped up by C that particularvalues of it are hardly ever discussed. Although used pri-marily as an index of isolation or "islandness," the parameterz must also be influenced by other factors.

A biogeographic analysis of Hawaiian birds byJuvik and Austring (1979, fig. 2) is a good ex-ample of the shortcomings of the species-areaequation discussed by Haas. Juvik and Austringdo not give confidence intervals, and their datafor the two smallest islands are essentially mean-ingless, or at least are not readily comparable tothose of the six larger islands, because the num-bers of species of birds and the areas of the islandsinclude drastic changes in magnitude (from 1 to20 species and from 0.77 to 10,464 km2, respec-tively). Further, their graph has only 8 points perregression line. Finally, as I will discuss further,their biogeographic analyses failed to anticipatehow much richer the Hawaiian avifauna was inprehistoric times.

Sauer (1969) made another detailed critism ofthe equilibrium theory of MacArthur and Wilson

(1967) that deserves more attention than it hasreceived. He wrote (1969:590, 591, 593):

In short, the equilibrium model and its derivatives sufferfrom extreme oversimplification by treating islands as func-tional units with no attention to internal habitat diversityand by treating species as interchangeable with no allowancefor genetic or geographical diversity. This is not even goodas a first approximation, because it filters out the interpret-able signal instead of the random noise. The authors are insuch a hurry to abandon the particulars of natural historyfor universal generalization that they lose the grand themeof natural history, the shaping of organic diversity by envi-ronmental selection. A model that visualizes various sizes ofassemblages of characterless species on various sizes of fea-tureless plains is essentially absurd, since it excludes the verybasis of genesis and continued coexistence of multiple spe-cies.

MacArthur and Wilson recognize the impracticality ofmeasuring immigration and extinction rates in their modelsbut suggest that these rates might be deduced from variationamong islands in equilibrial species number and speciescomposition. They are also hopeful about direct measure-ment of colonization rates. However, all these approachesrequire synoptic tabulations of entire biotas. Available tab-ulations, including most of those the authors cite, are usuallyaccumulations of observations over extended periods of timeand, even so, omit parts of the biotas. For an island of anysize and complexity, a complete census-type enumerationwould be a fantastic undertaking, particularly if the criterionof species presence were based on propagules rather thanon established populations. The whole approach smacks ofornithology and is reasonable enough for creatures that flyabout advertising their presence, especially if they are iden-tifiable by a body of volunteer watchers. However, enum-erating the entire biota would require a massive collectingjob to be undertaken by specialists. The entire biomasswould have to be screened, the forests felled, and the soilsifted for seeds, spores, and whatnot, with each tested forviability and cultured until mature enough to be identified.Each census would produce the suggested artificial Krakatoaand whatever was left of the biota could hardly be pursuedto the next census as an equilibrium system.

At present, I believe, biogeography would accomplishmore by using its concepts and tools than by redesigningthem. No matter how cleverly derived, our models willremain soft and amorphous until calibrated with real values.We need to work out enough solid cases of species patternsand the processes shaping them to get beyond vague banal-ities and isolated details. This is a grand enterprise in whichany number can join. Eventually, work will have to be pushedwith some taxonomically difficult organisms in some unpleas-ant habitats. At the moment there are still plenty of easilyrecognized species and many beautiful islands that no bio-

Page 84: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

78 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

geographer has claimed. There are good practical reasonsfor continuing to draw on islands for case studies, thoughthere is no longer any excuse for segregating them fromcontinents in biogeographical theory

Sauer's criticisms are no less valid today thanthey were in 1969.

The species-area relationship has recently beencriticized from another standpoint by Connorand Simberloff (1978), who discussed inadequa-cies in studies of species-area relationships ofGalapagos plants and birds. They noted (p. 219):"Generalizations about factors determining spe-cies numbers that are based on multiple regres-sion and correlation are precarious." Further,Connor and Simberloff (1978) stated that thevalues used for numbers of species may be faultyfor the following reasons, which are more or lessthe same as some of Sauer's criticisms: all speciesare treated equally, when in fact some are muchmore abundant or otherwise biologically impor-tant than others; criteria for residency are am-biguous (for example, is a single propaguleenough to be counted, or must there be a breed-ing population?); and different islands have notbeen collected as thoroughly as others, so thatspecies lists vary in their levels of completeness(see the example below on bats of St. Bartholo-mew). Connor and Simberloff justifiably criti-cized previous attempts to model the biogeogra-phy of the Galapagos, but then they went aheadand proposed their own, admittedly inadequate,alternative models.

Connor and McCoy (1979) analyzed the dataused in 100 species-area curves from all over theworld in reviewing the statistics and biology ofthe species-area relationship. Their conclusions(pp. 814, 815) should be sobering to many the-oretical biogeographers:

Our discussion of the theoretical basis of the species-arearelationship was basically inconclusive. The two most fre-quently proposed hypotheses, habitat diversity and area perse are both possibly correct, yet the results of either mecha-nism is neither qualitatively nor quantitatively different. Onevirtually always observes a positive correlation between spe-cies number and area, regardless of the mechanism. . . .

In general, we have found that published predictions andinterpretations concerning both the slope and intercept pa-

rameters [z and C] are not supported by the available evi-dence. Many other predictions and interpretations are eitherlogically untestable or require additional data for an ade-quate test. Because of these results, we are skeptical that anybiological significance can be attached to these parametersand recommend that they be viewed simply as fitted con-stants devoid of specific biological meanings.

I do not believe that the findings of Connorand McCoy (1979) were as inconclusive as theyclaimed. Rather, I believe that they corroboratedthe thesis of Sauer (1969) and Haas (1975),namely that biogeographers have yet to blendquantitative theory satisfactorily with empiricaldata.

To summarize thus far, the criticisms by Sauer(1969), Haas (1975), Connor and Simberloff(1978), and Connor and McCoy (1979) cast muchdoubt on the significance of modern quantifiedisland biogeography. The derivations of C and zare simply too vague and shallow to permit quan-titative comparisons of insular biotas that arebased on the calculated values of these constants,which cannot account for all of the factors thatactually do influence the number of species onislands. Common sense tells us that both areaand habitat diversity are related to the numberof species of organisms on islands, but neither issacred in itself.

That Sauer's suggestions have been ignored isparticularly unfortunate; instead of rigorous sur-veys of insular biotas, many biogeographers dolittle if any field work and thus are not familiarwith the methodology and validity of the datasets that they manipulate. Most workers still op-erate with the illusion that the systematics anddistribution of vertebrates are so well known thatone can go merely to check-lists and find defini-tive data to be analyzed. For the past 10 years,various field parties from the Smithsonian Insti-tution have been surveying the vertebrates of theWest Indies. We have visited over 40 islandsaltogether, and in nearly every case we haveadded species to the island's faunal list, whetheramphibians, reptiles, birds, or mammals. Muchof this information is still unpublished, but ourfindings clearly show that the distribution of

Page 85: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 79

vertebrates (considered to be the best-knowngroup of animals) in the West Indies (consideredto be one of the best-known groups of islands) isin fact too poorly known to attempt any detailed,quantitative analyses.

For example, Fleming (1982) studied severalaspects of the distribution of birds and bats inthe West Indies; his computations dependedupon knowing how many resident species occuron each island considered. There is such scatterin the points of his species-area graphs (Fleming'sfigs. 1, 2) that any line drawn through thesepoints has confidence intervals so large as to bemeaningless or very nearly so, sensu Haas (1975).One reason why these points are so scattered isthe inconsistency of the data-gathering process,exemplified as follows. In October 1982, ourfield party (R.I. Crombie, L.K. Gordon, G.K.Pregill, and I) collected three species of bats{Brachyphylla cavernarum, Artibeus jamaicensis,and Molossus molossus) on St. Bartholomew thathad not been recorded there previously, therebyraising the number of species of bats known fromthis island from one to four. If three nights (sixnet-nights) of suburban mist-netting can quad-ruple an island's chiropteran fauna, we shouldnot be surprised that the points in Fleming'sgraphs are so scattered. Nevertheless, typical ofecologists, Fleming laments that (1982:59) "allthat is known about them [the bats] is theirgeographic distribution." Only through muchmore field work can we piece together the dis-tribution and species-level systematics of WestIndian vertebrates, and the same could be saidfor just about any other group of islands.

P.R. Grant has recently stated (in Lewin,1983:1411): "There are so many factors influ-encing the morphology and distribution of or-ganisms that it has been very difficult to generategeneral theories of community structure." Thisstatement is certainly true, but should we beconcerned that functional generalizations are dif-ficult to formulate? As pointed out by Sauer(1982:62, 63), generalized models are seriouslyflawed because "they eliminate the environmen-tal variables that are the basic causes of such

[biogeographic] patterning." Whether islands orcontinents or somewhere in between, each re-gion of the world is biologically unique becauseits geography, geology, and climate, both pastand present, have interacted with plants andanimals. The potential complicating factors areinnumerable, so why should generalization beour main goal if each situation is unique? Com-parisons are usually informative, but our effortsshould not focus solely upon the search for sim-ilarities. Instead we should evaluate each set ofdata on its own merit, in realization that we willnever solve all of the mysteries, nor will gener-alized models explain specific situations.

For example, we should not be surprised thatDarwin's finches "fail to conform" to species-arearelationships as predicted by the MacArthur andWilson model. Yet Juvik and Austring (1979)were concerned and puzzled by this "failure."They claimed that the earliest reliable distribu-tional data on birds of the Galapagos were thosecompiled by the California Academy of SciencesExpedition in 1905-1906, before which timeJuvik and Austring regarded the extent of man'sinfluence on avian distribution in the Galapagosto be (p. 212) "only a matter of speculation." Byfinding and studying fossils, I have shown that itis possible to reconstruct the avifauna of theGalapagos well before 1905, and to reveal thenature of the fauna before human impact. Morethan previously realized, we can determine theextent of man's influence on insular life.

Most biogeographical studies, from Mac-Arthur and Wilson up to the present, have ig-nored the prehistoric and historic factors thathave influenced insular biotas. To some degree,man has had an impact on the biota of nearlyevery island in the world before the initial collec-tion of data that could be used in quantifiedinsular biogeography. We must attempt to deter-mine the severity of these influences if our goalis a realistic comparison of truly natural biotas.Written documentation exists in varying levelsof accuracy and completeness for man's activitiesof the past several hundred years on many is-lands, and these writings should be an important

Page 86: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

80 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

source of information for the biogeographer. Forinformation that precedes written records, wemust turn to the fossil record, which is the onlyway to determine which species occurred at agiven locality at any time that predates writtenaccounts. ("Fossil" herein refers to bones foundin either a paleontological or archaeological con-text.)

Admittedly, most biogeographers prefer toanalyze today's floras and faunas, not those ofthe past. But such studies are meaningless if thefauna of a relatively undisturbed island is com-pared to that of another island that has beengreatly altered by man. The modern vegetationmay not reveal the history of habitat disturbancesand associated extinctions, for habitat destruc-tion that occurred hundreds or thousands ofyears ago can be masked completely by succes-sional recovery of the vegetation. Nor does thepresent number of human inhabitants necessarilyreflect past numbers. (For example, severalhundred people lived on Floreana in the 1830's,compared to approximately 50 today. Hender-son Island, in the Pitcairn Group, has been un-inhabited in historic times and is often cited asone of the few examples of a "pristine" island.However, Fosberg et al. (1983) reported Polyne-sian archaeological sites from Henderson, inwhich occur four species of birds that no longerlive on this island (Steadman and Olson, 1985).Even Henderson Island, therefore, can no longerbe regarded as "pristine.") In his critique of theMacArthur and Wilson equilibrium theory, Wil-liamson (1981:83) also noted that this theorydoes not consider historical factors. But William-son then stated that historical phenomena arefrequently unimportant in studying modern dis-tributions. I believe that any biogeographicalanalyses, whether on continents or islands, areunreliable if they are not backed by historicalinformation.

Although the arguments in this paper are ap-plicable in concept to any group of organisms,they pertain in practice only to those groups forwhich a significant fossil record is obtainable,such as vertebrates, mollusks, and certain groups

of insects. This discussion will deal mainly withterrestrial avifaunas for several reasons.

1. The number of species of birds on an islandmay be high enough to have at least some math-ematical potential for quantitative manipulation.These analyses would be less well suited for in-sular mammals, for example, because of theirlower species diversity.

2. Birds are an easily accessible group tostudy, and living insular avifaunas thus are oftenrelatively well known compared to other groupsof organisms (as Sauer stated, 1969:591: "Thewhole approach smacks of ornithology. . .").Nevertheless, as noted previously, more thor-ough distributional data are needed even forbirds.

