The Biological Investigation of Malpelo Island, Colombia

JEFFREY B. GRAHAM EDITOR

SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY • NUMBER 176 SERIAL PUBLICATIONS OF THE SMITHSONIAN INSTITUTION The emphasis upon publications as a means of diffusing knowledge was expressed by the first Secretary of the Smithsonian Institution. In his formal plan for the Insti- tution, Joseph Henry articulated a program 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 year in all branches of knowledge." This keynote of basic research has been adhered to over the years in the issuance of thousands of titles in serial publications under the Smithsonian imprint, com- mencing with Smithsonian Contributions to Knowledge in 1848 and continuing with the following active series: Smithsonian Annals of Flight Smithsonian Contributions to Anthropology Smithsonian Contributions to Astrophysics Smithsonian Contributions to Botany Smithsonian Contributions to the Earth Sciences Smithsonian Contributions to Paleobiology Smithsonian Contributions to Zoology Smithsonian Studies in History and Technology In these series, the Institution publishes original articles and monographs dealing with the research and collections of its several museums and offices and of professional colleagues at other institutions of learning. These papers report newly acquired facts, synoptic interpretations of data, or original theory in specialized fields. These pub- lications are distributed by mailing lists to libraries, laboratories, and other interested institutions and specialists throughout the world. Individual copies may be obtained from the Smithsonian Institution Press as long as stocks are available.

S. DILLON RIPLEY Secretary Smithsonian Institution SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY • NUMBER 176

The Biological Investigation of Malpelo Island, Colombia

Jeffrey B. Graham EDITOR

SMITHSONIAN INSTITUTION PRESS City of Washington 1975 ABSTRACT

Graham, Jeffrey B., editor. The Biological Investigation of Malpelo Island, Colombia. Smithsonian Contributions to Zoology, number 176, 98 pages, 35 figures, 1975.—The results of joint Smithsonian Institution and U. S. Navy ter- restrial and marine investigations of Malpelo Island, Republic of Colombia are reported in 15 papers in this volume. A new of (Phyllodactylus), a new (Tamaria), two new species of shrimp (Alpheus and Synalpheus), and a new species of fish (Chriolepis) are described. The terrestrial ecology of Malpelo and the behavior and natural history of the Anolis agassizi and Diploglossus millepunctatus are described and discussed. Genie variability in A. agassizi has been investigated and karyotypes of A. agassizi and D. millepunc- tatus are reported. The ecology of the island's benthic marine communities is detailed and papers listing and discussing zoogeographically interesting features of the island's crustacean, starfish, and fish species are included. The geology of Malpelo is briefly described and an improved map of the island is presented. The importance of Malpelo Island in the understanding of biogeographic prob- lems in the eastern tropical Pacific Ocean is reviewed.

OFFICIAL PUBLICATION DATE is handstamped in a limited number of initial copies and is recorded in the Institution's annual report, Smithsonian Year. SI PRESS NUMBER 5095. SERIES COVER DESIGN: The coral Montastrea cavemosa (Linnaeus).

Library of Congress Cataloging in Publication Data Main entry under title: The Biological investigation of Malpelo Island, Colombia. (Smithsonian contributions to zoology, no. 176) Supt. of Docs, no.: SI 1.27:176. "The results of joint Smithsonian Institution and U.S. Navy terrestrial and marine investigations . . . are reported in 15 papers in this volume." I. Natural history—Malpelo Island. 2. Marine biology—Malpelo Island. 3. Malpelo Island. I. Graham, Jeffrey B., ed. II. Smithsonian Institution. III. United States. Navy. IV. Series: Smithsonian Institution. Smithsonian con- tributions to zoology, no. 176. [DNLM: 1. Ecology. 2. Lizards. 3. Marine biology. W1SM454N no. 176 1974/QL244 B6134 1974] QL1.S54 no. 176 [QH121] 591'.08s [WQ.V%6V5\ 74-23%

For sale by (he Superintendent of Documents,, U.S. Government Printing Office Washington, D.C. 20402 - Price $2.20 (paper cover) Contents

Page INTRODUCTION, by Jeffrey B. Graham 1 TERRESTRIAL BIOLOGY OF MALPELO ISLAND: A HISTORICAL REVIEW, by George C. Gorman and Terence L. Chorba 9 RECONNAISSANCE AND MAPPING OF MALPELO ISLAND, by A. Ross Kiester and Jeffrey A. Hoffman 13 FIELD OBSERVATIONS ON THE GEOLOGY OF MALPELO ISLAND, by Jeffrey A. Stead 17 THE ECOSYSTEM ON MALPELO ISLAND, by Henk Wolda 21 NATURAL HISTORY, BEHAVIOR, AND ECOLOGY OF Anolis agassizi, by A. Stanley Rand, George C. Gorman, and William M. Rand 27 NOTES ON THE NATURAL HISTORY OF Diploglossus millepunctatus (Sauria: Anguidae), by A. Ross Kiester 39 A NEW GECKO FROM MALPELO ISLAND (SAURIA: GEKKONIDAE: Phyllodactylus), by Raymond B. Huey 44 ELECTROPHORETIC ESTIMATES OF GENIC VARIATION IN, AND THE RELATIONSHIPS OF, Anolis agassizi, by T. Preston Webster 47 NOTES ON THE CHROMOSOMES OF Anolis agassizi (SAURIA: IGUANIDAE) AND Diploglossus millepunctatus (Sauria: Anguidae), by Brad Stamm and George C. Gorman 52 SWBTIDAL COMMUNITIES OF MALPELO ISLAND, by Charles Birkeland, David L. Meyer, James P. Stames, and Caryl L. Buford 55 THE MACRURAN DECAPOD CRUSTACEA OF MALPELO ISLAND, by Lawrence G. Abele ; 69 ASTEROIDEA FROM MALPELO ISLAND WITH A DESCRIPTION OF A NEW SPECIES OF THE Tamaria, by Maureen E. Downey 86 FISHES COLLECTED AT MALPELO ISLAND, by John E. McCosker and Richard H. Rosenblatt 91 A NEW SPECIES OF GOBY FROM MALPELO ISLAND (TELEOSTEI: : Chriolepis), by Lloyd Talbott Findley 94

in

The Biological Investigation of Malpelo Island, Colombia

Introduction

Jeffrey B. Graham

On 28 February 1972 a group of 17 scientists U.S.S. York County; J. A. Stead, U.S. Navy Hydro- representing the Smithsonian Tropical Research graphic Center. R. J. Kinney, USN, served as ex- Institute, the Republics of Panama and Colombia, pedition physician. The geology of Malpelo was and several U.S. universities embarked on a six studied and a more accurate map of the island was day expedition to investigate the marine and ter- made by an exploration team. restrial biota of Malpelo Island, Colombia, a small The scientists were transported from Panama to isolated Pacific island 270 miles to the west of Co- Malpelo by the U.S.S York County (LST 1175) lombia and south of Panama. The expedition, a commanded by LCDR Lauren Seeber. At Malpelo, joint project of the U.S Navy and the Smithsonian the scientists were supported by York County crew- Tropical Research Institute, was undertaken to men who did everything from prepare box lunches gain a more comprehensive understanding of to operate small boats and assist in the collection ecological processes and the natural history of and sort of specimens. While on station the York plants and on Malpelo. The scientific County conducted sea trials and made numerous party consisted mostly of biologists who made col- depth soundings and positional sightings of the lections, carried out field observations, and con- island. This work, while helping to establish that ducted experiments on the island and in the waters the peaks on Malpelo are some 125 meters higher around it. (The Malpelo scientific party consisted than previous maps indicated, will also improve of C. Birkeland, J. B. Graham, D. L. Meyer, A. S. the hydrographic charts of this area. Rand, A. Rodaniche, W. L. Smith, J. P. Stames, The Smithsonian-U.S. Navy Expedition to Mal- and H. Wolda, all of the Smithsonian Tropical pelo marked the fourth time that Smithsonian Research Institute (STRI); O. Arroyo, Colombian scientists and their colleagues have been able to Institute of Natural Resources (INDERENA); J. Bar- use a U.S. Navy vessel for research purposes. In reto S., Universidad George Tadeo Lozana, Colom- March 1970 biological investigations of the Secas bia; G. C. Gorman, University of California Los Islands in the Gulf of Chiriqui, Panama, were con- Angeles (UCLA); J. A. Hoffman, Smithsonian As- ducted by scientists aboard the U.S.S. Traverse trophysical Laboratory; A. R. Kiester, Museum County. Later in the same year the U.S.S. Wal- of Comparative Zoology; D. B. Macurda, Univer- worth County supported marine studies at Coiba sity of Michigan; W. M. Rand, Massachusetts Island, Panama. In 1971 the U.S.S. Terre Bonne Institute of Technology (MIT); T. L. Chorba, Parish carried a research team to Bocas del Toro, Jeffrey B. Graham, Smithsonian Tropical Research Institute, Panama, for three days of field work. In the sum- P. O. Box 2072, Balboa, Canal Zone. mer of 1973, while the volume on Malpelo was SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY being compiled, the fifth and most ambitious ex- Island Biology and Its Biogeographical pedition was undertaken when the U.S.S. Spartan- Significance burg County took 25 scientists to Cocos Island, Costa Rica, 550 miles southwest of Panama. The biological investigation of islands has These expeditions have been fruitful. The larg- made a valuable contribution to the studies of est and best developed coral reef formations to be evolution, biogeography, and ecology. Islands, by found thus far in the eastern tropical Pacific were virtue of their differences in size, shape, climatic discovered and subsequently studied by Peter conditions, and distances from continents, present Glynn and colleagues during the Gulf of Chiriqui a stepped progression of natural experimental trip (Glynn, Stewart, and McCosker, 1972). Previ- areas (MacArthur, 1972). Darwin's ideas on the ous to these investigations it was thought that no natural variability and selection of species devel- well-developed coral reefs occurred in the eastern oped in part through his observations at the Gala- tropical Pacific. The coral-eating starfish Acanth- pagos Islands. Some biologists investigate islands aster planci was found in great abundance and to test different aspects of modern evolutionary several species of Indo-Pacific corals were also dis- theory. Islands, however, are intrinsically inter- covered by Glynn and his co-workers during the esting because each is unique, and they afford bio- same expedition. Our understanding of the faunal logists the chance to study the effects of isolation composition and zoogeography of eastern tropical on the resident species. Island populations may Pacific fishes has benefited significantly from Navy- have greatly expanded or otherwise modified supported research in this region (Rosenblatt, niches and may have decreased levels of genetic McCosker and RubinofF, 1972). variability or, if isolated for a long period, endemic The use of Navy ships is a valuable addition to species may be formed (MacArthur and Wilson, the research capacity of the Smithsonian Tropical 1967). Research Institute. Navy vessels allow many scien- Malpelo Island (3°51'07"N, 81°35'40"W) is one tists to participate in the expeditions and it is of several oceanic islands in the eastern tropical possible to visit localities beyond the range of small Pacific. It is the only island on the Malpelo Ridge research vessels. Needless to say, costs and the (Figure 1) a solitary volcanic submarine ridge that logistical support problems are also vastly reduced. extends in a northeast-southwest direction with a ACKNOWLEDGMENTS.—It is a pleasure to acknow- length of 150 miles and a width of 50 miles (Chase, ledge Admiral C. H. Griffiths, Commander of the 1968). An oceanic island, Malpelo occurs outside United States Navy, Fifteenth Naval District, Fort the continental shelf and it is separated from the Amador, Canal Zone, and his staff who made mainland of Central America by depths greater arrangements for the Malpelo Expedition. We also than 1800 fathoms. Malpelo has never been con- thank Captain Lauren Seeber and the officers and nected, even by shallow water, with any other men of the U.S.S. York County for supporting our islands or the mainland. investigations. We express our appreciation to the Malpelo's distance from the mainland and the Foreign Ministry of the Republic of Colombia for depths of water between the island and the main- permitting us to visit Malpelo Island and to the land are significant barriers to the colonization of U.S. Army Southern Command, 352nd Aviation both terrestrial and shallow-water marine organ- Detachment, for providing an aerial reconnaissance isms. Colonization therefore must occur through of the island. I would like to thank my colleagues: sea surface currents which would transport rafting L. G. Abele of Florida State University, D. Dexter terrestrial organisms and the planktonic larvae and of San Diego State University, J. McCosker of the subadults of marine forms to the island. Malpelo California Academy of Sciences, and A. S. Rand, is closer to the mainland than Cocos and the Gala- I. RubinofF, and R. R. Warner, all of the Smith- pagos islands (Figure 2) and it is reasonable to sonian Tropical Research Institute, for their ad- assume it serves as a stepping stone for marine vice and assistance in the preparation of this vol- forms enroute, via oceanic currents, to these ume; I acknowledge the technical assistance of islands. I. Downs, A. Kourany, and V. Vergel of the Smith- Consideration of the complex pattern of surface sonian Tropical Research Institute. circulation in the eastern tropical Pacific Ocean NUMBER 176

(Wyrtki, 1965) provides insight as to how surface gyre that borders the mainland and encompasses currents may serve for transport from the mainland Cocos and Malpelo islands is also formed at this to Malpelo and beyond to Cocos and the Gala- time. If winds blow strongly for long periods the pagos islands. Only general features of surface condition known as "El Nino" occurs. This results circulation in the area of these islands for the in the transport of a larger than normal amount months of January and July are presented (Figure of warm surface water, which displaces the Peru 2) and the reader is referred to Wyrtki's paper for current to the southwest away from South America details on monthly current vectors, velocities, and and the Galapagos Islands. During £1 Nino the the persistence of flow for the different areas in the Galapagos Islands may be completely bathed in eastern tropical Pacific. warm tropical waters. In the months from January to March the trade- The Equatorial Countercurrent (Figure 2) does winds cause a strong current flow (0.5 knots) out not penetrate very far into the eastern tropical of the Gulf of Panama towards Malpelo and the Pacific from January to April, but from May to Galapagos islands (Figure 2). A small anticyclonic December this current is strong and reaches the

FIGURE 1.—Submarine topography of the central eastern tropical Pacific Ocean in the vicinity of Malpelo, Cocos, and the Galapagos islands. The lines are 600, 1000, and 1200 fathoms. Data are from Chase (1968). SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY 20*

10* NORTH EQUATORIAL CURRENT

EQUATORIAL COUNTERCURRENT t t C

SOUTH EQUATORIAL CURRENT

120' 110 IOO« 90' 80* 70* 20*

NORTH EQUATORIAL CURRENT

EQUATORIAL COUNTERCURRENT

SOUTH EQUATORIAL CURRENT

10 < FIGURE 2.—Surface currents around the major oceanic island in the eastern tropical Pacific Ocean. The islands indicated are Clipperton (CI), Cocos (C), Galapagos (G), and Malpelo (M). Data are from Wyrtki (1965). NUMBER 176 coast of Central America where it divides. The that lichens and mosses constitute the important major part of the Equatorial Countercurrent lower trophic levels for the land community. flows northwest along Central America, while the Malpelo's endemic lizards were well studied on remainder flows south and then turns west as part this expedition. A new species of gecko (Phyllo- of the South Equatorial Current. From May to dactylns transversalis) was discovered and this is December Clipperton, Cocos, and Malpelo islands described by Huey. This species, while similar in lie in the swath of the Equatorial Countercurrent, many ways to Peruvian species has its closest affinity and the Galapagos Islands occasionally receive to Mexican Phyllodactylus. The depth of biologi- water from the southern part of this current from cal knowledge about mainland and island Anolis, August to October. The combination of strong together with their extensive research experience southern surface flow in the months from January with anoles enabled Rand, Gorman, and Rand to to May and the circulation pattern of the Equato- ask specific questions about the ecology and be- rial Countercurrent probably account for the dis- havior of Malpelo's A. agassizi. The Malpelo anole persal of mainland marine organisms to these is larger than most other anoles isolated on islands islands. and it reproduces seasonally. Its social behavior is Another phenomenon that is of special interest strikingly different from that of other Anolis; it to marine zoogeographers is the presence and dis- is not territorial, there is little aggression, and its persal patterns of Indo-Pacific organisms (corals, displays are few and simple. Anolis agassizi is an mollusks, crustaceans, and fishes) at the eastern active, curious lizard that is readily attracted to tropical Pacific islands and on the Central Ameri- things that appear in its habitat: scientists, their can mainland (Costa Rica and western Panama). equipment, and their fruit and candy as well. This has been recently reviewed for fishes by Their curiosity and the remarkable tolerance of Rosenblatt, McCosker, and Rubinoff (1972). Lar- individuals for each other, these authors believe, vae of these animals are apparently transported enables A. agassizi to take advantage of resources across 1000 miles of open sea by the Equatorial such as water and food that have variable distribu- Countercurrent, which, as was shown above, ex- tions in time and space on Malpelo. Kiester de- tends to the mainland of Central America and scribes Diploglossus millepunctatus as curious sweeps across some oceanic islands for long periods and opportunistic and draws basically the same of the year. It is not presently known how many conclusions about the ecological significance of species have successfully completed this transit or this behavior as were made for A. agassizi. Interest- whether they more frequently colonize oceanic is- ingly, Kiester postulates that D. millepunctatus, lands or the mainland. because of its dependence on birds, may have its reproductive seasonality synchronized with that of the nesting seabirds. This awaits further study. Summary of the Malpelo Expedition Webster measured a low degree of genie variability The amount of information that was accumu- in A. agassizi. From what is presently known this lated in the four full days of field work on Malpelo seems to be a characteristic of old, solitary anole is gratifying. Part of the expedition's success lay in populations. Stamm and Gorman found that the what was learned about the island by its earlier karyotypes of A. agassizi and D. millepunctatus explorers. Gorman and Chorba have traced the are similar to the more primitive groups in these history of Malpelo's exploration, which goes back two genera. Taken together these observations to 1542. Early visitors to Malpelo had the impres- might suggest that Malpelo's Anolis and Diplo- sion that the flow of energy to the terrestrial eco- glossus have been in residence for a very long time. system had to be sustained by the sea and that this But as Webster and Stamm and Gorman point out, was done mainly by birds. While acknowledging our present knowledge is too limited to say this that the success of the land crab (Gecarcinus) and with certainty. the large lizard (Diploglossus) definitely depends Kiester and Hoffman surveyed the island and, on the activity of birds, Wolda has found that the with special climbing equipment, scaled its sum- terrestrial community is much too large and com- mits. Sightings from the peaks and from the bridge plex for this to be the only pathway. He concludes of the York County enabled them to determine SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 3.—Left: Aerial view (4000^ of Malpelo Island. The northern end of the island is in the foreground. Right: A York County landing craft approaching the west side of Malpelo Island. that the highest elevations of Malpelo are almost the discovery of the new Phyllodactylus species at 125 meters higher than early maps indicated. Stead Malpelo may have traveled faster than any other conducted a geological survey of Malpelo. His sjiecies discovery in history as it was flashed the description of the island's weathering, together same day to the ship's home port at Little Creek, with a compilation of the York County's depth Virginia. Sorting, preservation, and labeling con- soundings around Malpelo enables us to appre- tinued past dark while at the same time, in other ciate the original size of this rock. Once 8 to 10 parts of the ship, other men worked, and scuba times its present size, Malpelo has been steadily tanks, thermos bottles, and lunch containers were worn by the persistent pounding of the sea. When refilled for use the next day. When work was we brought our small boats close to shore and finished for the day, showers, warm food, a recount- tried to land on the island we keenly sensed the ing of the day's adventures, a soft bed, and last, intensity of the struggle between Malpelo and the but by no means least, the evening movie fully sea (Figure 3). restored us for the next day. Rough water and the island's steep slope made Investigation of Malpelo's marine habitat was getting from the boat to shore and returning a also rewarding. The water was so rough, however, major undertaking for the terrestrial party. Ex- that the small boats we had brought for the diving ploration ended prematurely for one scientist in teams were unsafe and diving operations had to this group who severly injured his arm while trying be staged from one of the York County's boats. We to get from the island into the boat. In spite of initially tried to use the ship's landing crafts but these obstacles none in the terrestrial group elected these had so much freeboard that they were not to spend their nights on Malpelo and, at days end, practical. After struggling through a day's diving all were glad to return to the haven of the York from the landing craft, I related this difficulty to County. The warship took on a curious appearance Captain Seeber, who immediately offered us the at this time of day. Its main deck was littered with use of his launch. Thus it came to be that the biological specimens and collecting equipment and elegantly furnished Captain's Gig became our was crowded with seamen who were rapidly trans- diving boat. No doubt the launch has seen better formed into eager assistants, audiences, or both. days both before and since we loaded it with divers, The crewmen relished this and took great pleasure their gear, and specimens. But it proved an invalu- in hearing about the discoveries and adventures able asset at Malpelo. There were many sharks of the day. Accounts were passed rapidly to the around the island; they were seen at the surface sailors on watch and expedition progress reports and on every dive. No particularly dangerous en- were included in the ships daily bulletin. News of counters occurred with sharks during our dives, NUMBER 176 but their omnipresence was sobering and at times thought to be endemic at Cocos or the Galapagos their proximity was slightly unnerving. islands or were known only from these two locali- Birkeland and colleagues completed their ambi- ties. These findings strongly suggest that Melpe- tious project of studying and characterizing the lo is a stepping-stone for marine mainland subtidal benthic communities at Malpelo. The species enroute to more distant islands and, as the smooth vertical subtidal walls are dominated by surface current patterns indicate, for Indo-Pacific large, sparsely distributed barnacles. Recruitment species as well. Some of the terrestrial invertebrates success for transforming barnacle larvae is low but listed by Wolda appear to be most closely related once they are established, adults may live for a to Indo-Pacific groups, which prompts the con- long time. On the gently sloping southeastern edge clusion that these were rafted from the old world of Malpelo, Birkeland's group found extensive to Malpelo via the Equatorial Countercurrent. coral development. Although they do not consti- The exchange of species between Malpelo, Cocos, tute a true coral reef, these corals are luxuriant and the Galapagos Islands is implied by their simi- and depth zonation is marked for different species. lar fish and decapod faunas and by the marked The occurrence of corals to as deep as 30 meters, faunal similarity observed by Birkeland et al. be- about the deepest known in the eastern tropical tween Malpelo and the Galapago Islands. Further Pacific, is attributable to the clear water around work with these faunas will clarify these relation- Malpelo. A list of all the marine invertebrates ships. known from Malpelo is appended to the Birkeland Although much was learned at Malpelo the et al. paper. Abele records 43 species of decapod Crustacea (18 species of macrurans) at Malpelo short time for study has left questions and the biota and describes two new species of snapping shrimp. of the island would doubtlessly be better under- Abele lists the distribution of each species and, by stood if more sampling, observation, and experi- drawing upon his other recently completed studies, mentation could be conducted. It was agreed that he makes an interesting observation about the additional visits should be made, ideally at dif- species diversity of decapod crustaceans at Malpelo, ferent times of the year so that the possible influ- Cocos Island, and several other eastern tropical ence of seasons could be assesed. That there is Pacific localities. The diversity of starfish at Mal- much to learn is evident when it is pointed out pelo is low and Downey lists only six species in that we still do not know how much rain annually her paper. Three species occur throughout the falls on Malpelo. Murphy (1945) suggested that eastern tropical Pacific and a new species of Tam- the island never receives rain, but Mapelo does aria and a new subspecies of Narcissia are de- lie within the range of the convergence of the trade scribed for Malpelo. Until Downey's work, one winds (Intertropical Convergence Zone, Wyrtki, Malpelo starfish (Leiaster) was previously known 1965), which suggests a seasonal rainfall. The is- to occur only at the Hawaiian Islands. land does have weathered, phosphatized rock Our knowledge of the Malpelo fish fauna is still (McConnell, 1943), and seepage marks, dry incomplete but McCosker and Rosenblatt report stream beds, and pools of water were found by the that 70 species of fish are now known for the island. Smithsonian-Navy expedition. All of these suggest In the specimens collected in the Smithsonian- that rain is not uncommon but expeditions to the Navy Expedition, McCosker and Rosenblatt found island have practically all been during the months several new species, as well as a new genus of fish; of the dry season (December to March) and we all awaiting description except the new species can only speculate about amounts of rain in the of goby (Chriolepis) described by Findley. The rest of the year. Abele, and McCosker and Rosenblatt papers dem- This volume, by reviewing and compiling Mal- onstrate several interesting features that the deca- pelo's literature and by reporting results to date pod crustaceans and fishes of Malpelo have in of the most thorough investigation of the island common. The majority of species in both groups sets the stage for future exploration. It describes are Central American mainland forms; there are Malpelo but it cannot convey the impressive also Indo-Pacific species and some endemics. Both scenario of that island. When viewed objectively groups contain species that were previously it is an isolated, inhospitable island, which by these 8 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY virtues has remained a natural laboratory for the MacArthur, R. H. study of evolution. 1972. Geographical Ecology: Patterns in the Distribution When seen first hand, however, Malpelo is a of Species. New York: Harper and Row. large grotesque rock that rises rudely, ominously MacArthur, R. H., and E. O. Wilson out of a monotonous sea. As we contemplated this 1967. The Theory of Island Biogeography. Princeton: rock we reminded ourselves, and took pleasure in Princeton University Press. the fact, that it was a sight that few people have McConnell, D. 1943. Phosphatization at Malpelo Island, Colombia. Bul- ever seen. letin of the Geological Society of America, 54:707- 716, 2 plates. Murphy, R. C. Literature Cited 1945. Island Contrasts. Natural History, 15:14-23. Chase, R. E. Rosenblatt, R. H., J. E. McCosker, and I. Rubinoff 1968. Sea Floor Topography of the Central Eastern 1972. Indo-West Pacific Fishes from the Gulf of Chiriqui, Pacific Ocean. US. Bureau of Commercial Fisheries, Panama. Contributions in Science of the Los An- circular 291, 33 pages. geles County Museum, 234:1-18, 3 figures. Glynn, P. W., R. H. Stewart, and J. E. McCosker Wyrtki, K. 1972. Pacific Coral Reefs of Panama: Structure, Distribu- 1965. Surface Currents of the Eastern Tropical Pacific tion, and Predators. Sonderdruck aus der Geolo- Ocean. Bulletin of the Inter-American Tropical gischen Rundschau, 61:483-519, 15 figures. Tuna Commission, 9:271-304. Terrestrial Biology of Malpelo Island: A Historical Review

George C. Gorman and Terence L. Chorba

and to attempt to understand the flow of energy ABSTRACT in the system. The expedition will have been suc- cessful if, for no other reason, others are inspired Records of the exploration of Malpelo Island date to visit this unique massive rock to pursue further back to 1542. Early biological investigations mainly some of the interesting problems raised in these concerned the terrestrial fauna, particularly the island's lizards. The complete list of Malpelo's pages. avifauna is presented. The etymology of Malpelo The first record of a landing on Malpelo that is discussed. we know of took place in the late 18th century. The Spanish Commodore, Alejandro Malaspina (in Malaspina and Bustamente, 1885:551) quot- That the biology of Malpelo would be poorly ing one of his officers wrote: known should surprise few people. The island One of the shipmasters in the local trade, favoured by fine couples geographic isolation with hazardous acces- weather and a smooth sea, once landed on a shelf of rock on sibility. It also lacks spectacularly unique denizens the north face of the island, and after climbing thirty steps such as Komodo dragons or giant tortoises. Thus down hewn out by hand, came upon a large pool of rain- interest in Malpelo is primarily restricted to the water, which was not overclean and had some bird feathers dedicated naturalist who can appreciate that the in it; but .... he used the opportunity to fill some casks. very isolation makes it a laboratory of evolution, This .... is all that can be said about Malpelo. and who realizes that a species need not be the At the time there was more that could have been largest or the strangest to be of genuine biological said, but Malapsina either was not familiar with interest. The joint Smithsonian-U.S. Navy Expedi- any previous mention of the island by the chroni- tion to Malpelo was not the first to visit this unique clers, or did not consider what they said to be of corner of the world; others have stopped on the sufficient importance. island, perhaps while enroute to the Galapagos, or Pedro de Cieza de L£on (1881:92) wrote that while collecting marine life off the coast of South the Spanish colonial administrator in Peru, Cristo- America. Our expedition probably was the first to bal Vaca de Castro, en route from Panama to be dedicated to the Malpelo ecosystem per se, to Buenaventura in 1542 "came to a rocky islet which attempt to collect and identify as many of the the sailors call Mai Pelo." living forms as was feasible, to examine the be- Literally, Malpelo means "bad hair." The origin havior and ecology of at least some of the species, of this name is unknown but the island appears on a parchment map of the world of 1550 (Desceliers, George C. Gorman, Department of Biology, University of 1550) that is in the British Museum. The name on California, Los Angeles, California 90024. Terence L. Chorba, student in Final Honour School of Psychology, Philosophy, the map is "ye mallabry." The term "malabrigo" and Physiology, Oxford University, and member of Brasenose was frequently used by cartographers to designate College, Oxford, England. islands and bays. Literally, "malabrigo" means 10 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY shelterless. Such a name would be well deserved. which Slevin (1928) named Celestus hancocki. He Another possibility, however, is that the word claimed that this lizard was most closely related to "Malpelo" has roots in the Latin malveolns (liter- Celestus millepunctatus O'Shaughnessy known ally "inhospitable" or "spiteful"), which in the from a single specimen of unknown provenance, vernacular could have degenerated to "malbolo" but that it differed from C. millepunctatus in and at some point been interpreted as "malpelo." several features of scutellation and limb develop- At the time that Vaca de Castro sailed past ment. Malpelo, though perhaps unknown to the Crown, Slevin did not explore the island. All his speci- the island was a territory of Spain by virtue of the mens of Celestus were taken within a few feet of Papal Bull issued in 1493 by Pope Alexander VI, the water's edge on a rocky ledge. When wounded, and the Treaty of Tordesillas (1494) under which he claimed that they would take to the water. Spain and Portugal divided the non-Christian Stomach contents consisted of feathers and remains world in half. Subsequently, Malpelo passed auto- of crabs. matically to Peru and later to Colombia without The George Vanderbilt South Pacific Expedition dispute. Inaccessible, uninhabitable, lacking arable of 1937 stopped at Malpelo in February. Some field land and suitable anchorage, it is not the kind of notes by the Vanderbilt's collector Ronald Smith real estate for which one battles furiously. were published by Fowler (1938). One of the first scientists to discuss life on Mal- Smith was confused in his , since he pelo was C. H. Townsend (1895), who visited believed that there were five species of lizards the island in March 1891 on board the U.S. Fish roaming about. But from his descriptions it is ob- Commission Steamer Albatross. During the brief vious that there were only the two known forms. visit, Townsend collected a Swallow-tailed Gull Sexual dimorphism, and perhaps distinctive size (Creagrus furcattis) and noted that seabirds of classes led to his erroneous conclusion. Since we several species swarmed about the inaccessible can easily determine when Smith was referring to summit. Anolis, and when to Celestus, we might mention Townsend also collected the first with a few of his ecological observations. He claimed the locality data "Malpelo Island." These lizards that the Anolis was found throughout the island, were described as Anolis agassizi (Iguanidae) by all the way to the summits, but that the Celestus Stejneger (1900:163), who reported: was limited to the height of 150 meters. He also reported a habitat difference. The anoles were Mr. Charles H. Townsend, who collected these specimens in Malpelo, informs me that they were running over the rocks seen mostly amongst loose rocks and exposed sur- near the water. The island was too steep to afford a landing, faces, whereas the Celestus were found primarily but the lizards were shot off or whisked off the face of the in small cave-like formations. The lizards were cliffs, thus falling into the water, whence they were secured described as fearless and quite bold. by the collector. The Vanderbilt expedition collected four speci- Two years prior to the publication of Town- mens of Anolis and seven Celestus, and this served send's report, Faxon (1893) described the common as impetus for Dunn (1939) to reconsider the land crab of Malpelo as Gecarcinus malpilensis lizard fauna of the island. He made two major (Gecarcinidae) from a single male specimen col- decisions at the generic level, one that has stood lected by the same expedition in which Townsend the test of time, the other that has not. Dunn collected the lizards. No notes other than a descrip- pointed out that the genus Celestus could not be tion were provided. maintained as distinct from Diploglossus. He as- In November and December of 1927, Captain signed the large anguid to Diploglossus, the generic A. G. Hancock made a trip to the Galapagos. J. R. name that it still bears. More interestingly, he Slevin of the California Academy of Sciences was demonstrated that the previously described C. mil- along, and when the boat passed near Malpelo, lepunctatus was not distinct from Slevin's C. han- Slevin and a colleague were able to effect a landing cocki. Thus the specimen described by O'Shaugh- from a small skiff on the northeast corner of the nessy in the mid-nineteenth century was almost island. They collected 27 Anolis (a medium-sized certainly obtained from Malpelo, although no re- lizard) and 10 specimens of a large anguid lizard cords of an expedition to the island have been NUMBER 176 11 found. The second decision was to raise the anole bird list of Malpelo (Red Billed Tropic Bird, to a distinct genus (Mariguana). This decision Phaethon aethereus); examined the stomach con- has not been accepted (Etheridge, 1960). tents of a gull and found that it was eating Bond and deSchauensee (1938) presented an marine crabs; and, most important, noted the pres- annotated checklist of the birds of Malpelo based ence of a diversity of invertebrates including upon the observations and collections of the Van- spiders, pseudoscorpions, beetles, flies, ants, and derbilt expedition and incorporating data from other insects. Samples of Malpelo rocks were ob- Townsend (1895). It is reprinted below in its tained by Murphy and a study of phosphatization entirety. at Malpelo was published by McConnell (1943). The land crabs (Gecarcinns) impressed Murphy ANNOTATED LIST OF THE BIRDS (1945:16) who described them as OF MALPELO ISLAND ghoulish in appearance . . . , fat and bloated looking crea- 1. Sula dactylatra granti Rothschild. (Sulidae, Pelicani- tures with shells of a ghostly white. They may be quite formes) Masked Booby. Two adult males were collected harmless toward human beings, yet they seem to eye a visitor February 8, and an immature female February 9. with an intent that grows uncomfortable. If you sit long 2. Fregata sp. (Fregatidae, Pelicaniformes). None collected enough the crabs will move up closer and closer, as if with but Smith notes: "Man o'War Birds 50; small colony on whetted appetites, and I have a feeling that they would one of the outlying rocks on the southwest of the island." make life miserable for anyone who had to sleep ashore. 3. Falco peregrinus anatum Bonaparte. (Falconidae, Fal- coniformes) Duck Hawk. Not collected but two birds Garth (1948) provided important information observed twice by Smith on Malpelo Island. 4. Actitis macularia (Linnaeus). (Scolopacidae, Charadrii- on this crab. He did not agree with Faxon (1893) formes) Spotted Sandpiper. Not collected but "ten seen that it was an endemic species, but considered it in company with tattlers." G. planatus Stimpson, a species widespread from 5. Heteroscelus incanus (Gmelin). (Scolopacidae, Charadrii- Baja California to Acapulco, that has successfully formes) Wandering Tattler. One male and two females, colonized the and Clipper- in winter plumage, were collected February 8 and 9. ton, but that is notably absent from the Galapagos. About twenty-five individuals were seen. 6. Creagrus furcatus (Neboux). (Laridae, Charadriiformes) There have been additional landings on Mal- Swallow-tailed Gull. An adult male and female as well pelo, some documented by specimens. For example, as a nestling were collected February 8 and 9. Smith the San Diego Society of Natural History has records a colony of thirty birds on the island. lizards collected in December 1931 and January 7. Anous stolidus ridgwayi Anthony. (Laridae, Charadrii- formes) Noddy. An adult female, the only specimen 1933, by Cyrus Perkins. There is, however, little seen, was secured February 9. It was not in breeding additional published information about the ecol- condition. ogy and natural history of the island or its inhab- 8. Anous minutus diamesus (Heller and Snodgrass). (Lari- itants. What inspired the present expedition? Why dae, Charadriiformes) Black Noddy. An adult male col- travel hundreds of miles to visit a barren rock? lected February 9, was the only specimen seen. The bird The very harshness of the environment raises inter- was not in breeding condition. 9. Hirundo rustica erthrogaster Boddaert. (Hirundinidae, esting questions. In addition to sea birds, and a Passeriformes) Barn Swallow. An immature female was few nonmarine birds that are very strong flyers but secured February 8. Four birds were seen flying out of presumably not residents, only three common ter- a cave. restrial residents, Anolis and Diploglossus (lizards) 10. Progne sp. Martin. (Hirundinidae, Passeriformes). Re- and Gecarcinus (a crab) had been described and corded by Townsend but none secured. Smith did not see identified. What species of plants (if any) and any martins. invertebrates would be found? What are their bio- During March 1941, the schooner Askoy ap- geographic affinities? How does such a community proached Malpelo and Dr. Robert Cushman maintain itself? Murphy and another member of his party made a Our working assumption was that this was a ter- landing. His popular article (Murphy, 1945) pro- restrial ecosystem dependent upon primary pro- vides the best summary of the natural history of duction on the sea and not on the land itself. the island. He estimated the booby population at Presumably the birds which fed upon marine life about 25,000; added an additional species to the provided food resources for small terrestrial inver- 12 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY tebrates (parasites, carrion eaters, egg predators, Etheridge, R. E. etc.) which in turn provided sustenance for the 1960. The Relationships of the Anoles (Reptilia: Sauria: Iguanidae): An Interpretation Based on Skeletal lizards. Morphology. Unpublished Ph.D. Dissertation, Uni- Several of us were specifically interested in the versity of Michigan. Ann Arbor: University Micro- Anolis lizards. The genus is comprised largely of films Inc. tropical arboreal species. They tend to be highly Faxon, W. territorial, defending fixed perch sites with elabo- 1893. Preliminary Descriptions of New Species of Crus- rate displays. Here was an environment devoid of tacea. Part 6 in Reports on the Dredging Opera- tions off the West Coast of Central America to the the typical plant cover that we associate with Galapagos, to the West Coast of Mexico, and in Anolis. Did they have any special physiological the Gulf of California, in Charge of Alexander adaptations that permitted life on warm, exposed Agassiz, by the U.S. Fish Commissioner Steamer rocks? Was there any modification of the social "Albatross," during 1891, Lieutenant Commander structure of populations in the absence of typical Z. L. Tanner, U.S.N., Commanding. Bulletin of the Museum of Comparatix>e Zoology at Harvard Col- perch sites that are associated with their terri- lege, 24:149-220. toriality? Fowler, H. W. These are the kinds of questions that were on 1938. The Fishes of the George Vanderbilt South Pacific our minds when we first discerned the form of Expedition, 1937. Monographs of the Academy of Malpelo Island, gray itself against the gray of the Natural Sciences of Philadelphia, 2: V+1-349 pages, 10 plates. [Malpelo, pages 5-6.] early dawn on the sea on 29 February 1972. Ans- Garth, J. S. wers are what we sought when we braved the steep 1948. The Brachyura of the "Askoy" Expedition with swells and jumped off the landing craft to begin Remarks on Carcinological Collecting in the Pan- serious exploration of the island and its life. ama Bight. Bulletin of the American Museum of Natural History, 92(l):l-66, 4 figures. Malaspina, A., and J. de Bustamante Literature Cited 1885. Viaje politico—Cientifico alrededor del Mundo por las corbietas "Descubierta" y "Atrevida": desde 1789 a 1794. Madrid: Publicado por Dom Pedro Bond, J., and R. M. deShauensee de Nova y Colson. 1938. Zoological Results of the George Vanderbilt South McConnell, D. Pacific Expedition of 1937, Part II: The Birds of 1943. Phosphatization at Malpelo Island, Colombia. Bul- Malpelo Island, Colombia. Proceedings of the Acad- letin of the Geological Society of America, 54:707- emy of Natural Science of Philadelphia, 90:155-157, 716. plates 9-11. Murphy, R. C. Cieza de Leon, Pedro de 1945. Island Contrasts. Natural History, 54:14-23. 1881. Guerras Civiles de Peru: Guerra de Chupas. Ma- Slevin, J. R. drid: Publicado por Dom Pedro de Novo y Colson. 1928. Description of a New Species of Lizard from Mal- [English translation by Sir Clements Markham and pelo Island. Proceedings of the California Academy reprinted by the Hakluyt Society, London, 1918.] of Science, fourth series, 16 (21):68I-684, plates 25, Desceliers, P. 26. Stejneger, L. 1550. Faicte a Arques . . . Ian 1550. Add. MSS 24065. 1900. Description of Two New Lizards of the Genus British Museum (Natural History), London. Anolis from Cocos and Malpelo Islands. Bulletin Dunn, E. R. of the Museum of Comparative Zoology, 36 (6): 1939. Zoological Results of the George Vanderbilt South 161-164, 1 plate. Pacific Expedition of 1937, Part III: The Lizards Townsend, C. H. of Malpelo Island, Colombia. Notulae Naturae of 1895. Birds From Cocos and Malpelo Islands, with Notes the Academy of Natural Sciences of Philadelphia, on Petrels Obtained at Sea. Bulletin of the Museum 4:1-3. of Comparative Zoology, 27:121-126, 1 plate. Reconnaissance and Mapping of Malpelo Island

