BULLETIN of the FLORIDA STATE MUSEUM Biological Sciences Volume 19 1975 Number 5 ECOLOGICAL ANALYSIS OF THE CAYMAN ISLAND AVIFAUNA DAVID W. JOHNSTON e 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: JON C, BARLOW JOHN WILLIAM HARDY 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 $2,093.41 or $2.093 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: May 5, 1975 Price $215 ECOLOGICAL ANALYSIS OF THE CAYMAN ISLAND AVIFAUNA DAVID W. J0HNSTON1 SYNOPSIS: Ecological studies of the avifauna carried out over the past eight years in the three Cayman Islands of the Caribbean Sea revealed the presence of 151 species, mostly transients. Grand Cayman has 39 species of breeding birds ( 26 terrestrial), Little Cayman 29 ( 17 terrestrial), and Cayman Brac 30 ( 20 terrestrial). Each island also supports an additional 40 North American migrants in winter. Seasonal occurrences and relative abundances of birds are described, particularly for seven upland ecological formations in which the greatest breeding species diversities occur in logwood-thatch palm-red birch and limestone forests. In one sere, from pastureland through limestone forest, bird species diversity and total abundance in- crease with age of the community. High species diversity and density in the lime- stone forest are associated with some semblance of stratification but more so with a greater number of feeding niches. For the most part, the land birds breed in May and June at the onset of a rainy season. In the past 30-40 years, tw6 bird species have become extinct on Grand Cayman, whereas two others have at least attempted colonization. The problems of inter- island distributional patterns are discussed in an attempt to explain the presence and absence of different species on different islands. No evidence was found to support the idea that the absence of one species on an island permitted another one on that same island to broaden its ecological niche; rather, the absence of woodpeckers, a tanager, flycatcher, dove, finch, and others on the two small islands strongly sug- gests vacant niches on those islands. Competition for environmental resources among the resident terrestrial birds was analyzed, especially between congeners and between species of similar ecologies. The four ( or five) sympatric dove and pigeon species are separated chiefly by habitat and food differences; the two woodpeckers by food; four flycatchers by com- binations of bill size, habitat, feeding height, and food type; the vireos by body size and bill size; and two Dendroica warblers by habitat and feeding height. With the exception of the scarce Barn Owl, no significant vertebrate predators on birds occur on these islands. The largely ornithophagous owl probably repre- sents the greatest single biological control of the avifauna. 1 The author is a Professor of Zoology, University of Florida, Gainesville, Florida 32611. Manuscript accepted 26 April 1974. JOHNSTON, DAVID W. 1975. Ecological Analysis of the Cayman Island Avifauna. Bull. Florida State Mus., Biol. Sci., Vol. 19, No. 5, pp. 235-300. 235 236 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 5 Feeding ecologies, habitat distribution, and taxon cycles are specified, where known, for all the resident terrestrial species. Taxon cycles appear to be similar to those of other insular avifaunas, despite the absence of montane refugia on the Cay- mans. Species in Stage I ( evidently the recent colonizers ) are more common in early seral stages ( "marginal habitats"), whereas the endemic subspecies ( Stage IV) are more abundant in the mature forests. Because of the relatively large number of available habitats, Grand Cayman has 3.8 habitats per species, an exceptionally high figure when compared with mainland and other insular populations. This value plus other data indicate a remarkable generalization for the Cayman Island birds. The large number of wintering species appears to influence the residents very little because the wintering forms usually occupy feeding niches different from the residents. It is suggested that the winter is characterized by an abundant food supply ( chiefly insects) that is incompletely exploited by the resident avifauna. TABLE OF (ONTENTS INTRODUCTION . 236 ACKNOWLEDGMENTS. ..... . 237 GENERAL DESCRIPTION OF THE ISLANDS AND THEIR AVIFAUNA . 238 EC:OLOGICAL FORMATIONS _ ..... 243 AV[FAUNAL DISTRIBUTION IN THE ECOLOGICAL FORMATIONS 253 POPULATION DENSITIES ON CAYMAN BRAC 259 BIRD POPULATIONS AND UPLAND SECONDARY PLANT SUCCESSION... 261 BREEDINC, SEASONS .. ...... 265 INTER-INIAND DISTRIBUTIONAL ~ATTERNS 268 IMMIGRATION AND EXTINCrioN . ....... ...... 272 COMPETITION FOR ENVIRONMENTAL RESOURCES .4.#1036 1HE RESIDENT TERRESTRIAL Avit-AUNA · 275 PREDATION AND POPULATION CONTROL . 281 A GENERAL ASSESSMENT OF FEEDING Ec:oLOG,ES ANI) HABIT.4-r DISTRIBUTIONS . 283 TAxoN CYCLES .,- 283 HABITAT DISTRIBUTIONS OF INSULAR AND MAINLAND BIRD POPULATIONS. 288 THE WINTER AviFAUNA ,. 291 LITERATURECITED- -- 293 APPENDIX 1: SCIENTIFICAND COMMON NA.ME.% oF BIRDS APPEARING IN TEXT . 296 APPENDIX II: SCIENTIFIC ANI) COMMON NA1!ES OF PLANTS APPEARING IN TEn 298 AppEN'Dix III: ~TOM:\CH (ONTENTS OF (AYNi.AN ISLAND BIRDS 299 INTRODUCTION Despite the recent burgeoning of interest in island biogeography, few complete ecological investigations have been conducted on insular forms of virtually any taxon level. Theoretical treatments of insular population biology ( e.g., MacArthur and Wilson 1967) necessarily relied heavily upon many literature sources in the construction of species-area curves, extinction-immigration curves, and other ecological models. This is not meant to imply that such derived models are necessarily incorrect or in- adequate, but simply to underscore the need for more raw data on island 1975 JOHNSTON: CAYMAN ISLAND AVIFAUNA 237 populations so that ecological models and generalizations might be more rigorously tested, amended, or even rejected. Since complete data for avian insular populations are particularly sparse, scattered in the litera- ture, and often incomplete, the ecological model-builder or synthesizer must consult a multitude of scientific papers by different authors, each using different techniques and reporting usually single ecological param- eters from widely different islands. For the entire complex of West Indian Islands, not one island has yet been subjected to a thorough, com- prehensive investigation of its avifauna, despite-the distributional survey by Bond ( 1971) and the general assessments of taxon cycles by Ricklefs and Cox ( 1972). Accordingly, this report is an assemblage of published data, several years of personal field experience and collecting, and a thorough analysis of as many ecological parameters as possible for a single insular avifauna, namely that of the three small isolated Cayman Islands. Attention is focused on "standard" ecological parameters such as population densities, competitive interactions, food and feeding behavior, reproductive cycles, predation and other population controls, habitat and stratal distribution, and secondary succession. Even this comprehensive report, which cor- roborates -and augments ideas from other insular studies, is admittedly incomplete in one important aspect-the impact on resident bird popula- tions by a large contingent of North American birds that overwinter in these islands. ACKNOWLEDGMENTS Through the years a number of organizations have provided financial support for these investigations. They include a Biomedical Institutional Support Grant from the Division of Sponsored Research of the University of Florida, the Bradley Fisk Fund, the American Philosophical Society ( Johnson Fund, Penrose Fund), and National Science Foundation ( GB-2114) Partial subsidy for publication came from the Bradley Fisk Fund and the Division of Biological Sciences, University of Florida. Assistance in Seld observations and collections came from Jon C. Barlow, Charles H. Blake, Donald W. Buden, Alexander Cruz, Erma J. Fisk, and Albert Schwartz. David May made the insect identifcations and Albert Laessle most of the fruit and seed identifcations. Walter Auffenberg, Pierce Brodkorb, and Ronald Pine kindly assisted in the identification of vertebrate prey items. I am particularly indebted to Alexander Cruz, Daniel Simberloff, Carmine Lanciani, and Robert Ricklefs for spirited and helpful discussions that shed much insight on ecological problems concerning insular avifaunas. C. D. Hutchings, Chief Agricultural officer, and R. F. Pocock, Chief of Police of the Cayman Islands, were both cooperative in many ways espe- cially in the collecting of birds. Finally, of inestimable value has been the lifelong field experiences of the islands' principal naturalist, Ira Thompson. 238 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 5 GENERAL DESCRIPTION OF THE ISLANDS AND THEIR AVIFAUNA The three Cayman Islands ( Grand Cayman, Little Cayman, Cayman Brac) lie in the northwestern Caribbean Sea where, as a group, their remote position is rather extreme among the many West Indian islands. Grand Cayman ( 19°20'N,81°20'W) is approximately 290 km ( 180 mi.) south of Cuba, about the same distance northwest of Jamaica, and 480 km northeast of Honduras, the nearest point in Central America. Car man Brac ( 19°43'N,79°50'W) is 89 km east of Grand Cayman and 8 km east of Little Cayman. These three limestone islands are of similar geological structure and represent the projecting peaks of the ancient submarine Cayman Ridge extending from near British Honduras to Cuba ( Richards 1955). Steep-sided submarine slopes occur around the islands, with a 100-fathom line lying only a few hundred meters offshore. Bart- lett Deep, a 6,200 m trench, is found just south of Grand Cayman. Further indications of the isolation of these islands are the many well- marked animal species and subspecies that have been described from them, including birds ( Johnston et al. 1971), insects ( Clench, H. 1964), mollusks ( Clench, W. 1964), and reptiles ( Grant 1940). These islands are typically low and flat, Much of Grand Cayman ( 185 lan: ) is less than five meters in elevation, although an east-west forested ridge on its north side reaches 20 m in places. Low-lying la- goons and inland swamps abound ( Fig. 1). Chiefly because of a beauti- ful beach of coral sand on its western side, most of the human popula- 81°20 81'10 j ~,-4f' ~114 SOUNDWEST BAY , ,_1· p -+F-41 K-4 41 1Flj -t.~- -r- 4-P A -r GEORGETOWI~»R. p 1 4-6-'-1 EAST END -< r \=4244 .-/ p w 1 UMESTONE FOREST SCALE 51 MANGROVE FOREST r=r~=1 I POND 3 KM FIGURE 1.-Map of Grand Cayman showing general distribution of major forest types and ponds. P = pasturelands/clearings; R = residential areas. 