LMNG WORLD

Journal of The Trinidad and Tobago Field Naturalists' Club 1997-1998

Natura Maxime Miranda in Minimis

Published March 1998 Living World is published biennially by the Trinidad and Tobago Field Naturalists' Club All rights reserved.

layout by Michael Tikasingh EDITORIAL

As new Editor of Living World I a m conscious of the na expansion following fires in the Aripo Savannas. An enormous effort {he previous Editor, Dr. Victor Quesne l has account on the food and feeding behaviour of the bat Carollia put into t he lournal in making it the s u ccess it has been over perspicillata is given by Victor Quesne l. Migrat or y patterns of the years. He revived t he journal in 1956 and was c lose ly butterflies are difficult to study, so we welcome the observa­ associated with the production of most of t he issues published tions made by Dave C h adee o n Plloebis statira. Matthew Cock so far. T here were other EdilOrs of course, notab ly Ri chard cont inues wi th hi s ser ies on the Hesperi id ae b utterfl ies . He ffrench and Peter Bacon. who kept the Journa l goi ng wh i le describes a new species for T r in idad, Clito trinida(lensis. He maintai ning high standards. The new Editor pledges to at least expects to comp lete the series of studies in our next issue and maintain the h igh standar ds set by previous Edi t ors. then begin a series of u pdates o n t h ese butterflies. In thi s issue, we presen t a series of artic les by Ri chard Our next issue, 1999 . 2000, might well be a special issue ffrench, one giving a brief h i story of o rn ithol ogy in Trin idad and it is hoped that we wi ll have contribut ions from more and the others updating the stat us of certain bird species in member s of the Clu b. Deadli n e for the nex t issue is December T r inid ad a n d Tobago. I a n L ambie and V ish nu De b ie prov ided I , 1998. the data o n bird counts fo r the years 1976 to 198 1. Pau l Comeau and C.C. C l ubbe give an interest ing account of savan- ES T

THE TRINIDAD AND TOBAGO FIELD NATURALISTS' CLUB T h e Tri n idad and Tobago Fiel d Natu ralists' C l ub was excursions are h eld on t h e last S u nday of every mon th founded on 10th J u l y 189 1. Its name was i nco rpo rated by exce p t Decemb er. a n Act o f P arli a m e nt (Ac t NO.17 o f 199 1). T h e obj ects o f t h e Members h ip is open [ 0 al l pe rso n s of at l east fi ftee n C lu b are t o bri n g togeth e r person s in t erested i n th e st u dy of year s of age, w h o subscribe to t he obj ects of the Club. Datural hi story, the d i ffu s io n of k n ow ledge thereof a n d the A ll enqu i r i es co n cer nin g t h e Club or its 1 0u rn a l shoul d b e conservati on o f natu re. addressed t o the Hon orary Sec retary, P .O. Box 642, P o rt o f Month ly lecture meetings are hel d at S l. Mary's Spai n, Republ i c of Tri n idad a n d Tobago Coll ege, on the second Thursday of every m onth, w hile fiel d e m ail: [email protected]

M an age men t C ommitt ee 1 997 1 9 98

Pre s id e nt Ed wo u d H eester man Carra I A lexande r V i c ePr es ident She ld on E d wards L a u ren t de Verteu i l Honorar y Sec r e tar y Courte n ay Rooks Courtenay Rooks Honorary A s s i s tant Secret a ry L a ure nt d e Verteuil Cal ista Pie rre T re as urer Selwyn Gomes Selwy n Gomes Co mmitt ee Me mb e r s Vi cky Banchard G r aham Wh i te Detta Buc h S h eldon Edwards D an lagge r nau t h Dan Jaggern au lh

E ditori a l C ommitt ee E l isha Tikasingh (Ed itor), Victor Quesnel, Paul Comeau

Cov er Ph o to: M arginal veget atio n a t Aripo Savanna V. (See P.23) P hoto by Paul Comeau

ACKNOWLEDGEMENTS

Apprec iati on i s extended to t he T ri nidad P ubli s hin g Co. Lt d. for their assistance in the prod uction o f the negat ives used i n thi s issue of the Journal , and to our proof reader s Cal ista Pierre, Rupert Mends a nd Laurent de Verteu il. A REVIEW OF THE ORNITHOLOGY OF TRINIDAD AND TOBAGO 1950-1985 Richard ffrench Toftingal Laurieston Road, Gatehouse of Fleet, Castle Douglas, Scotland DG7 2BE

Introduction In presenti ng this historical review of the progress of The first book to attempt to remedy th is situation was ornithology on Trin idad and Tobago over a 35-year pe riod , I produced locally in 1950, not by an ornitho logist but b y an am conscious that I may have omitted mentio n of work elderly Catholic priest, Father Raymund Devas, who lived in unknown to me, or which I have been unable to remember. If Grenada. It cont ai ned virtua ll y no illustrations, was very so, I apologise to those concerned. Nevertheless, I was res i­ short and fairly inaccurate, but it did contain an a l most com­ dent on Trinidad for nearl y 30 of those years and actively plete list of species and it was a start. A s imilar attempt to in volved in ornithology, so I me t, corresponded or worked en li ghten interested visitors was made by Mrs. K. Alford with with the majority of those mentioned be low . I hope 1 have notes on birds in a Tobago guide-book at about thi s time. d o ne them justice. Popular articles illustrated by photographs and very inade­ quate paintings appeared in oil company magazi nes written The Fifties by J o hnson ( 1956-57) and Saunders ( 1956, 1957), and th ese In common with most neolropical cou ntri es at the mid­ may have helped to arouse the latent interest in birds amo ng century, Trinidad and Tobago cou ld produ ce little evidence of the people of Trinidad and Tobago. Two indi vidua ls, R.E. ornithological aeti vi ty up to 1950 other than the wo rk of col­ Johnson ( 1937) a nd E. Chenery, had indeed studi e d local lectors. Even the extraordinarily detailed and valuable work birds s ince the 1930s, but there was little publis hed work. The of Belcher and Smooker (1934- 19 37) was based so le ly on revival of the Trinidad a nd Tobago Field Naturalists' Club for~ the ir collection of nests and eggs. The age of the collector was tunately led to publication of a Club journal at varyin g inter­ indeed not quite over, since sizeable collections of bird speci­ va ls from 1956 up to 1965 , and ever y two years after that. mens were made on Trinidad and Tobago during 1950-1951 This led to early contributions o n birds from Chenery ( 1956) by A. Ra sool and R. Plowden-Wardlaw for the Yale Peabody on nes ting tou cans, and by Quesnel (1956) on the den si ty of Museum in Massachusetts, and by G. F . Mees in 1953-54 for Great Kiskadees in Port of Spain. the Leiden Museum in Holland . During this decade a consid­ It is perhaps not s urpri sing, therefore, to find that the e rable number of bird skin s were also obtained by Wilbur major contributions to ornithological science at thi s time Downs and Thomas Aitken in the course of their work at the emanated from expatriates, mainly Briti sh, for Trinidad and Trinidad Regional Virus Laboratory (TRVL), a nd deposited in Tobago was th e n s till a co lony, or North American. First to the collect ion now held by the Caribbean Epidemiology appear was a useful volume by Junge and Mees ( 195 8) giving Centre (CAREC) in Port of Spain. This s hould form the basis a n account of the taxonomic work arising from the laller's col­ of a national collection, if s uch a project is ever contemplat­ lections five years earl ier. Brief descriptions of all s pecies ed. were included, but there were no illu stration s. Next came a Apart from Leotaud's work of 1866. written in French book from Geoffrey Herklots, principal of The Imperial and now virtually unobtainable, the re was up to 1950 no College of Tropica l Agriculture (lCTA) which preceded The descriptive account of the birds of Trinidad a nd Tobago. University of the West Indies (U.W.I.) at St Augustine. Dr. James Bond's monumental study of West Indian birds, begun Herklots was a British plant phy sio logist with an inte rest also in the 1920s, pointedly drew a line between the Antilles and in birds a nd painting. He attempted to fill the need for an Tobago, in acknowledgement of the fact that the avifauna of illustrated fi e ld g uide of local birds , but unfortunately the Trinidad a nd Tobago were basically continenta l in nature. work ( 1961) was not a success . The descriptions were far too rather than oceanic, and his 1936 book omitted the birds of lengthy and detailed. being clearly based on specimens in the these islands. Thus anyone living in or vis iting Trinidad to hand, and so were in many cases useless for field wo rk; worse study birds could identify s pecies only by using books on the st ill , the author's ow n paintings were too often grotesque a nd birds of British Guiana, now Guyana, the Antilles, Panama, inaccurate, usually depic tin g only the anterior half of the Mexico or even the U.S.A.; or alternatively, by relying on bird, and were well below the s tandard by now to be expect­ museum collections mos t of which were to be found in the ed in an international market. Herklots' field experience also metropolitan countries of the north. Naturally, this restric­ was far too limited to e nable him to provide adequate tre at­ tion tended to deter resi dents of Trinidad and Tobago from ment for the majority of the species accounts. taking an interest in orn ithology, for the basi s of most s tudies Probably th e most s ignifi cant development for loca l depends on identification. ornithology during thi s decade came abo ut from the deci sion 4 Living World 1997-1 998 of the famous American naturali st and ex plorer William Beebe Meanwhile, there had been brief visits to Trinidad and to set up a tTopical field research station for the New York Tobago by sc ientists with specific interes ts, which re sulted in Zoological Soc iety a l hi s properly, "Simla", in the Arima publications in the international sc ie ntific press; these includ­ Valley. Founded in 1950, it was to last over twenty years, and ed Darnton (1958) on manakins, Gilliard (1958) on biq:ls-of­ served main ly to attract researchers to s tudy tropical biology paradise in Tobago, Tashian (1957) on oropendolas, and in a setting that benefited from convenient access and mod­ Gross (1958) on the s leeping habits of Bananaquits. During e rn facilitie s rarely avai lable to field workers elsewhere in the the 1960s the Simla s tation facilitated important research by neotropies at that time. Although Dr. Beebe was already of Collins ( 1968-1974), mai nly on various swi ft s, Lill (1974) on advanced years and produced litt le sig nificant work him self evol utionary aspects of manakin courting behaviour, and at Simla. his paper on the ecology of the Arima Valley ( 1952) Drury (1962) , on oropendola ne s tbuilding. Independent provided a use fu l base for fu ture research. research led to the publication of papers by other workers, A few years laler, Beebe 's far- s ightedness was rewarded such as Lanyon (1963), on the obscure Myiarchus flycatchers, when Simla, and its neighbouring estate of Sl. Pat's, became Gochfeld (1972,1973) on marsh land and unusual species, and the base for research conducted over five years by David and Nottebohm (1969), on parrot learning behaviour. Barbara Snow, professional ornithologists from Britain, who On the more popular front, two i mportant books appeat­ soon became, along w ith Alexander Skutch in Costa Rica, so me ed which certainly put Trinidad and Tobago into the limelight of the most outstanding persona lities in the world of neotrop­ to some extent. These were the work of J an Lindblad, the ical fie ld ornithology. One of the great benefits of the Snows ' Swedish naturalist and photographer, o n the scarlet ibis in situation was that, unlike the majori ty of visiting expatriate Caroni Swamp (1969), and Brooke Worth (1967), a scie nti st researchers , they were able to work over a lengthy and con­ working with Wilbur Downs at the TRVL, whose perceptive ti n uou s period, and thus achieve an indepth understanding of and amusing book on his adventures in Trinidad is a classic the eco logy a nd other biologica l aspects of the area. The prin­ of its kind. cipal subj ects of their studies were the lek-forming manak ins One interesting development followed the destruction of (1962a, 1962c. 1963, 1971), be llbirds (1970) and h umming­ Tobago's fo rests in 1963 by Hurricane Flora. Perhaps typical­ birds (1972, 1973, 1974) as well as the o ilb ird (1961, 1962), l y, the government of the day was less interested in the almost certain swifts (1962b) and th rus hes (1963b); a rising fro m the total extinction of severa l hill fores t species there, including above came continued work on the significance of fruit-eating the White-tailed Sabrewi ng, than in the effect on the islands in tropical birds and its connection with the evolution of lek tourist trade of the similarly near annihilation of the intro­ be hav iour (1962a, 1962c). The Snows' work a lso embraced duced Greater Bird-of-Paradise on Lillie Tobago Island. many other branches of ornithology, and provided extremely Sensing an opportu nity, I and so me others were ab le to bring valuable primary data, meticulously collected, on breeding about the establishment on Little Tobago, for one year, of a seasons (1963a, 1964, 1973, 1974), social organization ( 197 1, graduate researcher from Wisconsin, Jim Dinsmore, whose 1973), feeding niches (1972) and ecology ( 197 1), longevity work (1967-1972) not only dealt with the ecology of the ( 1974), and the general biology of species (1963a, 1964, island, and its doubtful suitab ility as a haven for birds-of-par­ t 985). To fUrlh er their ecological studies they also collected ad ise, but also i nc luded the f irst comprehensive account of plant material widely for identificafion purposes, since litera­ the main seab ird islands off north-east Tobago. ture on neot ropicai plants was then as s parse as that on b irds. D uri ng the 1960s I had not been idle. Fo llowi ng the pre­ Although the Snows' residence on Trinidad ended in cepts and example of the Snows, my wife and I began as early 1961,their association w ith neot ropical ornithology has as 1958 to catch and band a variety of species, with part ic u­ end ured for more than three decade s to the present time . by lar emphasis on the Dickcissel (1967a), whose erratic winter many visits to continenta l countries. I myself benefited great­ irruptions into south T rinidad reached a peak at this time. ly from the Snows' generous shari ng of primary data to assist We worked al so on migrant shorebirds, which conveniently my own studies, as well as through s timulating companion­ roosted and fed in great numbers on Pointe-a-Pierre mud-flats sh ip in the field. For a young ornit ho logist starling o ut in the near our home, in the days before the oil company reclaimed neotropics, the opportunity to learn as a kind of apprentice that land for a tank farm. and on the scarlet ibi s (1970), from such experts was a wonderful stroke of l uck. which st ill nes ted in Caroni Swamp up to 1969. We a lso paid a number of visits to offshore islands, principally Soldado The S i x ti es Rock, where we banded and studied Sooty Terns and Brown Since we were all disappointed by the inadequacy of Noddies (1989, 1991a), and the various Bocas island s in the Herklot' s book, it is not surprising that in the earl y 1960s the course of Field Natu ralists ' Club exped itions (1965, 1967b, Snows and I began to talk about a joint project to produce a 1969). publicatio n wOrlhy of the extraordinary avifauna of this coun­ In 1965 1 finally began work on actually writing my try. To begin with, this consisted mainly of amassing data on guide to local birds. David Snow, who left the country in 1961 , the general biology of species, and the understanding was had decided he was too busy elsewhere to participate, but that I would concentrate my efforts on lowland areas, swamps gave much-valued he lp. I was a lso fortunate enough to and coasts, while the Snows would deal with hill forests. acquire the cooperation of Don Eckelberry and John O'Neill, Review Trinidad ornithology 5 twO of the foremost bird artists in the U.S.A., and, c rucially, I had followed up my more general work by concent rat· financial backing for the publication from a friend. 1 must ing o n more spec ific topics, such as the life history of the mention here that at no time d id I aspire to write a field guide pearl kite (1982), the bi rd-life of an abandoned estate at which exists principall y or even so le l y to enable identifica­ G rafton, T obago (1978), a nd on a more s usta ined attempt to tion. Birds are far too interesting, indeed fascinating, to be rel­ bring an interest in history into the li ves of the wider publi c egated to the leve l of names or numbers to be collected by of Trinidad and Tobago through magazine and newspape r those with nothing betler to do. Thus I determined to include articles, and eventua ll y a less cos tly illu strate d booklet on the and present in su mmary form all the info rmation 1 could commo ner birds of the coun try ( 19 86). glean on the 400 odd species known to occur on Trinidad and Moving into the e igh ties, we find conti nuing work on Tobago. It took me eight yea rs to co mplete (1973), and a fur­ migrant seabi rds by a Canadian team led by Blokpoel and ther 18 year s to update (1991 b). Of course we had been col­ Morris (1982, 1984), a useful but far too brief study of th e lecting a wide variet y of data for ten years before I began writ­ impact of chachalacas o n c rop predation in Tobago by ing the text and thi s has cont inu ed right up to the present. Diamond ( 19 33), and so me interesti ng comparative work by The emphasis was on life history material, s uch as vocalisa­ Wunderle on the avifaunas of Tobago and Grenada (1985). In tions, food, breeding and o ther behaviour. From the o utset 1 the later years of the decade. work was forthcoming on rare aimed to provide a reference book which could be useful for or threatened species such as the piping-guan , scarlet ibis and residen ts and visitors on Tri nidad and Tobago, but wh ich other wetland species, but I am reaching beyond the scope of would also provide a basis for field workers in neighbouring this study. countries. Thus I included a wide-ranging bibliography of It cannot have escaped notice that little ornitho logica l some 330 titles, which I had consulted in the course of my research was produced during the period under review by compilation. I believe it was all this information, collated for native-born residents of Trinidad and Tobago. Exceptions to th e first time in a general study of the birds of a neotropical this rule include some members of th e Trinidad and Tobago country, which gave my book its main va lue. Field Natural ists' C lub, such as Elisha Tikasingh and Clyde Although it had a l ways s urprised me tha t few members Crichlow; bu t no one in that Club wou ld Ileed to be reminded of the Field Naturalists' Club had taken much interest in bird that one principal contributor, espec ia ll y in latter days. has study, there had neve rt heless been pl enty of interest in other been that a ll -round naturalis t of the "old schoo !" , ViclOr branches of natural history, and particularly in the conserva­ Quesnel. who followed up his original work from 1956 on th e t ion of wildl ife. This latler concept was probably the pri ncipal Kiskadee by artic les on hummingbirds (1977), ni g ht jars motivator i n the establi shment in 1967 of the Asa Wright ( 19 85), and in a hilarious if controve rsial work on the song of Nature Centre, the fi rst of its kind i n the neotropics, which the peppershrike (1987). was based largely on the initiative of Don Eckelberty, the I said earlier that William Beebe in hi s establishment of American artist, who spear-headed the e ffort s to set it up. Simla laid down an important foundation; so I return finally Shortl y after this, th e New York Zoological Society decided to the Arima valley, and to the other hi gh ly successful venture after the death of William Beebe to give up its field stat io n, of the Asa Wright Nature Centre, which has drawn onitholog­ and the property was eventually taken over to be run on a ica l pilgrims from all parts of the world. This has led to a valu­ limited budget by the Nature Centre. Thus the facilities able tourist asset, and has contributed towards the awakening remained in place to e nable important work to be do ne there of an awareness, through education and public ity, of wild li fe during the 1970s and beyond. By this time th e University of and especially birdlife in this country. In the ten years since the West Indies had estab lished itself securely i n various cam­ I found myself unfortunately con strained to leave my resi­ puses, and support there had been available for visi ting or dence of 30 years in the West Indies, I have seen the contin­ re sident fie ld workers in orni th ology. ued expansion of this process, not only at the Centre, but also in other areas of both islands, and into valuable and praise­ The Seventies and ea rl y E i ghties worthy efforts by a number of individuals a nd organ izations Valuable work was done over a lengthy time by a tea m in the fi elds of eco- tourism , photography, publication and from Florida, led by Peter Feinsinger, mainly on th e subject of research. It is my hope that th e latter field wi ll now begin to the ecological relationships of hummin gbirds and flowering be developed even more widely. plan ts ( 1978, 1982); also by the Wileys (1980) and Mano li s I am proud to have been part of the development of ( 1982) on the breeding strateg ies of certain icterids, by ornithology in Trin idad and Tobago over th e last 40 years . I Morris (1984) o n breeding seab irds on Little Tobago, and by have made many fine friends a nd acquaintances during that Williams and o th e rs on bird migra tion th at had been moni­ time. and I would li ke to express my very si ncere appreciation tored on radar (1977). One of the most useful pieces of for all that they have done with me a long the way. research, fro m my personal poi nt of view, was carried out by Ke e ler-Wolf (1982) o n comparative avian ecology in the hill forests of Trinidad and Tobago, particularly as it related to the recovery of Tobago's hill forest after 1963. 6 Living World 1997·1998