A. Ross Kiester and Jeffrey A. Hoffman

ported reaching the summit of the rock, but had ABSTRACT not indicated which of the two summits they con- sidered this to be. Indeed very little general Daily observations of the exploration party are re- description of the topography of the island existed ported. Early maps of Malpelo show it to have one in the literature, and no mention was made of the peak with a maximum elevation of about 300 m. The exploration party surveyed and mapped the existence of the two peaks. Because of this con- island and found that the island has twin peaks, fusion, a general survey of the topography of the which are both approximately the same height and island was one of the expedition's scientific about 150 m higher than previous maps indicate. objectives. An improved map of Malpelo is given. It is be- lieved that the Smithsonian exploration team is In pursuit of this objective, two members of the the first to reach the island's summit. terrestrial party of the expedition extensively ex- plored the entire island, while other members of the party carried on intensive studies at one site. In this report we give a general description of the Introduction island as obtained by the survey party, and discuss An overflight of Malpelo Island, arranged the mapping of the island. through the courtesy of the U. S. Army, in prepa- ration for the U. S. Navy-Smithsonian Expedition Mapping Technique to Malpelo showed serious defects in the official map based on a 1954 aerial survey. The published As the party traversed the island, measurements U. S. Navy Hydrographic Office map (No. 1685, were made of the horizontal and vertical angles North Pacific Ocean, Isla del Coco, Malpelo Island between the various summits and landmarks. All inset) showed a single major peak of 260 meters measurements were made with a marine sextant (845 feet) on the island. Observers in the overflight and a Brunton compass. Further surveying was reported two distinct peaks of nearly the same done from the ship. Sextant measurement of the height, read as roughly 390 meters (1200 feet) on angles between the various features of the island the plane's altimeter. Initial observations from the can give the height of these if the distance to the ship on arrival at Malpelo did not reveal which of island is known. This method, however, limited the peaks was higher. They were so close in height the survey to the various peaks and ridges visible that whichever peak was closest to the ship ap- in profile from the ship. An accurate knowledge peared the highest. A previous expedition had re- of the location of each feature on the island was necessary for the reduction of sextant sights. While A. Ross Kiester, Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138. Jeffrey A. Hoff- the ship was circling the island, azimuth bearings man, Smithsonian Astrophysical Observatory, Harvard Uni- were taken with a pelorus at regular intervals on versity, Cambridge, Massachusetts 02138. all of the main features, including the north and

13 14 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY south tips of the island. An attempt to use gun- the southern end. Beside the peaks the most con- ranging radar to fix the distance from the island spicuous feature of the island is a large arm ex- was frustrated by an uncertainty in determining tending out on the west side of the island and form- the radar reflection point on the island. Lacking ing a small bay (Figure 4). The bay is narrow and this distance measurement, the position of the ship the cliffs surrounding it are very sheer, giving a was fixed by the intersection of the azimuth bear- fjord-like appearance. Several sea caves and arches ings to the tips of the island and to the "South were also seen on the circuit. Besides the main Peak," which position was shown on the old map. island itself, there are several other rocks and sea Further surveying had to accept the island outline stacks in the vicinity. The two largest groups of and scale of the published Hydrographic Office these rocks are clustered to the north and to the map. But these are less likely to be in error than south of the island. the photographically determined contour lines. A total of three days were spent exploring the The intersections of the azimuth bearings to each main island and one day was spent exploring the of the features from the various ship positions lo- largest of the offshore rocks. cated each feature on the map. Heights were After the circuit of the island, a landing site determined from vertical sextant measurements of was chosen on the south side of the end of the spur the angles between the features and the waterline, which forms the fjord-like bay on the west side of and horizontal angle measurements of the width the island. This site became the base for most of of the island. An azimuth bearing to the South the exploration and the location of most of the Peak combined with the horizontal angle fixed intensive studies of the terrestrial biology of the the ship's position during this phase of the survey- island. This landing and study site was designated ing. The linear heights of the features were then Landing Site 1 (Figure 4). After landing the en- obtained by solving a simple proportional equation tire party and its equipment, the exploration team and correcting for the height of the observer above started climbing up the spur. The vista up the the waterline of the ship.

Itinerary of the Exploration Party North Racks 5 Stools A preliminary idea of the topography of the island was gained the first morning by making a complete circuit of the island close into shore. This survey also provided an opportunity to pick out possible landing sites. The island is quite steep along the sides and has a rather flat-topped appear- ance. It has three distinct peaks (Figure 4). At the north end of the island is the smallest of the three which was termed "North Peak." This summit is sharply set off by nearly vertical walls all of the way around it. The top, however, appears to be relatively flat. This peak is separated from the next by an east-west fault. This next peak south is cy- lindrical and butte-like, and is much higher than the first. It appears to have equally as steep sides and is also flat on top. This peak was referred to as "Plug." To the south of it, separated by a huge east-west fault which split the entire island is a .""Shoolt more rounded peak which appeared to be of the same height as Plug. This summit was called "South Peak." From it the island slopes clown to FIGURE 4.—Reference map of Malpelo showing the various landing sites (1-4) mentioned in the text, the routes taken the south until it reaches the rather sheer cliffs at by the exploration party, and significant landmarks. NUMBER 176 15 spur toward the saddle between Plug and South top of the two peaks, as were the land crabs (Gecar- Peak was a rocky slope completely devoid of vege- cinus malpilensis). Blue-faced Boobies (Sula dacty- tation. Climbing up the spur the party found sev- latra) were also omnipresent. At the top of both eral caves of various sizes, including one which peaks a strong breeze was blowing continuously in went through the spur from one side to the other. contrast to the still conditions lower down on the The only vegetation that could be found was algae leeward side of the island. From Plug the party which was growing inside of the caves and crevices climbed back down to the saddle and thence down where some moisture was present. After some to the landing site. scrambling, the saddle was reached. Here a cairn On the second day the party landed on the with a bottle containing the names of several Co- northwest edge of the island (Landing Site 2, Fig- lombians from two mapping and survey expedi- ure 4). While two members of the party worked tions was found. In the saddle ttte lava was shot near the shore, two others climbed towards North through with holes of varying diameter, making Peak. Initially the peak was attempted by climbing movement slow. From here the party ascended to around the west side to its north face, which had the summit of South Peak. Once on the summit several large chimneys. Also in the area is a spec- the party had a view down the entire south end of tacular spire isolated from the main peak. The the island. This view corresponded exactly to the route was frustrated because the chimneys were photograph taken from the summit by the Vander- seen to contain formidable chock-stones. On the bilt Expedition of 1937 (Bond and deSchauensee, less steep slope to the northwest of the North 1938, plate 10B). Thus it was concluded that the Peak were some of the few patches of grass to be Vanderbilt Expedition had in fact climbed South found on the island. The largest of these patches Peak. was not more than several square meters in area. As sighted from South Peak, the Plug had an Retracing their steps the party then climbed to elevation of 1° to 1.5° horizontal, showing it to be below the saddle between Plug and North Peak. the true summit of the Malpelo rock massif. As Here there was a gully which led up to the saddle. surveyed from the ship, Plug was determined to In the gully several large pools of water were be approximately 6 meters higher than the South observed. One of these was perhaps one-half meter Peak. The root mean square errors for the mea- deep and was the largest body of standing water surements of the two peaks from various ship posi- seen on the island. The pools contained some tions are about equal to the difference in height aquatic insects which unfortunately could not be of the two peaks. The 1° to 1.5° elevation of the collected. Continuing up the gully into the fault northern peak as sighted from the southern corre- proved difficult and roped climbing was necessary sponds to a height difference of 6-9 meters, which to scale a 25-meter lava wall. From the saddle it agrees with the difference as determined by sextant was a relatively easy scramble up to North Peak. sights from the ship. Again both species of lizards and numbers of From South Peak the party returned to the sad- boobies were present all the way up to the summit, dle and climbed up to the base of Plug. From this and the wind was blowing steadily in contrast to point the climbing became more difficult. After the leeward side. From here the party climbed some searching, a chimney on the east side of Plug down along the east side of Plug and over the was found that afforded a relatively easy, but none- saddle south of Plug and down to Landing Site 1. theless technical (class 5), climb up to the top. On the last day of exploration of the main is- No bottles or cairns were found on this summit land, a party of two landed on the east side of the plateau. None of the previous parties climbing on island (Landing Site 3, Figure 4). This landing the island had mentioned the necessity of technical site was used by a Colombian expedition which rock climbing for the routes they used. Thus, the had placed a bronze plaque to mark the site. The Smithsonian party seems to have been the first to party made its way up to the saddle south of Plug reach the summit plateau of Plug. and then up to South Peak. From there they As with the southern summit, Plug was devoid climbed down to the southern end of the island. of vegetation. However, both species of lizards The area southwest of South Peak is relatively (Anolis and Diploglossus) were common up to the the flattest part of the island. The slope leading 16 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY from South Peak down to the end of the island numbers. No land crabs appeared to be on the was a scree slope with small rocks cemented by rock. Several small bushes and a fern were growing solidified lava. Lava bombs were strewn about at the top of the rock. The bushes frequently liberally, and all over the lava had solidified into showed large trunk masses with only small green cubist architectural forms. The south end of the shoots. This was suggestive of large seasonal varia- island was extremely dry. Few lizards were seen on tions in rainfall. the central dome of the south ridge, but both A Red-footed Booby (Sula sula) was sighted species as well as the land crab were abundant on perched in the middle of the bushes. This species the extreme south end. There was a small patch habitually perches in trees in contrast to the Blue- of grass on the southernmost cliff. From the south faced Booby, and has not previously been reported end the party hiked back up over South Peak and from Malpelo Island. Positive identification was then back down to Landing Site 1. made when one of the members of the party was On the fourth day, the party visited the largest able to climb down to within a few meters of the of the offshore rocks (Landing Site 4, Figure 4). bird before it flew, and secure several good photo- This rock is in the South Rocks group and like the graphs showing the characteristic red feet of this other rocks it is quite steep. A landing was difficult species. and roped climbing began from the tiny shelves As with the landing, departure from the rock near the water. Near the top of the rock (45m = was somewhat difficult due, in part, to the nearby 130 ft) the slope decreased appreciably. Rather shoals. surprisingly, a large patch of grass was found here. This was the largest and lushest stand of vegeta- Conclusion tion found anywhere. Anolis were common in the grass as well as on much of the top of the rock. The survey undertaken by the exploratory party Diploglossus were also found, but not in large resulted in the map shown in Figure 5. As men- tioned above, the outline of the island and the scale are taken from the old Hydrographic Office map, but the topography and contour lines are the result of the new survey. It should be borne in mind that the map may not be entirely accurate due to the rugged nature of the terrain, and to the limited time and equipment of the survey party. Future visitors to Malpelo should be able to make substantial improvements in the map, partic- ularly with regard to geological details. Any exten- sive on-site surveying, however, could not be accom- plished without a good deal of technical climbing. We hope that this map will be of some help to any future visitors. Finally we remark that the rather dry descrip- tion given here in no way does justice to the sus- tained surrealism of this unique and unusual island.

Literature Cited Bond, J., and R. M. deSchauensee 0 1938. Zoological Results of the George Vanderbilt South Pacific Expedition of 1937, Part II: The Birds of FIGURE 5.—A revised map of Malpelo showing summit alti- Malpelo Island, Colombia. Proceedings of the Acad- tudes and land contours based on the surveying of the STRI emy of Natural Science of Philadelphia, 90:155-157, exploration party in March 1972. plates 9-11. Field Observations on the Geology of Malpelo Island

Jeffrey A. Stead

Several types of igneous rocks are present on ABSTRACT Malpelo. Field identifications are as follows: da- cite, trachyte, tuff, basalt, and andesite. Up to an Malpelo Island had a volcanic origin. Depth sound- altitude of 210 to 240 meters, the island appears ings around the island indicate that it was once to be mostly trachyte, with lesser amounts of 8 to 10 times larger than its present size. Marine weathering has eroded the island and has formed dacite and tuff. This lower zone is intruded by at steep cliffs and sea caves along its sides. Several least two sets of dikes made up of basic rocks. The types of igneous rocks and minerals occur on the top of the island is covered by an andesite cap. island. The distribution of major joints and faults Field identification of minerals is as follows: is shown. Specimens of dacite collected near the landing site contained amygdules of quartz and apatite. One amygdule contained fine, hairlike, zeolite crystals Published accounts of the geology of Malpelo coated with hematite. Small amounts of chrysocolla Island include McConnell's (1943) analysis of were found in a joint at an altitude of about 240 three rocks from the island and the field observa- meters. Guano occurs in limited amounts on all tions of Murphy (1945, also in McConnell, 1943). parts of the island. The amygdaloid specimen assayed by McConnell Fresh water seepage was observed in several (1943) showed a small amount of phosphatization places. Red streaks of iron oxide are deposited on (replacement of natural minerals by phosphate the rock faces where seepage has occurred. Some compounds derived from guano) while this process of these streaks represent principal areas of runoff was very advanced in another rock (phosphate following rainstorms rather than permanent seep rock). The specimen of augite andesite was only age areas. slightly altered by phosphates. Murphy reported Steep cliffs around the perimeter are probably that guano was not very concentrated on the island maintained by the combined effects of wave action, and McConnell (1943) concluded that phosphati- which loosens and removes rock in low areas zation occurs mainly in porous areas of the island (Steers, 1953), rain, and alternate heating and (along faults and fractures) where guano could be cooling, which loosen the soft lavas (mostly carried and concentrated by rains. trachytes) above the normal splash zone until they Malpelo is volcanic in origin, the present island fall off along joints. The lava cap is a very resistant is the remnant of a much larger structure. Most of rock, and probably does not fall until the rock the island's perimeter consists of nearly vertical underneath it has been weathered away. rock faces which extend from 60 to 230 meters Sea caves (Figure 6) are formed by the scouring above sea level (Figure 6). action of waves on brecciated rocks around faults (Guilcher, 1958). There are many small faults on Jeffrey A. Stead, Defense Mapping Agency, Hydrographic Cen- ter, Rodman Office, Box 2023, Rodman, Canal Zone. Present Malpelo, most of which intersect the waterline at address: Hahnemann Medical College, Philadelphia, Pennsyl- large (60°-90°) angles. The largest cave was vania 19102. estimated to be 9 meters wide and 13 meters high

17 18 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

(above sea level) and to have a 10-fathorn water nent feature of the fjord, strikes 2O5°T and dips depth at its mouth. Some of the caves have a white 78°NW. The largest sea cave on the island is at formation on their ceilings which resembles flow- the apex of the fjord (Figure 7). stone. This is probably composed of insoluble sub- The U.S.S. York County made continuous depth stances from guano, washed down along the faults soundings around Malpedo Island. These sound- and redeposited on the ceilings of the caves. ings, plus those from a 1952 U.S. Navy sketch sur- Small amounts of soil, consisting of variable vey, are shown in Figure 7. mixtures of clay and feces, are found in Bottom depth falls off greatly to 50-60 fathoms sheltered areas of the island. Soil is found nowhere on all sides of the island, but the bottom gradient in large amounts, but increases in abundance with increases rapidly at 60-70 fathoms (Figure 7). increasing altitude, and rarely occurs below 90 The area out to the 60-fathom curve may represent meters. The absence of soil in the lower levels can a wave-cut plateau formed during the Pleistocene be attributed to steep slopes, rain, and to wave (Shepard, 1963). If this is assumed, the 60-fathom action which, during severe storms, can splash curve would indicate the minimum size of the origi- water to considerable heights (Johnson, 1919). nal volcano. A line has been drawn (Figure 7) to Two major sets of joints and faults were meas- indicate a depression in the sea floor. This may be ured in the fjord area (Figure 7). Their average a fault. Its trend, 218°T, is close enough to the measurements were as follows: strike 210°T, dip 210±T average of one of the major joint sets on 79°NW; strike 110°T, dip 82°SW. The eastern the island to strongly indicate a structural relation- wall of the fjord is the scarp of a major fault, repre- ship. The dashed line trends 136°T, and also may sented by a cliff over 60 meters high and half a indicate a fault, or it may mark the division be- kilometer long. This fault, which is the most promi- tween two previously existing cones.

FICI'RE 6.—The southwestern tip of Malpelo Island showing the lava cap, the steep cliffs, and narrow caves that occur at the watcilinc all around the island. NUMBER 176 19

FIGURE 7.—Map of Malpelo Island showing depth soundings from a U.S. Navy 1952 sketch survey (boldface numbers), those from the T.S.S. York County (lightface numbers) and depth profiles around the island. The location of two submarine faults (218°T and 136°T) are indicated. 20 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

ACKNOWLEDGMENTS.—I thank D. Meyer of the Johnson, D. W. Smithsonian Tropical Research Institute, B. Ma- 1919. Shore Processes and Shoreline Development. New curda of the University of Michigan, and R. Ste- York: John Wiley and Sons. wart of the Panama Canal Company, Engineering McConnell, D. Division, for reading this manuscript and offering 1943. Phosphatization at Malpelo Island, Colombia. Bul- suggestions. Macurda assisted me with field work letin of the Geological Society of America, 54:707- and Steward helped me interpret photographs of 716, 2 plates. the island. Murphy, R. C. 1945. Island Contrasts. Natural History, 15:14-23. Shepard, F. P. Literature Cited 1963. Submarine Geology. New York: Harper and Row, Guilcher, A. Inc. 1958. Coastal and Submarine Morphology. London: Steers, J. A. Methuen and Co., Ltd. 1953. The Sea Coast. London: Collins Clear-Type Press. The Ecosystem on Malpelo Island

Henk Wolda

the Pacific (Clapp, 1972) are known to contain a ABSTRACT variety of species and all kinds of animals are able to reach islands more remote than Malpelo (Carl- A rather high percentage of predators was found quist, 1965). Moreover, one would expect at least among the terrestrial invertebrates of Malpelo Is- one of the lizards (Anolis agassizi) to feed pre- land. Except for the birds and one of the lizards, almost all of the animals depend on a purely ter- dominantly on insects. Indeed, Dunn (1939) found restrial economy based on numerous algae, lichens, "two or three species of insects" in stomachs of and mosses. Anolis. Bond and deSchauensee (1938) report an abundance of "small black gnats or flies" and a chirping, as if by crickets. The most encouraging Studies on the biota of isolated islands and statement is that of Murphy (1945), who, in the groups of islands have produced exciting results, stomachs of Anolis, found "remains of small crabs, influencing and even initiating the theory of evo- spiders, pseudoscorpions, beetles, flies and fly lar- lution. Almost all existing islands have now been vae, ants and other insects." The problem to be surveyed, at least to some extent, but until 1972 Malpelo remained a notable exception. Except for >

21 22 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY investigated is what animals occur on the island gramma dealbata (Presl) Tryon. On the main and how do they make a living. The near-absence island, however, lower forms of plant life were of plant life reported in the literature suggested abundant. On vertical rock faces moist with fresh that the economy of Malpelo is entirely based on water and under ledges, algae, mosses, and lichens the birds bringing in food from the ocean, their could be found everywhere (Figures 10, 11). A few droppings, and dead bodies. pieces of rock were collected and eight species of In the first week of March 1972, I was able to lichens were identified from them: Caloplaca test this idea. My collections and observations were sp., Candelbria sp., ?A.spicilia sp., ?Diploschistes sp., made in localities indicated on Figure 8 and I Lecidca sp., Graphidaceae sp., fBasidea sp. and examined some material collected from the South Pyxine cf. glebosa Tuck. Considering the smallness Rock. of the sample, the number of sj>ecies present may The general impression one obtains from the be much larger than this. main island is that it is bare rock (Figure 9), with This vegetation is the basis for quite a complex no higher plants except for very few small patches community of animals. In crevices and, especially, of a yet unidentified grass. On the South Rock this under rocks, a multitude of animals was discovered. grass was fairly abundant and here also a shrub My collection probably does not cover all the occurred, possibly a legume, and a fern, Pityro- species present, but the fact that mine was about as good as the anoles (see below) suggests that I may not have done too badly. I did not make ex- tractions from soil under the rocks and any micro- arthropods living there would have been missed. The abundance and diversity of the invertebrates obtained (Table 1) is rather impressive. At least 17 species of insects were found, belonging to 8 orders. The "black flies" mentioned by Bond and deSchauensee (1938) represent a new genus of Chloropidae and the crickets they heard belong to the genus Hygronemobius, related to H. liura described from British Guyana (Gurney, pers. comm.). These two species were among the most abundant insects, together with a little black ant (Pheidole sp.) and staphylinid beetles. The two species of snails were fairly common in clusters under rocks at locality 1. There was an extra- ordinary variety of spiders, some eleven species of Araneida, none of which were webbuilders. Worms and isopods were abundant under rocks at locality 2, far above the intertidal zone. It is much too early to start speculating about the origin of the species found, but there are indi- cations that some of them may have come from the Indo-Pacific. After a preliminary investigation of the centipeds, Crabill (pers. comm.) suggests that the closest relatives of some of these may not be South American species. Representatives of the opilionid family Assamidae seem to be mostly living in the Indo-Malayan region (Brues, Melan- der, and Carpenter, 1954). The land crab and all FIGURE 9.—General view of site 1. three species of lizards (including the new gecko NUMBER 176 23 described by Huey, pp. 44-46 herein) are all found main island, plus five from the South Rock. These on Malpelo only, but how many of the other were kindly given to me by A. S. Rand. Virtually species are endemic is not yet known. all the items listed in Table 1 were observed, in- To depict the structure of the ecosystem on cluding some juvenile crabs. The only obvious ex- Malpelo, i.e., to construct a foodweb, roles had to ception were the earthworms, presumably because be assigned to each of the species concerned. The they are living in soil under rocks and thus inacces- classification of these species as a herbivore, a pred- sible. It is certainly not because they are distaste- ator, or a scavenger is based on what is known ful. When I removed the rocks, several anoles came about relatives living elsewhere and on some ob- in immediately and feasted on the worms. The servations. The classification of the staphylinids anoles from the South Rock also had a variety of as herbivores, for instance, was made because they arthropods in their guts, but differed from those were very abundant and found mostly among from the main island in that they had their rec- mosses. The beetle, Enochrus, is classified as a pred- tums filled with grass seed. ator, although the adults may be omnivorous. The Both individuals of the gecko, Phyllodactylus, results are included in Table 1 and the tentative had eaten crickets, a few ants (probably Strumi- foodweb is illustrated in Figure 12. The landcrab genys) and a lepidopteran larva, possibly Ereun- Gecarcinus malpilcnsis, was very common and was etis. The number of species per individual was frequently observed nibbling on stones, presum- much smaller than that observed in Anolis. Ants ably feeding on lichens and algae. Remains of in- (Pheidole sp.) and staphylinid beetles, both of sects were often found in their feces. Many of these which made up the bulk of the diet of anoles, were may have been scavenged or obtained while eating absent in the geckos. These differences could be the feces of lizards, but the crabs can also act as accidental, but are probably real, Anolis being predators. On one occasion I saw two crabs that diurnal and Phyllodactylus nocturnal. were holding a male Anolis, which was still very Nematodes, probably several species, were found much alive, although one of its hindlegs was al- in 60 percent of the anoles, averaging three per ready eaten. A third crab was eating the tail. individual (range 1 to 14). The geckos had 11 and These crabs also feed on dead birds. They are 12 nematodes, respectively. opportunistic omnivores, but probably mostly The large lizard, Diploglossus millepunctatus, scavengers. has a quite different way of life. The feces that I I had the opportunity to examine the gut con- examined did not contain any arthropods at all. tents of 38 individuals of Anolis agossizi from the They were observed feeding on fish regurgitated

"••&•

FIGURE 10.—Moist rock face and land crabs at site 1. Fici'RF. II.—Moist wall of empty creek bed at site 2. 24 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY by boobies (Sula dactylata granti), on fish leftovers species living on Malpelo, an unusually high per- after the boobies had fed their young, and on fresh centage are predators, many of which may be gen- bird droppings. Slevin (1928) and Dunn (1939) eralists rather than specialists. Two of the herbi- found crab remains in Diptoglossus' stomachs and vores, the snails, seem to be eaten exclusively by Garth (1948) suggested that this lizard feeds on the lizards, not by any of the invertebrate preda- both crabs and bird eggs. The latter may be true, tors. These predators are all eaten by Anolis, but but I have no evidence to support this idea. the bulk of their food consists of herbivores. More The foodweb (Figure 12) shows two interesting exhaustive collections, made at different times of features. The first relates to the rather large variety the year, are needed to determine whether indeed of predators, some 24 species, including two lizards. the predators make up such a large proportion of In contrast I found only nine species of herbivores the fauna. If this is so, it might have some bearing and eight scavengers. This suggests that among the on theoretical considerations which state that the

TABLE 1.—List of invertebrates found on Malpelo Island with estimate of their food habit (H = Herbivore; P = Predator; S = Scavenger; Par. = Parasite; O = Omnivore)

Oligochaeta 1 (?)sp. S Nematoda spp. Par. Gastropoda 2 spp. H Diplopoda sp. H Geophilomorpha Chilenophilidae Ribautis sp. P Scolopendromorpha Cryptopidae Scolopocryptops sp. P Cryptops sp. P Decapoda Gecarcinidae Gecarcinus malpilensis O Isopoda sp. S Chelonethida sp. P Araneida Oecobiidae sp. P Selenopidae Selenops sp. P Agelenidae spp. P Lycosidae spp. P •• Salticidae about 3 spp. P Segestriidae PAriadna sp. P Pholridae sp. P Opilionida Assamidae Paramitraceras sp. P Schizomida sp. P Acarina sp. Par. Thysanura sp. H Collembola Entomobryidae Lepidocyrtus caprelsi S " Lepidocyrtus sp. S Orthoptera Gryllidae Hygronemobius sp. H Hemiptera Enicocephalidae Systelloderes sp. P Coleoptera Hydrophilidae Enochrus sp. P Staphylinidae sp. H Scydtnaenidae PEuconnus sp. P Diptera Ephydridae Nostima sp. S " Scatella sp. s Sciaridae Bradysia sp. s Chloropidae sp. H Lepidoptera ?Pyralidae sp. H " Tineidae Ereunetis sp. S Hymenoptera Formicidae Odontomachus sp. P Pheidole sp. H Strumigenys sp. P NUMBER 176 25

number of predator species can never exceed the anoles seem to limit their density through behav- number of prey species (cf. Levins, 1968), except ioral means, as they are nonterritorial (see Rand, in some special circumstances (Stewart and Levins, Gorman, and Rand, pp. 27-38 herein). Their gut 1973). If the phenomenon is real, it could be that contents did not suggest that they were experienc- some of the necessarily simplified assumptions on ing food shortage, but in other seasons this could which such theories are based are violated. It could be different. be that predation by the lizards on the inverte- The other interesting feature of the foodweb is brates, prey and predators alike, is so intense, at that my original idea, that the entire economy of least during some times of the year, that competi- the island was based on the sea, may be wrong. tion between the invertebrate predators is greatly Only the birds and Diploglossus depend completely, reduced, thus enabling more of them to coexist. and the landcrab partly, on income from the In that case, the problem boils down to what limits ocean. All the other species seem to depend on a the numbers of lizards. We found no evidence of purely terrestrial economy. I would predict that if predation by birds on the lizards, nor do the the birds, which provide the link between the

TERRESTRIAL SOUTH ROCK GRASS

V A ALGAE LICHENS HERBIVORES NEMATODES MOSSES 9spp. PREDATORS V 22spp. ANOLIS PHYLLODACTYLUS V V A A V? FECES SCAVENGERS V CORPSES 8spp.

