I. t#42 / .64 V . 0- 0 + r C - a* '/ ':.--»T T KT 1% „@,D.T.T .. %" r /= 4- of the FLORIDA STATE MUSEUM Biological Sciences Volume 19 1975 Number 4 THE ROCK IGUANA, CYCLURA PINGUIS, ON ANEGADA, BRITISH VIRGIN ISLANDS, WITH NOTESON CYCLURA RICORDI AND CYCL'URA CORNUTA ON HISPANIOLA W, MICHAEL CAREY 4 IFSMI '. 1% 'rt, tr *• -4' " ss-.~ ,s~ ,~ * S~L *, 6*, f. P f UNIVERSITY OF FLORIDA GAINESVILLE Numbers of the BULLETIN OF THE FLORIDA STATE MUSEUM, BIOLOGICAL SCIENCES, are published at irregular intervals. Volumes contain about 300 pages and are not necessarily completed in any one calendar year. CARTER R. GILBERT, Editor RHODA J. RYBAK, Managing Editor Consultants for this issue: WALTER AuFFENBERG HENRY FITCH WILLIAM MII.STEAD DONALD TINKLE Communications concerning purchase or exchange of the publications and all manu- scripts should be addressed to the Managing Editor of the Bulletin, Florida State Museum, Museum Road, University of Florida, Gainesville, Florida 32611. This public document was promulgated at an annual cost of $2249.10 or $2.249 per copy. It makes available to libraries, scholars, and all interested persons the results of researches in the natural sciences, emphasizing the Circum-Caribbean region. Publication date: April 25, 1975 Price: $2.25 THE ROCK IGUANA, CYCLURA PINGUIS, ON ANEGADA, BRITISH VIRGIN ISLANDS, WITH NOTES ON CYCLURA RICORDI AND CYCLURA CORNUTA ON HISPANIOLA W. MICHAEL CAREYI SYNopsm: A field study of the ethoecology of the rock iguana, Cyclura pinguis, on Anegada, British Virgin Islands was conducted from 27 March to 5 May 1968. The sympatric Hispaniolan species, C. ricordi and C, comuta, were observed from 12 to 16 May 1968. The results of these studies and all known ethoecological literature on the genus are discussed. Cyclum pinguis inhabits primarily xeric limestone areas, is heliothermic, and exhibits behavioral thermoregulation. Retreats nearly always consist of natural cavities in the limestone or beneath boulders. C. ricordi and C. comuta are primarily burrowers. Adult of all three species are principally herbivorous, although some animal matter is taken. Probably all C!/clum juveniles are primarily insectivorous. Parasitism from ticks and nematodes is a common occurrence in Cyclura. Five adult males on the main study plot ranged from 522 to 545 mm (f- 534.60 + 3.88 mm ) in snout-vent length and from 5,95 to 7.17 kg (x-6.72 + 0.21 kg) in body weight; five adult females ranged from 426 to 487 mm (x-468.00 + 10.87 mm) in snout- vent length and from 4.31 to 5,13 kg (x-4,758= 0.13 kg) in body weight, The five males occupied home ranges from 116.06 to 985,11 m2 (x- 546.61 * 160.46 mz ) in area; home ranges of the five females measured from 155.47 to 412.31 m2 (r- 276.28+ 41.62 m:). Extensive unutilized zones existed between home ranges of iguanas of, the same sex. Aside from food, available shelter seems to be the most importait factor governing selection of home range locale. Cyclura pinguis exhibited a density of 2.03 iguanas/hectare (0.82/acre), the lowest density figure 1The author is currently engaged in free-lance writing. The paper was submitted to the Uni- versity of South Florida, Tampa, in partial fulfillment for the degree of Master of Arta. Manu- script accepted 10 July 1972. Carey, W. Michael. 1975. The Rock Iguana, Cyclura pinguis, on Anegada, British Virgin Islands, with Notes on Cvclura ricordi and Cyclura cornuta on Hispaniola. Bull. Florida State Mus., Biol. Sci., Vol. 19, No. 4, pA 189-234. 189 190 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 4 of any lizard for which data are documented, The sex ratio was equal. Evidence suggests monogamy. Three adult females had 12, 14, and 16 yolked ovarian follicles. Adulthood appears to be attained between 7 and 9 years of age and between 350 and 400 mm snout-vent length. The reproductive cycle in C. pinguis suggests synchronization with the two rainfall maxima in the Virgin Islands. Adult iguanas far outnumbered juveniles (87.5% vs. 12.5% ) suggesting decline of the population. This disproportionate age structure seems to be the result of interactions ( competi- tion/predation) with domestic livestock. With the resultant loss in fecundity, relatively few adults now exist to continue the population. A conservation program is outlined whereby the population might be restored. This program might be applicable to other decimated Cvclura populations as well. TABLE OF CONTENTS INTRODUCTION 190 ACKNOWLEDGMENTS... 192 DESCRIPTION OF THE ISLAND AND STUDY SITES 193 METHODS AND MATERIALS 194 ENVIRONMENTAL CONDITIONS . . .. 195 GENERAL FEATURES OF THE LIZARDS 200 THERMAL RELATIONSHIPS . 203 RETREATS 207 FOOD AND FEEDING BEHAVIOR. 209 HOME RANGE 211 POPULATION DENSITY AND BIOMASS 216 POPULATION STRUCTURE AND REPRODUCTION 218 DISCUSSION 225 CONSERVATION . 228 LITERATURE CITED .. 229 INTRODUCTION Rock, or ground, iguanas ( genus Cyclura Harlan) are large, primarily herbivorous lizards endemic to the Bahamas, Greater Antilles, Cayman Islands, and Virgin Islands. The current distribution of extant species is spotty, but Pleistocene and Recent fossil remains have been found on several islands no longer supporting populations: e.g., St. Thomas ( Miller 1918), Puerto Rico ( Barbour 1919), Great Exuma ( Allen 1937), New Providence ( Etheridge 1966), Navassa ( Patton 1967), and possibly Bar- buda ( Etheridge 1964). These remains clarify several distributional pat- terns. Only Hispaniola supports sympatric forms. C!/clura ricordi and the rhinoceros iguana, C. cornuta, occur in portions of the Cul-de-Sac-Valle de Neiba plains of the Dominican Republic and in scattered areas farther south on the Barahona Peninsula. These are xeric salt plains that were once part of a Pleistocene waterway separating the then north and south islands ( sensu Williams 1961) of Hispaniola. This apparently comprises the total range of C. ricordi, whereas C. cornuta ranges over much of Hispaniola and several of its satellite islands. 1975 CAREY : CYCLURA PINGUIS 191 Cyclura belongs in the iguanine line of the family Iguanidae ( Savage 1958, Avery and Tanner 1971), but intrageneric relationships are poorly understood. Hence nomenclature in this paper essentially follows that of the last systematic treatment of the group ( Barbour and Noble 1916) and subsequent species descriptions. A study of the evolutionary relation- ships among the species is in progress ( Schwartz and Carey, in prepara- tion). No previous detailed studies on the ecology and behavior of any Cyclura species have been done, though several short papers describe various aspects of their activity. Gosse ( 1848) first reported habits of the Jamaican C. lophoma (.