R e f e r e n c e s

B ee b e, W. 1952. I n troduction to the eco logy of t he Arima ffr e n c h , R . P . and M, f f r e n c h . 1978. The birds of Graflon Valley, Trinidad. Zoologica 37: 157·183. Esta te, Tobago. Living World. J . Trinidad Field B e l c her , C . a nd G .D . S m o ok er . 1934-[937. Bi rds of the Natu ralisl Cl ub. p p . 19-24. colony of Trini d ad and Tobago (in 6 parts). Ibis (13)4: ffr e n c h, R.P. 1982. The b reeding of the Pearl Kite in 572-595 el se q. Trinidad. Living Bird 19: 12 1-131. B l okpoe l , H ., R .D. Mo r ris and G.D. Tess i er. 1984. ffre n c h , R . P . 1989. The birds and other vertebrates of F ield investigation s of the biology of Common Terms Soldado Rock. T rinidad. Living World, J. Trinidad Field win te ring in Trinidad. 1. Field Omi/h. 55: 424-434. Naturalists' Club. pp 16-20. C h en c ry, E.M. 1956. The Sul ph ur-and-white-b reasted ff re n c h , R . P . 1991a. Synch ronous bree di ng and mou lt in Toucan. J. Trinidad Field Naturalists' Cluh. pp. 4 - 11. the Brown Noddy tern on Soldado Rock. Tri nidad. Livillg eo II in s, C. T. 1968a. The comparali \Ie biology of two World, J. Trinidad Field Nalllralists' Club. pp. 39-4 1. species of swifts in Trinidad, West Indies. Bull. Fla. Stale f fr enc h , R . P . 199 1b. A Guide to the Birds of Trinidad alld Mus. 11(5): 257-320. Tobago (2nd edition). Cornell Univ. Press, Ithaca, N.Y. Co lli ns, C.T. 1968b. Notes on the biology of Chapman's 426 pp. Swift. Am. Mus. Novit. 2320. 15pp. Gilliard, E. T. 1958. Feathered dance rs of Litl le Tobago. Co ll ins, C . 1974. Su rvival rate of the Chestnut-collared Nat. Geog. Mag. 11 4: 428-440. Swifr.Wesl. Bird Bander 49: 50 - 53. Goc h fc ld , M. 1972. Observations on the status, ecology Dar nto n, 1.1958. The display of t he manakin, Manacus and behavior of Soras wintering in: Trinidad, West manacl/s. Ibis 100: 52-58. Indies . Wi/SOli 81/11.84: 200 - 20 I . Devas, R. 1950. Visitor's book of birds. Trinidad alld Goch f e l d, M. 1973. Observat ions on new or unus ual birds Tobago. Port of Spain. Muir Marshall 76 PP. fiom Trinidad , and comments on the genus Plegadis in Diamo n d, A.W.1983. Censusing and managing cage-birds. Venezuela. Condor 75: 474-478. seabirds, and birds harmful to agriculture i n Trinidad Gross, A. O. 1958. Life history of the Bananaquit of Tobago a nd Tobago. F.A.O. field document FO:TRII79/011. is land. Wilsoll Bull. 70: 257-279. D in s m ore, J . J. 1967. Ecology and behavior of the Greater Her kl ots, G . A.C . 1961. The Birds of Trinidad alld Tobago. Bird-of- paradi se on Little Tobago island. M.S. thesis. London: Collin s. 287 pp. Uni v. of Wisconsin. J o h nson, R.E. 1937. The nest of the Common Poloo. D i ns m ore, J . J . 1972. Avifauna of Littl e Tobago island. Q.J. Jlll/strated Londoll News. March 13. p.436. Fla. Acad. Sci. 35 (I): 55-71. Jo hnson, R . E. 1956-57. Trinidad birds (i n 4 paris). Shell Drury, W . H . 1962. Breeding activit ies, especia l ly nest- Trill. Mag. (various issues). bui lding. of t he Yellowtail OSt;IIOI) S decllmallus in Jun ge, C.C . A. and G.F.Mees. 1958. Th e Al' ijalllla of Trinidad , West I ndies. Zoologica 47: 39-58. Trinidad alld Tobago. Zoologische Verilan(l elingell 37. Fe i ns in ge r , P . and R.K . Co l we ll. 1978. Community Leiden. 172 pp. organization among Neotropica l nec tar- feeding birds. K ee l er-Wo lf, T . H. 1982. Reduced avian dens it ies in Am. Zool. 18: 779-795. Tobago lower montane rain fores t ; a re sou rce-based Fe i ns i nger, P . , J. A. Wolfe and L.A. Swarm . 1982. s t udy. Ph.D. diss. Univ. of California, Santa Cruz. 305 Island ecology: reduced hummingbird diversity and the pp. pollination biology of plants. Trinidad and Tobago. L a n yo n , W . E. 1963 . Experimen ts on discrimination in Ecology 63: 494-506. Myiarchus flycatchers. Alii. MilS. Novit. 2 126. 16 pp. ffre n ch, R.P. 1965. Notes on the avifauna of Grand Fond, Leo t aud, A. 1866. Oiseaux de l'ile de la Trinidad. POri of Monos. J. Trinidad Field Naturalists' Club. pp. 28-50. Spain: Chroni c le Press. 560 pp. ffre n ch, R.". 1967a. The Dickcissel on its wintering L ill , A. 1970. Nidification in the Channel -billed Toucan in grounds in Trinidad. Living Bird6: 123-140. Trinidad. Condor 72: 235-236. ffre n ch, R . P. 1967b. The avifauna o f Chacac h acare is land. Lill , A . 1974. Sexual behaviour of the lek-forming White­ J. Trinidad Field Naturalists' Cillb. pp. 45-46. bearded Manakin. Z. Tierpsychol. 36 ( 1-5 ) : 1-36. f fre n c h , R . P. 1969. The avifauna of Saut d'Eau is land. J. L i n d b l a d , J. 1969. Jourlley to Red Birds. Hill &Wang. New Trinidad Field Natllralists' Club. p. 16. York. 176pp. ffte n c h , R .P. and F. Haversc hmi d t. 1970. The Scarlet Ma n ol i s, T. 1982. Host relation s hips and reproductive Ib is in Surina m an d Trinidad. Living Bird9: 147-165. stra tegies of the Shiny Cowbird in Trinidad and Tobago. ffre n c h , R.P. 1973. A Gllide to the Birds of Trinidad and Ph.D. diss. Univ. of Colorado, Boulder. 136pp. Tobago. Livi ngston Pub. Co., Wyn newood. PA. 470 pp. Morris, R. D . 1994. Breed i ng c h ronology and reproductive s uccess of seabirds on Little Tobago. Cololl. lVaterbirds 7: 1-9. Review Trinidad ornithology 7

Nottebohm, F. and M. Nottebohm.1969. The parrots of Snow, D.W. 1962c. A field study of the Golden-headed Bush-Bush, AI/imal Killgdom 72: 18-23. Manakin in Trinidad. Zo%gica 47: 18 3 ·198. Quesnel, V.C. 1956. The density of the population of Snow, D.W . 1963. The display of the Blue-backed Manakin breeding Ki skadees in Port of Spain. 1. Trinidad Field in Tobago. Zoologica 48: 167-176. Naturalists' Club. pp. 24-25. Snow, D . W. 1971. Social organization of the Blue-backed Quesnel, V.C.1977. Stay-at- home Starthroat. Living World, Manakin. Wi/son Bull. 83: 35-38. 1. Trinidad Field Naturalists' Club. p. 1 I . Snow, D.W.1985. The web o/adaptation: bird studies in the Quesnel, V.C. 1985. Why do night jars si t on the road at Americ a" tropics. Ithaca: Cornell Un iv. Press. 192 pp.

night? Living World, J. Trinidad Field Natllralist.~' Club. Snow, D.W . and A. Li11.1974. Longevit y records for some pp. 19-23. Neotropical landbirds. Condor 76: 262·267. Quesnel, V.C. 1987. The songs of the Rurous-browed Snow, D.W. and B.K. Snow. 1963a. Weights and wing­ Peppershrike Cyclarllis gujanellsis. Livillg World, 1. lengths of so me Trinidad birds. Zo%gica 48: 1-1 2. Trinidad Field Naturalist Club. pp. 43-45. Snow, D.W. and B.K. Snow. 1963b. Breeding and the Saunders, J.B. 1956. The Scarle t Ibis. Trill. Regent News annual cyc le in three Trinidad thru shes. Wilson Bull. 75: 6(4),10-12. 27-4 I . Saunders, J. B. 1957. Bird Life on Soldado Rock. Trill. Snow, D.W. and B.K. Snow. 1964. Breeding seasons and Regent News 7(3): 4-7. the annual cycles of Trinidad landbirds. Zo%gica 49: 1- Snow, B.K. 1970. A field stud y of the Bearded Bellbird in 39. Trinidad. Ibis 11 2: 299 -3 29. Snow, D.W. and B.K. Snow. 197 3. The breeding of the Snow, B.K. 1973. Social organisation of th e Hairy Hermit Hairy Hermit Claucis hirsllta in Trinidad. Ardea 61: Clallcis hirsllta. Ardea 61: 94-105. 106-122. Snow, B.K, 1974. Lek behaviour and breeding of Guy's Tashian. R.E. 1957. Nesting behavior of the Crested He rm it Phaethornis guy. Ibis 116: 278-297. Oropendola in northern Trinidad. Zo%gica 42: 87-98. Snow , B.K. and D . W. Snow.1971. The feeding ecology of Wiley, R.H . and M.S . Wiley . 1980. Spacing and timing in tanagers and honeycreepers in Trinidad. Auk 88: 291- the nesting ecology of a (ropical blackbird; compari son 322. of populations in different environment s. Ecol. Monogr. Snow, B.K. and D . W . Snow . 19 72. Feeding niches of hum­ 50, 153-178. mingbirds in a Trinidad valley. 1. Anim. Ecol. 41: 471- Williams , T.C., J.M. Williams, L.C. Ireland and J.M. 485. Tea I. 1977. Bird migration over the western North Snow, D . W. 1961 and 1962. The natural history of the Atlantic Ocean. Am. Birds 31: 251-267. Oilbird Steatornis caripellsis, in Trinidad. (in 2 parts). Worth, C.B. 1967. A Nafltralist ill Tri"idad. Lippincott Co. Zoologica46: 27-48; and 47: 199-221. Philadelphia. 29 1 pp . Snow, D . W . 1962a. A field study of the Black-and-whi te Wunde r le, J.M.1985. An ecological comparison of the avi­ Manakin in Trinidad. Zo%gica 47: 65-104. faunas of Grenada and Tobago . Wilson Bull. 97: 356-365. Suo w, D. W . 1962b. Notes on the biology of Trinidad swifts. Zoologica47: 129-139. DOUBTFUL ORIGIN IN SOME BIRD SPECIES RECORDED FROM TRINIDAD & TOBAGO

Richard ffrench Toftingal Laurieston Road, Gatehouse of Fleet, Castle Douglas, Scotl and DG7 2BE

Intr o du c tion abraded by contact w ith ils cage. In other cases we can only The sc ience of ornith o logy operates w ith i n certain recog­ surmise, taking into accou nt aspects suc h as t he dis tance nised paramet ers. All wild species in a p art ic u lar a rea a r e from the bird's natu ral range, its normal tende ncy towards accorded indi vidual status, s uch a s res ident. migrant visi tor, vagrancy or no t, along with other known examples wi t hin the vag ran t, elc. Bu t a ny s pecies whose origi ns a re suspected 10 spec ies, and a ny evi dence relating to local captive coll ections. be artificia l, e.g. as a res u l t of in troduct io n or the prod uct of Some g ro u ps of birds are often kept i n captivit y, so nal­ captive breeding. usually is treated on a di ffere nt le vel from ura lly they form t he la rgest proport ion of t he fo l lowi n g l is t. tr ul y w i ld b irds in a ny s t udy of the area's ornithol og y. They inclu de wate rfowl, the parrots and the seed-eaters. It is therefore valuabl e from t ime to t i me to exami ne a s Apa rt fro m well -known collections of such bi rds a t the c r itically as possible the stat us of a n y species t hat might rea­ Emperor Val ley Zoo in Port of Spain a nd at t he Po inte-a-Pierre sonabl y be suspected of dubious origin. so as to p reven t these W il dfowl Trust, a large n u mber of birds are kept in captivity susp ic i o ns from be ing clouded and eventually forgo tt e n with (but rarely encou raged to breed) by individua l e nthu siasts on the passing of time. In most cases introduced species do not Trinidad and Tobago. Curre n tly there ex is ts no mechan ism su rv ive fo r long, becau se of ecological constra ints or compe­ by which b irds t hat esca pe, or are re leased into Ihe wild, t it io n wi t h ind ige no us s pecies; but t here are a few exam ples f rom t he above so urces a re iden ti fied as such, a nd it seems where s uch species have been s uccess fully nat uralized , so unli ke ly that th is s ituat io n wil l change. A ll the more need. t ha t t hey have now establi shed sizeable populations that are therefo re, for the ornithological comm un i ty to be aware of like ly to remai n so . These includ e the House Spa rrow Passer t he co nseque nces. domesticlls and the Commo n Starling Stllrn/ls vil/garis in

North A m erica ~ a nd the Canada Goose Branla canadensis and Annot a t e d Li s t t he Rudd y Duck Oxyura jamaieellsis in Brita i n ( th o ugh the lat­ (* Denotes a spec ies consid ered here as not being wild or feral ter' s p osi tion in Eu rope has recently come under threat f rom with in Trinid ad or Tobago). con servat ion i s ts. I . Du cks. The avowed a im of the Poi nte-a-Pierre Wildfowl We should not be s urpri sed that in nearl y every c ase of Trust at i ts i ncept ion in J anuary 1967, when I served as successful naturalization serious objections have arisen, a s Honorary Secretary, was to breed i n capti vity various species negative impacts on the natu ra l environ ment come to be of the i ndigenous w i ldfowl of t he country, mos t of which were recognised. Among the disadvantages are the effects of com­ t hen ex t remely rare, and through carefu l re- introduc tion to pe tition eliminating indige nous species, t he over-abundance restock the local population. Pri ncipa ll y thi s applied to two of the newco mer ari si ng from t he lack of natural predators, spec ies of Detldroeygtla: D. Olitll",,,o/is the Black - betlied and the detrimen tal results s uffe red by farmers and o t hers W h istling-Duck, and D. hieolor the Fulvous Whistling-Duck. when crops, gardens, parks and even buildings are damaged Although natural force s may a lso have played t heir part. by Ihese birds. there see ms no doub t that th e recovery of the former on both The following account deals separate ly with more than 30 islands over the last 30 years is partly attributable to the fact spec ies that have occurred o n Trinidad or T o bago over the thaI several h undred D. alitultltlalis ducklings we re released las l 50 years. The great majority are species known to be into the wild. To a le sser ex tent thi s may have happe ned wi t h kept i n ca ptivity by avicu lturi s ts . zoos. or just as pets. D. hie%r. The thi rd species , D. vidllat(l. the White-faced Sometimes indi vidual s esc ape . others are d e liberatel y freed. Whis tling- Duck proved harder to breed in capti v ity. but and in a s mall number of cases this has led to the establi sh­ even t ually t his was successfu l ly achieved, and may we ll have me nt of a viable feral population. breeding in the wild. but been the origin of records suc h as two of thi s spec ies see n by so metime s depe ndent to a varying ex te nt on humans for su r­ myself and Graham While at the Caroni Rice Project on 6 July v i val. Obviously, w ithout a considerable effort it would be 199 1; further records o f four birds were seen by G. Wh ite on impossible to be absolutely certa i n that a particular individ­ 8 June 1992 and 18 J une 1995 (pers. comm.). Ano t her ual bi rd originated from an artificial background; but some­ s pecies commonly bred in capt i vity, both at t he T rust a nd by t imes such a bird betrays its origins by s uch s igns as unusual many a hou seho lder, is the Muscovy Duck Coirina ltIosehara. tameness or plumage tha t has faded in captivity or beco me Any record of this species in the wild s tate has therefore to be Doubtful origins of some T&T bird species 9 viewed with great caution. Unfortunately, o ther non-native when a bird was seen at a farm by members of the Club on an spec ies o f waterfowl have occasionally been bred in captivity, expedition to SW Trinidad . The bird's ex treme tameness, and it is not impossible thaI individuals emanating from suc h combined with an injury to one foot, leads me to suspect that origins may have found their way to freedom. this individual is like ly to have been capt ured or wounded; in Venezuela the species is normally quite sedentary and is 2. Black Vulture Coragyps arratlls. Though com mon r arely seen to fly far from its riverside habitat. throughout Trinidad, Ihis species is rarely found o n T obago. Around 1959 a few were released by those filming " Swiss 7. Feral Pigeon Columba livia. Probably the best known Family Robinson" on Tobago, remaining at large for several of all introduced species and common in most countries; it is years. I saw three th e re in 1965, and si ng le birds in April descended from the Rock Dove, a truly wild bird which fre­ 1967 and August 1968. Isolated birds have continued to be quents c liffs, caves and islands in the more remote regions of seen up to April 1989 (D. Rooks). Perhaps we should not the world. Domestication of this species began in the Middle assume that all these Tobago records are attributable to Walt East at least 5000 years ago, giving rise to the enormous vari­ Disney! The species is commonly found a t Toco, from where ety of plumages found in the domes ti c bird; whereas the wild its keen eyesight and flying ability should not preclude a visit bird is uniformly bluish-grey with black wing-bars and a pal e to Tobago. It is in fact surpris ing that this vulture is not more rump, with a glossy iridescent neck. Many birds are kept as co mmon on Tobago. free-flying pets or used for racing, while a great number of feral birds frequent the large grain-stores on the outski rts of 3. *C rested Guan Penelope purpurascens. In 1962 J was Port of Spain and elsewhere. It seem s unlikely that any indi­ summon ed to the Emperor Valley Zoo to examine a bird vidual s of this species on Trinidad and Tobago could be truly brought in from south Trinidad under questionable circum­ wild birds. s tances. The keepers thought it "might be a Pawi" (Pipile), but it turned out to be this mu ch larger species. Although 8. Parrots. This fami ly. cont aining many popular cage­ fairly widespread in northern Venezuela, includin g th e birds, provides the greatest source of controversy concern in g Orinoco delta (Meyer de Schauensee & Phelps 1978), il seems the orig in of individuals. Apart from the six species known to hi ghly unlikely that such a heavily-flying bird would be capa­ exist as truly wild birds on Trinidad or Tobago, there are sev­ ble of undertaking a fli ght of several kilometres over the open eral others whose origins are more obsc ure and which req uire sea to Trinidad. Guans are often kept captive by Amerindians individual treatment. on the mainland. (a) Blue-aDd-yellow Macaw Ara ararallna. This mag­ nificent large macaw maintained a precarious foothold in the 4. Red Jungle·fowl Gallus gallus. Apart from the vast Nariva Swamp area of Trinidad up to the 1960s. 1 saw a numbers of thi s species reared in the course of the poultry group of 13 at Cal too Trace in 1959. There were still a few trade, a s ma ll but interesting feral population exists on the in the area of Bu sh- Bu sh, where F. and M. Nottebohm (1969) island of Little Tobago, The species was introduced o n the estimated the total population to be 15 in early 1968. BUI a island by an early resident about 1875, going wild after his destructive forest fire in 1970 and relentless enc roachment death (I ngram 1913). During most of the present century by squatters into the Nariva area over the following 25 years their descendants have flourished o n Little Tobago, benefit­ have certai nly wiped out that small remnant. Yet the species ing from the bananas and other frui ts planted for the use of is regularly seen in back-yards, sometimes in a cage, but also birds-of-paradise. Most individual s have reverted to a wild as a wing-clipped individual hopping about in semi-freedom. state, avoiding human contact since th ey were freque ntly In the 1970s and 1980s C. Turpin had severa l free-flying pet hunted for food; recently , with the proliferation of tourist birds living in his garden at Charlotteville, Tobago. Others traffic, a few birds have become tame. taking advantage of were released by R.S.W. Deane and E. Lau at Speyside in a mis­ the scraps provided at the landing-place. But th e fine guided attempt to "re-in troduce" these swamp-dwelling plumage and excellent condition of the wild "chi ckens" that macaws inlo Tobago (where there is no ev idence they ever ha ve seen on the island s how that they have adapted well. actually lived as wild birds). Clearly there is much illegal traf­ fic in and ou t of Trin idad, mainly via the Icacos and Cedros 5. *Crestcd Bobwhite Coli n us cristallls. About 20 years area, where fishing boats can eas ily ply their trade unob­ ago I was notified of the occurrence of a banded bird of this served between Venezuela and Trinidad. Duri ng the nine s pecies which had been found wandering beside a railway months Oc tober 1979 to June 19 80 more than 125 of these track in centra l Trinidad. Its behaviour indicated captive ori­ macaws were exported from Trinidad to the U.S.A. (Roet et.al. gin, but I was never able to discover its previous owner. 1981), ostensibly as nati ve-bred birds, even thou gh by then Quails are some times kept in captivity to provide both eggs the local population was undoubtedly extinct. With the high and me a t for the table. prices paid by U.S. traders, it is nOI surprising that such pre s­ s ure o n the remaining wild popul at ion in South America is 6. *Sunbittern Ellrypyga helias. Only known on Trinidad hard to eradicate. There has been some local movement from one record at Icacos on 25 April 1985 (TTFNC Bulletin), towards captive breeding and re-introduction into the wild, 10 Living World 1997· 1998 but th is cou ld on ly work if a sufficiently large area of appro­ sma ll Austral ian parrot is o ne of the most widely kept cage- pria te habitat i n Nariva Swamp were to be reserved and e ffec­ birds in the world. Occasionally escaped individuals are tively wardened for the macaws, an unlikely prospect. found, but rare ly survive for lo ng in the wild. (b) *Sca rlel Macaw Ara macao. The few r ecords of Ihi s (g) *Sulphur-crested Coc katoo Gacallw galerita. Thi s large macaw arc mostly unsatisfactory. Two birds reponed large white cockatoo is also native to Austral ia. For some from Nariva Swamp in October 1934 (Belcher and Smaoker years between 1988 - 1993 an escaped in dividu al was to be 1936) were see n by an un-named observe r , not by the seen at Crown Point, Tobago, where it frequently associated autho rs; other birds were only heard; a grou p of five were w ith Cattle Egrets. reported from Waller Fie ld by Abbott (Herklols (961 ) without ( h ) Yellow-crowned Parrol Ama zona ochrocephala. details; morc recently there have been isolated records of The precise status in Trinidad of this well-known spec ies (also individua ls from the SW pe ninsu la. While it is not imposs i­ called Ye ll ow- headed Parrot) is one of the more difficult ble for some of these records to h ave involved wi ld birds. we assessments to make in this study. During the peri od 1956 mus t agai n co nsid e r the li kelihood of esca pes f rom captivi t y, to 1976 I searched assiduously for wild indi vid ual s withou t for the species is com mo n ly kept as a pet and is traded like success. The on l y birds found were living in urban or s ub· the previous species through Trinidad (Roet el.al. 198 1). urban surroundings in Po rt of Spa i n, S t Augusti ne or Pointe­ (c) *Red-shouldered M a caw Ara nobi/is. There is no a-Pierre, where capt ive or tame birds were frequently seen o r confirm ed loca l record of wild birds of thi s species a lso known heard. Occasiona l reports from the SW peninsula could not as Hahn 's Macaw. Belcher and Smooker (1936), who never be properly subs tanti ated. and on all visits I made to that personally r ecorded the locally com mon A. manilala, quote area I could only find the nat ive A. amaZOllica. Reports were a n extremely ten uous s i gh t- record by LG.Myers in Octo ber further compli cated by the propens ity of some local bird­ 1934, and also refer to a specimen of doubtful origi n in the keepers to di vide amazonica birds into either "Blue- head s" or Brit is h Museum. I was the refore astonished when I was told "Yellow·heads'· accordi ng to the variabl e amount of yellow by R .S. W. Deane during the 1970s of the intention o f a British each bird had on its head; bu t "Ye llow -head" did not usually aviculturist, the late L. Hill, to "re-introduce" the s pecies to refe r to ochrocephala, although somet imes it did! In 1995 I T rinidad and Tobago by releasing capt i ve-bred birds. Two found a sizeab le population ofochrocephalaapparently li ving birds were actually released at Poin te-a-Pi erre, but seem not in a feral state at the U.\\1'.1. campus, St Augustine. There is to have su rvived. A bird see n by G. White at Waterloo on 20 no doubt that many cap ti ve birds are traded through October 1994 was presumed also to be a n escaped ca ptive. Trinidad , e.g. over 1000 of this species were exported from (d) Brown-throated Parakeet Aralillga perfinax. Small Trinidad to U.S.A. during the 9-mon th period October 1979 to groups of thi s species began to be seen regularly from about June 1980 ( Roet et. af. 1981). It is highly l ikely that th ese 1986 by G. White (1987), ma i nly i n suburban areas of nOrE h birds originated in Venezuela. where regulations forbid Trinidad, and in the Caura and Maracas Valleys (R. Neck les exportation, probably passing easi l y if illegally i nto Trinidad ill /itl. ) at least up to 1994. The species is common in si mi­ v ia the SW peninsula. I conclude the refore that whereas feral lar habi tats throughout north ern Venezuela (Meyer de groups have become established in suburb a n district s of Schaue nsee and Phelps 1978). Observers have reported these Trinidad , the species' orig ins as a wild bird can not be satis· birds to be quite tame, a ll ow in g an approach to five metres. factorily proven, so long as large numbers of thi s highly pop­ While this wou ld normally indicate likely domest icated origin, u lar pet bird are passing through the terri tory. it does seem in this case to be a characteristic of the species. In addition, t his d oes not seem to be a species common ly kept 9. *Red-billed Toucan Ramphaslos lucanl/s. During the in captivi t y an d comparati ve ly few are traded (R idgely 1981, 1960s a bird of this spec ies was brought to the Emperor Valley Niles 1981). Therefore, while one cannot rule out the possi­ Zoo in Port of Spain . It was sai d to have been '·found" in south bili t y that these record s involve feral birds that were origi­ T rinidad by a hunter. Almost certain ly it originated in its nally introduced, I am inclined to consider records of this nati ve country of Venezuela, where it is com mo n in Bolivar s pec ies with i n Tri ni dad to involve genuine ly wild birds. (Meyer de Schaucnsee and Phelps 19 78). and was probably (e) Green-r umped Parrotlet Forpus passerillus. kept in capt i vity after being taken from a nest. Toucans make Although now a common residen t of open habitats in popular pets. Trinidad. the species appears not to have been known before 1916 an d is not represented in any collec tions made before 10. Great Kiskadee Pi/an glls slilphu ratli s. Maybe the that time. This seems to indicate likely i ntroduction from the best-known locally of a ll Trinidad's birds, even though oeca­ ma inland, where it is commo nl y kept as a pel. Int roduction sionally co nfu se d with th e Baal-b illed Flycatcher from Trinidad to Tobago certainly happened during the Megarynehus or the Tropical Kingbird Tyrall/I/IS me/ancho/i· 1950s and 19605. and the species is now widespread on ells. Unfortunately, this very popularity may have led the Tobago in sui table habitats, e.g. Mount Irvi ne. It s hould be spec ies to be introduced to Tobago , where R.S. W. Deane and c lassed as feral, for while it is sti ll frequently caged in both the late E. Lau released birds into the wild at Speyside about islands, it clearly holds its ow n in the wild state . 1970 . The species has now s pread from Speyside as far as (f) *Budgerigar Melopsittaclts Illldillatils. This well -known Louis d 'Or, where D. Rooks has found nesting birds. The suc- Doubtful origins of some T&T bird species 11 cess of these feral birds can be attributed to the notable have been correct. aggressiveness of the species in defending its territory, and predictably may lead through competition to the demise of IS. Purple Honeycreeper Cya"erpes caeruiells. A com­ other species among the indigenou s avifauna of Tobago . mon resident on Trinidad , frequenting forests and adjoining Thoughtless introductions of this natu re are to be deplored. cultivation. In the 1960s I saw a few individuals near Charlotteville, Tobago, but the only records since then have II. Greater Bird-or-Paradise Paradisaea apoda. The been occasional sightings by David Rooks mainly in the area well-known history of this species from its introduction in of Gilpin Trace. Since I know that the late M. Turpin kept thi s 1909 from Indones ia to the island of Little Tobago need not species captive at Charlotteville, where the 1963 hurricane be repeated. After the 1963 hurricane a length y study on destroyed her aviaries , I assume th e above records re la te to Little Tobago by J.J. Dinsmore (1967) found a maximum of escaped cage-birds or their descendants. But it is of course seven individuals, which graduall y dwindled, and the last possible that there is a small indigenous population of the confirmed sighting on the island was in February 1981 , species living in the Main Ridge forest on Tobago. though some claims for later occurrence have been made. People with vested interests have attempted to import 16. Golden-rumped Euphonia Euphonia cyanocephala. replacements from Indonesia for this " island zoo", in spite of This s mall tanager, locally called " Ttte Bleu", seems to be a opposition from the ornithological and conservationist com­ rare resident, turning up here and there at various times in munity. It is to be hoped that those in authority will uphold Trinidad's Northern Range. Although there is no reason to the claims of the natural world against the interests of dollar­ doubt any particular record , one has to beware of assuming orientated entrepreneurism. that all occurrences are natural, owing to the ex is tence of caged individuals which may well gain a temporary freedom. 12. *Red-winged Blackbird Agelaius phoeniceus. A male of this s pecies was found living in Caroni marshes 17. *Trinidad Euphonia Euphonia trinilatis. More com­ between June 1980 and mid-1981 (ffrench and Mano lis monly found o n Trinidad than the previou s spec ie s. A si ngle 1983). It associated with the native A. icterocephallts and record in 1964 from Tobago, where the bird is not known to established its own territory. Although the species is known be indigenous, probably refers to an escaped cage-bird, as to breed as far south as Costa Rica and Cuba, the populations such pets are commonly carried by boat between th e islands. of those areas are not known to migrate south, as do the northern forms breeding in the U.S.A., which reach no further 18. Palm Tanager Thrallpis palmarllm. One of the com­ than California and Texas. J . Bond (in litt.) pointed out the monest birds o n Trinidad, where it frequents a variety of overwhelming likelihood that this isolated bird was trans­ habitats. First noti ced in the Scarborough area of Tobago by ported to Trinidad by boat. Numerous documentation exists D. Rooks and others about 1982, since when it has spread of such instances, and I am inclined to agree that this is the fairly rapidly through suitable areas of the island. Although most likely explanation. hardl y known as a pet bird , the sudden appearance of this sedentary spec ies leads me again to suspect artificial intro­ 13. * Troupial Icterus icterus. There are quite a few duction by some misguided person for an entirely inexplica­ records of thi s species occurring on Trinidad over the year s ble motive. It has been suggested that the species may have (a lthough some of them have turned o ut to be the migratory reached Tobago via a boat travelling from Trinid ad, but I I. galbula). Because the great majority come from urban/s ub­ have no evidence that it has been observed on board any boat urban districts, and in view of the great popularity of this even in harbour. spec ies as a c agebird in Venezuela, I am in c lined to suspect that a ll refer to escaped cagebirds. Otherwise, why are there 19. Finches and Seed-eaters. One of the most regret- no records from the wilder a reas of Trinidad? table developments for local ornithology over th e last 40 years has bee n the progressive disappearance of nearly all the 14. Red-breasted Blackbird SllIrnella militaris. Fairly members of thi s group. There is little doubt that this is widespread in sa vannas on Trinidad, but not recorded from attributable to the depredations of the pet-trade, combined Tobago till November 1974 (ffrench 1991 ). Records of sev­ with the destruction of habitat and general ease of access to eral birds conti nued in the Lowlands area up to about 1990, more remote areas, which previous ly served as a refuge and but have recently ceased (D. Rooks pers . comm.). Bearing in reservoir. As a result, so few wild birds are left of the for­ mind that there is no s uitable habitat along the e ntire north merly widespread Sporophila / Oryzoborllsspecies that nowa­ coast of Trinidad, one wonders how s uch a movement of a days if an observer is lucky enough to find a bird, it is nec­ normally sedentary species cou ld have been initiated. J. essary to ask whether it may not be an escape from captivity. Bond (i" litt.) suggested that there might have been an artifi­ (a) *Blue-black Grosbeak Cyallocompsa cyalloides. A cial introduction, but I have never known instan ces of this male of this beautiful species was recorded near POrt of Spain species in captivity. Howev er , its apparent di sappearance in Jul y 1990 (R. Neckles ill litt.). Th e circumstances and tame­ from Tobago does lead me to wonder whether Bond may not ness of the indiv idual indicate artificial origin, especially as 12 Living World 1997-1998 sing ing males are highl y popular a nd valuable o n t he main­ R e fe r e n ce s. I a nd. (b) "' Lesser Seed-Finch OryzoborlH (llIgo/ensis. On at Belcher , C. a nd G.D.Smooker. 193 6. Birds of t he co lo ny least two occas ions, c.g. in September 1966 and August 1968, of Trinidad & T obago Part III. Ibis 13 (6): I - 35. I encoun t e red singing males of this s pec ies in the area of Dinsmore , J.J. 1967. E col ogy and behavior of the G re ater Gilpin Trace, Tobago. The s pecies was hith e rto unkn own on Bird-of- Paradise on Little T o bago island. M.S. thesis, Tobago, though s tili at that time presen t on Trinidad in sma ll U ni v. o f Wisconsin. numbers. I regard t he most likely explanati on of these occu r­ ffrench, R.P. 19 91. A Guide to the Birds of Trillidad & re nces to be escapes from capti vi t y. especia ll y as I foun d the Tobago (2nd edition). Cornel l U ni vers ity Press. Ithaca. species in similar circumstances on Barbados in 1956. N.Y. 426pp (c) * Java Sparrow Padda oryzivora. Th is popular Asian rfren ch, R.P. and T. Manolis. 1983. Notes on some seed-eater is a co mmo nl y kepi pel, a nd this seems the most birds of Trinidad wetl ands. Living World. J. Trinidad &. likely explanati on for one seen at Turt le Beach. Tobago on II Tobago Field Na lltralis/'s Club. pp 29 - 3 1. November 1980 (C.E.Keller ill titt). Herklots, G . A.C. 1961. The Birds oJ Trillidad and (d) *Co mmon Wax bill ESlrilda aslrild. Another com- Tobago. London. Coll ins 287pp. manly ke pt cage-b ird , native t o Africa sou th of lhe Sahara. Ingram, C . 191 3. Birds-of-Paradise in the West Indies. S m a ll numbers appeared near the Caroni Ri ce Project as early Avicull. Mag. 2, 5: 35-4 1. as 18 September 1990 (G .Whit e pers. comm.), and up to 40 Meyer de Scha uens ce, R. a nd W.H. P help s . 1978 . A have been see n in the area of Trincity water treatment plant guide to rile Birds of VeneZllela. Princeton. Princeton as recently as January 1996 (D. Finch ill lin). It may now be Univ. Press. 424pp. establi shing itself as a fera l s pecies . Niles, J.J. 198 1. The status of p sittacine birds in G u yana . (e) *Saffron Finc h Sica lis Jlaveola. Alt h o u gh com m on 111 R.F. Pasquier (ed). "Conser vati o n of New World locally in western Trinidad, it is not distributed generally Parrots", Smithsoni an In s t. Press. I. C.B.P. thro ug h out the country. Th e species was introduced to Nottebo hm , F. and M. 1969. The parrots of Bush Bush. Charlotteville, Tobago by the la te M. Turpin, a nd a few birds A ni mal Kingdom 52, I : 18 23. escaped after the hurricane, being late r seen agai n in captiv­ Ridgely, R. S. 1981. The c urrent distribu t ion a nd status of ity. It is no longer extan t on Tobago. mainland neotropical parrots. III R.F. Pasquier (ed). (f) *S iskin sp . Carduelis sp. A bird of this genus was seen "Con servat ion of New World Parrots". Smith soni a n In s t. by a British ornitholog ist at Mount St Bened ict on 23 March Press. I. C. B.P. 1992 (J. Hornbuckle jn lift.). Not known fo rmerl y from Roet , E.C. D.S . Mack and N. Duplai s. 198 1. Trinidad, though the Yellow-bellied Siskin C. x01l1hogastra P sittaci nes imported by the Uni ted States (Oc tober 1979 ranges through northern Venezuela east to Paria peninsula. - J une 1980). III R.F. Pasquier (cd). "Conser vation of New Possibly a vagran t , b ut just as likely to h ave been an escaped World Parrots", Smithsonian Inst. Press. I.C.B.P. cage-b ird . White , G. 1987. Noteworthy Bird Records. Livillg World, 1. Trinidad & Tobago Field NaIurn/isls' Club. p 42. Acknowledgeme n ts