3. Deposits of vertebrate fossils on islands usu-ally include many birds.

In the few instances where the fossil record ofa particular island is well documented, one canreconstruct that island's past avifauna in a man-ner that at least approaches the natural situation.The best-documented is in Hawaii, where Olsonand James (1982a, 1982b, 1984) have shown thatlate Quaternary extinction of birds occurred ata much greater rate than ever before suspected.They have collected rich deposits of avian fossilsfrom most of the major Hawaiian islands. Thesefossils document for the first time the occurrenceof many new species, as well as many new recordsof species otherwise known only from other is-lands in the archipelago. Altogether, fossils havemore than doubled the number of previouslyknown species of terrestrial Hawaiian birds(1982a, table 1; 1982b, table 5), and new speciesare still being discovered. Olson and James' fossilrecord includes many species that apparently be-came extinct through hunting and habitat de-struction of the past 1000 years by Polynesians.It is still uncertain how much of this extinctionoccurred in prehistoric times (i.e., before A.D.1780) versus early historic times (ca. A.D. 1780to 1880), because few specimens were collectedduring the first hundred years after Europeancontact. Regardless, the wave of extinctionamong Hawaiian birds during the past century

Page 87: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 81

was not a unique event within or outside ofHawaii. Elsewhere in Polynesia, levels of extinc-tion in birds as great or greater than those inHawaii are now being discovered for the CookIslands (Steadman, 1985) and the Marquesas(Steadman, personal observation of previouslyunstudied archaeological bones). The prehistoricrecord is even more important than the historicrecord in determining the natural avifauna ofPolynesia. Biogeographic studies of the Polyne-sian avifauna that do not consider the fossil rec-ord are not only incomplete, they are downrightmisleading.

Turning to Floreana, we see that late Holo-cene fossils from the Post Office Bay region haveconfirmed that Tyto punctatissima, Mimus trifas-ciatus, Geospiza nebulosa, and G. magnirostris,each of which is now extinct on Floreana, wereonce resident there. These findings should notbe unexpected, for Floreana has suffered asmuch or more human-related biological disturb-ance than any other island in the Galapagos (see"Human History," and "Introduced Mammals").The fossil record of Floreana differs from thatof Polynesia, however, in two major ways—thechronology and the extent of the extinction. Theextinctions on Floreana occurred later in timethan those in Polynesia. So far, no prehistoricextinctions are surely known from Floreana,whereas numerous prehistoric (= before Euro-pean contact) extinctions occurred in Polynesia.As noted above, many extinct birds are associatedwith Polynesian cultural sites, and therefore musthave became extinct within the last 1000 years.Thus the chronology of extinction in Polynesia,as in the Galapagos, is related to the arrival ofman. The Galapagos lacked pre-European inhab-itants, and we find little if any prehistoric extinc-tion there. Note that this discussion is confinedto the Holocene, thereby eliminating any effectsof climatic change near the Pleistocene-Holoceneboundary.

Aside from the chronological difference, theextent of extinction was much greater in Poly-nesia than in the Galapagos. Several factors maybe involved here. One is the much greater hu-

man population of Polynesia both in prehistoricand historic times. The climate and soil in mostof Polynesia are generally much better suited forhuman occupancy than that of the Galapagos. Itfollows that higher human populations will tendto have a greater impact on an area than smallerpopulations. For example, essentially all of theremaining native forest in Hawaii is in humid,mountainous regions. The Polynesians had de-stroyed essentially all of the lowland forest beforethe arrival of Europeans in the late 1700s, andin doing so they wiped out numerous species ofbirds. The human population of the Galapagosmeasured only in the hundreds from its begin-ning in 1832 until the 1940s. Today approxi-mately 7000 persons live in the Galapagos, morethan ever before, but nevertheless a relativelylow figure. Many of the islands in the Galapagosretain pristine or nearly pristine lowland habitats.Another factor is how long the islands have beeninhabited. Human residency in the Galapagosspans only 150 years, although these islands havebeen visited sporadically for over 400 years,whereas most of Polynesia, including Hawaii,Marquesas, and Cook Islands, has supported peo-ple for at least 1000 years. Lastly, certain Po-lynesian birds were more vulnerable to extinc-tion simply because of what they were. It is easyto imagine flightless geese, ibises, and rails beinghunted to extinction by Polynesians. The onlyflightless birds in the Galapagos are marine spe-cies—Spneniscus mendiculus (Galapagos Penguin)and Phalacrocorax harrisi (Flightless Cormorant),both of which are confined to the shorelines ofFernandina and Isabela in places that are tooinhospitable to be colonized by humans. Never-theless, feral dogs are now a threat to Galapagospenguins, at least on Isabela (Barnett, 1982).

The Galapagos fauna has suffered extinctionwithout replacement; human impact seems tohave wiped out more species than have beenreplaced by natural colonization. This samephenomenom has occurred in the West Indies(Steadman et al., 1984), as well as in Polynesiaand other oceanic islands. As noted by Johnson(1983) for the Channel Islands, the concept of

Page 88: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

82 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

natural turnover (if such exists) is not applicableto islands that have been significantly disturbedby man. Olson and James (1982b:53) point outthat: "Data [on numbers of species] from islandsthat have suffered from human disturbance,which would include most of the islands in theworld, should not be relied upon for numericalstudies of island biogeography unless the impactof such disturbance is known." I would add thatthe fauna of the Galapagos is nowhere near"equilibrium" (if such exists), regardless of theextent of historic extinctions. The Galapagos lackspecies from most families of Neotropical landbirds. For example, I see no reason why theGalapagos could not support indigenous popu-lations of falcons, goatsuckers, woodpeckers, co-tingas, wrens, or tanagers, just to mention a fewpossibilities. Cuckoos and warblers were prob-ably unrepresented in the Galapagos until onlyhundreds or at most thousands of years ago (seebelow). Thus faunal diversity in Galapagos birdslargely depends upon availability of colonists.These islands are large enough and diverseenough environmentally to support many morespecies than exist now.

Another problem in analyzing numbers of spe-cies of land birds on islands is deciding whichresident birds should be included. For instance,should one include aquatic birds such as ducks(Anas bahamensis), gallinules (Gallinula chloro-pus), or shorebirds (Himantopus mexicanus,Phoenicopterus ruberj? Are certain rails (Laterallusjamaicensis, Neocrex erythrops) non-aquaticenough to be considered among the land birds?Is Geospiza pallida, known from only a singlehistoric Floreana specimen, to be included? ForFloreana, Harris (1973) listed 16 species as"breeding now" and 22 species as "ever re-corded." The two rails were not included inHarris' numbers, but I have included them inmy own analysis of the land birds of Floreana(Table 12) because these rails occur mainly inthe moist highlands and because they are preyedupon by barn owls (Groot, 1983; personal obser-vation).

Concerning Darwin's finches, B. R. Grant and

TABLE 12.—Status of resident land birds of Floreana (23species either breed now, have bred in the past, or may havebred in the past, on Floreana).

Species

Buteo galapagoensisLaterallus jamaicensisNeocrex erythropsZenaida galapagoensisCoccyzus melacoryphusTyto punctatissimaAsio flammeusPyrocephalus nanusMyiarchus magnirostrisProgne modestaMimus trifasciatusDendroica petechiaGeospiza nebulosaG. fuliginosaG. fortisG. magnirostrisG. scandensG. crassirostrisG. parvulaG. pauperG. psittaculaG. pallidaG. olivaceaTotal

Occursas afossil

X

X

XX

X

XXXXXXXX

X14

Extinct onFloreana

X?

X

X

X

X

X?X?

4 to 7

Recentcolonizer

ofGalapagos

X

X

X?

X

3 or 4

P.R. Grant stated (1982:654):

Extinctions are not likely to be observed, but they can beinferred from extrapolations of observed fluctuations. Tounderstand the absence of species on islands and the ecolog-ical and morphological differences between those present,there is no real substitute for direct studies of food availa-bility and use by coexisting species over a long period oftime.

I do not deny the importance of direct foodstudies; indeed such studies, as well as studies ofnesting habits, have helped in suggesting causesof extinction for Geospiza nebulosa and G. mag-nirostris on Floreana. Reciprocally, ecologistsneed the fossil record to document the chronol-ogy and extent of the extinctions in the firstplace.

Hamilton and Rubinoff (1963, 1964, 1967)

Page 89: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 83

studied factors (area, isolation, plant diversity)that are potentially related to number of speciesand degree of endemism in Darwin's finches.They cited Lack (1947) and Bowman (1961) inrecognizing 9 species of Darwin's finches to benative to Floreana. But Lack (1947, table XVII)listed only 8 resident species, noting that if G.difficilis (= nebulosa) and G. magnirostris werefound to have been once resident on Floreana,this figure would increase to 10. Bowman (1961,table 7) recognized 9 species, omitting G. nebu-losa because of apparent uncertainties in its lo-cality of collection (discussed in Lack, 1945:14,15). (Contrary to Hamilton and Rubinoff (1963),Bowman (1961) did not follow the taxonomicconclusions of Swarth (1931), who recognized 28species of Darwin's finches in the Galapagos ascompared to Bowman's 13 species.) Hamiltonand Rubinoff (1963) produced an equation topredict the number of Darwin's finches on anyisland in the Galapagos. For Floreana, this equa-tion predicted 6.2 species, far short of the 10 or11 species now known to have ocurred there(Sulloway 1982a,b; this study). Harris (1973:274)listed 8 species of Darwin's finches as currentlyresident on Floreana, with G. nebulosa regardedas "probably once resident, now not present" G.magnirostris regarded as "has been recorded,probably stragglers." Power (1975), Connor andSimberloff (1978), and Alatalo (1982) used thedata of Harris (1973) for their quantitative anal-yses of the Galapagos avifauna. Unavoidably,these data were compiled in the absence of pa-leontological information, so they do not repre-sent the undisturbed fauna of the Galapagos.

To summarize, biogeographical equationslook good only when they produce results thatare consistent either with previously known em-pirical data or with the researcher's preconcep-tions. When these circular equations fail to "pre-dict" the desired numbers, the unfortunate re-sults are written off as being due to such thingsas statistical errors, inadequate data, or "un-known factors." In effect, these reasons onlyrepresent an admission of our ignorance of bio-logical phenomena. Even on islands, biogeogra-

phy is too complex to be explained by a fewmagic numbers.

After 20 years of emphasis on theoretical re-search in the field of island biogeography, theemerging paleontological record may spark areturn to empiricism. Fossils permit us to deter-mine the natural fauna of Floreana and otherislands with more confidence than ever before.The fossil record from Floreana seems to be tooyoung to reveal evolutionary changes; instead,its forte is in revealing extinctions and past rela-tive abundances. As fossil records emerge fromother islands in the Galapagos, we will be able tomake more meaningful inter-island faunal com-parisons.

ANIMALS NOT RECORDED AS FOSSILS.—Notincluded among the fossils are certain speciesthat either still live on Floreana, once lived there,or may be suspected to have once occurred there.In order to interpret a fossil fauna as thoroughlyas possible, one must attempt to account forabsences in the record by answering some basicquestions. Is a certain species unrepresented inthe fossil fauna because it simply did not occurthere at the time of deposition of the fossils? Ifnot, are there any behavioral or ecological traitsof the species that may help to explain its ab-sence?

Cricetine rodents and land iguanas (Conolo-phus) fall into the first category. They are notknown historically from Floreana, and I regardtheir absence as fossils as strong evidence thatthey in fact never occurred there. Various ryto-derived fossil sites on San Cristobal, Santa Cruz,Rabida, and Isabela are dominated by indigenousrodents; had rodents been present on Floreana,the barn owls would have preyed upon them anddeposited their bones in the caves. Land iguanasoccur as fossils regularly but in low numbers onSanta Cruz, Rabida, and Isabela. Young landiguanas are preyed upon and deposited in cavesby barn owls, whereas adults are trapped withinthe caves by falling through vertical roof col-lapses. One of these methods probably wouldhave entombed land iguanas in the Floreanacaves, had they been present. Porter (1822:163)

Page 90: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

84 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

noted that land iguanas were not found on Flo-reana in 1813, a time when tortoises were stillabundant and before the takeover by feral mam-mals.

Nine of the 23 species of birds in Table 12were not recorded as fossils. Of these, the ab-sence of Buteo galapagoensis (Galapagos Hawk)and Asio flammeus (Short-eared Owl) may be ex-plained by their being too large to be eaten bybarn owls, as well as the fact that they do not usecaves for any of their own activities. The sevenother species of birds are small enough to beeaten by barn owls, so different reasons must besought for their absence. Four of these species,Laterallus jamaicensis (Black Rail), Neocrexerythrops (Paint-billed Crake), Geospiza psittacula(Large Tree Finch), and G. pallida (WoodpeckerFinch) occur mainly in the humid highlands ofFloreana, outside of the hunting territory of barnowls that roosted near Post Office Bay. In addi-tion, N. erythrops is undoubtedly a very recentcolonizer of the Galapagos, for these reasons: itis totally undifferentiated from mainland forms;it was not recorded in the Galapagos until 1952;and it does not occur in the huge fossil samplefrom Cueva de Kubler, Santa Cruz. The faunaof Cueva de Kubler does include, although insmall numbers, fossils of L. jamaicensis, G. psitta-cula, and G. pallida. The presence of these spe-cies in Cueva de Kubler but not in the fossilfaunas of Floreana may only reflect the muchlarger number of fossils from Cueva de Kublercombined with its location farther inland and ata slightly higher elevation than the Floreana sites,although Cueva de Kubler is still in arid zonevegetation.

The situation for G. pallida is more compli-cated, for this finch is known on Floreana fromonly a single specimen, collected in the moisthighlands on 11 October 1905 (Gifford,1919:253, 254; Swarth, 1931:243). Thus thepossibility exists that G. pallida never was resi-dent on Floreana. Harris (1982:152) noted with-out detail a "few sight records" of G. pallida fromFloreana, presumably within the past 20 years.Specimens are essential to document any new or

unusual island records of Darwin's finches; sightrecords are not reliable. Although G. pallida isapparently very rare or extinct today on Flo-reana, it probably once occurred there as a resi-dent, only to be victimized somehow by the large-scale biotic changes that have occurred in theFloreana highlands during the past 150 years.Geospiza pallida is still common today in thehighlands of Santa Cruz and Isabela, but thehabitat destruction on these two larger islandshas not been as complete as on Floreana. Lack(1969:263) attributed the absence of G. pallidaon Floreana to "unknown reasons (but just con-ceivably linked with the co-existence there, andnowhere else, of C. psittacula and C. pauper)."The natural status of G. pallida on Floreanaremains a mystery.