1975 JOHNSTON: CAYMAN ISLAND AVIFAUNA 239 tions are currently on that portion of the island, namely at Georgetown or West Bay or between the two. ( In the early 19705, however, an extensive construction "program" was rapidly expanding eastward and progressively engulfing formerly untouched natural communities.) Ex- cept along the western "seven-mile beach" a typical coral reef fringes the island. A honey-combed bluff limestone formation, often with an intermittent ridge and swale effect, characterizes the central, eastern, and southern parts of the island, whereas "ironshore" formation is espe- cially prominent adjacent to the shoreline ( Doran 1954). Pockets and crevices in the underlying rock frequently include a reddish soil formed by erosion of the original limestone. Topographically and geologically, Little Cayman ( 24 kmz ) resembles Grand Cayman ( cf. Fig. 1 and Fig. 2), except that relatively few places on the smaller island are over 4 m in elevation and the highest point is only 14 m. The few human inhabitants, largely clustered at the south- west end, have disturbed Little Cayman very little. A recently con- structed road nearly encircling the island's shoreline forebodes "develop- ment." Cayman Brac ( 31 km2 ) differs physiographically much from the other two islands, chiefly by the presence of an ascending ( west to east) plateau with bluffs that reach 43 m at the island's eastern end. Abutting both the north and south sides are bluffs ( Figs. 3 and 4). Closely asso- ciated with the existence of the high bluffs is a marked reduction in la- goons and mangrove swamp forests on Cayman Brac. The human popu- 19'43 j %2 LIMESTONE FOREST SCALE Il MANGROVE FOREST I POND IKM 80'05' i 80000' t FIGURE 2.-Map of Little Cayman. Legend as in Figure 1. 240 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 5 02 LIMESTONE FOREST 19'4512 MANGROVE FOREST SCALE I POND 1 KM '-' - I , b.-. AREAS *~ . USS-CULTIVATEZ 19°41 79'50'i 79'45' FIGURE 3.-Map of Cayman Brac. Legend as in Figure 1. FIGURE 4.-Precipitous bluff on east end of Cayman Brac; sole breeding sites of Sula leucogaster and Phaethon tepturus in the Cayman Islands. lation of a few hundred persons is now chiefly restricted to the north coast and western tip, although coconut plantations were once common on both coasts. Largely because of their relative inaccessibility, the bluff or limestone forests in the central highlands of the island have been little disturbed, except for scattered, small cultivations of vegetable crops. 1975 JOHNSTON: CAYMAN ISLAND AVIFAUNA 241 The climate of the Cayman Islands is characterized by mild to hot temperatures and a distinct dry season. A weather station maintained at Owen Roberts airfield on Grand Cayman provided the temperature and rainfall data in Table 1. From the data supplied by the weather station personnel, the mean annual high temperature was calculated to be 30°C. Seasonal fluctuations in temperature are not extreme. On the other hand, both rainfall ( annual mean of 154.9 cm) and prevailing winds do change seasonally. From May to October the prevailing winds are easterly, but through the winter months they tend to be from the northeast or northwest. A dry season extends from November through April, and periods of drought are frequent at other times. De- spite an annual wet season during the summer and autumn months, the islands' low relief, desiccating winds, shallow soils, and porosity of the limestone formation preclude the humid, tropical, luxuriant vegetation characteristic of many West Indian islands. In fact, much of the natural upland vegetation ( see later discussion), including the introduced species, takes the form of low xerophytic scrub, many of the plants pos- sessing sharp spines and small leaves. Occasional autumnal hurricanes buffet the islands, and their effects may be quite severe, as was the hurri- cane of October 1944. TABLE 1.-TEMPERATURE AND RAINFALL MEASUREMENTS AT OWEN ROBERTS AIR- FIELD, GRAND CAYMAN, 1957-1970. Mean temperatures (C) Mean rainfall ( cin) maximum minimum rnean extreme rnean extrenie January 28 32 21 13 4.39 February 28 31 21 11 3.30 March 29 32 21 13 1.57 April 30 32 21 15 4.52 May 30 33 22 14 19.25 June 31 34 24 21 23.93 July 32 33 24 21 15.06 August 32 33 23 22 15.82 September 31 33 23 20 21.16 October 30 32 23 21 27.25 November 29 31 22 15 12.70 December 28 30 22 14 7.11 Through 1971, 151 species of birds had been recorded from the Car mans ( Johnston et al. 1971). Most of these are North American migrants either in-transit or overwintering in the islands. A breakdown of the avifauna is as follows: 242 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 5 Grand Cayman Little Cayman Cayman Brac 185 km2 24 km2 31 km, known or suspected breeding species: a. aquatic 13 12 10 b. terrestrial 26 17 20 total 39 29 30 These data fall within the expected ranges on a species-area curve deter- mined by Ricklefs and Cox ( 1972) for small West Indian islands. Species numbers ( 12-47) for the small ( 18-658 km2 ) satellite islands off Hispaniola also generally conform to this species-area curve ( Schwartz 1969). Bond ( 1934: 345) believed that "Grand Cayman has received most of its bird life from Jamaica and Cuba, whereas Little Cayman and Cayman Brac have derived theirs from Cuba alone. . . ." Certainly the breeding terrestrial avifauna of Grand Cayman has many species in common with Cuba or Jamaica. But of the two smaller islands only one species ( Mimocicha mbripes), which is restricted to Cayman Brac, has amnities with Cuban birds, whereas the remaining 18 species on the smaller islands have affinities with forms occurring in both Cuba and Jamaica. Furthermore, the Grand Cayman avifauna includes more species related to the Cuban avifauna than do the smaller islands. The extent to which the origins of all the Cayman Island avifauna can be assessed accurately will always be speculative, partly because such an evaluation would depend upon the taxonomic level under con- sideration. For example, most biogeographers consider the family Paru- lidae to be of North American origin ( Mengel 1964), but Dendroica petechia eoa, the Cayman Island form, also occurs on Jamaica. So, did this Yellow Warbler reach the Caymans from North America ( perhaps via Cuba) and then spread to Jamaica, or follow the opposite route, or neither one? We really do not know the answer to this and related questions. It is nevertheless possible to assess affinities of avifaunas in a general way, as Bond ( 1934, 1966a) has attempted, and one would have to agree with Bond that by and large the Cayman Island avifauna probably originated chiefly from Central American elements. In this group, I would include the Columbidae, Cuculidae, Vireonidae, Coere- bidae, Icteridae, and some Fringillidae. On the other hand, the closest relatives to the Caymanian Amazona, Mimocichla, Colaptes, Centurus, and Melopyrrha are presently Cuban. Contributions from Jamaica de- pend partly upon one's viewpoint but quite likely these include Chor- deiles minor, most or all of the four species of Tyrannidae, Dendroica petechia, and perhaps Spindalis and Quiscalus. 1975 JOHNSTON: CAYMAN ISLAND AVIFAUNA 243 My own field work and collections on these islands were initiated in 1965 and have continued intermittently for eight years, covering all sea- sons and all three islands, although most of the data were taken on Grand Cayman. Additional valuable field notes; specimens, bird stomachs, and other experiences on the islands in the past decade have been kindly shared with me by Albert Schwartz, Erma J. Fisk, Alexander Cruz, Ronald Pulliam, Jon C. Barlow, and Donald W. Buden. ECOLOGICAL FORMATIONS So that the distribution of the avifauna can be analyzed generally, major ecological formations for the islands are proposed, defined, and de- lineated here. Fundamental to the definition of the formations are major and minor plant associations, distinctive topographic features, and perti- nent geologic, edaphic, and hydrologic conditions. These ecological for- mations are similar to, if not identical with, some of the major plant communities described by Asprey and Robbins ( 1953) from Jamaica. Implied for each formation are those general and specific niche require- ments essential to the successful occurrence of each of its avifaunal com- ponents, including food and other trophic relationships, available nesting and feeding sites, and protective cover. In addition to the plant com- munity analyses by Asprey and Robbins ( 1953), other reports have proven valuable in the definitions of these formations, namely those by Doran ( 1954) and Richards ( 1955) on geology, and Swabey and Lewis ( 1946) on forestry. In evaluating insular avifaunas it is sometimes useful and desirable to include a "marine littoral element, essentially pelagic birds depending upon offshore waters for food and the island for breeding sites. Although thi5 contingent is not considered here as a distinct, important ecologic formation, it is emphasized that at least four species of birds breeding on the Cayman Islands would be considered pelagic during the non- breeding season-Phaethon lepturus, Sula leucogaster, S. sula, and Fre- gata magnificens. FRINGING REEFS AND LAGOONS.-For the most part the three islands possess fringing coral reefs that enclose shallow lagoons or sounds ( Fig. 5), although on Cayman Brac the reefs are limited to the southwestern tip. The reefs, usually 30 to 100 m offshore, typically vary from about 0.5 m below sea level to about 0.3 m above. Water in the enclosed la- goons may be 3 m or more deep but is usually less than 1 m deep, De- spite the incessant pounding of surf on the reefs and water depth of the lagoons, both are visited periodically for food by some herons, egrets, and terns, plus an occasional shorebird, especially at low tide. 244 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 5 ..... --. -- -- FIGURE 5.-Typical fringing reef, lagoon, and sand-coral beach, Grand Cayman. A distinctive feature of Grand Cayman is its immense North Sound (Fig. 1), embracing some 64 km2 and partially separated from the open sea to the north by a broken reef and shoals. The greatest depth of the sound ( near its center) is only about 6 m; much of its shallow floor abounds in turtle grass ( Thalassia testudinum). Although the boundary of the North Sound with Iand is not always clear-cut, nonetheless much of its periphery is composed of red mangroves ( Rhizophora mangle), with occasional mangrove islands in the sound. The sound and its fring- ing red mangroves are not generally used by aquatic birds, but at least one island ( Booby Cay) is a well-known heron rookery, whereas other small mangrove islands provide temporary roosting sites for frigate birds, herons, and egrets. SAND-CORAL BEACH.-The coastlines of the islands are composed of coral cobbles, sandy beaches, or steep bluffs. The beaches are best de- veloped for about 5 km on the western end of Grand Cayman, a popular resort area. Steep bluffs ( 6-30 m) dropping off into deep water are best developed along the northeast coast of Grand Cayman and at the eastern end of Cayman Brac. The sand-coral beach formation ( Fig. 5) consists of blown coral sand and well-rounded coral fragments piled up and worn by winds, waves, storms, and hurricanes. This formation, some 3-10 m wide, has little covering vegetation, but some pioneer plants grow on it including 1975 JOHNSTON: CAYMAN ISLAND AVIFAUNA 245 Sesuvium portulacastrum, Ipomea sp., and Sporobolus virginicus. On the landward side the sand-coral beach formation is continuous in many places with a rugged, irregular shore platform termed "ironshore." It slopes gradually to a height of 4-5 m, may extend inland some 10-50 m, and also possesses sparse or no vegetation, except in certain areas where woody thickets of Caesalpinia bonduc have developed. SEA GRAPE-ALMOND WOODLAND.