My thanks to Graham White for i n format ion and for sugges­ ti o ns to improve this paper; also 10 severa l o ther corres po n­ d ents who p r ovided information and records. SOME OUTSTANDING PROBLEMS IN THE ORNITHOLOGY OF TRINIDAD AND TOBAGO

Richard ffrench Toftingal Laurieston Road, Oatehouse of Fleet, Castle Douglas, Scotland D07 2BE

Introduction. It i s presumptuo u s to su p pose t hat the clutch es of up to 10 eggs from Caroni Swamp. More recently orn ithology of any area can be completel y understood. it has been recorded breeding on Tobago, as well as in ca p­ Nevertheless, for an area as small as Trinidad a nd Tobago, tiv ity at Pointe-a-Pierre. Does it stil l breed i n o ther parts of one which h as been fairly well studied over the las t 40 years, Trinidad ? it ma y be helpful if from time \0 time some of the ou ts tand­ ing problems a r c brought to the attention of the o rnithologi­ Masked Du ck Nomonyx dominica . Although it is c lear cal communi t y, in o r der that perhaps they may be scrut i­ that th is species does breed on Trinidad for groups of i mma­ nized under a more intense focus. To this end I exam ine here ture birds have been found, so metimes acco mpan ied by an so me thi rty species, for which we need answers, mai nl y per­ adult female the egg size remains somew hat a nomalo us . t ai n ing (0 their s tatu s in the cou ntry. Belcher/Smooker refer to tw o clu tches of buffy white eggs, For several species I draw attention \0 cases of possible with granul ated s he ll s, 4 avera g ing 60 x 46 mm. In misidentification in the past, sometimes where very s imilar November 1967 a similar clutch of 4 was brought from Caroni species are involved, and in other cases w here the previous marshes to John Cambridge at Pointe-a-Pie rre, who was well evidence may have been faulty. For instance, the series o f experienced with the locally breeding ducks. He put th e eggs generall y excellent papers on the nesting of local birds by the into a n incubator. where three du ckl ings hatched a day or so late authors, Sir Charles Belcher and G. D. Smooker (1934 - later. I saw th e remains of the eggs heJt s and verified their 1937), were later found 10 have conta ined a few mistakes. approximate size as match i ng the prev ious c lutches . They ascribed nests of the an t-tan ager Habia rubica to the Unfortunately the ch icks soon died, bu t al th o ug h I exam­ antbird Myrm eciza IOl/gipes, and those of the crake Lateralllls i ned th em carefully and prese rved the specimens in alcohol exilis 10 another . larger rail. Furthermore, in a number of for transhipmen t to an American museum, by so me mi s­ instances they admitted that their identification was based o n chance the container was lost. James Bond ( 1978) cha ll e nged the word of a th ird party, often "a hunter", who broug ht in the iden tification of all these clutches on Ihe grounds that eggs or a nes t that he had found. but may well not ha ve iden­ they were too large by compari son with o thers from Cuba. tified th e parent bird correct ly . Collectors not infrequently More evidence is needed! e mpl oy field assistants to find material, and there is no way of ensuring th at al l such material is properl y named. II is Gray -h eaded Kite Leptodon cayanensis. Although this therefore necessary to re-examine so me o f those records, in spec ie s is fairly regularly seen over the Northern and Cen tral o rder to confirm or reject their aut he nt icity. Ranges, the on ly actual record of breeding comes from some For o ther s pec ies I merely remind readers that we require eggs purportedly collected on Trinidad, which reached a co l­ further data to cor roborate what has previously been assert­ lector in Europe (Schonwetter 1961). As egg-co llectors arc ed or assumed. Sometimes eco logical changes or persecution sometimes fo und to be unre liable fo r purposes o f sc ient ific may have al te red the statu s of spec ies, and o nly s ustained authenti c ity , 1 fee l we need firmer evidence that this species observat ion and serious effort ca n s how what is now Ihe tru e breed s on Trinidad. picture. It is my hope that the inc reased number of observers visiting or resident in the country will resul t in answers being Hook-billed Kite ChOl/drohierax IIl/eil/alllS. After appar­ found 10 many of the following probl ems. ent ly escap ing noti ce on Trinidad for over 75 yea rs, thi s species again showed up in 1978, and has si nce been Rufescent Tiger-Heron Tigrisoma lilleatltm Nowadays observed on a number of occasions. Sedenta ry and sluggish, rarely seen in marshland (ffrench 1991). Both Leotaud it may well be breeding on Trinidad. but again we need di rect (1866) and Belcher/Smooker consi dered it a resident spec ies. evidence of this. Eggs, apparen tly from T rinidad, were But the set of eggs c ited by the laller (and brought by some­ described i n Ihe Schon wetter 1961 account, bUI they are sus­ o ne to Smooker) are much too s ma ll for this species. Yet the picious l y large, compared with authentic eggs from Suriname species is not particularly known to mi grate. Does it still (ffrench 1991). breed on Trinidad? Whit e-t ailed Hawk Buteo albicaudatlts. This sava nn a White-checked Pintail Alias bahamellsis. Belcher / s pecies is very rarely found o n Trinidad nowadays. although Smooke r knew this as a resident o f Trinidad and recorded I believe that some obse rvers have mistaken ly confused it 14 Living World 1997~ 1998

wi th a somew hat similar species, the far com moner S h ort~ tai led Hawk Buteo brachYllrlls, which main ly frequents Ye ll ow·bilJed Tern Sterna superciliaris. Thi s sma ll river fores ted country. It is interesting that the latter species was tern is fa irl y common on Trinidad during the same period of never recorded at all by Bel c her/Smooker; thi s makes me the year as P/iaetllsa, but it was not recorded by wonder about the authenticity of their two nests a t Moun t Belcher/Smooker, who may possibly have thought it was th e

Hope . attributed to albicaudatlls. a species the y term "a resi~ migratory Leas t Te rn S. albijrons. They considered the lat­ dent of fairly frequent occurrence". ter might have b red in June. and I suspect the se observations might well have referred to S. sllperciliaris. It is well worth Broad - w in ged Hawk Buteo platypterlls. Another species looking for nests of thi s s pec ies in Caroni marshes. never recorded by Belcher/Smooker (!). but fairly common today. especially on Tobago. Known definite ly as a winter L in ed Q u a il ·Dove Geotrygoll Iinearis. Thi s sec retive visitor (nominate race) from the north, this spec ies is found dove, usually found on or near the ground in high fore st , has on Tobago in all months. Does it nest there, and if so, which been recorded on Tobago, but to my knowledge there are no subspecies is invo lved? A male specimen taken on Trinidad, record s si nce Hurricane Flora in 1963. Is it still extant on now in the American Museum of Natural History. was identi­ Tobago ? fied as from the Antillean race antil/arum, but this was chal­ lenged by James Bond (in lin), who said it was from the nom­ O il b i rd Stealorllis caripensis. One of the mos t ext raord i­ inate northern race. Clearly winter visi tors from the north are nary records of the past few yea rs was that of a group of these al so regularl y seen , often flocking on migration; so it is pos­ birds at Hi llsbo roug h Dam, Tobago in September 1988 sib le that two separate populations are involved. (ffrench 1993). Although the species is know n to roost in sea­ caves on the north coast of Trin idad , and may well commute

As h ~t hr oated C r ake Porzalla albicollis. There are no between Huevos and the mainland , it seems almost incredi­ recent authe nt ic records of this species, which was collected ble that it should fly across 30 miles o r more of open sea on Trinidad in the 19th century. Belcher/Smooker consid- between Trinidad and Tobago. An alternative theory put fo r­ ered it a rare resident, but the three nests they mention con­ ward by Hans Boos, one of the observers at Hill sborough, is tained eggs that were clearly too s ma ll , probably belonging to that O ilb irds may be resident on Tobago, roosting in so far the tiny crake Lateralfus exilis. P. albicol/is much rese mbles undi scove red caves on the sea coast or inland, an extremely the immature form of the migratory Sora P. carolilla, and this unlikely hypothesis in my opinion. The s mall s ize of the may have caused confusion . But is it still found on Trinidad? island, coupled with the propensi ty of thi s spec ies to attract human predation. and to excite the attention of folk-Iorists, Azure Ga ll inule Porplryruiajlavirostris. Thi s species was seems to me to rule out any possibility that it could have just not recorded on Trinidad before 1978, and was unknown to not been noticed over the years. So pending fu rther obser­ fie ld-workers from the Trinidad Regional Virus Laboratory vations from Tobago, I have to assume that the 1988 record (now the Caribbean Epidemio logy Centre), who worked inten­ invol ved vagrants from Trinidad. Much remains to be discov­ s ively du ring the 1960s in the area of Bush- Bush Forest in ered about the species' moveme nts outside the breeding sea­ Nariva Swamp, where the species is now apparently resident. son, though it is known to migrate within the con tine nt. However, a lthough it is regularly seen. no direct evidence of breeding is yet forthcoming. Observer s should look ou t for Ringed K i ngf i sher Ceryle torquata. A regular but fairly this. especially in the Melon Patch area. rare visitor to Trinidad from the mainland. The si ngle breed­ ing record at Caron i of Belc her/S mooker in June involved a Col l ared P l over Charadrills col/aris. Thi s little plover is si ngle egg, unlikely to have been a full c lutch. Similarly th e fairly commonly seen on both is land s from May to November, Amazon Kingfis her CI!loroceryle ama'l.Olla was also found and occasio nally in o ther months. Th e only breeding records breeding once by Belcher/Smooker at Madamas in May, are those of Belcher/Smooker in June and July. Is this spec ies although it ha s h ardly been recorded otherwise from just nn off-season visitor from South America, or does it Trinidad. Do either of th ese large kingfi shers s till breed local­ breed here? If lhe latter, where is it nesting. on beaches, ly? Or is it that river condition s are very different now from recla imed land, or poss ibly sava nna edges at Piarco or Wa ll er 60 odd years ago? Field? F l yca t chers. There are several prob lems involving correct La r ge-bi l led Tern P/ia€lIl sa simplex. Thi s s trikingly pat­ identification in Ihis difficult family , for which the best diag­ terned b ird is a fam iliar visitor to western coastal areas of nostic feature is often the call-note or song. But often visitors Tri nidad, as well as inland mars hes and reservoirs . Reco rded to the islands have insufficient time or opportunity to become in almost every month but less com mon from November to acquainted with the calls, and I be lieve some ide ntifications Feb ru ary, when it re turns to breed on the con ti nent. O ne on the bas is of plumage only may have been mistaken. nest was recorded by Belcher/Smooker in May. Is it possib le Among the few species that vis it from the south of the that it still breeds on Trinidad ? continent during the austral winter be twee n April and Problems of ornithology in Trinidad and Tobago 15