GECARCINUS

A

FECES BIRDS FOOD REMAINS DIPLOGLOSSUS CORPSES V

TICKS

MARINE

FIGURE 12.—Tentative structure of the foodweb on the island of Malpelo. The four question marks refer to uncertainties in utilization of grass by the herbivores, crabs eating living insects, Diploglossus taking living crabs, and the importance of bird droppings in providing nutrients for the plants. For simplicity, some relationships are omitted. These include lizards and crabs providing food for the scavengers and ticks being eaten by Anolis. 26 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY ocean and the land, disappeared, Diploglossus Brues, C. T., A. L. Melander, and F. M. Carpenter would soon be extinct and the land crab would 1954. Classification of Insects. Bulletin of the Museum of reach lower densities, but that this would have Comparative Zoology at Harvard College, 108:1-917. Carlquist, S. little, if any, effect on the rest of the fauna. The 1965. Island Life. Garden City, New York: Natural His- only condition under which this would not be true tory Press. is when the weathering rock would not provide Clapp, R. B. enough nutrients for the plants to grow, so that 1972. The Natural History of Gardner Pinnacles, North- fertilization (e.g., phosphates) by the birds is of western Hawaiian Islands. Atoll Research Bulletin, vital importance. 163:1-25,7 figures. ACKNOWLEDGMENTS.—I wish to express my grati- Dunn, E. R. tude to all those who identified the specimens: 1939. Zoological Results of the George Vanderbilt South Pacific Expedition of 1937, Part III: The Lizards Drs. M. J. Dibben and T. Esslinger of Duke Uni- of Malpelo Island, Colombia. Notulae Naturae of versity for the lichens, Dr. D. B. Lellinger of the the Academy of Natural Sciences of Philadelphia, National Museum of Natural History, Smithson- 4:1-3. ian Institution, for the fern, Dr. L. Abele of the Fowler, H. W. Smithsonian Tropical Research Institute and Drs. 1938. The Fishes of the George Vanderbilt South Pacific R. E. Crabill, W. B. Peck, D. L. Wray, A. B. Expedition, 1937. Monographs of the Academy of Gurney, J. L. Herring, P. J. Spangler, J. M. King- Natural Sciences of Philadelphia, 2: v + 349 pages, 10 plates. [Malpelo, pages 5-6.] solver, W. W. Wirth, R. J. Gagne, C. W. Sabrosky, Garth, J. S. D. C. Ferguson, D. M. Weisman, and Dr. D. R. 1948. The Brachyura of the "Askoy" Expedition with Smith, all of the Department of Agriculture, for Remarks on Carcinological Collecting in the Pan- identification of the items listed in Table 1. ama Bight. Bulletin of the American Museum of I also gratefully acknowledge Drs. N. G. Smith Natural History, 92(1): 1-66, 4 figures. Levins, R. and J. B. Graham of the Smithsonian Tropical 1968. Evolution in Changing Environments. Princeton: Research Institute and G. C. Gorman of the Uni- Princeton University Press. versity of California at Los Angeles for critically Murphy, R. C. reading the manuscript. 1945. Island Contrasts. Natural History, 54:14-23. Slevin, J. R. 1928. Description of a New Species of Lizard from Literature Cited Malpelo Island. Proceedings of the California Academy of Sciences, fourth series, 16:681-684, Bond, J., and R. M. deSchauensee plates 25, 26. 1938. Zoological Results of the George Vanderbilt South Stewart, F. M., and R. Levins Pacific Expedition of 1937, Part II: The Birds of 1973. Partitioning of Resources and the Outcome of Malpelo Island, Colombia. Proceedings of the Acad- Interspecific Competition: A Model and Some emy of Natural Sciences of Philadelphia, 90:155- General Considerations. The American Naturalist, 157, plates 9-11. 107:171-198. Natural History, Behavior, and Ecology of Anolis agassizi

A. Stanley Rand, George C. Gorman, and William M. Rand

Introduction ABSTRACT Almost all islands in the Caribbean have been Anolis agassizi is found throughout Malpelo Is- colonized by one or more stocks of Anolis lizards. land, and in places is extremely abundant. In one Patterns of distribution, ecological relationships, study area we estimated one lizard per two square and behavior have been extensively studied in meters. Social behavior contrasts markedly with these West Indian anoles by Ruibal (1961), Rand other anoles in the apparent lack of territoriality, infrequent aggressive encounters, and reduced dis- (1964), Gorman (1968), Schoener (1969), Schoe- play repertoire. There is a strong tendency for ner and Schoener (1971), Lazell (1972), Williams social facilitation—the species is aptly described as (1972) and many others. The Pacific waters adja- "curious." Home ranges are large, and lizards tend cent to the tropical American landmass are rela- to move long distances during the course of the tively impoverished in islands, and only two are day. Resources such as food, water, and perch sites inhabited by Anolis. One is Cocos Island, and the may be shared by many individuals. There appear to be no special adaptations to temperature ex- second is Malpelo. Malpelo is more isolated than tremes nor to desiccation. Reproduction is appar- any of the West Indies, which have been naturally ently highly seasonal, as evidenced by the complete colonized by anoles, although strictly in terms of absence of small juveniles in museum collections distance it is within the limits known for coloniza- made in December and January, and in our col- tions by Caribbean Anolis (Williams, 1969; Rand, lections of February and March. In our sample about half the females were carrying eggs. The 1969). Malpelo is also more devoid of vegetation species is larger than most solitary anoles (adult than any West Indian island. Caribbean anoles males often > 100 mm) and markedly sexually are mostly arboreal, in the strictest sense, the ex- dimorphic for size (males larger). Adult males ceptions usually perch on other sorts of vegetation fall into two discrete reproductive categories that (bushes, grass, cactus, etc.). There are a few anoles have morphological correlates. Some have black heads and permanently erect nuchal crests—these associated with rocks, such as A. rimanim on boul- have active testes. Others lack the nuchal crest ders in Hispaniola and A. bartschi on cliffs around and have spotted heads like the females—these cave entrances in Cuba; but none is from an envi- have completely regressed testes. ronment so devoid of higher plants as is Malpelo Island. Previous to the expedition reported here, our knowledge of Anolis agassizi was limited, but the A. Stanley Rand, Smithsonian Tropical Research Institute, very nature of the habitat suggested that this P. O. Box 2072, Balboa, Canal Zone. George C. Gorman, species might be rather different in behavior and Department of Biology, University of California, Los An- geles, California 90024. William M. Rand, Department of ecology from better known anoles of the Antilles. Nutrition, Massachusetts Institute of Technology, Cambridge, The species was first described by Stejneger (1900: Massachusetts 02139. 161-163) with a detailed account of external mor- 27 28 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY phology. The paper included a superb watercolor A. hendersoni of Hispaniola or A. bartschi of Cuba. plate of an adult male. Stejnegar also provided The color pattern of A. agassizi is unusual, con- the few field observations available: "Mr. Charles sisting of minute light spots on a very dark ground H. Townsend, who collected these specimens in color, but many species of Anolis on small islands Malpelo, informs me that they were running over have unique color patterns. the rocks near the water. The island was too steep We were able to verify these attributes and to to afford a landing, but the lizards were shot off or discover additional peculiarities particularly in whisked off the face of the cliffs, thus falling into the social behavior of A. agassizi during our visits the water, whence they were secured by the col- from 29 February to 3 March 1972. lector." Slevin, in 1927, landed briefly on Malpelo ACKNOWLEDGMENTS.—We are grateful to Drs. and collected a series of A. agassizi, but he did not Thomas A. Jenssen of the Virginia Polytechnic explore the island. Institute and Ernest E. Williams of Harvard Uni- Ronald Smith of the George Vanderbilt South versity for reading and commenting on the manu- Pacific Expedition of 1937 did explore the island script. Travel funds for G. C. Gorman were pro- and found anoles all the way to the summit. He vided by a grant from the National Geographic described them as fearless and quite bold (Fowler, Society; partial support for A. S. Rand was pro- 1938). Dunn (1939) utilized the Vanderbilt Expe- vided by National Science Foundation grant No. dition specimens to reconsider the status of the B019801. lizards of Malpelo. Because of peculiarities, par- ticularly a small throat fan, he erected a new, monotypic genus for agassizi. This he called Mari- Distribution and Numbers guana. He reported that the agassizi stomachs con- Anolis agassizi were found everywhere on Mal- tained small insects of two or three species. He pelo from sea level to the highest peaks, and also remarked that the unusual color pattern (see be- on the only one of the small rocks south of the is- low) was similar to that of the other totally unre- land that was explored. They were rare on the lated lizard species on the island, Diploglossus south plateau and uncommon on the black rocks millepunctatus. in the lowest 3 meters. Etheridge (1960) considered the relationships Anoles seemed less common on the large smooth of all available species of anoline lizards including rock faces than on the more irregular rocky areas agassizi. He showed that this species fit squarely which provided more hiding places and probably into the latifrons species group of Anolis, and he more food and water. They certainly did not avoid therefore did not recognize the validity of Mari- either the vertical cliffs or the more nearly hori- guana. The latifrons series is an osteologically zontal edges. The concept of structural niche as primitive group of anoles otherwise restricted to defined by perch dimensions (Rand, 1964) fails the southernmost of the Lesser Antilles and the completely in an environment such as Malpelo. tropical American mainland from Costa Rica southward. Anolis agassizi is virtually indistin- guishable from the Lesser Antillean members of Home Range Size and Population Density the latifrons series on osteological characters alone, Our home range data are based on marked in- but genetic similarity evidence (Webster, p. 50 dividuals. On the first afternoon we marked all of herein) implies that it is not particularly close to the anoles that we could catch in an area of per- either the island or continental species examined. haps 20 m long by 10 m high (Figure 13). We Compared with West Indian anoles, the Malpelo visited this area on three succeeding days and re- species was known to be unusual but not unique corded the numbers of anoles seen, as well as their in several aspects of its biology. Anolis agassizi is activities. larger than most anole species that occur on is- We marked 29 anoles. Of these, we saw 21 again lands without congeners (Schoener, 1969:389), during the next three days. Of the six large adult but it is somewhat smaller than Anolis ferreus of males marked, five were resighted. The marked Marie Galante. Anolis agassizi has a very reduced lizards did not seem to prefer specific perches; dewlap in both sexes, but not more reduced than rather they moved back and forth across the rocks, NUMBER 176 29 some moving 5 to 10 meters during a morning. One average of 16 anoles) of which 0 to 31 percent lizard was seen about 150 meters away from where (average of 16 percent) were marked. If we were it had been marked; however, another marked liz- seeing about the same proportion of unmarked as ard was seen at virtually the same spot on three of marked anoles, and assuming that of the 29 successive days. marked anoles, 20 remained in the area, then the Anoles gathered from distances as great as 10 to total number of anoles in our area would have 15 meters at concentrated resources, such as the been about 120. With this small sample and these oranges and jellied candy ("Chuckles") that we many assumptions, we hesitate to calculate con- offered, and at water seeps. Home ranges are large fidence intervals. We estimate, however, that anoles and, even for adult males, are widely overlapping were concentrating at the oranges from an area of and seem little defended. The long distance moved perhaps 500 to 1000 square meters. This would by one anole and the fact that several marked ani- indicate a population density of about one anole mals were not resighted suggests that there is a per 5 to 10 square meters. Although not all areas floating population like those postulated for of the island are equally habitable by Anolis, there other anoles (e.g., Philibosian, 1972). must be at least 100,000 Anolis agassizi on the During our observations, and when lizards con- island. gregated at sliced oranges we had set out (p. 31), Measurements were taken from 45 animals, we made periodic counts of the numbers present which were preserved and deposited in the Muse- and the proportion of these that were marked. um of Comparative Zoology. This series contained Eleven counts showed 8 to 25 anoles present (an four juveniles whose sex could not be determined, 14 adult females (snout-vent length, SV x 77.1 [71-84] SD 4.6) and 27 males. Fourteen of the males had enlarged testes, dark heads, and well- developed nuchal crests (SV x 97.6 [89-114] SD 7.3). The remaining 13 had small testes, female- like coloration, and small, if any, nuchal crests (SV x 84.2 [71-99] SD 7.9). (The latter males are discussed on p. 32). Weights of these animals just before preserva- tion were found to be highly correlated with snout- vent length in each sex (r = 0.% for males, r = 0.85 for females). The lower correlation for females is due primarily to their smaller size range. The linear regression equations for weight (W) on snout-vent length (SV) are: males, W = — 40 + 0.62 SV; females, W = -18 + 0.35 SV. Only a small number (13.4 percent) of the lizards captured had unbroken tails, and an additional 6.8 percent were broken in capture. The sexes do not differ in the percentage of tails broken (p. 32). In the 18 lizards with unbroken tails, the tail was slightly more than twice SV (mean = 2.19; SD = 0.12). The regression of tail length (T) on SV is T = 22 + 1.9 SV, with an r of 0.92. Schoener (1969) has shown that Anolis species which live on islands without congeners tend to fall into a restricted size range. The largest third FIGURE IS.—Part of the study area on Malpclo where Anolis of the males usually have a snout-vent length be- agassizi were marked and observed and the orange and tween 55 and 90 mm, and the females between 40 "Chuckles" experiments conducted. and 60 mm (for any given species, the female is 30 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY always significantly smaller). Anolis agassizi is an bers than any other prey. The next most common exception. The largest third of the males averages items were beetles, both larvae and adults, which 105.4 mm SV, the maximum is 114 mm. The com- occurred in 72 percent of all guts. Together, bee- parable female average is 85.2 mm SV with a re- tles and ants made up the majority of the food corded maximum at 87 mm. Thus A. agassizi is eaten and probably more than half of the biomass. about the size of the Lesser Antillean giants A. The food items were small, nothing longer richardi and A. bimaculatus, both of which are than 20 mm, few over 10 mm. The majority of the sympatric with smaller species. items and probably of the volume was in the 5 mm Two hypotheses come to mind to explain the class. larger than predicted sizes of A. agassizi: (1) There This range of prey corresponds with our sub- was a congener on the island that is now extinct. jective impressions of what was available. We saw This is unknowable, but unlikely. The very re- almost no invertebrates longer than 20 mm except moteness of the island makes the probability of a earthworms, which were perhaps too well concealed second colonization extremely low; and insertion to be captured, and the large land crabs, which of a new colonist onto a small, already occupied were too big to be taken. island is the exception rather than the rule for As previously pointed out, A. agassizi is larger West Indian Anolis, where competitive exclusion than might be predicted on the basis of Schoener's seems to prevail (Williams, 1969; Gorman and (1969) study of single species anole communities. Boos, 1972). (2) The average prey size is larger on The feeding data, however, do not support the hy- Malpelo than in the Antilles. This is simply not pothesis that A. agassizi is larger because it feeds the case (see below). on unusually large prey. No precise comparisons Any number of additional hypotheses to explain can be made, but the food of A. agassizi is certainly the body-size anomaly could be conjured up (e.g., smaller than that of the smaller mainland species selection to resist predation; intense sexual selec- A. polylepis (Andrews, 1971), and on the same tion), but we have only our ignorance of the his- order as that of the smaller A. roquet, which is torical factors that have acted during the evolution alone on the island of Martinique (Schoener and of this species with which to discuss such hypothe- Gorman, 1968). ses. Thus we resign ourselves to the observation Anoles made repeated efforts to catch a hookless that A. agassizi is large. trout fly which we presented to them. They jumped into the air for it when the artificial fly was dan- gled overhead. One suspects that they may gather Food and Feeding at a bird carcass to feed on the invertebrates which FIELD OBSERVATIONS AND ANALYSIS OF STOMACH also gather there. In this situation, they may leap CONTENTS.—The stomachs of 43 anoles were ex- into the air to catch circling flies. amined and the contents identified by H. Wolda, A small series of anoles was taken from one of who collected invertebrates on Malpelo (see the south rocks, where grass and bushes grew much Wolda's paper herein). Most of the species of in- more abundantly than on Malpelo itself. All of vertebrates that he found on Malpelo, except these anoles had their lower intestines packed with earthworms, were represented in the Anolis' guts. grass seeds which appeared to be undigested. Major groups taken by the anoles include: snails, About 60 percent of the guts contained a few juvenile land crabs, isopods, millipedes, centipedes, (1 to 4) small, presumably parasitic, nematodes. spiders, pseudoscorpions, ticks, and insects. The FEEDING EXPERIMENTS.—As pointed out above, insects included beetles, flies, true bugs, caterpillars, the anoles seemed particularly bold and curious, ants, crickets, and Thysanurians. Many of these and often came to us and even jumped on us. animals are cryptic. They were found under rocks Casual observation implied that they were particu- and in crevices. Anoles certainly do enter crevices larly attracted to the color orange. Thus, they and holes in the rocks where they probably catch seemed especially interested in a Kodak film pack- some prey. age and in an orange screw-cap on a container of The most common items in the stomachs were suntan lotion. This led us to perform two very ants. These occurred in all guts and in larger num- simple field experiments. NUMBER 176

The Orange Experiment: On several occasions in and near the study area, we set a half of an orange on a rock. Anoles from the surrounding areas gathered at the orange to lick repeatedly at c . \> the cut surface, and to bite, pull off, chew, and swallow, bits of the pulp (Figure 14). They also licked the peel, but did not persist in this. The anoles were intially attracted by the bright color of the orange, and persisted because of its taste. We saw lizards come to the orange from as far as 15 meters across the rock face. Five or ten indi- viduals would feed at the orange itself at one time, while another 10 to 20 remained within a 2 meter radius of it. There was almost no aggression shown at the orange though a male occasionally spread FIGURE 15.—Anolis agassizi making a choice between two his small throat fan or bobbed. equal sized pieces of "Chuckles." The anoles seemed attracted to the area of the orange by the activities of the anoles already there, as well as by the orange itself. On one occasion we this candy contains five colors (flavors): red, saw a succession of anoles drinking (1 to 4 at a green, black, orange, and yellow. Comparisons of time) from a small wet spot on the rocks. Again we different colored candy were made by setting out had the impression that these lizards were attracted pairs of equal sized bits of candy close together in by the activities of other individuals. areas where we had not conditioned the lizards to Anoles did not stay at the orange for more than oranges. We noted which bit of candy was selected a few minutes though the same lizard might have by the first lizard to approach (Figure 15). returned several times in the course of an hour or The results (Table 2) show that yellow and two. orange are equivalent and that they both are The "Chuckles" Experiment: To test the possi- chosen more frequently than the other colors (p< bility that the anoles actually showed a color pre- 0.01). Of the remainder, red and green are about ference in making initial feeding choices, we used equal, and both preferable to black, though the the jellied candy called "Chuckles." A package of difference is not significant (p>0.05). We hasten to add that this appears to be a color preference and not a taste preference, for once the less opti- mal candy was licked by the lizard, it was immedi- ately eaten. We cannot separate hue from brightness in these color choices, but all candies presented were conspicuous.

TABLE 2.—Candy color preferences of Anolis agas- sizi (numerator is the number of times "Color selected" was chosen before "Other color in test pair"; denominator is the total number of trials)

Color selected Other color in test pair Yellow Red Green Black Orange 3/7 3/3 6/6 6/8 Yellow 5/6 4/6 6/6 Red - 2/6 5/6 - - 4/6 FICURE 14.—Anolis agassizi gathered about an orange half. Green SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

The preference for orange and yellow seems tail breakage and intensity of predation in lizards most peculiar because there are certainly no fruits of the western United States. of this color on the island. The only objects that If, as it seems likely, crabs and Diploglossus are we could find in this color range were the large the important predators of anoles, it is likely that land crabs, and the anoles showed not the least they catch more small than large ones. The color interest in these. One possible explanation is that patterns of Anolis are highly cryptic on the rocks, the anoles are attracted to and feed on the yolks of suggesting selection by a visually hunting predator. broken seabird eggs. Many boobies nest on Mal- Adult males with their permanently erect nuchal pelo and broken eggs may be a frequent occurrence crests and black heads are very striking and are at certain times of year. more conspicuous than females and juveniles. Adult males of many Antillean Anolis also are more conspicuous than females, presumably be- Predation cause sexual and social selection favors conspicu- The most conspicuous potential anole predators ousness in males and the larger size of the males on the island are the large lizards, Diploglossus, reduces their predation and consequently the selec- the land crabs and the birds. The seabirds could tion against being conspicuous. certainly eat anoles, but Neal Smith (pers. comm.) says this is unlikely. The peregrine falcon (Falcon Temperature peregrinus) has been recorded on the island (Bond and deSchauensee, 1938), but it is normally a bird An open question in our minds, before seeing predator. The land crabs were everywhere. We the anoles of Malpelo in the field, concerned tem- saw a small group tearing apart and eating a still perature adaptations. This species lives on bare living anole (see p. 23). How it was caught is un- rock near the equator. Would A. agassizi prefer known for the crabs are relatively slow moving high body temperatures? Would it be able to resist animals. high environmental temperatures? The Diploglossus appear to depend largely on Our short-term, relatively simple studies indicate the exuvia of boobies and on land crabs for food, conclusively that with respect to temperature A. and they gather where an adult booby is feeding agassizi is a rather typical anoline lizard. Forty its young. They do occasionally eat anoles, how- body-temperature measurements of active animals ever, for a specimen in the American Museum of were taken in the field with a Schultheis thermo- Natural History collections had one in its stomach meter. The mean temperature, representing pooled (C. Meyers, pers. comm.). The Diploglossus could samples of several mornings and early afternoons not be induced to eat oranges, but one big male was 30.6°C (24-38°C) with a standard deviation repeatedly approached an orange that had at- of 1.8. We believe that this approximates the pre- tracted a large group of anoles, and the Anolis ferred body temperature as, quite obviously, the hurriedly fled at each approach. Though this Dip- lizards could easily have had higher body tempera- loglossus was much swifter than the land crabs, tures by sunning themselves briefly on exposed the anoles were even faster and easily evaded him rocks. Conversely, by spending more time in cre- by keeping about a meter ahead of him on the vices, lower body temperatures could have been rocks. maintained. The A. agassizi temperatures were There is strong circumstantial evidence that very similar to the Lesser Antillean species A. rich- predation is important. Approximately 85 percent ardi and A. aeneus (Schoener and Gorman, 1968), of the anoles that we examined had their tails and comparable to, but slightly lower than, A. broken and regenerated. Given the low level of homolechis on Cuba (Ruibal, 1961), a filtered aggressiveness that we observed and the fact that sun animal. females showed as high a precentage of breakage Anolis agassizi does not show a temperature as males, the implication is that the breakage was adaptation to particularly high or low tempera- not caused by interaction with conspecifics, but tures. This is what one might expect from the rather by predation attempts. Pianka (1970) has structure of the island, which receives a great deal argued for a correlation between frequency of of insolation but also provides a great many hiding NUMBER 176

places to escape the sun. The most extensive slopes light source at 30 cm, and in less than two minutes and cliffs orient north and south so that large parts with the light source at 42 cm. The temperature of of the island are in total shade for either the onset of panting was not correlated with body size, morning or the afternoon. This is probably the and only slightly with the distance between heat most peculiar aspect of this anole's thermal ecology. source and cage. With the heat source at 30 cm, The A. agassizi were active during most of the day ten lizards began panting at a mean body tempera- in the shade and were seen basking only briefly ture (MBT) of 34.8°C; with the heat source at 42 when the sun first reached them. Since we worked cm, seven lizards began panting at MBT of 33.7°C. on the west side, there were large areas where we We scored trials only when the lizard moved, saw anoles active in the early morning and which then panted, the reason being that occasional ani- they abandoned fairly shortly after the sun reached mals opened the mouth soon after being dropped them (see p. 35). into the cage. In all cases where mouth opening The situation is quite unlike that which is usual preceded movement, the time of the trial was very for a tree lizard who seldom needs to move more short and the body temperature was well below than a few inches to get into the shade. An A. that of all other trials. To confirm that movement agassizi must move several meters if it is to exploit was a response to the heat, a transparent heat filter many of the rocks and cliffs. In the absence of fly- was inserted between the light and the cage. In ing predators it does not need to worry about keep- eight of 10 cases the lizard did not move at all for ing close to shelter. more than 2.5 minutes (trial was discontinued at Anoles that live in areas of high insolation tend 3 minutes). to have either enlarged scales or black peritoneal This temperature at which panting begins pigment. Anolis agassizi has the latter. The walls (pooled mean, 34.2°C) is considerably lower than of the body cavity are pigmented, most heavily on published data for several Puerto Rican species of the posterior half. There is some pigment on the Anolis (Heatwole et al., 1969), and is between data heart and on the last 1 or 2 cm of the intestine. In for two Cuban species (Ruibal, 1961), A. allogus some examined A. agassizi the entire dorsal surface of deep shade (30.3°C) and A. homolechis of fil- of the lower intestine is pigmented. tered sun (36.2°C). To determine adaptations to high temperatures, Quite clearly, A. agassizi is not specifically a simple heating experiment was performed on adapted to high body temperatures. Because of the animals brought to Panama from Malpelo. These north-south orientation of Malpelo and the great experiments were conducted immediately upon contrast between sun and shade on each part of our return, and the lizards could not have accli- the island with the passage of each day, some of mated to the new climatic regime. Individuals the social organization discussed below may be re- representing all size classes were tested. In these lated to the lack of temperature adaptation. An experiments we suspended a bright lamp above a animal cannot afford to defend a restricted terri- cage that measured approximately 30 X 30 X 45 tory if it is unliveable for half of every day. cm. The cages were constructed of wood with two glass sides. Two parallel series of experiments were run: in one, the lamp was placed at a height of Water and Drinking 30 cm above the cage top; in the second, at 42 cm. Many anoles that occur in very arid micro- Each individual was dropped into the cage, the habitats (A. poncensis in Puerto Rico, A. auratus light turned on, and the response observed for 3 in Panama) tend to have enlarged dorsal scales. minutes. Anolis agassizi does not. This reinforces our im- It has been observed that when an anole is pression that Malpelo is not as dry as the lack of dropped into a novel situation its typical first re- vegetation suggests. Lying within the doldrums and sponse is to "freeze." As it warms up under these being of some height, Malpelo probably receives experimental conditions, it starts moving about frequent rain during most of the year. Water prob- with the onset of discomfort; finally it opens its ably does not stand very long on the exposed bare mouth and begins panting (Ruibal, 1961). surfaces of cliffs and rocks. But fresh water was All lizards moved within 50 seconds with the found spottily over most of the island during our 34 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY visit, even though it did not rain. The volcanic 1971) indicated that within a population some fe- rock of which the island is made is porous and males were likely to be reproductive (with an there are many small seeps, springs, and rock pools. oviducal egg or an enlarging follicle) throughout We saw anoles drinking together at a small seep the year, the percentage seemingly affected by rain- and it may be that much of their water comes from fall. Thus in wet months 100 percent of the fe- such limited and shared sources. males might be scored reproductive, but in exten- Initial tests done by Paul Licht of the Univer- sive dry seasons, reproductivity could approach sity of California at Berkeley on captive A. agassizi zero. suggest that they lose water under drying condi- The reproductive condition of the females on tions much as typical Antillean anoles do, and Malpelo also hints at seasonality. In our collec- have no special adaptations either to prevent desic- tion, five females had oviducal eggs, the smallest cation or to tolerate water loss. Several specimens female being 72 mm SV; two additional females collected on Malpelo immediately excreted fluid had enlarging follicles, these were both large speci- urine, implying no special need for water con- mens (86 and 87 mm). Without adequate samples servation. we cannot determine the size at sexual maturity. Since our smallest reproductive was 72 mm, let us consider this our cut-off point. There were six Reproduction, Sex Ratio, and Size Class females larger than 71 mm which were completely Distribution nonreproductive. Thus, in early March, about 50 Our collection on Mapelo leads to the conclu- percent of the collected female A. agassizi (6 of 13) sion that the lizards must be highly seasonal (or at were nonreproductive. In the museum samples least periodic) in their reproduction, or that we (December-January) six of eight females (75 per- were guilty of very biased sampling, or both (they cent) larger than 71 mm were nonreproductive. are not mutually exclusive). The smallest lizard To determine whether there is true seasonality or collected was 41 mm (SV). This is well above the whether there is merely a low level of reproduc- hatching size of 25 to 30 mm (p. 35). Thus, we tive activity parceled out throughout the year failed to find evidence of any really young lizards. would take further collections of A. agassizi. It is often the case that Anolis are more secretive The males presented the most aberrant and un- as young juveniles and special efforts are needed expected findings. They could be divided into to collect them, but we had enough people spend- three groups. There were large males with the ing enough time in prime lizard habitat to make obviously developed secondary sexual characters, us believe that hatchlings were truly scarce, if not e.g., black head and erect nuchal crest. These totally absent. lizards ranged from about 90 to 114 mm SV and One of us (GCG) examined the two large muse- in weight from 13 to 34 grams. There were also um series of A. agassizi: the Slevin specimens in the smaller males (71 to 85 mm SV and 7 to 13 g) that California Academy of Sciences, collected in Dec- were sexually mature, but lacked the impressive ember 1927; and the Cyrus Perkins specimens in secondary sexual characters of their larger neigh- the San Diego Society of Natural History, collected bors. In other words, they looked like females. in December 1931 and January 1933. Slevin (1928) This is not particularly surprising, for in other wrote of his trip to Malpelo, whereas no informa- studies of Anolis' reproductive cycles (Licht and tion is available on the two journeys made by Gorman, 1970) it has been noted that males are Perkins. Both these series then were made at ap- physiologically sexually mature long before they proximately the same time of year (December- are fully grown and, in fact, before they could be January) and ours was made in the first days of categorized as socially mature, i.e., assuming the March. The smallest lizards in these collections social role of an adult male. were 64 mm (San Diego) and 56 mm (California But there was a remaining class of animals that Academy). Thus nothing is known about the time puzzled us in the field. Six male specimens (75, 85. of hatching or hatchling ecology in nature. 86, 90, 94, and 100 SV) lacked male secondary Studies on the reproductive cycles of tropical sexual characters. Of interest is that three of these Anolis (Licht and Gorman, 1970; Sexton et al., were 90 to 100 mm long and were as large as some NUMBER 176 35

black-headed males, yet all six had completely sured 25 to 30 mm SV (x = 27.7) and weighed regressed gonads, to the extent that we at first 0.42 to 0.81 g (x = 0.62). labeled them "intersexes." Microscopic examina- tion led to the conclusion that these were fully regressed males. Nothing like this has been seen Social Interactions in studies of Caribbean Anolis. Although most tropical anole species show a moderate seasonal The anoles literally swarmed in our study area. cycle for gonad size, the testes are never completely The most impressive observation was the relative regressed; they are always obviously testes, and in lack of aggressive encounters. There appears to be most species they produce sperm throughout the a very strong tendency towards social facilitation, year. Thus there is a testes weight cycle in Malpelo i.e., when an anole took the initiative to approach Anolis. Furthermore, the population rarely shows an object, others in the area were likely to follow a large variation for testes condition, i.e., the males suit. They also appear to have a strong tendency are generally in phase with one another. In con- to examine any unusual object, making it very dif- trast, on Malpelo, we found some males that ap- ficult to perform unbiased observation without a peared fully reproductive and others that appeared blind. Thus the anoles frequently made active ap- fully regressed. proaches to the observers, on several occasions The museum collections also had specimens even climbing on the camera and tripod set up to with fully regressed gonads. In the San Diego col- film their behavior. We could do little to avoid lection there were putative males of 69, 78, 79, 82, them—backing off is not recommended on Malpe- and 83 mm SV that were fully regressed, yet males lo's steep cliffs. as small as 74 that were reproductive. There is no We interpret the social facilitation as an adapta- way to determine whether this is a seasonal phe- tion to patchy but superabundant food sources. nomenon or whether it is a population phenome- Thus it is likely that the carcass of a bird may non. The general lack of aggression that we often serve as a focus for feeding of many anoles in observed (p. 36) might be related to the large a given area. In the West Indies, a well-established number of nonreproductives in the population. method for attracting large numbers of anoles is Male Anolis are often more conspicuous than to break open a termite nest. Under such condi- females and more likely to hold their ground tions, large numbers will often gather and feed against an approaching collector. Thus they have actively with little aggressive interaction. The en- greater likelihood of ending up in museum jars. tire social organization of A. agassizi seems to be Our collection consisted of 27 males and 14 fe- based upon this type of situation. males, a sex ratio of almost 2:1. A pool of our sam- Home ranges were large and activity was not ple with the museum collections (and restricting restricted to a few preferred perches. The follow- ourselves to animals 72 mm and greater) shows a ing anecdotal observations support the conclusion sex ratio of 62:27, favoring males. We find it hard that home ranges overlap widely with residents to believe that this is simply sampling artifact, but showing little aggression. Up to 8 anoles simulta- serious demographic study would have to be under- neously would lick an orange, with up to 15 within taken to resolve the problem. a meter of it. One morning, approximately a half dozen lizards that had been active on a smooth rock ledge moved laterally about 10 meters to a Reproduction in Captivity cliff face. This occurred about a half hour after the sun reached the ledge. Presumably, they were Four females returned to Panama laid eggs, one abandoning a foraging area that had become too at a time. Time intervals (n = 18) between suc- hot. Many then gathered to try to catch a dangling cessive eggs varied from 4 to 27 days (x = 13.6). trout fly without showing any aggressive inter- Eggs were kept under ambient climatic conditions action. On one accasion, a group of about a dozen in Panama (which must be close to those on Mal- anoles assembled at our portable radio and climbed pelo) and 13 hatched with an incubation time of all over it, though it provided neither food nor 48 to 69 clays (x = 57.8). The hatchlings mea- water. After some moments, all drifted away. This 36 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY is an example of the social facilitation referred to There was not an initial territoriality that subse- above. quently broke down under repeated attacks as has H. Wolda reported that as he turned over rocks been described for gulls by Drury and Smith looking for invertebrates, he was followed by a (1968). group of anoles that fed on the food that he ex- At the orange, when a new lizard arrived with a posed. The A. agassizi in a walk-in cage in Panama rush, one already there might run off. In general, gather around anyone who enters, as if expecting large anoles supplanted smaller ones, but no or- food. ganized dominance hierarchy was evident. Occasionally we saw aggressive interactions be- Introducing anoles tethered to a long stick at tween individuals, including chasing and some various spots both close to and far from other display. The instances involved crested vs. crested anoles produced a little interaction and in one males, crested vs. noncrested males, and noncrested case an attack and bite. It also produced a little vs. noncrested individuals of unknown sex. But we display. The residents did not seem to be afraid of emphasize that we saw remarkably little display in the intruder, nor of the way it was presented but comparison to our experiences with West Indian in most cases they were just not very interested. anoles. The only display that we saw was a very The social behavior in Anolis agassizi differs simple multiple bob. This varied somewhat in from that described for any other species of anole. amplitude and it was given sometimes with and This is most readily interpretable in terms of the sometimes without the dewlap being spread. The atypical structural niche, viz., in place of discrete display was usually seen in seemingly aggressive defensible perch sites (tree trunks, fence posts, situations, but was also occasionally given in a etc.) there are large areas of cliff face that are ex- somewhat less vigouous form when a lizard moved posed to complete shade or complete sun for and stopped. many consecutive hours each day. In addition, food The captive anoles in Panama have shown little resources are presumed to be much patchier than in the way of additional kinds of display. We have in a typical tropical environment. seen males jaw fencing, jaw locking and biting one The basic difference in social organization of another. More elaborate displays have not been this species lies in the reduced overt aggressive seen, nor any other simple displays. Males in cap- interactions among individuals, and the high de- tivity that are courting approach the female with gree of social facilitation such that directional the simple multiple bob, followed by pauses. This movements of one individual are likely to prompt is similar to courtship displays of other Anolis similar investigatory movements in several other species. Anolis agassizi has a display repertoire that individuals. In anthropomorphic terms, this species is more reduced than is typical for the genus and is extremely curious. appears more like a sceloporine lizard. This is prob- Both males and females seem to have overlap- ably a secondary simplification. ping home ranges. They do not defend territories In a large outdoor cage in Panama, adult males or individual areas. There is a high degree of tol- show evidence of having developed a rather com- erance for the close approach of conspeciflcs at a plex dominance hierarchy. common resource, be it food, water, or perch site. On Malpelo, the anoles exhibited direct compe- There are occasional overt aggressive interactions tition for food. Frequently, when an anole had when "individual distance" appears to be viola- worried off a bit of orange pulp or picked up a ted, but aggression is seemingly not associated with piece of "Chuckle" too big to swallow, immediately violation of specific space (territory). Widely over- it would run off with it. Sometimes it was pursued lapping home ranges without aggression have been by several other lizards, including much smaller observed in females of A. valencienni in Jamaica individuals that would try to seize the food from (R. Trivers, Harvard University, pers. comm., and its mouth. In these cases there was very little ag- A. biporcatus in Panama, A. S. Rand and pers. obs.). gression shown. But in both these species the males are much less No anoles attempted to monopolize the resource tolerant of members of the same sex. by chasing or displaying at others. This was true Anolis agassizi not only has a reduced frequency even of the first individual to arrive at a resource. of display interactions but also seems to have a re- NUMBER 176 37 duced display repertoire. This is in strong contrast there is more variability between species of Anolis to the anoline species in the West Indies (Gorman, in social organization than had been realized. We 1968) and to A. townsendi of Cocos Island (Car- have attempted to show in this paper that the penter, 1965). Our tentative explanation is: first, rather unusual social organization of A. agassizi that, with reduced aggression and territorial de- is an adaptation to its unusual enviroment. We fense, there are fewer bits of information to be predict that A. agassizi will remain an extreme in communicated; and second, that the striking, per- the continuum of social organization among anoles, manently displayed color pattern and erect nuchal but that intermediates between it and the more crest of adult, reproductively active males assume common types exist and, further, that the conti- much of the communication function otherwise nuum will be expanded in several quite different transmitted by complex movements of head, dew- directions as species adapted to unusual ecologies lap, and tail (display action patterns). The only are described. West Indian anole known to have reduced display is Anolis ferrens of Marie Galante which is like A. agassizi in being unexpectedly large and in having Literature Cited a very striking permanently erect crest (in this case on the tail) in adult males. Andrews, R. M. 1971. Structural Habitat and Time Budget of a Tropical We would suggest that the reduced aggression Anolis Lizard. Ecology, 52 (2):262-270. is a direct adaptation to living under conditions Bond, J., and R. M. deSchauensee where resources are scattered so that a large home 1938. Zoological Results of the George Vanderbilt South range is necessary to include them. Furthermore, Pacific Expedition of 1937, Part II: The Birds of the resources are often so abundant that there is Malpelo Island, Colombia. Proceedings of the Acad- enough for all and consequently little advantage emy of Natural Sciences of Philadelphia, 90:155-157, gained by attempting to monopolize them. The plates 9-11. reduced amount of display permits efficient utili- Carpenter, C. C. 1965. The Display of the Cocos Island Anole. Herpeto- zation of common resources. The social facilitation logica, 21 (4):256-260. ("curiosity") aids in the discovery of these re- Drury, W. H., and W. J. Smith sources. 1968. Defense of Feeding Areas by Adult Herring Gulls The social organization that we have described and Intrusion by Young. Evolution. 22 (I): 193-201. for A. agassizi is very different from that described Dunn, E. R. in studies of A. sagrei (Evans, 1938), A. carolinensis 1939. Zoological Results of the George Vanderbilt South Pacific Expedition of 1937, Part III: The Lizards (Greenberg and Noble, 1944), A. lineatopus of Malpelo Island, Colombia. Notulae Naturae of (Rand, 1967), and A. nebulosus (Jenssen, 1970). the Academy of Natural Sciences of Philadelphia, From these studies we have come to expect anoles 4:1-3. to show a more or less similar pattern. In these Etheridge, R. species the adult male maintains a small home 1960. The Relationships of the Anoles (Reptilia: Sauria: range centering on one or more preferred perches. Iguanidae): An Interpretation Based on Skeletal Morphology. Unpublished Ph.D. Dissertation, Uni- The home range is defended against other adult versity of Michigan. Ann Arbor: University Micro- males as a territory and the resident gives frequent films Inc. assertion displays. An adult female maintains a Evans, T. L. still smaller home range which she may defend 1938. Cuban Field Studies on Territoriality of the Lizard, against other females and against similarly sized Anolis sagrei. Journal of Comparative Psychology, subadult males. The territory of an adult male 25:97-125. Fowler, H. W. typically overlaps those of one or more adult fe- 1938. The Fishes of the George Vanderbilt South Pacific males. All of the species showing this general pat- Expedition, 1937. Monographs of the Academy of tern of social behavior have similarities in their Natural Sciences of Philadelphia, 2: v + 349 pages, ecologies. Evidence is accumulating from recent 10 plates. [Malpelo, pages 5-6.] studies of species with different ecologies (e.g., A. Gorman, G. 1968. The Relationships of Anolis of the roquet Species aeneus, Stamps, 1973; A. valencienni, Trivers, pers. Group (Sauria, Iguanidae), III: Comparative Study comm.; A. biporcatus, A. S. Rand, pers. obs.) that of Display Behavior. Breviora, 284:1-31. 38 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Gorman, G., and J. Boos Ruibal, R. 1972. Extinction of a Local Population of Anolis Lizards 1961. Thermal Relations of Five Species of Tropical through Competition with a Congener. Systematic Lizards. Evolution, 15:98-111. Zoology, 21:440-441. Schoener, T. W. Greenberg, B., and G. K. Noble 1969. Size Patterns in West Indian Anolis Lizards, I: Size 1944. Social Behavior of the American Cameleon (Anolis and Species Diversity. Systemic Zoology, 18:386-401. carolinensis Voigt). Physiological Zoology, 17 (4): Schoener, T. W., and G. C. Gorman 392-439. 1968. Some Niche Differences in Three Lesser Antillean Heatwole, H., T. H. Lin, E. Villal6n, A. Muftiz, and A. Matta Lizards of the Genus Anolis. Ecology, 49 (5):819-830. 1969. Some Aspects of the Thermal Ecology of Puerto Rican Anoline Lizards. Journal of Herpetology, Schoener, T. W., and A. Schoener 3 (l-2):65-77. 1971. Structural Habitats of West Indian Anolis Lizards, Janssen, T. A. I: Lowland Jamaica. Breviora, 368:1-53. 1970. The Ethoecology of Anolis nebulosus (Sauria, Sexton, O. J., E. P. Ortleb, L. M. Hathaway, R. E. Ballinger, Iguanidae). Journal of Herpetology, 4 (1-2): 1-38. and P. Licht Lazell, J. D., Jr. 1971. Reproductive Cycles of Three Species of Anoline 1972. The Anoles (Sauria, Iguanidae) of the Lesser An- Lizards from the Isthmus of Panama. Ecology, tilles. Bulletin of the Museum of Comparative 52(2):2Ol-215. Zoology, Harvard University, 143 (1): 1-115. Slevin, J. R. Licht, P., and G. C. Gorman 1928. Description of a New Species of Lizard from 1970. Reproductive and Fat Cycles in Caribbean Anolis Malpelo Island. Proceedings of the California Lizards. University of California Publications in Academy of Sciences, fourth scries, 16:681-684, Zoology, 95:1-52. plates 25, 26. Pianka, E. R. Stamps, J. A. 1970. Comparative Autecology of the Lizard Cnemidoph- 1973. Displays and Social Organization in Female Anolis orus tigris in Different Parts of Its Geographic aeneus. Copeia, 1973 (2):264-272. Range. Ecology, 51 (4):702-720. Stejneger, L. Philibosian, R. 1900. Description of Two New Lizards of the Genus 1972. Territorial Behavior and Population Regulation in Anolis from Cocos and Malpelo Islands. Bulletin Anoline Lizards. Unpublished Ph.D. dissertation, of the Museum of Comparative Zoology, 34 (6): 161- University of California, Riverside. 164, 1 plate. Rand, A. S. Williams, E. E. 1964. Ecological Distribution in Anoline Lizards of 1969. The Ecology of Colonization as Seen in the Zoo- Puerto Rico. Ecology, 45 (4):745-752. geography of the Anoline Lizards on Small Islands. 1967. Ecology and Social Organization in the Iguanid Quarterly Review of Biology, 44:345-389. Lizard Anolis lineatopus. Proceedings of U. S. 1972. The Origin of Faunas: Evolution of Lizard Con- National Museum, 122 (3595): 1-79. geners in a Complex Island Fauna: A Trial Analy- 1969. Competitive Exclusion Among Anoles (Sauria: sis. Pages 47-89 in volume 6 in T. Dohzhansky, M. Iguanidae) on Small Islands in the West Indies. Hecht, W. Steere, editors, Evolutionary Biology. Breviora, 319:1-16. New York: Appleton, Century, Crofts. Notes on the Natural History of Diploglossus millepunctatus (Sauria: Anguidae)