- C. collei). Nearly a century later Grant ( 194Oa) and Lewis ( 1944, 1946) noted the apparent impending extinc- tion of this species on Jamaica and Coat Island. Short works on C, cornuta ( Noble 1923, Klingel 1929) and reports dealing with the Ba- hamian forms C. rilegi (Stejneger 1903), C. figginsi (Bailey 1925), and C. carinata bartschi and C. nuchalis ( Cochran 1934) also have appeared. Activity of the Cuban C. macleayi macleal/i has been reported by Buide ( 1951), Street ( 1952), Sutcliffe ( 1952), Hardy ( 1956), and Cooper ( 1958). The forms on Cayman Brac and Little Cayman ( C. maclea!/i cal/manensis) and Grand Cayman ( C. macleayi lewisi) were studied by Grant ( 194Ob) and Carey ( 1966) . Grant ( 1944) also studied C. steine- geri on Isla Mona, off the western coast of Puerto Rico. In addition to the Cyclura species already extinct ( mattea, nigerrima, portoricensis, and a few unnamed forms), several others may face similar fates, because they inhabit tiny islands and cays-a situation that keeps many populations, although dense, precariously small. On the Greater Antilles, where ecological diversity is greater, Cyclura occurs only in restricted xeric habitats, and their decline or extinction has been caused primarily by man's activities. Because of the generally precarious status of Cyclura species, all are listed in the International Union for the Con- servation of Nature and Natural Resources ( IUCN) RED DATA BOOK on endangered amphibians and reptiles ( Honegger 1968). The present report and that by Carey ( 1966) are parts of a projected series of studies concerning the ecology and behavior of West Indian rock iguanas. The results will provide means for interpreting ethoecological components of the generalized C!/clura niche in these discrete but roughly similar insular communities. This report on the ethoecology of the rock iguana, Cyclura pinguis Barbour ( Fig. 1), on Anegada, British Virgin Islands includes remarks regarding the sympatric Hispaniolan forms, C. ricordi ( Dum6ril and Bibron) and C. cornuta ( Bonnaterre) and reviews all known literature dealing with ethoecology of Cyclura. 192 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 4 *Y>6, 23842'"~Im .--- FIGURE 1.-Adult male Cyclum pinguis from Anegada, British Virgin Islands. Note the prominent dorsaI ridge. Literature on C. pinguis is scarce. Schomburgk ( 1832) first reported " iguanas on Anegada when he mentioned having seen Iguana sapidissima on the island, but it was not until 1916 that the first specimen ( an adult female) was collected by James L. Peters. Barbour ( 1917) subsequently described the species as new, and in later papers he ( 1917, 1919, 1930, 1935, 1937) mentioned, or implied, its extreme rarity. Today his com- ments seem unwarranted, although the species is not common. Grant ( 1937) gave a brief account of stomach contents of adults, and Under- wood ( 1962) supplied brief notes on their habits. ACKNOWLEDGMENTS I wish especially to thank my faculty advisor, Roy W. MeDiarmid, for his con- tinued patience and interest in my research and for critically reviewing the man- uscript. I also thank Derek C. Burch, Wade C. Sherbrooke, Richard Thomas, and Walter G. Whitford for their helpful suggestions concerning various sections, Albert Schwartz aided in many ways, primarily through providing contacts in the Domini- can Republic. I am indebted to several island residents for their aid during my stay on Anegada, especially the Levans family who provided my room and board. Aubrey Levans and "Brasso" Norman provided much companionship and help in the field. During my stay in the Dominican Republic, similar provisions were made by the family of Sixto Inchaustegui of Santo Domingo. In addition, they made possible my trip to the vicinity of Lago de Enriquillo. James Feigl aided in preparation of the illustrations. The line drawing is by James Seagle. A major part of the fieldwork for this study was made possible through a Grant- in-Aid of Research from the Society of the Sigma Xi. To this organization I extend my sincere thanks. 1975 CAREY: CYCLURA PINGUIS 193 N 64° 20' 0 100 CS km 0 a %0 18°43' 0 0 4 kilometers 1 FIGURE 2.-Map of Anegada, British Virgin Islands showing ponds and study areas. Blackened areas are ( from left) Windberg Key, Citron Bush, and The Settle- ment. Inset shows location of Anegada (arrow) in relation to Puerto Rico ( PR) and other islands on Puerto Rican Island shelf. DESCRIPTION OF THE IsLAND AND STUDY SITES Anegada lies on the Puerto Rican Island Shelf, which consists primarily of Puerto Rico, Vieques, Culebra, and the Virgin Islands ( Heatwole and MacKenzie 1967). It is the northeasternmost island on the shelf (Fig. 2), lying about 24 km north of Virgin Gorda, and is about 15 km long and 3 km wide at its widest point. Although lying on the same bank as the other Virgin Islands, which are volcanic in composition and remarkably hilly, Anegada is composed almost entirely of lime- stone, is essentially without relief, and has a maximum elevation of about 9 m. Most of the western end is a sandy plain or salt flat with many highly saline ponds that often inundate the area during rainy periods. The eastern and central portions are largely eroded limestone plains, honey-combed with cavities leading beneath the surface. A total of 40 days ( 27 March-5 May 1968) was spent on Anegada, and pre- liminary investigations for -a later intensified study were carried out on the two sympatric forms, C. ricordi and C. cornuta. This represents about 215 hours of actual field time. Five days ( 12-16 May 1968) were spent in the Dominican Republic in the vicinity of Lago de Enriquillo, just east of Jimant near the Haitian border. This desolate portion of the Cul de Sac is some 40 m below sea level. The main study area on Anegada was located about 3.5 km northwest of The Settlement ( Fig. 2), a point where sandy and rocky areas meet. This plot will be referred to as the "Citron Bush," a local name applied to the general area of which the study site was a part. This area is roughly rectangular and encompasses about 4.93 ha ( ca 12.19 acres). Boundaries are natural and may somewhat