Septembe r is the Va riegated Flycatcher EmpidollolllUS varills. Tobago. with possibly as many as fou r separate forms resident which is intermediate in s ize between two other flycatchers in or visiting our islands. Once more. it is probably risky to that are heavily streaked below, the Streaked Flycatcher differentiate between these in the field. but in the hand care­ MyiodYllastes maculalus and the Piratic Flycatcher LegalllS ful meas urements and examination may s uffice. Ca ll - notes lellcophaius. To make matters more comp licated two differ­ a lso vary s lightly. but the birds are usually s ilent outside e nt races of E. vari/iS may be found on Trinidad. rufinlls their breed in g season. Althoug h some au thors me rge all which is resident in nOrlb Venezuela, and the sou thern forms of chi vi with the North American Red-eyed Vireo oli- migrant race. which is slightly larger. Unlike MyiodYllastes vacells, this is a matter of opinion. It would still be good to and Legatlls. E. varius is very q uiet and it s weak. thin psee is understand the local position of these vireos more clearly. of little help in identification. Observers should always con­ sider calls if possib le when disti nguish ing between these three Ye ll ow Wa rbl e r Delldroica petechia This common nOrlh­ species. ern migrant is a familiar winter resident. but in various forms Another identification problem involves severa l membe rs is known also to breed as far south as Peru. Venezuela and i ts of the genus M y iarclllls. I am usually a little doubtfu l when I offshore islands, as well as a num ber o f West Indian islands, hear of s igh t identifications of M . .oI'a;1Isolli, vellezlleiellsis. inc luding Barbados . A few years ago Dav id Rooks (ill lilt.) tyranllllllls or evenferox being made on the basis of plumage. claimed a poss ib le nest at Grafton, Tobago but adequate evi- Matters are com pl icated by differences in immature plumage. dence was lacking. It is highly unlikely that the migratory The exhaustive study of thi s genus by W. E. Lanyon (1963) for m aes tiva would breed on Tobago. and if the resident form s howed that the birds them se lves discriminate d between the petec hia were to breed there. it would be sedentary and we various sy mpatric spec ies mainly on the basi s of different ca ll­ s hould expect olher records on Tobago between mid-April notes. But aga in this does no t help us in the case of swainsoni, and August. It is also likely that males of such a local breed­ the sout hern migrant that has rarely been recorded on ing form would be di stinctive in p lumage. e.g. with a chest­ Trinidad. since it is very quiet on its nOrlhern winter quar­ nut cap. It is well worth keeping a look out for suc h an inter­ ters. M. veneZIle/ellsis. on the other hand, can readily be dis­ esting occurrence. No te. A very recent record of a singing ti nguished on Tobago by its call from tyraflfluills, but it is warbler on Tobago in July (Hayes. ill lilt.) may indeed throw probably not sa fe to identify these two spec ies by plumage, more lig ht on this situation. except maybe in the hand. Another poss ibility that has not yet been explored is thatferox, which J ame s Bond and others Pu rple Ho n eycreeper Cyallerpes caeruiells. Soon after mistakenly cla i med to be loca ll y resident (before Lanyon Hurrica ne Flora in early 1964 I found a few individual s at s howed those records to pertain to vellezuelensis or slVa ill­ Pigeon Peak. Tobago. but have never seen them on Tobago sOlli), may turn up on the Bocas Islands. since it is known s ince . I understood at the time that the late Mrs Mavis Turpin from the Paria penins ula. Observers should listen for i ts dis­ had kept some in capt ivity, and that the hurricane damage tinctive call-note, a short, rolling prrrl. had released her birds. David Rooks (ill lilt.) has been see ing The other rIycatcher proble m relates to anot her difficult the species ··regularly" in the Main Ridge forest, but I know genus. Elaellia. While the common Jlavogaster presents no of no other sigh tings. Observers should be on the look out difficulty, it is quite difficult to tell apart in th e fie ld the for thi s species. which in good conditions can eas il y be dis­ sma ll e r species. chiriquen.ds and parviroslris. The latter is tinguished from it s congene r . the resident Red-legged a southern breeder, migrating mainly no rth to Amazonia. Honeycreeper C. cyan ells. and very occasionally reaching Trin idad. Although it is somew hat larger than chiriquen s is. it is probabl y unsafe to G old e n-ru mped Eup h on i a Ellphollia cya nocephala. identify it in the field, especially as it is unlikely to call. Originally known as the Blue-hooded Eupliollia. and locally as Observers should look out for nests of chiriquellsis in areas Tete Bleu, this species is very elusive nowaday s. possibly like Aripo Savanna; but again some members of thi s species because of the depre dations of bird-catche rs. Although may migrate into Trinidad fro m the main land. BelcherlSmooker recorded one nes t in J uly. the irregular sightings on Trin idad point to the likelihood that this species B lu e - and-w hi te Swa ll ow NotioclJelidoll cyalloleuca. is an off-season visitor from the mainland. All authe nt ic Ano ther southern migrant that is usually fou nd fro m J une to sigh tings should be recorded. September in su itable habi ta t in wes tern Trinidad . this s pecies was also found breeding by BelcherlSmooker in late Sooty G r ass q u i t Tiaris fllligillosa. Fairly common on March: this might poss ibly have refe rred to the re sident nom­ Trinidad nowadays, where it is su bject to seasonal mig ration. inate race, that ranges from Costa Rica south. It is probably Care must be taken not to confu se this spec ies with its con­ i mpossible now to prove th is. for no other nests have bee n gene r T. bicolor. which is common on Tobago and found loca lly . Have they been looked for? Chacachacare. No authentic records yet from To bago. where sl ight variations in male plumage in bicolor may have led to C hi v i V i reo Vireo child. In my book (1991) I s e t ou t the mis-identification. Although bicolor is s light ly larger than complex nature of this species' s tatus in Trinidad and fllligillosa, the latte r' s wing averages 10 mm lo nger. wh ich 16 living World 1997-1998 should help identification for birds in the hand . Bond , James. 1978. 22nd Supplement to the Check-list of Birds of the West Indies ( 1956). Acad. Nat. Sci. Orange· fronted Yellow· Finch Sica lis columbiana. Philadelphia. 2 0pp. Another oddity, in that the o nly records for Trinidad are (french , Richard. 1991 . A Guide t o the Birds of Trinidad those of Smaoker, who collected this species near San Juan i n and Tobago. 2 nd edition. Cornell Univ. Press, Ithaca, September 1926 and recorded nesting at the same lime . The New York. 426pp. species may easily be confused with the Saffron Finch S. ffrench, R i chard. 1993 Further records of birds on flaveola, but the latter is noticeab ly larger. There seems no Trinidad and Tobago. Living World, J. Trinidad and reaso n w h y columbiana s hould not occur o n Trinid ad, so Tobago Field Na turalis ts' Club. pp. 28 - 3 1. o bservers s hould look out for it in ranch-la nd and savannas in Lanyon , W. E . 1963 Experiments on Species Discrimination the west. in M yia rcllU s Flycatchers. Am. Mu s. Novi!. 2126. 16pp. Leotaud, A. 1866. Oiseaux de t'ile de la Trinidad. Port of References. Spai n. Chronicle Press. 560pp. Schonwetter, M. 1961. Handbuch der Oologie, part 3. Belcher, C.and G. D. Smaoker. 19 34 - 1937. Bird s of t he Berlin. 184pp. colony of Trinidad and Tobago (i n 6 parts). Ibis ( 13) 4: 572 - 595 el seq. A RECONSIDERATION OF SOME CAPRIMULGIDS ON TRINIDAD AND TOBAGO

Richard ffrench Toftingal Laurieston Road, Oatehouse of Fleet, Castle Douglas, Scotland D07 2BE

In t ro du e t ion. T he Ca pri m ul g id ae, inc ludin g t he su b·g ro ups L. semitorq ua tu s n aftereri (whi c h i tsel f may be a separa te, ni ght h a wks a n d ni g ht ja rs, a rc diffic ult birds 10 stud y, large ly c lose ly re l ated spec ies) in Argentina (S t raneck et al. 19 87). because they are noctu rna l o r, in some cases, c repuscu lar. In Pe nd i n g f u rth e r ev ide nce, we s ho ul d mai nta in a n o pen mi nd ad di tion, t heir mo tt led, c ryptical l y pattern ed plu mage c reates on Herklo ts' record , but it seems likely th at t h is species is a n p roblems of ide ntifi cation for man y o b ser vers, espec ia ll y i n arboreal nester. and observers would do we ll to look fo r nest s d i m light. i n trees. The six (or seven) species known to occur on Tri n idad In Venezue la Paul Sch wan z (pers. comm.) reported that andlor T obago (ffrench 1991) inc lude some fo r w hich detai ls t hree differe n t -sized Lllrocalis ni g hth awks occ u rred in cloud of statu s a nd dis t ribu tion a re far from c lear . In th is paper I am forest at Ranch o G rande. He cons ide red t he m d iffere n t s ub- out l i n ing some of t he problems. i n the hope that f u t ure spec ies, i ncludi ng t he nomi na te race (whic h occ u rs on observers may s ucceed in unrave lli ng them. T wo s pecies, the T rin idad), schaeferi a n d the sout hern m i g ra n t nattereri Pauraque Nyctidromlls a/bicollis and t he White-tai led N ight jar (occ urri ng in Ju ly and A ugust). A fourth. larger race rufiven­ CaprimJl/gJls cayellllellsis are comparat ive ly wel l-k nown, t he Iris is found in the Venezuelan A ndes. Seu tin a nd Le tzer fo rmer hav in g been the su bject of recent papers in t hi s jou r­ ( 1995) remarked t hat k now n egg s izes of the a bove Lurocalis na l (Q ues nel 1985, 1989 . 1993). I s ha ll deal mai nl y w ith the va ri e d s i g nifica ntly, b ut poi nted o ut t hat the egg fo u n d b y o t he r s pec ies . He rklo ts was small e r tha n would be expected. If it wa s not a Lllrocalis e g g. it coul d o n l y have come fro m Ca p ri mulg u s 1. Short - tail e d Ni g hth a wk Lurocalis sem i rorqJlat Jls . caye n nen s is, which is a c o mmo n b reedin g res ide nt i n the Thi s s mal l n ig hth awk, f orme r ly k nown as Sem icoll ared W alle r Fie ld area . Nighth awk or (erroneous l y) N i ght jar, is w ide l y d is tribut ed on L. sem i to rqllatll s has bee n o bser ved in ever y mo nth o f the T rin idad but canno t be ca ll e d common. It is regul ar ly seen year o n T r inida d , but it is certainly not c lear w hethe r the over fo rest i n the Ari ma va lley a t a lt itudes up to 450m, but sa me race is invol ved all the time. Migran ts f rom t he south also a t sea-level i n the Oropouche Lagoon, w he re I fo und small would be expected be t ween May a nd October , a nd I w o uld bands haw kin g a t du sk over small ri vers near mangroves . It is expect b reedi ng (if a n y) to occ ur betwee n February an d June. c haracte rised by its small s ize, generall y dark colour wi th Regular obser vati o ns of t hi s s pecies in its known ha unts w h i tis h c rescent-shaped ma rki ngs o n t he second aries a nd shou ld reveal a ny s i g n ifi ca nt seasonal move me nts. covens. but 110 pale bar Oil the primaries, (wh ich all other locally occ urring nighth awks have ). A no the r d iag nostic fea­ 2. C om m on Nighthawk Chordeifes mino r. ture is t he short, square-ended ta i l, so t h at at rest its long The o nl y l ocal record is t hat of Ki rk fro m To bago ( 188 3), b u t w ings protrude beyond the ta il. In fli g ht it is c harac te ri sed by h i s com ment th at it occurred " fr o m Jul y to Octobe r" prompts its e rra tic. bat- li ke mo veme nt above the trees o r over w a te r. I s pec ulat io n th at Kirk may ha ve c on f used t hi s no rthe rn have never heard a son g, t houg h I have heard its fli g ht c all , a m i g r a nt w ith it s co n ge ne r C. aCll tipe nnis (see be low ). li ght, hissi ng whick-whick o r wee- it. In Venezue l a i t has appar­ Certain l y the North A me r ic an C. minor does mig r ate so u t h e n t ly been r ecorded c alling a re peate d cu-ick o r clle-wit inlO the A n tillean is lands (whe re I saw it on Barbad os) and ( t houg h t hese call s may poss ibl y have co me f rom a d iffe re nt Sou t h A merica, so c ould be ex pecte d o n Tr i nidad a nd To bago. species). I have seen bird s o f t hi s genus fl yi n g over C row n Poi nt, but at T he s ta tu s o f t h is spec ies in Tri n id ad is mad e uncertai n the time was never ab le to identi f y it adeq ua te ly. The mos t by r ecent doubts over the onl y breeding record, th at o f like l y mo nth fo r s ig htin gs is October, and t he time duri ng the Herk l o ts ( 196 1) a t Aripo Savann ah o n 28 Ma rch 1954. A n egg last hour o r so o f d ay li gh t. Its close s imilari ty to t he sl ig htly meas uri ng 23.5 x 16. 0 m m was be in g inc u ba ted o n the s maller C. aCll tipenn is means t ha t identification may well g rou nd, b u t Herk l ots, as us ua l, d i d not s tate whet her or n o t he necess itate a bi rd in the ha nd. collec te d the adul t (a nd in fac t i t seem s he rare l y p reser ved his speci mens). Recently Seutin and Le tzer ( 1995) fo und a 3. L e ss e r Night h awk Chordeiles aClltipennis. nest of th is s pec ies (subs p. nocti vag us) in Panama. It w a s on a T his is the spec ies t hat is co mmo n ly seen in Trin id a d during h o ri zo nta l branc h 6 m up in a n E ry thrina tre e , w here t he AuguSl to O c tober, fee din g hi g h over the Caro ni sa va nnas a nd a u tho rs o b se rved t he d evel o pme nt of a chi c k over several mar shes duri ng t he l ast ho u r of d ay li g h t (un less some of these days. Th is nes t locatio n matc hed those of thr ee oth er nests o f a r e C. minor !) . At suc h a time o ne can see as man y as fift y in 18 livi ng World 1997-1998 the s ky at once, tw isting and turning as t hey feed up to 30m two ni ghthawks nest i ng, but no-one e lse has, could be ecolog­ above ground. After d a rk they can be seen along tracks a nd ical c h a nges - drainage, agricultural and economic develop­ side-roads in the savannas, their eyes gleaming in the reflec­ ment - that have occurred in the Caroni marshes over si x o r tion of car head-lights. All specimens of Chordeiles so fa r col­ seven decades. But large numbers of both species still occur lected on Trinidad and Tobago have been acu t ipellllis. du ring the migrat ory period , when we should expec t that Discrimi nation of the two Chordeiles species s h oul d be rea­ most bi rd s visiting Trin idad (or Tobago) from South Ame rica sonably easy in good conditions, but of course mostly they are would be south e rn breeder s. On balance I believe that both seen in the last hours of daylight or after dark. these nighthawks are non-breeding visitors. The nesting of acutipenllis on Trinidad has been record­ ed o n ly by Bel cher and Smaoker (1936), who fou nd four nes ts 5. Rufous Night jar Caprimulglls rufus. fro m February to mi d-May, a time when I personally h ave One of the mys teries about th is large ni gh t jar is that it has so found the s pecies scarce, to say the least. I wonder therefore rare ly been seen . Yet it seems to be not uncommon in its habi­ if t he ne sting (and the eggs) attri buted by t hese autho rs to tat of deciduous scrub a nd li g ht woodland in NW Trinid ad. this species m ay not have been those of Caprimulgus cayen­ Durin g 1965 - 1969 on a number of overn i ght v i si ts t o t he lIellsis, as suggested earl ier for Lurocalis. It is perhaps signifi­ Bocas Isl a nds I heard many call ing birds between April and cant that Smooker had spent quite a few years collecti ng on J u ne, and the s peci es has been heard i n the Northern Range Trinidad w ithout ever encounteri n g cayellnells;s; he eventu­ foothills as far east as St Augusti ne . Yet no calls are heard out­ ally fo u nd out from Roberts (1934) that it was res ident. The s ide the breeding season, a nd the species i s certain ly difficult eggs and nests of the two s pecies are ver y similar. to find when it is not call ing. In spite of considerabl e efforts If I a m wrong in my s u s pi cions , then we shou ld regard C. (even including attempts a t mi s t-netting) we never ac tuall y aeutipenllis as a rare resid e nt. In that case, why does o ne located a si ngle Rufous N i g ht jar on t hose Bocas visits, and no never hear its call? This might be described as a churring trill specimen has ever been co ll ected there. l as t ing abou t 10 seco nds and often repeated. I t must be Belc h er a n d Smooker (1936) recorded fou r n ests admitted that s u ch a sound at ni ght i n marsh or savannah (February to May), and one bird was actuall y seen. There is co untry migh t be interpreted as comi n g from a n amph ibian; in deed no mistaking the breeding song of the species, wh ich but the songs of the various spec ies of frog s and toads in resembles that of its congener the Chuck-wills-widow, bu t can Trinidad have been intens ivel y stud ied (e.g. Kenn y 1969, safel y be separated from that species. David Rooks (pers. 1971), so that a n y si milar song coming from a night jar wou ld comm.) claimed to have found thi s s pecies nesti ng o n Tobago, almost certain ly have been n o ticed. I am therefore i nc l ined t o identifying it only by its plumage. He did not "notice" any believe that C. aeutipellnis is a common. regu lar migrant to call, and I believe that confirmation can onl y follow if thi s T rinidad (a nd le ss common on Tobago), but is unlikely to be diagnostic feature can be safel y identified. a resident at all. Identification of Ca primul gi d s. 4. Nac unda Nig hthawkPodager naeunda. It is a comm o n feature amongst bi rd s of crypt ic plumage The statu s of this considerabl y larger nighthawk is simi l ar to (e.g. night jars, so me antbirds and flycatchers) that they com­ that of the previous species but different ! Li ke C. aeutipennis municate wi th each other quite effective ly i n condit io n s of i t i s most com monly fo u nd between J une and October, and thi ck u nderg rowth or i n darkness, largely by distinctive ca ll ­ rare ly at other times. Belcher and Smooker (1936) recorded notes , ra t her than by d ist inctions of plumage. Human t hree nests with eggs in Caroni marshes du ring April. In thi s observers of these groups of birds might do we ll to e mulate case the l arge si ze of the eggs prec ludes any mis take. But the birds themselves, and rely for secure identifi cat ion on sat­ again there are few si g hlings of thi s spec ies except during the isfac tory understand in g and app reciation of th eir so ngs or period when most south te m perate mi grants move to the calls. Of the three commonly breed ing s pecies of caprimul gid s north of the continent. on Trinidad and Tobago, all the calls are quite distinct, even The sma ller race of this species, minor, breeds in t h e with t hei r sl ight variations . northern part of Sou t h America, so one assumes t h at Although Victor Quesnel ( 1993) has distinguished up t o Smooker 's obser vat ions of breeding bird s refer to th is race. seven separ ate "ca ll s" fo r the Pauraque Nyctidrolllus a!hieo!- Certai nly mos t of the collected spec imens are s mall enough, lis, by far the co mmonest is the far-carrying w h istle, resem­ a l though two birds have wing chords measuring 236 a nd 238 bling wee-oo, which may be preceded by one or t wo prelimi­ mm , large enough to f it the mig ratory nominate race. The lat­ nary notes that are occasi onally added. It s other ca ll s are ter i s known to reach Venezuel a from late May (Friedmann mostl y si mi l ar. and Smith \950). There a re no specime n s availab le from T h e Whi l e- tailed Night jar Caprimulglls cayennellsis Tobago, and to m y knowledge no records from this cent ury, found on both is lands, m ainly cal l s an extreme l y h igh-pi tched but Kirk ( 1883) claimed to h ave collec ted two in September double note - chi-peeer, quite un li ke the lower-pitched song and O c tobe r ; it seem s unlikely that he was mistaken in t he of Nyctidromus. But both these species make a whistling call. identity of this s trikingl y large nighthawk. The Rufous Night jar C. rufus has a completely different One solution t o the p u zzle as to why Smooker found t hese breeding call from the above two s pecies, being a rhythmical , Caprimulgids in Trinidad and Tobago 19 but not al all musical, series of four or five notes, chuck-wit- Kenny, J . S.1971. A further contribution on the amphibia of wit-wee-oo, repeated every few seconds. would never Trinidad. Trinidad alld Tobago Field Nataturalists' Club describe this sound as a whistle, and I ca nn ot understand Jourllal. pp. 24 - 25. Belcher's description of it ( 1936) as a "characteristic hiss". Kirk, J. 1883 . List of birds in L. G. Hay ' s Tobago Handbook. Once heard, it can hardl y be forgotten, for during the breed­ N.p. ing season males seem to call constantly through the hours of Quesnel, V. C. 1985. Why do ni ght jars s it on the road at darkness. night? Living World, Trinidad and Tobago Field Not e. Most night jar species also use their wings in Nataturalists' Club Journal. pp. 19 23. courtsh ip to crea te a drumming or whirring sound , quite sep­ Quesnel, V. C .. 1989. Observations on the Pauraque night­ arately from their calls. jar, Nyctidromus albicollis, on the roads at night. Living World, Trinidad and Tobago Field Nataturalists' Club Journal. pp. 12 - 15 . R e fe r e nee s. Quesnel, V.C. 1993. Patte rn s in the calling activity of the Pauraque night jar, Nyctidromus albicollis. Living World, Belcher , C. and G.D. Smooker. 19 36. Birds of the colony Trinidad and Tobago Field Notaruralists' Club Journal. of Trinidad and Tobago, Part 3.lbis (13) 6, I : 1- 35. pp. 42·47. (hench , R. P. 1991. A Guide to the Birds of Trinidad and Roberts, H.R. 1934. List of Trinidad birds with field notes. Tobago (2nd edition). Cornell University Press, Ithaca, Trop. Agric. Trin. 11 (4): 87-99. New York. Seutin, G. a nd M. Letzer. 1995. The Short-tailed Friedmann , H. and F.D. Smith. 1950. A contribution to Nighthawk is a tree nester. J. Field Orn. 66 (I): 30 - 36. the orn ithology of northeastern Venezuela. Proc. U.S. Straneck, R., R. Ridgely , M. Rumboll and J. Herrera. Nat'l Museum 100: 411-538. 1987. EI nido del Atajacamino castano Lurocalis not­ Herklots, G.A.C. 1961. The Birds of Trillidad and Tobago. tueri (Temminck) (Aves, CaprimuJgidae). Collins, London. Commullicaciones del Museo Argentino de Ciencias Kenny, J . S . 1969. The Amphibia of Trinidad. Studies 011 the Naturales "Bernardino Rivadavia " 4: 133 - 136. Fauna of Curacao and other Caribbean islallds, No. 108 . The Hague, Netherlands. SPECIES AND INDIVIDUALS RECORDED ON CHRISTMAS BIRD COUNTS 1976 - 1980 Ian Lambie and Vishnu Debie Asa Wright Nature Centre, Arima Valley, Trinidad

In the 1985 ~ 1986 issue of Living World magazine, the results of SPECIES TYEAR -- f976 19n 1978l--f979 1980 the Annual "Christmas Bird Counts" for the years 1 969~ 1975 and r--:: I I I I I 198 1- 1983 were published by Richard ffrench, fa - ~ ita . ?-_ The results of the "Christmas Bird Counts" for the years 31 Hook-"i1IiiOi

I~ HUOOY IUms one Ib rna eo oanoplper _'-""- 77 ammon ompe " 0·180WICer e ow- I ern IlU oco uove 1<' 14 <'4 l' 81 <:aleC ~~_ '" 4 " " 8 en eon 4 - ar • - ~ommon rouna-uove .... l3ln t:lreastea I founa uove lU ~ t-::g:86~H~UU~ddlOy~'(j~'r~ounj-< rTITIOn 000 eiiiiCo1 ar I a llU aui'iique 1 111 Whlte- allea Nllhtjar 18 oare WI es nu - 0 ar WI _b_ 114 raY-Hum e WI "" "" '" , 41 115 tsand-Humpea oWin "" '" <'4 { eras 40 - al WI esser wa ow- al WI ol1

Number of different species for years 1969 to 1996 = 265

CW = Count Week SAVANNA EXPANSION IN TRINIDAD, W.I.