Progne modesta (Galapagos Martin) is a rarebird on Floreana, although it does occur in thelowlands. This martin has not been recordedfrom any modern or fossil barn owl roost in theGalapagos. Progne modesta roosts mainly alongcliffs near the sea, places that make it inaccessibleto barn owls.

The absence of the two remaining birds, Coc-cyzus melacoryphus (Dark-billed Cuckoo) and Den-droica petechia (Yellow Warbler) cannot be ex-plained by biases of habitat or predation. Thesetwo species are common today in the lowlands ofFloreana. In fact, the Yellow Warbler may be themost common lowland bird. Furthermore, boththe cuckoo and warbler occur regularly in mod-ern pellets of barn owls from Santa Cruz andIsabela. In Cueva de Kubler, Santa Cruz, forexample, C. melacoryphus and D. petechia arepresent in fresh barn owl pellets, but are notamong the thousands of avian fossils studied thusfar, which were collected only 50 m away. Thelack of fossil remains of cuckoos and warblersfrom both Floreana and Santa Cruz suggests thatthese two currently common species colonizedthe Galapagos only very recently, perhaps evenin the last several hundred years. The fossil siteson Floreana range in age from modern to at least2400 years B.P., whereas Cueva de Kubler hasyielded radiocarbon ages ranging from modern

Page 91: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 85

to 1750 years B.P. The ages from Cueva deKubler are unreliable, however, because of ap-parent contamination from surface testing of nu-clear weapons in the Pacific (Robert Stuckenrath,pers. comm.). Thus the actual age of the fossilsin Cueva de Kubler is at least 1750 years B.P. andmay be much older. For now, I can say only thatthe conspicuous absence of C. melacoryphus andD. petechia from all fossil sites on Floreana andSanta Cruz suggests that these species were notpresent during the time of fossil deposition.

Although both seem to be relatively recentcolonizers, the cuckoo is likely to have been morerecent than the warbler, based on date of firstcollection of historic specimens and on the levelof morphological differentiation of the Galapa-gos populations. Coccyzus melacoryphus may nothave colonized the Galapagos until the 19th cen-tury, for it was not collected there until 1888,on Floreana and San Cristobal (Ridgway, 1890).Rothschild and Hartert (1902:404) described itas "somewhat rare and apparently a recent im-migrant." In 1905-1906, the California Acad-emy of Sciences Expedition recorded cuckoosfrom six different islands (55 specimens), sug-gesting a rapid increase in numbers and distri-bution, typical of geometrically expanding pop-ulations of a newly arrived species. Specimens ofC. melacoryphus from the Galapagos do not differsignificantly from those of mainland SouthAmerica (Swarth, 1931:71, 72; personal obser-vation), inferring that there has been very littletime available for morphological changes to de-velop between the two populations.

Dendroica petechia was first recorded from Flo-reana in 1852 (Sundevall, 1871), although itmust have been there in 1835, for Gould (1841)noted that warblers were common throughoutthe archipelago during the voyage of the Beagle.Yellow Warblers of the Galapagos and CocosIsland are essentially identical to each other andare traditionally regarded as forming their ownsubspecies, D. p. aureola. Salvin (1883:420) ap-parently regarded the birds from the Galapagosas identical to those of the mainland, for herecorded "Dendroeca aureola" from coastal Ec-

uador and Peru, as well as the Galapagos. Hell-mayr (1935:383) and Harris (1982:129) notedthat D. p. aureola is very similar to D. p. peruvianaof the Pacific coast of Colombia, Ecuador, andPeru. I have compared specimens of D. p. peru-viana with D. p. aureola from both Cocos Islandand the Galapagos, finding the two insular pop-ulations to be indistinguishable from each otherand slightly distinct from, but very similar to, thebirds from coastal Peru. Dendroica p. aureolaseems to be a barely recognizable race that isderived from D. p. peruviana. Additional fossilsites from other islands in the Galapagos shouldeither strengthen or discredit my suggestion ofrecent colonization by cuckoos and yellow war-blers.

Some historic colonization in the Galapagosmay have been aided by ships travelling from themainland, as well as within the islands. Harris(1974, 1982:120) has suggested that the severalrecent records of anis (Crotophaga sulcirostris andC. ani) in the Galapagos are possibly due tohuman agency. Other species may also be gettingfree rides to the Galapagos.

THE PALEONTOLOGICAL POTENTIAL OF THEGALAPAGOS.—The opportunity to elucidate an-cient and therefore undisturbed insular faunasdepends upon locating suitable accumulations offossils. The Galapagos lack calcareous dunes(aeolianites), which have yielded Quaternaryfaunas on Hawaii, St. Helena, and Fernando deNoronha (Olson, 1975, 1981; Olson and James,1982b). They also lack muck deposits, known tobe very rich in bird bones in parts of New Zea-land (Gregg, 1972; McCulloch and Trotter,1979), as well as phosphate deposits and fine-grained alluvium, which can be fossiliferous onAscension and St. Helena (Olson, 1975, 1977).Limestone caves, crevices, and sinkholes oftencontain fossils elsewhere, e.g., Hawaii, Cook Is-lands, Bermuda, and the West Indies, but theseweathered karst features are also absent in theGalapagos. To date, no fossil sites are knownfrom fumaroles in the Galapagos, as have beenfound on Ascension Island (Olson, 1977). Fine-grained volcanic ashes, such as those containing

Page 92: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

86 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

bird skeletons on Isla San Benedicto, Revillagi-gedos Islands, Mexico (Brattstrom and Howell,1956), are rare in the Galapagos and have notbeen explored paleontologically. Peat bogs occurin the highlands of the larger Galapagos islands,but they have not been excavated for fossils.

Deposits in lava tubes are the one outstandingopportunity for bone preservation in the Gala-pagos, although several vertebrate fossil siteshere are in large fissures or spatter cones. Lavatubes are present in good numbers on San Cris-tobal, Santa Cruz, Santiago, Isabela, and Fernan-dina, as well as Floreana. In future attempts todescribe the natural fauna of the Galapagos,these islands will provide rich opportunities forprospecting. The first four islands either are orhave been inhabited by man, and thus have theirfair share of introduced plants and animals. Aswas the case on Floreana, a fossil record fromthese islands would be expected to yield remainsof animals that have become extinct within thepast 200 years. For example, fossils from SanCristobal could provide information on the pres-ently uncertain status of Geospiza nebulosa, G.magnirostris, and G. psittacula on that island. OnSanta Cruz, I have excavated several sites (brieflysummarized in Steadman, 1981), a preliminaryanalysis of which reveals three species of extinctrodents and all 10 species of Darwin's finchesthat are known historically from Santa Cruz, butin different relative abundances than exist today.On Pinzon, one should keep an eye out for fossilsof rodents, mockingbirds, and Geospiza scandens.Santiago is yet unexplored paleontologically, buthas much potential for important discoveries,especially of rodents. Isabela, for which there isa limited fossil record, already has yielded newrodents (Steadman, 1981). Fernandina has neverbeen inhabited by humans, and does not harborany introduced mammals. The high levels ofhistoric volcanic activity (Simkin and Howard,1970) may be why the status of many vertebrateson Fernandina is poorly known. Natural extinc-tions are entirely possible. Fossils from Fernan-dina could shed light on the former status oftortoises, rodents, Geospiza magnirostris, G. scan-dens, and G. pallida, among others.

The vertebrate fossil record of the Galapagoshas not gone with certainty beyond the Holo-cene. Older lava flows should be checked care-fully for lava tubes, for Pleistocene fossils may bethe key to many intriguing evolutionary ques-tions. Within the next decade we can expect theemergence of a much more complete picture ofthe natural vertebrate fauna of the Galapagos.

Summary and Conclusions

Isla Floreana (Charles Island) is a relativelylarge, diverse island in the south-central part ofthe Galapagos Archipelago. I have collected andidentified over 20,000 late Holocene vertebratefossils from Floreana, representing more than1100 individual animals of 24 native species. Thisstudy is the first faunal survey of fossil verte-brates from the Galapagos Islands. The fossilswere collected from both surface and sub-surfacelevels of sediments from four lava tubes in thearid Post Office Bay region. None of the fossilsites is more than 1 km from the ocean, or morethan 60 m in elevation. The lava tubes are inflows of Brunhes normal magnetic polarity, andtherefore less than 0.79 million years old. Theoldest of six radiocarbon age determinationsfrom surface remains in three of the caves was2400 years B.P. Although fossils in the underly-ing sediment must be older than the surfacematerial, the sub-surface levels yielded no or-ganic material suitable for radiocarbon dating,and thus their age is not known precisely. Thereis no reason to believe, however, that any of thefossils are pre-Holocene in age.

Except for non-hatchling tortoises, nearly allof the fossils probably were deposited in the cavesas pellets regurgitated by the Galapagos BarnOwl (Tyto punctatissima). A detailed analysis ofthe fossil fauna elucidates the past feeding habitsof barn owls on Floreana. Because Floreanalacked native rodents, the barn owls there fedmainly on reptiles and small birds. The fossilfauna includes six indigenous species of verte-brates that are now extinct on Floreana: theGalapagos tortoise (Geochelone elephantopus), Flo-reana snake (Alsophis biserialis), Galapagos Barn

Page 93: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 87

Owl (Tyto punctatissima), Floreana Mockingbird(Mimus trifasciatus), Sharp-beaked Ground Finch(Geospiza nebulosa), and Large Ground Finch(Geospiza magnirostris). Based on calculated min-imum numbers of individuals, these extinct spe-cies are, respectively, 1st, 7th, 16th, 6th 15th,and 2nd in abundance among the 24 speciesrecorded as fossils. Together the extinct speciesmake up 57.2% of the individuals in the entirefossil assemblage. Furthermore, the 3rd and 4thmost numerous fossil taxa, the lava lizard (Tro-pidurus grayii) and Galapagos Dove (Zenaida gal-apagoensis), are exceedingly rare on Floreanatoday. Of the 7 most common fossil taxa, only asmall gecko (Phyllodactylus baurii) survives inwhat may approach natural numbers. If percentof occurrence in the fossil record approximatesrelative population size, then historic extinctionhas changed the composition of Floreana's ver-tebrate fauna even more than might be suggestedby the number of extinct species alone. The fossilrecord shows that the fauna we see today onFloreana is drastically different from that of pris-tine times.

While details of the process are unknown andthe evidence is largely circumstantial, I believethat all extinction on Floreana may be related tohuman impact, including direct predation, habi-tat alteration, and the introduction of alien mam-mals. Floreana's first large human settlementbegan in 1832, accompanied immediately or veryshortly thereafter by alien mammals such as blackrats, house mice, cats, dogs, pigs, goats, cattle,horses, and donkeys. Direct human predationmay have been involved in all extinctions to someextent, but it probably was the main cause onlyfor the tortoise and the Galapagos Hawk (Buteogalapagoensis), the latter not recorded as a fossil.Predation by feral mammals, particularly rats,cats, dogs, and pigs, may also have been involvedto some extent in all extinctions except perhapsthe hawk. Extinction of the Galapagos Barn Owlwas probably due to predation in combinationwith reduction or loss of preferred prey species,such as snakes, lava lizards, doves, mockingbirds,and certain finches, especially the Large GroundFinch. Extinction on Floreana of the Floreana

Mockingbird and Large Ground Finch may haveaccompanied the depletion of prickly pear cactus(Opuntia megasperma var. megasperma) in the low-lands by feral goats and donkeys. Both the mock-ingbird and finch seem to have been dependentupon Opuntia for nesting and feeding, so thesevere reduction in cactus was devastating tothese specialized forms. The loss of the Sharp-beaked Ground Finch on Floreana may havebeen due to habitat change in the highlands,accompanied by a sudden burst of competitionfrom other finches that moved into the highlandsas man altered the vegetation. I believe that allextinction on Floreana occurred in historic times;whether or not this holds for other islands in theGalapagos awaits more digging for fossils.

Because of a paucity or lack of specimens withunequivocal locality data, scientists have ques-tioned whether or not most of the extinct speciesdiscussed actually once occurred on Floreana.The fossils demonstrate that they did. The ab-sence of fossils of the Dark-billed Cuckoo (Coc-cyzus melacoryphus) and the Yellow Warbler (Den-droica petechia) is evidence that these species arevery recent colonizers of the Galapagos. Thecuckoo has not differentiated perceptibly fromits mainland ancestors, and the warbler is onlyslightly differentiated from populations fromcoastal Ecuador and Peru, so the absence of thesetwo species from fossil deposits even as young aslate Holocene is not unexpected.