restrict movements of iguanas to and from the area. To the south and west are Iarge, nearly denuded sandy areas, to the north mostly sandy beach and the Atlantic Ocean, and to the east an extremely dense thicket. Only adult iguanas inhabit the Citron Bush. The substrate consists almost entirely of weathered limestone projections, soil-filled pockets, and countless natural cavities leading to caverns beneath the 194 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 4 limestone. In the southwestern portion of the study plot is a small sandy area with numerous scattered limestone boulders. The ground cover is mostly dead vegeta- tion; e.g. leaves, cacti, logs, and brush. The second study area was located about 2.5 km west of the Citron Bush. This plot is known locally as Windberg Key and is a tiny, egg-shaped islet lying about 100 m off the western shore of Red Pond on the main island ( Fig. 2). This islet is 90 m long and 50 m wide at its widest point and about 0.28 ha ( ca 0.69 acres). Windberg Key was inhabited only by juvenile iguanas. The substrate is heavily eroded and cracked aeolian limestone with few natural cavities and little sand. Limestone boulders are scattered over much of the key. Except during the dry season when Red Pond dries up considerably, the isolation of Windberg Key is nearly complete. Additional observations were made at other localities whenever the opportunity arose; the principal ones being Cedar Well Key, Low Key, and much of eastern Anegada, The first two areas are actually peninsulas jutting into the north and south sides of Flamingo Pond, just west of Red Pond. Each key is about 150 m long and 100 m wide. The substrate on each is eroded limestone with many boulders. In the Dominican Republic, observations were made along roads at localities that appeared suitable. This was done on a day to day basis to obtain general ideas regarding the ethoecology of C. ricordi and C. cornuta. METHODS AND MATERIALS An analysis of plant composition in the Citron Bush was obtained by randomly selecting ten points in the area and counting the numerical abundance of each species within a ten meter radius of each point; frequency, relative abundance, and cover were determined. On Windberg Key this was done for every plant. An entire day was usually spent at the plots, observations being made between 0700 and 1800 hours ( roughly the diel activity period of C. pinguis). Investigations of lizard population structure, density, biomass, and home range usually employ mark and recapture techniques or at least some means of individual recognition. Past experience with adult Cyclura species, however, argues against mark and re- capture because the initial capture frightens an individual so much it may abandon the area. On Cayman Brac, on two separate occasions, adult male C. macleavi cal/- manensis were captured for inspection; upon release they fled and were not seen again. A similar experience at the beginning of my study occurred at the east end of Anegada. Fortunately, these lizards usually exhibit several distinctive features ( scars, missing digits, color pattern, etc.) making recognition simple, even from a distance of several meters. Most juveniles on Windberg Key were not distinctly featured; hence marking was essential. Various designs were painted at the base of their tails for recognition. Iguanas and their main retreats were assigned a number. Characterizing fea- tures, sex, age class, approximate snout-vent length ( SVL) and weight ( juveniles were measured and weighed), air and substrate temperatures, position in relation to the "center of activity" ( to be discussed below), time of day, and date were recorded. Adult C. pinguis were secured only with some difficulty. The easiest way was to chase them into their holes and ( if the holes were short) either noose them or annoy them until they surged from the hole and could be grabbed. Pulling iguanas from holes often required the assistance of a native. Cyctura. as do Conolophus in the GaMpagos Islands ( Carpenter 1969), often entangle their leg5 among roots on the sides of the holes. This, in addition to inflation of the body, makes them difficult to extricate. C. pinguis adults could never be approached close enough in the open to be caught by hand or noose. Juveniles were collected by any of the above methods or by hand as they hid beneath rocks. All measurements were recorded to the nearest millimeter, weight to the nearest 1975 CAREY: CYCLURA PINGUIS 195 0.10 of a pound and converted to kilograms. These data were secured for adults in the Citron Bush at the conclusion of the study. All temperatures were measured in the shade with a Schultheis quick-recording thermometer. Body temperatures of the lizards were taken cloacally and externally on the flank. Substrate temperatures were measured at the point occupied by the observed iguana; air temperatures about 200 mm above that point. Water temper- atures were taken at a depth of 150 mm. Humidity data were gathered with a standard meteorological sling-psychrometer. Readings were taken four times daily at a central point on the main study plot. Food preferences were determined by analysis of stomach contents and fecal droppings. Plants seen being consumed by C. pinguis were identified. Unless otherwise indicated, ranges are followed by means in parentheses. Pre- served material from Anegada and the Dominican Republic is deposited in the collection of Albert Schwartz. Sample sizes in this study were small. For example, only five animals were examined for stomach contents. As inadequate as this might appear, it is absolutely essential when investigating most Cyclura populations because of low population numbers. Many populations are dense but very small because of specialized habitat preferences. Sampling to the extent that most biologists would consider satisfactory wouId obliterate some populations ( e.g., C. macteayi Zewisi on Grand Cayman or any of the tiny-island populations in the Bahamas). ENVIRONMENTAL CONDITIONS CLIMATE.-In general, the climate of the Virgin Islands is tropical, with insolation reaching its greatest intensity in the spring ( Stone 1942). The mean annual rainfall varies greatly, ranging from 88.9 to 177.8 cm 0700 1100 1500 1900 1 1 1 1 1 1 1 1 - 100 35 - - 90 TE M PE R AT U R E °C 30 - 80 AllO IIN AH 3 A lln E IH - 70 25 - 1_ - 60 - 50 30 20- - - - 40 11 1 1 1 1 0700 1100 1500 1900 TIME OF DAY FIGURE 3.