PAUL L. COMEAU and COLIN C. CLUBBE c/o National Herbarium, U.w.I., St. Augusti ne, Trinidad. W.I.

Introduction The t en savannas at Afipo. coveri n g 267 ha, are si t uated Neolropical savannas have received co nsiderable atten­ in the Northe rn Basin or Clfoni Sy nc line of Trinidad. They are lion over the years because of their ecology. specially adapt­ located within the Long Stretch Forest Reserve (Fig. I ), eleva­ ed plants and unusual soil conditions (Myers 1933, Beard tion 38 m. which contains the Mallicaria-Jessellia-Ellterpe 1953. Bl ydcllstein 1967, Ahmad a nd Jones 1969a, 1969b, association, Calophyllum faciation (Gal ba-Pa[m ) of Marsh Eilen 1972, Sarmiento and Monasterio 1975, Medina 1982, Forest (Beard 1946) or the Calophyllllll1-Palmae (Galba- Palm) Sarmiento 1984 ) , In Trinidad, most of the published wo r k on type (Marshall 1934). The savannas include palm islands and this unique ecosystem (Beard 1946. 1953, Ri cha rd so n 1963, open treeless expanses cove red by sedges, grasses and other Quesnel 1979) has concen t nl(cd on th e vegetation dynamics herbaceOlls plants. The s urroundin g fore st. dominated b y of th e open Silvan na s with recent studies focusing on the role palms in the und erstory. together wi t h the savann as form a fire plays in altering species composition (Sc hwab 1988, vegetat ion mosaic th at occurs on fiat topography with rainfall Co meau \990). As a result. little attention has been directed ranging between 250 to 280 em per a n num. The sava nn a so il towards the marginal or ecotona l areas ( Leolaud 1992) except belong s to the Order: Ultisol, Suborder: Aquulls. Great Group: to point out the sharp boundaries that exist between forest P lin t haquults (Ahmad and Jones 1969b) which has a fragipa n and sava nn a (Marshall [934. Beard 1946. 1953). The f re­ i. e., iron pan, restricting dra inage. quent burning of savanna margins is ment i oned by Richardson ( 1963) but no evidence is presented to indicate Met h o d s that sava n na plants are coloniz ing these da m aged areas. Surveys we re carried o ut along the marginal areas of the Recent field work by the authors al the Aripo Savan nas has largest savan nas at Aripo (I, II, III. V. VI II ) as well as Savan nas shown th at few s harp ecotona[ boundaries remain, that a lot VI and V II . The sampling was done over a seven year period o f the marginal area has been damage d by fire and thaI evi­ (1986-1993) during the months of November. December and dence exists which supports savanna expansion. January - end of wet to beginning of dry season; April, Ma y, and June· end of dry to beginning o f wet season: plus J uly Background and August - middle of wet season. A ll spec ies g ro w ing around the perimeter of these savannas were recorded and notations KEY mad e on abundance, habit, morpholo gy, phenol ogy and habi­ - Roods •• ' Trolls tat. Those plants that cou ld n ot be iden t ified ill Silll were later ""'- Rivrrs keyed out us in g the Flora (Williams e t al.1928-) a nd reference .~~,., ArUliciot channet ...... Nawral donne! co ll ection of t he Nat ional Herbarium. Occasiona l forays were ...... WClttrshc.d made into the forest during th e m a rginal survey to determine the ex t en t of the ecotone whenever it became diffu se. Six 10 x 10 metre sq uare plots were established to obtain quantitative data (Fig. 2). Three of these plots were con t rols located in Sl able co mmunitie s. o ne from t he o p en sava nn a o f V, designated OS5, and two from sharp forest/savanna eco­ lones o n Ihe east a nd west sides o f Savanna VI , namely FS6E and FS6W. These twO plots we re t hen di vided in to CFP6 fo r the co mbined forest phases a nd CSP6 for the combined savan­ na phases. The rema i ning three p lots were loca ted i n diffuse unstable transi t ion zo n es between savanna and forest. one in Savanna V I , designated TS6, and two in Savanna V, one locat­ ed in the sou th -eas t section a nd th e o ther in the n orth-west sec tion of that savanna. designated TS5S and TS5N respec­ tive ly. A ll s peci es in s id e the plots were noted and v isual est i­ mates were made on occ urre n ce. whe ther ab un dant. common or rarc, as well as life-for m, i.e. tree, s hrub, herb, epiphyt e, or c limbe r.

Figure 1; The ten savanas al the Long Stretch Forest Reserve Map Ob serv ation s on mar g i nal ,' ariabi l it y Adapted from Schwab (1988) A wide variety of marginal conditions occur with respec t t o s pecies composition and habitat features. Amidst Ihis variation. 24 Living World 1997-1998

the high degree of fire disturbance that occu r s on a pe ri odic basis (Sc hwab 1988). Some of th e best s harp-bounda r y exam­ FS6W o 0.5 1km FS6E ples are found in Savanna V I ( Pl ate I) whe re two plots we re T56 set up. In the absence of disturbance, ego fire and cutt ing, common mature fore st trees near sava nna margins include VI !lex ar;mellsis ( Bi scuitwood). Is ertia parviJ/ora (Wild ixora) and Parinari clI lI1fJ estris (Bois bande) together with all the dominant palms l is t ed by Beard (1946) for Marsh Fore s t • T55N except Atta/ea lI1aripa (Coco ri te palm ) w hich may be absent N beca u se edap hic conditi o n s arc tOO wet in ecotonal areas . •055 Mature trees of the mars h fore s t reac h an average height of 15m, the t a ll est being Mallritia j lex ll osa ( Moriche palm ), Symphollia globuliJera ( Ye ll ow mangue), Terminalia amazo­ Ilia ( Whi te olivie r) and Parjllari. Along some forest margins V 1 Chrysobalall ll s ;eaeo ( F at pork) is the domina nt tree. As these trees m at u re they lean towa rd the light, b ecome top heavy and topple over which cou ld explai n t heir absence in some Figure 2: The positions of the sampling plots in Savannas V and VI area s. at the Long Stretch Forest Reserve In the open treeless savanna (Plate 2), which represents Legend: the ot h er s table community, the ground cover is dominated FS6W= Forest/Savanna 6 west by th e sedges RhYllcllOspora Cllrv llla a nd R. barbara, the grass FS6E=ForestlSavanna 6 east Pa spalllll1 fJllchellllll1 and the insectivorous Drosera capillaris TS6=Tran sition Savana 6 TS5N=Tran sition Savanna 5 north-west OS5=Open Savanna 5 TS5S= Transition Savanna 5 south-east re peated p atterns we re o b served and these formed the basis fo r establishi ng three s uccession h ypo theses and co mpari ng them wi t h the stable co mm unities whic h served as control s ites. T o understa nd the dy n am i cs o f savan n a expa ns i o n a nd/or con­ trac ti on it is necessary to first clearly define what arc the more stab le commun it ies in the savannatrores l complex. The most s table co m mu nit ies e ncou ntered during the s u r­ vey were th e margi nal areas w here the sha r pes t ecoto nes occur bet ween fores t a nd sava nn a, and th e open sav anna away from ma rgi n a l di s turbance. Sharp ecoton es are not a common fea ture aro und a n y of the Aripo sava nn aS reflecting Plate 2: Treeless open expanse of Savanna V with stable ground cover dominated by grasses and sedges

(S u ndew). A lth o ug h th ese spec ies are the most ab un dant. the vegetation d oes no t form a continu o u s cover but is p atchy wi th u p t o 40% of the grou nd s urface being d evoid of plant s, except for algae, attesting to competit io n for l i m i ted nutrients and moisture ext remes. The only sh rubs p resent in open sava nna are Chrysob(l!alllfs ieaeo (Fat pork ) and Byrsollima era.uiJolia (Savan n a serrette). The next s tep was to com pare t hese comm unit ies wi t h the u n s t ab le asse m b lages along savan n a margins i n o rder to gauge deviation away from o r progress ion towa rds stab ili t y. I n th e t ransit ion p lot in Savanna V I which appears t o be reverting back to forest. the most abundant p l ants are t h e com p osite Wedelia ca raeasolla. the sedge Scleria bracleata, sharp ecotonal boundary between mature marsh and t h e melastome Micol1ia ciliata, and H elicon ja psittacorulII. savanna on the western side of Savanna VI Th e form e r t wo arc ag ressive co mpetito rs taki n g advantage of increased light when the m arg in get s opened up by fire Savanna Expansion 25

Plate 3: A diffuse margin on the western side of Savanna VI where Plate 4: The north-west corner of Savanna V where expansion is relict Moriche palms are the on ly marsh forest trees to have survived taking place the fi re damaged habitat. Sh rubs beneath the palms, such as /lex arimensis (Biscuitwood) and Isertia parviflora (Wild ixora) are good indicators of the re-establishment of forest. d isturbance while the HelicoIJia is an i n vas i ve weed. Mauritia flexllosa ( Moric h e palm) was the only mature tree ( Pl ate 3) indicating its ability to wi th stand fire as there was ev idence of burning in and arou nd the plot. Once the ca nopy is open ed up by fire, the increased li gh t produces a dense mid -story herbaceous layer made up mainly of sedges (Diplacrlllll [ongi­ folium, Lagellocarplls gllialleIJsis and Scleria) that allow litt le light penetration to the grou n d . In well-shaded areas, beneath s h rub s and immature trees, forest species such as the woody vine ROllrea slIrinamellsis, a n d MaproliJ/ea gllianensis ( Petite­ feuille) ga in a foothold. The two lransition plots in Savanna Yare composed of Plate 5: The gap between the Moriche palms in the foreground once plants th at show s trong aff inity with savanna vegetation . In contained trees which were destroyed by repeated fires. Grasses and the plot in the sout h -east sector, the mos t abundant plants are sedges such as Paspalum pulche/lum, Rhynchospora barbarta and the marginal melastome C011101ia verollicaefolia, the grass Lagenocarpus rigidus, once established, favour the development of savanna. Pa spal1l 111 plllchelllll1l and the sed ge RIz YllchosfJo ra filiformis, the latter two bei n g open savanna s p ecies. Th e melas tome Both the transi tion plots in Savanna Y occur in areas appears to be in d ec l ine being replaced by the grass and where forest once exte nded in a n arrow band ac ross the land­ sedge. The microtopography, in thi s and other marginal sc ape isolating smaller pockets of savanna from larger ones . areas, is hummock y due t o worm and termite activity and Recurring fires eliminated the forest trees and opened a p er­ possibly soil creep (Com eau 1990) . Rel i cts of the forest pers ist manent gap , allowing the isolated pockets of savanna to coa­ with s pecie s such as Tabebllia stellocalyx (Wild calabash), lesce with the main savanna ( P lates 4 and 5) . Chrysobalalllls icaco ( Fat pork ) and Mauritia flexllosa (Moriche palm), the latt er producing pneumatophores i.e . Marginal stability versus sava nna expansion spike- like aerial roots at ground level , within the plot. Repeated observations along marginal areas at the Aripo Th e other tra n sit ion plot in Savanna Y, which is located in Savannas showed a great variety of successio nal or retrogres­ the north-west sect or, has two abun d ant plants, Paspalllfll plI[­ sional s tages that d eviated from the sharp l y defined ecoton e chelflllll an d Rh yJ/chosp ora globosa both of which are tru e that charac terizes a stable climax for the forest/savanna com­ savanna s p ecies . In addition t o Paspa/llm, several tall grasses plex ( Plate I ). These d eviation s a re a result of fire disturbance form a dominant herbaceous cover, PaIJiCll11l cyaliescellS, P. (Sc h wab 1988, Comeau 1990). From this var i ety, three ge ner­ {)arl'ifolium, Alldropogoll virgat lls and A. lellcostachYlIs, al patterns emerged. First, tran sition areas that are reverting which see m s to have s lowed succession toward s savanna. back to fores t ( Pl ate 3); seco nd , tran sition a reas that are Hummocky microtopogra p hy characteristic of marg inal areas , becoming more sava nna -l ike ( Pl ates 4 and 5); and third, tran­ combined wi t h hi gh water tables, creates wet hollows where si tion areas that are leaning in the direction of sava nna but algae flourishes. T he relict fo res t tree Mauritia flexl/osa have become arrested in their development b y species domi­ ( Moriche palm) adapts to thi s wetness by producing pneu­ nance and pronounced micro-relief (Plate 6). matophores. To test this patte rn hypothesis, the plots set up in the 26 Living World 1997-1998

Table 1 - Total species numbers and species percentages within the plots; duplicate species are excluded.

Spec i es P lot s FS6E FS6W TS6 TS 55 T55N 055 To t a I 63 67 44 42 36 22 Total F 35 36 1 8 4 5 1 Tota I F+S 9 10 8 10 6 2 Total M 8 8 9 8 6 0 Total S II 1 3 2 1 7 1 5 1 8 To t a I D 0 0 7 3 4 1 % F 55.6 53. 7 40.9 9.5 1 3.9 4.5 % F+S 1 4. 3 14. 9 I 8.2 23.8 1 6. 7 9 .1 % M 1 2. 7 1 I. 9 20.5 19.0 1 6. 7 0.0 %S 1 7.5 19. 4 4 . 5 40.5 41.7 8 I. 8 % D 0.0 0 . 0 1 5. 9 7 .1 1 I. 1 4.5

L egend: Species category F::: forcst species S = savan n a species M= marginal species F+S = forest and savanna species D = disturbance specics

Pial designation FS6E = Forest/savanna 6 east FS6W = F orest/savanna 6 west TS6 = Transition Savanna 6 TS5S = Transition Savanna 5 south-east TS5N = Transition Savanna 5 north-west OS5 = Open Savanna 5

therefore, the least sim il ar in species composition. Conversely, OS5 has the greatest similarity (index 41) with the combined savanna phase of the control plOlS in V I ( CSP6), again what would be expected. What is significant is the wide Plate 6: Savanna expansion has been slowed by the dominance of discrepancy, i.e., difference, between the low similarity (ind ex th e tall grasses in the foreground, such as Panicum cyanescens, P. 6) of OS5 and the transition plot in Savanna VI (TS6), and the parviforum, Andropogon virgatus and A. leucostachyus. higher similarities (31 and 34) be t ween OS5 and the transi­ tion plots in the SE and NW sect ions o f Savanna V. TS5S and transition zones were compared w ith the cont rol plots (see TS5N respectively. T his substa ntiate s the contention that TS6 Table I). The transition plot in Savanna VI ( TS6) which re p re­ is revening back to forest and that TS5S and TS5N and sents re-es t ablishmen t of forest has 41% forest species and becoming more savanna- like in their species composition. To only 4 .5 % savanna species while the transition plot in the SE further reinforce this contention, a similar discrepancy exists sec t ion of Savanna V (TS5S) which represents est ablishment between the low similarity (index 11) of CSP6 and TS6, and of savanna has 9.5% forest species and 40.5% savan na species, the much higher similarities (46 and 43) between CS P6 and a reve rsal of the trend in TS6. The transition plot in the NW TS5S and TS5N respectively. section of Savanna V (TS5N) which represents arrested devel­ opmen t of s avanna has 14% forest species and 42% savanna Ta bl e 1[ - S~rensen Similarity Matrix species. percentages which show simila ri ties to those for TS5S. Plot s The trends are what would be expected from the observed CFP6 TS6 TS5S TS5N CSP6 OS5 general pallerns. The three transition plots (TS6, TS5S, TS5N) CFP6 also have the highest percentages of marg i nal and distur­ TS6 47 bance species compared with the con lrol plots, namel y FS6E. TS5S 24 35 FS6W and OS5 (see Table I). In fact. the two forest/savanna TS5N 19 32 52 control plots of V I have no disturbance species while the open CSP6 28 II 46 43 savan na plot of V has no marginal s pecies. This reinforces the OS5 2 6 3 1 l4 41 validity of the controls. If we look at th e S0rensen similarity matrix (S0rensen L egend: CFP6 = Combined Fores t Phase 6 1948) for the various plots (seeTable 11), the least simi larity TS6 = Transition Savanna 6 (index 2) occu rs be t ween the combi ned forest phase of the TS5S = Transition Savanna 5 south~easl control plots in Savanna VI (CFP 6) and the open savanna con­ TS5N = Transition Savanna 5 north-west trol plot in V (OS5). This result is expec ted as these plots rep­ CS P6 = Combined Savanna Phase 6 resen t the extreme opposites of stable vege t ation and are, OS5 = Open Savanna 5 Savanna Expansion 27

Key species in the transition process Comeau, P.L. 1990. Savannas in Trinidad. Living World, J. T here a re sever al good indicator species that o nce we ll Trinidad & Tobago Field Naturalists' Cil/b. pp. 5-8. establ ished in ecotonal areas a ffect the direction in which suc­ Eiten, G. 1972. The cerrado vegetati on of Brazil. Bot. Rev. cession wi ll proceed. These inc lud e /l ex arill/ens;s 38: 20 1-34l. (Siscuitwood) and Isertia parviflora (W ild ixora) which favour Leotaud , N. 1992 Ecotonal dynami cs between Aripo the re-esta bli shment of forest while the grass Paspalllt1l pul­ Savannas and sur rounding mars h fore s t. (Research chellum, and the sed ges Rh y n chospora barbarta and Report BT39A), The University of the West In dies, Lagelloca rplls rigidll s would influence succession to proceed Trinida d. toward sava nn a-l ike conditions . Marshall , R .C. 19 34. The physiography a nd vegetation of Tinidad and Tobago - A study in pl a nt ecology. Oxf ord Conclusions Forest. M em. 17. Stro ng floristic ev idence exists for savanna expans ion fol­ Medina, E. 1982. Ph ysio log ica l ecology of neotropical lowing repeated disturbance in marginal co mmun iti es at th e savan na plants. In Huntley, B.J. and Walker, S .H. (eds). Aripo Savannas. The freq uency of fire seems to be one of the Ecology of tropical savannas, Ecol. Stl/d. 42, pp. 308-335. determining factors that cont rol the direction s uccess ion will Springer Verlag, Berlin. take. Time- lapse aeri al photography wou ld provide addition­ Myers, J.G. 1933. Notes on th e vegetation of t he a l s upport for thi s evidence. Ve nezue lan lla nos. J. £col. 2 1: 335-349. Quesnel, V . 1979. T he Aripo Savannas. Trinidad Natl/ralist Acknowledgements 2(11 ) :25-29. T he a uthors wis h to th ank Winston J oh nso n, chief tec hni­ Richardson , W.D. 1963. Observations on the vegetation cal officer at th e National Herbarium , for his generous sup­ a nd ecology of the Aripo Savannas, Trinidad. J. Ecol. 51: port in providing field assistance and helping with pla nt iden­ 295-3 13. tificat ions, and to Yasmi n Comeau, Curator of th e He rbariu m, Sarmiento, G. 1984. The ecology of neotropical savannas. for prov iding us wi th working space and techn ical assistance, Harvard Univ. Press, Cambridge. and fo r proof-reading the manuscript. Sarmiento, G. and Monasterio, M.1975. A critical con­ sideration o f the e nv iro nm e ntal conditions associated with the occurrence of savan na ecosystems in tropical References America. In Golley F.B. and Medina, E. (eds). Tropical Ecological Systems - Trends in Terrestrial and Aquatic Ahmad, N. and Jones, R . L . 1969a. A plinthaquult of the Research, II , pp. 223-250. Springer-Verlag, N.Y. Aripo Savannas, North Trinidad: I. Propeni es of the soil Schwab, S.1. 1988. Floral and faunal compos iti o n, pheno lo­ and chemica l compos it ion of the natural vegetat ion. Soil gy, a nd fire in the Aripo Savannas Scientific Reserve, Sci. Soc. Amer. Proc. 33(5): 762-765. Trinidad, West Indies. (M.Sc. Thesis). UI/iv. Wise. Stevens Ahmad, N. and Jones , R.L. 1969b. A pl inth aquul t of t he Point. Aripo Savan nas, North Trinidad: II. Mineralogy a nd gen­ Sorensen, T. 1948. A method of establishing g roups of esis. Soil Sci. Soc. Amer. Proc. 33(5): 766-768. equal amplitude in plant sociology based on similarity of Beard, J.S . 1946. The natural vegeta ti on of Trinidad . species content. K. Dansk. Videllsk. Selsk. Bioi. Skrift Oxford Forest. Mem. 20. 5(4): 2- 16 , 34. Beard, J.S. 1953. The savanna vege tation of nort hern trop­ Williams, R.O. et al. 1928- The Flora of Trinidad and ical Ameri ca. Ecol. Monogr. 23(2): 149-215. Tobago. Governmen t Printery, Port of Spain . 3 Vo lu mes. Blydenstein, J. 1967. Tropical savanna vegetation of the llanos of Colu mbi a. Ecology 48( 1): 1-15. FOOD AND FEEDING HABITS OF THE FRUIT-EATING BAT CAROLLIA PERSPICILLATA