A. Ross Kiester

of the validity of the genus Celestus requires fur- ABSTRACT ther attention (Myers, 1973). Diploglossus millepunctatus is among the largest The natural history of the large anguid lizard of the anguids, ranging up to 250 mm snout-vent Diploglossus millepunclatus from Malpelo Island length with a tail length of 162 mm and a weight is described. This species is characterized by its abundance, flexible thermal regime, euryphagy, of 268 grams. It is rather heavy-bodied with short and curiosity. limbs. Older males possess the enlarged wide heads that are characteristic of some other anguid genera such as Gerrhonotus (Stebbins, 1954). The body has a very shiny metallic appearance and like that Introduction of other anguids is encased in a solid armor of The large anguid lizard Diploglossus millepunc- osteoderms. The tail tends to be rather blunt in tatus is one of the most conspicuous elements of adults but is more typically tapered in the young. the fauna of Malpelo Island. Because the island The color pattern is dark black or brown-black is seldom visited little has been reported of its with small yellow or yellow-white flecks in great natural history. Here we begin to fill in some of profusion, as the specific name suggests (Figure the gaps in our knowledge of this species. 16). Diploglossus millepunctatus was first described by O'Shaughnessy (1874) who listed the type- locality as "the north-west coast of America." Slevin (1928) described Celestus hancocki from Malpelo Island, noting that it was similar to the unique type of O'Shaughnessy. Dunn (1939), who had additional material, synonomized Slevin's name on the basis of similarities of his series with the type of millepunctatus and the fact that there was some indirect historical evidence that the orig- inal specimen could indeed have been collected on Malpelo. Dunn also placed the species in the genus Diploglossus arguing that Celestus was not recognizable. While Dunn's generic allocation of the species millepunctatus has prevailed, the issue Fir.tRF. 16.—Diploglossus niillepurictatus aggregated about a A. Ross Kiester, Museum of Comparative. Zoology, Harvard young Blue-faced Booby (Sula dactylalra) which has just University, Cambridge, Massachusetts 02138. been fed.

39 40 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

ACKNOWLEDGMENTS.—I wish to thank T. L. Activity and Thermal Relations Chorba of Oxford University, G. C. Gorman of the University of California, Los Angeles, J. A. Hoff- Although no nights were spent on the island, it man of the Smithsonian Astrophysical Observatory, seems likely that D. millepunctatus spends nights A. S. Rand of the Smithsonian Tropical Research inside crevices. Upon arriving on the island at 0800 Institute; and W. M. Rand of the Massachusetts or earlier we observed few lizards out in the open Institute of Technology for assistance in collecting but saw many in crevices. Later, the lizards moved data, and R. B. Huey of the Museum of Compara- out to bask in the midmorning sun. By midday tive Zoology for aid in the preparation of this most had retreated back into the crevices. During manuscript. the afternoon most remained in or adjacent to crevices and only an occasional individual roamed in the open. Habitat Overall mean body temperature (cloacal tem- perature ± se) of 12 D. millepunctatus during Diploglossus millepunctatus is found virtually activity was 27.5 ± 0.64°C (range 24.8 to 32°C). everywhere on the island from next to the water's However, body temperatures shifted upward be- edge to the top of the summit (376.4 m). Lizards tween morning and afternoon. The mean body usually did not occur far from the crevices into temperature of six lizards active between 0838 and which they retreated. Otherwise, they scramble 0910 was 25.8 ± 0.3°C, whereas the mean body over the bare rocks and cracks, sometimes moving temperature of six lizards active between 1324 and with agility in an almost serpentine fashion, but 1342 was 29.3 ± 0.7°C. There was no overlap in also occasionally falling. body temperature between these two samples, and Although these lizards live next to the ocean the difference between the samples is significant and are capable of swimming, we were unable to (p = 0.01, Kolmogorov-Smirnov test, Siegel, 1956: get any individuals to enter the water voluntarily 127). as Slevin (1928) reported they would do. Among anguids only Ophisaurus attenuatus is reported to have a higher body temperature (Fitch, 1956:444). However, as with Anolis agassizi (p. Population Size and Structure 33), D. millepunctatus appears in general to have These animals are extremely common. No quan- no special physiological adaptations to the ex- titative measurements of population density were treme thermal environment in which it lives. The attempted, but rough counts indicated that a den- ready availability of the cooler crevices seems to sity of 1 per 10 square meters was perhaps a rea- afford D. millepunctatus a means of cooling itself sonable estimate. The population seems to consist when the temperatures on the exposed parts of the mostly of adults. Of the literally hundreds of ani- island are extreme. On the other hand, much of mals that were observed only three young were the island remains in the shade until the very late observed. Except for these individuals, the entire morning, by virtue of the steep rock walls, and population fell into a size range of approximately this species can also be active at lower tempera- 180-260 mm snout-vent length. The juveniles were tures. Thus it appears to be flexible with regard approximately 75 mm snout-vent length and no to the temperatures at which it is active. animals of an intermediate size were seen. This striking size distribution might possibly be due to the fact that younger animals tended to remain Drinking and Water Relations hidden in crevices. However, the fact that some Diploglossus millepunctatus drinks readily in young were seen while no intermediate sized ani- captivity. On Malpelo standing fresh water was mals were found seems to indicate that the obser- observed in a few locations and water seepages vations reflect a real phenomenon, namely that were seen in many of the crevices, although most reproduction is seasonal or erratic and that our of the island was dry. It is impossible to know from sample was taken prior to a period of reproductive our observations what the seasonal pattern of water activity. availability is, or whether there is water available NUMBER 176 41 in the deeper crevices. Rock surfaces on Malpelo crabs in the stomachs of the animals they exam- ranged from very porous to extremely smooth and ined, and Slevin also reported finding feathers. We hard. In either case it seems that water would not observed this species feeding on carrion consisting stand on the surface of the island for any appreci- of both dead crabs (Gecarcinus malpilensis) and able time after a rain storm, either percolating a Blue-faced Booby (Sula dactylatra). In the case down into the rock, running off rapidly, or evapo- of the booby, the lizards may have also been feed- rating. Although it did not rain during our visit, ing on the insects attracted to the carcass. captive animals displayed an extremely rapid re- Most of the feeding observed was connected with sponse to water. When kept in a large pen with no the activities of living boobies. On two occasions available water, the animals would respond to the animals partially consumed fresh booby feces. The first drops of water sprayed into the cage by run- most spectacular feeding behavior occurs when an ning immediately to the nearest puddle that was adult booby is feeding its young regurgitated fish. forming. This response seemed to be about as fast When an adult bird returns and alights near its a movement as the animals were capable of. They young, the young bird begins to squawk loudly. responded by coming out of crevices or down off This noise attracts the attention of any nearby D. the tops of large bricks that were in the cage. This millepunctatus which immediately run to the vi- rapid response appears to be a specific behavioral cinity of the two birds. If any of the fish is dropped, adaptation to a transiently available water supply. the waiting lizards immediately snatch the food and drag it away. We observed this complete se- quence of behavior on two occasions, and saw Food and Feeding Behavior lizards attracted to calling boobies on several others. Although this species will eat the usual kind of In one case, more than seven adult lizards that had prey items that are to be expected for a lizard of aggregated about two birds successfully carried off this size, such as insects and invertebrates, the and consumed an entire fish (Figures 16, 17). Dur- bulk of the food taken by adults is quite out of the ing such activity the birds appear oblivious to the ordinary for a lizard. There are small insects on lizards. In captivity these lizards would readily eat Malpelo, and in captivity D. millepunctatus will canned tuna or fish-flavored cat food. readily take crickets and mealworms. Actual feed- It appears that the boobies provide a substantial ing on insects on the island was not observed, how- amount of food for the lizards, at least during the ever, probably because the insects occur only inside nesting season. Obviously this source of food is the crevices. Out in the open the lizards were ob- not available year round. The lizard's large stumpy served to feed on a variety of bizarre foods. Both tail may possibly serve as a storage organ, and Slevin (1928) and Dunn (1939) reported finding several of the individuals which we examined had large fat bodies in the abdominal region. Presum- ably the young feed more on the insects that are available in the crevices and that may not be as seasonal as the supply of fish.

Social Behavior In general this species does not appear to have a complex social behavior system. Individuals fre- quently ignore each other, although short chases were occasionally observed. The only activity that could be interpreted as a display is a particular "head-up" posture in which a lizard sits with its head held back so that it points almost directly upwards. Although two lizards within eyesight of FICURE 17.—Diploglossus millepunctatus feeding on a fish each other could be observed to do this, no obvious regurgitated by the booby. reaction was observed. Further observations will 42 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY be needed to confirm that this is indeed a display. Conclusions Social facilitation for feeding appears to be well developed in this species. A running individual This species appears to be very opportunistic, seems to attract the attention of others who fre- ready to take advantage of a transiently available quently give chase. This reaction appears to be in resource on split second notice. Associated with contrast to the response elicited by a slowly moving this opportunism is a high level of curiosity and a individual. When food or a source of food is pres- tendency to respond to anything novel or disturb- ent, it is not clear how much of the attraction is ing in the environment. We found that individuals due to the food itself and how much to the other of D. millepunctatus would approach human lizards; however, aggregation can be extremely beings or their artifacts with little regard for pos- rapid. As mentioned above, seven or eight animals sible danger. This curiosity seems linked to the aggregated about two boobies, and the aggregation unpredictable and meager food supply on Malpelo took place in a radius of over 5 meters in about 2 Island where any disturbance may likely mean minutes. Rapid aggregation in itself is evidence for food rather than danger for the lizards, since there social facilitation (Kiester and Slatkin, 1974), but are no natural predators (with the possible excep- this point needs to be investigated experimentally. tion of migrating raptors; Bond and deSchauensee, One pair was observed copulating. They were 1938). Thus the benefits of curiosity are great found in a crevice in the shade. The male held the while the disadvantages normally associated with head of the female in his jaws, biting from the this behavior are reduced or absent. In this way side. His body was positioned laterally and intro- D. millepunctatus is typical of predators on remote mission took place from the side. The pair was islands (Carlquist, 1965). found in copulation in the midmorning and re- mained in the same position for two and one-half Literature Cited hours. The position is similar to that described for Gerrhonotus multicarinatus (Fitch, 1935:18). Bond, J., and R. M. deSchauensee 1938. Zoological Results of the George Vanderbilt South Pacific Expedition of 1937, Part II: The Birds of Reproduction Malpelo Island, Colombia. Proceedings of the Academy of Natural Sciences of Philadelphia, 90: A series of 18 specimens collected on a single 155-157, plates 9-11. day during the visit (early March) showed marked Carlquist, S. reproductive activity. Females had many well- 1965. Island Life. Garden City, New York: The Natural developed embryos, and males had large testes. History Press. Females brought to Panama and maintained in Dunn, E. R. captivity gave birth to live young two months 1939. Zoological Results of the George Vanderbilt South later. Other females in this same group and some Pacific Expedition of 1937, Part III: The Lizards females taken to Cambridge, Massachusetts, which of Malpelo Island, Colombia. Notulae Naturae of initially appeared gravid showed no embryos on the Academy of Natural Sciences of Philadelphia, examination after their death some weeks later. 4:1-3. Thus it may be possible that resorption of embryos Fitch, H. S. can occur. 1935. Natural History of the Alligator Lizards. Trans- actions of St. Louis Academy of Sciences, 29:1-38. We may conjecture that D. millepunctatus will 1956. Temperature Responses in Free-living Amphibians have its reproductive cycle tied to that of the and Reptiles of North-eastern Kansas. University booby. From our collections we know that the of Kansas Publications of Museum of Natural His- lizards were in reproductive condition (and often tory, 8:417-476. fat) near the end of the fledging period of the Kiester, A. R., and M. Slatkin boobies. If the amount of food available to the 1974. A Strategy of Movement and Resource Utilization. adult lizards is markedly increased by the appear- Theoretical Population Biology. [In press.] ance of nestling boobies which need to be fed, then Myers, C. W. breeding by the lizards might be expected to follow 1973. Anguid Lizards of the Genus Diploglossus in Pan- ama with the Description of a New Species. this period. American Museum Novitates, 2523:1-20. NUMBER 176 43

O'Shaughnessy, A. W. E. Slevin, J. 1874. Descriptions of New Species of Scincidae in the 1928. Description of a New Species of Lizard from Mal-

Collection of the British Museum. Annals and pelo isiand. Proceedings of the California Academy

Magazine of Natural History, fourth series, 13: of Sciences fourth series> 16:681-684, plates 25, 26. 298-301. Siegel, S. Stebbins, R. C. 1956. Nonparamelric Statistics for the Behavioral Sci- 1954. Amphibians and Reptiles of Western North Amer- ences. New York: McGraw-Hill. ica. New York: McGraw-Hill. A New Gecko from Malpelo Island (Sauria: Gekkonidae: Phyllodactylus)

Raymond B. Huey

and E. E. Williams of the Museum of Compara- ABSTRACT tive Zoology for helpful comments on the manu- script. I acknowledge the collectors of the type- Phyllodactylus transversalis, a new gecko from series, A. S. Rand and H. Wolda of the Smithson- Malpelo Island, Colombia, is characterized by ab- ian Tropical Research Institute. sence of tubercles on arms, legs, rear of head, and tail; absence of ear denticulation; absence of ab- dominal plaque; slight enlargement of dorsal tu- bercles, which form only two distinct paravertebral Phyllodactylus Iransversalis (new species) rows; terminal lamellae of digits distinctly widened and truncate; 13 to 15 fourth toe lamellae; ground FIGURE 18 color medium gray-brown with dark chocolate bands. Source of this species is uncertain. HOLOTYPE.—Adult female, Museum of Compara- tive Zoology (MCZ 130042), collected by A. Stanley Rand from under a rock near the first collecting area on Malpelo Island, Colombia, February 1972. Introduction PARATYPE.—Immature specimen (MCZ 130043) The gekkonid genus Phyllodactylus is widely collected from under a rock on Malpelo Island by distributed in the western hemisphere. Recent Hendrik Wolda, February 1972. reviews by Dixon (1962, 1964a, 1964b) and Dixon DIAGNOSIS.—Differs from the Central American and Huey (1970) have revised all New World tuberculosus group, the Mexican delcampi group, members of the genus except species from the and most South American groups by having small Galapagos Islands. Some 45 species are currently and scattered dorsal tubercles forming only two recognized. paravertebral rows; differs from unctus group in The collection of lizards from the Smithsonian- possessing dorsal tubercles; differs from gerrhopy- U. S. Navy Expedition to Malpelo Island, Colom- gus group by absence of abdominal plaque; from bia, included two specimens of Phyllodactylus, a the species microphyllus by having expanded ter- genus not collected by the few previous expedi- minal lamellae and a small nostril; and from in- tions to the island. Morphological characteristics aequalis in having dark chocolate dorsal bars, more indicate distinctness of these specimens, which I fourth toe lamellae (13 to 15 vs. 10 to 12), smaller here describe following the methodology of Dixon head scales (third labial snout scales 32 vs. 20 to (1964a). 24; scales between eye and nostril 14 to 15 vs. 11 ACKNOWLEDGMENTS.—I thank J. R. Dixon of to 12). Texas A & M University; G. C. Gorman of the DESCRIPTION OF HOLOTYPE.—Rostral twice as wide University of California at Los Angeles; J. B. as high, its dorsal edge almost straight with a dor- Graham and A. S. Rand of the Smithsonian Tropi- sal vertical groove one-half depth of rostral; 2 cal Research Institute; A. E. Greer, A. R. Kiester, internasals, somewhat rounded, their median edges Raymond B. Huey, Museum of Comparative Zoology, Har- in broad contact, bordered posteriorly by 7 small vard University, Cambridge, Massachusetts 02138. granules and a postnasal on each side; nostril

44 NUMBER 176 45 surrounded by rostral, labial, internasal, and 2 flattened, overlapping, those on posterior edge postnasals; first supralabial in narrow contact smaller and granular; rear of head granular with- with ventral edge of nostril; shallow depression out intermixed tubercles; 5 supralabials and 4 in- between internasals, slight depression in frontal fralabials to a point below center of eye; mental region; 15 scales on line between nostril and eye; bell-shaped with posterior median edge sharply scales in posterior loreal region slightly larger than angular, as wide as long, bordered posteriorly by scales in mid-orbital region; 32 scales across snout 2 postmentals; postmentals slightly longer than at level of third labial; 22 interorbital scales; eye wide, their median edges in broad contact, fol- large, its diameter contained in snout length about lowed by transverse row of 5 flattened scales. 1.5 times; eyelid with 2 inner rows of granules, a Dorsum with scattered, small, conical tubercles; median row of larger scales, and 1 row of large only 2 paravertebral rows evident, extending onto scales on the edge, the last 8 to 9 are pointed; ear neck; paravertebral rows separated by 3 to 5 irreg- large, its diameter contained in the eye diameter ular rows of small granules; 40 paravertebral tu- 2.7 times; scales on anterior edge of ear opening bercles between axilla and groin, separated by 0 to 2 granules; remaining dorsal tubercles not in distinct rows; dorsal granules irregular; postanal tubercles 3 on each side, not distinct; enlarged abdominal plaque absent; venter with 27 scales across belly, from throat to vent 69. Dorsal surface of upper arm with rounded, slightly elevated scales; forearm with slightly smaller scales of similar shape; dorsal surface of thigh and tibia with granular scales; claw long, visible from above and below; terminal toepad greatly enlarged, longer than wide, truncate; fourth toe lamellae 13 to 14; tail partially regene- rated, original tail stub with a few wide median scales on ventral surface, devoid of tubercles on dorsal surface, but with scattered, flattened tuber- cles at base. Measurements (in mm): Snout-vent length 57, axilla-groin length 26, length of leg 25, length of arm 17, length of tail 22 + 18, length of head 15.5, depth of head 5.7, width of head 11.2, length of snout 6.4, diameter of eye 4.3, diameter of ear 1.6, distance from eye to ear 5.7. Color in Alcohol: Dorsum medium gray-brown ground color with white speckles; 8 distinct broad chocolate brown transverse bars from base of head to base of tail, somewhat broken along midline; width of bars slightly larger than ground inter- spaces; lateral area of trunk as dorsal ground; arms and legs with dark chocolate ground suffused with light scales and small light spots, without definite pattern; top of head with suffusion of dark brown and small light blotches, no definite pattern; dark chocolate brown stripe from nostril to eye, blend- ing with head color posterior to eye, bordered be- FIGURE 18.—Holotypc of Phyllodactylus transversalis, adult low by a pale whitish line from first infralabial female, MCZ 130042, dorsal view. through ear; belly pale yellow. 46 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

VARIATION (based on paratype).—Similar to acters suggest long isolation from source areas holotype except in the following characters: 14 and hinder analysis of affinities. scales between nostril and eye; 24 interorbital Phyllodactylus transversalis shares many exter- scales; 6 supralabial and 5 infralabial scales to a nal morphological characters with the inaequalis point below center of eye; mental nearly triangu- group of Peru (Dixon and Huey, 1970:69). After lar; no contact between postmentals, followed pos- examining the type and paratype, however, J. R. teriorly by transverse row of 6 flattened scales; Dixon (pers. comm.) suggested that P. transver- venter with 30 scales across belly, from throat to salis may be closer to certain Mexican species (P. vent 67; 39 tubercles in paravertebral row between paucituberculatus, P. unctus, and P. delcampi). axilla and groin; fourth toe lamellae 15-14; tail There are no obvious affinities with Phyllodactylus regenerated 3 + 19; snout-vent length 36 mm. from the Galapagos. The source of this gecko is NATURAL HISTORY.—Both geckos were collected thus equivocal. during the day under rocks. Because of the topo- ETYMOLOGY.—The specific name refers to the dis- graphy of Malpelo, these geckos are necessarily tinctive dorsal cross-bands (Figure 18). scansorial. Dixon and Huey (1970:66) have shown that scansorial Phyllodactylus have expanded toe- pads, a characteristic of P. transversalis. Literature Cited Two large diurnal lizards are also present on Dixon, J. R. Malpelo (Diploglossus millepnnctatus and Anolis 1962. The Leaf-toed Geckos, Genus Phyllodactylus, of agassizi) and eat crustaceans as well as booby regur- Northeastern South America. Southwestern Natur- gitations and feces. Ants and crickets were found alist, 7:211-226. in the stomachs of P. transversalis (A. S. Rand, 1964a. Further Data on the Geckos (Phyllodactylus) of pers. comm.). Islands of the Extreme Southern Caribbean. South- western Naturalist, 9:203-206. Accompanying papers in this volume describe 1964b. The Systematics and Distribution of the Lizard Malpelo Island (Kiester and Hoffman, pp. 13-16) Genus Phyllodactylus of North and Central Amer- and the biology of the diurnal lizards (D. mille- ica. New Mexico State University Science Bulletin, punctatus, Kiester; A. agassizi, Rand, Gorman, 64:1-139. and Rand). Dixon, J. R., and R. B. Huey 1970. Systematics of the Lizards of the Gekkonid Genus ZOOGEOGRAPHIC RELATIONSHIPS.—The dorsal cross- Phyllodactylus of Mainland South America. Con- banding and the two rows of dorsal tubercles of tributions in Science of the Los Angeles County P. transversalis are distinctive. These unique char- Museum, 192:1-78, 14 figures. Electrophoretic Estimates of Genie Variation in, and the Relationships of, Anolis agassizi

T. Preston Webster

have been no recent immigrants bearing new al- ABSTRACT leles. On such a small island, secondary reductions in population size are possible. The founders prob- From data on proteins representing 30 genes, an average individual of Anolis agassizi is estimated ably encountered more extreme conditions on to be heterozygous at 2.1 percent of its loci. This Malpelo than those to which they were adapted, low value is similar to that found in other solitary even if the source population, like those anoles anoles and in some anole populations of rela- that have been successful colonists in the Carib- tively recent origin, but it contrasts sharply with bean (Williams, 1969), occurred in open, edge results from some Greater Antillian species. Very situations. Anolis agassizi, however, is a long estab- little genetic similarity was found in the compari- lished population that has perhaps experienced son of A. agassizi to five congeners. pressures and had time to reacquire heterozygosity. In this study, genie variation in A. agassizi was estimated by electrophoretic examination of enzy- Introduction matic and non-enzymatic proteins. Since there is An insular lizard population may have relatively no prediction to which this result can be compared, little genie variation as a result of random pro- it is considered in the context of similar estimates cesses and natural selection acting during its estab- from other anole populations. Secondarily, many lishment. The small size of the propagule, perhaps of the same proteins were used to estimate the only a single impregnated female, should limit genetic difference between A. agassizi and five variation. This holds whether the alleles carried congeners. by the founder or founders are a random or a se- ACKNOWLEDGMENTS.—I thank the following for lected sample of those in the parental population. their contributions to this study: G. C. Gorman of Subsequently, random allelic loss (genetic drift) the University of California at Los Angeles, T. A. during the initial or secondary periods of small Jenssen of Virginia Polytechnic Institute and State population size and selection would further reduce University, A. R. Kiester of the Museum of Com- variation. A low level of heterozygosity could en- parative Zoology, and A. S. Rand of the Smithson- dure, but in a persistent population many new ian Tropical Research Institute for collecting polymorphisms might accumulate. specimens. J. B. Graham and A. S. Rand of the A considerable loss of genie variation may have Smithsonian Tropical Research Institute, and E. E. occurred during the arrival and life of Anolis Williams of the Museum of Comparative Zoology agassizi Stejneger on Malpelo Island. Malpelo may commented on the manuscript. This study was have received anoline immigrants only once and, conducted in part while I was an NSF predoctoral as the source population is extinct, there certainly fellow and was supported by NSF grant B 019801X to E. E. Williams, Museum of Comparative T. Preston Webster, Museum of Comparative Zoology, Har- vard University, Cambridge, Massachusetts 02138. Zoology.

47 48 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Materials and Methods tion). Single individuals or pooled extracts were used in these comparisons, but the variation in Plasma and hemolysate were obtained from 32 each population has been studied and considered. A. agassizi collected at the first landing site on the west side of Malpelo. Six individuals were used for chromosome studies and preserved, and the re- Genie Variation maining 26 were homogenized in distilled water In A. agassizi variation was detected at 4 of the (1:2, W:V). Homogenates were centrifuged and 30 loci. One of 26 individuals was heterozygous stored at -80°C. for Protein B, 2 of 26 were heterozygous for one of Electrostarch was used for horizontal starch gel the malate dehydrogenases, and 2 of 26 were heter- electrophoresis. Each protein was examined in ozygous for one of the peptidases. Transferrin was as many electrophoretic environments as conven- represented by 3 variants, with 14 of 32 individuals ient and feasible. In several cases this multiplicity heterozygous. The remaining 26 genes appear to of observation increased the number of mobility be monomorphic. Thus the average individual of differences detected. Many of the buffer systems this population is estimated to be heterozygous at and protein assays were those previously used in 2.1 percent of its loci (H = 0.021). studies of Anolis (Webster, Selander, and Yang, This result can be evaluated within the context 1972; Webster and Burns, 1973). Additional tech- of a broad range of estimates of H from anole niques are those commonly used in studies of ver- populations (Table 3). Since a large proportion tebrate proteins (Brewer, 1970; Shaw and Prasad, of protein polymorphism is expected (Lewontin 1970; Selander et al., 1971). and Hubby, 1966) and known (Boyer, 1972) to be Twenty-eight enzymatic and nonenzymatic pro- undetectable by electrophoresis, such values are teins were examined: leucine aminopeptidase, two surely underestimates of the actual extent of genie peptidases, fumarase, creatine kinase, phospho- variation. Their comparison requires both that the hexose isomerase, eight dehydrogenases (lactate, electrophoretically determined H be considered two malate, two isocitrate, alcohol, a-glycerophos- as having some value as an index to actual genie phate, and 6-phosphogluconate), indophenol variation among structural loci or within the oxidase, two aspartate aminotransferases, two phos- genome as a whole and that some standard error phoglucomutases, an esterase, three "general pro- or arbitrary criterion for difference be assigned. teins" (A,B,C), hemoglobin, transferrin, and three From a joint consideration of morphological plasma proteins. Like other anoles (Webster, Sel- variation and electrophoretically determined H, ander, and Yang, 1972), A. agassizi appears to have Soult et al. (1973) concluded that H has consid- four or five genes encoding the polypeptides that erable power as an index to genie variation, even are subunits of hemoglobin, but it is treated here when only 20 to 25 loci are examined. Two other as if it were under two-gene control. Lactate dehy- studies support this relationship. Kluge and Ker- drogenase is encoded by two genes. Since each of foot (1973) examined several sets of vertebrate the remaining proteins appears to be encoded by populations, finding that within each the average a single gene, the products of 30 genes have been variability of a morphological character is corre- examined. lated with the extent to which there has been diver- Many of the proteins studied in A. agassizi were gence in its mean. Webster, Selander, and Yang, compared with their apparent homologues in these (1972) provided the complementary relationship, congeners: A. garmani (Jamaica: Mandeville), although their measure of morphological diver- A. luciae (St. Lucia: Vigie Beach), A. squamulatus gence was relatively crude. They found that among (Venezuela: Parque Nacional Rancho Grande Anolis species, those with higher values of H were "Henry Pihier"), A. marmoratus (Guadeloupe: differentiated into more geographic races. The 3.6 km SW Capesterre), and A. singularis (Haiti: genetic base for both morphological variation and Savane Zombi). To obtain an estimate of differ- divergence probably involves many more genes entiation between members of the same species than the handful examined in the direct estima- group, A. luciae was compared to A. extremus (St. tion of H. But whether these relationships are Lucia: Vigie Beach, a recently introduced popula- simply evidence for the strength of H as an index NUMBER 176 49 to heterozygosity at structural loci or if they extend treatment. I suggest that an H of 0.02 should be it to other components of the genome is a matter considered indicative of less genie variation than for further study. occurs in a population with an H of 0.08 or more. The estimates of H for anoles are based on simi- The reader is encouraged to provide his own lar sets of genes that received similar technical criterion.

TABLE 3.—Electrophoretic estimates of genie variation (H) in populations of Anolis lizards

Species Populations Loci H Source

CONTINENTAL SPECIES

A. carolinensis 29 0.036-0.057 Webster, Selander, and Yang, Southeastern U.S. 1972:529

GREATER ANTILLEAN SPECIES

A. stratulus 22 0.024 Soute, et al., 1973, fig. 1 Puerto Rico A. poncensis 22 0.035 Puerto Rico A. evermanni 22 0.070 Puerto Rico A. gundlachi 22 0.082 Puerto Rico A. krugi 22 0.104 Puerto Rico A. pulchellus 22 0.129 Puerto Rico A. porcatus 24 0.079 Webster, in prep. Havana, Cuba A. distichus dominicensis . 21 0.090 Bon Repos, Haiti

SOLITARY SPECIES A. extremus 22 0.036 Soute, et al., 1973, fig. 1 Barbados A. roquet 21 0.041 Martinique A. luciae 27 0.012 Webster, in prep. St. Lucia A. marmoratus 21-23 0.014-0.033 Guadeloupe A. agassizi 31 0.021 Webster, herein Malpelo

RELATIVELY RECENT POPULATIONS A. carolinensis 29 0.064 Webster, Selander, and Yang, South Bimini 1972:529 A. angusticeps 25 0.000 South Bimini A. distichus 27 0.043 South Bimini A. sagrei 26 0.009 South Bimini A. sagrei 28 0.052 Lister and Webster, in prep. Swan Island 50 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Unfortunately, H has not been estimated for interesting to note that no variation was observed any of the South or Central American anoles. for three esterases (only one of which was in- Whether A. carolinensis, a secondarily continental cluded in the estimate of H). These enzymes usu- species, has more genie variation than A. agassizi ally show complex patterns of variation in anole is questionable (Table 3). populations, including some with little genie Many of the Greater Antillean species represent variation. That the estimates of H for A. agassizi stocks long established on islands so large and bio- and several other solitary anoles are fairly uniform logically diverse that they probably have a conti- suggests that the net influence on genie variation nental influence on genie variation. Estimates of may be quite similar in these older populations. H have a broad range, but many of the values are If this consistency is confirmed by further studies higher than those obtained from anoles in any of solitary species, it will be appropriate to con- other evolutionary or ecological context. Source sider selective explanations. island populations of two good colonists in the Caribbean, A. porcatus and A. distichus, appear to be more variable than A. agassizi. Genetic Differentiation Like A. agassizi, many of the species of the Lesser On the basis of a protein phenotype including Antilles have been long established and occur in the products of 24 genes, A. agassizi is quite distinct the absence of congeners. Their islands, however, from all five of the species with which it was com- tend to be larger than Malpelo and provide a more pared (Table 4). Two of the species are also mem- typical anole environment. Estimates of H from bers of the latifrons series of the alpha section four species are nonetheless similar to that from (Etheridge, 1960): A. luciae (St. Lucia, Lesser A. agassizi. Antilles), like A. agassizi, has retained caudal auto- The South Bimini and Swan Island popula- tomy, but A. squamulatus (Mainland, Venezuela) tions represent colonizations that are relatively re- has lost it. Since the latifrons series is defined by cent (Williams, 1969). It is in such populations character states that are ancestral for the alpha that any loss of heterozygosity during and soon section (Etheridge, 1960), it is not surprising that after founding should be most visible. The popula- the species within it should have diverged to such tions of A. sagrei and A. angnsticeps on South an extent. Even two members of the apparently Bimini have little or no detectable genie variation, monophyletic roquet species group (Gorman and although the A. angusticeps have some esterase Atkins, 1969), A. luciae and A. extremus (native variation that was refractory to interpretation and to Barbados, Lesser Antilles), differ completely at was omitted from the reported H of 0.00. Else- 17 of the 29 genes examined. Anolis singularis is where in the Bahamas and the West Indies, how- as distinct from A. agassizi as are members of the ever, these species have more variation (Webster, latifrons series; and the fourth alpha section species, in prep.; Lister and Webster, in prep.; see Table A. marmoratus, has no shared electrophoretic mo- 3, A. sagrei, Swan Island). These results and those bilities. The beta section A. garmani is no more for insular populations of Peromyscus (Selander distinct than some of the alpha section species. et al., 1971) and cave populations of Astyanax These findings shed no new light on the affinities (Avise and Selander, 1972) suggest that a major of A. agassizi. Except for very closely related popu- loss of heterozygosity may frequently occur during lations, electrophoretically determined genetic dif- the colonization of islands; however, the South ference is always an underestimate of genetic dis- Bimini populations of A. distichus and A. caroli- tance. When the species compared are as distantly nensis seem to have been less affected. related as these, it even becomes unreliable as an Since several Greater Antillean species appear index to genetic distance. Thus the small differ- to have as little genie variation as A. agassizi, there ences in similarity that were observed in this study is no compelling reason for concluding that it lost should be considered meaningless. As concluded much during and subsequent to its colonization of previously (Webster, Selander, and Yang, 1972), Malpelo. If it did, it is likely that both random when only 20 to 30 proteins are examined, electro- and selective processes played a role, but their rel- phoresis has little value in anoline taxonomy be- ative importance cannot be determined. It is yond the species group. NUMBER 176 51

TABLE 4.—Similarity (j*,; Nei, 1972) of eleclrophoretic mobilities of proteins in comparisons between Anolis agassizi and five congeners

A. A. A. A. A. Protein garmani marmoratus singular is squamulatus luciae aspartate aminotransferase-1 0.00 0.00 0.00 1.00 0.00 phosphohexose isomerase 0.00 0.00 1.00 0.00 1.00 leucine aminopeptidase 0.00 0.00 1.00 1.00 1.00 isocitrate dehydrogenase-2 1.00 0.00 0.00 0.00 0.00 lactatc dehydrogenase-1 0.00 0.00 0.00 1.00 0.00 lactate dehydrogenase-2 0.00 0.00 0.00 1.00 0.00 malate dehydrogenase-1 1.00 0.00 1.00 0.00 0.00 malate dehydrogenasc-2 0.00 0.00 0.00 0.00 1.00 ct-glycerophosphate dehydrogenase 0.00 0.00 0.00 0.50 1.00 Sum of Similarities 2.00 0.00 3.00 4.50 4.00

Values for the following are 0.00: 6-phosphogluconate dehydrogenase, aspartate aminotrans- ferase-2, peptidase, isocitrate dehydrogenase-1, two phosphoglucomutases, fumarase, indophenol oxidase, alcohol dehydrogenase, albumin, proteins A, B, and C.