-Diel ranges and means of air temperature ( black bars) and relative humidity ( white bars) in the Citron Bush, Anegada, British Virgin Islands, 27 March-5 May 1968. 196 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 4 at the various stations. Two rainfall maxima occur annually, a small one in May or June and a larger one in October. During this study, shade temperatures were taken in the Citron Bush at least four times daily at 0700, 1100, 1500, and 1900 hours and were incidental to studies of activity cycles of C. pinguis. The maximum temperature recorded during the period of observation was 33.4°C on several occasions; the minimum was 21.1°C during three late-March mornings. Ranges, means, and humidity data are given in Fig. 3. In the Dominican Republic, temperatures in the shade ranged from lows of 29.4°C to 31.1°C at the beginning of Cyclura activity ( around 0700 hours) to highs of 37.4°C to 41.0°C between 1300 and 1600 hours. Rainfall was not measured, but showers were frequent during the first week of the study. After 3 April it rained only five more times. The entire West Indian region had been suffering under drought con- ditions for some time prior to the study. FIGURE 4.-Typical xerophytic thorn forest habitat of Cyclura in the West Indies. Photo taken 10 February 1965 on Cayman Brac, British West Indies. VEGETATION,-The vegetation on Anegada is typical Antillean xero- phytic thorn forest or scrub formation ( Fig. 4). The beach around most of the island supports either shore scrub and pioneer vegetation or mangrove, Rhizophora mangle. Vegetation on the sandy plain to the west is rather dense and coppice-like in some areas. The loblolly tree 1975 CAREY. CYCLURA PINGUIS 197 (Pisonia rotundata), boxwood ( Bumelia obouata), and seagrape ( Cocco- loba uvifera) are common. Because of its porosity, the eastern rocky plain is more arid. Principal plants are seagrape, agave ( Agaue sp.), cacti ( Melocactus intortus and Pilosocereus ro!/enii) and poilmerboy C Pithecolobium bahamense). Epiphytes such as Tillandsia utriculata are common on trees all over the island. The greatest concentrations of C. pinguis were in the xeric limestone regions. Iguanas were relatively rare on the sandy plain, except on the numerous limestone outcroppings forming peninsulas or islets in the salt ponds. Rocky areas were possibly inhabited in preference to sandy ones because of the abundance and diversity of refuges. The Citron Bush is an ecotone between the sandy and rocky plains. A composition analysis of its vegetation and that of Windberg Key ap- pears in Table 1. The vegetations of Cedar Well Key and Low Key are similar to that of Windberg Key. VERTEBRATE ASSOCIATES.-Few other vertebrates inhabit Anegada. Apparently there are no naturally occurring predators or competitors to adult C. pinguis. However, introduced domestic animals have become TABLE 1.-PERCENT RELATIVE ABUNDANCE, FREQUENCY, AND COVER OF VEGETATION ON THE Two STUDY PLOTS ON ANEGADA, BRmSH VIRGIN ISLANDS, 27 MARCH-5 MAY 1968. Citron Bush Windberg Key Percent Relative Percent Percent Relative Species Abundance Frequency Abundance Conocarpus erectus 2.7 30 62.4 Cassine x!/locarpa 1.4 50 Colubrina arborescens 9.1 100 Lantana feticulata 11 .6 70 Tabebuia heterophvIla 3.5 60 Plueria alba 6.3 80 Crossopetalum rhacoma 6.0 80 Stfumphia maritima 2.9 50 Coccoloba uvifera 1.8 50 Pithecolobium bahamense 12.8 100 24.8 Pisonia rotundata 3.4 30 Vallesia antillana 0.1 10 Croton discolor 8.8 50 Randia aculeata 2.4 40 Am!/ris diatrupa 5.2 60 Ziziphus rignoni 1.4 70 Eugenia procera 0.1 10 Regnosia uncinata 0.3 10 Zanthoxylum 8auum 0.7 40 0.8 Agaue sp. 4.9 70 Bursera simarouba 0.2 20 0.8 Bumelia obuata 0.2 20 11.2 Melocactus intortus 5.4 90 Opuntia sp. 1.8 30 Pilosocereus royenii 12.8 100 Percent cover 46 61 198 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 4 TABLE 2,-VERTEBRATE ASSOCIATES OF Cyclura pinguis IN THE CITRON BUSH, ANEGADA, BRIMSH VIRCIN ISLANDS, AND THEIR KNOWN OR POSSIBLE INTERRELATIONSHIPS. Possible or Known Possible or Known Relationships Competitors Predators Unknown Mammals Bos taurus Homo sapiens Artibeus iamaicensis Sus scrofa Sus scrofa Mus musculus Equus asinus Felis catus Capm hircus Canis familiarls Ouis aries Rattus noruegicus Birds Falco apart)erius Falco peregrinus Zenaida asiatica Mimus polyglottoa Falco sparcerius Zenaida macfoura Tgrannus dominicensis Buteo iamaicensis Columbigallina passerina Crotapht/ga ani Anthrocothorar dominicus Fregata magniticens Coereba flaveola Mimus polyglottos Dendroica petechia Snakes Alsophis portoricensis Alsophis portoricensis Typhlops richardi Lizards Anolis cristatellus Ameiua erm Z Sphaerodactylus macrolepis Anolis pulchellus Ameiva exsul Mabuga mabouga Frogs Let*odact!/lus albilabris increasingly important in these respects ( Table 2). Under pristine con- ditions C. pinguis, at present omnivorous ( albeit primarily herbivorous), probably occupied several trophic levels concurrently. Juveniles and eggs were probably subjected to some natural predation just as they are today. The colubrid snake, Alsophis portoricenas anegadae, was com- monly reported by natives to feed on juveniles, as was A. cantherigerus on the Caymans ( Carey 1966). Alsophis portoricensis also has been reported to feed on other lizards similar in size to juvenile C!/clum ( Schmidt 1920, Grant 1932 ). Predatory birds such as Falco sparverlus and Buteo iamaicensis are almost certainly a constant threat to juveniles. Hawks, owls, gulls, and herons feed on young marine iguanas, Am- blyrhynchus cristatus, in the Galdpagos (Carpenter 1966). Hatchling Cyclura on tiny islands in the Bahamas are almost certainly vulnerable to such predators, especially gulls and herons. The teiid lizard, Ameiva exsul C Wolcott 1924), and several resident passerines ( e.g., Mimus) are occasional lizard-egg feeders and possibly prey on eggs of C. pinguis. Mockingbirds ( Mimus) feed on Ambly- rhynchus eggs in the Galdpagos ( Carpenter 1966). Important natural competitors of C. pinguis were presumably few and essentially the same as those today ( Table 2). Seemingly, only herbivorous birds and insects could have competed with adult C. pinguis 1975 CAREY: CYCLURA PINGUIS 199 for food, and these were probably of minor influence. Juvenile iguanas were probably subjected to considerably more competition since they were msectivorous as well as herbivorous. They