Victor c. Quesnel po. Box 47 Port of Spain

Introdu c tion cayenllensis has slightly larger fruit than Vismia falcara. However, I On 16 Jul y \992 I realised that a small group of bats, perhaps three could not absolutely identify the two species by th e sizes of the or fo ur, had been hanging up in a corner of my bedroom to feed. They discarded fruit skins, so I decided on the species collected by looking at did not roost there during the day but used the spot as a te mporary the species in fruit at the time. The two species of Piper were identified roost for feeding only. Getting rid of th em I thought might be difficult from plants grown from th e seeds of discarded fruit and faeces, as well so I decided to study their feeding behaviour instead. To do that I sim­ as from recognisable portions of di sc arded fruits. I could not reliably ply put a sheet of newspaper below the roost to coll ect the discarded identify the species from the appearance of the seeds alone so in Table food and faeces and examined the collection later. At first I did Ihis the I the two species were not separated. Pothomorphe seeds also could next day; later the collections accumulated over several days. My inter­ not be reliably distinguished from Piper so I relied on the examination est at the start of the project was simply the food they ate, but when of full- grown plants for ripe fruit when deciding presence or absence. casual obserations suggested that feeding acti vity was much greater A single spat, dark blue in colour, contained seeds resembling those in afte r mid-night than before, I changed the paper at mid-night so I could th e family Melastomataceae and this is e nt ered in Table 1 with a separate the coll ections into one before mid-night and one after. At first question mark. I could not identify all the fruit or eliminated seeds I found, but by com· The "while pulp" referred 10 in Table I sometimes contained seeds, paring them with fruit I could pick on nearby trees and by growing sometimes not. Four plants were grown from the seeds and later iden­ plants from the seeds, 1 eventually identified almost all of the species ti fi ed as Philodendron acuratum. It is probable that this pulp at differ­ involved. enl times cont ain s the fruit of other monocot species such as Monsrera All the observations were made at thi s o ne site at Haven Hi ll Farm, spp and Musa spp; I have seen the bats eat ripe plantains. I have also Talparo. On 16 May 1993 the bat was identified as Carrolfia perspiciflata seen the bats collect the fru it of Passiflora tuberosa a nd Pothomorphe by April Allgaier. peltata which aided the identi fication of some of the droppings. Not all th e food is fruit. Partly-eaten leaves of Cassia bacillaris The food turned up so frequently that it is hard not to think of them as deliber­ Most of the food consists of fru it, and Table I records th e species, ately sought rather than accidently gathered. C. bacillaris flowers, and their monthly distribution throughout the study . The droppings though not so frequent as the leaves, also seem to be deliberately consisted of whole green fruit, portions of uneaten fruit, fruit skins that sought. The re were at times other leaves, which remain unidentified. were more or less complete, bits of skin, calyces, seeds, pulp, and fae­ During May and June when beetles are usually common, wings, elytra ces containing seeds and pulp. Some frui t were comparati vely easy to and legs of a fairly large reddish brown beetle were frequently found recognise, some were not. and again it is hard to think of these beetles as not deliberately sought. There were two species of Vismia and two of Piper. Vismia The condition of the uneaten parts of the fruit sometimes gave h-able I. Foods eaten by Garoflia perspicillata from 1992 to 1994 Year ,. Species J.," A•• 'ep 000 Nov 000 Jan" Feb MU Ap< May JUn JU' ". 'ep 00< Nov Oeo J~ Feb Ku Ap< Kay J .. Vismia lalcata b • • • • Piper spp c • • • • • • • • • • • • • • • • Banara guianensis b • • • • Pothomorphe peltata c • • • • Cecropia peltata c • • • • • • • • • • Rollinia exsucca c • • • • • • • • • • Vismia cayennensis b • • • • • • • • • • • • • • • • • Cassia bacillaris 1 • • • • • • • MWhite pl.llpM P . • aCl.ltatl.lm c • • • • • • • • • p • • • • Castilla elastica • • • • • • • • • Pllssillora tuberosa b • • • Melastome ? b • • • trigonata f • Ficus • Psidium guajava b • Acnistus arborescens b • leaf, n • Cassia hacillaris • Insect parts • • • • • • • • • • • • • • • • • • • •

b _ berry , c • composite fruit, f • fig , , ~ legume , p • fruit of partially fused carpels, n flower Feeding habits of Garollia perspicillala 29 information about the bats' manner of feeding. The uneaten parts of Table II Numbe rs a nd distri bution of v: fa/cala fru it eaten by the Vismia fruits consisted of the skins with perhaps the calyx and stalk C. perspicillala in the period 30 Juiy-24 Aug. 1992 NM 30 J uly, attached; the bats evidently squeeze the fruits, suck out the con tents FQ 5 August, FM 13 August, LQ 21 August and spit out the skins. The fruits of Banara guianellesis were t reated in the same way but because the fruit skin is not so tough as that ofVismia Per iod B e for e mi dnig ht Af t e r mid nig h t the skin was often torn into several pieces. On the other hand, the berry of ACllistus arborescens seems as a rule to be swallowed whole because 30 July · 6 August· 26 III I have never foun d discarded skins. The fruits of Rollinia exsucca were 7 Augus t - 12 August+ 48 163 bitten open, the soft central parts with the seeds consumed and the 13 August· 19 August# 161 harder material discarded. Although the entire fr uit of Cecropia pelta­ 20 August - 24 August "\3 79 ta seemed to be eaten sometimes, at other times the central columns (rhachides) were discarded. The difference in treatment may be due to '"Excluding I Aug .. 2 Aug.; +Excluding 10 Aug.; NExcluding 16 Aug. the difference in size of the bats in the group fo r it seemed to consist of a mother and juveniles. The rhachides of Piper and PotllOmorphe were 429 fruits consumed in the firs t period and 229 in the second. Tested regularly discarded. The seeds of Cassia bacillaris with adhering pulp by X2 (X2 = 60.8) they were significat ly different at Panimal 0600 , rather than a leaf-eating one; fruits have evolved to make themselves , attractive to herbivores whereas leaves have evolved mechanis ms to , , di scourage herbivory. , The finding that more food was consumed in the s ix hours after , , midnight than th e six hours before came as a surprise. I draw atlention , l< , to the fact that I did not actu ally watch the bats feed. More bats may , "~ have come to the roost after midnight than before, and thus account 0 2400 ------~------w MOONLlGHT!HOURS ! for the difference, but I think this unlikely. Goodwin and Greenhall ::; F ( 1961) state that "Carol/ia perspicillata probably feeds twice during · each nigh!." The present results may be regarded as giving some s up­ · port to this supposition th ough there is nothing in them to rule out the · possibility of more o r less continu ous feeding. But even so, they don' t · give any clue as to why they shou ld feed in two bou ts, or why there ·: 1800 should be more feeding in the second bout than the first. How do the NM FM NM bats spend their "free" time in the period before midnight? In increased a La searching for food or in some other entirely unrelated behaviour? In a MOON PHASE population feeding night after night o n the same food as th ese bats were doing, it would seem that the bats should know where to find th eir Figure 1: Idealised diagram to show the re lationship between food and should have to spend relatively litlle time on searching for moon phase and the number of hours of moonlight between dusk and dawn. new sources. So my belief is that they spend most of the "extra" time in NM - New Moon FQ - Rrst Quarter FM - Full Moon searching the ir surroundings for new roosts, in dominance rivalries and LQ - Last Quarter in sexual behaviour. The increase in the amount of food taken at full moon compared References with new moon also came as a surprise. The authors of two popular books (Pete rson 1964, Schober 1984) mention that the eyes of the fru it­ Freeland, W.J.and D.H. Janzen . 1974. Strategies in her­ eating bats of the Old World tropics have special adaptations for seeing b i vory by mammals: the role of p lant second ary in the dark and that these adaptations help the bats to find their food. com pound s. Amer. Na t. 108(961); 269-289. There is no mention of anything similar in the fruit-eating bats of the Gooodwin, G.G. a nd A.M. Greenhall. 196 1 A review of New World, bu t the present result s show that they too use their eye­ the bat s of Tri ni dad and Tobago. D escript ions. rabies sight. By the X2 test th e increase in feedi ng near fu ll moon as compard infection, and ecology. Bull. Amer. Mu s. Nat. His!. with that near new moon is significant and related to the presence of 122(3); 191 -301 p l us plates 7-46. moonlight. From first quarter to (ull moon th e moon is visible during Greenhall, A . M. 1956. The food of some Trinidad fruit bats the whole period of darkness up to midnight and with each passing (Artibeus a nd Carollia). J. Agric. Soc. Trin. Tobago 56; night becomes visible for an increasing ly long time in the period from 424-447. midni ght to dawn until at full moon it is visible all night long. From full Greenhall. A.M. 1957. Food p references of Trini dad fruit moon lO last quarter the moon is visble during the whole period from bats. J. Mammal. 38(3), 400-401. midnight 10 dawn but is visible in the period from dusk to midnight for Moroney, M.J. 1956. Facts from Figures, 3rd Ed., Penguin decreasing lengths of time. During Ihe whole period from first quarter Books, H armmdsworth, 472 pp. to lasl quarter there is much more moonlight than in the period from Peterson, Ru sse ll. 1964. Silently by Night. Mc Graw Hill last quarter to first quarter (Fig. I) and the bats increase their feeding Book Company, New York. significantly in this period. The conclusion mu st be drawn that the Schober, Wilfried 1984. Th e Uves of Bat s. Croom H elm. moonlight helped the bats to fi nd their food through the use of their Lo nd on and Canbe rra. eyes. OBSERVATIONS ON THE MIGRATORY PATTERNS OF THE BUTTERFLY PHOEBIS STATlRA (CRAM) IN NORTH AND SOUTH TRINIDAD

Dave D. Chadee Insect Vector Control Division, Ministry of Health, 3 Queen Street, St. Joseph, Trinidad, West Indies and Trinidad Public Health Laboratory, 16-18 Jamaica Boulevard, Federation Park, Port of Spain, Trinidad, West Indies.

The yel low migrant , Phoebis statira (Cram) i s d ist ri bu ted small s warm fl yi ng from west to east. The time of thi s activity throughout Latin America and the Caribbean region, in c lud­ was 11 .45 a m. Figure 1 s hows the fli g ht palCern s obser ved fo r ing the Guianas, Brazi l , Venezuela, Ecuador, Peru , Panam a and P. statira at Table land and at C hamps F leurs. Trinidad , West Trinida d a nd T o b ago ( Barcan!, 1970). PllO ebis s ralira u sua ll y Indie s. are g reatl y a ltracted to flo wer s and are fo und generally in Based on the observations of these three days it wou ld open a reas with s un s hine. In addition . they u s ually mi grate appear that this annual m igratio n of P. statira fo ll owed t he west each year w ith the o n set of the ra iny seaso n and thi s migra­ to east d i rection ( Fig . l ) but occ urred in August ( 1995) rather tory behaviour h as b een documented in 1918, 1919, 1926, tha n in June, t he mo nth rainy season begins, as reported b y 1932,1947,1954,1962 (Ques nel 1971) and in 1960 i n R io Barcant ( 1970). It is c lear that the migration o f this butterfly Claro ( Barcan! 1970). In 1969, obser vations we re made al populatio n must be advantageous to the s pec ies. For example, Nariva, P eti t Valley, Valsayn Park, Sangre Grande. Manzani lla, migrati on prov ides a means of avoiding adverse environmen­ a nd Mayaro (Quesnel 197 1). Thi s pape r d esc ribes th e mi gra­ tal co ndition s a nd thus may e nhance fi nd ing alternative habi­ t ory patte rns of P. statim observed over three day s in Augus t tats, there by inc reas in g th e butterfl y'S li fe -hi story options. 1995 in the s mall village of Tableland , locat ed eas t of San However , migration carries costs as well as benefits. Fernando on the Naparima-Mayaro Road and a t Champ Fl e urs, a town loc ated a lo ng the Eastern Main Road in north 6,°I Trinidad ( Fig. I ). 45' Fro m 10.40 a.m. on Saturday 19 August, 1995 a swarm o f P. statira fir s t appea red in a band or co lumn a pproxim a t e ly 50 m e tres wide a nd from approximately four m eters above ground leve l to jus t below the tree tops circa 25 m. The s un­ s h ine was brilliant a nd their characteristic bri g ht yellow green coloration was ver y evid ent. Swarms of five to 10 butterflies were seen while ma ny were scatte red in the moving col u mn , wi th a definite lead group and f requent changi ng of posi- tions as indi v idual s moved in a zigzag p attern withi n the 10°____ + ______) group. It i s es tim a te d th at over 9.000 P. statim butt e rflies 3d we re observed mi g rating in a n easterly di rec t ion with numer- o u s pairs observed copulating o n the wing. A t n oon rai n N began falling w hic h resulted in the complete cessation of flight. As the rains s ubs ide d at 12:45 p.m . the swarm in g activ- dl{;J,'fi:,p.!;(.: :...... ity re-em e rged until 2 .05 p.m. after w hic h o nly a f ew s trag­ ~:'i;.{,,~i(;: •..: ' ...... '. ... -', ..: .... t, .• ~~ .: :.; .. ',.:.:.:::;:'" g le rs could be observed. O n Sunday 20 August, 1995 act i vity bega n a t 10.00 a.m. Moyoro but th e den s it y of the s wa rm was considerably reduced. On both days nume rous butte rflies were observed f eed in g on the nectar of both orna me ntal and wi ld fl owers alon g the road­ s ide. The co lumn of P. statira was followed by car a long the Napa rim a M ayaro Road and numerous clusters were observed drinki n g f ro m puddl es and p ools o f water at Poo le and San ~ Possible source of butterflies Pedro. I n add itio n , m a n y P. statim were seen as fa r eas t as May aro, located o n th e south eastern coast of Trinidad. o Likely track of migrating buHerflies (Fi g. ] l. I n additi o n . o n Monday 22 Augu st 1995 while dri vi n g Figure 1. Map of Trinidad showing path of migration of Phoebis statira a long the Eas t e rn Main Road at Champ Fleurs I observe d a butterflies in August, 1995. 32 Living World 1997-1998

These include: diatel y and any reproductive delay due to mig ration are min­ a. t he actual metabolic cost wh ich involves the uti­ imized. In addition, based o n these three observations we are lization of energy to support flight: un abl e to d et erm i ne th e exact breed i ng habitats of P. statira b . risks o f increased predation and not finding a but their eas t ward migr a ti on suggest that new habitats are sui ta b l e habitat, and fou nd. However, it is unlikely that the next generation woul d c. potential reproductive cost owin g to increased migrate fu rther eastward because this wou ld ta ke the butter­ time to firs t oviposition, decreased energy reserves fly population over the Atlantic Ocean (Fig. I ). There is a pos­ ava i lable for reproduction, s h ortened lifespan, sibili ty that follow i ng t he eas t ward migration and copulation and/or a d ecrease in overall fecundity. the P. statira popu lation may migrate wes t wards but this However, the observations of cop ulating pairs of P. statira mi gration event has never been recorded and no literature is at Tab lela nd and Champ Fleurs sugges t a n oogenes is-fl ight currentl y available on this unique migration pattern. syndrome wh ich also has been observed amo ng ma n y in sects It is recommended that further o bservat ions be conducted (Rank in and Burchsted 1992). Th is behaviour activity (copu­ on the P. statira mi grati on patterns, population dyn amics and lat ion) s uggests that swarmin g increases or enhances maJe­ kinetics to examine f urther the oogenesis-fl i ght syndrome, fema l e contact and thus provides better co pul ating opportu­ the body size requirements for migration, metabolic costs, n it ies. In addition, with migration, the weak and genetically fli gh t fue ls and the re producti ve cost of flight. inferior indiv iduals usua ll y die, o r fa il to successfully join the swarm o r mate, thus, leaving o nl y robust males w ith repro­ References ductive vigour to copulate with females, thereby improving the P. statira genet ic const itut ion . Barcant, M. 1970. Butterflies of Trinidad and Tobago. It should be no ted that P. starira butterflies were observed Collins, London, UK. mi g rating a nd copu lati ng a t the same, time t hu s providing a n Quesnel, V . C . 1971. L ep idopteran migration i n 1969. The evol ved advantage. for in mos t populat ions flight and repro­ Trinidad and Tobago Field Naturalists' Club Journal pp . duction are physiologically an t ago ni sti c ( Rankin a nd 48-51. Burchsted 1992). Thu s, w hen the P. statira migrants arrive a t Rankin, M.A. a nd J. e.A. Burchsted. 1992. The cost of thei r new habitat they a re ready to begin reprodu ction imme- mi gr ation in insects. Ann. Rev. Ent.37:533-559. THE SKIPPER BUTTERFLIES (HESPERIIDAE) OF TRINIDAD PART 9, GENERA GROUP E CONCLUDED (THIRD SECTION) WITH A DESCRIPTION OF A NEW SPECIES OF CLITO

MATTHEW 1. W. COCK CABI Bioscience UK Ce ntre (Ascot) S ilwood Pa rk, BuckhurSl Road, Ascot, Berks SL5 7TA, UK

Introdu ct i on

This is the n i nt h in a con tinuing series on the ide n tificat ion and b iology o f t he Tri nidad Hesperiidae. It continues direct­ ly from part 8 (Cock 1996) . In the neXI part I plan 10 treat the remaining Py rgi nae. comprising Evans' (1953) ge nera groups F and G. I reiterate my thanks to Dr C. Denni s Adams, Mrs Yasmin Comeau, Bharai Kallaa an d Winst on Johnson of the National He rbari u m who identified the plan ts from which I reared Plate 1 Milanion hemes a, Cumberland Hill, 14.xii.1978 Hesperiidae in Trinida d . Mr Carlos Lo pez-Vaamonde kindly c hecked m y tran slation of the description o f the biology of specimens, a nd all 24 s pecime ns for wh ich I have data we re Aflligonus erosus from Comstock a nd Vazquez ( 1961 ). The captured between lale November and mid-Apri l , s ugges ting foll owi ng have very kindl y assisted in prov id ing access to the that it fl ies predomi nantly durin g the dry season . coll ection s in their care: Dr George McGav in of the Hope At Be lem, Moss (1949) records the food plant to be Entomological Coll ections, O xford Univers ity Museum (HEC), Rollillia orthopetala and o ther Annonaceae, but gives no other Dr Phi ll ip Ackery of t he Natural His tory Museum (NHM) (for­ details. In Trinidad , SAS and I have t he food plant to be R. merly British Museum (Natural History», Dr Mark Shaw of the multiflora, known locally as w il d cash imar o r wild sugar apple Royal Scottish Museum (RSM ), Mr Scot! Alston-Smith to his (Freeman a nd Williams 1928). In February 1994 I co ll ected pri va te co ll ection (SAS), Professor Julian Kenn y and Dr Gene larvae a nd pupae on this food plant above Mount S I Poll ard of the University of the West Indies , St. Augustine Benedict's a nd above Fort George, an d reared o ut several (UW I). specimens. The plants used were a ll saplin gs growi ng beside I especia ll y thank Scott A ls ton-Smi th who has read and paths; they were no more than a metre tall , with leaves 25-35 commente d on this paper , and provided additional records mm w ide and 80-150 mm long. The detail s of the larval a nd from his collecti n g, and observ ations a nd food pl an t record s pupal s helters are no tewo rth y and distincti ve. wh ich have not previousl y been published (indicated as SAS The eggs are small wit h about 14 r ibs; they are laid on the in text). leaf lami na UNS, toward s t he base, 1-2 per leaf. The final la r­ val instar (Plate 2) is about 20 mm lon g; the head is chordate 10 7 . E46/1 Milanion "emes " emes Cramer 1777 in shape, and rounded; the s urface is covered with rugose Pl ales 1-5 dark projections; the head is li gh t brown, with the posterior This s u bspecies occ urs i n Venezuela , Trinid ad, the marg in dark brown; the face is covered with pale speck les, Gui anas (TL S urinam) and at th e mouth of the Amazon. Kaye and the most conspic uo us feature is a whi te, vert ical s treak (1914) records two Trinidad specimens in the coll ection of H near the apex o n each s ide of the face. The body is dull g reen J Adams; su bsequently (Kaye 192 1) he records addi ti o nal with a heavy overlay of white s peck les. The pupa is 14 mm s pecimens from Fondes Amandes (E J Patterson) a nd A riap ita long, rounded wi th a s ho rt , blunt frontal sp ike ; cre master Road (W J Kaye). long ( I mm) and s le nde r ; colour yellowish white, except The sexes are si milar. Thi s speci es is dark brown above abdominal sp iracles, probosc is sheath di s tally, a nd three w ith wh ite ma rkin gs; below it is a lighter shade of brown, but inter-segme nt a l areas a re black. t he markings are sim il a r apart from a n expa nd ed white spot The stage I shelter (Plate 3) is a 6 mm diameter ci rcular in s pace I B UNF, and the bases of both wings are sli ghtly flap cut fro m the middle of the leaf lamina and fo lded unde r paler. Illus tra tions in Barcant ( 1970, Fig.4, No.2) and Lewis as a con vex c up ; all of the flap and the co rresponding leaf UPS ( 1973. Plate 83, No. 38). No costal fold; F 013 mm, 'fl 12- 13 area except the line along which the larva rests are ske le­ mm. toni sed . There is littl e o r no fee ding associa ted with the stage This see ms to be an occasiona l, but widespread s pecies , I shelter apart from the ske letonising of the shelte r, suggest­ although perhaps over-represented in collect ions since it is so ing th at during thi s stage the larva may not leave the shelter, dist inc tive. It is perhaps more com mon in the north than in and only feeds within it. the south. Most record s are from fore sts a t intermed iate a lti- The stage II s helter ( Pla te 4) is normally made upon the tude, the h ighest altitude fro m which I have seen a specimen same leaf as the stage I s he lte r. It is about 15 mm x 7 mm; a be ing 518 m o r 17 00 ft. Al most a ll captures are of isolated wide flap is cut fro m the edge of the lamina to the leaf mid - 34 Living World 1997·1998

Plate 2. Milanion hemes fifth instar larva collected on Rollinia multiflora, Plate 4. Milanion hemes stage II shelter on Rollinia multiflora, behind behind Mount St. Benedict's on MI. Tabor Track, 28.ii.1994 Mount 8t. Benedict's on Mt. Tabor Track, 28.ii.1994 rib and is folded under along a s hort joi n incorporati ng a s he lter flap is s kele tonised apart fro m the two loose bi ts of main ve in, a nd then s haped by the larva as follows. At the leaf at each end; the s helter leaf UPS is ske leton ised in irregu­ di s tal e nd a notch about 3 mm lon g is cu t into the end , par­ l ar p atches a long each margin of the s heller. leaving the cen­ a l lel to the mid rib, and th e si des pu ll ed together to form a tral a rea under w h ich th e larva rests and the pupa is formed. keel. usually wi lh the leaf-flap on onc sid e of the notch pro­ Further feeding d is ta l and basal to t he s heller leaves only the jecting. whic h causes t he shel ter to be raised from the leaf basal portion of the leaf intact. so that the she lte r then looks UPS. At the basal end an arc is cut from the si de away from like a ske letoni sed fragme nt attached to the mid -rib . At pupa­ the shel ter hinge. about 3 mm short of the basal end of the ti on the inside of the shelter, but not the pupa, is lined with a loose white waxy powde r. I h ave exami ne d Moss's material in the N HM . The pupa i s formed in a globular s he lter up to 3 em long, with a fine lat­ ti ce-work of small holes except over the area whe re the pupa occurs. Thus. there see ms to be differences between t he pupal s helters from the two areas, perhaps associated with the diffe re nt food plant species.