Fossils also allow us to compare the skeletalmorphology of populations on Floreana to thosefrom other islands, as well as from the Neotrop-ical mainland. Based on plumage and osteology,I regard Tyto punctatissima, Pyrocephalus nanus(Galapagos Vermilion Flycatcher), P. dubius (SanCristobal Vermilion Flycatcher), and Mimus tri-fasciatus as full species, the latter restricted toFloreana (formerly) and the nearby islets ofChampion and Gardner-near-Floreana. Geospizanebulosa nebulosa, a large race of Sharp-beakedGround Finch, is recognized for Floreana only;it may have occurred also on San Cristobal, butproper documentation is lacking. I regard Geo-spiza magnirostris magnirostris as having occurredonly on Floreana and San Cristobal. This largest

Page 94: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

88 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

of all Darwin's finches was probably the mostcommon bird in the lowlands of Floreana beforehuman contact. The Red Bat, Lasiurus borealis,was the only indigenous mammal recorded fromthe fossil sites. Floreana apparently lacked nativerodents, for had they been present, the barn owlwould have preyed upon them and depositedtheir bones in the caves, as is the case elsewherein the Galapagos, e.g., San Cristobal, Santa Cruz,Rabida and Isabela. The absence of fossils ofland iguanas (Conolophus) is evidence that thisendemic Galapagos lizard also never occurred onFloreana.

The fossils from Floreana allow us to recon-struct the natural, unperturbed (= pre-human)vertebrate fauna of this island much more com-pletely than previously possible. This exposes amajor limitation in the popular "equilibrium the-ory of island biogeography," namely that quan-titative inter-island comparisons of faunas usuallyare made with no historical basis. Without a fossilrecord, one cannot determine with confidencethe extent of man-related extinction on an island.As a result, biogeographers generally have littleidea how natural their faunas are. Other theo-retical and practical shortcomings show thatquantitative biogeographical models oversim-plify natural phenomena to the point where littleor no meaningful information is generated.MacArthur and Wilson have initiated a great

amount of interest in islands, but much of thisinterest has been channeled into the testing ofgeneralized models with poorly controlled datarather than the gathering of new data. By puttingthe theoretical cart before the empirical horse,biogeographers of the past two decades seem tobe overlooking the complexities of insular biotas.Such concepts as turnover and biotic equilibriummay eventually be demonstrated or refutedthrough long-term field studies, whether paleon-tological or neontological. A sufficiently rich andwell-dated Holocene fossil record, for example,might disclose natural turnover. To date, noneis evident, and instead we find unnatural extinc-tion without replacement.

Availability of fossil sites limits the extent towhich paleontology can reveal the natural faunaof the Galapagos or any other islands. The Ga-lapagos lack the aeolianites, fine-grained allu-vium, calcareous mucks, phosphate deposits, andthe limestone caves, crevices, and sinkholes thathave produced rich fossil sites on other oceanicislands. Lava tubes are the primary source ofvertebrate fossils in the Galapagos, althoughother volcanic features may serve as natural trapsor as owl roosts, such as earthquake crevices andsteep-walled spatter cones. Expansion of the Gal-apagos fossil record to other islands may be lim-ited somewhat by the occurrence of lava tubes.

Page 95: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

Literature Cited

Abbott, Ian, and L.K. Abbott1978. Multivariate Study of Morphological Variation in

Galapagos and Ecuadorean Mockingbirds. TheCondor, 80(3):302-308, 3 figures, 2 tables.

Abbott, Ian, L.K. Abbott, and Peter R. Grant1977. Comparative Ecology of Galapagos Ground

Finches (Geospiza Gould): Evaluation of the Im-portance of Floristic Diversity and InterspecificCompetition. Ecological Monographs, 47(2): 151-184, 20 figures, 14 tables.

Abs, Michael, Eberhard Curio, Peter Kramer, and JochenNiethammer

1965. Zur Ernahrungsweise der Eulen auf Galapagos:Ergebnisse der deutschen Galapagos-Expedition1962/63, IX. Journal fur Ornithologie, 106(l):49-57, 2 tables.

Agassi/., Alexander1892. Reports on the Dredging Operations off the West

Coast of Central America to the Galapagos, to theWest Coast of Mexico, and in the Gulf of Califor-nia, in Charge of Alexander Agassiz, Carried onby the U.S. Fish Commission Steamer "Albatross,"Lieut. Commander Z.L. Tanner, U.S.N., com-manding, II: General Sketch of the Expedition ofthe "Albatross," from February to May, 1891.Bulletin of the Museum of Comparative Zoology atHarvard College, 23(1): 1-89, 22 plates.

Alatalo, Rauno V.1982. Bird Species Distributions in the Galapagos and

Other Archipelagoes: Competition or Chance?Ecology, 63(4):881-887, 2 figures, 2 tables.

Allen, Joel Aseph1892. On a Small Collection of Mammals from the Ga-

lapagos Islands, Collected by Dr. G. Baur. Bulletinof the American Museum of Natural History, 4:47-50.

Alpert, Leo1963. The Climate of the Galapagos Islands. Occasional

Papers of the California Academy of Sciences, 44:21-44, 11 figures, 8 tables, appendix.

Ashmole, N.P.1963. Sub-fossil Bird Remains on Ascension Island. The

Ibis, 103b(3):382-389, plate 10.Atkinson, I.A.E.

1977. A Reassessment of Factors, Particularly Rattusrattus L., That Influenced the Decline of EndemicBirds in the Hawaiian Islands. Pacific Science,31(2): 109-133, 1 figure, 1 table.

Banning, George Hugh1933. Hancock Expedition to the Galapagos Islands,

1933: General Report. Bulletin of the ZoologicalSociety of San Diego, 10:1-31, 25 figures.

Barber, Richard T., and Francisco P. Chavez1983. Biological Consequences of El Nino. Science,

222(4629): 1203-1210, 7 figures, 1 table.Barlow, Nora, editor

1963. Darwin's Ornithological Notes [with introduction,notes, and appendix]. Bulletin of the British Museum{Natural History), Historical Series, 2(7):201-278.

Barnett, Bruce D.1982. Feral Dogs of the Galapagos Islands. Informe An-

ual 1981 Estacion Cientifica Charles Darwin, pages205-216, 3 figures, 1 table.

Basilewsky, Pierre1968. Les Calosomes de iles Galapagos (Coleoptera, Car-

abidae). In Mission Zoologique Beige aux iles Gala-pagos et en Ecuador (N. etj. Leleup, 1964-1965);Resultats Sdentifiques, pages 179-207. Bruxelles:Institut Royal des Sciences Naturelles de Belgique.

Baskin, Jonathan N., and Ernest E. Williams1966. The Lesser Antillean Amewa (Sauria, Teiidae):

Re-Evaluation, Zoogeography, and the Effects ofPredation. Studies on the Fauna of Curacao andOther Caribbean Islands, 23(89): 144-176, figures42, 43, tables 12-20, 2 plates.

Baur, George1889. The Gigantic Land Tortoises of the Galapagos

Islands. The American Naturalist, 1889:1039-1057.

1892. Das Variiren der Eidechsen-gattung Tropidurusauf den Galapagos-Inseln. In Festschrift turn Sieben-zigsten Geburtstage Rudolf Leuckarts, pages 259-277, 3 figures. Leipzig.

1895. The Differentiation of Species on the GalapagosIslands and the Origin of the Group. In BiologicalLectures Delivered at the Marine Biological Labora-tory of Wood's Holl [sic] in the Summer Session of1894, pages 67-78, 3 unnumbered tables.

Beebe, William1924. Galapagos: World's End. 443 pages, frontispiece,

82 figures, 8 color plates, 2 appendices. New Yorkand London: G.P. Putnam's Sons.

Bell, Thomas1843. Reptiles. In Charles Darwin, editor, The Zoology of

the Voyage ofH.M.S. Beagle, under the Command ofCaptain Robert Fitzroy, R.N., during the Years 1832

89

Page 96: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

90 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

to 1836. part 5: 51 pages, 20 plates. London:Smith, Elder, and Co.

Boersma, P. Dee.1978. Breeding Patterns of Galapagos Penguins as an

Indicator of Oceanographic Conditions. Science,200(4349): 1481-1483, 2 figures.

Bonnott, Paul1928. An Outlaw Barn Owl. The Condor, 30(5):320.

Boulenger, G.A.1891. On the Galapagos Lizards of the Genus Tropidu-

rus. The Annals and Magazine of Natural History,series 6, 7 (58):501-503.

Bowman, Robert I.1961. Morphological Differentiation and Adaptation in

the Galapagos Finches. University of California Pub-lication in Zoology, 58:1-326, 74 figures, 63 tables,22 plates.

Bowman, Robert I., and Anne Carter1971. Egg-pecking Behavior in Galapagos Mocking-

birds. The Living Bird, 10:243-270, 16 figures, 7tables.

Brattstroni, Bayard H., and Thomas R. Howell1956. The Birds of the Revilla Gigedo Islands, Mexico.

The Condor, 58(2): 107-120, 5 figures.Brodkorb, Pierce

1963. Catalogue of Fossil birds, Part 1. Bulletin of theFlorida State Museum, Biological Sciences, 7(4): 179-293.

Broom, Robert1929. On the Extinct Galapagos Tortoise That Inhab-

ited Charles Island. Zoologica, 9(8):313-320, fig-ures 306-309.

Brosset, A.1963. Statut Actuel des mammiferes des lies Galapagos.

Mammalia, 27(3):324-338, plate 12.Buden, Donald W.

1974. Prey Remains of Barn Owls in the Southern Ba-hamas Islands. The Wilson Bulletin, 86(4):336-343, 1 figure, 1 table.

Cane, Mark A.1983. Oceanographic Events During El Nino. Science,

222(4629); 1189-1195, 6 figures.Carpenter, Charles C.

1966. Comparative Behavior of the Galapagos Lava Liz-ards (Tropidurus) In Robert I. Bowman, editor,The Galapagos: Proceedings of the Galapagos Inter-national Scientific Project, pages 269-273, 1 figure,1 table. Berkeley: University of California Press.

1970. Miscellaneous Notes of Galapagos Lava Lizards(Tropidurus—Iguanidae). Herpetologica, 26(3):377-386, 2 figures, 1 table.

Chambers, Steven M., and David W. SteadmanIn press. Holocene Terrestrial Gastropod Faunas from

Isla Santa Cruz and Isla Floreana, Galapagos.Transactions of the San Diego Society of Natural

History, volume 22.Chestennan, Charles W.

1963. Contributions to the Petrography of the Galapa-gos, Cocos, Malpelo, Cedros, San Benito, TresMarias, and White Friars Islands. Proceedings ofthe California Academy of Sciences, series 4,32(ll):339-362, 7 figures.

Chubb, Lawrence John1933. Geology of the Galapagos, Cocos, and Easter Is-

lands. Bernice P. Bishop Museum Bulletin, 1 1 0 : 1 -44, 65-67, figures 1-8, tables 1-3, plates 1-5.

Clark, Deborah A.1984. Native Land Mammals. In Roger Perry, editor,

Galapagos, pages 225-231, 1 table. Oxford: Per-gamon Press.

Coblentz, Bruce E.1978. The Effects of Feral Goats (Capra hircus) on Island

Ecosystems. Biological Conservation, 13(4):279-286, 2 figures.

Colinvaux, Paul A.1968. Reconnaissance and Chemistry of the Lakes and

Bogs of the Galapagos Islands. Nature,219(5154):590-594, 7 figures, 3 tables.

1969. Paleolimnological Investigations in the GalapagosArchipelago. Mitteilungen der Internationalen Ver-einigungfur Limnologie, 17:126-130, plate 1.

1972. Climate and the Galapagos Islands. Nature,240(5375): 17-20, 3 figures.

Colinvaux, Paul A., and Eileen K. Schofield1976a. Historical Ecology in the Galapagos Islands, I: A

Holocene Pollen Record from El Junto Lake, IslaSan Cristobal. Journal of Ecology, 64(3)989-1012,6 figures, 5 tables.

1976b. Historical Ecology in the Galapagos Islands, II: AHolocene Spore Record from El Junco Lake, IslaSan Cristobal. Journal of Ecology, 64(3): 1013-1026, 3 figures, 1 table.

Connor, Edward F., and Earl D. McCoy1979. The Statistics and Biology of the Species-area

Relationship. The American Naturalist, 113(6):791-833, 5 figures, 7 tables, appendix with 3tables.

Connor, Edward F., and Daniel Simberloff1978. Species Number and Compositional Similarity of

the Galapagos Flora and Avifauna. Ecological Mon-ographs, 48(2):219-248, 3 figures, 10 tables.

Conway, Ainslie, and Frances Conway.1947. The Enchanted Islands. 280 pages. New York, G.P.

Putnam's Sons.Cookson, W.E.

1876. [Letter from Commander W.E. Cookson to Dr.A. Gunther of May 29, 1876.] Proceedings of theZoological Society of London, 1876:520-526.

Corbet, Gordon Barclay1978. The Mammals of the Palaearctic Region: A Taxonomic

Page 97: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 91

Review. 314 pages, 104 maps, 1 table. London:British Museum (Natural History).

Coulter, Malcolm C.1982. The Dark-rumped Petrel, Pterodroma phaeopygia,

on Isla Floreana, Galapagos, lnforme Anual 1981Estacion Cientifica Charles Darwin, pages 170-173,1 table.

Cox, Allan1983. Ages of the Galapagos Islands. In Robert I. Bow-

man, Margaret Berson, and Alan E. Leviton, edi-tors. Patterns of Evolution in Galapagos Organisms.568 pages. San Francisco: American Associationfor the Advancement of Science, Pacific Division.

Cox, Allan, and G. Brent Dalrymple1966. Paleomagnetism and Potassium-argon Ages of

Some Volcanic Rocks from the Galapagos Islands.Nature, 209:(5025):776, 777, 2 figures, 2 tables.