would have been sen- sitive not only to the feeding activities of adults and their competitors, but also to the numerous resident insectivores ( Table 2), including birds, several lizards, and possibly a frog. About 300 years ago the ancestors of the present native human colony ( which now numbers about 200, all living in The Settlement) arrived on Anegada and brought dogs, cats, goats, cattle, donkeys, pigs, and sheep with them. Although most of these animals abound on the island today, it is impossible to detect what efFect they may have had on the relation- ships of C. pinguis with the natural fauna. Iguanas continue to occupy the same consumer levels, but apparently now lead a more competitive existence. Natives say that dogs feed on both the eggs and juveniles of C. pin- guis and occasionally will run down and kill adults. I witnessed this on one occasion on Anegada. Dogs and domestic pigs are known to prey on eggs and juveniles of other large iguanines ( Hirth 1963a, Carpenter 1965, 1966, Rand 1968). Cats probably affect only juveniles. Herds of goats and cattle overlap considerably with the greatest concentrations of iguanas, feeding on brushy undergrowth and possibly increasing exposure of juvenile iguanas to predators. In addition, elim- ination of this brush may increase the difficulty of food procurement for the young iguanas directly and through resulting decreases in insect density. Windberg Key apparently supports very little vertebrate life other than C. pinguis. Only Anolis cristatellus wilegae, Sphaerodactylus ma- crolepis macrolepis, Alsophis portoricensis anegadae and the Antillean mango, Anthracothorax dominicus were in evidence. No mammals were seen. As no domestic animals inhabit Windberg Key, it seems probable that resident iguanas lead a more natural life than their main-island counterparts. If the domestic animals invaded Windberg Key, C. pinguis would very likely be eliminated. On 13 April, I observed the morning activity of a female in the Citron Bush. She faced southwest about 20 m from her retreat. A bananaquit, Coereba #aveola, Hitting about had no apparent influence on the iguana until it emitted a shrill whistle. The iguana immediately turned 180°, fled to its retreat, and entered after hesitating for a moment at the en- trance. A blind snake, Typhlops richardi catapontus, was collected about 80 mm below the surface just inside a cavity occupied by a male iguana. In the Dominican Republic I frequently egcountered curly-tailed lizards, 200 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 4 Leiocephalus schreibersi, about 2 m inside iguana burrows, generally during the heat of.the day or at night. PARASITES.-No dissections were made for endoparasites in C. pinguis, but other forms of Cyclura C carinata [Dosse 1938], macleagi cayman- ensis and macleayi lewisi [Grant 194Ob], and macleayi macleayi [Barus et al. 1969] ), host several nematode species. Cyclura ricordi and C. cornuta, autopsied upon return from the Dominican Republic, also hosted several nematodes. Those in C. cornuta were identified as Atractis cru- ciata (John Lewis, pers. comm.). I have observed ticks of various species on all of the above iguanas, as well as on C. rileyi, C. #gginsi, and C. steinegeri. Several are dis- cussed by Robinson ( 1926 ). A series of 59 ticks taken from C. pinguis represented a new species, Amblyomma antillorum C K6hls 1969). About 20-30 ticks per lizard were found only on adult C. pinguis, most fre- quently in the area of the femoral pores and in the thoracic region. No ticks were found on 17 C. ricordi examined. C. cornuta (N= 10) was infested by Amblyomma albopictum in the same areas as de- scribed for C. pinguis. GENERAL FEATURES OF THE L~ARDS Cyclura pinguis is one of the largest members of the genus. There was a highly significant difference ( t= 5.772; P<.01) in SVL between adult males and females in the Citron Bush ( Table 4), and for all adults for which data were obtained ( t= 4.359; P<.017 [data extracted from Fig. 5]). Adult males in the Citron Bush averaged 534.50 * 3.88 mm SVL; adult females 468.00 + 10.87 mm. There also was a highly significant difference (t= 8.000; P< .01 ) in body weight between adult males and females in the Citron Bush ( Table 4), and for all adults for which data was obtained (t= 6.621; P< .01 [data extracted from Fig. 5] ). Adult males in the Citron Bush aver- aged 6.72 +0.21 kg in body weight, adult females 4.75 * 0.13 kg. Adults of both sexes are dull, olive-gray above with columns of tur- quoise ( sometimes green) dots ( ca 2 mm in diameter) extending down- ward from the dorsal crest to the lateral fold. Beginning at or below the fold are reticulations of turquoise and brown, fading on the belly to blue- cream. The head is olive-brown, the throat and gular pouch lighter. The tail is blue dorsally, fading to dull brown posteriorly. Limbs are blue above, each scale bordered in black, and tan to cream below. Spines of the dorsal crest vary considerably from shades of blue to black. Aside from being larger ( Fig. 5), males have proportionally much larger heads, with the temporal region and jowls obviously being more "swollen." Both sexes possess a prominent, fatty dorsal ridge ( Figs. 1, 1975 CAREY: CYCLURA PINGUIS 201 600 - 550 - - 0 6,0. 500- ' 08 0 450 - - 0 400 - 0 SN O U T- VE N T LE NG TH ( m m ) 350 - 0 300 - d: o 250 - CO 200 - 150 - a 12345678 BODY WEIGHT (kg) FIGURE 5.-Length-weight relationship for Cyclura pinguis (N-22), Anegada, British Virgin Islands. Black dots, males; white dots, females; triangle, sex unknown. 7), although in males the ridge is somewhat higher than in females. The above characters permit immediate recognition of sexes. In addi- tion, the longest spines of the dorsal crest are usually about 18 mm in males, whereas they are about 10 mm in females. Also, the tail base of males is more laterally compressed than in females. Cyclura ricordi is somewhat smaller than C. pinguis; the adult males range from 315 to 400 mm (Y= 355 mm, N = 4) SVL, and adult females range from 308 to 365 mm ( x- 340 mm, N-3). Cyclura ricordi is powder blue to tan above, darker to black below. 