Literature Cited Lister, B. C, and T. P. Webster In prep. Genie Morphologic Variation in Island Popu- Avise, J. C, and R. K. Selander lations of the Lizard Anolis sagrei. 1972. Evolutionary Genetics of Cave-Dwelling Fishes of Nei, M. the Genus Astyanax. Evolution, 26:1-19. 1972. Genetic Distance between Populations. The Ameri- Boyer, S. H. can Naturalist, 106:283-292. 1972. Extraordinary Incidence of Electrophoretically Si- Selander, R. K., M. H. Smith, S. Y. Yang, W. E. Johnson, and lent Genetic Polymorphisms. Nature, 239:453-454. J. B. Gentry Brewer, G. J. 1971. Biochemical Polymorphism and Systematics in the 1970. An Introduction io Isozyme Techniques. New York: Genus Peromyscus, I: Variation in the Old-field Academic Press. Mouse (Peromyscus polionotus). Studies in Genetics Etheridge, R. (University of Texas Publication 7103), 6:49-90. 1960. The Relationships of the Anoles (Reptilia: Sauria: Shaw, C. R., and R. Prasad Igiianidae): An Interpretation Based on Skeletal 1970. Starch Gel Electrophoresis of Enzymes: A Compila- Morphology. Unpublished Ph.D. dissertation, Uni- tion of Recipes. Biochemical Genetics, 4:297-320. versity of Michigan. Ann Arbor: University Micro- Soule, M. E., S. Y. Yang, M. G. W. Weiler, and G. C. Gorman films Inc. 1973. Island Lizards: The Genetic-Phenetic Variation Gorman, G. C, and L. Atkins Correlation. Nature, 242:191-193, 2 figures. 1969. The Zoogeography of Lesser Antillean Anolis Liz- Webster, T. P. ards: An Analysis Based upon Chromosomes and In prep. Electrophoretic Estimates of Genie Variation Lactic Dehydrogenases. Bulletin of the Museum of in Populations of West Indian Anolis Lizards. Comparative Zoology, 138:53-80, 15 figures. Webster, T. P., and J. M. Burns Kluge, A. G., and W. C. Kerfoot 1973. Dewlap Color Variation and Electrophoretically 1973. The Predictability and Regularity of Character Detected Sibling Species in a Haitian Lizard, Anolis Divergence. The American Naturalist, 107:426-442, brevirostris. Evolution, 27:368-377. 10 figures. Webster, T. P., R. K. Selander, and S. Y. Yang Lewontin, R. C, and J. L. Hubby 1972. Genetic Variability and Similarity in the Anolis 1966. A Molecular Approach to the Study of Genie Lizards of Bimini. Evolution, 26:523-535. 1 figure. Heterozygosity in Naural Populations, II: Amount Williams, E. E. of Variation and Degree of Heterozygosity in Nat- 1969. The Ecology of Colonization as Seen in the Zoo- ural Populations of Drosophila pseudoobscura. geography of Anoline Lizards on Small Islands. Genetics, 54:595-609. Quarterly Rex'ieu' of Biology, 44:345-389, 17 figures. Notes on the Chromosomes of Anolis agassizi (Sauria: Iguanidae) and Diploglossus millepunctatus (Sauria: Anguidae)

Brad Stamm and George C. Gorman

have this same karyotype. Although many species ABSTRACT groups of Anolis show highly derived chromoso- mal morphology, a large section of the genus has The karyotypes of Anolis agassizi and Diploglos- been conservative and retained the ancestral karyo- sus millepunctatus are very similar. When com- type (for review, see Gorman, 1973). pared to other species in their respective groups, the karyotypes of the two Malpelo lizards are con- Anolis agassizi is also a member of the latifrons sidered primitive. group. This species group has representatives on the South American mainland as well as the Lesser Antilles (the term roquet group follows Under- During the past ten years description of reptil- wood, 1959). The Malpelo anole and the roquet ian karyotypes has moved from relative novelty to group all possess fracture planes in the caudal ver- virtually routine. Chromosome preparations of tebrae, hence caudal autotomy is commonplace. Malpelo lizards were made directly from testes fol- This character appears to have been secondarily lowing methods described by Gorman, Atkins, and lost among living mainland forms. Thus the pre- Holzinger (1967). sumed ancestors of A. agassizi and the Lesser Antil- Both Anolis agassizi and Diploglossus millepunc- lean latifrons anoles are extinct, and the living ctatus have kartotypes consisting of six pairs of South American anoles are more derived (Ethe- metacentric macrochromosomes and 12 pairs of ridge, 1960). microchromosomes, 2n = 36 (Figure \9a,b). Chromosomes tend to support this contention. Among anguid lizards previously studied, this Many members of the roquet group have a karyo- karyotype was found in Diploglossus costatus (the type virtually indistinguishable from A. agassizi. only other Diploglossus studied to date) but not Only a single mainland latifrons member has been in the genera Gerrhonotus, Anguis, and Ophisau- previously studied, A. jacare, and it has a derived rus (Bury, Gorman, and Lynch, 1969). McDowell karyotype of 2n = 32, with a reduction in chromo- and Bogert (1954) considered the diploglossines some number (Williams et al., 1970). For compari- the most primitive of the anguids. son we add data on two other mainland latifrons Similarly, representatives of the most primitive species, A. frenatus from Panama, and A. squamu- group of Anolis, the Lesser Antillean roquet group latus from Rancho Grande, Venezuela. Webster, (alpha group, latifrons series; Etheridge, 1960), Hall, and Williams, (1972) argued convincingly that within Anolis, 12 metacentric macrochromo- somes and 24 distinctly smaller microchromosomes Brad Stamm, Museum of Comparative Zoology, Harvard Uni- versity, Cambridge, Massachusetts 02138. George C. Gorman, is the primitive condition. Anolis squamulatus (not Department of Biology, University of California, Los Angeles, illustrated) has the typical primitive complement. California 90024. Anolis frenatus has 2n = 36, which is the typical

52 NUMBER 176

ft*

IOJJ B S IIU « «"

• FIGURE 19.—Lizard karyotypes: A, Anolis agassizi £, 2n = 36, 12 ca. metacentric macrochromo- somes and 24 microchromosomes (pair 6 appears heteromorphic in this preparation, but we lack sufficient numbers of metaphase plates, and have no data on females to confirm this); B, Diploglossus millepunctatus $, 2n = 36, 12 ca. metacentric macrochromosomes and 24 microchromosomes; c, Anolis frenatus $, 2n = 36, size break between pairs 4 and 5, macro- chromosomes do not break sharply from microchromosomes; pair 7 is intermediate in size. saurian diploid number; but there is an enlarged among the most primitive representatives, as deter- metacentric pair of "microchromosomes," and a mined osteologically (Gorman, 1973). break in size between pairs four and five implying ACKNOWLEDGMENTS.—Travel to Malpelo by Gor- a derived karyotype (Figure 19c). man was supported by a grant from the National The similarity between A. agassizi and the Geographic Society. Laboratory work supported by roquet group of Lesser Antillean Anolis in both NSF Grant B 019801X to Dr. E. E. Williams. osteology and karyotype substantiates the belief that both retain primitive character states. We do Literature Cited not argue for common origin between these two groups. The similarity in karyotype between the Bury, R. B., G. C. Gorman, and J. F. Lynch 1969. Karyotypic Data for Five Species of Anguid Lizards. Diploglossus and Anolis of Malpelo is fortuitous. Experientia, 25:314-316. This chromosomal formula is common to repre- Etheridge, R. sentatives of numerous lizard families, often 1960. The Relationships of the Anoles (Rcptilia: Sauria: 54 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Iguanidae): An Interpretation Based on Skeletal McDowell, S. B., and C. M. Bogert Morphology. Unpublished Ph.D. dissertation, Uni- 1954. The Systematic Position of Lanthanotus and the versity of Michigan. Ann Arbor: University Micro- Affinities of the Anguinomorphan Lizards. Bulletin films Inc. of the American Museum of Natural History, 105: 1-142. Gorman, G. C. Underwood, G. 1973. The Chromosomes of the Reptilia: A Cytotaxo- 1959. The Anoles of the Eastern Caribbean, Part III: nomic Interpretation. Chapter 12 in A. B. Chiarelli Revisionary Notes. Bulletin (Museum of Compara- and E. Capanna, editors, Cytotaxonomy and Verte- tive Zoology), 121:191-226. brate EvohUion. New York: Academic Press. Webster, T. P., W. P. Hall, and E. E. Williams 1972. Fission in the Evolution of a Lizard Karyotype. Gorman, G. C, L. Atkins, and T. Holzinger Science, 177:611-613. 1967. New Karyotypic Data on 15 Genera of Lizards in Williams, E. E., O. A. Reig, P. Kiblisky, and C. Rivero-Blanco the Family Iguanidae, with a Discussion of Taxo- 1970. Anolis jacare Boulenger: A "Solitary" Anole from nomic and Cytological Implications. Cytogenetics, the Andes of Venezuela. Brevoria (Museum of 6:286-299. Comparative Zoology), 353:1-15. Sub tidal Communities of Malpelo Island

Charles Birkeland, David L. Meyer, James P. Stames, and Caryl L. Buford

the Malpelo subtidal consists almost entirely of ABSTRACT vertical rock walls, surface area data were easily acquired from underwater photographic transects, The subtidal communities of Malpelo are described using color and black-and-white film. Relative sur- in terms of the pattern of spatial coverage of sessile face area coverage was then estimated by tabula- organisms. The Malpelo subtidal is characterized by a prevalence of large barnacles (Balanus penin- tion of randomly located points on the projected sularis) on steeply sloping rock walls and a highly transparencies and by planimetry (Tables 5,6,8). structured community of hermatypic corals on Photographs used for the quantitative survey are gradually sloping substrata, both communities ex- on file with C. Birkeland. We found that such a tending to a depth of about 30 meters. The rock method was most efficient for acquiring basic infor- wall communities of Malpelo have significantly more free space available for recruitment (38.6%) mation from a large, unfamiliar, area within a than do comparable communities in continental limited time. Photographic transects have the fol- regions (7.3%). Size distribution suggests that re- lowing additional advantages for primary survey of cruitment is a greater risk, but conditions for adults an unexplored underwater area: First, size distri- are better on Malpelo than in continental regions. bution of the predominant sessile organisms can Observations suggesting the importance of fish sometimes be used to deduce aspects of the age or grazing are discussed. An expanded depth range at Malpelo provides clearer zonation of the coral recent history of a community, such as the fre- community than along the Panamanian coast. Al- quency or heterogeneity of disruptive factors. though coral growth is rich in terms of surface Second, the extent of available unoccupied space coverage and size of colonies, a true reef is not in a habitat is often an important limiting factor formed. for sessile organisms. Finally, the habitat or nur- sery of certain motile animals is often defined by specific sessile organisms and, therefore, the sessile Introduction organisms may provide the best definition of the community as a whole (e.g., coral reef, oyster bed, The subtidal communities at Malpelo had never kelp bed, sea grass bed, etc.). been observed before our brief visit, and thus our Dives were made at each of eight stations (Fig- objective was to describe these communities in ure 20). At all stations except Station 6, the basal- quantative terms that would be useful for compari- tic rock wall sloped steeply, sometimes vertically, sons to other areas and for any return visits to to a depth of about 37 meters on Malpelo itself or the island. about 49 meters on the offshore pinnacles (Stations We have utilized the relative surface area cover- 4 and 5). Sometimes a talus slope of large boulders age of macroinvertebrate species as the quantita- provided a more gentle slope in deeper water (be- tive expression of community composition. Since low about 20 m; Stations 3, 7, and 8). Station 6 was the only area found with a gradual slope and it Charles Birkeland, David L. Meyer, James P. Stames, and Caryl L. Buford, Smithsonian Tropical Research Institute, was the only location of a rich community of her- P. O. Box 2072, Balboa, Canal Zone. matypic corals. 55 56 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

TABLE 5.—Percent coverage of primary substrata on rock walls of Malpelo Island, 1 to 3 March 1972

Depth range (meters) 0-3 3-6 6-27 27-37 37-50 No. of points (sample size x 50) 450 1350 2800 500 350

"OPEN SPACE" bare rock or sediment .. 1.1 1.6 4.6 5.2 5.4 •crustose coralline algae 17.8 30.1 34.0 24.6 20.3

"OCCUPIED SPACE"

"Balanus peninsularis (live) 5.3 14.0 5.8 1.8 1.1 *Balanus peninsularis (dead, test) 16.8 23.6 8.7 6.6 2.8 Balanus peninsularis (total) 22.1 37.6 14.5 8.4 3.9 *Lobophora variegata 11.6 3.8 1.0 •"algal scuz" 4.9 6.0 5.2 18.8 12.0 "red algal crust" - 0.7 1.2 2.2 1.4 branching coralline algae (mostly Amphiroa spp.) 5.1 0.1 0.8 _ - *Polyfibrospongia sp 30.7 6.7 1.1 •orange flat sponge _ 0.1 1.6 6.0 13.1 yellow flat sponge - 0.1 0.5 0.2 0.6 red flat sponge 0.4 0.1 1.6 - 0.6 •white flat sponge _ _ 0.3 6.2 11.1 white lumpy sponge _ _ 0.5 - 0.6 blue sponge _ 0.1 - - - greenish gray sponge _ _ _ 05 - gray-brown sponge _ _ - - 0.3 hydroid _ _ - - 1.7 bryoz'oan _ _ _ _ 7.7 Lichenopora sp 0.1 + - — Pacifigorgia spp 2.9 4.0 10.0 2.6 0.0 Lophogorgia alba _ 0.6 - - gorgonacean _ - 6.6 Errinopora pourtalesii _ 0.3 7.6 2.3 Tubastraea aurea 0.3 2.4 0.6 1.1 Tubastraea aurea (dead) 0.1 0.5 0.2 - Pocillopora robusta 0.4 0.9 - - Porites spp _ 9.0 - - Astrangia sp _ - 1.4 - Pavona varians - 0.5 - - Ostrea irridescens - 3.0 6.4 8.0 Ostrea irridescens (dead) 0.1 - - - Ostrea fisheri - - 0.8 - Didemnum candidum 3.3 7.2 1.7 5.4 2.6 Didemnum sp 0.4 3.3 1.4 + Polyandrocarpa tincta _ _ 0.8 _

• Predominant categories. An organism that occupies over 10 percent of the primary substrata is arbitrarily denned as predominant. + indicates that the organism was present in photographs of the area, but it did not appear as data in surface-coverage counts from contact of random points. NUMBER 176 57

TABLE 6.—Proportion of space open for invasion, as opposed to space occupied by macroscopic sessile organisms other than crustose coralline algae, comparing Malpelo Island and the continental coast of the eastern tropical Pacific (coverage estimated by number of rectangular coordinate points—taken from a random number table—falling on the image of each category; 50 points tallied from each quadrat photograph; several points in the continental photographs fell into shadow so were not determined)

No. of Open Occupied Total no. Mean percent Region quadrats space space of points of open space Oceanic (Malpelo) 56 1081 1719 2800 38.6 Continental (Taboguilla Is., Panama and Playas del Coco, Costa Rica) 213 2705 2918 x* = 800, 1 d.f. Reject Ho: conclude significantly more "open space" exists in subtidal communities of Malpelo than in comparable communities ©t mainland.

ACKNOWLEDGMENTS.—We were assisted during crabs; and Janet Haig, also with the Allan Han- diving work at Malpelo by D. B. Macurda, Jr., of cock Foundation, identified the porcellanid and the University of Michigan and W. L. Smith of diogenid crabs. L. G. Abele of Florida State Uni- the State University of New York at Stony Brook. versity completed the list of crabs and provided a We thank D. Dexter of San Diego State University, complete list for the decapods. Joseph Rosewater and J. B. Graham of the Smithsonian Tropical of the National Museum of Natural History, Research Institute for constructive review of the Smithsonian Institution, confirmed and corrected manuscript. The list of invertebrates was compiled the lists of mollusks. Alan J. Kohn of the Univer- with the generous assistance of ten authorities. sity of Washington verified the species of Conus Maureen Downey of the National Museum of Nat- and Dora P. Henry, also from the University of ural History, Smithsonian Institution, contributed Washington, identified the barnacles. Peter W. her list of asteroids to the Appendix of this paper. Glynn of the Smithsonian Tropical Research Insti- John S. Garth of the Allan Hancock Foundation tute confirmed and corrected our identification of identified the xanthid, dynomenid, and majid scleractinian corals. J. W. Wells of Cornell Uni-

TABLE 7.—Colony-polyp-number distributions in populations of Tubastraea aurea at Malpelo compared with colony polyp number distributions in two areas near the continental coast (counts were made from quadrat photographs)

A B C Mean no. Total no. No. solitary No. colonies polyps per Region colonies polyps with 10 or colony examined more polyps (±SE)» Oceanic (Malpelo) 372 88 102 11.6±0.8 Continental (Taboguilla Is., Panama and Playas del Coco, Costa Rica) 408 135 92 10.0±0.8 • Solitary polyps not included in analysis (N = A —B). X* = 6.78, 2 d.f. (categories compared were B, C, and A-B-C). Reject Ho: conclude significantly fewer solitary polyps in proportion to number of larger colonies (> 10 polyps) in a single sample of Malpelo populations when compared with a 3 year sample of several mainland populations. 58 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

TABLE 8.—Zonation of the coral community of Malpelo Station 6: percent cov- erage by each species, bare rock, or dead coral substrata (data acquired by plani- meter measurements on tracings of projected photographs)

Depth range (meters) 92 & 10.7 123 13.8 & 15.4 16.9 18.4 26.1 & 27.6 No. of planimeter samples 8 10 13

PERCENT OF SURFACE

Live hermatypic corals 57.2 69.3 74.8 89.2 58.6 42.7 Dead coral substrata 24.5 27.6 22.7 10.8 29.8 18.0 18.3 3.1 2.5 — 11.6 38.4 Rock or sediment substrata

PERCENT OF LIVE CORAL SURFACE 94.3 80.5 14.4 2.4 Pocillopora robusta 5.7 19.5 77.3 43.2 80.5 47.4 Porites spp _ _ 8.3 54.4 19.5 _ Pavona clivosa 52.6 Agariciella planulata

versity identified the hydrocoral and George Hech- tel of the State University of New York at Stony Brook identified the sponge.

Rock Wall Communities The steep rock walls (Figure 21) were predomi- nantly occupied by the large barnacle Balanus peninsularis (Table 5). Sixty to 80 percent of the Balanus space was occupied by empty tests (Figure 22) which remained on the walls after the death of the barnacles because their predators were snails (probably Thais planospira and possibly Purpura pansa) rather than sea stars. These tests provide a major source of shelter or habitat for many poly- chaetes, mollusks, crustaceans, and fishes (Ento- macrodus chiostictus) and a substratum for hy- droids, sponges, and tunicates. The prevalence of barnacles is reflected in the white sand sediment which encircles the island at the base of the rock wall. This sand is composed almost entirely of frag- ments of Balanus tests. A clear zonation was observed on the vertical 500 Meters walls (Table 5). To a depth of 3 meters, the large black sponge Polyfibrospongia sp. occupied over 30 percent of the space. The brown alga Lobo- phora variegata was also prevalent above 3 meters. o Both Polyfibrospongia and Lobophora decreased rapidly in abundance between 3 and 6 meters 4 where Balanus tests occupied nearly 37 percent of FIGURE 20.—Subtidal collection stations at Malpelo Island. the space. From 6 to 27 meters the substratum was NUMBER 176 59

FICURE 21.—Vertical rock wall community; Station 3, 15 m depth. Prevalent animals are Pacifi- gorgia media (left), Porites sp. (right), and Balanus peninsularis. 60 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

very clear of "scuz" (filamentous algae, hydroids, panamense, and the sea stars, Tamaria sp. and etc.) and was covered mostly with crustose coral- Narcissia gracilis. line algae and scattered patches of anthozoans, i.e., The most striking aspect of the occupation of hermatypic corals (mostly Poriles spp.) an aher- the subtidal substrata at Malpelo was the amount matype (Tubastraea aurea), and gorgonaceans of "open" or "free" space. Table 6 compares the (Pacifigorgia spp.). Below 27 meters, filamentous amount of open space in 59 permanent quadrat algae, small flat encrusting sponges, oysters (Ostrea photographs from small Pacific islands less than irridescens), and the violet branching hydrocorals 14 km from the mainland in Costa Rica and Pana- (Errinopora pourtalesii) became prevalent. Al- ma with the amount in 56 survey photographs though occupying only 7.6 percent of the substra- from Malpelo. Bare rock or space occupied by tum, the vivid violet hydrocorals appeared as a crustose coralline algae is available as attachment "field of Errinopora" below 27 meters. Certain space for other sessile or sedentary organisms. Be- motile animals were also found only at depths of tween the depths of 6 and 27 meters near the main- 27 meters or below: the basket star, Astrodictyum land, an average of 7.S percent of the space is

FICURE 22.—A typical example of the rock wall community; Station 7, 10 m depth (1 = Balanus peninsularis, 2 = Porites sp., 3 = Tubastraea aurea, 4 = Pacifigorgia media). NUMBER 176 61

available as open space in rock wall communities. lives under less crowded conditions. In other words, In comparable rock wall communities at Malpelo, an individual may be subjected to greater risk ini- 38.6 percent of the space is available. tially, but may attain greater fecundity if it sur- Many of the prevalent sessile organisms were vives until adulthood. common to continental and oceanic rock wall com- The reason for reduced recruitment success is munities. The oceanic Malpelo populations for unknown but two possible causes can be hypothe- both Tubastraea aurea and Lophogorgia alba ap- sized: isolation from other populations and heavy peared to be characterized by size distributions in grazing by fishes. Isolation may reduce recruitment which larger individual colonies were more preva- of shallow benthic organisms because most, per- lent and less recruitment was observed in compari- haps 85 percent of the species (Thorson, 1961:456), son with the continental populations. The number have pelagic larval stages for which the mean dura- of polyps in each Tubastraea aurea colony in Mal- tion is about 2 weeks. Only about 30 percent have pelo photographs were counted and the polyp- planktonic stages longer than 5 weeks (Thorson, number distribution (presumed to be a form of 1961:459). Malpelo is small, isolated, and sur- age distribution) was compared with polyp-number rounded by deep water. It seems likely that a large distribution in Panama mainland quadrats (Table portion of the larvae may be carried away before 7). The distributions differ significantly by chi settlement can occur. The water moving past Mal- square when the proportions of recruitment for pelo comes predominately from the northeast solitary polyps, colonies of 2 to 9 polyps, and colo- where the nearest shallow water is around the Azu- nies of 10 or more polyps are compared. The main- ero Peninsula, 367 km away. The fauna of Cobb land data are from several areas through all sea- Seamount, 430 km west of Washington State in the sons over a 3-year period while the Malpelo data northern Pacific, has large populations of brood- are from a single survey. The significant difference ing species usually found only in the intertidal or in age distribution could be the result of a particu- shallow subtidal (Birkeland, 1971). Once a popu- larly poor year, or time of year, in which the Mal- lation of a brooding species is established on an pelo survey was made, rather than to differences isolated pinnacle, it may have a generally better between oceanic and continental populations. If chance of maintaining a population than species the solitary polyps are excluded from the analy- with planktonic larvae. The emplacement of set- sis, the polyp-number distributions do not differ tling plates at Malpelo would provide particularly significantly by Mest. Nevertheless, we predict that interesting quantitative data for comparison with more data from further surveys would confirm a data from plates set nearer the mainland. general trend towards proportionately less recruit- The sessile ciliate protozoan Folliculina was ment and longer adult life in oceanic populations. found on some of the Malpelo algae and it may A similar trend is evident for closely related have arrived by rafting on drifting algae. However, species living on the island and mainland, respec- Scheltema (1973) has recently presented data show- tively. For instance, Balanus trigonus (adults ca. ing that dispersal of this organism can also occur 0.8 cm basal diameter) is the prevalent species in by attachment to the shells of veliger larvae. Fol- continental areas examined while Balanus penin- liculina is often found on the larger species of sularis (adults ca. 6 cm basal diameter) is the pre- veligers which are "long-distance larvae from shoal- valent species at Malpelo. The same generalization water bottom-dwelling organisms of the shelf" could be drawn comparing colonial tunicates of (Scheltema 1973:123). the genus Didemnum. The effects of grazing fish in producing open In summary, the oceanic rock wall communities space is implied by some general observations con- at Malpelo had over five times as much open space cerning algae and is based on the assumption that as the continental rock wall communities. Since grazers of algae also remove recently settled ben- space is a limiting resource for sessile benthic or- thic animals. Algae and juvenile sessile organisms ganisms on solid substrata, the abundant avail- were very sparse on Malpelo (Tables 5, 6), yet the ability of open space at Malpelo implies that re- exceptionally clear water should provide favorable cruitment is a more difficult problem for these pop- growth conditions for algae. As an example, one ulations. Once established, however, an individual small sprig of Sargassum was found at 46 m. The 62 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

I

FIGURE 23.—"Shingle" formation of Porites sp. on the rock wall community. NUMBER 176 63 substrata in shallower water presumably had the major factor contributing to the strong faunal enough available light and other conditions for similarities between Malpelo and the Galapagos growth of Sargassum. The large schools of acanthu- Islands. Many familiar tropical eastern Pacific rids may have kept the algae (and settling larvae) mainland species were conspicuous by their ab- under heavy grazing pressure. This hypothesis sence: Muricea and antipatharians (the usually could be simply tested by the emplacement of fish present arborescent coelenterates), and the sea stars exclusion cages for 4 to 12 weeks. Pharia and Phataria. Scarid fishes, hermit crabs, A species list from the collections and observa- and gastropods were remarkably scarce. tions made during the survey is given in the Ap- pendix. The most prevalent of all animals, Balanus peninsularis (Table 5), is known from the Gala- Coral Communities pagos but not from the mainland. Since its test On the steeply sloping rock walk typical of the provides a major source of shelter or habitat for Malpelo subtidal, hermatypic corals occur only motile animals and a substratum for sessile orga- as a veneer interspersed with other growth (Figure nisms, the prevalence of B. peninsularis may be 22) or in some cases as shingled, overhanging

FIGURE 24.—Porites heads in the coral community; .Station 6, 18 m depth. 64 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

masses (Figure 23). Table 5 shows that the grand average surface area coverage by hermatypic corals on steep rock walls is only about 10 percent. In contrast, a more gradual slope at Station 6 supports an extensive coral community (Figure 24) extend- ing to 30 m depth. Table 8 shows that hermatypic corals dominate the surface area coverage of this slope. At 15 meters depth, a continuous growth of Porites sp. measured 6.7 meters across by 2 meters high. At 28 meters, a continuous growth of Agari- ciella planulata covered an area of 3 by 1.8 meters. Figure 25 illustrates that coral coverage at about 30 meters is extensive and not merely confined to isolated colonies. The rich development of hermatypic corals at Station 6 suggests that conditions for coral growth are highly favorable at Malpelo. The lack of coral development elsewhere around Malpelo may be principally due to two factors, both related to the prevalence of vertical rock walls around the island. Vertical surfaces are frequently shaded due to sun angle and substratal irregularities; thus, much of FIGURE 25.—Agariciella planulata; Station 6, SO m depth. the light falling on them is reduced in intensity

depth in feet water level

100

FIGURE 26.—Profile diagram of the coral community at Station 6. As the diver faces shore, transect A is to the left and transect B is to the right with a sand and cobble channel separating them (r = Pocillopora robusta, s = Porites spp., c = Pavona clivosa, and v = Agariciella planulata). NUMBER 176 65

FH.URE 27.—Nicks from fish grazing on Porites sp. and quality by scattering or by passing obliquely corals can build a reef. Figure 23 shows that corals through the water mass. By contrast, intensity on growing on the vertical faces tend to extend out horizontal surfaces at a similar depth would be into the water mass to form a more horizontal sur- greater. Calcification in corals is enhanced by the face which would better intercept light. This pro- photosynthesis of their symbiotic zooxanthellae cess can proceed only so far until erosion by bur- (Goreau, 1959) and it is likely that the small rowing animals causes collapse. Furthermore, sub- amount of coral noted along vertical walls at Mal- tidal talus piles and the eroding aspect of the cliffs pelo is partly due to lower light intensities. above suggest that rockfalls are a frequent occur- The second factor possibly contributing to the rence at Malpelo. Some talus appeared to be of lack of extensive coral development at Malpelo is recent origin, judging from the low numbers of simply the lack of horizontal support on which large encrusting organisms. These rockfalls prob- 66 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY ably constitute a major disruptive phenomenon 1973. Distribution of Coral Reefs in the Pearl Islands (Gulf of Panama) in Relation to Thermal Condi- which prevents dense coral buildup. tions. Limnology and Oceanography, 18:367-379, 8 Even at Station 6 the prolific coral growth was figures. not developed above 10 meters depth, and it was Glynn, P. W., R. H. Stewart, and J. E. McCosker not clear that the living corals were growing on a 1972. Pacific Coral Reefs of Panama: Structure, Distribu- pre-existing coral framework. We thus hesitate to tion and Predators. Sonderdruck aus der Geolo- refer to Station 6 as a coral reef. It is possible that gischen Rundschau, 61:483-519, 15 figures. the development of a true reef is prevented at Goreau, T. F. 1959. The Physiology of Skeleton Formation in Corals, Malpelo by other factors, one of which may be I: A Method for Measuring the Rate of Calcium temperature. Surface temperatures were usually Deposition by Corals under Different Conditions. about 26.5°C, but an abrupt thermocline was Biological Bulletin, 116:59-75, 5 figures. formed at various depths. Temperatures below the Scheltema, R. S. thermocline were as low as 19.5°C and it has been 1973. On an Unusual Means by which the Sessile Marine suggested that this is cool enough to limit coral Ciliate Folliculina simplex Maintains Its Wide- spread Geographical Distribution. Netherlands growth (Glynn and Stewart, 1973). Journal of Sea Research, 7:122-125, 2 plates. To the casual observer, the pattern of coral zona- Thorson, G. tion is more clearly expressed at Malpelo than in 1961. Length of Pelagic Larval Life in Marine Bottom continental eastern Pacific coral communities be- Invertebrates as Related to Larval Transport by cause the greater depth range at Malpelo allows Ocean Currents. In M. Sears, editor, Oceanography. American Association for Advancement of Science the lower zones to spread out over a larger area Publication, 67:455-474, 3 figures. (Figure 26). For instance, Pocillopora spp., Pavo- na gigantea, and Pavona varians are all intermixed between 6 and 7 meters at Uva Island, near the Appendix coast of Panama (cf. Figure 8 in Glynn, Stewart, and McCosker, 1972) while Pocillopora robusta, A List of Marine Invertebrates Identified from Pavona clivosa, and Agariciella planulata are sepa- Malpelo Island rated by depth at Malpelo (Figure 26). PROTOZOA As on coral reefs in continental regions of the tropical eastern Pacific (Glynn, Stewart, and Mc- Folliculina sp. Cosker, 1972), signs of the effects of grazing by PORIFERA fishes are prevalent. The edges of the ridged exten- sions of the large Pavona clivosa colonies were Polyfibrospongia sp. lined with nicks scraped by browsing fishes. Fig- COELENTERATA ure 11 in Glynn, Stewart, and McCosker (1972) Hydrozoa is a typical illustration of this phenomenon. Oc- casionally, concentrations of nicks also occurred Errinopora pourtalesii (Dall) on sheets of Porites (Figure 27). Only three indi- Anthozoa viduals of the scarid fish, Scarus rubroviolaceus, Pacifigorgia media (Verrill) were seen at Station 6 and none were observed at Pacifigorgia sp. other stations. The tetraodontid, Arothron melea- Lophogorgia alba (Duchassaing and Michelotti) gris, and balistids were common at all stations and Pocillopora robusta Verrill probably these fishes, rather than scarids, were re- Pavona clivosa, Verrill sponsible for most of the damage to corals (Figure Pavona varians Verrill Agariciella planulata (Dana) 27). Porites spp. Astrangia sp. Literature Cited Tubastraea aurea (Quoy and Gaimard) MOLLUSCA Birkeland, C. 1971. Biological Observations on Cobb Seamount. North- Gastropoda west Science, 45:193-199, 2 figures. Diodora saturualis (Carpenter) Glynn, P. W., and R. H. Stewart Seila assimilata (C. B. Adams) NUMBER 176 67

