Journal of the Field Naturalists' Club

Natura Maxime Miranda in Minimis

Published July 1986.

LIVING WORLD is published biennially by the Trinidad and Tobago Field Naturalists' Club. This issue is dedicated to Rtchard ffrench. All rights reserved. ·setting, design and page mechanicals by 8M Publica· tions, 20 Collens Road, Maraval, Trinidad.

Dedication

RICHARD FFRENCH was born in England and shortly after gra­ duating from Oxford University came to the West Indies with his wife Margaret. He lived in Barbados for three years and then came to Trinidad to teach at St Peter's School, Pointe-a'-Pierre. His interest in was alreadY well developed bu t so was his interest in music and he contributed much to the cultural life of Trinidad and Tobago in both of these fields. He is a past president of the Trinidad and Tobago Field Naturalists' Club and edited its jour­ nal for a number of years. He was also chairman of the Board of Management of the Asa Wright Nature Centre for some years and served as a member of the board after his term as chairman. From his arrival in Trinidad to his departure in April 1985 he studied the avifauna of our two islands and produced his Guide to the Birds of Trinidad and Tobago in 1973. It is now in its fourth edition. A smaller (and cheaper) guide to the common birds has just been published and shDuld introduce the study of birds to an even wider pUblic. He has published several papers in this journal and in others and with Peter Bacon wrote Nature Trails of Trinidad. In all of these endeavours he has been ably helped and encouraged by his wife Margaret and we wish both all possible success in whatever new ventures they undertake in their native land. ·In recognition of his contribution to the study of natural his­ tory this issue of Living World is.gratefully dedicated to him.

3 Contents Cover

Front: Anolis chrysolepis plalliceps, the only native anole (Photo by Hans Boos) DEDICATION ...... PAGE 3 Back: A peripatus from Mt. Aripo new to the island (Photo by MOVEMENTS OF SEABIRDS OFF CROWN POINT, TOBAGO V.M. St. J. Read) By Richard ffrench ...... PAGE 5

ADDITIONAL NOTES ON THE BIRDS OF TRINIDAD AND TOBAGO By Richard ffrench ...... PAGE 9

CHRISTMAS COUNTS OF BIRDS IN TRINIDAD 1969-1975 AND 1981-1983 By Richard ffrench ...... PAGE 12

WHY DO SIT ON THE ROADS AT NIGHT? By Victor C. Quesne1...... PAGE19

THE UNDERGROUND LIFE OF THE TRINIDAD WORM­ LIZARD AMPHISBAENA ALBA The Trinidad and By John Riley ...... PAGE 24

NOTES ON CURIOUS METHODS OF LOCOMOTION IN TWO Tobago Field IGUANIDS FROM TRINIDAD, WEST INDIES By julius O. Boos ...... PAGE 26 Naturalists' Club A PRELIMINARY STUDY OF CULEX MOSQUITOES ATTRACTED TO A TURTLE IN TRINIDAD, W.l. By Dave D. Chadee, Robin C. Persad, Wayne Ramnath and THE Trinidad Field Naturalists' Club was founded on the 10th R. Ganesh ...... PAGE 27 July 1891. Its name was changed to the present one in 1974. The objects of the club are to bring together persons AN ONYCHOPHORAN FROM THE SUMMIT OF MT ARIPO, interested in the study of natural history, the diffusion of the TRINIDAD, WITH NOTES ON OTHER FROM knowledge thereof and the conservation of nature. THAT LOCALITY At present the club has an enrolment of over four hundred By V.M. St. J. Read ...... PAGE 28 members comprised of nature-lovers and professional and amateur naturalists. NOTES ON HIE LIFE HISTORY OF CATOBLEPIA Monthly lecture meetings are held at St. Mary's College BERECYNTHIA BERECYNTHIA (CRAMER) (: on the second Thursday of the month while field excursions are BRASSOLIDAE) held on the last Sunday of each month, except December, when By julius O. Boos ...... PAGE31 no official club activities are organised. Membership is open to all residents of Trinidad and Tobago, THE BUTTER FILES (HESPERIIDAE) OF TRINIDAD. of at least fifteen years of age, who subscribe to the objects of the PART 4: PYRGINAE (SECOND SECTION) club. By M.J.W. Cock ...... PAGE 33 The club's management committee is: Dr. V.c. Quesnel, President; T. F. Farrell, Vice-president; Luisa Zuniaga, Honorary THE FAMILY ARISTOLOCHIACEAE IN TRINIDAD WITH Secretary; Lana Salandy, Honorary Assistant Secretary; John SPECIAL REFERENCE TO ITS MEDICINAL USES, ITS Hilton, Honorary Treasurer; Colin Agostini, Glenn Wilkes, FOLKLORE AND ITS USE AS A LARVAL FOODPLANT BY Graham White. Its editorial committee is: Victor Quesnel, Julian P APILIONIDAE Duncan, Richard ffrench, Hans Boos. By julius O. Boos ...... PAGE 48 All enquiries concerning the club or its journal shOUld be addressed to the Honorary Secretary, 1 Errol Park Road, St. BOOK REVIEW ...... PAGE 52 Ann's, Trinidad W.l.

4 I

Movements of seabirds off Crown Point, Tobago

By Richard [french Beech Ridge, Bussage, Stroud, Glos., England

VERY little has been done to date in Trinidad and Tobago in the Dividing the day into four sections (A - D), denoting early way of in-depth studies of seabirds, in spite of the fact that morning, late morning, early afternoon and late afternoon, - Tobago in particular possesses what are probably the most in each case within half an hour of 07000, 1030, 1400 and 1730 important seabird breeding colonies in the south Caribbean. Un­ - I recorded over a 30-minute period all seabirds within my view, fortunately such studies require plenty of time, since by their na­ noting whether they were moving to the north or south of my ture seabirds move about a lot in somewhat inaccessible areas, position. and in almost every case the researcher requires a boat. Since I I covered each section of the day three times within a fort­ have lacked both these essentials, I have had to content myself night over approximately the same period of the year (early over the last 27 years with random observations, some banding at April) for three consecutive years, so the results (Tables 1 - 3) the colonies, and a study of moult that bogged me down in awe­ should give adequate coverage of seabird movements just off some statistics. But I have immensely enjoyed myself working western Tobago during early April. with seabirds, especially on Tobago. Although some birds are likely to have been counted more Here I present some observations, based on some systematic than once in a 30-minute period, as they first flew in one direc­ coast watches at Crown Point, Tobago during 1979, 1980 and tion, then returned the other way, I am convinced that this was 1981. Sometimes one cannot immediately recognise the signi­ extremely uncommon. Such behaviour would indicate actual ficance of certain statistics, so I include them in detail, at least foraging. But when seabirds are actually feeding, or going directly partly in the hope that others may be stimulated to extend or to a shoal, they fly faster and more purposefully than usual, in develop future watches in a more meaningful way. a manner that is instantly recognisable to fishermen or those experienced with seabirds. The great majority of the birds that I METHODS counted seemed to me to be cruising steadily and unhurriedly, as if journeying to a distant point. stationed myself with binoculars at Sandy Point beach on In order to set up some comparison with the April figures, I the extreme western tip of Tobago (see Figure 1), so as to gain an also watched twice for 30-minute periods in December and 0 unobstructed view of the seacoast over nearly 180 . Visibility January. The figures from these "winter" watches (not included was almost always very good, so with lOx binoculars I could scan here) showed a complete absence of four of the eight , the horizon and recognise any birds within 1 V2 kilometres of the namely the gulls, Sandwich, Roseate and Noddy Terns, and far shore. fewer numbers of most of the other species. Only the Royal Tern and the Brown Pelican were well represented in December. FIG. 1 To Pigeon Pt. RESULTS

1. The observations (Tables 1-3) show that eight species of sea­ were involved in the movements past Crown Point. Al­ though some birds were seen passing at distances up to 1 V2 km from shore, the majority passed fairly close to shore. Brown Pelicans (Pelecanus occidentalis), Laughing Gulls (Larus articilla), Royal and Sandwich Terns (Sterna maxima and S. sandvicensis) mostly passed within 200 metres, with Roseate Terns (Sterna dougallii) a little farther out. Brown Noddies (A nous stolidus) 1 :25,000 and Brown Boobies (Sula leucogaster) were mainly seen about Hotel s View poi nt 600 metres out, while Magnificent Frigatebirds (Fregata magni­ [icens) were the most variable, often passing over at a consider­ L______Airport Runway _ able height. Very few of the seabirds, other than some frigate­ birds, crossed the point over land, SQ those seen from the view­ o point probably constitute the vast majority of seabirds moving Buoy north or south in that area.