-This coastal vegetation formation is characteristic of the West Indies ("strand woodland" of Asprey and Robbins). From the "ironshore" or, where it is absent or poorly de- veloped, from the sand-coral beach formation, the gradually ascending beach ridge supports a distinctive plant community dominated by Cocco- loba uvifera, Termin4>4>4 Z Tyrannus caudifasciatus Mviarchus stolidus NNN Elaenia martinica Mimus polyglottos ???? Z z z Mimocichia plumbea Vireo crassirostris Vireo magister N Dendroica petechia NY NY Dendroica uitellina N Y Coereba flaueola N N NY N NNNN N Spindalis zena N NY Quiscalus niger Y Tiaris olivacea N N N Melopurrha nigra YY Y Totals (N and/or Y) 1+ 2+ 8+ 11+ 15 21 8 4 1 1 2 5 *includes all breeding species for which data are available, based upon nests (N) and/or dependent young (Y). 268 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 5 ably Mimus poll/glottos. Crotophaga is obviously an omnivore ( see Ap- pendix III), Tyto preys on birds and rodents, and Coereba is chiefly a nectarivore. Considering the small populations of Tyto on these islands, an apparently adequate food supply, and the absence of a competitor, the breeding season of this owl should not a priori be restricted by either a wet or dry season. Crotophaga is more abundant, occurring in seat- tered but small and discrete groups in many ecological formations; be- cause of its omnivorous habits, it too would hardly be restricted in breeding by wet or dry seasons. Perhaps a similar argument could be made for Mimus, but less is known of this species' annual cycle. Coereba is an interesting case, because throughout its range in the Caribbean re- gion it is renowned for a protracted breeding season. In the absence of hummingbirds ( at least for the Cayman Islands) this Bananaquit can utilize a unique food resource ( nectar and tiny insects) to support a large and widespread population at all seasons. INTER-ISLAND DISTRIBUTIONAL PATTERNS Archipelagos are renowned for their interrupted distributional pat- terns among the avifauna. In this context, examples are found in indi- vidual species of Galapagos finches ( Geospizidae) and Hawaiian honey- creepers ( Drepaniidae). However, patterns above the family level, al- though often well known for a given archipelago, have largely been over- looked in the literature, and since the Cayman Island avifauna contains conspicuous examples of distributional -anomalies," an analysis of these distributional patterns is discussed here and summarized in Table 14. One is faced, of course, with the obvious question-why is species A restricted to a given Cayman island and yet species B, whether a con- gener or not, occurs on all three islands? At the outset, several basic premises must be recalled and at least temporarily accepted: ( 1) the three Cayman islands are of a similar geological age; ( 2) with relatively minor exceptions ( introduced logwood and its scattered stands limited to Grand Cayman), ecological formations or habitat structures appear to be identical, or nearly so, among the·three islands ( see previous discussion); ( 3) the three islands are each approximately the same distances from both Cuba and Jamaica, the potential if not actual sources of most of the avifauna; (4) Grand Cayman has a much greater land area than the smaller islands, and certainly the areal expanse of each ecological forma- tion is far greater on the larger island; and (5) there is no evidence that hurricane tracks or wind or water currents are extraordinarily restricted to any one of the islands. Pertinent to these disjunct distributions are the statements by White- head and Jones ( 1969: 176): if one assumes that the rate of move- 1975 JOHNSTON: CAYMAN ISLAND AVIFAUNA 269 TABLE 14.-SOME INTER-ISLAND BREEDING DISTRIBUTIONAL PATTERNS. Species Grand Cayman Little Cayman Cayman Brac Phaethon Zepturus breeds Sula leucogaster +0 breeds Sula zula breeds Fregata magnificens breeds + Dendrocugna arborea breeds breeds Columba leucocephata breeds + breeds Leptotita jamaicensis breeds Amazona teucocephala breeds + breeds Colaptes auratus breeds Centurus supercilians breeds Mgiarchus stolidus breeds Mimocichla plumbea + breeds Vireo altiloquus + breeds breeds Vireo magister breeds Spindolis zena breeds Quiscalus niger breeds breeds + Melopwrha nigra breeds Totals = 17 11 5 6 Breeding endemic on 7 2 3 only one island 0 + - present, but contemporary breeding unconfirmed ment of propagules... across a given segment of ocean surface is es- sentially constant through time, it follows that larger islands will intercept a larger number of disseminules per unit time. On this basis one would expect Grand Cayman to support a larger resident avifauna ( 39 species) than the two smaller islands ( Cayman Brac 30, Little Cayman 29). This is only a partial explanation, however, because habitat areal expanse and standing water with its concomitant mangrove swamps are also important factors supporting a diversity of bird life. With these ideas in mind, we can initiate an analysis of the inter- esting, interrupted distributional patterns as summarized in Table 14. Phaethon Zepturus and Sula leucogaster prefer or require for breeding the high bluffs occurring only on Cayman Brac. Sula sula and F1·egato magnificens currently maintain a moderate-sized breeding colony in the landward mangrove fringe of a large shallow lagoon on Little Cayman. The physiognomy of this lagoon site appears similar to those on Grand Cayman, but these birds have not bred on Grand Cayman, at least in historical times. Perhaps quite local prevailing winds and the lack of human disturbances are primary factors in the restricted breeding distri- butions of these two species on Little Cayman. Dendrocygna arborea is evidently confined as a breeding bird to Grand Cayman and Little Cayman where mangrove swamps and ponds are common ( cf. Figs. 1,2, 270 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 5 3); breeding of this species has not been confirmed for Cayman Brac, an island where open water and mangrove swamps are quite limited. Columba leucocephala and Amazona leucocephala evidently do not currently breed on Little Cayman, although both species are at least moderately common on the other two islands. Longtime human residents of Little Cayman report that both of these species breed on Cayman Brac and that small flocks of each undertake daily round-trip Hights to Little Cayman for feeding purposes only. Leptotila famaicensis, a large ground- feeding dove, is a species with one of the most restricted ranges of all the Cayman Island avifauna, being found only in the remote inland lime- stone forests of Grand Cayman. Why it does not occur on the other two islands is unknown. Quiscalus niger, a ubiquitous and conspicuous resi- dent of both Grand Cayman and Little Cayman, is virtually unknown on Cayman Brac. Small groups sometimes fly back and forth between the two smaller islands ( 8 km) but there is presently no rational expla- nation for its absence as a breeding bird on Cayman Brac. The eight remaining species in Table 14 present puzzling distribu- tional patterns among the three islands . Why are Colaptes, Centurus, Myiarchus, Vireo magister, Spindalis, and Melopyrrha all restricted to Grand Cayman? Why does not Mimocichla plumbea occur permanently on Grand Cayman and Little Cayman, the latter only 8 km from Cayman Brac? Apparently, favorable habitats for this species would be the same for all three islands. One is tempted to explain these patterns by sug- gesting that the original successfully breeding pair (s) reached only one island ( usually the largest) from the source population. In subsequent times, perhaps fortuitously, propagules have failed to reach the other islands, and the populations now established on the single island became sedentary. This could be the explanation for these eight species. Is it probable, however, that of the four forms with definite Cuban amnities ( Colaptes, Centurus, Melopyrrha, and Mimocichla), all but the latter one should colonize only Grand Cayman? In reviewing all these distributional patterns, I am inclined to support a chance-colonization hypothesis for a number of reasons. In the Brst place, preferred ecological formations on Cayman Brac, for example, ap- pear to be potentially suitable to all these terrestrial species currently restricted to Grand Cayman, and it is likely that they simply never reached the other islands. Alternatively, populations could have become extinct. Second, two species ( Zenaida asiatica and Mimus polyglottos) have within quite recent years spread from island to island and now maintain breeding populations on islands where each was absent as recently as 25 years ago. Third, Mimocichla plumbea has been recorded as a vagrant on Grand Cayman ( Johnston 1969) where an unmated fe- 1975 JOHNSTON: CAYMAN ISLAND AVIFAUNA 271 male, previously unrecorded from this island, even constructed unsuccess- ful nests in at least two successive years. The point is that with the pos- sible exception of the congeneric vireos in which competitive exclusion may be an important limiting factor ( see discussion beyond), most if not all of the terrestrial birds currently restricted to a single island could probably successfully colonize any of the other islands ( contra Bond 1934: 345-346). Perhaps the sedentary habits of species now restricted to Grand Cayman and the distance ( 97 km) to the other islands have served as barriers to their dispersal to the smaller islands. This latter point, emphasizing a "poor" immigration rate of birds such as wood- peckers, is in substantial agreement with the views of Ricklefs and Cox ( 1972: 215). Among the most curious distributional features of the Cayman Island avifauna is the absence of resident hummingbirds. This fact is especially anomalous when it is recalled that (1) every other major island and land mass east, west, and north of the Caymans all have several resident hum- mingbird species ( Moynihan 1968; Bond 1971 ), (2) hurricane paths in the past argue for the possibility of widespread dissemination of birds in the Caribbean region, and (3) occasional hummingbirds are known to migrate through Grand Cayman ( Johnston et al. 1971). Collectively, these points tend to negate an hypothesis that hummingbirds could not or have not reached the islands in the past. The problem is, I believe, whether or not conditions are currently favorable for their existence in the islands. The absence of breeding hummingbirds is related to two factors, a dependable food supply and the possible role of potential competitors. As pointed out elsewhere in this paper, the small size, low relief, desiccating winds, sparse soils, and porous limestone base of the Caymans preclude the support of luxuriant tropical vegetation, despite the mod- erately heavy, though seasonal, rainfall. The extensive expanses of lime- stone and mangrove forests of the Caymans ( Figs. 1, 2, 3) are not characterized by the diversity, distribution, or abundance