Epitonium billeeanum (DuShane and Bratcher) ALPHEIDAE Hipponix grayanus Menke Alpheus lottini Gue"rin, 1830, Indo-Pacific-western America Hipponix panamensis C. B. Adams A. malleator Dana, 1852, western America, eastern & western Crepidula aculeata (Gmelin) Atlantic Capulus sp. A. grahami Abele, new species, endemic to Malpelo (see Erato (?columbella Menke) paper by Abele in this volume) Cypraea albuginosa Gray Synalpheus charon (Heller, 1861), Indo-Pacific-western Cypraea isabellamexicana Stearns America Cypraea teres pellucens Melvill S. nobilii Coutiere, 1909, western America Simnia aequalis (Sowerby) S. digueli Coutiere, 1909, western America Cymatium pileare (Linnaeus) S. biunguiculatus (Stimpson, 1860), Hawaii, Clipperton, Colubraria ochsneri Hertlein & Allison Malpelo Muricanthns princeps (Broderip) 5. bannerorum Abele, new species, endemic to Malpelo (see Coralliophila sp. paper by Abele in this volume) Coralliophila Qneriloides [Lamarck]) Pomagnathus corallinus Chace, 1937, western America Quoyula madreporarum (Sowerby) Thais planospira (Lamarck) HlPPOLYTIDAE Acanthina brevidentata (Wood) Lysmata galapagensis Schmitt, 1924, Galapagos Island, Mal- Purpura pansa Gould pelo Morula lugubris (C. B. Adams) L. trisetacea (Heller, 1861), Indo-Pacific-offshore islands Anachis (?coronata [Sowerby]) Palinuridae Mitrella sp. 1 Panulirus penicillatus (Oliver, 1791), Indo-Pacific-offshore is- Mitrella sp. 2 lands Fasciolaria princeps Sowerby PORCELLANIDAE Conus diadema Sowerby Conus dalli Stearns Petrolisthes haigae Chace, 1962, western America Tylodina fungina Gabb P. glasselli Haig, 1957, western America Chromodoris sedna (Marcus) Pachycheles biocellatus (Lockington, 1878), western America Clastotoechus diffractus (Haig, 1957), western America Bivalvia PAGURIDAE Area mutabilis (Sowerby) Pagurus sp., western America Barbatia reeveana (Orbigny) Septifer zeteki Hertlein and Strong DlOCENIDAE Lithophaga aristata (Dillwyn) Lithophaga plumula (Hanley) Aniculus elegans Stimpson, 1858, western America Isognomon recognitus (Mabille) Trizopagurus magnificus (Bouvier, 1898), western America Ostrea fisheri Dall Ostrea Qiridescens Hanley) DYNOMENIDAE

ARTHROPODA Dynomene ursula Stimpson, 1859, western America

Cirripedia PORTUNIDAE

Balanus peninsularis Pilsbry Euphylax dovii Stimpson, 1860, western America Tetraclita stalactifer (Lamarck) Heteralepas quadrata (Aurivillius) XANTHIDAE Lepas anserifer Linnaeus Qitadrella nitida Smith, 1869, western America Decapoda Trapezia digitalis Latreille, 1825, Indo-Pacific-western America PALAEMONIDAE T. ferruginea Latreille, 1825, Indo-Pacific-western America Brachycarpus biunguiculatus (Lucas, 1849), circumtropical Menippe obtusa Stimpson, 1859, western America Harpiliopsis depressus (Stimpson, 1860), Indo-Pacific-western Domecia hispida Eydoux and Souleyet, 1842, Indo-Pacific- America western America Pseudocoutierea elegans Holthuis, 1951, western America Medaeus spinulifer Rathbun, 1898, western America Veleronia laevifrons Holthuis, 1951, western America Pilumnus pyginaeus Boone, 1927, western America Peridimenaeus hancochi Holthuis, 1951, western America Globopilutnnus xantusii (Stimpson, 1859), western America Ozius perlatus Stimpson, 1859, western America GNATHOPHYLLIDAE Carpilodes cinctimanus (White, 1847), Indo-Pacific-western Gnathophylloides mineri Schmitt, 1933, circumtropical (?) America 68 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

GRAPSIDAE Asterope carinifera Miiller and Troschel

Grapsus grapsus (Linnaeus, 1758), western America, eastern OPHIUROIDEA and western Atlantic Astrodictyum panamense (Verrill) GECARCINIDAE Ophiocoma aethiops Liitken Ophiocoma alexandri Lyman Gecarcinus mapilensis Faxon, 1893, endemic to Malpelo Ophiactis savignyi (Miiller and Troschel) MAJIDAE Ophiotheta mirablis Verrill

Teleophrys cristulipes Stimpson, 1959, western America ECHINOIDEA Mithrax pygmaeus Bell, 1835, western America Microphrys platysoma (Stimpson, 1859), western America Diadetna mexicanum A. Agassiz Lissa tuberosa Rathbun, 1898, western America Eucidaris thouarsii (L. Agassiz and Desor) ECHINODERMATA Tripneustes depressus A. Agassiz Echinometra sp. ASTEROIDEA Leiaster callipeplus Fisher CHORDATA Narcissia gracilis malpeloensis, new subspecies ASCIDIACEA Tamaria stria Downey, new species, endemic to Malpelo (see paper by Downey in this volume) Polyandrocarpa (Eusynstyela) sp. aff. tincta (Van Name) bradleyi Verrill Didemnum (Didemnum) sp. aff. candidum Savigny Nidorellia armata (Gray) Didemnum (Didemnum) sp. aff. moseleyi (Herdman) The Macruran Decapod Crustacea of Malpelo Island

Lawrence G. Abele

G. malpilensis as a species endemic to Malpelo. In ABSTRACT 1948, Garth reported on the Askoy collections, which included three species taken at Malpelo: an Eighteen species of Macrura are reported from intertidal crab, Grapsus grapsus, the land crab, Malpelo Island. Two new species of Alpheidae, Gecarcinus malpilensis, and a pelagic portunid, Alpnens grahami and Synalpheus bannerorum, are described and S. nobilii and S. digueti are rede- Euphylax dovii, which was abundant in the waters scribed. Notes are given on Periclimenaeus han- around the island. No additional material of deca- cocki, and Gnathophylloides mineri is recorded pod crustaceans appears to have been collected at from the eastern Pacific. Malpelo until the Smithsonian Tropical Research Ecologically about half of the 43 decapods known Institute and U. S. Navy expedition visited the from Malpelo are commensals with other inverte- island in 1972. brates, mostly Pocillopora coral. Two species are semiterrestrial or terrestrial and the remaining Several members of the expedition independ- species occur in a variety of subtidal habitats. ently made collections of decapods and sent them Zoogeographically decapod fauna of Malpelo (43 to various individuals for study. I was unaware that spp.) is most closely related to that of the western both Janet Haig and John S. Garth were studying American mainland with 62 percent or 27 species material from Malpelo when I began studying the occurring there. Nine species (20%) occur in the present collections. My study was then restricted Indo-Pacific region and seven of these reach the western American mainland. There are three en- to the macrurans, the subject of the present report. demic species, two circumtropical species, and two A complete list of the decapods reported from Mal- species known only from offshore islands. pelo is included in the Birkeland et al. appendix The very limited data available at present sug- (pp. 66-68, herein), and, therefore, it is unneces- gest a species-area effect on the shallow water sary to repeat here. decapods associated with oceanic islands. The macrurans collected at Malpelo Island form a small but interesting collection of 18 species. Two species were previously undescribed, three species Introduction are recorded for the first time since the original description, the genus Gnathophylloides is re- The first report of a decapod crustacean from corded from the eastern Pacific for the first time Malpelo Island was that of Faxon (1893) who de- and many of the remaining species were previously scribed a land crab, Gecarcinus malpilensis, based known from only a few records. on material collected by the Albatross. This species was considered to be a synonym of G. planatus by Collections of macrurans came primarily from Rathbun (1918), but Turkay (1970) recognized two localities on Malpelo Island referred to in the descriptions as: "Malpelo coll. no. 3," which was made on the southeastern section of the island Lawrence G. Abele, Smithsonian Tropical Research Institute, among coral in 10 m depth on 2 March 1972, and P. O. Box 2072, Balboa, Canal Zone. Present address: Depart- ment of Biological Science, Florida State University, Talla- "Malpelo coll. no. 4," which was made on the west hassee, Florida 32306. side of the island in 15 m depth on 3 March 1972. 69 70 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

"Malpelo coll." refers to general collecting around in shallow waters in a wide variety of habitats: on the island from 29 February to 3 March 1972. The coral reefs, in tide pools, on grass flats, and among abbreviation cl refers to carapace length excluding subtidal rocks (Chace, 1972). the rostrum, tl to total length excluding the ros- trum unless otherwise specified and USNM to the Periclimenaeus hancocki Holthuis, 1951 National Museum of Natural History, Smithsonian Institution, Washington, D.C. All of the material FIGURE 28B-D will eventually be deposited in that institution. ACKNOWLEDGMENTS.—I thank Jeffrey B. Graham Periclimenaeus hancocki Holthuis, 1951:97, pi. 29a-k. for entrusting me with the study of this interesting MATERIAL.—1 male, 1 ovigerous female, Malpelo collection. Charles Birkeland supplied me with a coll. collection of shrimps and the color slide of Panu- MEASUREMENTS.—Male cl 6.0 mm, ovigerous fe- lirus penicillatus. The following three individuals male cl 8.4 mm. contributed so much to this study that completion DISTRIBUTION.—The unique holotype was col- would not have been possible without their aid: lected from Pinas Bay on the Pacific coast of Pana- Dora M. Banner and Albert H. Banner of the Uni- ma. It is now reported from Malpelo Island. versity of Hawaii, examined alpheid material and REMARKS.—The male specimen agrees with the loaned material of Alpheus crockeri. Fenner A. description of the unique male holotype given by Chace, Jr., of the Smithsonian Institution, has Holthuis (1951). The lamella of the scaphocerite, once again given his time and energy to the com- however, does not extend as far beyond the lateral pletion of a study. He loaned material of Pericli- tooth as figured by Holthuis. The female specimen menaeus pearsei and compared material of Pseu- differs in several respects. The rostrum is relatively docontierea, Veleronia, and Gnathophylloides with shorter than that of the male and it is armed with type and other material in the Smithsonian Insti- four rather than five dorsal teeth. Holthuis had tution collections. only the male holotype of P. hancocki available when he noted certain differences between this PALAEMONIDAE

Brachycarpus biunguiculatus (Lucas, 1849)

Palaemon biunguiculatus Lucas, 1849:45, pi. 4: fig. 4. Brachycarpus biunguiculatus (Lucas).—Holthuis, 1952:3, pi. 1—Chace, 1962:606; 1966:625; 1972:18.

MATERIAL.—3 males, 6 non-ovigerous females, 5 ovigerous females, 1 juvenile, Malpelo coll. no. 3; 2 males, 2 non-ovigerous females, Malpelo coll. no. 4. MEASUREMENTS.—Males tl (including rostrum) 36.1-40.7 mm; non-ovigerous females tl 22.7-28.7 mm; ovigerous females tl 31.5-41.3 mm. DISTRIBUTION.—This species is circumtropical in distribution occurring on most oceanic islands. REMARKS.—All of the Malpelo material agrees with the description and illustrations of B. biungu- iculatus, which has an extensive synonomy dis- cussed by Holthuis (1952). A second recognized FICURE 28.—Periclimenaeus pearsei Schmitt: A, chela of first species of Brachycarpus, B. holthuisi, has been pereiopod. Periclimenaeus hancocki Holthuis: B, chela of first pereiopod; c, telson and left uropod of female; D, anterior described from deeper waters (30-60 m) off Brazil portion of carapace of ovigerous female. [Scale = 1 mm for by Fausto Filho (1966). The present species occurs A, B; 2 mm for c, D.] NUMBER 176 71 species and the closely related species, P. pearsei Guerrero south of White Friars. Ecuador, Galapa- Schmitt, 1932 from the western Atlantic and these gos Islands, Bindloe Island. It is now reported from are listed in his key (1951:78). The present female, Malpelo Island. and to a lesser extent the male, bridge these differ- REMARKS.—The present material differs in sev- ences with one important exception: the form of eral respects from the description of P. elegans the dactylus of the chela of the first pereiopods. In given by Holthuis (1951). All of the Malpelo speci- P. pearsei (Figure 28A), of which two males and mens are smaller (tl 7.1-8.6 mm) than those listed two females (USNM 65113, 65114) were available by Holthuis (tl 9-16 mm). The lamella of the sca- for comparison, the dactylus is high and convex phocerite is less produced in the Malpelo speci- while in P. hancocki (Figure 28B) it is low and not mens than those figured by Holthuis. In the orig- nearly as wide. This distinct and so far constant inal description he notes that larger specimens difference seems to warrant continued specific re- tend to have the dactylus of the larger second pere- cognition of these two closely related species. iopod robust with a large proximal tooth. Even the The eggs of both species are small (about 0.45 smallest of the Malpelo specimens have this char- X 0.3 mm) and numerous. acteristic. The rostrum of the Malpelo material seems to be shorter than Holthuis' material. The Malpelo males lack a spine on the posterior mar- Harpiliopsis depressus (Stimpson, 1860) gin of the third pleuron, which is present in Hol- Harpilius depressus Stimpson, 1860:38. thuis' figure (see Chace, 1972:45). Dr. Fenner A. Periclimenes pusillus Rathbun, 1906:921, fig. 71, pi. 24: fig. Chace, Jr., compared Malpelo material with para- 7 Bruce, 1970:306, fig. 1. types of P. elegans and noted that the Malpelo Harpiliopsis depressus (Stimpson) Holthuis, 1951:70, pi. specimens seem to have the cornea set less obli- 21, pi. 22a-f. quely on the eyestalk and the general body form MATERIAL.—1 male, Malpelo coll. no. 3. more depressed. MEASUREMENTS.—Male cl 2.5 mm. Given, however, the relatively few specimens DISTRIBUTION.—Red Sea and Seychelles Islands, available from such widely scattered localities and islands of the central Pacific and the west coast of the variation known in other pontoniids, it seems America from the Gulf of California to Colombia. best to refer the present material to P. elegans REMARKS.—Bruce (1970) examined the status of until more material is available. Periclimenes pusillus Rathbun from Hawaii and concluded that it represents a juvenile of Harpili- opsis depressus. Two color forms of this species Veleronia laevifrons Holthuis, 1951 occur on Pocillopora coral heads in Panama and Australia. Their status is being investigated by A. Veleronia laevifrons Holthuis, 1951:199, pi. 63f-m. J. Bruce. MATERIAL.—3 males, 3 non-ovigerous females, 4 The translucent form of this species is one of the ovigerous females, Malpelo coll. most abundant Pocillopora commensals in Pana- MEASUREMENTS.—Males tl 4.2-4.9 mm, non- ma (Abele, 1972). ovigerous females tl 6.4 mm, ovigerous females tl 5.4-6.6 mm. Pseudocoutierea elegans Holthuis, 1951 DISTRIBUTION.—Ecuador, off Cape San Francisco and Santa Elena Bay, Galapagos Islands (Hood Pseudocoutierea elegans Holthuis, 1951:182, pi. 55a-r [pisIsland). . The species is now reported from Malpelo 55 and 57 reversed, legends are not].—Chace, 1972:45. Island. MATERIAL.—3 males, 6 non-ovigerous females, 7 REMARKS.—Dr. Fenner A. Chace, Jr., compared ovigerous females, Malpelo coll. the present material with paratypes of V. laevi- MEASUREMENTS.—Males tl 7.1-7.3 mm, non- frons from Hood Island and concluded they were ovigerous females tl 8.1-8.4 mm, ovigerous females conspecific. // 7.4-8.6 mm. Schmitt (in Holthuis, 1951:200) gives color notes DISTRIBUTION.—California, Santa Catalina Island. on specimens of this species that were found on a Mexico, Baja California off Ildefonoso Island, gorgonian. 72 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

GNATHOPHYLLIDAE Alpheus malleator Dana, 1852 Gnathophylloides mineri Schmitt, 1933 Alpheus malleator Dana, 1852:557; 1855, pi. 31: fig. 9a-h— Crosnier and Forest, 1966:240, fig. 10. Gnathophylloides mineri Schmitt, 1933:7, fig. 3; 1935:167, fig. 31.—Lewis, 1956:288, figs. 1, 2.—Chace, 1972:52 Bruce, MATERIAL.—11 males, 2 non-ovigerous females, 1973:27. 2 ovigerous females, 7 juveniles. MEASUREMENTS.—Males cl 7.8-11.9 mm, non- MATERIAL.—1 ovigerous female. ovigerous females cl 5.2-5.4 mm, ovigerous females MEASUREMENTS.—Ovigerous female cl 1.9 mm, cl 7.8-11.4 mm. tl 5.5 mm. DISTRIBUTION.—Western Atlantic: southern Flor- DISTRIBUTION.—Southeastern Florida, Yucatan, ida to Sao Paulo, Brazil. Eastern Atlantic: Senegal and Caribbean Sea (Chace, 1972). The species is to the Congo. Eastern Pacific: Gulf of California now reported from the eastern Pacific at Malpelo to Ecuador, the Galapagos Islands, and now Mal- Island. pelo Island. REMARKS.—Dr. Fenner A. Chace, Jr., compared REMARKS.—None. the Malpelo specimen with Atlantic material of G. mineri and was unable to find any obvious dif- ferences such as those noted by Bruce (1973) in Alpheus grahami new species his comparison of G. mineri with G. robustus. In the Atlantic the species occurs under stones FIGURE 29A-C, F.-J or in coral, but it is usually found on the spines of MATERIAL.—2 ovigerous females (smaller speci- sea urchins (Tripneustes and Lytechinus). A men is holotype), 2 juveniles, Malpelo coll. no. 3. species of Tripneustes, T. depressus, is known to MEASUREMENTS.—Ovigerous female holotype cl occur at Malpelo Island (Birkeland et al., Appen- 11.1 mm, paratypes, ovigerous female cl 12.3 mm, dix, this volume). juveniles cl 3.7, 4.2 mm. TYPE-LOCALITY.—Southeastern side of Malpelo ALPHEIDAE Island, Colombia in 10 m depth among coral. DISTRIBUTION.—Known only from the type- Alpheus lottini Guerin-Meneville, 1830 locality. Alpheus lottini Guerin-Meneville, 1830, pi. 3 Holthuis, DESCRIPTION.—The rostrum is short and triangu- 1958:22. lar with an acute apex; it extends to about the mid- Alpheus ventrosus H. Milne Edwards, 1937:352.—Banner, dle of the basal antennular segment. The orbital 1958:164, fig. 4. teeth are triangular, subacute, and extend almost Alpheus laevis Randall, 1839:141. to the apex of the rostrum. Crangon ventrosa (H. Milne Edwards).—Banner, 1953:84, fig. 28. The first three abdominal pleura are rounded; Crangon latipes Banner, 1953:82, fig. 27. the fourth and fifth are bluntly angled and the sixth is subacute. The sixth segment is slightly MATERIAL.—2 ovigerous females, 1 juvenile, longer than the fifth and is about 0.7 of the length Malpelo coll. no. 3. of the telson. The telson has the length about 1.8 MEASUREMENTS.—Ovigerous females cl 5.2, 8.2 times the width. The distal lateral angles are mm, juvenile cl 2.9 mm. armed with a pair of spines, the inner one being DISTRIBUTION.—Red Sea and South Africa, across longer and stronger. Two pairs of dorsal spines are the Central Pacific to western America where it present; the anterior pair is placed slightly less occurs from Baja California to at least Panama. than half the distance from the anterior margin It is now reported from Malpelo. and the posterior pair is located about three- REMARKS.—This species is a characteristic mem- fourths the distance from the anterior margin. The ber of the Pocillopora coral community where it uropods extend well beyond the distal margin of usually occurs in male-female pairs. Patton (1973) the telson. The lateral uropod is armed with a has dealt with aspects of the biology of this color- strong movable spine at the distal lateral angle ful species. that is adjacent to a smaller immovable spine. NUMBER 176 73

The stylocerite extends distinctly beyond the does not extend beyond the antennular peduncle. distal margin of the basal antennular segment. The The third maxilliped extends beyond the carpo- visible portion of the basal segment is about three- cerite by the distal third of the ultimate segment. fourths the length of the second segment. The The distal portion of the ultimate segment is third segment is about three-fourths the length of armed with strong setae but no spines. the second segment. The first pereiopods are strong and unequal. The basicerite is unarmed dorsally, the angle The ischium of the major cheliped is armed with being blunt. The lateral spine is short and acute, about four spinules on the inferior margin and a extending to about half the length of the styloce- single spine on a lobe on the distal superior mar- rite. The scaphocerite is widest proximally and gin. There are about 10 movable spinules on the

FIGURE 29.—Alpheus grahami, new species: A, anterior portion of carapace; B, third pereiopod; c, small chela of first pereiopod; E, telson; F, third through fifth abdominal pleura; c, anterior portion of major chela of first pereiopod; H, dactylus of third pereiopod; i, second pereiopod; j, major chela of first pereiopod. Alpheus crockeri (Armstrong): D, anterior portion of major chela of first pereiopod. [Scale -• 2 mm for A, B, D, E, I; 4 mm for c, F-H, J.] 74 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY inferior margin of the merus; the distal margin is The third pereiopod has the ischium less than unarmed. The major chela is about three times as half the length of the merus. The merus is slightly long as high. There are no superior notches but a less than five times as long as wide; it is about 1.5 sulcus is present on the medial surface below the times the length of the carpus, about 1.3 times the superior margin. It extends from the base of the length of the propodus and about 5 times the movable finger to about the middle of the chela length of the dactylus. The propodus is armed with where it bifurcates; the upper portion extends about nine movable spines on the inferior margin. only a short distance but the lower portion extends The dactylus has a minute accessory tooth, sur- downward almost to the proximal margin of the rounded by setae, on the superior margin. The chela. The lower margin of the chela is compressed eggs are small and numerous. proximal to the base of the immovable finger re- REMARKS.—This species seems to be most closely sulting in the formation of a weak inferior notch related to Alphens crockeri (Armstrong, 1941) adjacent to two weak tubercles. The distal margin (= Crangon tuthilli Banner, 1953) known from of the chela is armed with a pair of large, acute the Central Pacific Ocean, Thailand (Banner, teeth, one on either side of the movable finger. Be- 1953; Banner and Banner, 1966) and the eastern low the lateral (outside) tooth, the margin of the Atlantic (Crosnier and Forest, 1966). Three speci- chela leading to the immovable finger is high and mens, through the courtesy of D. M. and A. H. crenulate. The movable finger is high, slightly com- Banner, of A. crockeri were available for compari- pressed, and blunt, although in the single chela son: 1 male cl 8.0 mm, 2 non-ovigerous females cl available the apex may be damaged. The ischium 5.0, 6.9 mm from Oahu, Hawaii. The two species of the minor chela is armed with about six spinules are very similar, and were it not for the extreme on the inferior margin and a spinule on a lobe on differences in the form of the major chela, other the distal superior margin. The merus is armed differences would have been attributed to varia- with about 11 spinules on the inferior margin; the tion. In A. crockeri (Figure 29u), the movable fin- distal margin is unarmed. The chela is about 1.7 ger is greatly compressed, with the superior margin times as long as high. The movable finger is al- being defined by a knife-like ridge; it is twisted most 0.6 the length of the chela and is somewhat in a horizontal plane and the apex is bulbous. The flattened dorsally. The distal margin of the palm immovable finger has a large, high proximal tooth is armed with two strong teeth, one on either side present and is twisted in a horizontal plane. In A. of the movable finger. The tips of both fingers are grahami (Figure 29c, F), the movable finger is acute; the hand is not balaeniceps in form. The robust and only slightly compressed; it is not second pereiopods are subequal and extend beyond twisted and the width is the same throughout. The the carpocerite by the distal three segments of the immovable finger lacks a proximal tooth and is carpus. The ischium is slightly longer than the not twisted. merus and both are shorter than the carpus. The Other differences between the two species are carpus is subdivided into five segments; the first slight but are constant in the material examined. is the longest being slightly longer than the sec- In A. crockeri the stylocerite is narrow and extends ond and the fifth, which are subequal in length. almost to the distal margin of the basal antennu- The third and fourth segments are subequal in lar segment, while in A. grahami the stylocerite is length and are slightly less than half the length of more robust and extends distinctly beyond the the first. The movable finger is slightly more than basal segment. The antennular peduncle of A. half the length of the chela. The fingers are un- crockeri is slender with the second segment having armed. The third through the fifth pereiopods are the length about twice the width, while in A. gra- similar in form. Three of the four available speci- hami the length is distinctly less than twice the mens have a small movable spinule at the base of width. The scaphocerite of A. crockeri is slender the ischium of pereiopods three and four. This and extends to or beyond the distal margin of the spinule is relatively large on the two small speci- antennular peduncle, while in A. grahami it is mens (cl 3.7, 4.2 mm), but it is small and difficult shorter than the antennular peduncle. The pereio- to see on one of the larger females (cl 12.3 mm) pods of A. crockeri are slender with the length of and it is absent on the other female (cl 11.1 mm). the merus being about six times the width while in NUMBER 176 75

A. grahami the length is about five times the width. Specimens were also examined from Punta Paitilla, From illustrations in the literature (Banner and Panama. Banner, 1966; Crosnier and Forest, 1966) it DESCRIPTION.—The rostrum is narrow and acute, would appear that the rostrum of A. crockeri is not extending to the distal margin of the basal an- longer and more acute than that of A. grahami, tennular segment. It is on a higher level and ex- but the specimens of A. crockeri from Oahu, tends beyond the orbital teeth, which are small Hawaii have a rostrum similar to that of A. gra- acute extensions on the orbital hoods. hami. The third abdominal pleuron is obtusely ETYMOLOGY.—The species is named for Jeffrey rounded, the fourth and fifth are bluntly angled B. Graham, who collected the specimens and en- and the sixth is acute. The fifth segment is slightly trusted them to me for study. shorter than the sixth, which is slightly more than half the length of the telson. The telson is slightly longer than its anterior width. It narrows distally Synalpheus biunguiculatus (Stimpson, 1860) and has a distinct constriction proximally. The Alpheus biunguiculatus Stimpson, 1860:31. telson is armed with two pairs of spines at the dis- Synalpheus biunguiculatus (Stimpson) Banner, 1953:32, tal lateral angles; the lateral spine is about one- fig. 9. third the length of the medial. There are two pairs of dorsal spines present; the first pair is located MATERIAL.—1 ovigerous female, 2 specimens, slightly anterior of the middle and the second is Malpelo coll. no. 3; 1 specimen, Malpelo coll. no. located about four-fifths of the distance from the 4; 1 specimen, Malpelo coll. anterior margin. The uropods extend well beyond MEASUREMENTS.—Ovigerous female cl 4.2 mm; the distal margin of the telson. The lateral branch other specimens cl 2.7-5.7 mm. is armed with a strong movable spine situated be- DISTRIBUTION.—Hawaiian Islands (Banner, 1953), tween two immovable spines at the level of the Clipperton Island (Chace, 1962), and now Malpelo diaeresis. Island. The stylocerite is long and slender extending to REMARKS.—The Malpelo material all lack the about the middle of the second antennular seg- major chela making the identification somewhat ment. The visible portion of the basal segment is uncertain. They agree in most respects with the slightly longer than the second segment, which is description and illustrations of Banner (1953). distinctly longer than the third segment. The carpocerite in the Malpelo material is some- The basicerite is armed with strong dorsal and what longer than the spine of the scaphocerite, but lateral spines; the dorsal spine is about half the the species is variable in this and other characters length of the lateral. The lateral tooth of the scaph- (Chace, 1962). ocerite extends beyond both the antennular pe- duncle and the carpocerite. The third maxilliped has the apex armed with Synalpheus nobilii Coutiere, 1909 nine strong, dark-colored spines arranged in a cir- FIGURE 30 cle. The ultimate segment extends slightly beyond the scaphocerite. Synalpheus nobilii Coutiere, 1909:40, fig. 22.—Schmitt, 1924: The first pereiopods are strong and unequal. 162, fig. 39.—Sivertsen, 1934:22 Schmitt, 1939:12, 24.— Chace, 1962:613. The superior distal margin of the merus is armed with a tooth. The major chela has the length MATERIAL.—23 specimens (6 ovigerous), Malpelo slightly less than three times the height. The distal coll. no. 3; 4 specimens, Malpelo coll. no. 4; 10 margin of the palm is unarmed. A tubercle is pres- specimens, Malpelo coll. ent on the medial margin of the palm just below MEASUREMENTS.—Ovigerous females cl 4.1-4.9 the distal margin. The immovable finger is armed mm, other specimens cl 1.8-6.2 mm. with two teeth proximally. The movable finger is DISTRIBUTION.—St. Helena, Ecuador (Coutiere, slightly less than one-third the length of the chela. 1909), Galapagos Islands (Schmitt, 1924), Clipper- The superior distal margin of the merus of the ton Island (Chace, 1962) and now Malpelo Island. smaller chela is armed with a tooth. The smaller 76 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY chela has the length slightly more than three times times as long as wide and is unarmed. It is about the height. The distal margin of the palm is un- twice the length of the carpus and about 1.3 times armed. The fingers are unarmed; the movable one the length of the propodus. The propodus is armed is about one-third the length of the chela. The sec- with six movable spines, the sixth being paired, ond pereiopods are equal and extend beyond the on the inferior surface. The carpus has a distal scaphocerite by the distal segment of the carpus. extension on the surperior margin and a pair of The ischium is about three-fourths the length of spines on the distal inferior margin. The dactylus the merus, which is slightly shorter than the car- is bifid with a small ventral protuberance; the ven- pus. The carpus is subdivided into five segments; tral tooth is wider than the dorsal and slightly the first is slightly less than five times the length shorter. The eggs are small and numerous. of the second, third, and fourth, which are sub- REMARKS.—Specimens of this species have been equal, and about twice the length of the fifth. The collected from the rocky intertidal zone in the Bay movable finger is about half the length of the che- of Panama. The material from Malpelo Island la. The third through the fifth pereiopods are came from among coral and coral rubble in about similar. The merus of the third is about three 10 m depth.

FIGURE SO.—Synalpheus nobilii Coutiere: A, anterior portion of carapace; B, second pereiopod; c, third pereiopod; D, small chela of first pereiopod; t, anterior portion of major chela of first pereiopod; F, distal portion of dactylus of third pereiopod, c, major chela of first pereiopod; H, telson and left uropod. [Scale = 4 mm for E, C; 2 mm for A-D, H; 1 mm for F.] NUMBER 176 77

Synalpheus digueti Coutiere, 1909 and the Pearl Islands, Panama, was also examined during the present study., FIGURE 31 DESCRIPTION.—The rostrum is narrow, subacute, Synalpheus digueti Coutiere, 1909:48, fig. 28 Chace, 1937: and short; extending to about one-fourth of the 123. length of the basal antennular segment. The orbit- al teeth arise from the anterior margin of the cara- MATERIAL.—4 ovigerous females, 36 specimens, pace; they are triangular, subacute, almost twice Malpelo coll. no. 3; 3 specimens, Malpelo coll. no. as broad as the rostrum, and are subequal in 4; 16 ovigerous females, 40 specimens, Malpelo length to it. coll. The third through the fifth abdominal pleura MEASUREMENTS.—Ovigerous females cl 5.2-8.4 are bluntly angled; the sixth is acute. The fifth mm, other specimens cl 3.0-7.6 mm. segment is almost equal in length to the sixth DISTRIBUTION.—Baja California (Coutiere, 1909; which is slightly more than half the length of the Chace, 1937). The species is now reported from telson. The telson is longer than wide with two Malpelo Island. Material from Taboguilla Island pairs of spines on the distal lateral angles; the

FIGURE 31.—Synalpheus digueti Coutiere: A, anterior portion of carapace; B, anterior portion of carapace of aberrant specimen; c, second pereiopod; D, minor chela of first pereiopod; E, third pereiopod; F, telson and left uropod; c, anterior portion of dactylus of third pereiopod; H, major chela of first pereiopod; i, anterior portion of major chela of first pereiopod. [Scale = 2 mm for A-F; 1 mm for c; 4 mm for H, I.] 78 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY lateral spine is slightly more than half the length extension on the distal superior margin and a pair of the medial. The distal margin of the telson is of spines on the distal inferior margin. The carpus convex and extends to about two-thirds the length is about half the length of the merus. The propo- of the medial spine. There are two pairs of dorsal dus is about four-fifths the length of the merus spines; the first pair is located at about the middle and is armed with seven movable spinules on the of the telson and the posterior pair is located inferior margin; the last spinule is paired. The slightly more than two-thirds of the distance from dactylus is bifid; the superior tooth is slightly the anterior margin. The uropods extend beyond longer than the inferior tooth and they are sub- the distal margin of the telson. The lateral branch equal in width at the base. is armed with a strong movable spine situated be- The eggs are small and numerous. tween two immovable spines at the level of the In fresh material the body is translucent, the diaeresis. appendages are blue, and the tips of the fingers of The stylocerite is long and acute, extending the first pereiopods are red. slightly beyond the middle of the second antennu- REMARKS.—The rostrum was somewhat variable lar segment. The first antennular segment is about in form often being wider than figured here. One 1.4 times the length of the second and about twice specimen (Figure 31B) has a highly aberrant ros- the length of the third. trum but was normal in other features. Three of The basicerite is armed with dorsal and lateral the 99 specimens examined lack a tooth on the spines; the dorsal spine extends to about one-fifth distal margin of the major chela. In other features of the length of the lateral. The lateral tooth of the species was not variable. the scaphocerite is long and acute and extends In the Gulf of Panama Synalpheus digueti beyond the antennular peduncle. The carpocerite occurs as male-female pairs (the female is usually is long and broad, extending well beyond the lateral ovigerous) in Pocillopora coral heads. tooth of the scaphocerite. The first pereiopods are robust and unequal. Synalpheus charon (Heller, 1861) The merus of the major cheliped is unarmed. The chela is about 2.6 times as long as high. The distal Alpheus charon Heller, 1861:27. margin of the palm is armed with a downwardly Synalpheus helleri de Man, 1911:246, pi. 8: fig. 37. directed spine. There are two tubercles on the Synalpheus charon (Heller).—Banner, 1953:37, fig. 11.— lateral margin of the palm at the insertion of the Banner and Banner, 1964:88. Synalpheus charon obscurus Banner, 1956:329, fig. 5. movable finger. The distal portion of the immov- Synalpheus charon charon (Heller) Banner, 1956:331. able finger is compressed. The movable finger is slightly more than one-fourth the length of the MATERIAL.—3 ovigerous females, 11 specimens, chela. The merus of the smaller cheliped is un- Malpelo coll. no. 3. armed, as is the chela. The length of the chela MEASUREMENTS.—Ovigerous females cl 4.6-5.3 is about 2.5 times the height. The movable finger is mm, other specimens cl 2.6-4.9 mm. about half the length of the chela. The second DISTRIBUTION.—Red Sea across the Pacific to pereiopods are subequal. The ischium is unarmed Baja California (Banner and Banner, 1966; Chace, and is about three-fourths the length of the merus. 1937). During the present study, material was also The merus is unarmed and is about four-fifths the examined from the Pearl Islands and Taboguilla length of the carpus. The carpus is subdivided into Island in the Gulf of Panama. five segments; the first is over five times the length REMARKS.—This species is a characteristic mem- of the second, third, and fourth which are subequal ber of the Pocillopora coral community, occurring and which are half the length of the fifth. The deep inside the coral at the base of the branches movable finger is slightly more than half the (Banner and Banner, 1964). It is common in this length of the chela. The third through the fifth habitat in the Gulf of Panama where it usually pereiopods are similar. The ischium and merus are unarmed. The ischium is slightly less than one- occurs in male-ovigerous female pairs. The body third the length of the merus. The carpus has an and appendages of fresh material is an even red color. NUMBER 176 79

Synalpheus bannerorum, new species The first five abdominal pleura are rounded; the sixth is subacute. The fifth segment is subequal in FIGURE 32 length to the sixth, which is three-fourths the MATERIAL.—5 ovigerous females (holotype tl 11.1 length of the telson. The anterior width of the tel- mm, cl 4.4 mm), 23 specimens, Malpelo coll. no. 3 son is greater than the length; it narrows distinctly (paratypes); 2 specimens, Malpelo coll. no. 4. in the posterior half. The posterior margin is less MEASUREMENTS.—Ovigerous females cl 4.0-4.7 than one-half the width of the anterior margin. mm, tl 9.9-11.1 mm, other specimens cl 3.3-5.0 mm, The distal lateral angles of the posterior margin // 8.5-11.4 mm. are acute and armed with a pair of spines; the TYPE-LOCALITY.—Southeastern side of Malpelo lateral pair being shorter and more robust than Island, Colombia, in 10 meters depth among coral. the medial. Both extend well beyond the lateral DISTRIBUTION.—Known only from the type- angles of the telson. The dorsal surface of the tel- locality. son is armed with a pair of strong spines; the first DESCRIPTION.—The rostrum is narrow and acute; pair is located about two-thirds of the distance it does not extend to the distal margin of the basal from the posterior margin and the second pair is antennular segment. It is on a higher level than located about one-third of the distance from the the orbital teeth, which are small, acute, directed posterior margin. Both pairs are set in well from slightly medially, and do not extend to the apex the lateral margins. The lateral angle of the outer of the rostrum. uropod is armed with a long movable spine set