CROWN PO INT, TOBAGO. 5 TABLE 1

Movements of seabirds at Crown Point during April 1 - 14, 1979

Direction Periods of Day Species A B C D Total Laughing Gull N 146 218 418 16 6 9 19 39 72 340 432 322 2051 S 16 22 9 11 3 42 40 145 21 44 13 367

Royal Tern N 36 21 10 15 4 8 25 7 54 28 13 222 S 1 4 3 2 5 4 5 4 3 4 36

Sandwich Tern N 2 8 6 18 S 7 9

Roseate Tern N 26 26 S 3 17 4 2 2 29

Brown Noddy N 6 3 4 14 S 1 2

Brown Pelican N 20 21 12 1 8 3 15 11 92 S 3 5 7 6 3 1 2 21 6 5 60

Magnificent Frigatebird N 4 3 3 8 15 33 2 10 2 3 6 89 S 3 2 5 4 6 4 7 31

Brown Booby N 7 9 10 9 2 69 3 113 S 14 3 10 22 14 1 64

TABLE 2

Movements of seabirds at Crown Point during April 6 - 19, 1980

Direction Periods of Day Species A B C D Total

Laughing Gull N 27 45 52 6 11 20 19 5 14 175 77 109 560 S 1 5 27 7 2 5 23 34 35 2 16 3 160

Royal Tern N 10 26 43 3 4 13 9 7 6 24 21 15 181 S 5 7 2 3 5 5 3 4 4 11 50

Sandwich Tern N 2 S 4 6

Roseate Tern N 2 21 4 5 6 39 S 4 7 3 14 I Brown Noddy N 4 7 1 S I 1

6 Brown Pelican N 11 6 8 4 2 2 5 4 2 4 6 55 S 7 7 4 10 6 4 4 1 33 23 26 125

Magnificent Frigatebird N. 3 8 2 196 9 63 2 6 7 170 467 S 1 6 12 3 1 3 5 6 3 40

Brown Booby N 3 3 3 42 11 11 74 S 8 1 3 3 3 15 34

TABLE 3

Movements of seabirds at Crown Point during March 30 - April 10, 1981

Direction Periods of Day Species A B C I D Total I Laughing Gull N 12 27 47 15 14 8 17 19 8 265 152 40 624 S 17 26 103 3 5 27 18 18 13 7 15 252

Royal Tern N 12 2 2 6 9 7 14 22 16 9 101 S 13 8 3 10 3 4 11 9 6 2 69

Sandwich Tern N 3 5 3 11 S 6 6

Roseate Tern N 4 14 5 32 2 4 13 72 S 49 33 10 4 96

Brown Noddy N 3 4 S 1 3

Brown Pelican N 2 8 5 1 2 6 5 13 14 6 63 S 2 4 9 6 7 2 32

Magnificent Frigatebird N 3 15 7 34 259 7 5 4 335 S 209 3 23 2 238

Brown Booby N 89 12 4 107 S 5 3 9 11 5 4 26 65

2. The height of travel above water-level varied from just above tropics. To a certain extend this can be associated with a daily the waves (especially Brown Noddies and boobies) to about 30 commuting from roost to feeding area in the open sea and back metres above the water. The main exception to this was the again. This tendency was especially marked in the boobies, which Frigatebirds, that tended to soar in circles, drifting gradually in clearly moved southwards during much Of the day, returning one direction and often congregating in thermal currents, at north in the eVening to their roosts on the rocky coasts of north heights of up to several hundred metres. and northeastern Tobago and the offshore islands. 3. Movement was more pronounced in the early morning and But the same cannot be said for the gulls and terns, which late afternoon, just as is the usual case for land birds in the moved northwards in greater numbers than in the other direction.

7 I am inclined to think that the Laughing Gulls and Royal Terns, would have to be much more comprehensive to tell us anything which were the two most numerous species, were actually in the meaningful. course of a northward migration at this time. It would indeed be Similarly, we learn little here about the pelicans, other than expected that many seabirds wintering off the northeastern coast the fact that they are more sedentary than the other species. of would be returning to their breeding grounds in Many of them spent much of their day resting on the buoy just the West Indies and North America in April . .Both these species off the point (see Fig. 1), fishing nearby or commuting up or down. More than the other seven species, the pelicans' move­ are also known to breed on our islands, although actual nesting ments appeared to be especially influenced by the fishing condi­ colonies of Royal Terns have not yet been found on Tobago. tions in the immediate neighbourhood. The small numbers seen of Sandwich Terns and Brown 5. The state of the tide at the period of observation appeared to Noddies do not warrant any conclusions here, and I found only make little difference, as one would expect with the very small fairly small numbers of Roseate Terns, although occasionally a tides in this part of the world. Moreover, no significant changes large flock would pass by. The recent discovery of a small nesting were apparent as a result of the occasional rainy spell; however, colony of Roseates near Buccoo would of course lead us to the weather was in fact predominantly fair. expect some movement of this species in the area. 6. I suggest that future studies of seabird movements in the 4. Frigatebirds also roost off northeastern Tobago, and the area might encompass different times of year, especially daily southwestern movement of the species to feeding grounds September/October, when migration would be expected. Also and back is a well-known phenomenon. But owing to their habit valuable data might be obtained from regular watches at points of congregating in thermal currents, it is more a matter of chance like Toco, Galeota and Icacos. Notwithstanding the above, the whether birds passing Crown Point in a specific direction will seabird area that is crying out for an in-depth study is the area of appear there in large numbers. Thus a general tendency one way St Giles Islands and Little Tobago. However, the logistics of such might easily be counter-balanced by only one large concentra­ a study are daunting; I only wish I had another life-time in which tion of birds drifting the other way in a thermal. The statistics to attempt it!

8 I Additional notes on the birds of Trinidad and Tobago

By Richard ffrench Beech Ridge, Bussage, Stroud, Glos., England

AS I prepare to return to my native land after more than a quar­ properly documented record of the species from Tobago, and the ter century in Trinidad, I am pleased to submit to this journal a most southerly record in the West Indies, from which it was pre­ final update on the status of local birds, based on records over viously known south to St Vincent and Curacao. the period 1981 to 1984. I am particularly gratified that I can include a number of observations made by Trinidadian natura­ SNOW GOOSE Chen coerulescens lists, as this indicates that local interest in ornithology is develop­ Previously recorded only on a single occasion in Trinidad ing and will continue to grow. when one individual was seen at Caroni (ffrench 1977). Four The following list of twenty-two species includes two new immature birds of the "blue" phase of this species were seen in records for Trinidad and four for Tobago (some of which pre­ the Caroni marshes on 30 December 1984 by a large party of ob­ viously required confirmation); also one doubtful breeding re­ servers, including Winston Nanan, Victor Quesnel, my wife and cord for Tobago is now confirmed. myself. Some of us approached to within 50 or 60 metres and obtained good views of the birds which were in close proximity THE SPECIES LIST to some tree-ducks. A report of four "geese" in the Oropouche Lagoon at about this time may well refer to the same birds. The LEAST GREBE Tachybaptus dominicus species, which winters mainly in U.S.A., is extremely rare south It is now confirmed that this species inhabits Tobago, pro­ of Mexico and Cuba. bably as a resident. Following reports from visiting naturalists, I found at least one pair at Hillsborough Dam in 1983 on 11 April BLUE-WINGED TEAL Anas discors and 23 October. That the species is quick to exploit suitable An adult male was seen in the Caroni marshes on 26 June habitat is shown by the presence in 1984 of several breeding pairs 1983. This common North American duck normally winters here at small artificial ponds in Santa Cruz, Trinidad, not long after between October and April, earliest records being mid-September their construction. and the latest May 10. Presumably this unsual summer record is of a bird that may have been crippled by a hunter. LEACH'S STORM-PETREL Oceanodroma leucorhoa Though commonly recorded off Trinidad on migration, there GRAY-HEADED KITE Leptodon cayanensis were no Tobago records of this species before 9 April 1983 when The species should probably be considered an uncommon I saw four individuals from the ferry-boat a mile or two off (rather than rare) resident. It has in recent years been recorded Scarborough. There is no reason to believe that its migration past over , open woodland and swamp forest in various parts of Tobago is anything but a regular annual occurrence at this time. Trinidad. Victor Quesnel finds it a regular inhabitant of woodland near Talparo; and Tim Manolis found an adult at Carapo carrying RED-BILLED TROPICBIRD Phaethon aethereus probable nest material on 29 April 1981. The first record of this species in Trinidad offshore waters occurred on 23 April 1983 when I saw a bird settled on the sur­ YELLOW -HEADED CARACARA Milvago chimachima face of the sea, from which it shortly afterwards flew. I was at the Breeding of this species on Trinidad, formerly only sus­ time on the inter-island ferry a mile or two off Trinidad's north pected, was confirmed when a nest was found in early 1984 on coast. Usually the species prefers the clearer waters found off Waller Field by J ogie Ramlal and others. The species continues Tobago. to spread throughout Trinidad (ffrench 1985).