Crumly, Charles R.1984. The Cranial Morphometry of Galapagos Tor-

toises. Proceedings of the California Academy of Sci-ences, 43(9): 111-121,5 figures, 8 tables.

Cruz, Felipe, and Tina Beach1983. Weather Data on Floreana, 1982-83. Noticias de

Galapagos, 38:28.Darwin, Charles

1869. Geological Observations on the Volcanic Islands andParts of South America Visited during the Voyage ofH.M.S. Beagle. Third edition, 648 pages [pages110-131, figures 11-14 on Galapagos]. NewYork: D. Appleton and Co.

1871. Journal of Researches into the Natural History andGeology of the Countries Visited during the Voyage ofH.M.S. Beagle Round the World, Under the Com-mand ofCapt. Fitz Roy, R.N. Volume 2, 615 pages,unnumbered figures. New York: Harper andBros.

Davis, D.H.S.1959. The Barn Owl's Contribution to Ecology and

Paleoecology. The Ostrich, supplement 3:144-153,2 figures, 1 table.

Davis, John, and Alden H. Miller1960. Family Mimidae. In Ernst Mayr and James C.

Greenway, editors, Check-list of Birds of the World,9:440-458. Cambridge, Massachusetts: Museumof Comparative Zoology, Harvard University.

Dawson, E. Yale1962. Cacti of the Galapagos Islands and of Coastal

Ecuador. Cactus and Succulent Journal, 34(3):67-74, figures 42-47; 34(4):99-105, figures 64-70.

1966. Cacti in the Galapagos Islands, with Special Ref-erence to Their Relations with Tortoises. In Rob-ert I. Bowman, editor, The Galapagos: Proceedingsof the Symposia of the Galapagos International Sci-entific Project, pages 209-214. Berkeley: Univer-sity of California Press.

DeBenedictis, Paul1966. The Flight Song Display of Two Taxa of Ver-

milion Flycatchers, Genus Pyrocephalus. The Con-dor, 68(3):306, 307.

de Bruijn, O.1979. Voedseloecologie van de kerkuil Tyto alba in Ned-

erland. [Feeding Ecology of the Barn Owl Tytoalba in the Netherlands.]Limosa, 52(3-4):91-154,4 figures, 14 tables.

Delano, Amasa1817. A Narrative of Voyages and Travels in the Northern

and Southern Hemispheres Comprising Three Voyagesaround the World Together with a Voyage of Surveyand Discovery in the Pacific Ocean and OrientalIslands. 598 pages [pages 369-385 on Galapa-gos]. Boston: E.G. House. [Not seen by DWS.]

DeNiro, Michael J., and Samuel Epstein1978. Carbon Isotopic Evidence for Different Feeding

Patterns in Two Hyrax Species Occupying theSame Habitat. Science, 201:906-908.

De Paepe, Paul1967. Cavity Filling in Rocks from Lava Tunnels on the

Galapagos Islands. Noticias de Galapagos, 5/6:19,20.

De Ridder, C.1976. [Visit to Isla Champion, 21 January 1976]. [Un-

published field trip itinerary report, on file at theCharles Darwin Research Station Library, SantaCruz, Galapagos.]

Duffy, David C.1981. A Master Plan for Ornithology in the Galapagos

Islands. Noticias de Galapagos, 34:10-16,2 figures.Durden, C.J.

1978. Fossil Cockroaches from a 1554 Spanish Ship-wreck. In J.B. Arnold, III and R. Weddle, editors,The Nautical Archeology of Padre Island: The Span-ish Shipwrecks of 1554, pages 407-416. New York:Academic Press. [Not seen by DWS.]

Eckhardt, Robert C.1972. Introduced Plants and Animals in the Galapagos

Islands. BioScience, 22(10):585-590, 3 figures, 1table.

Fleming, Theodore H.1982. Parallel Trends in the Species Diversity of West

Indian Birds and Bats. Oecologia, 53(l):56-60, 3figures, 2 tables.

Fosberg, F. Raymond, Marie-Helene Sachet, and D.R. Stod-dart

1983. Henderson Island (Southeastern Polynesia): Sum-mary of Current Knowledge. Atoll Research Bul-letin, 272:47 pages, 2 figures, 12 plates.

Fritts, Thomas H., and Patricia R. Fritts, editors1982. Race with Extinction; Herpetological Field Notes

of J.R. Slevin's Journey to the Galapagos, 1905-1906. Herpetological Monograph, 1: 98 pages, 9

Page 98: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

92 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

figures, 3 tables, 4 appendices.Garman, Samuel

1892. The Reptiles of the Galapagos Islands. Bulletin ofthe Essex Institute, 24:1-15.

Gifford, Edward Winslow1913. Expedition of the California Academy of Sciences

to the Galapagos Islands, 1905-1906, VIII: TheBirds of the Galapagos Islands, with Observationson the Birds of Cocos and Clipperton islands (Col-umbiformes to Pelecaniformes). Proceedings of theCalifornia Academy of Science, series 4, 2( 1): 1 -132,plates 1-7.

1919. Expedition of the California Academy of Sciencesto the Galapagos Islands, 1905-1906, XIII: FieldNotes on the Birds of the Galapagos Islands andof Cocos Island, Costa Rica. Proceedings of theCalifornia Academy of Sciences, series 4, 2(2): 189-258.

Goodfriend, Glenn A., and Darden G. Hood1983. Carbon Isotope Analysis of Land Snail Shells:

Implications for Carbon Sources and RadiocarbonDating. Radiocarbon, 25(3):810-830, 4 Figures, 3tables.

Goodfriend, Glenn A., and Jerry J. Stipp1983. Limestone and the Problem of Radiocarbon Dat-

ing of Land-snail Shell Carbonate. Geology,ll(10):575-577, 2 tables.

Gould, John1837a. [Remarks on a series of "ground finches" from

Mr. Darwin's collection, with descriptions of newspecies.] Proceedings of the Zoological Society of Lon-don, 5:4-7.

1837b. [Two new species of the genus Sterna, a species ofcormorant, and three species of the genus Orpheusfrom the Galapagos, in the collection of Mr. Dar-win.] Proceedings of the Zoological Society of London,5:26-27.

1841. Birds. In Charles Darwin, editor, The Zoology of theVoyage of H.M.S. Beagle, under the Command ofCaptain Robert Fitzroy, R.N., during the Years 1832to 1836, part 3: 156 pages, 50 plates. London:Smith, Elder, and Co.

Gramlich, John W., Virginia A. Lewis, and John J. Naughton1971. Potassium-argon Dating of Holocene Basalts of

the Honolulu Volcanic Series. Geological Society ofAmerica Bulletin, 82(5): 1399-1404, 1 figure, 1table.

Grant, B. Rosemary1981. [Visit to Isla Champion, 5 January 1981.] [Unpub-

lished field trip itinerary report, on file at theCharles Darwin Research Station Library, SantaCruz, Galapagos.]

Grant, B. Rosemary, and Peter R. Grant1981. Exploitation of Opuntia Cactus by Birds on the

Galapagos. Oecologia, 49 (2): 179-187, 2 figures, 6tables.

1982. Niche Shifts and Competition in Darwin's finches:Geospiza conirostris and Congerers. Evolution,36(4):637-657, 5 figures, 8 tables.

Grant, Peter R.1975. Four Galapagos Islands. Geographical fournal,

141(l):76-87. 2 figures, 2 tables, 4 plates.1980. [Visit to Isla Champion, 6-23 August 1980.] [Un-

published field trip itinerary report, on file at theCharles Darwin Research Station, Santa Cruz,Galapagos.]

1981. Speciation and the Adaptive Radiation ofDarwin'sFinches. American Scientist, 69(6):653-663, 8 fig-ures, 1 table.

Grant, Peter R., and B. Rosemary Grant1980. The Breeding and Feeding Characteristics of

Darwin's Finches on Isla Genovesa, Galapagos.Ecological Monographs, 50(3):381-410,16 figures,19 tables.

Grant, Peter R., and K. Thalia Grant1979. Breeding and Feeding Ecology of the Galapagos

Dove. The Condor, 81(4):397-403, 1 figure, 4tables.

Grant, Peter R., and Nicola Grant1979. Breeding and Feeding of Galapagos Mocking-

birds, Nesomimus parvulus. The Auk, 96(4):723-736, 1 figure, 5 tables.

Gregg, D.R.1972. Holocene Stratigraphy and Moas at Pyramid Val-

ley, North Canterbury, New Zealand. Records ofthe Canterbury Museum, 9(2): 151-158, 3 figures.

Groot, R.S. de1983. Origin, Status and Ecology of the Owls in Gala-

pagos. Ardea, 71(2): 167-182, 3 figures, 6 tables,3 appendices.

Guilday, John E., Paul W. Parmalee, and Harold W. Ham-ilton

1977. The Clark's Cave Bone Deposit and the LatePleistocene Paleoecology of the Central Appala-chian Mountains of Virginia. Bulletin of the Car-negie Museum of Natural History, 2: 87 pages, 21figures, 23 tables.

Gunther, Albert1860. On a New Snake from the Galapagos Islands.

Annals and Magazine of Natural History, series 3,6:78, 79.

1875. Description of the Living and Extinct Races ofGigantic Tortoises, Part I: Introduction; Part II:The Tortoises of the Galapagos Islands. Philosoph-ical Transactions of the Royal Society of London,165(1)7:251-284, plates 33-45, unnumbered ta-bles.

1877a. Account of the Zoological Collection Made duringthe Visit of H.M.S. "Peterel" to the GalapagosIslands. Proceedings of the Zoological Society of Lon-don, 1877:64-91, plates 11-13.

1877b. The Gigantic Land-Tortoises Living and Extinct in

Page 99: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 93

the Collection of the British Museum. 96 pages, 54plates. London: British Museum.

1896. Testudo ephippium. Novitates Zoologicae, 3(4):329-334, plates 20-22.

1902. Testudo galapagoensis [sic]. Novitates Zoologicae,9:184-192, plates 16-21.

Haas, Peter H.1975. Some Comments on Use of the Species-area

Curve. The American Naturalist, 109(967):371-373, 1 figure.

Ham.nm, Ole1975. Vegetational Changes in the Galapagos Islands

during the Period 1966-1973. Biological Conser-vation, 7(l):37-59, 7 figures, 13 tables.

1979. On Climatic Conditions, Vegetation Types, andLeaf Size in the Galapagos Islands. Biotropica,11(2): 101-122, 19 figures, 6 tables, appendix.

1981. Plant Communities of the Galapagos Islands.Dansk Botanisk Arkiv, 34(2): 1-163, 54 figures, 20tables.

Hamilton, Terrell H., and Ira Rubinoff1963. Isolation, Endemism, and Multiplication of Spe-

cies in the Darwin's Finches. Evolution, 17(6):388-403, 6 figures, 4 tables.

1964. On Models Predicting Abundance of Species andEndemics for the Darwin's Finches in the Gala-pagos Archipelago. Evolution, 18(2):339-342, 1figure, 2 tables.

1967. On Predicting Insular Variation in Endemism andSympatry for the Darwin's Finches in the Gala-pagos Archipelago. The American Naturalist,101(918):161-171, 2 figures, 2 tables.

Harcourt, Sylvia1982. [Visit to Isla Champion, 4 January 1982.] [Unpub-

lished field trip itinerary report, on file at theCharles Darwin Research Station, Santa Cruz,Galapagos.]

Harlan, Richard1827. Description of a Land Tortoise, from the Galla-

pagos [sic] Islands, Commonly Known as the "El-ephant Tortoise."yourno/ of the Academy of NaturalSciences of Philadelphia, 5 ( 2 ) : 2 8 4 - 2 9 2 , p late 1 1 .

Harris, Michael P.1968. Egg-eating Behavior in Galapagos Mockingbirds.

The Condor, 70(3):269, 270, 1 figure.1973. The Galapagos Avifauna. The Condor, 75(3):265-

278, 2 figures, 3 tables.1974. A Field Guide to the Birds of Galapagos. First edi-

tion, 160 pages, 68 figures, 12 plates. New York:Taplinger Publishing Co., Inc.

1982. A Field Guide to the Birds of Galapagos. Secondedition. 160 pages, 68 figures, 12 plates. London:William Collins Sons 8c Co., Ltd.

Hatch, Jeremy J.1965. Only One Species of Galapagos Mockingbird

Feeds on Eggs. The Condor, 67(3):354, 355.

Heller, Edmund1903. Papers from the Hopkins Stanford Galapagos Ex-

pedition, 1898-1899, XIV: Reptiles. Proceedingsof the Washington Academy of Sciences, 5:39-98,unnumbered tables.

1904. Mammals of the Galapagos Archipelago, Exclusiveof the Cetacea. In Papers from the Hopkins Stan-ford Galapagos Expedition, 1898-1899. Proceed-ings of the California Academy of Sciences, thirdseries (Zoology), 3(7)233-250, plates 23, tables.

Hellmayr, Charles E.1934. Catalogue of Birds of the Americas, part VII. Field

Museum of Natural History, Zoological Series,8(330): 1-531.

1935. Catalogue of Birds of the Americas, part VIII.Field Museum of Natural History, Zoological Series,8(347): 1-541.

Hendrickson.John D.1966. The Galapagos Tortoises, Geochelone Fitzinger,

1835 (Testudo Linnaeus, 1758, in part). In RobertI. Bowman, editor, The Galapagos: Proceedings ofthe Symposia of the Galapagos International ScientificProject, pages 252-257, 1 table. Berkeley: Univer-sity of California Press.