of flowering trees and shrubs that one finds elsewhere in the West Indies. Although quantitative support cannot be marshalled for this statement on plant diversity and distribution for the Cayman Islands, it is obvious that most of the favorite hummingbird nectar-bearing plants of Jamaica ( Hibiscus, Bougainoillea, Nerium, Tamarindus, Psidium and many others) are at best limited chiefiy to scattered and relatively sparse house sites on the Caymans. Thus, I believe that even if potentially immigrant humming- birds did currently find their way to the Caymans, a dependable food supply would not be available. Another consideration in the lack of hummingbirds is the possibility of 272 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 5 competition specifically with the ubiquitous Bananaquit ( Coereba fla- veola). Throughout the rest of the West Indies and Central America where hummingbirds and Coereba are sympatric, there is much evidence that hummingbirds dominate Coereba at food sources ( see, for example, Wetmore 1927). Furthermore, hummingbirds and these bananaquits generally feed in different fashions, though often utilizing the same food source simultaneously. None of these features suggest that competition does or would play a signiBcant role in explaining the absence of hum- mingbirds on the Caymans. Intriguing questions on niche breadth, occupancy, and vacancy are invoked by these inter-island distributional patterns. Characteristic timber-probing woodpeckers ( Colaptes and Centurus) are absent from the two smaller islands. For these woodpeckers island size and habitat diversity are important limiting factors. Three species that are frugi- vorous in low trees or on or near the ground surface on Grand Cayman are Leptotila, Spindalis, and Melopyrrha. Are these feeding niches simply vacant on the smaller islands? They probably are, but not enough is known of comparative feeding behaviors of species with similar feeding patterns where a competitor is absent. For example, it is possible, though not documented, that Elaenia has a broader feeding niche on Cayman Brac ( both qualitatively and quantitatively) than on Grand Cayman, where it undoubtedly competes with Spindalis, Melo- pyrrha, and other species for small fruits. Similarly, the absence of the foliage-gleaning insectivorous Myiarchus on the two small islands might release potential competitors ( Viteo, Dendroica) to the extent that their feeding behavior patterns are broadened on Little Cayman and Cayman Brac. These and related points certainly merit future intensive studies. IMMIGRATION AND EXTINCTION An equilibrium model for the number of species on an island was de- veloped by MacArthur and Wilson ( 1967), who predicted that the num- ber of species on an island is determined by a balance between immigration and extinction rates. Our current knowledge of the Cayman Island avifauna provides some concrete examples in support of this model, despite the fact that few bird species and a relatively short period of time are involved. There is sufficient evidence, for example, that two birds ( Mimocichla rauida and Icterus leucopteryx bairdi) are now extinct on Grand Cayman, where formerly ( 1900-1916) both were mod- erately common and were last observed in the 1930s (Johnston·et al. 1971). Unfortunately, no precise information is available on the causes of their extinction, whether by hurricanes, partial habitat destruction, human dis- 1975 JOHNSTON: CAYMAN ISLAND AVIFAUNA 273 turbance, or a combination thereof. Another thrush, Mimocichla plumbea rubripes, became extinct on Swan Island, possibly because of forest de- struction ( Paynter 1956). Both species on Grand Cayman, incidentally, were undoubtedly in Stage IV of a taxon cycle because, among other characteristics, each was highly restricted in habitat distribution, the thrush probably to limestone forests and the oriole to sites on the north- ern coast. ( I. l. leucopteryx is still common on Jamaica as is I. l. lawrencii on St. Andrews Is.) Conversely, the White-winged Dove (Zenaida asiatica) has invaded all three Cayman Islands since about 1935 and is now a well-established and common resident, especially in more peripheral habitats of the islands. A spurious attempt at coloniza- tion by an unmated female Mimocichla plumbea on Grand Cayman was earlier documented by Johnston ( 1969). If one can include this last occurrence as a bona fide instance of immigration, then within the last 40 years Grand Cayman alone has experienced two instances of avian extinction and two of immigration. Even if one considers the Mimocichia incident as being an unsuccessful immigration, the data still fit the Mac- Arthur-Wilson model for a small island where the probability of extinc- tion increases with decreasing island size. Immigration and extinction are not well documented for birds on the smaller islands ( Cayman Brac and Little Cayman), chiefly because few ornithologists have carefully studied the birds of these islands until re- cently ( Johnston et al. 1971). The interrupted distribution of certain species on these islands is both curious and unexplained ( see Table 14), especially since histories underlying the current distributional patterns are largely unknown, and little information is available on food, habitat, and other ecological factors. Evidently Amazona leucocephala does not presently breed on Little Cayman ( if it ever did, despite old specimens and observations to the contrary), but small flocks regularly fiy over from Cayman Brac during the day to feed on ripe fruits. Conversely, the older ornithological records indicate that Quiscalus niger was once common on Little Cayman and Cayman Brac; today, small flocks Hy from Little Cayman to Cayman Brac to spend the day feeding, but evidently this grackle does not now breed on Cayman Brac. Mimo- cichla plumbea is moderately common at least on the lowland perimeter of Cayman Brac ( Table 11); yet it has never been recorded from nearby Little Cayman but has attempted to breed on Grand Cayman. As late as 1956, the Mockingbird ( Mimus polyglottos) was unknown from both Little Cayman and Cayman Brac (Bond 1956), even though the species has been a common resident of ruderal and roadside habitats on Grand Cayman at least since the earliest days of ornithological in- vestigations in 1886. By 1956 Mockingbirds were reported on Cayman 274 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 5 Brac by C. H. Blake ( Bond 1958), and the species became well estab- lished on that island around "inhabitated areas" by 1966 ( Harvey, in Bond 1967). Harvey also noted the species on Little Cayman in the same year. In the summer of 1971, I recorded nine well-spaced individuals on Little Cayman during two hours of observation aIong 8 km of that island's newly constructed perimeter road. It is apparent that construe- tion of this road has created open habitats preferred by this species. The mockingbird probably immigrated to the two smaller islands from Grand Cayman where it continues to increase in number, probably in response to continued clearing of mangrove swamps and logwood forests, thus increasing its preferred habitat. As discussed elsewhere in this paper, the distribution of the genus Vireo on these islands is curious, because V. crassirostris occurs on all three islands with either V. magister ( Grand Cayman) or V. altiloquus ( Little Cayman and Cayman Brac ). No more than two species of ViTeo are sympatric on a given island. Bond ( 1966b) mentioned a single old record of two V. altiloquus on Grand Cayman that probably represented vagrants. The fragmentary data available on the ecology of Mimocichla rauida ( English 1916) and Icterus leucopter!/x on Grand Cayman conform to Mayr's ( 1965) belief that small population sizes and probable genetic uniformity have made such populations exceptionally vulnerable to the smallest environmental change. This is another way of invoking genetic drift as a contributing factor in the extinction of these two forms on Grand Cayman. Both existed in quite restricted habitats ( thus reducing gene flow), and their population densities in any one year probably never exceeded 100 breeding pairs. In these small populations, by genetic drift some alleles favored by selection could have been lost and less favored ones, perhaps lethals, could have increased in frequencies. These birds might not have been able to adapt to ensuing environmental changes. As stated before, the nature of that change, or changes, is unknown, but possibilities would include some habitat disturbance or hurricane effects. Furthermore, each of these small Cayman Islands appears to exemplify Mayr's statement ( 1965: 1587) ". . . that the smaller the island the lower the percentage of endemic species.... In fact, plotting data from these small islands on Mayr's Sgure 2, the linear relationship between double logarithmic plottings for island area on percent of endemic species would become curvilinear, the line extrapolating to zero endemic species with an island area of approximately 70 square miles or less. According to this interpretation, none of the Cayman Islands should have any endemic species, and currently they do not. In the Brst analysis of the birds of these islands, Cory (1886) de- 1975 JOHNSTON: CAYMAN ISLAND AVIFAUNA 275 scribed 13 new species from Grand Cayman alone, and subsequently, three other species were recognized by early systematists from the other two islands ( for a list, see Johnston et al. 1971). Later taxonomists ( e.g., Ridgway and Friedmann 1901-1950) synonymized all of these other West Indian forms, except for Mimocichla rauida. Consequently, a modern treatment ( Johnston et al. 1971) recognized no current endemic species from the islands, but 16 endemic subspecies. This latter Egure appears to be high for small West Indian islands and suggests a relatively long period of isolation for the resident avifauna, with little or no recent im- migration. A similar example of insular equilibrium between extinction and im- migration is Diamond's ( 1969) analysis of the Channel Islands of Cali- fornia. Those islands lie from 8 to 61 miles from the mainland. Be- tween 1917 and 1968 from 17-62 percent of the breeding bird species disappeared, but an approximately equal number of new immigrants became established. COMPETITION FOR ENVIRONMENTAL RESOURCES AMONG THE RESIDENT TERRESTRIAL AVIFAUNA Ecological isolating mechanisms, the avoidance of competition, and related phenomena in birds have been discussed by numerous investi- gators and were summarized recently by Lack ( 1971). The thesis of Lack and others is that contemporary bird species can coexist in the same area only if they differ in some ecological factor (s). If an investigator interested in sympatric forms ( genera, species) A and B looks long and hard enough, presumably he will discover that their coexistence is made possible by differences in habitat, food, feeding methods, stratification, or the Iike, because the only other alternatives are competitive exclusion or extinction ( i.e., coexisting species will have different ecological niches). Theoretically, at least, the more niches available in a given community, the more