FIGURE 32.—Synalpheus bannerorum, new species: \, anterior portion of carapace; B, major chela of first pereiopod; c, anterior portion of major chela; D, second pereiopod; E, anterior portion of dactylus of third pereiopod; F, third pereiopod; c., telson and left uropod; H, minor first pereio- pod. [Scale = 2 mm for A-D, F, H; I mm for E, G.] 80 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY between a pair of smaller acute immovable spines length of the carpus and about three-fifths the at the level of the diaeresis. The medial uropod is length of the merus. The third through the fifth longer than the lateral, which is subequal in length pereiopods are similar and robust. The ischium is to the telson. unarmed. The merus of the third is armed with The stylocerite is acute and extends to about the four to six movable spines. The length is slightly middle of the second antennular segment. The more than three times the width. The carpus is basal antennular segment is longer than the second armed with two distal spines. The propodus is and third, which are subequal in length. armed with about nine movable spinules on the The basicerite is armed with strong, acute dorsal inferior margin. The dactylus has the superior and lateral spines; the dorsal spine extending to tooth longer and slightly more than half the width slightly over one-third of the length of the lateral. of the inferior. The merus of the fourth pereiopod The scaphocerite is reduced in size but the lateral is armed with three or four movable spines. The spine is strong, extending beyond both the anten- merus of the fifth pereiopod is unarmed. The eggs nular peduncle and the carpocerite. The carpoce- are small and numerous. rite extends slightly beyond the antennular pe- REMARKS.—This species differs from all known ducle. eastern Pacific species in the form of the meri of The third maxilliped has the tip of the apex the third and fourth pereiopods and in the form armed with six dark-colored spines. It extends be- of the major chela. It is the only western American yond the carpocerite by the distal third of the species in which the meri of the third and fourth ultimate segment. pereiopods are armed and the constriction of the The first pereiopods are unequal and robust. immovable finger of the major chela appears also The merus of the major cheliped is armed with a to be unique. There is one east American species tooth on the distal superior margin. The carpus of Synalpheus (S. dominicensis Armstrong, 1949) is armed with two acute extensions. The major from the Dominican Republic, which has the meri chela is about 2.3 times as long as high. The distal of pereiopods three and four armed. Synalpheus margin of the palm is armed with a small down- bannerorum can be distinguished from S. domini- wardly directed spine. There is a large tubercle at censis by the form of the anterior portion of the the base of the immovable finger which is followed carapace, by the armature of the telson, by the by a deep depression and then another tubercle. form of the major chela and by the form of the The other side of the finger is also depressed giving dactyi of pereiopods three through five. In S. domi- a pinched appearance to the base of the finger. nicensis the medial margins of the orbits meet the The immovable finger is armed with two strong, rostrum at a sharp angle while they blend into the acute teeth in the proximal portion. The movable carapace before they meet the rostrum in S. ban- finger is more than one-third the length of the nerorum. The dorsal spines of the telson are both chela and has two distinct tubercles on the proxi- located in the posterior half of the segment in S. mal dorsal surface. There is a series of about five dominicensis, while in S. bannerorum only the pos- unequal tubercles on the medial surface of the terior pair is located in the posterior half. The palm between the acute distal spine and the base major chela of S. dominicensis is not compressed of the immovable finger. The merus of the minor at the base of the immovable finger as it is in S. cheliped is armed with an acute spine on the supe- bannerorum. The dactylus of the third pereiopod rior distal margin. The carpus is cup-shaped. The in S. dominicensis has the lower tooth less diver- chela is unarmed; the movable finger is slightly gent than that of S. bannerorum. Synalpheus ban- less than half the length of the chela. The second nerorum is also related to Indo-Pacific members pereiopods are subequal. The fingers are unarmed of the genus, but the form of the chela appears to and the movable finger is slightly more than half distinguish it from all other species of the genus. the length of the chela. The carpus is subdivided into five segments; the first is about six times the ETYMOLOGY.—This species is named for Dora M. length of the second, which is subequal in length Banner and Albert H. Banner who have contri- to the third and fourth, which are about half the buted so much to the knowledge of the Alpheidae length of the fifth. The ischium is about half the and who have shown interest in the present study. NUMBER 176 81

Pomagnathus corallinus Chace, 1937 MEASUREMENTS.—Immature male cl 1.8 mm, Pomagnathus corrallinus Chace, 1937:124, fig. 5.—Schmitt, males cl 2.9-3.8 mm, non-ovigerous females cl 2.9- 1939:12.—Chace, 1962:612. 6.0 mm, ovigerous females cl 4.0-5.1 mm. DISTRIBUTION.—The species has previously been MATERIAL.—3 ovigerous females, Malpelo coll. reported from the Red Sea, western Indian Ocean, no. 3. the Malay Archipelago, Micronesia, New Zealand, MEASUREMENTS.—Ovigerous females cl 4.6, 4.7, Hawaiian Islands, Clipperton Island (Holthuis, 5.4 mm. 1958; Chace, 1962), and now Malpelo Island. DISTRIBUTION.—Arena Bank, Baja California Neither this species nor L. galapagensis appear to (Chace, 1937), Clipperton Island (Schmitt, 1939; occur on the western American mainland. Lysmata Chace, 1962). Specimens from Taboguilla Island californica (Stimpson, 1866) occurs in California in the Gulf of Panama were also examined during and is common in the Gulf of Panama while L. the present study. porteri (Rathbun, 1907) occurs in Chile. The re- REMARKS.—The present material agrees with the cord of Sivertsen (1934) for L. intermedia (Kings- description and illustrations of Chace (1937). The ley, 1878), a western Atlantic species, is based on species commonly occurs on Pocillopora coral in a single specimen from the Galapagos Islands, the Gulf of Panama. which appears to be within the range of variation of L. californica (see Sivertsen, 1934, pi. 2: figs. 9-15; and Limbaugh, Pederson, and Chace, 1961, HIPPOLYTIDAE fig. 7). REMARKS.—Chace (1962) discussed variation in Lysmata galapagensis Schmitt, 1924 this species and noted three differences between Lysmata galapagensis Schmitt, 1924:165, fig. 41 Sivertsen, material from Clipperton Island and material from 1934:22 Hult, 1939:6.—Schmitt, 1939:12—Chace, 1962: the Hawaiian Islands. In some Hawaiian specimens 616. the posterior ventral rostral tooth is behind or MATERIAL.—2 males, Malpelo coll. no. 3. opposite the anterior dorsal tooth while in most MEASUREMENTS.—Males cl 2.6, 3.5 mm. of the Clipperton specimens both ventral teeth, DISTRIBUTION.—This species was previously when present, are well in advance of the anterior known from two records off Santa Cruz Island dorsal tooth. In larger adults from Hawaii the in the Galapagos Islands, and is now reported from fused part of the upper antennular flagellum is Malpelo Island. not much shorter than the free portion of the REMARKS.—Both specimens agree in all respects shorter branch while in large Clipperton speci- with the description and illustration of Schmitt mens the fused portion is distinctly less than half (1924). The rostral formula of both specimens the length of the free portion. The average num- was one postorbital, five preorbital and one sub- ber of carpal segments in Hawaiian material (23) apical ventral teeth. The occurrence of this species tended to be higher than that of Clipperton mate- with L. trisetacea offers additional support (Sch- rial (21). mitt, 1939; Chace, 1962) for the specific distinct- These three characters were examined in the ness of these two species. Malpelo material with the following results. The rostral formula was (number of specimens in pa- rentheses): 2+2/2 (16), 2+3/2 (1), 2+2/3 (1), Lysmata trisetacea (Heller, 1861) 2+2/1 (1), 1+2/2 (1), 1+2/1 (1). All specimens Hippolyte trisetacea Heller, 1861:29. with two ventral teeth had them both placed well Hippolysmata paucidens Rathbun, 1906:913, pi. 24: fig. 4. in advance of the anterior dorsal tooth; those with Lysmata paucidens (Rathbun).—Schmitt, 1939:12. one ventral tooth had it placed well in advance of Lysmata trisetacea (Heller) Holthuis, 1947:19, 65— Chace, the anterior dorsal tooth. The Malpelo material 1962:614. agreed also with the Clipperton material regarding MATERIAL.—6 males, 1 immature male, 5 non- the relative lengths of the fused and free portions ovigerous females, 9 ovigerous females, Malpelo of the upper antennular flagellum. The number coll. no. 3. of carpal segments in Malpelo material varied 82 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY from 21-26 distributed as follows (number of speci- including the following: Harpiliopsis depressus, mens in parentheses): 24 (5), 22 (3), 21 (2), 23 Alpheus lottini, Synalphens charon, S. digueti, (2), 25 (1), 26 (1). The average was slightly more Pomagnathus corallinus, Petrolisthes haigae, P. than 24 agreeing more with the Hawaiian material glasselli, Pachycheles biocellatns, Pagurus sp., Tra- than with the Clipperton material. These differ- pezia digitalis, T. ferruginea, Domecia hispida, ences warrant further consideration when compar- Medaeus spimilifer, Carpilodes cinclimanus, Tel- ing material from throughout the range of L. tri- eophrys cristulipes, and Mithrax pygmaeus. Both setacea as suggested by Chace (1962) and Holthuis Veleronia laevifrons and Pseudocoutierea elegans (1947). are associated with gorgonians. Gnathophylloides mineri occurs on the spines of sea urchins and Periclimenaeus hancocki probably occurs with PALINURIDAE sponges. Two species, Grapsus grapsus and Gecarcinus Panulirus penicillatus (Olivier, 1791) malpilensis, occur intertidally and supratidally, Astacus penicillatus Olivier, 1791:343. respectively. Panulirus penicillatus (Olivier).—Holthuis, 1946:125.—Hol- The remaining species occur in a variety of sub- thuis and Loesch, 1967:217. pi. 8. tidal habitats, mostly with rocks and coral rubble. MATERIAL.—None, the species was identified from color slides. Distribution MEASUREMENTS.—Cl 91-137 mm (Holthuis and The decapod fauna of Malpelo Island has affini- Loesch, 1967). ties with both the western American mainland and DISTRIBUTION.—This species occurs in the Indo- Indo-Pacific decapod faunas (p. 67). Twenty-seven West Pacific region: Red Sea, Southeast Africa to species, 63 percent of the Malpelo fauna, occur on Korea, Formosa, and Polynesia. In the eastern Pa- the western American mainland. Nine species, 21 cific it occurs at Revillagigedo Islands, Clipperton percent of the fauna, occur in the Indo-Pacific re- Island, Cocos Island, Galapagos Island (Holthuis gion and seven of these reach the western Ameri- and Loesch, 1967), and now is reported from Mal- can mainland. The other two, Lysmata trisetacea pelo Island. It has not been found on the western and Panulirus penicillatus occur from the Indo- American mainland. Pacific only to the offshore islands of the eastern REMARKS.—Although no material of this species Pacific. Three species, Gecarcinus malpilensis, Al- was collected, comparison of color slides taken by pheus grahami, new species, and Synalpheus ban- Dr. Charles Birkeland with the color plate of this nerorum, new species, appear to be endemic to species given by Holthuis and Loesch (1967) per- Malpelo Island. Two species, Brachycarpus biun- mitted identification. guiculatus and probably Gnathophylloides mineri are circumtropical in distribution. Lysmata galapa- Ecology gensis is only known from the Galapagos Islands and Malpelo Island, while Synalpheus biunguicu- The short time available to members of the expe- latus is known from Hawaii, Clipperton Island, dition did not permit detailed collecting notes on and Malpelo Island. so many different groups of organisms (see Appen- The decapod fauna of Malpelo Island then has dix in Birkeland et al., this volume). However, its greatest affinities (63%) with that of the west- some observations are possible based on published ern American mainland, next is the Indo-Pacific information and my own experiences in Panama region (21%), then endemic (7%), circumtropical (Abele, 1972). (4%), and island species (4%). Perhaps the most interesting observation is the large percentage of decapods from Malpelo Island which are commensals. Almost half of the 43 Comparison with Other Islands species occur in close association with some other A detailed comparison, at the species level, of invertebrates. Most occur with Pocillopora coral, the decapod faunas of the offshore islands of the NUMBER 176 83

ing, although the slope (z=0.56) is higher than that reported for other area-species curves (z = 0.20-0.35; MacArthur and Wilson, 1967). The present curve may not be representative being based on only three islands and is presented here primarily to stimulate interest in examining this relationship in other marine organisms. fe 1.8. I Literature Cited 8 1.7 Abele, L. G. 1972. Comparative Habitat Diversity and Faunal Rela- tionships between the Pacific and Caribbean Pana- manian Decapod Crustacea: A Preliminary Report with Some Remarks on the Crustacean Fauna of

I.OC PERIMETER OF ISLAND Panama. Pages 125-138 in M. L. Jones, editor. The Panamic Biota: Some Observations Prior to a Sea FIGURE 33.—The relationship between island perimeter and Level Canal. Bulletin of the Biological Society of the number of decapod crustacean species present at (1), Washington, number 2. Malpelo, (2) Clipperton, and (3) Cocos islands. Refer to text Armstrong, J. C. for discussion. 1941. The Caridea and Stomatopoda of the Second Tem- pleton Crocker-American Museum Expedition to eastern Pacific will have to await more data on the Pacific Ocean. American Museum Novitates, 1137:1-14, 4 figures. other islands. 1949. New Caridea from the Dominican Republic. Some comparisons can be made, however, be- American Museum Novitates, 1410:1-27, 9 figures. tween numbers of species of decapods associated Banner, A. H. with oceanic islands (Figure 33). There are 25 1953. The Crangonidae, or Snapping Shrimp of Hawaii. species of decapods known from Saint Helena Pacific Science, 7 (1):2-144, 147, 50 figures, 1 frontis- (Chace, 1966; 1968), an isolated island in the piece. south Atlantic. Twenty percent, five species, are 1956. Collections from the Mariana Archipelago. Part I in Contributions to the Knowledge of the Alpheid endemic. There are 58 species of decapods known Shrimp of the Pacific Ocean. Pacific Science, 10: from Clipperton Island (Chace, 1962; Garth, 1965) 318-373, 23 figures. in the eastern Pacific. None of these species is 1958. On a Small Collection from Onotoa, Gilbert Is- endemic. I have found about 100 species lands. Part III in Contributions to the Knowledge known from Cocos Island in the eastern Pacific of of the Alpheid Shrimp of the Pacific Ocean. Pacific which at least six (6%) are endemic. Malpelo Is- Science, 12(2):157-169, 4 figures. Banner, A. H., and D. M. Banner land has a decapod fauna of 43 species with three 1964. Collections from the Phoenix and Line Islands. (7%) of them being endemic. Part IX in Contributions to the Knowledge of the Within a given region there generally exists an Alpheid Shrimp of the Pacific Ocean. Pacific orderly relation between numbers of species and Science 18:83-100, 5 figures. the size of the area sampled. This relationship has 1966. The Alpheid Shrimp of Thailand: The Alpheid Shrimp of the Gulf of Thailand and Adjacent been worked out for a number of terrestrial organ- Waters. The Siam Society Monograph Series, 3, isms occurring on islands (MacArthur and Wilson, vi + 168 pages, 168 figures. 1967). In comparing shallow water organisms as- Bruce, A. J. sociated with islands, it seems more appropriate 1970. On the Identity of Periclimenes pusillus Rathbun, to use island perimeter rather than island area. A 1906 (Decapoda, Pontoniinae). Crustaceana 19(3): comparison between island perimeter and numbers 306-310, 1 figure. of decapod species of three eastern Pacific islands 1973. Gnathophylloides robustus sp. nov., a New Com- mensal Gnathophyllid Shrimp from Western Aus- (Saint Helena, having a different faunal source, tralia, with the Designation of a New Genus is not included), Malpelo, Clipperton, and Cocos, Levicaris (Decapoda, Caridea). Crustaceana 24(1): is shown in Figure 33. The relationship is strik- 17-32, 9 figures. 84 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Chace, F. A., Jr. "Albatross," during 1891, Lieutenant Commander 1937. Caridean Decapod Crustacea from the Gulf of Cali- Z. L. Tanner, U.S.N., Commanding. Bulletin of the fornia and the West Coast of Lower California. Museum of Comparative Zoology at Harvard Col- Part VII in The Templeton Crocker Expedition. lege 24:149-220. Zoologica (New York), 22(2):109-138, 9 figures. Garth, J. S. 1962. The Non-Brachyuran Decapod Crustaceans of Clip- 1948. The Brachyura of the "Askoy" Expedition with perton Island. Proceedings of the United States Remarks on Carcinological Collecting in the Pan- National Museum, 113(3466):605-635; 7 figures. ama Bight. Bulletin of the American Museum of 1966. Decapod Crustaceans from St. Helena Island, South Natural History 92(1): 1-66, 4 figures. Atlantic. Proceedings of the United States National 1965. The Brachyuran Decapod Crustaceans of Clipper- Museum, 118 (3536):623-661, 15 figures. ton Island. Proceedings of the California Academy 1968. A New Crab of the Genus Cycloes (Crustacea; of Sciences, fourth series, 33(1): 1-46, 26 figures. Brachyura; Calappidae) from Saint Helena, South GueYin-Meneville, F. E. Atlantic Ocean. Proceedings of the Biological So- 1830. Crustaces, Arachnides ct Insectcs. Section 1 of part ciety of Washington, 81:605-612, 2 figures. 2 in volume 2 (zoologie) in M. L. I. Duperrey, 1972. The Shrimps of the Smithsonian-Bredin Caribbean Voyage autour du monde, execute par Ordre du Expeditions, with a Summary of the West Indian Roi, sur la Cornette de Sa Majeste, JM "Coquille," Shallow-water Species (Crustacea: Decapoda: Na- pendant Us annees 1822, 1823, 1824, et 1825. 319 tantia). Smithsonian Contributions to Zoology, 98: pages, 24 plates. Paris. 1-179, 61 figures. Heller, C. Coutiere, H. 1861. Synopsis der im rothcn Mccrc vorkommenden 1909. The American Species of Snapping Shrimps of the Crustaccen. Verhandelinger Zoologische botanische Genus Synalpheus. Proceedings of the United C.esellschaft (Wien) 11:3-32. States National Museum, 36 (1659): 1-93, 54 figures. Holthuis, L. B. Crosnier, A., and J. Forest 1946. The Stenopodidae, Nephropsidae, Scyllaridae and 1966. Crustaces Decapodes: Alpheidae. Part 19 in Cam- Palinuridae. Part I in The Dccopoda Macrura of pagnc de la Calypso dans le Golfe de Guinee et the Snellius Expedition. Part XIV in Biological aux lies Principe, Sao Tome et Annobon (1956), et Results of the Snellius Expedition. Temminckia, 7: Campagne aux lies de Cap Vert (1959). Fascicle 7, 1-178; 2 figures, 11 plates. volume 27, in Resultats scientifiques des Campagnes 1947. The Decapoda of the "Siboga Expedition, IX: The de la "Calypso". Annales de I'Institute Oceanogra- Hippolytidae and Rhynchocinetidae Collected by phique, Monaco, 44:199-314, 33 figures. the Siboga and Snellius Expeditions with Remarks Dana, J. D. on Other Species. Volume 39a8 in Siboga-Expedite. 1852. Crustacea, part I. Volume 13 in United States Ex- Leiden: E. J. Brill. ploring Expedition, During the Years 1838, 1839, 1951. The Subfamilies Euryrhynchinae and Pontoninae. 1840, 1841, 1842, Under the Command of Charles Part I in A General Revision of the Palaemonidae Wilkes, U.S.N. viii + 685 pages. Philadelphia: C. (Crustacea Decapoda Natantia) of the Americas. Sherman. Allan Hancock Foundation Occasional Papers, 11: 1855. Crustacea, Atlas. Volume 13 in United States Ex- 1-332, 63 plates. ploring Expedition, During the Years 1838, 1839, 1952. The Subfamily Palaemoninae. Part II in A General 1840, 1841, 1842, under the Command of Charles Revision of the Palaemonidae (Crustacea Decapoda Wilkes, U.S.N. 27 pages, 96 plates. Philadelphia: Natantia) of the Americas. Allan Hancock Founda- C. Sherman, tion Occasional Papers, 12:1-396, 55 plates. de Man, J. G. 1958. Crustacea Decapoda from the northern Red Sea 1911. Family Alpheidae. Part II in The Decapoda of the (Gulf of Agaba and Sinai Peninsula), I: Macrura. 1 Siboga Expedition. Siboga Expedite, 39a (2): 113— (Contributions to the Knowledge of the Red Sea, 327, 23 plates. no. 8.) State of Israel Sea Fisheries Research Station Fausto Filho, J. Bulletin, 17:1-40, 15 figures. 1966. Brachycarpus holthuisi, Nova Especie de Crustaceo Holthuis, L. B., and H. Loesch do Brasil (Decapoda PaJaemonidae). Arquivos 1967. The Lobsters of the Galapagos Islands (Decapoda, Estacao de Biologia Marinha Universidade do Ceara, Palinuridea). Crustaceana 12 (2):2I4-222; 1 figure, 6 (2): 123-125, 11 figures. plates 7-9. Faxon, W. Hult, J. 1893. Preliminary Descriptions of New Species of- Crus- 1939. Crustacea Decapoda from the Galapagos Islands tacea. Part VI in Reports on the Dredging Opera- Collected by Mr. Rolf Blomberg. Arkiv Zoologiske, tions Off the West Coast of Central America to the 3OA(5):1-18, 4 figures, 1 plate. Galapagos, to the West Coast of Mexico, and in Kingsley, J. S. the Gulf of California, in Charge of Alexander 1878. Notes on the North American Caridea in the Mu- Agassiz, by the U.S. Fish Commission Steamer seum of the Peabody Academy of Science at Salem, NUMBER 176 85

Mass. Proceedings of the Academy of Natural 1907. South American Crustacea. Revista Chile Historic Sciences of Philadelphia, 1878:89-98. naturale, 11:45-50; 1 figure, 2 plates. Lewis, J. B. 1918. The Grapsoid Crabs of America. Bulletin of the 1956. The Occurrence of the Macruran Gnathophylloides United States National Museum, 97, xxii + 461 mineri Schmitt on the Spines of the Edible Sea pages, 172 figures, 161 plates. Urchin Tripneustes esculentus Leske in Barbados. Schmitt, W. L. Bulletin of Marine Science of the Gulf and Carib- 1924. The Macrura and Anomura Collected by the Wil- bean, 6 (4):288-291, 2 figures. liams Galapagos Expedition, 1923. Zoologica (New Limbaugh, C, H. Pederson, and F. A. Chace, Jr. York), 5(15):161-171, figures 39-41. 1961. Shrimps that Clean Fishes. Bulletin of Marine Science of the Gulf and Caribbean, 11 (2):237-257, 1932. Coralliocaris pearsei. In Pearse, Inhabitants of 9 figures. Certain Sponges at Dry Tortugas. (Papers of the Lucas, H. Tortugas Laboratory, volume 23.) Carnegie Institu- 1849. Crustaces, Arachnides, Myriopodes et Hexapodes. tion of Washington Publication, 435:123, 124, fig- In Exploration scientiftque de VAlgerie pendant les ure 1. annees 1840, 1814, 1842. Zoologie I in Sciences 1933. Four New Species of Decapod Crustaceans from physiques. Volume I (403 pages, 8 plates) in His- Porto Rico. American Museum Novitates, 662:1-9, toire naturelle des animaux articule. 4 figures. MacArthur, R. H., and E. O. Wilson. 1935. Crustacea Macrura and Anomura of Porto Rico and 1967. The Theory of Island Biogeography. xi + 203 the Virgin Islands. Pages 125-227 in part 2 of pages. Princeton: Princeton University Press. volume 15 in Scientific Survey of Porto Rico and Milne Edwards, H. the Virgin Islands. 80 figures. New York: New York 1837. Histoire Naturelle des Crustaces, Comprenant Academy of Sciences. I'Anatomie, la Physiologie et al Classification de ces 1939. Decapod and Other Crustacea Collected on the Animaux. Volume 2, 532 pages; atlas, 32 pages, Presidential Cruise of 1938 (with Introduction and plates 1-14, 14 bis, 15-25, 25 bis, 26-42. Paris: Station Data). Smithsonian Miscellaneous Collec- Librarie Encydoped ique de Roret. tions, 98 (6): 1-29, 2 figures, 3 plates. Olivier, A. G. Sivertsen, E. 1791. Ecrevisse. Astacus. Pages 327-349 of volume 6 in 1934. Littoral Crustacea Decapoda from the Galapagos A. G. Olivier, Encyclopedic methodique, Histoire Islands. Part VII in The Norwegian Zoological naturelle, Insects. Expedition to the Galapagos Islands, 1925, con- Patton, W. K. ducted by Alf Wollebaek. Meddelelser fra det 1973. Community Structure among the Animals Inhab- Zoologiske Museum (Oslo), 38:1-23, 1 figure, 4 iting the Coral, Pocillopora damicornis at Heron plates. Island, Australia. Pages 214-238 in W. Vernberg, editor, Symbiosis in the Sea. Columbia: University Stimpson, W. of South Carolina Press. 1860. Prodromus Descriptionis Animalium Evertebra- torum, quae in Expeditione ad Oceanum Pacificum Randall, J. W. Septentrionalem, a Republica Federata missa, 1839. Catalogue of the Crustacea Brought by Thomas Cadivaladaro Ringgold et Johanne Rodgers Duci- Nuttall and J. K. Townsend, from the West Coast bus, Observavit et Descripsit. Proceedings of the of North America and the Sandwich Islands, with Academy of Natural Sciences of Philadelphia, 1860: Descriptions of Such Species as are Apparently New, 22-48. among Them Which are Included Several Species of Different Localities Previously Existing in the 1866. Descriptions of New Genera and Species of Macru- Collection of the Academy. Journal of the Academy rons Crustacea from the Coasts of North America. of Natural Sciences Philadelphia, 8:106-147, plates Proceedings of the Chicago Academy of Sciences, 3-7. 1:46-48. Rathbun, M. J. Turkay, M. 1906. The Brachyura and Macrura of the Hawaiian Is- 1970. Die Gecarcinidae Amerikas: Mif einem Anhang lands. Bulletin United States Fish Commission, 23: fiber Ucides Rathbun. Senckenbergiana biologica 827-930, 79 figures, 24 plates. 51 (5/6):333-354, 11 figures. Asteroidea from Malpelo Island with a Description of a New Species of the Genus Tamaria

Maureen E. Downey

OREASTERIDAE ABSTRACT Nidorellia armata (Cray) One new species, Tamaria stria, and a new sub- Pentaceros (Nidorellia) armata Gray, 1840:277; 1866:7, pi. 14: species, Narcissia gracilis malpeloensis, are de- fig. 1-3. scribed from shallow waters of Malpelo Island. A Oreaster armatus (Gray).—Mueller and Troschcl, 1842:52 — collection totaling six species of starfish is described, Lutkcn, 1864:148—Bell, 1884:79.—von Martens, 1865:433. with notes on distribution. Goniodiscus armatus (Gray).—Lutken, 1859:75. Goniodiscus conifer Mobius, 1859:10, pi. 3: fig. 5-6. Nidorellia armata (Gray).—Verrill, 1867:280—Perrier, 1876: 67—Viguier, 1878:193.—Leipoldt, 1895:634—H. L. Clark, 1910:332, pi. 4: fig. 2.—Doderlein, 1916:418; 1936:316, pi. Introduction 21: fig. l-6a.—Boone, 1926:3, pi. 1; 1933:73, fig. 2.— Ziesenhenne, 1937:216.—H. L. Clark, 1940:333.—Steinbeck In the course of the faunal survey of the shallow and Ricketts, 1941:381, pi. 10: fig. 1, pi. 11: fig. 2—Ely, waters (0-50 m) of Malpelo conducted by C. Birke- 1945:15.—Caso, 1943:9, 66, pi. 22: fig. 1-2, pi. 23: figs. 1-2; land and his colleagues, only six species of starfish 1953:221; 1961:63, figs. 22-23.—H. L. Clark, 1958:95.—Caso, were collected; however, all are species with inter- 1962:63, figs. 22-24. esting or unusual distribution patterns, and in- Nirodella armata (Gray).—Stanek, 1955:48. cluded in the collection are a new species of the Nidorellia armata is represented in this collec- genus Tamaria and a new subspecies of Narcissis tion by one specimen, quite large and heavy (R gracilis. The abbreviation R refers to the major 10 cm, r 6.5 cm, dry weight 177 grams). The mar- radius from center of disc to tip of arm; and r re- ginals, particularly the distal ones, are tremen- fers to the minor radius, from center of disc to dously swollen and most bear 1 or 2 stout spines or interradial margin. low tubercles. The much smaller principal plates ACKNOWLEDGMENTS.—I wish to thank Charles of the abactinal surface usually bear 1 or 2 low Birkeland and David Meyer of the Smithsonian tubercles. The papular areas of the abactinal sur- Tropical Research Institute for the opportunity to face are covered with pedicellariae of the split examine this collection, and Dr. David L. Pawson, granule type. This is a common shallow water Chairman, Department of Invertebrate Zoology, species of the eastern tropical Pacific, known from National Museum of Natural History, Smithson- lower California to Peru and the Hawaiian Islands. ian Institution, for reading the manuscript and offering helpful comments and suggestions. OPHTOIASTERTOAE Leiastef callipeplus Fisher Maureen E. Downey, Department of Invertebrate Zoology, National Museum of Natural History, Smithsonian Institu- Leiaster callipeplus Fisher, 1906:1083, pi. 30: figs. 1, la, pi. tion, Washington, D.C. 20560. 31: fig. 3.

86 NUMBER 176 87

Leiaster callipeplus is a Hawaiian species, re- dial, superomarginal, and inferomarginal plates ported by Fisher from Maui, Lanai, Kauai, and diamond-shaped, slightly overlapping distal corner Bird Islands, in 58-124 meters. One fine specimen, of preceding plate, connected transversely to adja- collected at Malpelo in 49 meters at the base of a cent row of plates by small, irregular, secondary rock wall (Station 5, Birkeland et al., Appendix, plates. Six rows of papular areas, 6-10 pores per this volume) measures R 20 cm, r 2.9 cm. Leiaster area. Three rows of rounded actinal intermediate callipeplus has also been collected near the central plates, becoming two distally. Entire body, up to American mainland in the Gulf of Chiriqui, Pana- adambulacral armature, covered with close uni- ma (Birkeland, pers. comra.). form coating of granules, those in papular areas slightly smaller than those on plates. Adambulac- ral furrow spines two, one small, acicular, the other Narcissia gracilis malpeloensis, new subspecies heavy, flared at tip, broad in plane of furrow. Sub- Narcissia gracilis malpeloensis is represented in ambulacral spines single, heavy, thick, longer than this collection by six specimens. The species, de- broad. No pedicellariae noted. Madreporite small, scribed by A. H. Clark (1916:58) has previously round, covered with fine radiating gyri. Anus been collected from Baja California to the Gala- subcentral, inconspicuous. Ocular plates large, pagos Islands. The average size of these specimens rounded, raised, mostly bare, with few granules. is R 12 cm, r 2.5 cm. The Malpelo specimens repre- Color in life is reddish orange; dried, bright sent a new subspecies of N. gracilis, differing from orange with blue papular areas. the type-specimen and Clark's description in the MATERIAL EXAMINED.—Holotype, USNM El 1838, following respects: Station 4, Malpelo Island, 49 m, on rubble near Narcissia gracilis gracilis: Pedicellariae abundant base of vertical rock wall. Paratypes, USNM on all surfaces, of 2 slender valves with expanded El 1839, 4 specimens, Stations 4 and 5, Malpelo toothed tips, lying in alveoli; granules hemispher- Island, 36-49 m. ical; subambulacral spines in 3 rows; madreporite ETYMOLOGY.—The species name is the latinized round; papulae single; and proximal marginals acronym for the Smithsonian Tropical Research longer than broad. Institute. Narcissia gracilis malpeloensis: Few or no pedi- DISCUSSION.—Only one other species of Tamaria, cellariae—when present, of 2 stout curved un- T. obstipa Ziesenhenne, has been described from toothed valves of uniform thickness, not in alveoli; the eastern tropical Pacific, Tamaria obstipa is granules flat-topped, polygonal; subambulacral known from the type-locality, Cocos Island, Costa spines in one row, plus other spines not in rows; Rica, and from James Island, Galapagos Islands. madreporite triangular; papulae usually double; The present species agrees closely with T. obstipa, and proximal marginals broader than long. but differs in having secondary connecting plates Although the above differences are minor, and not only intermarginally but between all abacti- abundant pedicellariae versus few or no pedicel- nal plate rows; in having a carinal series, lacking lariae might be accounted for by the difference in in T. obstipa; in the nature of the adambulacral size (holotype: R 54 mm, r 8.5 mm) between these armature; and in the lack of pedicellariae. The specimens and the type ( frequently have number of pores per papular area is smaller in T. many pedicellariae when young and few or none stria (6-10) but is within the range of T. obstipa when fully grown), the very dissimilar character of (6-24). the pedicellariae, plus the other variations listed above seem, in view of Malpelo's isolated position, to warrant separation at the subspecific level. MITHRODIIDAE Mithrodia bradleyi Verrill

Tamaria stria, new species Mithrodia bradleyi Verrill, 1867:288— Perrier, 1878:77.— FIGURE 34 Sladen, 1889:539.—Fisher, 1906:1094, pi. 36: figs. 1, 2, pi. 37: figs. 1-3; 1925:68; 1928:491.—H. L. Clark, 1910:327, pi. DESCRIPTION.—R 44 mm, r 8 mm. Disc small, 6: fig. 1.—A. H. Clark, 1946:9.—Steinbeck and Ricketts, arms five, long, blunt, cylindrical. Carinal, adra- 1941:380, pi. 23: fig. 3.—Ely, 1945:27, pi. 8A, B.—Caso, 1944: 88 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 34.—Tamaria stria, new species, holotype, USNM El 1838. NUMBER 176 89