Herrera, Carlos M., and Fabian M. Jaksic1980. Feeding Ecology of the Barn Owl in Central Chile

and Southern Spain: A Comparative Study. TheAuk, 97(4): 760-767, 1 figure, 3 tables.

Hertlein, Leo G.1972. Pliocene Fossils from Baltra (South Seymour) Is-

land, Galapagos Islands. Proceedings of the Califor-nia Academy of Sciences, series 4, 39(3):25-46, 39figures.

Hey, Richard, G., Leonard Johnson, and Allen Lowrie1977. Recent Plate Motions in the Galapagos Area. Geo-

logical Society of America Bulletin, 88(10): 1385-1403, 15 figures, 3 tables.

Heyerdahl, Thor1963. Archaeology in the Galapagos Islands. Occasional

Papers of the California Academy of Sciences, 44:45-51.

Heyerdahl, Thor, and Arne Skjolsvold1956. Archaeological Evidence of Pre-Spanish Visits to

the Galapagos Islands. Memoirs of the Society forAmerican Archaeology, 12:1-71, 48 figures, 2 ta-bles, appendix.

Hickman, Carole S., and Jere H. Lipps1985. Geologic Youth of Galapagos Islands Confirmed

by Marine Stratigraphy and Paleontology. Science,227:1578-1580, 2 figures.

Hooker, Joseph Dalton1847. Enumeration of the Plants of the Galapagos Is-

lands, with Descriptions of Those Which are New.The Transactions of the Linnean Society of London,20(2)9:163-233.

Page 100: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

94 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Houvenaghel, Guy T.1974. Equatorial Undercurrent and Climate in the Ga-

lapagos Islands. Nature, 250(5467):565, 566, 2figures.

Howell, John Thomas1933. The Cactaceae of the Galapagos Islands. Proceed-

ings of the California Academy of Sciences, series 4,21(5):41-54, plates 2, 3.

1934. Cacti in the Galapagos Islands. Cactus and Succu-lent Journal, 5(8):515-518, 6 figures.

1941. Hugh Cuming's Visit to the Galapagos Islands.Lloydia, 4(4):291-292.

Howmiller, Richard, and Keith Dahnke1969. Chemical Analysis of Salt from Tagus Crater lake,

Isabela, Galapagos. Limnology and Oceanography,14(4):602-604, 1 figure, 2 tables.

Jaksic, Fabian M., Robert L. Seib, and Carlos M. Herrera1982. Predation by the Barn Owl (Tyto alba) [sic] in

Mediterranean Habitats of Chile, Spain and Cali-fornia: A Comparative Approach. The AmericanMidland Naturalist, 107(1): 151-162, 1 figure, 3tables, appendix.

Jaksic, Fabian M., and Jose L. Yanez1980. Differential Utilization of Prey Resources by

Great Horned Owls and Barn Owls in CentralChile. The Auk, 97(4):895, 896, 1 table.

Janzen, Daniel H.1976. The Depression of Reptile Biomass by Large Her-

bivores. The American Naturalist, 110(973):371-400, 1 table.

Johnson, Ned K.1983. The California Islands: Proceedings of a Multidis-

ciplinary Symposium [Review]. The Auk,100(3):782-784.

Johnson, R.G.1982. Brunhes-Matuyama Magnetic Reversal Dated at

790,000 yr B.P. by Marine-astronomical Correla-tions. Quaternary Research, 17(2): 135-147, 11 fig-ures, 2 tables.

Johnston, David W.1972 [1974]. Food of the Barn Owl on Grand Cayman,

B.W.I. Quarterly Journal of the Florida Academy ofSciences, 35(4): 171, 172.

1975. Ecological Analysis of the Cayman Island Avi-fauna. Bulletin of the Florida State Museum, Biolog-ical Sciences, 19(5):235-300, 19 figures, 18 tables,3 appendices.

Juvik, J.D., and A.P. Austring1979. The Hawaiian Avifauna: Biogeographic Theory

in Evolutionary Time. Journal of Biogeography,6:205-224, 2 figures, 6 tables, appendix.

Koford, Carl B.1966. Economic Resources of the Galapagos Islands. In

Robert I. Bowman, editor, The Galapagos: Proceed-ings of the Symposia of the Galapagos International

Scientific Project, pages 286-290. Berkeley: Uni-versity of California Press.

Lack, David1945. The Galapagos Finches (Geospizinae): A Study in

Variation. Occasional Papers of the California Acad-emy of Sciences, 21: 159 pages, 4 plates, 16 text-tables, 10 main tables.

1946. The Names of the Geospizinae (Darwin's Finches).Bulletin of the British Ornithologicists' Club,67(469): 15-22, 1 table.

1947. Darwin's Finches. 204 pages, 27 figures, 3 plates,32 tables, 4 supplementary tables. Cambridge,England: Cambridge Univ. Press.

1969. Subspecies and Sympatry in Darwin's Finches. Ev-olution, 23(2):252-263, 2 tables, 3 appendices.

Lanyon, Wesley E.1978. Revision of the Myiarchus Flycatchers of South

America. Bulletin of the American Museum of Nat-ural History, 161(4):427-628, 89 figures, 48 ta-bles.

Laruelle, Jacques1966. Study of a Soil Sequence on Indefatigable Island.

In Robert I. Bowman, editor, The Galapagos: Pro-ceedings of the Symposia of the Galapagos Interna-tional Scientific Project, pages 87-92, 8 tables.Berkeley: University of California Press.

Leveque, Raymond1963. Le Statut actuel des vertebres rares et menaces de

l'archipel des Galapagos. La Terreet la Vie, 4:397-430, 12 figures, 9 maps, 4 tables.

Levinson, M.1982. Taphonomy of Microvertebrates—from Owl Pel-

lets to Cave Breccia. Annals of the Transvaal Mu-seum, 33(6): 115-121, 5 figures.

Lewin, Roger1983. Finches Show Competition in Ecology. Science,

219(4591):1411, 1412, 1 figure.Linsley, E.G.

1977. Insects of the Galapagos (Supplement). OccasionalPapers of the California Academy of Sciences, 125:50pages.

Linsley, E.G., and R.L. Usinger1966. Insects of the Galapagos Islands. Proceedings of the

California Academy of Sciences, series 4, 33(7): 113-196, 1 table.

Loomis, Leverett Mills1918. Expedition of the California Academy of Sciences

to the Galapagos Islands, 1905-1906, XII: A Re-view of the Albatrosses, Petrels, and Diving Pe-trels. Proceedings of the California Academy of Sci-ences, series 4, 2(2):1-187, plates 1-17, unnum-bered tables.

MacArthur, Robert H., and Edward O. Wilson1963. An Equilibrium Theory of Insular Biogeography.

Page 101: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 95

Evolution, 17(4):373-387, 5 figures, 3 tables, ap-pendices.

1967. The Theory of Island Biogeography. 203 pages, 60figures, 13 tables. Princeton, New Jersey: Prince-ton University Press.

MacFarland, Craig G., Jose Villa, and Basilio Toro1974a. The Galapagos Giant Tortoise (Geochelone ele-

phantopus). Part I: Status of the Surviving Popu-lations. Biological Conservation, 6(2): 118-133, 7figures, 8 tables.

1974. The Galapagos Giant Tortoise {Geochelone ele-phantopus), Part II: Conservation Methods. Biolog-ical Conservation, 6(3): 198-212, 5 figures, 10 ta-bles.

Maglio, Vincent J.1970. West Indian Xenodontine Colubrid Snakes: Their

Probable Origin, Phylogeny, and Zoogeography.Bulletin of the Museum of Comparative Zoology atHarvard University, 141(1): 1-54, 37 figures, 2 ta-bles, appendix.

Marchant, Stephen1958. The Birds of the Santa Elena Peninsula, S.W.

Ecuador. The Ibis, 100(3):349-387, 1 figure.1959. The Breeding Season in S.W. Ecuador. The Ibis

101(2): 137-152, 1 figure, 1 table.Markham, A.H.

1880. A Visit to the Galapagos Islands in 1880. Proceed-ings of the Royal Geographical Society and MonthlyRecord of Geography, new series, 2(12):742-755.

Marlow, Ronald W., and James L. Patton1981. Biochemical Relationships of the Galapagos Giant

Tortoises (Geochelone elephantopus). Journal of Zo-ology, London, 195(3):413-422, 1 figure, 4 tables.

McBirney, Alexander R., and Ken-ichiro Aoki1966. Petrology of the Galapagos Islands. In Robert I.

Bowman, editor, The Galapagos: Proceedings of theSymposia of the Galapagos International ScientificProject, pages 71-77, 4 figures, 2 tables. Berkeley:University of California Press.

McBirney, Alexander R., and Howel Williams1969. Geology and Petrology of the Galapagos Islands.

Geological Society of America Memoir, 118:197pages, 48 figures, 21 plates, 21 tables.

McCulloch, Beverly, and Michael M. Trotter1979. Some Radiocarbon Dates for Moa Remains from

Natural Deposits. New Zealand Journal of Geologyand Geophysics, 22(2):277-279.

Mearns, Edgar A.

1903. Feathers Beside the Styx. The Condor, 5(2):36-38.

Mertens, Robert1960. Uber die Schlangen der Galapagos. Senckenber-

giana Biologica, 41 (3/4): 133-141.Montoriol-Pous, Joaquin, and Oleguer Escola

1975. Contribucion al conocimiento vulcano-espeleolo-

gico de la isla Floreana (Galapagos, Ecuador).In V Symposium de Espeleologia. Speleon, 1:57-67.

Morton, S.R., and A.A. Martin1979. Feeding Ecology of the Barn Owl, Tyto alba, in

Arid Southern Australia. Australian Wildlife Re-search, 6(2): 191-204, 4 figures, 4 tables.

Murphy, Robert Cushman1936. Oceanic Birds of South America. Volume 1, 640

pages, frontispiece, 61 figures, 5 unnumberedcolor plates, 38 numbered black-and-white plates.New York: American Museum of Natural History.

Newell, Reginald E.1973. Climate and the Galapagos Islands. Nature,

245(5420):91,92, 1 table.

Niethammer, Jochen1964. Resultats de l'expedition allemande aux Galapa-

gos 1962/63, VII: Contribucion al connaissancedes mammiferes terrestres de Pile Indefatigable(=Santa Cruz), Galapagos. Mammalia, 28(4):593-606, 4 figures, 4 tables.

Norman, F.I.1970. Food Preferences of an Insular Population of Rat-

tus rattus. Journal of Zoology, London, 162 (4):493-503, 2 figures, 3 tables.

1975. The Murine Rodents Rattus rattus, exulans, andnorvegicus as Avian Predators. Atoll Research Bul-letin, 182:1-13.

Olson, Storrs L.1975. Paleornithology of St. Helena Island, South Atlan-

tic Ocean. Smithsonian Contributions to Paleobiol-ogy, 23: 49 pages, 10 figures, 6 plates, 8 tables.

1977. Additional Notes on Subfossil Bird Remains fromAscension Island. The Ibis, 119(1)37-43, 2 figures.

1978. A Paleontological Perspective of West IndianBirds and Mammals. In Frank B. Gill, editor,Zoogeography in the Caribbean: The 1975 LeidyMedal Symposium. Academy of Natural Sciences ofPhiladelphia Special Publication, 13:99-117, 1 fig-ure.

1979. Multiple Origins of the Ciconiiformes. Proceedings1978 Conference of the Colonial Waterbird Group,pages 165-170.

1981. Natural History of Vertebrates on the Brazilianislands of the Mid South Atlantic. National Geo-graphic Society Research Reports, 13:481-492.

Olson, Storrs L., and William B. Hilgartner1982. Fossil and Subfossil Birds from the Bahamas. In

Storrs L. Olson, editor, Fossil Vertebrates fromthe Bahamas. Smithsonian Contributions to Paleo-biology, 48:22-56, figures 2-12, tables 5-12.

Olson, Storrs L., and Helen F.James1982a. Fossil Birds from the Hawaiian Islands: Evidence

for Wholesale Extinction by Man before Western

Page 102: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

96 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Contact. Science, 217 (4560):633-635, 1 figure, 1table.

1982b. Prodromusof the Fossil Avifauna of the HawaiianIslands. Smithsonian Contributions to Zoology, 365:59 pages, 12 figures, 5 tables.

1984. The Role of Polynesians in the Extinction of theAvifauna of the Hawaiian Islands. In Paul S. Mar-tin and Richard G. Klein, editors, Quaternary Ex-tinctions, pages 768-780, 3 figures, 2 tables. Tuc-son: University of Arizona Press.

Orr, Robert T.1966. Evolutionary Aspects of the Mammalian Fauna of

the Galapagos. In Robert I. Bowman, editor, TheGalapagos: Proceedings of the Symposia of the Gala-pagos International Scientific Project, pages 276-281, 4 figures. Berkeley: University of CaliforniaPress.

Otteni, Lee C, Eric G. Bolen, and Clarence Cottam1972. Predator-prey Relationships and Reproduction of

the Barn Owl in Southern Texas. The Wilson Bul-letin, 84(4):434-448, 1 figure, 11 tables.

Palmer, Clarence E., and Robert L. Pyle1966. The Climatological Setting of the Galapagos. In

Robert I. Bowman, editor, The Galapagos: Proceed-ings of the Symposia of the Galapagos InternationalScientific Project, pages 93-99, 4 figures. Berkeley:University of California Press.