202 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 4 There are four to five white bands on the Hanks that are bordered in black and end on the belly. The arms are black above, lighter below; the legs olive-blue. The spines of the dorsal crest are blue or black ana about 15 mm high. Juveniles are generally somewhat darker than adults. Cyclura cornuta is the same general size as C. pinguis. Adults range from uniform gray-brown to nearly black dorsally and lighter ventrally. The juvenile pattern ( tawny bands interspersed with cream dots) is obscured in adults. The limbs are usually olive, the head rusty in juveniles and dark brown in adults. Rhinoceros iguanas are unique in that the sclerotic coats of the eyes are light metallic brown rather than some shade of red as in all other known extant Cyclura species. Individuals of both this and the preced- ing species are difficult to sex externally. The local name of all three of these lizards is "guana," this term apparently is uniform for all Cyclura populations throughout the West 40- 0 38- LI ZA R D T EM PE R AT U R E ° 36 - 34 - 32 - .. 30 - 28 26 ill 'll 'll// 26 28 30 32 34 36 38 40 AIR TEMPERATURE °C FIGURE 6.-Relation of ambient and body temperature at time of capture for active Cuctura pinguis (N==16), Anegada, British Virgin Islands. 1975 CAREY: CYCLURA PINGUIS 203 Indies, regardless of the country. On Anegada juveniles are called -four o'clocks." THERMAL RELATIONSHIPS FIELD WORK.-Because of the relatively few iguanas on Anegada and my reluctance to upset their daily routines, temperature data for only 16 animals were obtained. These data were procured on sunny days, usually after 1100 hours, so the lizards had ample opportunity to reach desired activity temperatures. The distinct differential between ambient and cloacal temperatures ( Fig. 6) clearly indicates that C. pinguis is a heliothermic lizard. Ih the strict sense, heliothermy merely implies that heat gain in an ectothermic animal is obtained principally through solar energy ( Cowles 1940). It does not necessarily suggest that some form of behavioral thermoregulation occurs, although that possibility is not ruled out. Heath ( 1964) pointed out that even though many reptiles are fully capable of regulating their body temperatures behaviorally, they can also attain and retain equable temperatures without need of active regulation. Perhaps the most striking of the several examples of behavioral thermoregulation described for lizards is that of the marine iguana on the Galhpagos. In this species prostrate" or "elevated" basking postures are assumed in accordance with time of day, intensity of solar radiation, and body temperature ( Bartholomew 1966). Behavioral thermoregulation can be seen in the daily activity pattern of C. pinguis, which agrees closely with that seen in C. macleayi ca!/- manensis ( Grant 194Ob, Carey 1966). Cyclura pinguis usually appears each day shortly after 0700 hours, when ambient temperatures average around 25°C. Upon emergence they walk to the nearest patch of sun- light and assume a prostrate basking posture similar to that described for Amblyrhynchus. Prostrate basking in C. pinguis consists of lying Rat on the belly with the limbs extended in "spread eagle" fashion ( Fig. 7). -Y FIGURE 7.-Adult male Cyclura pinguis in typical prostrate basking posture. Note extension of limbs. 204 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 4 In Amblyrhynchus the head and neck are laid flat; in C, pinguis the head is elevated. About 1 hour later, apparently after reaching a desired body tempera- ture, they rise on their arms to a semielevated position. At this time ( ambient temperature ca 27°C) they begin to move about in search of food, stopping every few steps to taste bits of material around them. Basking continues after feeding, rarely in full sun but in a sun-and-shade mosaic among bushes, similar to that utilized by the green iguana, Iguana iguana, at San Bias, Nayarit, M6xico ( MeGinnis and Brown 1966). Throughout most midafternoons, when ambient temperatures hovered around 30 ° C, Cyclura pinguis was seen in partial or full shade . In later afternoons, from 1500 to 1800 when temperatures generally dropped be- low 30°C, C. pinguis again came out into exposed places, usually near a retreat. No iguanas were seen after this time, when temperatures usually fell below 26°C. Coolest ambient temperatures at night ( usually after 0100) averaged 22°C. Iguanas observed after sunset were always in their holes and presumed sleeping. Elevated basking was not observed in C. pinguis, and probably would be of little use to a form able to utilize the shade and filtered insolation of brushy areas. On the other hand, Amblyrhynchus spends the entire day on the barren lava coasts of the Caldpagos Islands, with no shady retreats available. Elevated postures allow them to maintain body tem- peratures below the lethal threshold while in direct sunlight ( Bartholo- mew 1966). InsufRcient data for reliable determination of eccritic temperatures were obtained. Cloacal temperatures secured from 16 active animals ranged from 27.1°C to 39.8°C (35.29°C); 56% of the records were greater than 36°C, and 25% ranged from 39°C to 39.8°C. This contrasts with mean activity temperatures of 31.98°C for Conolophus subcristatus, 35.13°C for C. pallidus ( Carpenter 1969), and 36.1°C for green iguanas, Iguana iguana ( McCinnis and Brown 1966). Means slightly greater than 34°C were observed by Carpenter ( 1966) for marine iguanas on several islands. Bartholomew ( 1966) indicated a preferred temperature range of 35° to 37°C for these lizards. The first several days of the study (late March) were characterized by cool, blustery weather that had no profound effect on C. pinguis. Midmorning temperatures at that time averaged in the low 20's, and it was windy and overcast. Though air temperature may have an indirect effect on daily activity of these iguanas, it may not be the main influ- encing factor. Milstead ( 1957a, b) found that light intensity and/ or soil temperature governed Cnemidophorus activity in southwestern Texas, and this may also be true of C. pinguis. 1975 CAREY: CYCLURA PINGUIS 205 55 - limestone50 - 0 45 - TE M PE RA TU RE ° - skin - \40- 1 ,Qr' cloaca 35- ,- IP 1 ''' - IP air30 - T ,0 I . YZ 25 ll,,,IIII 0 10 20 30 40 50 MINUTES FIGURE 8.