253; 1953:214; 1961:92, figs. 34-36; 1962:92, figs. 34-36; Caso, M. E. 1963:300.—Engel, John, and Cherbonnier, 1949:1. 1943. Contribucion al conocimiento de los asteridos de Mithrodia clavigera Perrier, 1875:360.—Ives, 1889:171. Mexico. 136 pages, 50 plates. Thesis, Universidad Nacional de Mexico, Faculte de Ciendas. Three specimens of Mithrodia bradleyi were col- 1944. Estudios sobre asteridos de Mexico. Algunas lected at Malpelo at depths from 12 to 14 meters; especies interesantes de asteridos litorales. Anales the known distribution for this species is from Baja Instituto Biologico de Mexico, 15 (l):236-259. California to the Galapagos Islands. The Malpelo 1953. Estado actual de los conocimientos acerca de la specimens have many more pedicellariae than fauna de los equinodermos de Mexico. Memoria del Congreso Cientifico Mexicano, 7:209-222, 12 specimens from Baja California in the collections figures. of the National Museum of Natural History, 1961. Los equinodermos de Mexico. 388 pages, 124 fig- Smithsonian Institution, but correspond well with ures, 20 plates. Universidad Nacional de Mexico, material from the Colombian coast. Faculte de Ciendas. 1962. Estado actual de los conocimiento acerca de los equinodermos de Mexico. 388 pages, 124 figures, 20 PORANIIDAE plates. Universidad Nadonal de Mexico, Faculte de Ciencias. 1963. Estudios sobre equinodermos de Mexico. Contri- Asteropsis carinifera (Lamarck) bucion al conocimiento de los equinodermos de las islas Revillagigedo. Anales del Instituto Biologico Asterias carinifera Lamarck, 1816:556. de Universidad de Mexico, 33:293-330, 9 plates. Aster ope carinifera (Lamarck).—Mueller and Troschel, 1840: Clark, A. H. 104.—H. L. Clark, 1920:33, pi. v: fig. 2. Asteropsis carinifera (Lamarck): Mueller and Troschel, 1840: 1916. Six new Starfishes from the Gulf of California and Adjacent Waters. Proceedings of the Biological 322; 1842:63.—A. M. Clark, 1967:37.—McKnight, 1968:713. Society of Washington, 29:51-62. Gymnasterias inermis Gray, 1840:278. Gymnasterias spinosa Gray, 1840:278. 1946. from the Pearl Islands. Smithsonian Gymnasterias carinifera (Lamarck).—deLoriol, 1885:67, pi. Miscellaneous Collections, 106 (5): 11. 20: figs. 7-10.—Leipoldt, 1895:649, pi. 32: fig. 13. Clark, A. M. 1967. Echinoderms from the Red Sea, Part 2: Crinoids, One specimen of Asteropsis carinifera was col- Ophiuroids, Echinoids and More Asteroids. (Num- lected at Malpelo on a vertical rock wall at 36 ber 21 in Reports of the Israel South Red Sea Ex- pedition 1962.) Sea Fisheries Research Station Bulle- meters. It is distributed throughout the Indo- tin (Haifa), 41:26-58, 5 figures. Pacific and the tropical eastern Pacific. This species Clark, H. L. is apparently attacked by the shrimp Hymenocera 1910. The Echinoderms of Peru. Bulletin of the Museum (Birkeland, pers. comra.), well known as a pre- of Comparative Zoology at Harvard, 52(17):321- dator on Acanthaster plancii, the crown-of-thorns 358, 14 plates. starfish. 1940. Notes on Echinoderms from the West Coast of Central America. Zoological Society, 25 (3a):331-352, 2 plates. 1958. Estrellas de Mar (Asteroidea). Biota, 2:85-105, 13 Literature Cited figures. Bell, F. J. Doderlein, L. 1884. Contributions to the Systematic Arrangement of 1916. Ueber die Gattung Oreaster und Verwandte, Zool- ogische Jahrbuch (Systematischen), 40:409-440. the Asteroidea, II: The Species of Or easier. Pro- 1936. Die Asteriden der Siboga-Expedition, III: Die ceedings of the Zoological Society of London, 1884: Unterfamilie Oreasterinae. Pages 295-369 of volume 57-87. 46c in Siboga-Expeditie. Plates 21-32. Leiden: E. J. Boone, L. Brill. 1926. Echinoderms from the Gulf of California and the Ely, C. A. Perlas Islands. Bulletin of the Bingham Oceano- 1945. Shallow Water Asteroidea and Ophiuroidea of graphic Collection (New York), 2 (6): 1-14, 9 plates. Hawaii. Bulletin of the Bernice A. Bishop Museum, 1933. Scientific Results of Cruises of the Yachts Eagle 176(1942):l-63, 8 figures, 13 plates. and Ara, 1921-1928, Win. K. Vanderbilt, Command- Engel, H., D. D. John, and G. Cherbonnier ing. Bulletin of the Vanderbilt Marine Museum 1949. The Genus Mithrodia Gray, 1940. Zoologisch Ver- (Huntington), 4:68-164, plates 25-102. handelungen (Leiden), 2 (1949): 1-39, 12 figures. 90 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Fisher, W. K. McKnight, D. G. 1906. The Starfishes of the Hawaiian Islands. Bulletin of 1968. Some Echinoderms from Tongatabu Island and the the United States Fish Commission, 23 (3,1903):987- South Minerva Reef. New Zealand Journal of Ma- 1130, 49 plates. rine and Freshwater Research, 2:712-715. 1925. Sea Stars of the Tropical Central Pacific. Bulletin Mobius, K. of the Bernice P. Bishop Museum, 27:63-87, plates 1859. Neue Seesterne des Hamburger und Kieler Mu- 5-8. seums. 14 pages, 4 plates. 1928. Sea Stars from the Arcturus Oceanographic Expedi- Mueller, J., and F. H. Troschel tion. Zoologica, 8:487:493. 1840. [Not seen, title unknown.] Akademie Wissen- Gray, J. E. 1840. A Synopsis of the Genera and Species of the Class shaften. Monatsberichtc der Koniglichen Preuss. Hypostomata (Asterias, Linnaeus). Annals and 1842. System der Asteriden. 134 pages, 12 plates. Braun- Magazine of Natural History, series 1, 6:175-184, schweig: Friedrich Viewig und Sohn. 275-290. Perrier, E. 1866. Synopsis of the Species of Starfish in the British 1875. Revision de la collection de stellerides du Museum Museum. 17 pages, 16 plates. London: John Van d'Histoire Naturelle de Paris (Astcriidae, F.chinas- Voorst. teridae, Ophidiastcridae). Archives de Zoologie Ex- Ives, J. E. perimentale et Generate, 4, 5: 384 pages. 1889. Catalogue of the Asteroidea and Ophiuroidea in 1876. Revision dc la collection de Stellerides du Museum the Collection of the Academy of Natural Science d'Histoire Naturelle de Paris. Archives de Zoologie of Philadelphia. Experimental et Ginerale, 5:1-104, 203-309. Lamarck, J. B. 1878. £tude sur la repartition geographiquc des Astcrides. 1816. Histoire naturelle des animaux san vertebres. Edi- Nouvelles Archives du Museum, (2): 1-108. tion 1, volume 2, 568 pages. Paris. Sladen, P. Leipoldt, F. 1889. Report on the Asteroidea collected by H.M.S. Chal- 1895. Asteroidea der "Vettor-Pisani" Expedition (1882- lenger. Pages 1-893 of volume 30 in Report on the 1885). Mit Anhang: Die von F. Orsini im rothen Scientific Results of HM.S. Challenger, Zoology. 117 Meere gesammelten Asteroideen. Zeitschrift wissen- plates. schaften Zoologische, 59:545:654, plates 31, 32. Stanek, V. J. de Loriol, P. 1955. La Beaute de la nature. 373 pages, illustrated. 1885. Catalogue Raisonne des Echinodermes recueillis Prague: Artia. par M. V. de Robillard a Tile Maurice, II: Steller- Steinbeck, J., and E. F. Ricketts ides. Memoire Societe physique Histoire naturelle 1941. Sea of Cortez. Pages 10 + 598, 40 plates. New York. Geneve, 29 (4): 1-84, plates 7-22. Verrill, A. E. Lutken, C. 1867. Notes on Radiata. Transactions of the Connecticut 1859. Bidrag til Kundskab om de ved Kysterne af Mel- Academy of Arts and Sciences, 1 (2):247-613, plates lemog Syd-Amerika levende Arter af Sdstjerner. 4-10. Videnskabelige Meddelelser fra den naturhistoriske Viguier, C. Forening i Kjovenhavn, pages 25-96. 1878. Anatomie comparee du squelette des Stellerides. 1864. Kritiske Bemaerkninger om forskjellige Sostjerner Archives de Zoologie Experimentale de Genirale, (Asterider), med Beskrivelse af nogle nye Arter. 7:33-250, 11 plates. Videnskabelige Meddelelser fra den naturhistoriske Ziesenhenne, F. Forening i Kjovenhavn, 8-12:135. 1937. The Templeton Crocker Expedition: Echinoderms von Martens, E. from the West Coast of Lower California, the Gulf 1865. Ueber zwei Seesterne von Costa Rica. Annals and of California and Clarion Island. Zoologica, 22 (15): Magazine of Natural History, 1865:433. 209-239, figure 2. Fishes Collected at Malpelo Island

John E. McCosker and Richard H. Rosenblatt

of Mycteroperca olfax, Apogon atradorsatus, ABSTRACT Eupomacentrus arcifrons, and E. beebei, species otherwise known only from those localities. Other Seventy species of shore fishes are reported from specimens collected included fishes previously Malpelo Island, the majority of them previously unrecorded from that locality. Four species are though to be Galapagos endemics, e.g., Labrisomus undescribed and may be endemic to Malpelo. No dendriticus and possibly a deep water species of new taxa are described. On the basis of these col- Lythrypnus. The presence of certain reef-associated lections, it appears that the fauna is primarily Indo-Pacific species, e.g., Gymnothorax flavimar- eastern Pacific mainland in composition, although ginatus, Zanclus canescens, and Acanthurus glau- five species are shared only with the Galapagos and copareius, is relatable to hydrographic and ecologi- Cocos islands. cal conditions found on other eastern tropical Pa- cific offshore islands (Rosenblatt, McCosker, and RubinofF, 1972). Included in the Malpelo collec- The fish collections made by the expedition con- tions were undescribed, and apparently endemic, tain the first shore fishes taken by rotenone ichthy- species of Chriolepis, Acanthemblemaria and Axo- ocides and diving at Malpelo. Previous collections clinus, and an undescribed genus and species of of Malpelo fishes were limited to those caught by tripterygiid. hook and line (Fowler, 1938, 1944) or dipnetted The following annotated list of fishes from Mal- beneath surface night lights (Clemens, 1957; Cle- pelo Island is based primarily on those collected mens and Nowell, 1963). Important material repre- by J. B. Graham and party at two localities, and senting several new species and many new locality underwater photographs and sight records made records was obtained by the STRI expedition. The by C. Birkeland (pers. comm.). Fishes from the collections, however, were limited in scope and are first locality (SIO 72-96) were collected along a not sufficiently representative of the probable sloping basalt and sparsely covered Pocillopora total ichthyofauna to allow any definitive state- bottom between the surface and 6 meters along the ments concerning the faunal composition. western edge of the island. The second collection Malpelo occupies an important zoogeographic (SIO 72-97) is a composite of small east and west location in that it arises from deep water beyond shore stations, from the surface to depths of 20 the continental shelf, and could provide a meters. Fishes and photographs are deposited in "stepping-stone" to the Galapagos Archipelago for the Marine Vertebrates Collection of the Scripps shore fishes. The known fauna is mainly eastern Institution of Oceanography (SIO), University of Pacific mainland in origin. However, there are California at San Diego. Verified records from certain similarities to both the Galapagos and hook and line and night-light stations are also in- Cocos islands. This is evidenced by the presence cluded. The basis of identification is indicated by the following abbreviations: * specimens collected John E. McCosker, Steinhart Aquarium, California Academy and deposited at sio, s sight record by STRI diving of Sciences, Golden Gate Park, San Francisco, California party, p photographed underwater by C. Birke- 94118. Richard H. Rosenblatt, Scripps Institution of Ocean- 1 2 ography, La Jolla, California 92037. land, recorded by Fowler (1938), recorded by 91 92 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Fowler (1944), 3 recorded by Clemens (1957), C. tnelampygus Cuvier and Valenciennes P 4 recorded by Clemens and Nowell (1963). Elagatis bipinnulatis (Quoy and Gaimard) 2 LUTJANIDAE CARCHARHINIDAE Lutjanus Jordan; (Gilbert) l.P Carcharhinus galapagensis (Snodgrass and Heller)l Lutjanus viridis (Valenciennes) 1.2.P C. falciformis (Muller and Henle)2 as C. malpeloensis C. leucas (Muller and Henle) S POMACENTRIDAE Triaenodon obesus Ruppell S Chromis atrilobata (Gill) •*,• Eupomacentrus arcifrons (Heller and Snodgrass).'.• Known SPHYRNIDAE only from Malpelo, the Galapagos, and Cocos islands. Sphyrna sp. S E. beebei (Nichols).* Previously known only from the Gala- pagos and Cocos islands. MURAENIDAE Microspathodon dorsalis (Gill) *

Gymnothorax flavimarginatus (Riippell).P Abundant in the LABRIDAE West Indo-Pacific and Hawaii, and several eastern Pacific offshore and nearshore localities (Rosenblatt, et al., 1972). Thalassoma lucasanum (Gill) • G. castaneus Jordan and Gilbert P Bodianus diplotaenia (Gill) !»• G. dovii (Gunther)P Muraena lentiginosa Jenyns P SCARIDAE Scar us rubroviolaceus BleekerS

Myripristis leiognathus Valenciennes • 3 KYPHOSIDAE Holocentrus suborbitalis (Gill) S Sectator ocyurus (Jordan and Gilbert) S Kyphosus sp. S KUHLIIDAE Kuhlia taeniura (Cuvier) S.1,2 CHAETODONTIDAE Holacanthus passer Valenciennes I.* APOGONIDAE Heniochus nigrirostris (Gill) P Pomacanthus zonipectus (Gill) S Apogon atradorsatus Heller and Snodgrass.* Previously known only from the Galapagos and Cocos islands. CIRRHITIDAE SERRANIDAE Cirrhitichthys oxycephalus (Bleeker) • Cirrhilus rivulatus Valenciennes 1,2,* Epinephelus dermatolepis Boulenger • E. labriformis (Jenyns) • Paranthias furcifer (Valenciennes) *1 SCORPAENIDAE Mycteroperca olfax (Jenyns).l Previously known from the Scorpaenodes xyris (Jordan and Gilbert) • Galapagos, Malpelo, and Cocos islands. Phillip C. Heemstra Scorpaena plumieri Bloch • had kindly reexamined Fowler's specimens (ANSP 89121) and found their gill rakers and dorsal spine profile to AULOSTOMATIDAE agree with Galapagos specimens of M. olfax. Aulostomus chinensis L.S MULLIDAE FISTULARIIDAE Pseudupeneus grandisquamis (Gill).4 Based on surface dip- netted prejuvenile specimens. Fistularia petimba Lac^pede S Mulloidichthys dentatus (Gill) S GOBIIDAE CARANGIDAE Lythrypnus sp. A * Selar crumenophthalmus (Bloch) 1 Lythrypnus sp. B • Caranx caballus GiintherS Chriolepis lepidotus Findley (see paper by Findley in this C. vinctus Jordan and Gilbert 3 volume) NUMBER 176

GOBIESOCIDAE CANTHIGASTERIDAE

Tomicodon cf. petersi (Garman)# Canthigaster punctatissima (Giinther) • Arcos decor is Briggs • Labrisomus cf. dendriticus (Reid).* Previously known only TETRAODONTIDAE from the Galapagos Islands. Parastathmonotus culebrai (Seale) • Arothron meleagris (Bloch and Schneider) •

TRIPTERYGIIDAE BALISTIDAE

Axoclinus new species • Sufflamen verves (Gilbert and Starks) 1 Genus and species.* An undescribed species, closely related Melichthys niger (Bloch).l.P As M. buiva in Fowler, 1938. to Enneapterygius corallicola Kendall and Radcliffe from Canthidermis maculatus (Bloch). As C. sp. in Clemens, 1957. the Galapagos Islands, and an undescribed species from Cantherhines dumerilii (Hollard) S Panama and Costa Rica. Alutera scripta (Osbeck) S

CHAENOPSIDAE Literature Cited Acanthemblemaria new species • Clemens, H. B. 1957. Fishes Collected in the Tropical Eastern Pacific, BLENNIIDAE 1954. California Fish and Game, 43:299-307. Clemens, H. B., and J. C. Nowell Ophioblennius steindachneri Jordan and Evermann 3,4,* 1963. Fishes Collected in the Eastern Pacific during Tuna Entomacrodus chiostictus (Jordan and Gilbert) * Cruises, 1952 through 1959. California Fish and HypsobUnnius brevipinnis (Giinther) • Game, 49:240-264. Plagiotremus azaleus (Jordan and Bollman) * Fowler, H. W. 1938. The Fishes of the George Vanderbilt South Pacific ANTENNARIIDAE Expedition, 1937. Monographs of the Academy of Natural Sciences of Philadelphia, 2: v + 349 pages, Antennarius sanguineus Gill * 10 plates. [Malpelo, pages 5-6.] 1944. Results of the Fifth George Vanderbilt Expedition (1941). Monographs of the Academy of Natural Sciences of Philadelphia, 6:57-529. Acanthurus glaucopareius Cuvier • Rosenblatt, R. H., J. E. McCosker, and I. Rubinoff A. xanthopterus Valenciennes S 1972. Indo-West Pacific Fishes from the Gulf of Chiriqui, laticlavius (Valenciennes) S Panama. Contributions in Science of the Los An- Zanclus canescens (Linnaeus) S geles County Museum, 234:1-18, 3 figures. A New Species of Goby from Malpelo Island (Teleostei: Gobiidae: Chriolepis)

Lloyd Talbott Findley

ing persons: C. E. Dawson of the Gulf Coast Re- ABSTRACT search Laboratory, Mississippi, J. E. McCosker of the California Academy of Sciences, and D. A. Two specimens representing a new species of the Thomson of the University of Arizona for reading gobiid fish genus Chriolepis were collected during the 1972 Smithsonian Institution-U.S. Navy Expe- and improving the manuscript; J. B. Graham of dition to Malpelo Island. The new species is herein the Smithsonian Tropical Research Institute for described, and its taxonomic relationships are providing collection data and editorial advice; and briefly discussed. Jenean Thomson of the University of Arizona for the illustration. I thank C. E. Dawson and D. F. Hoese of the Australian Museum for encourage- Introduction ment, and R. J. Lavenberg of the Los Angeles County Museum of Natural History for examina- METHODS.—Measurements were made with dial tion of the type-species. This paper is a contribu- calipers to the nearest 0.1 mm, and are expressed tion from the University of Arizona Marine Bio- as percent of standard length (SL) in Table 9. logy Program. Counts and measurements requiring explanation are as follows: Head length is distance from ante- rior margin of upper lip to upper attachment of Chriolepis lepidotus, new species opercular membrane. Head depth is depth of head at the vertical, posterior margin of preoperculum. FIGURE 35 Head width is the maximum width between cheeks. HOLOTYPE.—USNM 211456, adult male, 30.0 mm Depth at anal fin origin is shortest distance be- SL, Malpelo Island, Colombia, east side of island, tween base of anal spine and dorsal fin base. Post- depth ca. 10 m, above the coral zone, cobble bot- orbital distance is the distance from posterior rim tom with some calcareous sand and boulders, very of eye to upper attachment of opercular mem- sparse algal growth; collected by J. B. Graham brane. Transverse scale rows is the number of scale using rotenone ichthyocide and SCUBA, 2-3 March rows, counted upward and forward, between the 1972. last anal ray and dorsal fin base. The last dorsal PARATYPE.—USNM 211457, adult male, 29.1 mm and anal rays, branched to their bases, were SL; collected with holotype. counted as one. DESCRIPTION.— (See Table 9 for meristic and mor- The holotype and paratype (the only known phometric data). A robust species. Head plump, specimens) are deposited in the National Museum with wide cheeks, more or less rounded; depth at of Natural History (USNM), Smithsonian Institu- preoperculum 0.71 (holotype) and 0.68 (paratype) tion, Washington, D. C. of head width. Mouth oblique, forming an angle ACKNOWLEDGMENTS.—Thanks are due the follow- about 40°-45° with body axis, the posterior angle Lloyd Talbott Findley, Department of Biological Sciences, reaching a point approximately under rear margin University of Arizona, Tucson, Arizona 85721. of pupil. Interorbital narrow, about one-third or

94 NUMBER 176 95 less of eye length. Pectoral base broad. Gill slit including mandibular area, and similar to other narrow; opercular membrane attachments at upper species in the genus (Figure 35). A low, shallowly and lower edges of pectoral base in advance of in- V-shaped, transverse, fleshy crest on head behind sertion of fin rays. eyes. A low, short, fleshy "keel" extending medially Gill rakers short on first arch, those on second between nostrils. Nostrils tubular; anterior tube and following arches shorter; 3+8=11 on outer narrow, its length about equal to pupil diameter; face of first arch; 9 on lower limb of second arch. posterior tube wider, its length about one-half that Pseudobranchiae in four small tufts. of anterior tube. Tongue tip truncate. Head without pores or barbels. Sensory papillae Teeth in rows in both jaws, relatively large, on head relatively large; in rows below eye, on sharp, mostly caniniform, slightly to distinctly re- cheek, at edge of preoperculum, on operculum, on curved, mostly depressible and with faint yellowish snout, and on dorsal and ventral surfaces of head apical tips; outer row teeth in upper jaw enlarged; outer and inner row teeth in lower jaw enlarged. Lower jaw with (1) abbreviate outer row of large, TABLE 9.—Counts and measurements of the holo- type and paratype of Ghriolepis lepidotus slender, recurved, well-separated teeth confined to anterior portion of jaw; (2) behind outer row an- Holotype Paratype teriorly is a patch of many small (about % to 14 Characters (USNM (USNM length of outer row teeth) closely set, less recurved 211456) 211457) teeth in irregular rows, becoming fewer in number, Sex $ more widely spaced, and confined to one row poste- Standard Length (mm) 30.0 29.1 riorly at side of jaw; (3) inner row of large, slender, COUNTS recurved, well-separated teeth anteriorly, becom- ing more widely spaced and stouter posteriorly, Dorsal fin rays VII-1,10 VII-1,10 Anal fin rays 1,9 1,9 until 2 to 3 much enlarged and recurved "fanglike" Pectoral fin rays (left/right) 20/20 20/20 teeth at side of jaw, followed by a few, much Pelvic fin rays 1,5 13 smaller, closely set teeth. Upper jaw with (1) outer Segmented caudal fin rays 17 17 row of large, slender, recurved, well-separated Branched caudal fin rays 15 14 teeth anteriorly and at side of jaw (not absent Procurrent caudal fin rays posteriorly as in outer row of lower jaw); (2) be- (upper/lower) 5/5 6/5 Gill rakers (outer face of first arch) 11 11 hind outer row anteriorly is a patch of many small, Longitudinal scale rows ca.34 ca.35 closely set, less recurved teeth in irregular rows, Transverse scale rows 13i/2 13i/2 becoming fewer in number posteriorly; (3) inner row of many small, closely set teeth, becoming more PROPORTIONS (% of SL) widely spaced posteriorly. No teeth on vomer or Head length 33.3 32.2 palatines. Head depth 15.5 15.6 Head width 21.8 23.5 First dorsal fin with seven flexible spines; bases Base of second dorsal fin 28.5 275 of first five spines equally spaced; the last two Depth at anal fin origin 16.8 175 spines widely separated from the first five, their Pectoral fin length 27.3 24.7 bases widely spaced. Third and fourth spines slight- Pelvic fin length 205 21.0 ly the longest, but no spine notably longer than Caudal fin length 24.7 23.5 Caudal peduncle length 16.0 13.6 any other, and none filamentous. Caudal peduncle depth 13.7 13.8 Second dorsal fin separated from first dorsal fin Eye length 6.7 6.2 by distance slightly more than eye length. First Snout length 7.6 6.9 element of both second dorsal and anal fins a flexi- Mouth length 13.4 125 ble spine; remaining elements are subequal, seg- Width between mandible tips 10.2 10.1 mented, branched rays, with the last branched to Interorbital width 2.4 1.9 Width between posterior nostrils 4.1 5.1 their bases. Length between anterior & Pectoral fins when adpressed reach past tips of posterior nostrils 2.9 2.3 pelvic fins to level of end of spinous dorsal; all Postorbital distance 19.6 19.1 pectoral rays branched. 96 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Origin of pelvic fins in advance of pectoral base. coming progressively smaller, thinner, more em- Pelvic fins short, not reaching anus; separated and bedded, and with fewer and shorter ctenii, until diverging, not united by a membrane into an ad- cycloid anteriorly. Ctenii becoming progressively hesive disc as in most gobiids; interspinal mem- restricted to central portion of posterior edge of brane absent, but pelvic bases slightly connected each scale proceeding anteriorly, until completely by a very low, inconspicuous frenum. Pelvic spine lost at about twelfth scale along midline in ad- flexible; segmented pelvic rays well developed and vance of hypural. Anteriormost scales cycloid, branched, most with three main branches, the minute, embedded, and difficult to see. innermost ray branching less than others, but not A naked area beginning near base of sixth dor- reduced or splinted to the fourth. Fourth ray sal spine, equivalent to one scale width from base slightly the longest, then the third, then the fifth of sixth spine to below base of fifth spine, to two (which is almost equal to the third). Caudal fin scale widths below fourth spine, becoming wider short, rounded. and sloping downward and forward to near upper Urogenital papilla of male long, slender, atten- margin of pectoral base. A lower naked area from uate; with a few scattered melanophores. under pectoral base sloping downward and back- SQUAMATION.—Four enlarged strongly ctenoid ward to near midline of venter, continuing as a basicaudal scales (some lost from both specimens); narrow strip around anus to about base of fourth the outer two (at upper and lower angles of caudal anal ray. Head, nape, chest, and venter scaleless. fin base) slightly narrower and longer than the COLORATION IN ETHYL ALCOHOL.—Background inner two. Body extensively scaled. Scales along color of head and body brownish yellow. Head midline extending forward from hypural to near with a number of prominent black spots, especially axil, beneath origin of spinous dorsal fin (Figure laterally and ventrally. An elongate spot at mid- 35), in about 34-35 more or less irregular longitu- posterior edge of eye (diffuse in paratype); one dinal rows; difficult to enumerate (some scales lost behind eye at dorsal edge of cheek near end of low from both specimens, but scale pockets usually transverse dorsal crest; one at dorsal margin of discernible). Fewer scales in anterior rows; in a preopercular-opercular junction; one at midante- wedge-shaped pattern as approaching axil. Trans- rior edge of operculum which begins a row of verse scales difficult to enumerate, about 13i/£ about 9 or 10 widely spaced spots lying in pre- in an irregular row (first scale at base of last anal opercular-opercular groove, following curvature of ray very narrow in comparison to scales above it, groove downward and forward and onto the imme- counted as i/2 scale). diately adjacent branchiostegal membrane, then Posterior scales large, with prominent ctenii; be- continuing forward ventrally below lower jaw to

FIGURE 35.—Chriolepis lepidotus, holotype, USNM 211456, male, 30.0 mm SL, Malpelo Island, Colombia. Most scales not shown; the dashed line indicates anterior limit of squamation. The row of specialized basicaudal scales is a composite from paratype and right side of holotype. Sensory papillae on head shown as small circles and are a composite from both specimens. NUMBER 176 97 near symphysis. This row is partially duplicated by Dorsal fins with about four or five oblique rows a shorter row of about 5 spots beginning at lower of dark lines, noticeable as prominent black spots posterior edge of preoperculum, running forward on spines and rays (membranes torn in both speci- below lower jaw to near symphysis, the spots lying mens). A dark horizontal bar on anal fin mem- close to and opposite the spots of row described brane, running the length of fin near its edge. A above. About 5-8 widely spaced spots scattered over few faint diffuse spots on caudal rays suggest three cheek (more prominent on paratype). A spot (on or four vertical lines. Pectoral fins slightly dusky paratype) at posterior end of interorbital, fol- on interadial membranes. Pelvic fins dark. lowed by profuse minute melanophores on the low Chest and venter whitish, with numerous minute transverse dorsal crest. A spot at posterior dorsal melanophores, appearing "peppered" on micro- corner of operculum. A prominent spot on upper scopic inspection. Inside of gill chamber with a and lower angles of outer pectoral fin base; a more dark horizontal stripe ventrally, in advance of distal irregular spot at bases of fifth to eighth pec- lower pectoral base; a few melanophores dorsally toral fin rays (slightly larger on paratype). A small in area surrounding pseudobranchial tufts. spot on inner pectoral fin base near upper margin. RELATIONSHIP.—The general head shape, rela- A black bar below anteroventral edge of eye, tively high number of fin rays, and general color passing obliquely downward across suborbital and pattern suggest a relationship with a species com- jaws, terminating in anteriormost ventral spot of plex of Chriolepis, most of whose members in the the series described above. A fainter dark bar ex- eastern tropical Pacific are undescribed. Among tends from midventral edge of eye downward across the described species, C. lepidotus is perhaps clos- suborbital region, and curves around end of lower est to C. minutillus Gilbert, 1891, and C. tagus jaw. Snout, nostrils, upper jaw, and anterior por- Ginsburg, 1953 (judging by original description tion of lower jaw densely peppered with minute only). Complete discussion of relationships within melanophores. the genus Chriolepis and allied genera of seven- Two wide, transverse, light brown bars across spined gobiids with separate pelvic fins is deferred dorsal surface of head, the posterior is across the pending completion of a revision in progress by nape and bisected at either side by pectoral fin. the author of all known species in the eastern trop- Five wide, vertical, light brown bands on body, ical Pacific. most noticeable dorsally as dark saddles, becoming COMPARISONS.—Chriolepis lepidotus differs from more diffuse ventrally, and terminating near mid- other described congeners in the eastern tropical ventral line. Brownish yellow interspaces between Pacific in the extent of its squamation; its scales bands with a few scattered melanophores. Each extend forward to beneath the origin of the spi- band with one or two adjacent, prominent, black nous dorsal fin. Among the other described species, spots at midline. Anteriormost body band crosses C. minutillus Gilbert, 1891 (Gulf of California), dorsum between first and fifth dorsal spines, and most closely approximates this condition; its scales passes downward behind pectoral fin; second band extend forward to beneath or slightly in advance between sixth spine and end of spinous dorsal fin; of midspinous dorsal. From the description of C. third band between second and fifth dorsal rays; tagus Ginsburg, 1953 (Galapagos Islands), and in fourth band between seventh and ninth rays; fifth C. zebra Ginsburg, 1938 (Gulf of California), band behind end of second dorsal fin on anterior scales only extend forward to beneath the end of portion of caudal peduncle. Dorsal segments of the spinous dorsal, or slightly posterior to this posi- bands 3 and 4 appearing as darker blocks offset tion in the latter species. posteriorly from their ventral counterparts below The new species also differs from other described midline, with dorsal block of band 3 most notice- eastern tropical Pacific congeners in having the able and lying over the interspace between third highest number of pectoral fin rays (20 vs. 19 in and fourth ventral blocks. An irregular dark blotch C. tagus; 17 or 18 in C. zebra; 15 or 16 in C. behind fifth band at dorsal edge of caudal pe- minutillus.) The prominent black spots on the duncle, adjacent to first procurrent ray. A narrow head further serve to distinguish C. lepidotus from vertical bar of darker pigment underlying basicau- the other species described from these waters. dal row of specialized ctenoid scales. Chriolepis lepidotus further differs from C. min- 98 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY utillus in having a larger mouth (posterior angle Species of Chriolepis are sublittoral in habitat extends to below posterior margin of pupil vs. to and secretive, hiding under rocks and in crevices. below midpupil in C. minutillus), shorter pelvic Until recently, few specimens were known, but fins (when adpressed, not reaching anus vs. reach- with the introduction of SCUBA and rotenone ich- ing or almost reaching anus in C. minutillus), and thyocide collecting techniques by diving ichthyo- more gill rakers on the outer face of the first arch logists, specimens have begun to accumulate. These (11 vs. 8 or 9 in C minutillus). will allow for a comprehensive systematic, zoogeo- The new species further differs from C. zebra in graphic, and evolutionary treatment of the genus having a rounded rather than a depressed head, a and its allies (briefly treated by Bohlke and Ro- larger mouth (posterior angle extends only to be- bins, 1968:129-131) in the eastern tropical Pacific. low midpupil in C. zebra), and more second dorsal There remain about five undescribed species of fin rays (11 vs. 10 in C. zebra). Chriolepis in these waters. Chriolepis lepidotus differs from the description of C. tagus in having fewer second dorsal fin rays (12 in C. tagus), and fewer anal fin rays (10 vs. 11 Literature Cited in C. tagus). DISTRIBUTION.—Chriolepis lepidotus is presently Bohlke, J. E., and C. R. Robins 1968. Western Atlantic Sevcn-spined Gobies, with De- known only from Malpelo Island, Colombia. scriptions of Ten New Species and a New Genus, ETYMOLOGY.—From the Greek lepidotos (scaly), and Comments on Pacific Relatives. Proceedings of referring to its squamation, the most extensively the Academy of Natural Sciences of Philadelphia, known in the genus. 12O(3):45-174. REMARKS.—The genus Chriolepis was erected by Gilbert, C. H. Gilbert (1891:557-558) for the species C. minutil- 1891. Scientific Results of Explorations by the United lus, which he described as being totally scaleless on States Fish Commission Steamer Albatross, XXII: Descriptions of Thirty-four New Species of Fishes the basis of a single specimen trawled by the Alba- Collected in 1888 and 1889, Principally among the tross in the Gulf of California. His description was Santa Barbara Islands and in the Gulf of Cali- in error, since the holotype retains a single basi- fornia. Proceedings of the United States National caudal scale, as mentioned by Ginsburg (1938:111; Museum, 14 (880):539-566. 1953:21), and confirmed by examination of the Ginsburg, I. type by D. F. Hoese and R. J. Lavenberg (pers. 1938. Eight New Species of Gobioid Fishes from the comm.). Recently collected material of C. minutil- American Pacific Coast. Allan Hancock Pacific lus under study by the author shows that, indeed, Expeditions, 2(7):109-121. squamation is extensive in this species, less so than 1953. Ten New American Gobioid Fishes in the United States National Museum, Including Additions to a in C. lepidotus, but more so than in C. zebra, for Revision of Gobionellus. Journal of the Washington example. Academy of Sciences, 43(1): 18-26.

U.S. GOVERNMENT PRINTING OFFICE: 197 46-365/31

Publication in Smithsonian Contributions to Zoology

Manuscripts for serial publications are accepted by the Smithsonian Institution Press, sub- ject to substantive review, only through departments of the various Smithsonian museums. Non- Smithsonian authors should address inquiries to the appropriate department. If submission is invited, the following format requirements of the Press will govern the preparation of copy. Copy must be typewritten, double-spaced, on one side of standard white bond paper, with Wz" top and left margins, submitted in ribbon copy with a carbon or duplicate, and accompa- nied by the original artwork. Duplicate copies of all material, including illustrations, should be retained by the author. There may be several paragraphs to a page, but each page should begin with a new paragraph. Number consecutively all pages, including title page, abstract, text, litera- ture cited, legends, and tables. The minimum length is 30 pages, including typescript and illus- trations. The title should be complete and clear for easy indexing by abstracting services. Taxonomic titles will carry a final line indicating the higher categories to which the taxon is referable: "(Hymenoptera: Sphecidae)." Include an abstract as an introductory part of the text. Identify the author on the first page of text with an unnumbered footnote that includes his professional mailing address. A table of contents is optional. An index, if required, may be supplied by the author when he returns page proof. Two headings are used: (1) text heads (boldface in print) for major sections and chap- ters and (2) paragraph sideheads (caps and small caps in print) for subdivisions. Further headings may be worked out with the editor. In taxonomic keys, number only the first item of each couplet; if there is only one couplet, omit the number. For easy reference, number also the taxa and their corresponding headings throughout the text; do not incorporate page references in the key. In synonymy, use the short form (taxon, author, date:page) with a full reference at the end of the paper under "Literature Cited." Begin each taxon at the left margin with subse- quent lines indented about three spaces. Within an entry, use a period-dash (.—) to separate each reference. Enclose with square brackets any annotation in, or at the end of, the entry. For references within the text, use the author-date system: "(Jones, 1910)" and "Jones (1910)." If the reference is expanded, abbreviate the data: "Jones (1910:122, pi. 20: fig. 1)." Simple tabulations in the text (e.g., columns of data) may carry headings or not, but they should not contain rules. Formal tables must be submitted as pages separate from the text, and each table, no matter how large, should be pasted up as a single sheet of copy. Use the metric system instead of, or in addition to, the English system. Illustrations (line drawings, maps, photographs, shaded drawings) can be intermixed throughout the printed text. They will be termed Figures and should be numbered consecu- tively; however, if a group of figures is treated as a single figure, the components should be indicated by lowercase italic letters on the illustration, in the legend, and in text reference!: "Figure 96." If illustrations (usually tone photographs) are printed separately from the text as full pages on a different stock of paper, they will be termed Plates, and individual components should be lettered (Plate 9b) but may be numbered (Plate 9: figure 2). Never combine the numbering system of text illustrations with that of plate illustrations. Submit all legends on pages separate from the text and not attached to the artwork. An instruction booklet for the preparation of illustrations is available from the Press on request. In the bibliography (usually called "Literature Cited"), spell out book, journal, and article titles, using initial caps with all words except minor terms such as "and, of, the." For capitalization of titles in foreign languages, follow the national practice of each language. Underscore (for italics) book and journal titles. Use the colon-parentheses system for volume, number, and page citations: "10(2) :5-9." Spell out such words as "figures," "plates," "pages." For free copies of his own paper, a Smithsonian author should indicate his requirements on "Form 36" (submitted to the Press with the manuscript). A non-Smithsonian author will receive 50 free copies; order forms for quantities above this amount with instructions for pay- ment will be supplied when page proof is forwarded. K 1

•

"iHr ^ ^MBI