species of birds that can actually or potentially coexist in that community. Interesting cases are those of insular birds with quite similar niches; these have been the subject of intensive studies among the closely related resident Cayman Island birds. A morphological indication of reduced interspecific competition be- tween closely related birds is difference in bill size, which, in turn, pre- sumably reflects differences in food particle size and hence partitioning of food resources. Grant ( 1968, 1969),for example, demonstrated that sympatric congeneric species of birds on islands differ in bill length and avoid competition by feeding in different habitats or by having different feeding habits. Schoener ( 1965), in developing this theme quantitatively, 276 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 5 suggested that a character difference ( ratio of larger to smaller bill length) greater than 1.14 was typical of sympatric congeneric insular birds. Congeners with such bill length differences would reduce com- petition by selecting different-sized food particles. Of the five congeneric pairs of birds on the Cayman Islands, two have a bill character difference less than 1.14: Dendroica petechia-D. uitellina ( 1.01) and Tyrannus dominicensis-T. caudifasciatus ( 1.02) ( bill measurements taken chieRy from Ridgway 1901- 1950 ). In the two Dendroica species stomach anal- yses revealed few qualitative differences in food choices ( see discussion beyond and Appendix III); competition between these forms is avoided primarily by differences in habitat choice as well as feeding heights. The Tyrannus species, although possessing similar bill lengths, (1) take dif- ferent foods, (2) are found in different habitats, and (3> obviously feed in different fashions. Bill size differences are, therefore, not always the most important means of avoiding competition in all the Cayman Island congeners. As will be discussed later in this paper, congeners and other closely related forms clearly have evolved a spectrum of mechanisms that in various combinations facilitates coexistence. It should be noted that in the relatively short time that birds have been studied on these islands, there is no assurance that competition is not causing a gradual exclusion of one species or a slow transition into a new ecologic niche for another. DovES AND PIGEONS ( COLUMBIDAE ).-A nearly linear relationship exists between island size and number of species of Columbidae in the West Indies ( A. Cruz, pers. comm.). Island size is not necessarily the prime factor, however, because certainly habitat diversity and distance to the source population are of major importance. The largest West Indian islands support the highest numbers of species of Columbidae ( Jamaica 10, Cuba 11, Hispaniola 10), but these islands also have greater relief and habitat diversities than are found on more xeric, flatter West Indian islands, such as the Caymans. Furthermore, on no island are the several dove and pigeon populations and densities necessarily the same; Columba inornata and Geotrugon passerina of Jamaica are both relatively rare, as compared with the common Columbina passerina and Columba leucocephala on that island. Relative population densities of the doves and pigeons on the Cayman Islands seem to be, in decreasing order of abundance: Zenaida asiatica, Columba leucocephala, Columbina pas- serina, Z. aurita, and Leptotila iamaicensis ( the latter restricted to Grand Cayman). Thus in terms of a given island, it is ecologically misleading merely to enumerate species without also considering relative or absolute abundances and habitat diversity. The question posed here is: how can each of the Cayman Islands sup- port its complement of four or five species of columbids? Or, how do 1975 JOHNSTON: CAYMAN ISLAND AVIFAUNA 277 these columbids partition resources so as to permit (1) sympatry on the island and (2) syntopy in given ecological formations? It is initially important to admit the paucity of information on their breeding habits, seasonality of breeding ( see Table 13), and feeding behavior. Never- theless, from the available information summarized in Table 15, it ap- pears that the columbid species partition resources in two principal fashions-by habitat selection and by food type. As breeding birds, no more than three species are found in a given ecological formation ( Tables 3-10). All these species are basically frugivorous and, for the syntopic species, limited stomach analyses ( see Appendix III) do not indicate any marked interspecific differences in food habits. TABLE 15.-NICHE CHARACTERISTICS OF DOVES AND PIGEONS ON GRAND CAYMAN. 1 Leptotila iamaicensis , 1 BS 2 Columba leucocephala FH 2 (H),BL BS,BL 3 Zenaida aurita 3FT FH,FT BL,H (BS) BL, (BS) 4 Zenaida asiatica FT? FT? FT? 4 BS,BL BS,BL BS,BL BS,BL 5 Columbina passerina H, FT H,FH FT (H),FH FT FT Key: BS ( body size), BL ( bill length), H (habitat), FH ( feeding height), FT ( food type ). A symbol in parentheses indicates that the difference is poorly developed. WOODPECKERS ( PICIDAE).-The two resident woodpeckers ( Colaptes auratus and Centurus superciliaris) of Grand Cayman were earlier be- lieved to be ecologically separated, chiefly on the basis of habitat, with Colaptes supposedly being a bird of the mangrove swamps and Centurus most common around human habitations ( Johnston 1970). Subsequent intensified study by Alex Cruz and me showed, however, that these species were not necessarily separated by habitat differences. As seen earlier in the distribution analyses of birds in the various ecological formations ( Tables 3-10), clearly both woodpeckers occur together in most of these wooded formations, although Colaptes predominates in mangrove swamps. Overall, the differences in habitat choice between these two woodpeckers are only of minor importance ( Table 16). The principal niche differences between these species are in feeding methods and food type. Colaptes is essentially a probing and drilling 278 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 5 TABLE 16.-NICHE CHARACTERISTICS OF CLOSELY RELATED RESIDENT BIRDS OF THE CAYMAN ISLANDS. body bill feeding feeding food® ° size length habitat height methods type Colaptes auratus 1° 20 2 1 30 4°Centurus superciliaris Tgrannus dominicensis Ty1annus caudifasciatus ~ 1 1 4 4 4 3 Elaenia martinica Mviarchus stolidus ~ 3 4 1 4 4 4 Vireo magister 1>GC 4 4 2 2 ? 2Vireo crassirostris 1 Vireo altiloquus >LC 4 4 2 2 ? ? ' CB Dendroica petechia ~ 2 1 4 4 1 1 Dendroica uitellina 0 Differences: (4) strongly developed, (3) moderately welI developed, (2) poorly developed, (1) absent or negligible. 0 ° Based chieRy upon percent differences in animal and vegetable material and not upon different species of animals or plant materials consumed. woodpecker, whereas Centurus spends much of its time in gleaning arthropods ( especially in bromeliads) and taking small fruits. Analyses of stomach contents indicate differences in foraging techniques. From Appendix III it can be seen that 19 Colaptes fed heavily ( 97 % of the diet) on arthropods, mostly on ants and termites. Although Centurus has a diet high in insects ( 56 % ), especially beetles, approximately one- half of its diet is fruit. In fact, the frequent occurrences of this wood- pecker at house sites are closely connected to the fruit-bearing trees, especially papaya, cultivated there. Of interest also are the "herptiles" taken by Centurus, the HI/la most likely captured by probing into brome- liads. FLYCATCHERS ( TYRANNIDAE).-Of all the closely related birds on the Cayman Islands, the four resident species of Tyrannidae possess the most interesting sets of ecological isolating features. No two species-pairs have evolved precisely the same mechanisms, which are based on dif- ferences in habitat choice, feeding height, feeding methods, and food types ( Table 16). Of particular significance is the fact that one of these "Hycatchers," Elaenia martinica, is chiefly a frugivore ( Appendix III), and as such does not compete with the other three insectivorous tyran- nids. Also, Tgrannus caudifasciatus consumes both lizards and frogs, a predatory habit not shared with the other tyrannids. Elaenia martinica is one of the most widespread and abundant members of the resident 1975 JOHNSTON: CAYMAN ISLAND AVIFAUNA 279 avifauna ( Tables 3-10) and is also extremely belligerent both intra- and interspecifically. In the hundreds of interspecific behavioral encounters I observed, it appeared that Elaenia was always dominant. Among the four species, within-habitat feeding strata clearly differed from species to species ( Figure 18 ); in no case did two species spend more than about 20 percent of the time feeding at the same stratum, and even when they did, each selected different foods. In summary, the tyrannids provide a classic case of the partitioning of environmental resources such that coexistence of four resident forms is possible on Grand Cayman. Elaenia is chiefly a frugivore, whereas the others are chiefly insectivorous; Tyrannus dominicensis is a large insectivore feeding at tree-top levels in open sites around towns and 100- A Elaenia G C 6-0 Elaenia C B • - • Myiarchus G C A Tyrannus dominicensis CB 75- 50- PE R C EN T O F TI M E FE ED IN G 25- S.0 S . 0- 1 , 1 0-Im 1-3m 3-6m >-6m FIGURE 18.-Feeding heights and percentages of time feeding for flycatchers. 280 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 5 houses; T. caudifasciatus is common in wooded formations where it consumes large insects and small vertebrates; Mytarchus is an inter- mediate-sized woodland and "edge" species that is more generalized in food habits ( insectivorous and frugivorous) than the other species. VIREOS ( VIREONIDAE).-The three resident vireos of the Caymans proved difficult to analyze for interspecific ecological differences, and the present summary should be considered tentative. As is true elsewhere in the distribution of the genus Vireo where sympatry and/or syntopy are evident, the usual case is the presence of an arboreal member of the subgenus Vireosyloa and a thicket-inhabiting member of the subgenus Vireo ( Hamilton 1962). Presumably these stratal and perhaps other differences effectively minimize interspecific competition. Over the years of observations on Grand Cayman, however, , it be- came apparent that V. crassirostris, although the only vireo inhabiting thickets and shrubby fields, coexisted with V. magister in early logwood successional stages. A similar relationship existed between V. crassirostris and V. altiloquus on the other two islands. Indeed, through the seral stages into a climax limestone forest formation, both forms could be found. Frequently both species were observed or collected in the same strata between 3 and 5 m. As a rule, V. crassirostris was most frequently encountered below about 4 m, whereas both V. magister and V. altiloquus tended to prefer foraging positions from 5 to 15 m. The relatively large number of stomachs examined ( 11 for V. crass- irostris and 9 for V. magister) and taken at the same season revealed food differences that (1) reinforce feeding strata differences and (2) provide the key to avoidance of interspecific competition ( see Appendix III). V. crassirostris consumes a higher percentage ( 77 % ) of animal food than V. magister (51%). Probably the species of Coleoptera or other taxa mutually eaten by both vireos were in fact different, although the insect fragments found in their stomachs could not be identified be- low the family level. Thus, the two vireo-pairings on these islands showed closer ecolog- ical niche characteristics than the other closely related resident avifauna, and should certainly be subjected to a more detailed scrutiny. WOOD WABBLERS (PARULIDAE).