Patton, James L., Suh Y. Yang, and Philip Myers1975. Genetic and Morphologic Divergence among In-

troduced Rat Populations (Rattus rattus) of theGalapagos Archipelago, Ecuador. Systematic Zool-ogy, 24(3):296-310, 2 figures, 7 tables.

Paynter, Raymond A., Jr.1970. Subfamily Emberizinae, Buntings and American

Sparrows. In Raymond A. Paynter, Jr., and Rob-ert W. Storer, editors, Check-list of Birds of theWorld, 13:3-214. Cambridge, Massachusetts: Mu-seum of Comparative Zoology, Harvard Univer-sity.

Pitt, William D., and Matt J.James1983. Late Cenozoic Marine Invertebrate Paleontology

of the Galapagos Islands [Abstract]. In WesternSociety of Malacologists Annual Report, 15:14, 15.

Porter, David1822. fournal of a Cruise Made to the Pacific Ocean by

Captain David Porter in the United States FrigateEssex in the Years 1812, 1813, 1814. Second edi-tion, 2 volumes. New York.

Power, Dennis M.1975. Similarity among Avifaunas of the Galapagos Is-

lands. Ecology, 56(3):616-626, 8 figures, 1 table.Pregill, Gregory K.

1981. Late Pleistocene Herpetofaunas from PuertoRico. University of Kansas Museum of Natural His-

tory Miscellaneous Publications, 71: 72 pages, 26figures, 5 tables.

1982. Fossil Amphibians and Reptiles from New Provi-dence Island, Bahamas. In Storrs L. Olson, editor,Fossil Vertebrates from the Bahamas. SmithsonianContributions to Paleobiology, 48:8-21, tables 1-4.

Preston, Frank W.1962a. The Canonical Distribution of Commonness and

Rarity, Part I. Ecology, 43(2): 185-215, figures 1-31, tables 1-13.

1962b. The Canonical Distribution of Commonness andRarity, Part II. Ecology, 43(3):410-432, figures32-37, tables 14-24.

Pritchard, Peter Charles Howard1967. Living Turtles of the World. 288 pages, color and

black-and-white figures and plates, appendix. Jer-sey City, New Jersey: T.F.H. Publications, Inc.

Quinn, William H., and Wayne V. Burt1970. Prediction of Abnormally Heavy Precipitation

over the Equatorial Pacific Dry Zone. Journal ofApplied Meteorology, 9(l):20-28, 3 figures, 1 table.

Racine, Charles H., and Jerry F. Downhower1974. Vegetation and Reproductive Strategies of Opun-

tia (Cactaceae) in the Galapagos Islands. Biotropica,6(3): 175-186, 5 figures, 4 tables.

Ramage, C.S.1975. Preliminary Discussion of the Meteorology of the

1972-73 El Nino. Bulletin of the American Meteor-ological Society, 56(2):234-242, 13 figures, 2 ta-bles.

Rasmusson, Eugene M., and John M. Wallace1983. Meteorological Aspects of the El Nino/Southern

Oscillation. Science, 222(4629):! 195-1202, 6 fig-ures.

Ray, Clayton E., and Frank C. Whitmore, Jr.1973. Paleontology. In Tom Simkin, William G. Reeder,

and Craig G. MacFarland, editors, Galapagos Sci-ence: 1972 Conference, pages 67, 68. Washington,D.C. [Printed for the Smithsonian Institution;sponsored by the National Science Foundationand the University of Wisconsin.]

Reynolds, T.N.1835. Voyage of the United States Frigate Potomac under

the Command of Commodore John Downes, Duringthe Circumnavigation of the Globe, in the Years 1831,1832,1833, and 1834. 560 pages. New York. [Notseen by DWS.]

Richardson, Constance1933. Petrology of the Galapagos Islands. Bernice P.

Bishop Museum Bulletin, 110:45-67, figure 9, ta-bles 4-7.

Ridgway, Robert1889 [1890]. Scientific Results of Explorations by the

U.S. Fish Commission Steamer Albatross, I: Birds

Page 103: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 97

Collected on the Galapagos Islands in 1888. Pro-ceedings of the United States National Museum,12(767): 101-128, 6 figures.

1893. Remarks on the Avian Genus Myiarchus [sic], withSpecial Reference to M. yucatanensis [sic] Law-rence. Proceedings of the United States National Mu-seum, 16(955):605-608.

1894. Descriptions of Twenty-two New Species of Birdsfrom the Galapagos Islands. Proceedings of theUnited States National Museum, l7(1007):357-370.

1897. Birds of the Galapagos Archipelago. Proceedingsof the United States National Museum,19(1116):459-670, 7 figures, plates 56, 57, un-numbered maps and tables.

1901. The Birds of North and Middle America: FamilyFringillidae—The finches. Bulletin of the UnitedStates National Museum, 50 (1): 715 pages, 20plates, unnumbered tables.

1907. The Birds of North and Middle America: FamilyTurdidae—Thrushes [through] Family Cotingi-dae—Chatterers. Bulletin of the United States Na-tional Museum, 50(4): 973 pages, 34 plates, un-numbered tables.

Robinson, B.L.1902. Flora of the Galapagos Islands. Proceedings of the

American Academy of Arts and Sciences, 38(4):77-269, 3 plates, 4 tables.

Robinson, William Albert1936. Voyage to the Galapagos. 279 pages, illustrations,

plates, maps. New York: Harcourt, Brace and Co.Rose, Ruth

1924. Man and the Galapagos. In William Beebe, Gala-pagos: World's End, pages 332-417, figures 78,79. New York and London: G.P. Putnam's Sons.

Roth, Louis M.1982. Introduction. In WJ. Bell and K.G. Adiyodi, edi-

tors, The American Cockroach, pages 1-14. Lon-don: Chapman and Hall. [Not seen by DWS.]

Rothschild, Walter1902. Note Regarding Testudo elephantopus. Novitates

Zoologicae, 9:448.1907. Extinct Birds, xii + 244 pages, 42 plates. London:

Hutchinson & Co.Rothschild, Walter, and Ernest Hartert

1899. A Review of the Ornithology of the GalapagosIslands, with Notes on the Webster-Harris Expe-dition. Novitates Zoologicae, 6(2):85-205, plates 5,6, unnumbered figures and tables.

1902. Further Notes on the Fauna of the GalapagosIslands: Notes on the Birds. Novitates Zoologicae,9:381-418, plate 10, 2 figures.

Ruprecht, Andrej L.1979. Food of the Barn Owl, Tyto alba guttata (C. L. Br.)

from Kujawy. Acta Ornithologica, 16(19):493-511,2 figures, 1 table.

Salvin, Osbert1876. On the Avifauna of the Galapagos Archipelago.

Transactions of the Zoological Society of London,9(9):447-510, plates 84-89, 1 map, unnumberedfigures.

1883. A List of the Birds Collected by Captain A.H.Markham on the West Coast of America. Proceed-ing of the Zoological Society of London, 1883:419-432.

Sauer, Jonathan D.1969. Oceanic Islands and Biogeographic Theory: A

Review. The Geographical Review, 59(4):582-593.1982. Cayman Islands Seashore Vegetation: A Study in

Comparative Biogeography. University of Califor-nia Publications in Geography, 25: xii + 161 pages,1 map, 25 profiles, 39 plates, 13 tables.

Schluter, D., and Peter R. Grant1982. The Distribution of Geospiza difficilis in Relation

to G. fuliginosa in the Galapagos Islands: Tests ofThree Hypotheses. Evolution, 36(6):1213-1226,6 figures, 5 tables.

Seebohm, Henry1881. Catalogue of the Passeriformes, or Perching

Birds, in the Collection of the British Museum:Cichlomorphae, Part II: Containing the FamilyTurdidae (Warblers and Thrushes). In Catalogueof the Birds in the British Museum, 5:426 pages, 18color plates. London: British Museum.

Seemann, Berthold Carl1853. Narrative of the Voyage ofH.M.S. Herald during the

Years 1845-1851 under the Command of CaptainHenry Kellett, R.N., Being a Circumnavigation of theGlobe and Three Cruises to the Arctic Region in Searchof Sir John Franklin. 2 volumes in 1, frontispiece,map. London.

Sharpe, R. Bowdler1888. Catalogue of the Passeriformes, or Perching

Birds, in the Collection of the British Museum:Fringilliformes, Part III: Containing the FamilyFringillidae. In Catalogue of the Birds in the BritishMuseum, 12: 871 pages, 16 color plates. London:British Museum.

Shurcliff, Sidney Nichols1930. Jungle Islands: The "Illyria" in the South Seas. 298

pages, 90 illustrations, 2 maps, unnumbered draw-ings and color plates, appendix. New York: G.P.Putnam's Sons.

Simkin, Tom1984. Geology of the Galapagos. In Roger Perry, editor,

Galapagos, pages 15-41, 10 figures. Oxford: Per-gamon Press.

Page 104: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

98 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Simkin, Tom, and Keith A. Howard1970. Caldera Collapse in the Galapagos Islands, 1968.

Science, 169(3944):429-437, 6 figures.Simkin, Tom, Lee Siebert, Lindsay McClelland, David

Bridge, Christopher Newhall, and John H. Latter1982. Volcanoes of the World. 232 pages, 6 figures, 2

tables. Stroudsburg, Pennsylvania: HutchinsonRoss Pub. Co.

Simpson, Beryl B.1975. Glacial Climates in the Eastern Tropical Pacific.

Nature, 253(5486):34-36, 1 figure.Slevin, Joseph R.

1931. Log of the Schooner "Academy" on a Voyage ofScientific Research to the Galapagos Islands,1905-1906. Occasional Papers of the CaliforniaAcademy of Sciences, 17: 162 pages, 16 plates, 1map.

1935. An Account of the Reptiles Inhabiting the Gala-pagos Islands. Bulletin of the New York ZoologicalSociety, 38(l):3-24, 25 figures.

1955. Charting the "Enchanted Isles." In Essays in theNatural Sciences in Honor Captain Allan Hancockon the Occasion of His Birthday,July 26, 1955, pages99-110, 4 maps. Los Angeles: University ofSouthern California Press.

1959. The Galapagos Islands: A History of Their Explo-ration. Occasional Papers of the California Academyof Sciences, 25: x + 150 pages, 31 figures.

Snodgrass, Robert E., and Edmund Heller1904. Papers from the Hopkins-Stanford Galapagos Ex-

pedition, 1898-1899, XVI: Birds. Proceedings ofthe Washington Academy of Sciences, 5:231-372,unnumbered tables.

Steadman, David W.1981. Vertebrate Fossils in Lava Tubes in the Galapagos

Islands. In Barry F. Beck, editor, Proceedings of theEighth International Congress of Speleology, 2:549,550. Americus: Georgia Southwestern College.

1982. The Origin of Darwin's Finches. Transactions ofthe San Diego Society of Natural History,19(19):279-296, 3 figures.

1984. The Status of Geospiza magnirostris on Isla Flo-reana, Galapagos. Bulletin of the British Ornitholo-gists'Club, 104(3):99-102, 1 table.

1985. Fossil Birds from Mangaia, Southern Cook Is-lands. Bulletin of the British Ornithologists' Club,105(2):58-66, 1 table.

Steadman, David W., and Storrs L. Olson1985. Bird Remains from an Archaeological Site on

Henderson Island, South Pacific: Man-caused Ex-tinction on an "Uninhabited" Island. Proceedingsof the National Academy of Sciences, USA,82(18):6191-6195, 2 figures, 1 table.

Steadman, David W., Gregory K. Pregill, and Storrs L.Olson

1984. Fossil Vertebrates from Antigua, Lesser Antilles:

Evidence for Late Holocene Human-caused Ex-tinctions in the West Indies. Proceedings of theNational Academy of Sciences, USA, 81:4448-4451,2 figures, 2 tables.

Steadman, David W., and Clayton E. Ray1982. The Relationships of Megaoryzomys curioi, an Ex-

tinct Cricetine Rodent (Muroidea, Muridae) fromthe Galapagos Islands, Ecuador. Smithsonian Con-tributions to Paleobiology, 51: 23 pages, 11 figures,1 table.

Steindachner, Franz1876. Die Schlangen und Eidechsen der Galapagos-In-

seln. Festschrift Deutsch Kaiserlich-koniglichen Zool-ogische-Botanische Gesellschaft in Wien, 1876:301-330, plates 1-7.

Stewart, A.1911. A Botanical Survey of the Galapagos Islands. Pro-

ceedings of the California Academy of Sciences, series4, l(l):7-288.

Strauch, Dore1936. Satan Came to Eden, xii + 274 pages, 15 photo-

graphs, 2 drawings, 2 maps. New York: Harperand Bros.

Stuiver, M., and H.A. Polach1977. Reporting I4C Data. Radiocarbon, 19(3):355-363,

1 figure, 1 table.Sulloway, Frank J.

1982a. Darwin and His Finches: The Evolution of a Leg-end. Journal of the History of Biology, 15(1): 1-53,8 figures.

1982b. The Beagle Collections of Darwin's Finches (Geo-spizinae). Bulletin of the British Museum (NaturalHistory), Zoology Series, 43(2):49-94, 6 figures.

Sundevall, CarlJ.1871. On Birds from the Galapagos Islands. Proceedings

of the Zoological Society of London, 1871:124-130.Svenson, Henry K.