Plate 3. Milanion hemes stage I shelter on Rollinia multiflora. behind Mount 81. Benedict's on Mt. Tabor Track, 28.ii.1994 she lter; the shelter is attached to t he leaf U PS a long this arc. leaving the basa l porti on of the flap free. The larva skele­ tonises the shelter flap and the leaf UPS except for a cen tral line un der which i t rests: there i s additional feeding nearby on the sa me leaf. The stage III shelter (Plate 5) is formed on a different leaf, Plate 5. Milanion hemes stage III shelter on RoJlinia multiflora, behind and it is also ll sed for pupation. A large flap about 35 mm Mount 81. Benedict's on Mt. Tabor Track, 28.ii,1994 lon g is cut from the margin to the mid-rib. and fo lded under joined a long most of it s length alo ng the mid-rib. old shellers may s plit along the mid -rib. The flap and l eaf are consumed MY/Oil to the extent that both are about 10 mm wide: an arc is cut All members of this genus a re predominant l y white with fro m about 5 m m from each e nd of the n ap, cu rved f rom ncar va rying diffuse brown markings. Two spec ies, lassia an d t he flap outer margin to the mid-rib; the shelter is then ander have w hite hyaline apical spots in F spaces 6-9. while formed by attach ing the inner edge of each arc to the leaf UPS. the re m aining three, menipplls, pe/opidas a nd jasoll do not. leaving the remainder of t he flap loose at eac h e nd of the shel­ M. lossia is smaller than M. onder and has a relat i vely clear­ ter, but joined to the s he lter at each side. The whole of the ly defined white di scal band which extends to the costa. In M. Skipper butterflies 35 ander, w hi ch is a darke r s pec ies. the discal band is shaded brown towards the cOS l a. M. menippus is si milar in size 10 M. ander, but not as dark. It has a more o r less c lear white discal band UPF, whereas in M. peJopidas and M. jason t h e band i s filled w ith d iffuse brown markin gs. M. pelopidas and M. jason are ver y si milar, although the genita li a are different: the cuiller of M. pelopidas is lon g. na r­ row and tapered, while that of M. jason is s h ort bro ad and rounded. The c haracter given by Evan s ( 1953) seems effec­ tive: the dark bar a l e nd of the F cell is direc ted to t he outer edge of the discal SpO I in space I B in M. pe/opidas, and towards the middle o r inner edge of t h is SpOI i n M. jason. The life history does not seem to have been recorded for

figures 1-2, Left Plate 7 Myfon lassia 9, Curepe, 9.xi. 1980 vaJve 01 male genitalia 01 Myton (1893) record ed fro m Trinidad and Kaye ( 1904. 1921) lists peiopdas Fabricius, wi thoul further comme nt. Kaye (1940) recorded M. lassia Chacachacare Is., under its correct name as " rather a common spec ies" with sev­ nr Iiglthouse on e ral r eco rds. flowers 01 The female of this species is larger than th e male and more 0mn0Iaena heavi l y marked in brown. The col ours are w hi te a nd brown as 0

Plate 8. My/on ander andrea d, Cumberland Hil l, 2.viii. 1981 Plate 10. My/on menippus r3, Rio Claro-Guayaguayare Road, milestone 4-5,17.ix.1978

Plate 9. My/on ander andrea 9, Cumberland Hill, 2.viii.1981

Plate 11. My/on menippus ~, Arima-Blanchisseuse Road, 19.iv.1982 on Morne Catherine (vii.1992), The young larvae are yellow green with a brown head . According to W i lliams ( 1929), "8. :.picara" is widespread on poor so i ls.

110. E50/8 My/oil me"ippus Fabricius 1776 Plates 10·14 Thi s com mon and widespread s p ecies is found from Mex ico to Argentina (TL Surinam). Kaye (1940) records it. under its syno n y m Elldamidas me/onder Cramer. as "a fairly common s pecies". wilh r ecord s from the soulh of Trinidad . A third white and brown s pecie s. and once again the female is larger and more heavily marked than the male. Illustration in Lewis ( 1973. Plate 83, No. 59. as M. me/allder, a synonym). No costal fold: F d 21 mm. ~ 22-23 mm. Thi s species seems to be more common in the sou th of Trinidad th a n in the north: I have II records from the south , one from Centra l and three from the north. Six of my seven specimens were taken at Eupatorium flowers , and I s u s pect from the dates of capture of the earlier collectors that they had similar experiences. All record s are from relatively undis­ turbed forest areas. Plate 12. My/on menippus male at flowers of Austroeupatorium inu/ae­ In October 1988, SAS reared this species from two folium, Brasso, 11.x.1993 Malpighiaceae on Morne Catherine: Hiraea /agi/olia and Skipper butterflies 37

Sligmapllylloll cOllvolvulifOlil/lII, while I have reared a specimen collected on H. fagifolia on MI. Taman a ( 14 .x. 1995). S. COII­ volvillifolillm is widespread in dry places in Trinidad, while H. fagifolia seems more local ised (Wi lliams 19 29). The following accoun t is based upon the specimen whi ch 1 reared. The final larval shelter was an elongate triangle cut diago­ nally from the edge of th e lea f lamina and fo lded over upwards with th e s hort si de adjacent to th e lamina edge; it was hin ged along a major leaf vein, and he ld over the leaf with several strands of silk extending from the margins of the flap over the leaf surface. The fina l inst ar larva ( Plate 28) measured 22 mm. Head chordate, broadly indent at vertex and flattened dorsally; ru gose, shiny: black, strikingl y marked in wh ite: a broad white band across face, ex tendi ng along epicranial suture to vertex, and lat­ erally with dorsal and ven tral extensions towards t he posterior margin of the head; within the white band is a black band across Plate 14. My/on menippus pupa, collected as larva on Hiraea fagifofia, the frons, extending laterally across the epicranium, this la teral Mt. Tamana, 14.x.1 995 portion deep ly divided by a white projection towards the frons. Tl conco lorous with body. Body dull, translucent green; do rsal This is anot her com mon and widespread species, found line s li g htly darker d ue to absence of underlying fat bodies; from Mexico to Paraguay (TL " Indiis"), Ka ye ( 1904) lists thi s broad, yellow dorso-Iateral line composed of vertical marks on anterior half of T2-A 8, a typical segment hav ing one quadrate mark and four narrow vertical bars. Spiracles pale, at centre of star of trachea, and linked by weak line of trachea visible through cuticle. Legs and prolegs concolorous. The pupa (Plate 14) measures 18 mm and is supported at the cremaster and with a Y -shaped si lk girdle. It is s moothly con-

Plate 15. My/on peJ10pidas Ch,.ealeh"eaJe Is., nr. lighthouse, 7.i. 1982

Plate 13. My/on menippus fifth instar larva collected on Hiraea fagifolia, Mt. Tamana, 14.x,1995 toured, the eyes slightl y protuberant; a few weak setae on eyes. The c uticle is transparent showing the underlying g reen colou r; b lack markings as follow s: a thin line along the costa F; spot at base of F; spot dorsally o n fron s: row of three spots across front of head. the central one di splaced ventrally; faint spot posterior and ventral to eye. Spi racle T I proj ecting , cons picuou s; black posteriorly and laterall y. pale anteriorly. Abdomen spiracles Plate 16. My/on pe/opidas !f!, Moruga Bouffe, 23.v.1982 concolorous. Pupat ion lasted 18 days, III. ESO/IO MY/Oil p e lopida s Fabricius 1793 s pecies un der its sy nonym Elldamidas 01.ema But l er and Plates 15, 16 reco rd s a captu re by WE Broadway, and another fr o m 38 Living World 1997-1998

Tunapuna; later he added th a i it i s co mmoner in the south ( Kaye 192 J ). It also occurs on Chacachacare Is la nd (Cock 1982.). 1 em This is a white species, overlaid w ith pale brown diffuse markings; the female is larger with heavier and less diffuse markings. No costal fold; Fa 19-21 mm , l? 21 mm. I agree with Kaye (1921) that thi s spec ies is commoner in the south, but even there it is not common. Several of the locality records s ugges t this species is more associated with d i stu r bed forest than the last , e.g. secondary fores t in San Miguel Valley, Arima, Maraval, Waller Field, near summit of Chacachacare Island, Moruga Bo uffe. Although I have s peci­ mens taken at flowers of Eupatorium and Cordia, this species does not seem to be so closely t ied with flowers as the last. SAS has reared this species from larvae collected on Combretllm fruticosllm (Combretaceae) at Los Bajos ( iL1990). Plate 17. CHlo lrinidadensis n. sp. paratype a , La Laja Ridge, 17.iii.1982 C. fruticosum occurs on hillsides and in thickets in moist and dry districts (Williams 1932). The larvae of M. pelopidas doubl e arc spot at centre, space 2 a basal V -s haped spot, have the head cho rdate, light brown , with the vertex dark space 3 a basal quadrate spot, spaces 4-8 a row of spots except brown a nd five streaks, one centrally, and t wo laterally, that of s pace 5 dis placed distally, a U shaped spot at e nd cell. angled inwards towards mouth parts; the body ground colour UPH brown; discal area, consisting of a patch in space Ie at is green, but an overlay of fine yellow dots renders the over­ about 11 3, distal half of cell, s pot near base of space 2, base all effect yellow-green; thin yellow lateral line. of spaces 3-6, a nd quadrate spot across space 7 at 1/2, pale , semi-hyaline, tinted viole t-bloe with brown veins; pale brown 112. E50/11 Mylon jason Ehrmann 1907 margina l and s ub- marginal fascias. UNF light brown; brown This s pecies is widespread from Guatemala to Paraguay sub-marginal fascia; margin crossed by brown veins; s pace I A (TL Venezuela), but not common in collections. Kaye (1940) to 4/5 near white; space I B distal to hyaline spot pale brown records it from Trinidad, under the name Elldamidas macaira except submargi nal line and vein I B which are brown. UNH jason, on the bas is of speci mens collected from Port of Spain (lO.viLI927, Forbes) and SI. Ann's (27.iiLI929, Huntingdon). I have no t seen the two specimens li sted by Kaye. Evans (1953) lis ts a female from Trinidad in the NHM collected by A 1 em Hall at St An n 's, xi-xiLl931. I know of no further captures. Very similar to the last, but distinguished as indicated under the generic account above. In the field the two speci es are likely to be indistinguishable. On recent visits to Trinidad I have made a point of co ll ecting specimens that appear to be M. pe/opidas in the hope of finding M. jason, but thu s far all material has proved t o be the former species, sugges ting that M. jason is not co mmonly encountered in Trinidad. Further collecting. part icularly on the hills around Port of Spai n may lead t o the redi scovery of this species.

113. E5211 Clito trinidadensis, n. sp. Plates 17, 18 In Cock (1982b), I li sted thi s species as C. lillera fitlera Mabille, a new record for Trinidad. On c loser exami nation, Plate 18. CHto trinidadensis n.sp holotype <1 UNS, Marne Catherine, 24.iii.1982; and in light of de Jong's recent description of C. jonkersi from Surinam (de Jong 1983), 1 am now of the opinion that the brown; spaces IA . IB and Ie pale brown in basal 2/3; disc as material from Trinidad also represents a new species, and UPH but paler v io let with brown veins; dark brown s ub-mar­ accordingly I describe it h ere. ginal fascia. Fringe of w ings brown. Body above brown, except margin of collar, eyes and seg ments of labial pal pi

Description. Male. Plates IO ~ ll. F 16-17 mm. UPFbrown; which are pale. Thorax be low grey-brown; head be low white. diffuse dark brown pre-discal fascia in space I; si milar sub­ Antennae dark except anterior margin of shaft which is che­

marginal fascia; margin slightly paler. crossed by dark brown quered with white, and pure white basally; nudum dark chest~ veins; white hyal ine spots in spaces 1-8 and ce ll: space I a nut brown, 21 segments of which 2-6 cou ld be considered on Skipper butterllies 39

Fogures ~ male genitalia of CHlo lrinidadensis n. sp. hoIotype (Trinidad, w.1., Marne ealhelire, 24.'i.1982, MJW. Cod<).

4, genitalia _ valves and aedea- gus removed, lateral view; 5, left va/ve ntemaJ view; 6, Lr1CUS and!1"'1hos _lieN; 4 , mm 7, aedeagus, _lieN; o 8, aedeagus, Ia!eralIieN from left { 7 " 8 the club. Costal fold 5-6 mm; reaches level of c lub. Discussion. de J ong ( 1983 ) pointed Oul th at Evan s' (1953) Ma l e genitalia. Figures 4 8. Uncus with a pair of lateral diagnos is of the gen us Cliro as having 16 nudum seg me nt s is lobes ncar base. Gnalhos broad, turned up s harpl y at apex. erroneous, based upon the material in the NHM arranged by Cucu llu s slender, the tip heavi ly toothed. No costal process. Evans which has 18-22 segment s. The nudum of 21 seg ments The ae d eagus is pointed, with an ang le d ventral p rojection for C. trillidadellsis fits well in the range for the genus. near apex , a sci erotized dorsal ridge towards the apex and a s trikin g pectina te cornutu s. Habitat. The localities for the type series include a hi ll top and a ridge top of the Northern Range of Trinidad. The south­ Female. Unknown. ern summit area of Marne Catherine at the time of these cap­ tures was a wide trail overshadowed by light semi-dec iduous Type material. Holotype male: TRIN IDAD, W .I. , Morne forest , a lthough in recent years, some of thi s Irail has been Cathe rine , 24.iii.1982, M.J .W. Cock. Parat y pe males: tarmaced and some has become overgrown. The La Laja Ridge TRINIDAD, W.I., Northern Range, La Laja Ridge, 17. iii.1982, site is a wi de , open track bordered by a mixture of, at that M.J.W. Cock; TRI NIDAD, W.I., Morne Catherine, 28.i.1980, S. time, abandoned s mallholdin gs and relatively undisturbed Alston-Smith; 3 ad TRINID AD, W.I., EI Tucuche. iii.1989, S. fore st. I don't know where on E I Tucuche the remaining A lston-Smith. Th e re are t hree spec imen s hom Trinidad in the paratype s were collected. The fact that at least nine speci­ A ngostura-Barcant collection in Lave ntille. Trinidad , but for mens of th is species are now known from the is land s uggests lack of rel iable data these are not included in the type series. this is not that rare a species within its se lected habi tat. The holotype will be deposi ted in the NalUral Hi s tory Museum, I retain the fir st paratypc, and the remaining 114 . E52/3 Clito clito Fabriciu s 1787 paratypes are in the coll ection o f S. Alston-S mith , Petit Valley, Plate 19 Trinidad . Apart from a s peci men from Guatemala, Ihi s spec ies is restricted to Trinidad , the Guianas (TL French Guiana) and Diagnosis. In appearance thi s spec ies is close to C. lirrera Mabille which occurs in Peru (ssp. liltera). French Guiana (ssp. a"da Evans) and Bo li via (ssp. lIebulosa Draudt) (Evans 1953) a nd C. jOflkersi de Jong from Suri nam (de Jong 1983). It dif­ fers in the dusky di sca l band of the hindwing which is white in C. jOllkersi an d C. liltera anda, and faint in C. lirteralittera. Hitherto, I have treated this species as C. Iillera littera (Cock 1982b), but the basal edge of the cell s pot of the fore w ing of C. littera littera is s traight and perpendic ular to the costa, Plate19. Clito clito r5, Spanish Farm, 12.x.1980; whereas that of C. tri"idadensis is so mewhat rounded, but clearly angled to the costa. Brazil; apart from a series fro m Belem, it is rare. I recently The male genitalia are intermediate between those of C. added thi s to the Trinidad li s t (Cock 1982b). Life hi story and jonkersi and C. iiflera. The uncus has a pair of late ra l lobes food plants unknown. appa rent in dorsa l a nd lateral view; C. jonkersi has s imi lar, This is a dark brown species with white markin gs . The but less well developed lobes. The valve is closer to C. iiflera, only species it might be co nfused with is MilaniOIl Ilemes, but although clearly differently proportioned, and lacks the costal they ca n be separated by the markings, size and wing s hape. process of C. jOllkersi. I do not know the female. Illust ration in Lewis (1973. Pla le 81. 40 Living World 1997-1998

No. 47). Costal fo ld very narrow , 3 mm long. F a 15 m m. intro duced thi s s pec ies t o the Trinidad li s t o n the basi s of a ma le which took at Cordia curassavica flo we rs ( 12.x. 1980) at Spanish Farm , Las Lomas (Cock 1982b). Recentl y I took a second male resting under a leaf in a sunlit c leari ng in Bu sh Bush (7 v.1995), SAS has seen specimens a l Valencia Fore st, and Las Cuevas Road. but has yet to c apture t hi s s pecies . C learly. a very rare speci es, pe rh aps associated w ith lowland forest. A.M. Moss reared thi s s pecies, although h e does not reco rd the food plant or an y detail s (Moss 1949) . Plate 22 Xenophanes tryxus r!, Plate 23 Xenophanes tryxus a I have examined hi s material in the N HM . The pupal s kin i s Cats Hill, 4.vii.1979; UNS, Moruga East, 24.ii.1980 transparent. and h e nce probably green in life; the TI s piracle projec ts a nd has a C s h ap e d rim. Illustration in Lew is ( 1973, Plate 88, No. 34. 9 ). No costal fold; F 0 14- 16 mm, 9 15- 15.5 mm. 114 a . ES2/S C lilo ze lot es Hewit so n 1 873 This s p ec ies is to be fo und occasionall y in t he low lands of l' latc s 20, 21 central and sou the rn T rinid ad, w ith ju s t two records fro m th e Thi s rare s pec ies is know n only from Fre nc h Guiana a nd va ll eys of the Northern Range. It is fo und in open , disturbed the Upper Amazon (Evans 195 3). It has no t previously been s ituation s inc luding o pen secondary forest. A di s proportion­ ate nu mber of captures see m to have been made c lose to the sea (e.g. Nariva Swam p. Iro i s Beach , Manzanilla, La Brea, Manzanilla Windbe lt Reserve). Moss (1949) found thi s s pec ies to feed upon Mafachra fasciata ( M alvaceae). a nd one or two o t her members of th is family . He records the larva lO be whit ish g ree n , f reckled a nd sli ghtly hairy with a roug h , sepia-brown h ead . and the pupa to be h a iry, warm brown. fre c kled and white powdered. In Rio Gra nde do Sui, Brazil, Biezanko (1963) and Biezank o and Plate 20 Clito zelotes rJ, Aripo Plate 21 Clito zelotes 9, Aripo Mielke ( 1973) record the food plant as Pa'llonia spin ifex Savanna, viii. 1986, S. Alston-Smith Savanna, viii.1986, S. Alston­ (Malvaceae). In Texas, which i s the northern limit of the Smith range of t his species, Ke nda ll a nd Ri c kard ( 1976) found the food p l ant t o be Mal 'lla'll i scus drllmmontiii (Mal vaceae). reco rd ed from Trinidad (Cock 19 82b). Ind eed , the fem ale In Trinidad , food plants incl ude Urena lobata, Mafachra d escribed a nd illu s trate d here does not see m to h ave been s p. and Hibiscus sp. (a ll Malvaceae). I have reared a male p rev ious ly recorded at a ll. Thi s new record fo r Trinidad i s found on U. fobara at Spani s h Farm (co ll ected l 3.x LI981 ), based upon capt ures by S Alston-Smith at Wall er Fie ld and and reared a l arva through to the distinctive pupa on Aripo Savanna. Malachra s p . from Saunder 's Trace (col lected Il .x. 1993), Male brown w ith wh ite hyali ne s pots. The broad pa le w hile SAS h as reared this s pecies from Hib i scus sp. On U. band H is yell ow in s paces I A-5, and w h ite in spaces 6 and 7. fobara, the larval shel te r was a nea rl y sq uare flap from t he The fe m ale has not been described b efore. It resembles the margi n of the leaf, fo ld ed over onto the top of the le af. Full m ale, but the band H is much red uced, a nd w hi te rather tha n g r own, the larva ( Pl ate 24) measured about 15 mm ; the h ead yell ow. Costa l fo ld narrow, 5 mm; F a 15 mm , 9 14 .5 mm. was rugose, black , a nd c hordate; body yellow-green dorsally, This rare an d locali sed s pec ies is only k now n in Trin idad with dorsal line darker; a thin white do rso- Iateral line; blue- from the l owla nd forest of east ern Trinid ad around Aripo Sava nn a a nd Wall er F ie ld. Captu res were made in Janu a ry 1986 (Waller Field) and August 19 86 (Ari po Savanna). Life h istory a nd foo d pl ants unknow n.

115. E53 Xellopllalles lryxlls Stoll 1780 Plates 22-24 This dist inc t species i s the only one of the genus. It is com mon and w idespread from Mexico to Argentina. Kaye (1904, 192 1) records it as "a fairl y common i nsect" . The grou nd co lour of thi s spec ies is brown , overlaid UPS wi th g rey scales. The extensive pale hya line marki ngs F and Plate 24. Xenophanes tryxus fifth instar larva collected on Malachra H a re di s ti nct ive. UN H broadly white except marg in and nar­ sp., junction of Rio Claro - Guayaguayare Road and Saunders Trace, r ow macular s ub -marg in a l line. A di s tinc tive s p ec i es. 11 .x.1993 Skipper butterflies 41 g reen laterally wi th scattered sho rt. fine, whi le setae arising from s ma ll w hite dOls; Tl and a ll legs concolorous. Mo ss's pupal case preser ved in th e NHM is transparent brow n, round e d a nd cove re d with lo ng, brown, proj ec ting hairs; loose while waxy patc hes ar e arranged in a patte rn on th e thorax and abdomen, patc hes on T, tra nsverse lin es on A 1-3, trans­ verse rows of pa tches on A4-8. My larva pupated o n 26 xi.1981, a nd an a dult ma le emer ged e ig ht days later. One 10 m m f ield co ll ec ted larva (S pani sh Farm, 13 .xi . J981) was p arasiti sed by la rvae of an ex ternally feedi ng c u lophid paras itoid , w hi c h pupa ted in the host s he lter.

Allfigollus Biezan ko ( 1963) records that another spec ies o f the gen us, Antigolllls liborius areta Evans 1953, feeds o n Ch orisia specio:w (Bo mbacaceae) and spec ies of Ablltiloll, Pavollia and Plate 26 Antigonus nearchus V, Parrylands, 2. ii.1980 Hibiscus (Mal vaceae). Bo th th ese families are quite c losely related to Ste rcu liaceae, the fam il y of th e food plant give n brown patches UPF sub-basally, te rme n, s ubapicall y a nd UPH below for Alltigolllls erOS Il S. a lo ng margin spaces 3-5 . UN S lighl yellow brown, with nar­ row, brown irreg ular band s across both wi ngs. The wi ngs a re 116. E5 5 / 1 Allfigolltls n ea rchtl s Latreille 1824 broader, an d the HW proj ec tion in space 6 is longer in the Plates 25, 26 female. The wi ng shape and colouring is di stinctive for thi s Thi s co mmon and widespread species is found from species . Illu Slrations in Riley ( 1975, Plate 22, Fig. 9, 9) and Mexico to Paraguay (TL South America). T wo males labell ed Lewis ( 1973, Plate 80, No. 29, d'). Narrow costa l fold; F a 2 1 J a mai ca in the N HM are most probably mis-labelled (Brown mm, 923 mm. and He ineman 1972). Kaye (1904 , 1921) records thi s as a I would normally regard thi s as an occas io nal spec ies, but common species a round Po rt of Spai n. sometimes i t ca n be commo n at Eupatorium flowers, especial­ This species s hows strong sexual dimorphi sm. Ma le dark ly in the south (e.g . in Pa rry la nds in January 1988). 11 is wide­ brown a bove wi th i ndi sti nct mark i ngs UPF: pale scales in basa l spread in o r near forests i n both the north a nd sou th . usual­ ba nd. a pi cal patch and sub- ma rgi nal above torn us; grey­ ly at low altitude, althoug h I do have a specime n from 701 m brown discal band fro m space 1 B to cos ta (paler at costa), o r 2300 ft. ( Mo rn e Bl eu), a nd there is a s pec i men labelled EI with a projecting extension in space 3; ches tnut scales in sub­ Tucuche in th e NHM (v iii.1 905, F Birc h). In additio n to regu­ di sca l patc h space I B and at marg in s paces I B and 2. UPH larly feeding on fl owers, males can sometimes be found fee d­ wi th indi sti nct brow n bands, c learest in spaces 2 and 3 . UNF ing on damp gravel beside streams (e .g. Arima Valley. c hestn ut brown , brow n basa ll y: indi sti nct di sca l band and 5.x.1979). Thi s and the next arc probably the only Trinidad a rced sub ma rgi nal band from apex to tornus. UN H costal ha lf species which I have observed to do this, although I have seen chest nut brown, s hading i nto pa le brow n at dorsum; irregul ar many other spec ies do so in South Ame rica. dark bands and patches except in spaces I A and I B. Female Kaye (1921) notes th at the larva feeds on cashew, lig ht purpli sh brown above wi th much clea rer markings : Allacardilll1l occidentale (Anaca rdi aceae).