-Relation of skin and cloacal temperatures to environmental temperatures in a juvenile Cyclura pinguis ( 772 g), Anegada, British Virgin Islands. Lizard was placed in full sunlight at 1400 hr, 2 April 1968. Thin cloud cover lasting 6 min began at T17; maximum wind velocity about 12 km/hr. 206 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 4 Although torrential rains drove Cyclura pinguis into their burrows, light steady rainfalls were apparently not annoying. Light rain may have a desirable cooling effect, as adults were seen in the open during several steady downpours. I have kept many Cyclum species in outdoor enclosures in southern New Mexico, where ambient temperatures usually exceeded 35°C by noon each summer day. Adults usually basked most of the morning. Often as I sprayed the iguanas with a garden hose, they followed the spray around the pen for about 10 minutes and then were wary of it. On several occasions they raised up on their toes, apparently 45 - 1.36 kg ~ ..3r 40 0.82 kg ~ 4.68 kg 0 0 .0 a TE M PE R AT U R E ' -00 air 25 / p- 20 , ,, I, I , ," ", i 0 10 20 30 40 50 60 70 MINUTES FIGURE 9.-Relation of heating rate to body size in Cyclura pinguis (N-3) on Anegada, British Virgin Islands. Lizards were placed together in full sunlight at 1000 hr, 20 April 1968; maximum wind velocity ab6ut 8 km/hr. 1975 CAREY: CYCLURA PINGUIS 207 to avoid contact with the hot, sandy surface, which averaged 60°C dur- ing the hottest part of the day. EXPERIMENTAL WORK.-Bartholomew ( 1966 ) indicated a need for additional experimental work on the thermal relations of large lizards. Although a comprehensive coverage of the thermal relations of Cyclura is beyond the scope of this paper, limited data on the heating rates of various sized C. pinguis in relation to environmental temperatures are useful ( Figs. 8,9). In both experiments, animals were tied at the waist and placed in full, midmorning sunlight. The method gave them free movement and allowed them to thermoregulate, Dorsal skin, air, and substrate temperatures were measured at 5-minute intervals until panting began, usually around 40°C. Panting apparently enables the animals to remain longer within their range of thermal tolerance. The rapid flow of air over the blood engorged tongue permits dissipation of slightly more heat than is produced metabolically ( Kour and Hutchison 1970). Data for marine iguanas closely agree with this figure ( Bartholomew 1966),but green iguanas begin panting at a slightly higher temperature (ca 42°C) ( MeGinnis and Brown 1966). Frantic efforts to escape usually accompanied panting in C. pinguis, and they were then quickly placed in the shade. A juvenile weighing 772 g was used to determine the relation of skin and cloacal temperatures to air and substrate temperatures. Primary sources of heat gain ( insolation and conduction from substrate) caused panting in forty-five minutes, at a temperature of 39.8°C. Although skin and body temperatures remained close throughout ( skin was al- ways at least 1 ° higher), they had nearly equalized when panting began ( Fig. 8). These data are in agreement with that obtained by Kour and Hutchison ( 1970) for other lizards. An indication of body size-heating rate relationships was obtained by tethering three animals of different weights ( 0.82, 1.36, and 4.68 kg) simultaneously. As in marine iguanas ( Bartholomew 1966), rates of heating were related inversely to body weights ( Fig. 9). RETREATS Little has been recorded concerning retreats of Cyclura Gosse ( 1848) did not mention holes as retreats of C. lophoma ( - C. collei) on Jamaica, but stated ( upon another's observations) that they lived in trees. Lewis ( 1944) observed C. lophonm in trees ( at least 3 m above ground) and C. macleayi caymanensis commonly basking on Conocarpus branches. This and the Grand Cayman form, C. macleayi Zewisi, usually favor natural cavities as retreats ( Grant 194Ob, Carey 1966). Climbing has also been documented for C. rileyi on San Salvador ( = Watlings 208 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 4 Island) in the Bahamas ( Stejneger 1903, Paulson 1966) and C, macleal/i macleal/i on Cuba ( Cooper 1958), but they too retreat into holes among the rocks. Cyclura cornuta on Ile de la Petite Gonave lives in the crevi- ces in coral rock ( Klingel 1929). I saw this lizard in the hills bordering Lago de Enriquillo in the Dominican Republic, but found that there they dig burrows in the sand as well as in the fossil coral cliffs as earlier reported by Noble ( 1923). Cyclura ricordi also burrowed in the hills bordering the lake. On Isla Cabritos, they lived only among crevices in rocks. Several Bahaman forms ( i.e. C. nuchalis and C. carinata bartschi [Cochran 1934], C. figginsi [Bailey 1925, Wayne King, field notes, Tony Gran65, pers. comm.], C. inornata [Rabb and Hayden 1957, Wayne King, field notes], and C. baeolopha [Wayne King, field notes]) prefer natural cavities. Sutcliffe ( 1952) and Hardy ( 1956) observed both burrowing and retreating to limestone cavities in C. macleayi macleayi in Cuba and adjacent islands. Burrows extended to nearly 3 m in length and had no enlarged chambers. Burrows of C. ricordi and C. cornuta were much longer ( ca 10 m) and also lacked chambers. In the Citron Bush, C. pinguis used burrows or natural cavities in the limestone, excavations beneath large limestone slabs, or cavities formed under large overlapping slabs that were raised slightly by the root system of a loblolly tree ( Pisonia rotundata) as refuges. Only one adult male was observed outside of a limestone area. His retreat was a bur- row, presumably self-excavated. Immatures on Windberg Key used only slabs as retreats. A single slab often harbored as many as five iguanas beneath it. A very young individual in The Settlement lived in a pile of debris in an abandoned hut, and I never observed it more than a few meters from the debris. Refuge entrance sizes ranged from holes just large enough for an adult to squeeze through to some much larger than an iguana. Refuges ran parallel to the surface, usually not more than 1 m deep, and varied in length from a minimum of about 2 m to an indeterminate maximum far into the limestone. Refuges contained moist soil and forest litter in the immediate interiors, with swarms of mosquitos at the entrances and in the moist interiors. Excavations in limestone slabs probably offer considerably less se- curity because of the short distance between entrance and terminal chamber ( often less than 2 m), but still provided sufficient concealment. A typical slab was about 4 mz and about .25 m thick. Cavities formed by roots of loblolly trees seem to offer much more security. The roots grew beneath and onto the slabs, forming extremely