WHITE-TAILED TROPICBIRD Phaethon lepturus COMMON PIPING Aburria pipile I explained in my book (ffrench 1973) and in an earlier Fonnerly known as the Trinidad (Pipile pipile) article (ffrench 1961) why I rejected previous reports of this or the Blue-throated Piping Guan (Pipile cumanensis) this species species on Tobago. I still reject those reports. However, on 29 has recently been reclassified as above (A.O.U. 1983). Recent December 1983 Bob Richardson found an adult of this species efforts by mem bers of the Forestry Division to locate remaining associating with a number of its congeners at Little Tobago populations of the species on Trinidad have succeeded in finding island. Careful documentation and illustrations accompany the re­ very small numbers in the northeastern during the early .cord, so I see no reason to doubt it. This seems to be the first dry season of 1983. Also two birds were seen by Frank Pitelka

9 (of California, U.S.A.) and Jogie Ramlal in the Aripo area in more commonly in the Caribbean area during recent years. The December 1982. only previous records from Trinidad waters made in 1969 (ffrench 1973) and 1982 (ffrench 1984) are now augmented by RUFOUS-NECKED WOOD-RAIL Aramides axillaris records made by Bob Richardson, who found seven individuals, There have been no records of this species on the mainland two of them in the light phase, just off Forest Point, Toco on 2 of Trinidad for many years, although it is not uncommon on off­ January 1983, associating with several other seabird species, shore islands in the northwest. However, Tim Manolis had a including three Parasitic Jaegers, S. parasiticus. Another good observation of this skulking rail near the Cacandee sluice­ pomarinus was seen on 9 January 1983 on the north coast near gate on 24 October 1981. The chestnut-brown head, neck, breast Yarra. Jaegers, or sku as, nonnally feed by attacking gulls and and wings were noted. On the mainland the species has been terns and forcing them to disgorge their prey; so they can usually found only among mangroves. be found in proximity to feeding groups of these smaller seabirds.

AZURE GALLINULE Porphyrula [lavirostris ROY AL TERN Sterna maxima Following a brief note of the first reported occurrence of An interesting banding recovery of this species occurred on this species on Trinidad in July 1978 (ffrench 1981), it has been 19 January 1983 when an oiled bird was brought to me at seen by numerous observers in the Nariva Swamp, especially near Pointe-a-Pierre; it bore a band which had been placed on it Bush-Bush, but also in the northern part near Sand Hill and as a chick at Cape Hatteras, North Caronina on 20 June 1976. Caltoo. As there is no record of migration in this species, and it This is only the third recovery of a foreign-banded Royal Tern seems unlikely that it would have gone unnoticed by trained on Trinidad. observers such as Wilbur Downs, Brooke Worth and Fernando Nottebohm, all of whom worked extensively in the Nariva ROSEATE TERN Sterna dougallii -Bush-Bush area between 1960 and 1970, the new records After searching in vain for more than two decades, I was probably indicate a range extension. In South America the species most gratified to locate a small breeding colony of this species appears to be locally abundant but spottily distributed, being on an island off Buccoo Point, Tobago on 22 May 1983. known from Delta Amacuro in , only 80 km distant Although a colony of unidentified white terns was found by from Nariva. Dinsm ore near Speyside in 1966, and I have several times witnessed copulation and other beeeding activity in the species, AMERICAN OYSTERCATCHER Haematopus palliatus this is the first indubitable record of nests and unfledged young Although apparently unrecorded on either island dUring this from Tobago. As I intimated earlier (ffrench 1979), it seems century, an individual of this species was found by David Rooks likely that intensive investigation into seabird breeding on Tobago on 31 August 1982 at Arnos Vale, Tobago, following a report and its offshore islands will also reveal nesting of the Royal Tern, from Mr and Mrs John Cartwright of Canada. David had a good S. maxima and the Sandwich Tern, S. sandvicensis, so far view of this distinctive shorebird at close range and heard its unrecorded. characteristic, penetrating call-note. The species is not commonly found in the southern Antilles, but I once saw a vagrant FORK-TAILED PALM-SWIFT Reinarda squamata individual on Barbados; it breeds in small numbers on Curacao, Tim Manolis first recognised that this species seems to be Aruba and Bonaire. much more widespread on Trinidad than hitherto, possibly because it was not always recognised (ffrench & Manolis 1984). MARBLED GODWIT Limosa fedoa In recent years it has been found by several observers, especially Following my last record of this rare species (ffrench 1981), David Rooks, not only at the "traditional" sites of Waller I am pleased to report that it has been recorded on three Field, Aripo Savanna, Nariva Swamp and the Botanic Gardens, occasions since then; a single bird seen at Carli Bay in September Port of Spain, but also all round the Queen's Park Savannah, 1.981 by T. Manolis, B. Mohan and G. Rarndeen, another seen nesting in various species of palms, as at Bishop Anstey's School by myself, Jim and Jane Yarrow at Pointe-a-Pierre on 7 October during September 1983 and October 1984.* Also on the south­ 1984, and a sizeable flock seen by David Rooks on the Savannah, western peninsula near Icacos and in the Santa Cruz Valley. Port of Spain at about the same time. I would like to think that Although moriche palms (Mauritia) are favoured as roost and nest this handsome large shorebird is becoming more common. sites, clearly the species also utilizes other palm species. * Victor Quesnel has records of this species at Woodbrook DOUBLE-STRIPED THICK-KNEE Burhinus bistriatus and Mucurapo dating back to 1954 - Ed. Some comment seems necessary on the interesting record of the first occurrence of this species on Trinidad in June 1983 SCALED ANT-PITT A Grallaria guatimalensis (Rooks 1984). Thick-knees and stone curlews resemble large It seems somehow fitting that I should leave Trinidad after plovers but form a separate family, represented in the Americas 27 years without having directly encountered at least one of its by only two species. This species is found from Mexico south to resident bird species. This one, represented by an endemic race, , also on the island of Hispaniola. In Venezuela it has eluded me, except that I strongly suspect it may have been occurs in open savannah, where I have seen it near Barinas, and it the author of an unfamiliar low hooting cal!, heard on the is also found on Margarita island. As it is not known to migrate, slopes of Naranja on the El Tucuche trail on 28 November 1982. the Trinidad record may represent a. post-breeding dispersal. Unfortunately I could not locate the bird, and have not heard Henceforth it may be worth looking out for it during June and the call since. However, Peter Hall from U.S.A. had an excellent July on the Port of Spain savannah, on golf courses and at Waller view of one, well documented and described, at Springhill estate Field or Aripo Savannah. on 4 January 1981. Though uncorroborated by other observers, his sighting was over several minutes at very close quarters in POMARINE JAEGER Stercorarius pomarinus good light conditions, so there seems no reason to doubt it. This large and powerful seabird has been seen and recognised Clearly the species remains extremely rare on Trinidad.

10 sclateri Middle America and West Indies). 6th edition. American Graham White has seen an individual twice in the same Ornithologists Union, Washington, D.C. location near Cacandee, which seems most likely to be this FFRENCH, RICHARD 1961. The Red-billed Tropicbird. J. species. The dates, 13 February 1983 and 11 March 1984, seem Trinidad Field Naturalists' Club. 9 - 10. rather early for what has been hitherto consid~red a rare migrant FFRENCH, RICHARD 1967. The Dickcissel on its wintering from South America. But it is possible that this tiny and incon­ grounds in Trinidad. Living Bird 6: 123 - 140. spicuous flycatcher has been resident here in small numbers all FFRENCH, RICHARD 1973. A Guide to the Birds of Trinidad the time. Certainly very little seems to be known about its life and Tobago. Livingston Publishing Co., Pennsylvania. 470 history. pp. FFRENCH, RICHARD 1977. Some interesting bird records from BAY-BREASTED WARBLER Dendroica castanea Trinidad and Tobago. Living World (J. Trindad and Tobago An individual of this species was seen at Grafton, Tobago on Field Nat. Club). 9 - 10. 11 November 1980 by Charles E. Keller of U.S.A. This is the first FFRENCH, RICHARD 1979. More records of rare birds in record for Tobago, though four have been reported from Trinidad Trinidad and Tobago. Living World (J. Trinidad and Tobago (ffrench 1973). Field Nat. Club). 25 - 26. FFRENCH, RICHARD 1981. Some recent additions to the DICKCISSEL Spiza americana avifauna of Trinidad and Tobago. Living World (J. Trinidad This migratory finch was conspicuously common in south­ and Tobago Field Nat. Club). 35 - 36. western and eastern Trinidad during the 1960's (ffrench 1967), FFRENCH, RICHARD 1984. Further notes on the avifauna of but numbers dropped in the 1970's and none were recorded Trinidad and Tobago. Living World (1. Trinidad and Tobago during the years 1975-1981. In early 1982 I encountered a single Field Nat. Club). 32 - 34. flock of about 50 birds in Nariva, but none in 1983; then in 1984 FFRENCH, RICHARD 1985. Changes in the avifauna of Trini­ I found about 100 Dickcissels at Nariva in late March, while dad. Neotropical Ornithology (ed. P. Buckley et al.) Orni­ flocks of several hundreds had been seen at Oropouche Lagoon thological Monographs No. 36. American Ornithologists since early January. It seems possible that a new cycle of winter Union, Washington, D.C. pp 986 - 991. invasions by this interesting migrant may be beginning once again. FFRENCH, RICHARD and MANOLIS TIM 1984. Notes on some birds of Trinidad wetlands. Living World (J. Trinidad and Tobago Field Nat. Clu b). 29 - 31. REFERENCES ROOKS, DAVID 1984. First record of the Double-striped Thick­ knee in Trinidad West Indies. Living World (J. Trinidad and A.O.U. 1983. Check-list of North American Birds (including Tobago Field Nat. Club). 3.