-The two resident warblers ( Den- droica petechia and D. vitellina ) were initially believed to be habitat restricted, with D. petechia being a mangrove swamp species and D. vitellina occurring principally in thickets, shrubland, and early logwood forests. Further observations showed that habitat separation was not complete because D. petechia is now known to be more widespread, occurring as a breeding bird in six of the upland ecological formations ( Tables 3-10 ). On the other hand, D. uitellina is found in only four of 1975 JOHNSTON: CAYMAN ISLAND AVIFAUNA 281 these. Attention must be given to the relative population densities of each species in each formation: where petechia is very common ( sea grape-almond woodland and mangrove swamps), D. uitellina is absent; where D. oitellina is fairly or very common (logwood, logwood-thatch palm-red birch, and limestone forests), D. petechia is at best uncommon. To some extent then, these species are segregated by habitat. Feeding heights of these warblers are also important, D. oitellina choosing the lower strata up to about 3 m, and D. petechia usually oc- curring from 3 m upwards. At present no available data support an hypothesis that the two species either have different foraging habits or select different foods. PREDATION AND POPULATION CONTROL Of special importance for the bird populations of the Cayman Islands is the obvious paucity of vertebrate predators. Grant ( 1940) docu- mented systematics of the herpetofauna but devoted little attention to their biological features. He did note ( p. 49), however, that the colubrid snake Alsophis angulifer is "diurnal, active, feeds on lizards, frogs, birds . . . Lewis took a number of specimens in the tops of trees, one over 30 feet off the ground... [where they were feeding on] Hylas." On Cay- man Brac a specimen of this snake, reported by Grant ( p. 50), had a honeyereeper in its stomach. In my several years of observing and col- lecting on these islands, I never encountered an Alsophis pursuing birds or with a bird in its stomach, nor did any of the local people indicate an ornithophagous habit for this species. Still, Grant does provide some evidence that the snake at least occasionally preys on birds. The abun- dant Anolis lizards, especially A. conspersus, might prey on birds' eggs ( there is no concrete evidence) because elsewhere in the West Indies Anolis occasionally eats eggs of honeycreepers ( Biaggi 1955). No mam- malian carnivores occur on the Caymans. Both Mus musculus and Rattus ( mostly R. rattus but some R. noruegicus) are found, but they are only locally abundant. Specimens of these mammals have been collected or observed chiefly in areas of human habitation and not in the "wilder" portions of the islands. A conspicuous hiatus in the avifaunal trophic structure is a diurnal predator. In winter an occasional hawk ( Circus c!/aneus, Falco sparver- ius and F. columbarius) does visit the islands ( Johnston et al. 1971). The food habits of these predatory birds in the Caymans are unknown, although elsewhere birds do constitute some portion of their diets. But even if they did prey on an occasional ( or more) land bird in winter, the small numbers and infrequent occurrences of these hawks would 282 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 5 tend to minimize their predation pressure on the avifauna. On the other hand, the resident Barn Owl ( Tyto alba) has been shown recently to be a potent predator on birds ( Johnston 1972), even though its numbers are low. The regurgitated pellets of these owls taken from five widely scattered sites on Grand Cayman, and quite likely representative of five different individuals, revealed a high proportion of avian remains among their prey, ranging in size from a medium-sized egret ( Egretta) to a honeyereeper ( Coereba). Furthermore, the prey items included at least eight genera of the resident birds. These data suggest that of the actual or potential vertebrate predators on the avifauna only Tyto alba is espe- cially important, and I suspect this predator represents the greatest single biological control of the resident avifauna. The ubiquitous hermit and land crabs are clearly potential scavengers or predators of any terrestrial animal life. Only the Nighthawk ( Chor- deiles minor) is a consistent ground-nesting bird on the Caymans, how- ever, and because this species is scarce and breeds only in bare Belds where crabs are also scarce, it appears highly unlikely that even this bird, its eggs, or young would be preyed upon by the crabs. Interest- ingly, where land crabs are most abundant on the islands ( rocky barren roadsides and limestone forests), no ground-nesting birds are known, unless it is a very occasional dove ( Columbina passerina or Zenaida aurita). The widespread and abundant Coereba #aueola is conspicuous throughout the Caymans and elsewhere in its range. In most West In- dian island avifaunas, it is the most abundant terrestrial resident species. Controlling mechanisms for this species were documented by Gross ( 1958) and include bird predators ( Quiscalus, Crotophaga) as well as ants and lizards. In the absence of concrete evidence, we can only as- sume that similar predators prey on Coereba in the Cayman Islands. Coereba is an occasional victim of the uncommon Barn Owl ( Appendix III). The extent to which other biological and physical factors exercise any control on population size of birds is as yet unassessed. For example, for the Caymanian avifauna no quantitative data are available on annual population fluctuations, clutch size, hatching success, fledging success, or, indeed, natality rate for any bird species especially in comparison with other insular or mainland populations. Two species are increasing in numbers and range in the Caymans, Mimus polyglottos and Zenaida asiatica. Conversely, the only documented recent losses there are of Mimocichla ravida and Icterus leucopteryx bairdi; causative factors bear- ing on their extinctions have been discussed in an earlier section. 1975 JOHNSTON: CAYMAN ISLAND AVIFAUNA 283 A G~NERAL ASSESSMENT OF FEEDING ECOLOGIES AND HABITAT DISTRIBUTIONS It is instructive now to generalize and summarize data on feeding ecologies or feeding niches of the resident terrestrial avifauna. As shown in Table 17, most of the bird species ( 57 % ) are arboreal foragers, either on arthropods or fruits. Furthermore, with the expected exception of pastures and cultivated areas, these same species are each widely dis- tributed in the seven upland formations and have comparable frequency of occurrence and relative abundance therein. Conversely, ground-feed- ing forms ( 21 % ) are relatively less common as a whole in the Caymans and are virtually restricted to the non-forested formations. The latter distribution strongly suggests the scarcity of insects and seeds on the ground in the several forest formations. Indeed, ground cover is notice- ably scarce in the several forest formations due to a combination of factors ( edaphic karst topography, light penetration). The relatively high proportion of arboreal insectivores and frugivores, augmented by the timber-probing and -gleaning woodpeckers would be expected on islands, such as the Caymans, where forest formations of several types predominate in areal extent ( see Figs. 1-3). Bird species diversity is a known function of habitat or plant species diversity. Although plant species diversity was not determined quantita- tively in the present study, it is nonetheless empirically true that in the comparisons of avifaunal compositions of the several ecological forma- tions, bird species diversity does attain its peak in those forest formations with the greatest plant diversity. Complexity of floristic composition ( presence of many codominant trees vs. one or two dominants, height of vegetation, some semblance of stratifcation, etc.) may be sequentially arranged in descending order ( number of bird species in parentheses) as follows: towns and house sites ( 20), limestone forests ( 17), logwood- thatch palm-red birch forest ( 15), sea grape-almond woodland ( 16), mangrove swamps ( 11), pure logwood forests ( 11), and pastures and cultivated areas (11). A relationship between number of terrestrial bird species and habitat complexity is implied by these data. TAXON CYCLES A subject of recent interest to some island biogeographers is that of taxon cycles, and for the West Indian avifauna these have been explored generally and in detail by Ricklefs and Cox ( 1972). Whether an anal- ysis of taxon cycles for individual small islands or archipelagos is mean- ingful ( see Rickiefs 1970, for Jamaica) is a moot point because of great areal and relief differences in large island groupings, but an attempt at 284 B U LLE T IN FLO R ID A S TATE M U S E U M Vol. 19, N o. 5 TABLE 17.-FEEDING ECOLOGIES AND HABITAT DISTRIBUTIONS OF RESIDENT TERRESTRIAL SPECIES ON GRAND CAYMAN. Total Pastures Towns Sea Logwood- Species & and Grape- Pure Thatch Palm- Lime- Feeding Avail- Cultivated House Almond Logwood Red Birch stone Mangrove Ecology able Aseas sites Woodland Forests Forest Forests Swamps Predator on vertebrates 1 1 1 Aerial Insectivore 2 1(3)° 1(1) 1(1) Foliage feeding-nectar 1 1(4) 1(4) 1(4) 1(3) 1(4) 1(3) Foliage feeding-arthropods 8 2(3) 6(10) 4(8) 6(14) 7(15) 7(11) 4(10) C chiefly insects) Foliage feeding-fruits/seeds 8 2(3) 5(11) 5(11) 1(1) 5(8) 7(13) 4(7) Timber probing-insects 2 2(2) 2(2) 1(1) 2(3) 2(4) 2(2) Ground feeding-arthropods 3 3(9) 2(6) 1(1) 1(2) (chiefly insects) Ground feeding- other invertebrates 1 1(1) Ground feeding-seeds 2 2(4) 2(5) 1(1) 1(1) °Figure in parentheses = sum of number of species X total abundance scores (see footnote of Table 18) 1975 JOHNSTON: CAYMAN ISLAND AVIFAUNA 285 such an analysis for the three Cayman Islands has been undertaken, a summary of which appears in Table 18. There appears, first of all, to be only a weak correlation between feeding ecology and early stages of the taxon cycle, namely that the presumed recent colonizers ( Stage I) are mainly either frugivores or omnivores. On the other hand, other char- acteristically frugivorous birds, plus some insectivores, have also been classified as Stage IV forms. As a rule, specific feeding ecology cannot be closely correlated with a stage in the taxon cycle, at least for the Cayman Island avifauna. There is, unfortunately, no available informa- tion on food habits of Mimocichla rauida and Icterus leucopteryx, the two recently extinct birds from these islands, although it is certainly true that if this study had been made 40 or more years ago, both of these birds