sturdy refuges. The cavities were not deep, but wound around wherever roots were growing, to form a maze. 1975 CAREY : CYCLURA PINGUIS 209 Cyclura macleayi macleayi C Buide 1951) and C. macleayi cayman- ensis ( Carey 1966 ) flee from intruders in short spurts, repeatedly stop- ping to bobb and hiss, while Cyclura pinguis rushes directly to the hole, rarely hesitating at the entrance. Both C. ricordi and C. cornuta re- treat in spurts before Bnally entering their burrows, as does Conolophus in the Galapagos Islands ( Carpenter 1969). Once C. pinguis individuals enter their holes, they may not come out for a surprisingly long time. When I approached the Citron Bush at 0900 on 14 April, an adult female entered her refuge and stayed there. At 1345 I found her still lying in the terminal chamber. On 19 April an adult male reemerged 2 hours and 47 minutes after having taken refuge. It is well known that marine iguanas enter the sea for feeding pur- poses, but they will not willingly enter the sea for escape unless their body temperatures are approximately that of the water ( Bartholomew 1966). An iguana just leaving the sea will quickly drop back if ap- proached, but a basking individual cannot be induced to retreat to the water. Only C. cornuta ( Noble 1923) and C. macleayi macleayi ( Hardy 1956, Lando and Williams 1969) are known to retreat to water. Gosse ( 1848) stated that his captive C. lophoma ( = C. coll€i) readily took to water, but not necessarily as a retreat. An adult male C. pinguis on Cistern Well Key often entered the water when disturbed. Temperature seemed to have little influence on his behavior. On 7 April, after re- cording his cloacal temperature, I placed him on the ground, and he immediately took flight into Red Pond. At that time his cloacal tem- perature was 40.2°C and the water was 27.5°C. He tried to dive, but the water was too shallow. On the morning of 12 April, his tempera- ture was 37.0°C and the water was 27.9°C. Upon release his behavior was identical to that described above. On four subsequent occasions he entered the water upon my approach. I assume, therefore, that CY- clura, unlike Amblyrhynchus, can and will use water as a retreat regard- less of the difference between body and water temperature. FOOD AND FEEDING BEHAVIOR In nearly all C!/clura studies vegetation was recorded as the primary food in the wild. Cyclura lophoma ( - C. collei) was observed feeding on the "guinea-hen weed" ( Petiveria sp., Gosse 1848). On Cayo de la Piedra C. madeal/i macleayi reportedly feeds principally on Opuntia buds, but eats crabs also ( Sutcliffe 1952). On Little Cayman, C. ma- cleagi caymanensis has been observed feeding on bay vine ( Ipomoea pes-caprae), broad leaf ( Cordia cal/manensis), and guana berry ( Erno- 210 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 4 dea littoralis) ( Grant 1940b). On Cayman Brac they often raid culti- vated areas, feeding on potato vine, plums, mangoes, and fallen pawpaws ( Grant 194Ob, Carey 1966). Cyclura steinegeri on Isla Mona was ob- served feeding on Tribulus cistoides ( Martin 1966), The literature is extensive regarding the herbivorous habits of other iguanine genera. Examination of stomach contents of five animals revealed that vegeta- tion composed a major part of the diet of adult C. pinguis ( Table 3). Grasses and leaves were the most common items. Analysis of seat com- position supplemented this analysis, although most of the plant material was unidentifiable. One adult female ( 400 mm SVL) contained four lepidopteran larvae, each about 75 mm long and 12 mm in diameter. Grant ( 1937) found seagrape leaves, cactus fruit, and "wild nutmegs" in the stomachs of two adults. Underwood ( 1962) suggested that C. pinguis fed on cactus shoots, fruits, and seagrapes. TABLE 3.-STOMACH CONTENTS OF ADULT Cyclura pinguis (N-5), ANEGADA, BRITISH VIRGIN IsLANDS, INDICATING FREQUENCY AND VOLUME PERCENTS. Percent Percent Food Frequency Volume Plants Conocarpus erectus 60 19 Coccoloba uvifera 20 12 Lantana reticulata 20 4 Regnosia uncinata 40 20 Erithalis fruticosa 40 6 Grasses 20 10 Unidentified 100 14 Insects Lepidoptera larvae 20 15 Stomachs of juvenile C. pinguis were not examined. It is quite pos- sible that insects form the major food item. Juveniles of several Cyclura species ( cornuta, figginsi, inornata, macleagi, pinguis, ricordi) kept in the laboratory at various times have shown a marked preference for insects ( esp. crickets, grasshoppers, and mealworms) over plant material. In- sects are a major food source for Ctenosaura similis juveniles in Guana- caste Province, Costa Rica ( Roy W. MeDiarmid and Dennis R. Paulson, unpubl. data). In addition to most fruits and vegetables, captive C. pinguis adults feed on canned dog food and laboratory mice and rats. Cyclura pinguis is an active forager. Following basking in the morn- ing they move quickly about an area, hesitating every few steps to taste particular objects ( pebbles, plants, etc.) in their path. They feed in typical iguanine fashion, snapping off leaves and flowers with a twist of the head or by the jaws alone, often using their forefeet to pull a food item closer. They occasionally climb into low bushes to feed. I saw them actively feeding on the leaves of Conocarpus erectus, Coccoloba uvifera, and Pisonia rotundata. 1975 CAREY: CYCLURA PINGUIS 211 Stomachs of C. ricordi and C. cornuta contained exclusively fruits of unidentified cacti. None were observed feeding in the field. In cap- tivity, feeding is identical to that of C. pinguis. "Sneezing" fluid from the nostrils in the form of a fine salt spray was common in each of these rock iguana species. As in other iguanines ( Templeton 1964, 1967), this probably is an extrarenal mechanism for ridding the body of excess salt. HOME RANGE SELECTION OF HOME RANGEs.-Natural cavities (Fig. 10) and loblolly trees occurred in nearly all home ranges of adult iguanas, and because both form refuges it appears that shelter is an important factor in home range selection. In addition, iguanas were occasionally observed resting on the lower ( 2-3 m high ) branches of these trees, as well as in the shade beneath the trees. Though loblolly leaves were never found in stomachs, an adult male was once seen nibbling them. Selection of mates may also influence the choice of home range sites. An area must be chosen that has sufficient available food as well as A A A NE AA -- //51