11 Christmas counts of birds in Trinidad 1969 -1975 and 1981-1983

By Richard ffrench Beech Ridge, Bussage, Stroud, Glos., England

INTRODUCTION portion of Caroni Swamp, including the bird sanctuary. In prac­ tice the group concentrated on Las Lapas, the Asa Wright Nature Centre, La Laja, Waller Field and the Caroni Swamp with adjoin­ The results of ten "Christmas Counts" during the periods ing marshes. 1969-1983 are here presented and analyzed. Christmas Counts In 1981-1982 less time was spent on the above areas in order of birds are routinely made in many parts of North America to allow the inclusion of the Government Stock Farm in north under the auspices of the National Audubon Society. Starting Waller Field, St Raphael and the western section of the Arena with 27 observers on Christmas Day 1900, the counts now Forest, and Las Lomas. involve many thousands of cooperators, and the idea has spread In all cases the time spent observing approximated to twelve beyond the boundaries of U.S.A. into Canada, Latin America and hours, roughly 0600 to 1800, and the weather was usually fine, the West Indies. allowing maximum observation. However, there were sometimes The counts here described were first held under the auspices showers, and in two cases, 1971 and 1982, a great deal of light of the Florida Audubon Society as part of an initial attempt to rain fell, inhibiting the course of observation. incorporate regular Christmas Counts into the programme of In every case but two the observers worked in one single cooperators in neotropical countries. The results were published party, varying from three to six in number, but in 1981 and 1983 in a journal called "Pan American Naturalist". This arrangement thirteen and 27 observers formed two parties, one in the was soon terminated, so I carried on up to 1975 for my own Northern Range and one in the plains, and were thus able to interest. Since 1976 counts have continued and have been spend longer in their respective areas. However, no birds were published in "American Birds", an official publication of the counted outside the originally chosen circle. National Audubon Society, but there has been no comparative analysis of the figures to date. RESULTS AND DISCUSSION I here present the results of ten years' cO)lnts, these being the years when I myself attended and organised the studies. During Table 1 summarises the counts by years, giving total figures. the five intermediate years 1976 - 1980 counts were made, but Table 2 lists all 201 species recorded during the ten counts, and by very variable numbers of observers, and occasionally in a gives totals of individuals in each species year by year. slightly different area. As I feel unable to evaluate these five Table 2 indicates roughly those species that are usually counts objectively, I have omitted them from this study. abundant or commonly seen in the count area. Of the 201 species 52 were recorded in each count, ranging from the Scarlet Ibis METHODS whose numbers always topped 1000 to the Yellow-bellied Elaenia whose numbers ranged from 1 to 7. All of these 52 species (and These Christmas Counts follow the rules laid down interna­ several others that were recorded in nine out of ten counts) can tionally for all participants under the National Audubon Society's be considered "common" in the sense that one can expect to jurisdiction. One area, forming a circle with diameter of 15 miles, see them on any outing. However, a bird like the Osprey may be is studied each year at about Christmas time. On a single chosen invariably present without being common, since because of calendar day within the period, from midnight to midnight, all limited habitat and territory size it is never likely to be present in the birds found within the area are recorded. The number of large numbers. So the list may be useful, especially to beginners, participating observers is unlimited, and they may be divided since it gives some idea of those species with are likely to be seen into different groups, provided the same birds are not counted (or heard). by more than one group. All the species that are recognised are The figures also reflect some significant changes. In the recorded, along with the num bers of individuals. Although period 1981-1983, when north Waller Field, San Raphael and Las complete accuracy is the aim, in the case of large flocks of birds Lomas were included for the first time, species such as the a reasonable estimate of num bers is allowed. Southern Lapwing, Savannah Hawk and Yellow-rumped Cacique The circle chosen in Trinidad had its centre at Arouca (Fig. were recorded, all of them absent before. In addition, Wattled 1). Thus the places covered included the Arima Valley north to Jacana and Orange-winged Parrot numbers reflect these added the watershed up to Las Lapas Trace, La Laja road, the western habitats, and White-headed Marsh-Tyrants were now seen in much portion of Waller Field, Piarco, St Augustine and the eastern greater numbers. On the converse side, Common Potoos were no

12 longer seen, since their favoured habitat was Cacandee, an area no does thrive on aerial that associate with cattle pastures, longer covered after 1971. so it is not too surprising to find them common on the Waller More significant changes are seen for those species affected Field stock farm. by the salt-water infiltration into the savannah and fresh-water Finally, significant decreases can be seen in Scarlet Ibis and marsh east of the Caroni North-South Drain. Though this infiltra­ Ruddy-breasted Seedeater. The former has changed its dispersal tion predates 1960, its effect has been gradual (and to date con­ habits considerably since the late 1960's, and it is also a difficult tinues), killing the fresh-water marsh-plants and creating shallow bird to count, always seeming to be more numerous than it really brackish lagoons, fringed or covered with mangrove saplings. This is, possibly because of the sheer spectacle it affords when in large development has no doubt been at least partly instrumental in flocks. My personal view is that its numbers have been adversely causing the notable desertion of the Bird Sanctuary roost-site affected, though not seriously, by the failure of Government to by several heron species, including the Great Egret, Snowy Egret, manage the tourist traffic on a proper basis. The same authori­ Cattle Egret, Little Blue Heron and Tricolored Heron. All these ties are also indirectly responsible for the notable disappearance species now roost much further east, nearer to the highway. The of the once-common Ruddy-breasted Seedeater. Owing to the significant fall in numbers observed on the counts may not nec­ gradual destruction by trapping of several other finch species - cessarily imply that these birds have diminished; it may merely a practice that the Government has done nothing to reduce indicate a change to a new roost-site, less accessible to observa­ trappers have recently turned to this seedeater, hitherto con­ tion. The Yellow-throated Spinetail, another fresh-water marsh­ sidered too small to bother with, so that now it too will follow its bird, has similarly been driven east. congeners into oblivion within a decade or two. However, the above development has converted areas, hitherto covered by marsh-plants, to open brackish lagoons, ACKNOWLEDGMENTS where increased numbers of Pied-billed Grebes and Common Gallinules are now to be found. I should like to thank all these who have participated in the Certain gregarious species show striking increases in numbers counts. They were Margaret, Simon and Julie ffrench, Daniel for some years, apparently without cause. These include the Lehnnann, Patrick, Ruth and Andrew Haynes, Geoff and Kate Black Vulture and three , the Shiny Cowbird, Carib Gibbs, Mike Harris, D. and M. Watson, Ian Walker, Richard Grackle and Yellow-hooded Blackbird. These sudden, and not Barchet, Harvey Gilston, Delia, Jeremiah and Sam Adams, Paddy sustained, increases may simply be ascribed to the chance ob­ Bell, Bob and Janet Kennedy, Bob Merritt, Jogie and Roodal servation of large flocks on certain days. Ramlal, John Perreira, Victor Quesnel, Paul Christopher, Ian Some species, however, do seem genuinely to have increased Lambie, Edward Rooks, Sylvia and Ann Marie Kacal, John and in numbers. Among these, the Ornate Hawk-Eagle has been more Dorothy Truman, Graham White, Winston Nanan, Flossie and in evidence over the Northern Range in later years. Once its call Melissa Cowell, Betty and Leo Kaufman, Donna Peterson, Mary can be recognized it is likely that an observer will record this and Tom Wood, Leonard and Lillian Wurzel, Lynne Yaskot. species on any single day's outing in the hills; 15 to 20 years ago Winston Nanan very kindly allowed us the use of his boat on this could not be said. At Springhill the Oilbird population has three occasions. steadily increased, in spite of occasional setbacks, under the influence of protection afforded by the Nature Centre. My own Postscript feeling is that a number of 150 is probably saturation point, given the size of the appropriate habitat, and that an occasional Dare I hope that some members of the Trinidad and Tobago exodus of 30 or so subadult birds must take place from time to Field Naturalists' Club may be influenced by the above report to time; which could account for the fluctuations over 100. I have lend their energies towards future counts, so that the areas to admit to b~ing puzzled by the sudden appearance of the covered may be more thoroughly studied, and thus enable a migrant Barn Swallow on the list from 1974 onwards. What seems regular monitoring system of a least one part of Trinidad to be odd is that it was not noticed during 1969-1973. But the species established?

TABLE 1

Summary of Christmas Count data, Trinidad, 1969-1975, and 1981-1983.

Year Observers Parties Species Individuals Remarks

1969 4 117 5176 Fine; edge of Waller Field only 1970 6 108 9958 Few showers; included Cacandee and Centeno

1971 5 113 13249 Drizzle for half the day 1972 6 110 12255 2 hours rain midday. No.9 Drain instead of Cacandee

13 1973 5 116 9925 Fine 1974 5 116 7065 Fine 1975 3 112 11759 Mostly fine; rain at Las Lapas

* * * * * * * * * * * *

1981 13 2 135 8380 Fine; area included Las Lomas, Arena Forest, N. WalIer Field.