would have been in Stage IV using the criteria of Ricklefs ( 1970). Both were known, for example, to be quite restricted in habitat prefer- ences and areal distribution on Grand Cayman. Some problems arise in subjectively assigning a given bird species to a given stage in the taxon cycle, especially as regards Stage IV. Ricklefs ( 1970: 475) gives this definition: "The last stage of the cycle, before going extinct or recycling, is the endemic population. As the cycle pro- gresses, populations move from marginal coastal habitats into more central and montane habitats." Similarly, Ricklefs and Cox ( 1972: 195) note: "Finally, descendant populations, often subspecifically or specifi- cally distinct, are restricted to small refugia." The problems in diagnosing taxon cycle stages for the Cayman Island avifauna condense to two major points. First, as described previously, these islands have neither montane habitats nor discreet small refugia, unless in the latter case one considers the mature, more-or-less inland limestone forests. Second, these three islands currently have no endemic species of birds, but, according to Bond ( 1956) and Johnston et al. ( 1971), do contain 13 species with endemic subspecies. It is of interest to recall that earlier taxonomists ( for example, Cory 1886) regarded seven of these as distinct species. Based upon the contemporary belief that these ( see Table 18) are all well-marked, distinct subspecies, endemic to one or more of the three islands, I am tentatively assigning the 13 forms to Stage IV in the taxon cycle. Their occurrences on the individual islands were outlined earlier by Johnston et al. ( 1971). Taxon cycles for the Caymanian avifauna may be compared, albeit with some qualifcation, with those of Jamaican birds ( Ricklefs 1970) and Solomon Island birds ( Greenslade 1968). Although the Cayman Islands currently contain no endemic species, Jamaica does have 26 such species, many of which (a) occur in the interior montane forests and (b) are in Stage IV. Ricklefs ( 1970) gave 14 as the number of Stage IV endemics, 286 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 5 TABLE 18.-FEEDING ECOLOGIES, STAGES IN TAXON CYCLE, AND ABUNDANCE OF RESIDENT TERRESTRIAL SPECIES. Feeding Stage in Total t Average No. of§ Species Ecologyt Taxon Cycle Abundance Abundance habitats Columba teucocephala FF I 8 2.0 4 Zenaida aurita FF I 4 1.3 3 Zenaida asiatica FF (?) I 6 3.0 3 Cotumbina passerina CS II 5 2.5 2 Leptotila famaicensis FF IV 1 1.0 1 Amazona leucocephala FF IV 2 2.0 1 Coccyzus minor FI,CIR) II 2 1.0 2 Crotophaga ani GI,GV,GS I 6 3.0 2 Chordeiles minor AI I 1 1.0 1 Colaptes auratus TPI,FF IV 8 1.3 6 Centurus superciliaris TPI,FF,TV IV 5 1.7 3 Turannus dominicensis AI I 7 2.3 3 Tgrannus caudifasciatus FI,TV,FF IV 4 1.3 3 Mviarchus stotidus FI,FF III 8 1.6 5 Elaenia martinica FF,FI IV 17 3.4 5 Mimus polt,glottos GI,GO,GS I 10 2.5 4 Mimocichla plumbea unknown IV 2 2.0 1 Vireo crassirostris FI,FF II 6 1 .5 4 Vireo altiloquus unknown I 2 2.0 1 Vireo magister FI,FF IV 12 2.4 5 Dendroica petechia FI II 10 2.0 5 Dendroica oitettina FI,FF IV 12 3.0 4 Coereba #aveola FN,FI IV 22 3.7 6 Spindalis zena FF IV 5 1.7 3 Quiscalus nigel FI,TV,FF IV 13 2.2 6 Tiaris oliuacea CS II 5 1.7 3 Melophurrha nigra FF,FI IV 5 1 .7 3 f Adapted from Salt ( 1953, 1957) and Orians ( 1969): AI, aerial insectivore; PV, predator on vertebrates; PI, predator on invertebrates; FN, foliage feeding on nectar; FI, foliage feeding on insects and spiders; FF, foliage feeding on fruits ( and/or seeds); TPI, timber probing for insects; TV, timber gleaning for verte- brates; GV, ground feeding on vertebrates; GI, ground feeding on insects and spiders; GO, ground feeding on other invertebrates ( crustacea, mollusks); GS, ground feeding on seeds. When two or more feeding ecologies are given for a bird, the predominant type is listed first. t Scored by conversion of U ( uncommon ) - 1, FC ( fairly common) = 2, C ( com- mon) = 3, VC ( very common) = 4. See individual census tables for each habitat concerned. § Seven habitats available during breeding season. Stage of Taxon Cycle I II III IV Number of species and subspecies 8 5 1 13 Mean number of habitats occupied 2.6 3.2 5.0 3.6 Mean total abundance 5.5 5.6 8.0 8.4 Mean of the average local abundance 2.1 1.7 1.6 2.1 1975 JOHNSTON: CAYMAN ISLAND AVIFAUNA 287 but his censuses did not include the endemic columbids, parrots, cuckoos, and hummingbirds. Ricklefs' 14 Stage IV forms, as compared with the Caymanian 13, were less widespread in habitat occupancy and less abundant than Stage I types. The converse is true for the Cayman birds ( Table 18). Probably these differences are at least partially explicable in terms of the likelihood that the Jamaican avifauna is older than that of the Cayman Islands, has available well-developed refugia in the montane forests, and occupies a much larger island with greater relief and habitat diversity. An interesting ecological question is whether or not immigrating and hence colonizing species on islands are found chiefly in "marginal habi- tats." Because so many factors potentially determine the success of colonization ( competition, adaptability of the propagule, niche avail- ability, etc.), it is important to speculate on, or document, an ecological habitat that fts the needs of most immigrating forms. Another way of examining this problem is to assess the habitat distribution of species in Stage I of a taxon cycle. Do they, indeed, "occupy marginal habitats at the center of their expansion... [and] exhibit 'ecological release' on small islands due to lack of competition" ( Ricklefs 1970) ? For Solo- mon Island birds, Greenslade ( 1968) concluded that "the habitats of the species suggest that expansion occurs mainly in coastal situations while the rest of the cycle involves a shift to lowland rain-forest and, increas- ingly in the final stage, to montane forest. The Cayman Islands have neither rain forests nor montane forests. In terms of secondary succession they do have fields and pastures, early logwood forests, logwood-thatch palm-birch forests, and a ( presumed) climax limestone forest. If one can equate the fields and pastures ( many of which are at least near the coast) with marginal or coastal habitats as used by Greenslade, Ricklefs, and others, then the distribution of breeding birds of each taxon cycle stage in this and later successional stages should provide answers to the questions posed above. From data in Tables 12, 17, and 18 it can be determined that 50 % of the Stage I birds breed in the Belds and pastures and the other 50 % occur in the later forested seral stages. All of the Stage II birds are restricted to early seral stages, and at least 75% of the Stage IV types are in the later seral stages. Thus even in the absence of montane forest refugia on the Cayman Islands, the ecological distributions of the resident avifauna partially support the idea that the most recent colonizing species are found primarily in so-called marginal habitats, and the "older" en- demics are more characteristic of interior forests. On the other hand, colonization of these islands has yet a different facet. What considerations should be given to the presently forest- 288 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 5 adapted Stage IV species such as the Centurus woodpecker, Amazona parrot, and vireos? At the time that each form invaded these islands, did each undergo a taxon cycle beginning in a marginal habitat and ultimately become associated with the mature limestone forest? Con- temporary data do not support such a possibility. Rather David Lack ( in lit) appears to have a more plausible explanation to the effect that potential colonists on nearby islands would have to be best adapted to conditions on the Cayman Islands. For example, the Amazona parrot on the Caymans is especially characteristic of limestone forests ( Table 8); on Cuba its conspecific is found in remote woodlands from mountains to sea level. Elaenia martinica, so widespread and abundant in the ~ Caymans, is adapted to arid lowland woodlands elsewhere in the Carib- bean region. Mimocichla plumbea of Cayman Brac, Cuba, and Hispani- ola is a bird of forested regions, plantations, and gardens. These and ~ other examples strongly support the contention that colonization of new islands is not necessarily restricted to Stage I species but could be suc- cessful for species in any taxon cycle stage, provided they were pre- adapted to and could find suitable ecological conditions on a new island. HABITAT DISTRIBUTIONS OF INSULAR AND MAINLAND BIRD POPULATIONS The extent to which breeding bird species occupy all habitats avail- able to them has been discussed at least in part by MacArthur and Wilson ( 1967, Chap. 5). Clearly, some combination of competition, predators, immigration, vegetational complexity, and other factors play significant roles in restricting bird species to given habitats. In some insular avifaunas that have been investigated intensively ( e.g., those of Puerto Rico) individual bird species have undergone an ecological ex- pansion into many habitats, whereas species in the Panamanian tropical forests are much more restricted to a small number of habitats. As a rule, if an avifauna can specifically occupy all or most of the habitats avaiI- able, it can be described as eurytypic or "generalized; if the species occupies only a small number of available habitats, it can be considered stenotypic or "specialized." Some interesting figures emerge from the habitat distribution of the Grand Cayman avifauna. Considering seven upland terrestrial forma- tions ( Tables 3-10 ), the distributions of the 24 species occurring therein ( exclusive of the wide-ranging Barn Owl and Nighthawk) are as follows: 1 species, in only 1 formation 5 species, in only 2 formations 5 species, in only 3 formations 5 species, in only 4 formations 1975 JOHNSTON: CAYMAN ISLAND AVIFAUNA 289 3 species, in only 5 formations 4 species, in only 6 formations 1 species, in all 7 formations By extrapolation one arrives at an average figure of 3,9 habitats per spe- cies. Data for mainland populations ( Fig. 19) are in sharp contrast; for example in Georgia the figure is 1.6 habitats per species ( Johnston and Odum 1956) and for Panama about 1.3 ( MacArthur et al. 1966). On Puerto Rico the value is 2.5 ( Recher 1970) and Jamaica 2.9 ( Ricklefs 1970). For Grand Cayman the relatively large number of habitats oc- cupied by each species indicates either reduced interspecic competition per se, the evolution of mechanisms to avoid competition, or both. These properties are discussed at length in the previous section on com- petition. To exemplify these features more generally, a number of published breeding censuses from typical island and mainland populations have been reviewed with respect to number of available habitats occupied by the species. This assessment is graphically summarized in Figure 19. In regions where relatively few habitats are available ( three or four), mainland bird populations tend to be specialized, and as more habitats become available ( Bve or six), the mainland populations tend to become even more specialized. Large