1935. Plants of the Astor Expedition, 1930 (Galapagosand Cocos Islands). American Journal of Botany,22:208-277.

Swarth, Harry S.1931. The Avifauna of the Galapagos Islands. Occasional

Papers of the California Academy of Sciences, 18: 299pages, 57 figures, 1 map, unnumbered tables.

Tanner, Z.L.1888. [Report on the Galapagos Islands.] [Unpublished

Manuscript in the Library of the Museum of Com-parative Zoology, Harvard University.] 16 pages.

Thomas, Richard, and Kim R. Thomas.1977. A Small-vertebrate Thanatocenosis from North-

ern Peru. Biotropica, 9(2): 131, 132, 1 table.Thornton, Ian

1971. Darwin's Islands: A Natural History of the Galapa-gos, xiv + 322 pages, 36 photographs, 28 figures.Garden City, New York: Natural History Press.

Page 105: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 99

Townsend, Charles Haskins1925a. The Galapagos Tortoises in Their Relation to the

Whaling Industry. Zoologica, 4(3):55-135, figures20-32.

1925b. The Whaler and the Tortoise. The ScientificM onthly,21(8): 166-172.

1928. The Galapagos Islands Revisited. Bulletin of theNew York Zoological Society, 31(5): 148-169, 32 fig-ures.

1930. Astor Expedition to the Galapagos Islands. Bulle-tin of the New York Zoological Society, 33(4): 135-155, 30 figures.

Treherne, J.1983. The Galapagos Affair, xv 4- 223 pages, 33 photo-

graphs, 6 Figures. New York: Random House, Inc.Trost, Charles H., and J. Howard Hutchison

1963 [ 1964]. Food of the Barn Owl in Florida. QuarterlyJournal of the Florida Academy of Sciences,26(4)382-384, 1 table.

Vanalek, George J.1982. Nycteris cinerea (Lasiurus cinereus) of the Galapa-

gos. Informe Anual 1981 Estacion Cientifica CharlesDarwin, pages 189-191, 1 figure, 1 table.

Van Denburgh, John1912a. Expedition of the California Academy of Sciences

to the Galapagos Islands, 1905-1906, IV: TheSnakes of the Galapagos Islands. Proceedings of theCalifornia Academy of Sciences, series 4, 1:323-374,plates 22-30, unnumbered tables.

1912b. Expedition of the California Academy of Sciencesto the Galapagos Islands, 1905-1906, VI: TheGeckos of the Galapagos Archipelago. Proceedingsof the California Academy of Sciences, series 4,1:405-430, unnumbered tables.

1914. Expedition of the California Academy of Sciencesto the Galapagos Islands, 1905-1906, X: TheGigantic Land Tortoises of the Galapagos Archi-pelago. Proceedings of the California Academy ofSciences, series 4, 2(l):203-374, plates 12-124,unnumbered tables.

Van Denburgh, J., and J.R. Slevin1913. Expedition of the California Academy of Sciences

to the Galapagos Islands, 1905-1906, IX: TheGalapagos Lizards of the Genus Tropidurus; withNotes on the Iguanas of the Genera Conolophusand Amblyrhynchus. Proceedings of the CaliforniaAcademy of Sciences, series 4, 2(1): 133-202, plates8-11, unnumbered tables.

Vernon, C.J.1972. An Analysis of Owl Pellets Collected in Southern

Africa. The Ostrich, 43(2): 109-124, 11 tables, ap-pendix.

Voigt, Birgit1977a. [Visit to Isla Champion, 1-11 May 1977]. Unpub-

lished field trip itinerary report, on file at theCharles Darwin Research Station Library, SantaCruz, Galapagos.

1977b. [Visit to Isla Champion, 19 September to 6 Octo-ber 1977.] Unpublished field trip itinerary report,on file at the Charles Darwin Research StationLibrary, Santa Cruz, Galapagos.

Voigt, Birgit, and Aksel Voigt1977. [Visit to Isla Champion, 27 February to 6 March

1977.] Unpublished field trip itinerary report, onfile at the Charles Darwin Research Station Li-brary, Santa Cruz, Galapagos.

Vries, Tjitte de1975. The Breeding Biology of the Galapagos Hawk,

Buteogalapagoensis. LeGerfaut,65(1 -4):29-57, 13figures, 2 tables.

1976. Prey Selection and Hunting Methods of the Ga-lapagos Hawk, Buteo galapagoensis. Le Gerfaut,66(l-2):3-42, 6 figures, 9 tables.

Wallace, George J.1948. The Barn Owl in Michigan: Its Distribution, Nat-

ural History and Food Habits. Michigan State Col-lege Agricultural Experimental Station TechnicalBulletin, 208: 61 pages, 2 figures, 7 tables.

Wetmore, Alexander, and Bradshaw H. Swales1931. The Birds of Haiti and the Dominican Republic.

United States National Museum Bulletin, 155: iv +483 pages, 26 plates.

Wiggins, Ira L., and Duncan M. Porter1971. Flora of the Galapagos Islands, xx + 998 pages,

268 figures, 96 color plates, 2 tables. Stanford,California: Stanford University Press.

Williams, Howel1966. Geology of the Galapagos Islands. In Robert I.

Bowman, editor, The Galapagos: Proceedings of theSymposia of the Galapagos International ScientificProject, pages 65-70. Berkeley: University of Cal-ifornia Press.

Williamson, Mark1981. Island Populations. 286 pages, many tables and

figures. Oxford: Oxford University Press.Wittmer, Margret

1961. Floreana Adventure. 239 pages, 22 photographs, 2maps. New York: E.P. Dutton and Co., Inc.[Translated from the German by Oliver Coburn.]

Wooster, Warren S., and Oscar Guillen1974. Characteristics of El Nino in 1972. Journal of

Marine Research, 32(3):387-404, 11 figures.

Page 106: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

Plates

PLATE 1.—Coracoids of Tyto: A, modern specimen of T. alba from Maryland, USNM 500619,female; B, modern specimen of T. punctatissima from Santa Cruz, UCMVZ 140963, female; c,fossil of 7. punctatissima from Barn Owl Cave, USNM 331288; all specimens X 1.8.

100

Page 107: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 101

PLATE 2.—Fossils of Mimus trifasdatus compared to modern specimens of M. parvulus fromSanta Cruz, UCMVZ 140977, on the right: A, tibiotarsus, USNM 284371, CPOS; B, tarsome-tatarsus, USNM 284361, CPOI; c, mandible, USNM 284371, CPOS; D, humerus, USNM284371, CPS. (Note larger size of M. trifasdatus in each element; all specimens X 1.9.)

Page 108: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

102 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

PLATE 3.—Skulls of Geospiza magnirostris, in lateral and dorsal aspect: A, G. m. strenua, UCMVZ130170, Santa Cruz, male; B, G. m. strenua, UCMVZ 140993, Genovesa, male; c, G. m.magnirostris fossil, USNM 284374 (cranium), 28435 (rostrum), Cueva de Post Office (Superior),Floreana. (Note that fossil cranium lacks quadrates, jugals, pterygoids, and palatines, thusgiving false appearance of being shallower than modern specimens; also note large size of fossilrostrum; note in fossil cranium the large, strong origin for M. adductor mandibularis externussuperficialis on each side of temporal region, as well as wide interorbital area just behindrostrum; all specimens approximately X 1.5)

Page 109: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

NUMBER 413 103

PLATE 4.—Mandibles of Geospiza magnirostris, in lateral anddorsal aspect: A, G. m. magnirostris fossil, USNM 284433,Cueva de Post Office (Superior), Floreana; B, G. m. strenua,UCMVZ 140993, Genovesa.male; c, G. m. strenua, UCMVZ130170, Santa Cruz, male. (Note overall larger size of fossil,especially in elevated surangular region; also note morepowerful nature of fossil, especially in large internal process(mandibular articulation); all specimens X 1.5.)

Page 110: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have
Page 111: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

REQUIREMENTS FOR SMITHSONIAN SERIES PUBLICATION

Manuscripts intended for series publication receive substantivereview (conducted by their originating Smithsonian museums oroffices) and are submitted to the Smithsonian Institution Presswith Form SI-36, which must show the approval of the appropriateauthority designated by the sponsoring organizational unit. Re-quests for special treatment—use of color, foldouts, case-boundcovers, etc.—require, on the same form, the added approval ofthe sponsoring authority.

Review of manuscripts and art by the Press for requirementsof series format and style, completeness and clarity of copy, andarrangement of all material, as outlined below, will govern, withinthe judgment of the Press, acceptance or rejection of manuscriptsand art.

Copy must be prepared on typewriter or word processor,double-spaced, on one side of standard white bond paper (noterasable), with 1 VA" margins, submitted as ribbon copy (notcarbon or xerox), in loose sheets (not stapled or bound), andaccompanied by original art. Minimum acceptable length is 30pages.

Front matter (preceding the text) should include: title pagewith only title and author and no other information, abstract pagewith author, title, series, etc., following the established format;table of contents with indents reflecting the hierarchy of heads inthe paper; also, foreword and/or preface, if appropriate.

First page of text should carry the title and author at the topof the page; second page should have only the author's nameand professional mailing address, to be used as an unnumberedfootnote on the first page of printed text.

Center heads of whatever level should be typed with initialcaps of major words, with extra space above and below the head,but no other preparation (such as all caps or underline, except forthe underline necessary for generic and specific epithets). Run-inparagraph heads should use period/dashes or colons as neces-sary.

Tabulations within text (lists of data, often in parallel columns)can be typed on the text page where they occur, but they shouldnot contain rules or numbered table captions.

Formal tables (numbered, with captions, boxheads, stubs,rules) should be submitted as carefully typed, double-spaced copyseparate from the text; they will be typeset unless otherwiserequested. If camera-copy use is anticipated, do not draw ruleson manuscript copy.

Taxonomic keys in natural history papers should use thealigned-couplet form for zoology and may use the multi-levelindent form for botany. If cross referencing is required betweenkey and text, do not include page references within the key, butnumber the keyed-out taxa, using the same numbers with theircorresponding heads in the text.

Synonymy in zoology must use the short form (taxon, author,yearpage), with full reference at the end of the paper under"Literature Cited." For botany, the long form (taxon, author,abbreviated journal or book title, volume, page, year, with noreference in "Literature Cited") is optional.

Text-reference system (author, yearpage used within the text,with full citation in "Literature Cited" at the end of the text) mustbe used in place of bibliographic footnotes in all ContributionsSeries and is strongly recommended in the Studies Series:"(Jones. 1910122)" or Jones (1910:122)." If bibliographic

footnotes are required, use the short form (author, brief title, page)with the full citation in the bibliography.

Footnotes, when few in number, whether annotative or biblio-graphic, should be typed on separate sheets and inserted imme-diately after the text pages on which the references occur. Exten-sive notes must be gathered together and placed at the end ofthe text in a notes section.

Bibliography, depending upon use, is termed "Literature Cited,"References," or "Bibliography." Spell out titles of books, articles,

journals, and monographic series. For book and article titles usesentence-style capitalization according to the rules of the lan-guage employed (exception: capitalize all major words in English).For journal and series titles, capitalize the initial word and allsubsequent words except articles, conjunctions, and prepositions.Transliterate languages that use a non-Roman alphabet accordingto the Library of Congress system. Underline (for italics) titles ofjournals and series and titles of books that are not part of a series.Use the parentheses/colon system for volume (number):pagination: "10(2):5-9." For alignment and arrangement of ele-ments, follow the format of recent publications in the series forwhich the manuscript is intended. Guidelines for preparing bibli-ography may be secured from Series Section, SI Press.

Legends for illustrations must be submitted at the end of themanuscript, with as many legends typed, double-spaced, to apage as convenient.

Illustrations must be submitted as original art (not copies)accompanying, but separate from, the manuscript. Guidelines forpreparing art may be secured from Series Section, SI Press. Alltypes of illustrations (photographs, line drawings, maps, etc.) maybe intermixed throughout the printed text. They should be termedFigures and should be numbered consecutively as they willappear in the monograph. If several illustrations are treated ascomponents of a single composite figure, they should be desig-nated by lowercase italic letters on the illustration; also, in thelegend and in text references the italic letters (underlined in copy)should be used: "Figure 9fc." Illustrations that are intended tofollow the printed text may be termed Plates, and any componentsshould be similarly lettered and referenced: "Plate 9b.." Keys toany symbols within an illustration should appear on the art ratherthan in the legend.

Some points of style: Do not use periods after such abbrevi-ations as "mm, ft, USNM, NNE." Spell out numbers "one" through"nine in expository text, but use digits in all other cases if possible.Use of the metric system of measurement is preferable; whereuse of the English system is unavoidable, supply metric equiva-lents in parentheses. Use the decimal system for precise meas-urements and relationships, common fractions for approximations.Use day/month/year sequence for dates: "9 April 1976." Formonths in tabular listings or data sections, use three-letter abbre-viations with no periods: "Jan, Mar, Jun," etc. Omit space betweeninitials of a personal name: "J.B. Jones."

Arrange and paginate sequentially every sheet of manu-script in the following order: (1) title page, (2) abstract, (3) con-tents, (4) foreword and/or preface, (5) text, (6) appendixes, (7)notes section, (8) glossary, (9) bibliography, (10) legends, (11)tables. Index copy may be submitted at page proof stage, butplans for an index should be indicated when manuscript is sub-mitted.

Page 112: Holocene Vertebrate Fossils from Isla Floreana, Galapagos · 2019-12-14 · Holocene Vertebrate Fossils from Isla Floreana, Galapagos. David W. Steadman Introduction Biologists have

i