1982 3 125 6500 Showery 1983 27 2 148 5815 Mostly fine

TABLE 2

Species and individuals recorded on Christmas Counts 1969 - 1975 & 1981 - 1983

Species 1969 1970 1971 1972 1973 1974 1975 1981 1982 1983

Little Tinamou 6 2 2 4 2 5 Pied-billed Grebe 2 2 4 4 12 10 6 Anhinga 1 2 2 Magnificent Frigatebird 10 1 White-necked Heron 1 Great Egret 28 38 48 68 120 70 20 13 8 13 Snowy Egret 14 56 12 70 122 1 4 3 Little Blue Heron 212 113 123 70 118 21 18 23 86 22 Tricolored Heron 376 147 370 128 353 160 54 32 38 7 Striated Heron 3 2 8 2 2 2 4 2 Cattle Egret 1120 6982 5938 6700 4100 2500 5200 714 795 280 ~lack-crowned Night-Heron 1 12 Yellow-crowned Night-Heron 9 4 11 3 Boat-billed Heron Scarlet Ibis 1500 1446 4992 4162 3800 3248 5601 5000 3500 3500 Black-bellied Tree-Duck 2 American Wigeon 8 2 Blue-winged Teal 89 7 70 20 King Vulture 1 Black Vulture 46 77 127 59 165 123 73 1133 181 701 Vulture 3 2 6 3 3 5 7 7 7 19 Gray-headed Kite 1 Dou b1e-toothed Kite 3 Short-tailed Hawk 1 2 1 2 Gray Hawk 2 White Hawk 2 1 2 Savanna Hawk 5 6 4 Common Black Hawk 2 9 2 2 4 2 4 Great Black Hawk 1 Ornate Hawk-Eagle 4 Long-winged Harrier Osprey 2 4 2 5 5 5 5 2 6 8

14 Species 1969 1970 1971 1972 1973 1974 1975 1981 1982 1983

Yellow-headed Caracara Peregrine Bat Falcon 3 4 3 Merlin 2 1 2 Gray-breasted Crake Clapper Rail 4 2 2 2 2 2 Gray-necked Wood-Rail 2 Common Gallinule 4 4 12 2 11 4 49 55 38 Purple Gallinule 1 6 2 2 4 Wattled Jacana 8 31 21 2 6 6 8 38 59 49 Southern Lapwing 24 15 19 Black-bellied Plover 7 10 22 Solitary Sandpiper 6 2 3 3 5 Lesser Yellowlegs 10 79 49 5 6 8 9 Greater Yellowlegs 4 1 11 7 3 Spotted Sandpiper 21 30 58 30 71 14 35 30 20 41 Willet 2 6 Least Sandpiper 4 17 Pectoral Sandpiper 5 6 Semipalmated Sandpiper 40 2 Common Snipe 1 Short-billed Dowitcher Scaled Pigeon 3 5 12 22 Pale-vented Pigeon Eared Dove 7 5 21 2 2 8 4 2 Common Ground-Dove 5 7 4 4 Plain-breasted Ground-Dove 2 1 22 2 Ruddy Ground-Dove 21 25 34 29 28 46 13 56 58 36 White-tipped Dove 1 4 1 2 2 Gray-fronted Dove 2 2 2 2 2 7 Green-rumped Parrot1et 6 3 3 1 33 2 30 Lilac-tailed Parrot1et 14 4 14 12 4 Blue-headed Parrot 1 10 6 10 .Orange-winged Parrot 16 8 4 2 2 21 75 29 Mangrove Cuckoo 1 Squirrel Cuckoo 2 2 Greater Ani 6 2 Smooth-billed Ani 45 70 29 39 27 36 21 35 35 18 Striped Cuckoo 3 2 4 6 Ferruginous Pygmy-Owl 4 2 5 2 8 4 5 Oilbird 36 37 43 43 49 64 74 123 105 50 Common Po too 4 7 4 1 Semicollared Nighthawk Pauraque White-tailed White-collared Swift Chestnut-collared Swift Gray-rumped Swift 100 31 26 10 50 17 56 30 12

15 Species 1969 1970 1971 1972 1973 1974 1975 1981 1982 1983

Band-rumped Swift 29 2 2 7 2 6 5 10 Short-tailed Swift 45 9 32 32 58 6 45 3 6 21 Lesser Swallow-tailed Swift 2 1 Rufous-breasted 6 2 3 4 4 3 4 9 Green Hermit 2 7 9 18 6 6 7 4 4 3 Little Hermit 2 1 1 5 4 5 White-necked Jacobin Brown 4 4 Green-throated Mango 1 Black-throated Mango 2 5 6 10 14 4 2 4 2 2 Ruby Topaz 3 1 Tufted Coquette 1 2 1 Blue-chinned Sapphire 4 4 4 3 2 3 4 White-chested Emerald 6 5 3 5 2 3 12 5 Copper-rumped 8 10 7 6 6 8 10 9 4 9 Long-billed Starthroat White-tailed Trogon 3 2 Collared Trogon 5 1 Violaceous Trogon 3 3 3 Belted Kingfisher Pygmy Kingfisher Blue-crowned Motmot 2 2 Rufous-tailed Jacamar 1 2 Channel-billed Toucan 11 5 8 13 6 5 13 9 17 12 Golden-olive Woodpecker 3 3 3 2 4 3 6 4 3 Chestnu t Woodpecker 1 2 2 2 Lineated Woodpecker 1 3 2 Red-rumped woodpecker 1 Crimson-crested Woodpecker Plain-brown Woodcreeper 2 2 BUff-throated Woodcreeper 1 2 1 2 2 Pale-breasted Spinetail 2 2 1 2 6 4 1 Stripe-breasted Spinetail 1 1 Yellow-throated Spinetail 16 10 10 7 4 4 4 6 2 8 Streaked Xenops Great Antshrike 2 4 2 3 3 Black-crested Antshrike 3 2 2 2 3 2 Barred Antshrike 5 3 5 3 5 6 12 4 4 Plain Antvireo White-flanked Antwren 4 3 White-bellied 2 2 2 3 2 6 Black-faced Antthrush 4 3 3 2 2 1 3 4 4 Bright-rumped Attila 2 2 1 Black-tailed Tityra 5 3 1 Bearded Bellbird 7 12 10 6 13 10 8 5 6 Golden-headed Manakin 12 7 8 9 3 2 8 5 19 White-bearded Manakin 8 8 10 6 2 10 5 20 Fork-tailed Flycatcher 18

16 Species 1969 1970 1971 1972 1973 1974 1975 1981 1982 1983

Pied Water-Tyrant 6 5 10 8 2 10 8 12 8 White-headed Marsh-Tyrant 7 4 3 3 1 6 19 16 12 Tropical Kingbird 12 12 13 20 15 15 26 22 16 19 Gray Kingbird Boat-billed Flycatcher 3 5 5 5 8 3 6 I Streaked Flycatcher 1 2 2 1 7 4 6 16 18 18 10 21 17 16 Dusky-capped Flycatcher Olive-sided Flycatcher Tropical 4 5 3 2 2 3 3 Euler's Flycatcher Yellow-olive Flycatcher 2 Yellow-breasted Flycatcher 3 3 3 3 6 2 Yellow-bellied Elaenia 3 2 4 5 7 6 4 5 Forest Elaenia 2 Sou them Beardless Tyrannulet 2 8 Slaty-capped Flycatcher 2 Ochre-bellied Flycatcher 2 3 5 2 2 White-winged Swallow Gray-breasted Martin 11 42 17 25 12 3 4 54 5 Rough-winged Swallow 2 1 2 3 2 1 Barn Swallow 17 4 73 115 7 Rufous-breasted Wren 6 4 6 14 5 10 8 8 8 7 House Wren 3 6 14 10 8 8 26 5 7 Tropical Mockingbird 3 24 10 14 14 25 17 26 10 12 Yellow-legged Thrush 3 Cocoa Thrush 4 5 8 10 15 3 27 11 8 Bare-eyed Thrush 1 2 2 2 6 2 2 White-necked Thrush 2 3 2 5 2 2 7 Long-billed Gnatwren 4 3 1 3 3 2 8 2 5 Rufous-browed Peppershrike 8 9 8 11 15 11 13 11 7 7 Chivi Vireo Golden-fronted Greenlet 9 7 8 7 11 12 2 2 12 Shiny Cowbird 515 220 2 10 49 15 22 20 52 15 Giant Cowbird 2 5 5 2 15 3 3 Crested 24 13 44 65 64 16 24 61 59 18 Yellow-rumped Cacique 17 16 21 Carib Grackle 529 38 77 41 71 50 59 44 640 134 Yellow-hooded Blackbird 27 18 115 8 31 14 2 6 4 7 Yellow Oriole 4 6 4 3 2 9 4 3 Red-breasted Blackbird 6 6 17 3 9 8 5 7 4 5 Black-and-White Warbler 2 1 Tropical Parula 2 4 2 2 Yellow Warbler 3 3 8 2 6 4 15 2 5 Northern Waterthrush 7 6 4 13 4 6 2 8 Masked Yellowthreat 1 American Redstart 1 2 8 2 2 5 2 2 Golden-crowned Warbler 2 3 3 2 2 5

17 species 1969 1970 1971 1972 1973 1974 1975 1981 1982 1983

Bananaquit 25 71 42 109 88 98 51 41 26 33 Bicolored Cone bill 1 2 1 3 9 3 Purple Honeycreeper 6 4 3 6 8 3 2 16 4 5 Red-legged Honeycreeper Green Honeycreeper 4 14 7 16 16 2 4 2 12 Blue Dacnis 1 6 4 6 6 7 Trinidad Euphonia 1 Violaceous 'Euphonia 8 7 7 8 3 12 6 12 Speckled 6 2 2 2 4 5 Turquoise Tanager 6 14 5 7 15 14 16 Bay-headed Tanager 8 15 5 22 14 10 2 26 10 27 Blue-gray Tanager 7 20 20 32 19 7 12 12 22 14 Pahn Tanager 2 17 23 23 17 9 24 23 17 24 Silver-beaked Tanager 4 9 2 10 6 9 5 11 20 12 Hepatic Tanager 2 1 7 3 5 Summer Tanager Red-crowned Ant-Tanager 4 2 4 2 White-lined Tanager 9 13 10 11 8 8 15 13 10 8 White-shouldered Tanager 1 2 2 J Grayish Saltator 1 4 9 8 3 8 Dickcissel Blue-black Grassquit 10 22 31 37 22 21 6 33 6 6 Yellow-bellied Seedeater 1 1 1 Ruddy-breasted Seedeater 10 8 3 14 7 6 4 2

FIG. 1 Christmas Count area in N.W. Trinidad.