islands, such as Puerto Rico and Jamaica, include bird populations that are intermediate between the extremes of generalization and specialization. On small islands ( including Grand Cayman) the bird populations reach peaks in generalization. Grand Cayman is unique in this feature by having the highest known value of mean number of habitats per species, namely 3.9. Lack concluded ( 1969: 207) that ". . . the small numbers of resident bird species on islands are due...to ecological limitations, to which the islands' birds are often specially adapted, and which enable fewer species with broader niches to exclude a greater number of specialists. If I interpret this sentence correctly and if "ecological limitations are chiefly those of habitat diversification, then the graphic presentation of Figure 19 supports his contention only for small islands. In fact, with respect to number of habitats occupied by each species, the larger islands of Puerto Rico and Jamaica, with their large number of available habitats, resemble the mainland populations. Grand Cayman, Bermuda, and St. John each tend to have ( fewer?) species with "broader niches." The extent to which such ecological generalization might "exclude a greater number of specialists is certainly undemonstrable from present evidence. Another, yet different, approach to this subject is Simberloffs species/ genera ratios ( 1970). Similar S/G values for large and small islands would indicate, for example, that competition is at least as intense on the pa. PR JA GC 7-IOAA A - Mich. Me. 6- 00 - N U M BE R H AB IT AT S AV A IL A B LE CR Al. 5- 00 - 4 - ~ Go. A I. 8 / EX TR EM E SP EC IA LI ZA TI O N CD A -\ 4 0 ~ N.C. Ga. SJ OOA Y \ /ietenia mahagoni ( L.) Jacq. West Indian Mahogany. Hippomane mancineUa L. Manchineel. Ceratiola sp Michx. Rosemary. Mangifera indica L. Mango. Comocladia dentata Jacq. Maiden Plum, Muntingia calabura L. Strawberry Tree. Thespesia populnea ( L.) Seaside Mahoe. Hibiscus sp. L. Hibiscus. Clusia #aua Jacq. Balsam. Passi#ora sp. L. Passionflower. Carica papaya L. Papaya Cereus (.L.) Mill. Cactus. Rhizophora mangle L. Red Mangrove. Conocarpus erecta L. Buttonwood. Terminalia catappa L. AImond. Laguncularia racemosa Gaertn. F. White Mangrove. Cal!/ptranthes pallens Griseb. Pale Lidflower, Psidium guaiaua L. Guava. Eugenia sp. L. "Strawberry". Myrtus sp. L. Stopper. Ardisia escallonioides Schlecht & Cham. Marlberry. Manilkam zapoda ( L.) Naseberry. Nerium oleander L. Oleander. Ipomoea sp. L. Morning Glory. Tournefortia gnaphalodes (L.) B. Br. Sea Lavender. Avicennia nitida Jacq. Black Mangrove. Cirsium sp. Mill. Thistle. 1 Nomenclature from Swabey and Lewis ( 1946), Asprey and Robbins ( 1953), or Long and Lakela ( 1971). 1975 JOHNSTON: CAYMAN ISLAND AVIFAUNA 299 APPENDIx III: STOMACH CONTENTS OF CAYMAN IsLAND BIRDS The birds utilized in these analyses were collected in either April or August, chiefiy from Grand Cayman. The numbers in parentheses represent the numbers of bird stomachs examined, A= animal food, V- plant food, and an asterisk (®) in- dicates a predominant food item. When more than one stomach was examined, the percentages for A and V are mean values. Nomenclature for the insect items follows that of Borror and DeLong ( 1964) and for plant foods, that of Asprey and Robbins ( 1953) or Long and Lakela ( 1971). Butorides uirescens (2). A, 100% Isopoda; Odonata, Libellulidae; Orthoptera, Acrididae, Cyrtacanthacridinae; Hemiptera, Belostomatidae, Belostoma (?)*; Co- leoptera, Carabidae*, Dytiscidae. Porzana carolina (1). V, 100% Sedge seeds. Columba Zeucocephala (1).V, 100%. EIaeocarpaceae, Muntingia° . Zenaida aurita (1).V, 100% . Cyperaceae, Scleria*, Rhynchosporao; Poaceae, Panicum; legumes; Empetraceae, Ceratiola, Myrtaceae, Eugenia; Asteraceae, Cirsium. Columbina passerina (2). V, 100%. Cyperaceae, Scleriao; Commelinaceae, Com- melina°; Asteraceae, Cirsium. Leptotita iamnicensis ( 1). V, 100%. Cyperaceae, Scleria; Convolvulaceae, Ipomoea; legumes; Elaeocarpaceae, Muntingia; Rosaceae, Prunus; Empetraceae, Ceratiola. Amazona leucocephala (2). V, 100%. Caricaceae, Carica. Cfotophaga ani (3). A, 87%,V, 13%. Arachnida, Araneida; Orthoptera, Acrididae, Cyrtacanthacridinae; Odonata, Coenagrionidae( ?), Aeshnidae( ?); Lepidoptera, Noctuidae, Nymphalidae( ?); Coleoptera, Scarabaeidae, Chrysomelidae( ?), Cur- culionidae°, Tenebrionidae; Homoptera, Cicadellidae; Hemiptera, Pentatomidae, Coreidae( ?), Scutelleridae®, Chelysoma; Diptera head; Hymenoptera, Vespidae*. Vertebrate: Anolis conspersus. Sedge seeds ; Passifloraceae, Passiflora. Tuto alba ( ca. 50 pellets). Invertebrate- Paguridae, Coenobita culpeatus, Coleoptera; Reptilia- Aristelliger praesignis, Aves-Egretta thula, Centurus superciliaris, Etaenia martinica, Mimus polyglottos, Dumetella carolinensis, Dendroica sp.?, Coereba flaveola, Quiscalus niger, Melopyrrha nigra; Mammalia-Rattus rattus*, R. noruegicus, Mus musculus*, Bracht/Pht/lla nana, Artibeus iamaicensis. Colaptes auratus ( 19). A, 97%; V, 3%. Arachnida, Araneida; Coleoptera, Cer- ambycidae, Bostrichidae (?); Hymenoptera, Formicidae, Formica°, Isoptera, Rhinotermitidae°. Unidentified fruit fragments. Centurus supercitiaris ( 16). A, 56%; V, 44%. Arachnida, Araneida, Sevtodes fusca; Orthoptera, Gryllidae, Acrididae*; Coleoptera, Curculionidae, Otior- hynchinae, Tenebrionidae; Hymenoptera, Formicidae, Vespidae (?). Hyla septentrionatis, Sphaerodactylus lewisi. Moraceae, Ficus; Caricaceae, Carica*. Tgrannus dominicensis (1).A, 100%. Homoptera, Cicadidae'. Turannus caudifasciatus (5). A, 96%; V, 4%. Orthroptera, Gryllidae; Coleoptera, Curculionidae*, Carabidae, Buprestidae, Pol!/cesta; Hymenoptera, Vespidae, Vespula(?), Formicidae, Myrmicinae. Hyla septentrionalis, Anolis conspersus°. Burseraceae, Bursera simaruba; Passifloraceae, Passi#ofa. Mviarchus stolidus (4).A, 89% ;V, 11%. Arachnida, Araneida; Orthoptera, Gryllidae; Lepidoptera; Coleoptera, Cerambycidae°, Oedemeridae'; Hemiptera, Scutelleridae, Chelysoma( ?)*; Hymenoptera, Tiphiidae( ?), Andrenidae, Colleti- Jae, Colletes, Vespidae (?). Burseraceae, Bunera simamba; unidentified seeds. 300 BULLETIN FLORIDA STATE MUSEUM Vol. 19, No. 5 Elaenia martinica (9).A,6%;V, 94%. Coleoptera, Curculionidae; Homoptera, Flatidae. Moraceae, Ficus°; Burseraceae, Bursefa simamba; Passifloraceae, Pass#lora; many unidentified seeds°. Mimus polyglottos (1). A, 87% ;V, 13%. Crustacea, crab parts, Isopoda; Cole- optera, Scarabaeidae, Dyscinetus( ? ); Hemiptera, Pentatomidae: Hymenoptera, Formicidae, Solenopsis. Unidentified seeds. Vireo crassirostris (11). 4 77%; V, 23%. Arachnida, Araneidao, Orthoptera; Lepi- doptera larva°, Microlepidoptera, Geometridae; Neuroptera, Chrysopidae; Cole- optera®, Curculionidaeo, Oedemeridae, Cerambycidae: Staphylinidae( ?), Ela- teridae( ?), Chrysomelidae; Hymenoptera, Formicidae, Vespidae; Hemiptera, Pentatomidae, Phymatidae. Burseraceae, Bursera simaruba; many unidentifiable seeds. Vireo magister (9). A, 51%;V, 49%. Arachnida, Araneida; Orthoptera, Gryllidae; Lepidoptera larvae®, Geometridae; Neuroptera, Chrysopidae; Coleoptera', Ten- ebrionidae, Chrysomelidae, Curculionidaeo, Cerambycidae,. Scarabaeidae( ?); Oedemeridae; Hemiptera, Pentatomidae; Homoptera, Flatidae; Homoptera leg; Hymenoptera, Ichneumonidae, Formicidae, Cryptocerus. Burseraceae, Bursera simaruba; many unidentified seeds. Dendroica petechia (3).A, 100%. Coleoptera°, Curculionidae, Oedemeridae°; Hymenoptera, Formicidae. Dendroica oitellina (7).A, 97% ;V, 3%. Arachnida, Araneida°; Lepidoptera larvae; Orthoptera, Gryllidae( ?); Coleoptera, Curculionidae°; Homoptera°, Aphidae, Fulgoridaeo, Ceropidae; Diptera, Muscidae; Hymenoptera, Formi- cidae. Few unidentified seeds. Coereba flaveola (6). A, 50% ; V, 50%. Lepidoptera larvae°; Coleoptera, Chry- somelidae, Chalticinae, Curculionidae; Homoptera, Cercopidae (?); Hymenoptera, Chalcidoidea. Unidentified flower parts. Spindalis zena (4). V, 100%. Myrsinaceae, Ardisia; many unidentified seeds. Quiscalus niger (8). A, 87% ;V, 13% . Arachnida, Araneida; Orthoptera, Tetrigi- dae, Gryllidae(?), Blattidae; Lepidoptera, Noctuidae; Coleoptera, Carabidaeo, Oedemeridae, Curculionidae°, Alleculidae; Homoptera, Cicadidae; Hemiptera, Pentatomidae, Scutelleridae (?), Hymenoptera, Formicidae; Diptera, Culicidae H!/la septentrionalis°, Anolis conspersus°. Moraceae, Ficus; unidentified seeds and plant fragments. Tiaris olivacea (3). V, 100% . Poaceae, Panicum°; many unidentified seeds. Melop!/rrha nigra (4).A, 39%; V, 61% . Lepidoptera larvae*; Coleoptera®; Chrysomelidae(?). Unidentified seeds. Contributions to the BULLETIN OF THE FLORIDA STATE MUSEUM, BI- OLOGICAL SCIENCES SERIES, may be in any field of biology. Manuscripts deal- ing with natural history of systematic problems involving the southeastern United States or the New World tropics are solicited especially. Manuscripts should be of medium length-irca 35 to 200 pages (10,500-16,000 words). Examination for suitability is made by an Editorial Board. The BULLETIN is distributed worldwide through institutional subscriptions and exchanges. It is considered the responsibility of the author to distribute his paper to ali interested individuals. To aid in this the author( s) receives( s) 50 copies free, and he may purchase additional separates at cost if ordered when page proof is re- turned. The author is also responsible for any charges incurred for alterations made by him on galley or page proofs. The Museum will send an invoice to the author for this amount upon completion of publication. PREPARATION OF MANUSCRIPT Contributors should consult recent numbers of the BULLETIN for preferred style and format. Highly recommended as a guide is the CBE Style Manual, 3rd Edition, 1972 ( Washington, D.C., Amer. Inst. Biol. Sci.). MSS must be submitted in duplicate ( please no onionskin) and satisfy the following minimal requirements: They should be typewritten, double-spaced (especialty tables, figure captions, and "Literature Cited'*), on one side of numbered sheets of standard (8-1/2xll in.) bond paper, with at least one-inch margins all around, Tables ( which should be unruled) and figure legends should be typed on separate sheets. All illustrations are referred to as figures. They must comply with the following standards: Photographs should be sharp, with good contrast, and printed on glossy paper. Drawings should be made with dense black waterproof ink on quality paper or illustration board. All illustrations should have a cover sheet, All lettering will be medium weight, sans-serif type ( e.g. Futura Medium, News Gothic) in cutout, dry transfer or lettering guide letters. Make allowance so that after reduction no lowercase letter will be less than 1 mm high ( 2 mm is preferred) nor any capital letter greater than 5 mm high. The maximum size for illustrations is 8-5/8x14 in. ( twice typepage size); illustrations should not be less than typepage width ( 4-5/16 in.). Designate the top of each illustration and identify on the back with soft pencil by author's name, MS title, and figure number. Manuscripts and alt editorial matters should be addressed to: Managing Editor of the BULLETIN Florida State Museum Museum Road University of Florida Gainesville FL 32611