11 7. E 55 12A llfi go tllt s erOS Ii S HObn e r 1812 P l ates 27-35 This, like th e last. is a comm on and wides pread species . It is found from Mexico to Paraguay (TL unspecified). Kaye (1904, 1921) found this spec ies frequent i n the north of Trinidad, a nd She ldon (1936, 1938) records it from Tobago (Cocoa Watt ie, Roxborough. Scarborough). T his spec ies is qu ite variab le in the dens ity of the mark­ ings and th e ex tent of th e hyaline markings of th e fema le. It shows s ligh t sexual dimorphism, the female normally having slight ly hya line pale ma rk ings in cell , and spaces 2 a nd 3, which are absent in th e male. Male UPS dark brown with a n overlay of grey-blue scales w hi c h de fin e g round co lo ur li nes and bars, of which UPF two bars ac ross e nd ce ll, and a broad Plate 25. Antigonus nearchus a, Arima-Blanchisseuse Road, milestone submarg inal line are the most disti nct. UNS brown , yellow 3.5,5.x.1979; brow n at margins; tw o vio le t discal spots space IB UN H, a nd 42 Living World 1997-1998

Plate 29. Antigonus erosus !j! , Morne Catherine, 6.viii.1979; Plate 27. Antigonus erosus rJ, Parrylands, 20.i.1988;

Plate 30. Antigonus erosus !j! , Port of Spain, 9.x.1995

Plate 28. Antigonus erosus 9, Port of Spain, reared from a larva col­ Scotland Bay on the mud behind the beach (9.i.198l); thi s is lected on Guazuma ulmifolia, 10.x.1995 probably tbe skipper which shows this behaviour mos t fre- quently in Trinidad. lance found a male feeding on fresh a black spot at termen UNH. Fema le UPS dark brow n , the donkey d r oppings on the B lanchisseuse-Paria Bay Track overlay of grey-blue scales s trong only in space 1 B UPF and (22.i. 1980); this is the on ly instance of a skipper feeding on UPH; margin s variable s haded chestnut brown. UNS light mammalian excreta which I have seen in Trinidad . brown, with irregular brown bands on both wings; l ight v iolet have noticed a characteristic resting stage when the d i scal bars across space I B UNF, space 1 A UNH yellow and adult is inacti ve, once by a fema le res ting for several hours on dark spot termen UNH. The wing shape and colouring is dis­ the Port of Spain Hilton roof walkway l ate in the afternoon, tinctive for this species ; NoclIlana sta tor Godman and Salvin and once by a male on the s ummit of Mt. Tamana du ring very (Cock 1992) perhaps comes closest, but the wing s hapes are overcast a nd showery weather. In thi s position ( Plate 32) the different, those of N. stator projecting at end of veins 3 and 4 fore wi ngs are held s ligh tly raised, wi th the do rsum at right H , and angled at end of vein 6 F. Illu stration in Smart ( 1976, angles to the body, while the hind wings are pressed against Page 113. Fig. 42, d'). Narrow costal fold; F d' 19 mm, !j! 20 the substrate and held with the dors um tight against the m m . abdome n. so that there i s almost no overlap between the fore In Trinidad, thi s species is common and widespread, espe­ wings and the hind wi ngs. cially in and around secondary forest. Although it i s mainly Comstock and Vazquez (1961) record and illus trate found in lowland s ituations. I h ave speci mens from Morne details of the biology of A. erosus in Mexico on an unidenti­ Cat h eri ne (457 m or 1500 ft.). Mt. Taman a (305 m or 1000 ft.) fied small tree. Subsequentl y, a larva l food plant in Mexico and the s ummit of Brigand Hill. A. erosus is attracted to flow­ was found to be Guazuma ulmifolia (Sterculiaceae) (Kendall ers and i s sometimes common at Eupatorium flowers in the and McGuire 1975). The sa me food plant is u sed in Trinidad south. have seen mal es drinking at mud puddles several where it is known as Bois l'orme or West Indian e l m (Freeman t i mes i n t he va lleys of the Northern Range, and once a t a nd Will iam s 1928). SAS and I found larvae on this host in Skipper butterflies 43

the lo nges t side, and folded over upwa rd s; the larva a te a row of small holes a long each side of the vein un der the s helter. The stage II shelter is a pocket (c. 20 x 15 mm) betwee n twO leaves one o n top of the other; the lower leaf has the shelter margin eaten very irregularly. and several holes on the UNS of the shelter, some bridged by one or two s trands of silk: upper leaf has three sma ll rou nd holes adjacen t to the edge of the shelter, and several large irregular holes adjacent to a vein on the opposi te margin of the shelter. Shelter II I is also be tween two leaves. the lower, sma ll leaf com pletely covered by the upper leaf; the larva rests on the undersurface o f the upper leaf. Three fina l larval shelters of the Poi nt Gourde material were also forme d by a tta.::hing one leaf on top of another; one Plate 31 . Antigonus erosus !j! in normal resting position (reared, Port of s helter had both leaves extensively perforated with round Spain 1.xi. 1995) ho les a nd li nes. abou t 2mm wide; the second resembled the first, but the upper leaf had a broad cen tral line without ho les and the larva rested on the undersurface of this: the third shelter had no perfora tio ns at all in the leaves. I suspect that these differences merely reflected the maturi ty of the larva in the shelter, and by the t i me the larva is ready to pupate the whole shel ter is perforated. T he final instar larva (P la te 33) measured 30 mm. The head is broad and chordate, slight ly indent at vertex, apical portions fl attened; pos terior openi ng very narrow, half the width of the head; covered with short. scattered, pale setae; ground colour light brown: posterior margin dark brown; the head is decorated with rows of smal l protruding bu mps (LUbercles): a trip le row paralle l to the ep icranial suture, con­ verging to a double row near frons. the apica l tubercles dark brown; a second triple row converging to a double row from the apex downwards and then angled inwards stopping short

Plate 32. Antigonus erosus !j! in long term resting position, Mt. Tamana, of the frons; the lateral portions of the epicrania are covered 12.xi.1995. with semi-regu larly arranged tubercles, but the pattern is not c lear; each side of the fro ns there is a short grey streak. T I Parrylands (ii.1988) but fa il ed to rear them. Subsequen tly conco lorous wi th body. Body dull green. covered with scat­ SAS rea red out a larva collected on this t ree on Point Gourde tered ye ll ow dots to give a ye llow. green appearance; dorsal (x . 1993), and I have reared it from th e same host and local i­ li ne only slig htly d arker; narrow ye llow dorsa-lateral line: ty (14.vii.1996) as well as from a larva collected al Port of narrow, diffuse, pale ventra-lateral line; body covered with Spain, above the Hilton Hotel (l0.)(.1995). In Trinidad, G. short, erect yellowish setae; anal plate slightly poi nted. ulmifolia is found o n hillsides a nd in forests in moist and dry districts (Williams and Cheesman 1929). There is substantial variation in the foliage (Wi ll iams and Cheesman 1929) which I tentatively associa te with plan ts growi ng in shade, small, thin, shiny leaves, and open sun, larger. thicker. mat leaves, so much so that initially I thought two different species were involved. T he larval description below is based on the female specimen which I reared from Port of Spain which was col­ lected from a plant growing beside the road above the Hilton in open sunl ight. The description matches that of Comstoc k and Vazquez (1961) closel y, although not exactly, e.g. the Mexican material has a narrow black dorsal plate o n TI, whereas this is concolorous in Trinidad larvae. The stage I shelter of the Port of Spain material is a small triangular flap (c. 8 x 4 mm) cut from the edge of the lamina Plate 33 . Antigonus erosus fifth instar female larva collected on near the base of the leaf; it is hinged on a leaf vein, opposite Guazuma ufmifolia, Po rt of Spain, 10.x.1 995 44 Living World 1997-1998

Spiracles more or less coocolorou5. Legs and prolegs coneol­ E58/2 Zopyrion satyrina Felder 1867 orOll S . There are two female specimens of Zopyrion satyrina in The pupa (Plate 34) was formed in the last larval shelter, th e Lamont coll ection a t UW I from Patos Is (Cock 1982b). lined with brown s ilk and s upported by the cremaster and a T hese were collected vi.19 18 by " CBW" (I assume this is C B V- s haped s ilk s upport. The pupa measured 18 mm ; rounded Williams), when the island be lo nged to Trinidad and Tobago. with no projections; brow n ; whole s u rface covered wi th erec t There are no records f rom th e country as recog ni sed today. I seta- like white waxy spikes, except for T I spiracles. costal am not familiar with thi s spec ies or genus in the fie ld . margin of wing cases and o ulline of appendages. The pupal Female UPS brown , light brown along margin ; UNS li g ht stage lasted 15 days. brown , st riated with brown and ches tnut brown UN H; a mar­ In co ntras t , matu re larvae from Po int Gou r de (Plat e 35) gina l dark s pot in each space UNF and UNH , strongest in were smalle r , measuring about 24mm. They resemble the spaces I C-3 and 5-6 UNH ; UN H with indi stinc t discal and sub­ marg in al ba nds of c hes tnut brown with dark margins. Illustra ti on in Lewis (1975, Pl ate 88, Fig. 36 UNS). Costal fold:

F ~ about 16 mm.

References

Bar can t, M. 1970. BUllerflies of Trinidad and Tobago. Coll in s, London . Biezanko , C . M. 1963. Hesperiid ae d a Zona S ueste do Rio Grande do SuI. Arquivos de £ntol1lologia. Pelota s . Serif A 6: 1-24. Biezanko , C .M. an d O.H.H. Mielke. 1973. Contribui cao ao est udo fa unistico dos Hesperi id ae Americanos. IV Especies do Rio Grande Do Sui, Brasil , com notas taxonomi­ Plate 34. Antigonus erosus pupa, collected as larva on Guazuma ulmifo­ cas e descricoes de especies novas (Lepidoptera). A cta lia, Point Gourde, 14.vii,1996 Biologia Paranaense, C"ritiba, 2 : 5 1- 102. Brown , F . M. a nd B. Heineman . 1972. J a m a ica and its Port of Spain larva except that the head is darker brown, e spe­ butterflies. E.W. Classey Ltd . London . c'iall y around the mou th parts, and the dorsal portion around Cock, M.J.W. 1982a. Butterfli es from C hacachacare Island eac h apex and across the vertex is black. inc luding three spec ies new to Trinid ad. Living Wo rld. 1. of Further inves ti gation is needed to understand the differ­ the Trinidad a nd Toba go Field Naturalists' Club, 1981 - 19 82, ences between the two Iypes of larvae descri bed above. As p. 25. Cock , M . J.W. 1982b. The skipper butterfl ies ( Hespe riid ae) of Trinidad. Pan II. Systematic list o f the Trinidad a nd Tobago species. Occasional papers of the Dept. of Zoology, UlVl , St. Augustine, No.5, 49 pp. Cock, M.J.W. 1984 . The ski pper butterfl ies (Hesperiidae) of Trinidad. Part 3 Pyrginae (first section) . Living Wo rld. 1. of the Trinidad alld Tobago Field Nt/turalists' Club, 19 83· 1984. pp. 38-4 2. Cock , M.J.W. 1992. The sk ipper butter flie s ( Hespe riidae) of Trinidad Part 7, genera group E (first section). Living World. J. of the Trinidad and Tobago Field Naturalists' Club 1991 - 1992. pp. 46-56. Cock, M.J . W. 1996. The s kipper butterflies (Hesperiid ae) o f Trinidad. Part 8, Gene ra group E (second sect ion). Living World. 1. of the Trinidad and Tobago Field Naluralists' Club Plate 35. Antigonus erosus fifth instar male larva collected on Guazuma 1995-1 996. pp. 27-35. ulmifolia, Point Gourde, 14.vii.1996 Co mstock, J.A . a nd L . Vazquez G . 1961. Estudios de los ciclos biologicos en Lepidopteros Mexicanos. Anales del noted, a ll three la r vae from Point Gourde were maies , where­ ln s tituto de Biologia 3 1: 349-448 . as that from Port of Spain was female , so possi bly these a re C ro wfoot, W . M. 1893. Preliminary list of Trinidad butte r­ sexua l differences. Equally, it may be that the larvae are vari­ flies. 1. of the Trinidad Field Naturalists' Club, I; 173-174. able, like the adu lts. Skipper butterflies 45

Eva n s . W .n. 1953. A catalogue of tire American Hesperiidae Kendall , R.O. and M.A. Ri cka rd . 1976. Larval food­ indicating tir e classification and nomenclature adopted in plants, s patia l and te mporal distribu tion for five ski ppers tire British Museum (Natural History). Part III (Groups E, F, (Hesperi idae) from Texas. J. of tire Lepidopterists' Soc. G) Pyrginae. Section 2. London; British Museum (Natural 30: 105-1 10. History), 246 pp, 53 plates. Lewis, H.L. 197 3. Butterflies of the World. Harrap, Lond on, d e Jong, R . 1983. Annotated list of the Hesperiidae 3 12 pp . (Lepidoptera) of Surinam, with descriptions of new taxa. Moss, A.M.1949. Biological Notes of some Hesperiidae of the Tijdsclrrift voor Entom%gie. 126 : 233-268. A mazon. Acta Zoologica Ulloana, Tucuman, 7, 27-79. Freeman, W.C. and R.O. Williams. 1928 The useful and Riley , N. D. 1975. A field guide to the butterflies of the West ornamental plants of Trin idad and Tobago. Memoirs of the I ndies. Collins, London. Dept. of Agric ulture, Trinidad and Tobago, 4. 192 pp. Sheldon, \V.G. 1936. Tobago and its butterfl ies. Kaye. W.J. 1904. A catalogue of th e Lepidoptera Elltomologist 69: 1-9. Rhophalocera of Trinid ad . Transactions of the Entomo. Soc. Sheldo n , W. G. 1938. Addit ions to th e butterflies of Tobago. of Londoll, 1904: 159-224. Entomologist 7 1: 29-3 1. Kaye. \V.J. 1914 . Additions and corrections to my catalogue Williams, R.O. 1929. Geraniales. Flora of Trinidad alld of the Lepidoptera Rhopalocera of Trinidad 1904. Tobago 1(2): 11 3-164. Transactions of tir e Entomo. Soc. of London 1913: 545-585. Williams, R.O. 1932. Myrtales. Flora of Trini dad and plate 30. Tobago 1(5): 322-332. Kaye , W.J. 1921. Cata logue of the Trinidad Lepido ptera Williams, R,O . and E.E. C he esma n . 1929. Malvales. Rhopha locera. M emoirs of tire Dept. of Agriculture. Flora of Trinidad alld Tobago 1(2): 72- 11 2. Trinidad alld Tobago: 2. 163 pp. K aye, W.J. 1940. Addit ions and correc tions to the recorded species of Trinidad butterflies. Tran saction of the Royal Elllomo. Soc. of London, 90, 551-573. Kendall , R.O. and W,W. McGuire. 1975. Larval food­ plants for twenty-one species of skippers (Lepidoptera: Hesperiidae) from Mexico. Bulletin of the Allyn Mu seum 27, 7 pp. FIELD OBSERVATIONS OF TROPIDACRIS COLLARIS (ORTHOPTERA: ROMALEIDAE) Christopher K. Starr Dep't of Life Sciences, University of the West Indies,St Augustine, Trinidad & 'Ibbago cstarr{jj}centre.uwi.tt

Tropidacris i s a neotropi c al genus of three known s peci es aggregation close to the ground on a small shrub c lose along­ that include the largest g r asshoppers in the world (Carbone ll side the gulch. I netted a sampl e of these, which disturbance 19 86). Two species, T. collaris and T. crisrata, have very caused the remai nin g individuals to scatter. Some time later broad ranges that include mos t of South America north of the I returned to that spot and found th e aggregation re-formed southern cone . The former is the species found on Margarita in a s imilar s ituation less than a meter from where I had first Is land, wh il e the range of the latter inc ludes Tri nidad and found it. Although I did not attempt to quantify ad ult densi­ Tobago. The two are readily distingui s hed by the fo llowing ty in any part of the gulch, they appeared to be most concen­ adult characters (Carbonell 1984.1986): a) a nte nnae enti rely trated within a very few meters of the aggregation of hoppers. yellow in T. collaris, basal two segments brown to black in T. I tas ted one hopper and found it to be very biller. approx­ cristata, b) dorsa l crest of pronotum continu ing o nto posteri­ imately like an adu lt mo narch butterfly (Danaus plexippus). ormas t lobe (metazona) in T . cristata but not T. collaris, c) Thi s distastefulness is consistent with the orange-and-black hindwings mainly orange to red in T. cristata, green Lo blue in body co lou rat ion . T. collaris. We will di s regard here the third, s maller s pecies, Casual observations els ewhe re around La Vecindad con­ T. de ca mps i, known from a si ng le locality in Colombia. firmed that th e adult grasshoppers were abundant over a Hoppers (larvae) of Tropidacris s how contrasting, evident­ large area. The local people with whom I spoke seemed ly warning colouration, whil e adults at res t are very well cam­ un surp ri sed by thi s and told me that it is a seasonal phenom­ ouflaged against vegetation . The forewing s bear a strong enon, although they were vague about the particular season resemblance to leaves, while the hindwings present s triking of appearance of adults or whether they are abundant every fla s h colouration. especially in T. cristata ( Rowell 1983 ; pers . year. They told me that T. collaris is locally known as lall­ obs.). gosta or fiangaragato. Adult Tropidacris are spectacular insects and are not rare Th e adults that I collected amounted to four fema les and throughout most of the genus's range . T. collaris is often 22 males with mean hind - femur le n g t hs of 38.8 mm ( range encountered a l ong much of the Caribbean coast and th e 36.5 - 42.0 mm) and 33.3 mm ( range 29 .0 - 37.5 mm ). respec­ ll anos region of Venezuela (F. Cerda, pers. co m m.). tively . Meas urement was of alcohol -preserved specimens with None thel ess, they are not often encountered in large num­ an ordi nary ruler to the nearest half millimeter. bers. I n Trinidad I have never seen two adult T. cristata in Di ssection of the females s howed that the two largest had one day and possibly not t wo in the same ca le ndar month. well-developed ovaries with apparently mature ova. The next Rowell 's (1983) brief account of T. cristata summarized the largest had slightl y developed ovaries. while the s mallest bi ological in for mation on record fo r the genus. ending wit h s howed n o ovarian development. t he remark that "Nothing very much is known of any of The sa mpl e of hoppers compri sed 19 fema les and 10 males . them." Carbonell ( 1986) s upplied s ome further details, not­ with hind-femur le ngths shown in Fig . I. Sex is readily deter­ ing that in both species the hoppers are g rega ri ous. genera­ mined by ventral examination of the abdo minal terminalia. tion s are not di screte, and all stages feed on a variety of Measurement was with an eyepiece micrometer to the nearest p l ants. He a lso noted that T. collaris s hows a very broad 0.16 mm; c ho ice of the left or righ t leg was according to con­ habitat range from humid forest to drier. more open forma­ ve nience . tions , wh il e T. c ristata is largel y abse nt from o pe n , dry for­ mations. It is this sparseness of biological information that justifies the present brief observations. All are fro m an area of cactus scrub outsid e the village of La Vecindad on the island of Margarita , Venez.uela, on 19 August 1997. Thi s was toward the end of a dry season th at lasts the greater part of the year I "6 on Margarita. Observation s were concentrated in a s hallow , .i \ dry gulch about 150 m long, lined with grasses, sedges, and z many herbaceou s broadleaf plant s but few trees. Hind-Femur Length (mm) M y attention was d raw n to T. collan's when I happened to see two adults in quick succession. I began a deliberate search and collected a ll adults that I could f ind . Figure 1: Frequency distribution of hind-femur length in 19 female and The on ly hoppers that I encountered were a s ing l e. dense 10 male hoppers of Tropidacris collaris from an aggregation. Tropidacris collaris 47

The size-freque ncy di stribution of hoppers indi cates that Thanks to C.S. Carbonell, F. Cerda and H.F. Rowe ll for crit­ these represent two ins ta rs, with males on average s li ghtl y ical comment on these observations. smaller th an same-i nstar females. Thi s conclu sion is corrob­ orated by consi sten t within-sex. differences in the form of the References abdominal terminalia bet ween th e two size-groups. The bas is of the strongly skewed sex-ratio in the sample of Carbone l l, C.S. 1984. Nomenclat ure and systematics of hoppers is not readily apparent. Given the sma ll data-set, I do Tropidacris a od EUlropidacris (O rthopte ra. Acridoidea. not discount the possibility that it is du e to sa mpling bias. Romaleidae). Norulae Naturae (Philadelphia) (461): I ­ Taken by themse lves , the observations of adults are cons is­ ll. tent with th e hypothesis of a hi g hly seasonal life cycle, with Carbonell, C.S. 1986. Revi sion of the neotropical genus males developing s omewhat faster than females and a con­ Tropidacris (Orthoptera. Acridoidea. Roma leidae. centration of breeding around the start of the rainy season. Romaleinae). Proc. Acad. lIat. Sci. Philadelphia 138:366- The presence of hoppers during the same time suggests that 402. T. collQr/'s breeds at two dist inct seasons on Ma rgarita. Rowe ll , H . F . 1983. Tropidacris crisrata (saltamonte 0 cha­ I regret that I was unable to extend these few opportunistic pulfn gigante. giant red-winged grasshopper). In D.H. observations and hope that they may stimulate someone else Janzen (ed.), Costa Rican Natural History. Univ. Chicago to undertake a more thorough s tudy. Press. Chi cago Pp. 772-773