L~TR,

ASA WR~GHT~,C, LA LAJA R ' • PORT-OF-SPAIN

ST. AUGUSTINE ARIMA • AROUCA WALLER ARONI • • L FIELD AMP

PIARCO

CAe NDEE CENTENO• ARENA • FOREST ' SAN RA :f(EL

Scale 5,072 m1 = 1 inch

18 Why do N ightjars sit on the roads at night?

By Victor C. Quesnel P.O. Box 47, Port of Spain, Trinidad, W.I.

ABSTRACT: Observations have shown that nightjars (Nycti­ always faced the centre. The remarks column came to contain dromus albicollis) sitting on the roads at night almost always sit additional information about the weather or the state of the road nearer the edge than the centre and face the centre. Numbers on at the site of observation ego whether there was grass growing at the road are influenced by moon phase being greatest near full the centre. All the observations were memorized and written moon and least at new moon. From these and other observations down immediately on returning home. Although on one arguments are developed to show that nightjars frequent the occasion eight observations had to be memorized usually roads primarily to feed but also to use them as stations from there were far fewer at anyone time. The burden on the which to call in establishing territories or in attracting mates. me,ory was lessened by the fact that the day's weather would That nightjars sit on the roads at night is well known to be the same for all and usually also the state of the road. In time naturalists. Their eyes shine red as they reflect the light from the one word "standard" would signify a position nearer the edge motor vehicles and the birds can usually be seen as they fly off on and the bird facing the centre and this too eased the burden on the approach of the car. When I went to live in Talparo in June the memory. There were occasions when some detail or other 1982 I saw them regularly as I travelled at night. I noticed that was forgotten and these suspect data were omitted from the the birds always seemed to sit near the side of the road facing the record or entered with question marks and treated appropriately centre so that in flying off they flew in front of the car from one in the analysis. Althcugh most observations were made from a side of the road to the other. Was this behaviour constant I moving car some were made from a stationary car some on foot wondered and why do the birds sit on the roads? ffrench (1973) and three from my study window. says of the Lesser Nighthawk, Chordeiles acutipennis; "At night Most of the observations (161,77%) were made on Leotaud they commonly sit on the wann pitch roads which cross the more Trace or the roads of Haven Hill Farm situated about 1 km from isolated savannahs" which suggests that they can detect the the main road on Leotaud Trace. Others (29,14%) were made on wannth of the roads and that they like it. Chapman (1894) says the 1 km stretch of the Tumpuna Road passing through El of the Pauraque, Nyctidromus albicollis: "they feed at night in Carmen Estate and some (17, 8%) on the Caroni South Bank the roads and footpaths". Here were two ideas to explore. Road at Las Lomas. One bird was seen on Lady Young Road but For the next six months I would be travelling back and forth no birds were seen on the Churchill-Roosevelt Highway between from Talparo to Petit Valley mainly at night and about five times Port of Spain and Arima, the Old Southern Main Road, the a week. This seemed to provide opportunity enough for a study. Caroni South Bank Road other than at Las Lomas and the This phase ran from 1 July 1982 to 31 Dec. 1982 after which Tumpuna Road between the Churchill-Roosevelt Highway and travelling became slightly less frequent and the series of observa­ Village. tions was discontinued. However, other series were begun later Arising out of the first series of observations a second series and there were many observations that were not part of the main was made in which records were kept of the estimated distance series. of the car from the bird at the instance it flew. In this series a mini tape recorder was used for recording the observations. This METHODS series was begun on 31st Oct. 1983 and ended on 13th Feb. 1984. The observations were confined to Leotaud Trace where Methods were simple. For the first series of observations a the state of the road made slow travel desirable. This restriction table was drawn up and the observations recorded by ticking or was imposed to reduce variation in the speed of the car. Lights writing in the appropriate column. The column headings were as were always on high beam. follows: Date, Time, Place, Day wet or dry, Road (at observation) A third series of observations on the sex ratio of birds seen wet or dry, Bird nearer centre or side, Bird facing centre or side, on the road was confined to Leotaud Trace and the roads of Remarks. The first three headings need no comment; some of the Haven Hill Farm. This was because on these roads it was possible others do. Observations on the weather and the state of the road to approach a sitting bird slowly and stop the car just in front of were meant to detennine if birds sat on the road less frequently it without courting danger from other cars. The bird could then after wet days or on wet rather than dry roads. Observations on be examined carefully in the light from the car. By this time (see the position of the bird on the road were meant to detennine if later) I had concluded that all the nightjars in the area were in fact the birds always stayed nearer the side than the centre and Pauraques and I had also learned how to detennine the sex of the

19 bird if it was clearly seen. The sex of each sitting bird was noted buff. In the light from a car the bars of both sexes may appear together with other information such as location, time and white at a distance but the bars of a female in flight appear to be whether the determination was made from the bird when sitting narrower. If a sitting bird allows a close approach the wing bar of or flying or both. the male can be seen to be brighter than that of the female. Also Leotaud Trace and the roads of Haven Hill Farm are about the white feathers of the male's tail can almost always be seen so half the width of the other roads where observations were made. that the determination of the sex of a sitting bird can be made if Leotaud Trace is paved in patched which make up perhaps half its it is seen from close up i.e. from the driver's seat of the car if length. There are numerous ruts and depressions and grass grows the bird is within a metre of the front of the car. When the in a strip at the centre in many places. The roads of Haven Hill bird flies an opportunity occurs for confirming the determination Farm are mostly dirt roads but the driveway to the house is paved from the width and brightness of the wing bars and the pre­ with oilsand. Here too grass grows along the centre. sence or absence of the white stripes in the tail. It should be noted, though, that when birds are approached that closely Species and sex identification they often do not fly in front of the car bu t quickly turn and fly off the near side of the road and outside of the illuminated area. In the first series of observations I simply recorded any night­ jar on the road and thought for a time that there were two RESULTS species, a larger one with prominent white wing bars and a slight­ ly smaller one with less conspicuous and narrower wing bars and Position on the road with no white stripes in the tail. I now believe that all the obser­ vations, with the possible exception of the one on Lady Young In the first series there were 208 observations. Eleven with Road, refer to the Pauraque, Nyctidromus albicollis. The evidence some uncertainties are now omitted from the analysis; the re­ is as follows. maining 197 are shown on Table 1. In 141 observations (70%) Despite the fact that many of the observations were made the birds sat nearer the edge than the centre and in a further from a moving car there were times when I could approach the 16 they were halfway between edge and centre. In 195 observa­ bird very closely, stop the car less than a metre from it and study tions (99%) the birds faced the centre. Only two birds (1 %) were its shape, colour and pattern for several seconds in the light from nearer the centre than the side and faced the side. No bird in this the car. In every such instance the lJird was a long-tailed species series was seen nearer the side than the centre and facing the side that conformed in appearance with the descriptions and illustra­ bu t one was seen in this position after the series was completed. tions of the Pauraque in Herklots (1961), ffrench (1973) and Also, no bird was ever seen facing down the road i.e. with its Dunning (1982) and with the specimens of the Pauraque in the body axis parallel to the edge. collection at the Caribbean Epidemiology Centre, formerly the Trinidad Regional Virus Laboratory. The white bib was rarely seen because of the bird's "hunched down" posture on the road and the fact that it always presented its lateral aspect in these Table 1 situations. On the other hand, the white wing bars and the white stripes in the male's tail were often clearly seen as the bird flew. Num bers of nightjars observed on the roads in the positions indi­ I have several times watched birds on my driveway from my cated. study window and they all conformed to the description of the Pauraque and those that called gave the characteristic call of the nearer nearer species. (The white bib can be more easily seen· on calling birds side centre halfway Total because the throat swells outward with the call). My notes record that on one occasion, on 24th Sept. 1982, a bird on Leo­ Centre 141 38 16 195 taud Trace in front of my car gave the characteristic call of the Pauraque as I approached. There is the further fact that in three Side o 2 o 2 years of living at Haven Hill Farm I have never heard the call of any other member of the Caprimulgidae. I have heard the call of Total 141 40 16 197 the Common Potoo, Nyctibeus griseus, which is a larger bird of different habits that, as far as I know, has never been reported to sit on the roads. This evidence leads me to believe that all the observations made at Leotaud Trace and Haven Hill Farm (77%) refer to the Influence of the weather Pauraque. The stretch of the Tumpuna Road that passes through El Carmen Estate where 29 observations (14%) were made is only Twenty observations with some uncertainties are now 2 km from Leotaud Trace and Las Lomas where 17 observations ignored. The remaining 188 observations are divided between the (8%) were made is only 5 km away and at both localities the different com binations of weather and road conditions at the habitat is much like that at Leotaud Trace. Although the birds time of observation as shown in Table 2. Numbers on wet and at these localities were not studied as intensively as those at dry roads were practically identical indicating no preference for Leotaud Trace they were apparently the same species. It is dry roads. Forty two (22%) of observations were made after wet reasonable to conclude, therefore, that most if not all the obser­ days when there was little or no chance of the roads being detect­ vations can be attribu ted to the Pauraque. This conclusion had ably warm. Sixty seven (46%) of the remaining 146 were made been reached before the second series of observations was begun. after midnight when any warmth in the road would have been By the time the third series was begun the appearance of largely dissipated, so that in 68% of the observations the roads both sexes was familiar to me. According to Herklots (1961) the (in my view) would not have been warm enough to influence the wing bars of the male are white whereas those of the female are birds' behaviour.

20 Table 2. at or near full moon than at new moon. As my familiarity with their behaviour increased I came to Numbers of birds seen on the roads in relation to weather condi- believe that the birds always flew on to the roads at dusk regard­ tions. less of moon phase and thought that this would show up if only the observations made early in the evening were plotted. Figure Wet Dry Sl\owers Total IB shows the distribution obtained when observations made up to 1930 hrs are plotted against moon phase as before. There are still more birds on the road near full moon than near new moon Wet 36 5 54 95 but the difference does not seem to be so pronounced. However, only 31 observations are available and these are really too few to Dry 6 73 14 93 establish the birds' behavioural norm. It should be noted that 207 of the observations were made Total 42 78 68 188 between dusk and 0140 hrs. The remaining one was made at 0535 hrs - a bird on the driveway that called repeatedly just Influence of the moon before dawn. What happens between 0140 hrs and 0535 hrs I and how it might influence the shape of the histogram is un­ During the course of the study it began to seem that moon­ known. It should be noted too that no records were kept of light was influencing the birds' behaviour. When the observations whether the moon was actually visible at the time of observa­ are plotted with respect to moon phase the histogram shown in tion. Fig. 1 is obtained. In this figure details of shape, such as the trough between two peaks at full moon, are unimportant. These Flight in relation to the car details would have been influenced by the frequency of trips when observations could be made and the amount of disturbance Sometimes the birds allowed such a close approach I was of birds on the road by traffic ahead of me. It is the general shape afraid I might run over them. At other times they flew when that is important and this shows clearly that the birds are indeed the car was many metres off. This behaviour also appeared to be influenced by the moon. Birds were seen much more frequently related to moon phase with the closest approach being allowed 10 B

40 A

Q ->- -Q Z - 20 0z

10

5 10 is 20 25 30 DAYS

FIG. 1 Numbers of nightjars seen on the roads at different phases 9f the moon. (A) All observations. (B) Observations between sunset and 1930 hrs only. The cycle begins with new moon on day 1. The other phases vary over two days because the synodic lunar period of 29.53 days is not evenly divisable by 4 and because the actual moment of new moon on day 1 varies within a 24 hr period. 21 at full moon. The second series of observations was made to test places to sing from in proclaiming their territories and a fourth this possibility. The speed of the car (and the noise of the engine) as places for courtship and mating. Each will be discussed in are obvious factors in the situation and for this reason the obser­ turn. vations were restricted to Leotaud Trace where the condition of It has been shown above that in only 32% of the total num­ the road demanded a speed no greather than 24 kph. Even with ber of observations was there any reasonable chance that the this restriction I was not at all sure that meaningful results would birds were responding to the warm th of the roads. A considera­ be obtained. The distance from the car at which the bird flew tion of the influence of moonlight further reduces the possibility was plotted as before against moon phase. The points were so that the warmth of the roads is important. Figure 1 clearly shows scattered that no trend could be detected. tha t the birds are on the roads in greatest num bers at our near However, except at or near full moon the birds would not full moon and in greatly diminished numbers at new moon. Now, be in moonlight for the whole night. The observations were various studies in different parts of the world have shown that therefore divided into 3-hr periods of each day of the moon's rainfall tends to be greatest just before first quarter and just phase and those made when tde moon would be visible to the after last quarter (Quesnel 1972) and this seems to be true for birds were compared with those when the moon would not be Trinidad as well (Quesnel, unpublished 0 bs.) The driest days coin­ visible. Again, no difference was seen. The distance from the car cide roughly with full moon and new moon. Thus, the tWG times at which the birds flew varied widely for all conditions. Hence, of the lunar month when warm night-time roads would be most the hypothesis was not supported. common (full moon and new moon) are associated with both The lights of the car must surely seem less bright at full the maximum and minimum num bers of birds on the roads. moon than at new moon. If the speed of the car could be kept Clearly then the warmth of the roads is unimportant and the de­ absulutely constant would a difference show up? Such a study gree of moonlight is the major factor influencing the birds' has not been attempted. However, other observations made behaviour. after the second series had been completed suggest that individual Why is this so? Possibly, roads, especially dirt roads, re­ birds might react differently. Thus, when birds were seen at the flect light better than foliage (chlorophyll, after all, is designed same point on the road night after night and the probability is to absorb light) and the brighter the moon the greater the dif­ high that the same bird was being observed repeatedly, flight ference in the amount of reflected light. Unfortunately, light seemed to occur at more or less the same distance from the car meters in cameras are not sensitive enough to respond to this regardless of moon phase. This aspect requires further study. reflected light and I have not been able to test this possibility. I assume. that insects respond to the reflected light in the same DISCUSSION way they respond to electric light i.e. they fly towards it. Thus, the numbers of insects flying over roads should be greater than It is clear from the observations recorded above that the over foliage and the difference should be greatest at full moon Pauraque habitually sits nearer the edge of the road than the and least (perhaps negligible) at new moon. This then, is the centre, faces the centre rather than the edge and never sits with attraction of the roads for the birds. In my view Chapman its body axis parallel to that of the road. (1894) is right and the birds sit on the roads to feed. Forty seven of the observations were made on major, paved Terres (1980) says of the Pauraque in the United States: roads and the remaining 161 on Leotaud Trace or the roads "frequently flies low along road in front of car for 100 yds. or of Haven Hill Farm. In only two of the 47 observations more and can be seen feeding on insects in beams of car head­ (4.3%) was the bird nearer the centre than the edge. The other lights." This statement supports the conclusion just reached but 36 observations of birds nearer the centre than the edge (Table I am not sure I have ever seen a bird feed when flushed from the 1) were therefore made on the minor roads and constitute roadside by an approaching car. In my experience the bird 22% of those observations. Therefore, sitting nearer the centre almost always flies across the road in front of the car because than the edge is much more common on minor roads than on it almost always faces the other side. Only occasionally does major roads. In fact, although not rigorously documented, it is it turn and fly along the road or over the near side. clear that on major roads the birds usually sit very near the edge It is now possible to see why the birds sit at the edge of the i.e. within one or two body lengths of the edge. road facing the centre. If I am right about the road attracting One reason for this may be the simple fact that traffic is insects then in this position the bird is best placed to see the heavier on the major roads than on the minor roads; perhaps maximum num ber of them. If the birds were responding to the the birds have learned to keep out of the way. If this is so they warmth of the roads there is no obvious reason why the orienta­ are not always succesful for ffrench has seen dead birds on the tion on the road should be so constant. road (pers. comm.). Their behaviour on large open spaces where We can now return to the question' 'Why are birds nearer the there is no traffic is therefore relevant but I have made no ob­ centre more often on minor roads than on major roads?". My servations on this aspect of their behaviour. A second possi­ suggestion is that in this position they can detect both flying bility is that the birds need to be near cover. Eighteen of the 36 insects and those in the grass both in the centre of the road birds (50%) seen nearer the centre than the edge on minor roads and at the far side. If this is correct then the Pauraque must feed were at locations with grass growing at the centre. This may not at ground level as well as in the air. I know of no report that it have been accidental. The other eighteen, though not near grass does but Terres (1980) gives evidence of other nightjars feeding at the centre, were nearer the grass on the other side than they on the ground. Thus, of the Common would have been on a major road twice as wide. There is thus Nightjar, Chordeiles minor, he says that one bird's stomach some evidence for the second possibility. To consider a third contained 2175 ants. (Were these flying ants or not?). Of the possibility it is necessary to try to answer the question why does Lesser Nightjar, Chordeiles acutipennis, he says "after the Pauraque sit on the roads at night? dark also feed in open on ground" and of the Poor-will, One answer may be that it likes warm roads. This arises out Phanaenophilus nuttalis, "many insects picked up from ground or of the observation of ffrench (1973) quoted earlier. The second by leaping from ground into air for them". Thus, it is possible answer may be that of Chapman (1894) that they are on the tha t the Pauraque, too, feeds on the ground or picks insects roads to feed. A third may be that the birds use the roads as off low vegetation. This is obviously a subject for further study.

22 The birds fly out on to the roads at dusk and for some ning the third series of observations. It is still in progress at the minutes they may be observed before they become invisible in time of writing but the observations already made indicate that the gathering darkness. At such times I have watched the Pauraque both sexes are present with males being slightly more common from the window of my study and seen it "leap" into the air than females. This is the result expected if the roads are used (presumably to catch insects) as described by Terres for the both for feeding and for territorial advertisement. Poor-will. The "leap", really a short flight, usually takes the A final question may be raised. Did the Pauraque seek bird across the driveway and it returns to its position near the out open spaces before there were man-made roads or has the side of the road turning as it alights so as to face the centre Pauraque changed its habits (and habitat?) as roads became once more. The fact that this manoeuvre has taken place at available? We may never know the answer but the attempt to every return to the road from the opposite side confirms the discover it would surely reveal a lot more about the bird's conclusion arrived at from the first series of observations viz. biology. that the bird habitually faces the centre of the road. I A bird will frequently return many times to the same spot. REFERENCES however, it may also shift its position at every return. It can be seen, too, that a bird often sits in the same location, within a few CHAPMAN, F.M. (1894) On the birds of the island of Trinidad. centimetres, night after night for a week or ten days. It is rea­ Bull. Am. Mus. Nat. Hist. 6, 1-86. sonable to suppose that it is the same bird. Eventually it dis­ DUNNING, J.S. (1982) South American Land Birds - A photo­ appears and may not return to the same spot. However, the bird graphic aid to identification. Harrowood Books, Newtown that haunts my driveway has sat on the road opposite a con­ Square, Pennsylvania, 364 pp. spicuous white stone in the centre for several months. FFRENCH, R.P. (1973) A Guide to the Birds of Trinidad and Although this discussion has shown that feeding is the pri­ Tobago. Livingston Publishing Co., Wynnewood, Penn­ mary purpose for the presence of the Pauraque on the roads it is sylvania, U.S.A. 470 pp. not the only activity that takes place there. The Pauraque on HERKLOTS, G.A.C. (1961) The Birds of Trinidad and Tobago. my driveway calls repeatedly from its station opposite the stone Collins, St James Place, London, 287 Pp. and I have seen and heard others call from positions on roads. QUESNEL, V.C. (1972) Recent Studies of the moon's influence Roads seem to serve the same function for them as songposts on the weather. Astronews - Off. Pub. T'dad & T'bago. for in the establishment of breeding territories. Ob­ Asronom. Soc. servations now in progress indicate that the birds also use roads TERRES, J .K. (1982) The Audubon Society Encyclopedia of for courtship and possibly mating. North American Birds. Alfred A. Knopf. New York. 1110 pp. After the first draft of the paper had been written and the conclusion reached that feeding was the main reason for the pre­ sence of the Pauraque on the roads at night, it occurred to me Acknowledgment: I thank Richard ffrench for reading the first that a sex ration far from 1: 1 for birds on the road would be draft and making many useful suggestions for the improvement evidence against the conclusion. This was the reason for begin- of the paper.

23 The Underground Life of the Trinidad Worm - Lizard Amphisbaena alba

By John Riley Department of Biological Sciences, The University, Dundee

AMPHISBAENIANS, popularly known as worm-lizards (or in underground chambers interconnected by wide passageways (a Trinidad, two-headed snakes), belong to the reptile order large nest can contain several hundred chambers, each as big as a Squamata where they share an equivalent status with snakes football) and, as the compost becomes exhausted it is cut up and and lizards. There are about 140 species, found mainly in the stored in even larger underground cham bers which serve as refuse Americas and Africa, and all are blind burrowing animals. Many dumps. posses keel- or shovel-snouted heads to facilitate movement An early dig (by hand) of an Atta nest in the Maracas Valley through soils and, although very little is known of their ecology, produced a clear and unmistakable sighting of a large amphis­ most appear to exploit the properties of tunnel systems, feeding baenian travelling rapidly along a gallery beyond a fungus on a variety of prey species such as worms and . chamber which had just collapsed. Although we failed to catch it, Two amphisbaenians, Amphisbaena alba and A. fuliginosa, it seemed that the story of an association with ants was true. In a occur on the island of Trinidad and also on the mainland of subsequent excavation in the Arima Valley, this time with an South America. They are unlikely to be confused as adults since overloader driven by Krishna Ramdial of the SRC Quarry, A. alba is up to 70 cm in length, creamy white, pink or red-brown Blanchiseusse Road, another large specimen was dug out in colour, whereas A. fuliginosa is usually no longer than 30 cm unharmed. In all, seven excavations have yielded 3 worm-lizards. and is strikingly blotched in a random black and white pattern. These findings raise an obvious question; simply put, What is Amphisbaena alba, the subject of this communication, is one A. alba doing in ant nests? Is it feeding on ants and their grubs? of the world's largest worm-lizards and, though much feared in To answer these questions we began a careful analysis of the in­ Trinidad, is entirely without vemon. However, it does possess testinal contents, or faeces, of several amphisbaenians derived powerful jaws and will bite if carelessly handled although only from a variety of sources (road kills, excavated specimens, and able to inflict superficial wounds. Despite its large size and uni­ preserved museum specimens). Ants are frequently found in the form colouration it is rarely seen in Trinidad: it can occasionally intestines of these animals (in 7 out of 11 specimens) but usually be glimpsed crossing roads or jungle paths, particularly during the in very low num bers and they do not seem to be a significant wet season, and its unusual mode of progression makes it instant­ food item. Beetles and their larvae, by contrast, were quite ly recognizable. It does not wriggle; instead, waves of muscle con­ commonly found and these comprise an important part of their traction pass along the body enabling it to progress in a straight diet. line. This type of rectilinear locomotion is unique amongst the Further research showed that certain of the beetles found vertebrates and allows amphisbaenians to move backwards or in the intestines of A. alba (most notably the three-horned forwards with equal facility. If annoyed or alarmed the head and rhinoceros beetle Coelosis biloba. L.) only occur in the nests of tail are held aloft. Since both are bluntly rounded and the eyes leaf-cutting ants where they apparently feed on the spent com­ are scarcely developed, this behaviour is thought to confuse an post stored in the underground refuse cham bers. Our evidence aggressor into attacking the "wrong end" so that the head can indicates that the worm-lizbrds inhabit ant nests because these then be used in defence. sites support this protected food source of beetles and their A search of the scientific literature revealed that virtually larvae. The ants themselves appear to be consumed only nothing is known of the habits of A. alha, apart from two very accidentally by the worm-lizards. Other surface-dwelling beetles obscure references to its possible association with the nests of are also preyed upon and we surmise that the animals also leaf-cutting (bachac) ants. Subsequent conversation with Trini­ occasionally forage above ground. dadians well acquainted with the jungle confirmed this fact and The eyes of Amphisbaena alba are vestigial and the ques­ accordingly, in 1980, we began a study of this animal by (in the tion of how the worm-lizard locates ant nests was solved by first instance) attempting to dig it out of ant nests. an interesting series of experiments. In 1983 three specimens The ant in question, Atta cephalotes, builds large nests, up were caught whilst on the surface in a cocoa plantation in the to 30 m across, that appear in the jungle as conspicuous mounds Maracas Valley and they were kept in captivity for some months of excavated earth often covered by defoliated vegetation. The in order to study their behaviour: they readily adapt to captivity ants cut leaves of forest plants into fragments and take them into and can be maintained on a diet of liver. We were able to show the nest where they use them to make a compost upon which a that the lizards could follow the foraging trails of ants by using special fungus is cultivated; the fungus is used by the ants as their odours liberated by the ants themselves. sole food source. Fungus cultivation is carried out in numerous The worker ants of Atta cephalotes do not forage in the

24 jungle at random but follow well defined trails, which are nor­ fected is eaten by the final host can the parasite develop mally kept free of debris, to specific feeding sites. Ants lay down to maturity and begin laying again). a chemical trail, secreted from special glands in the tail, which is Ants cannot fulfill the role of intermediate hosts because followed and reinforced by other workers. In the dry season ants they can only suck fungus sap and are incapable of ingesting the forage mainly at night and by day the trails, are deserted, although parasite eggs even though these measure only 0.13 mm in dia­ the chemical trail apparently still persists. We were able to show, meter. But we have demonstrated experimentally that A. alba quite conclusively, that worm-lizards placed near to such trails faeces, introduced into laboratory colonies of Atta cephalotes, could detect chemical scent using their tongues and were able are rapidly cut up and thrown onto the refuse dump. Here, under unerringly to follow the ant trails. They could even follow the natural circumstances, they will be accidentally eaten by beetle trails by day when ants were absent. Simulated trails, made by larvae. We fed larvae of the three-horned rhinoceros' beetle (Coe­ scraping a cutlass through forest floor litter, were not followed losis bi/oba) on faecal material contaminated with parasite eggs and the worm-lizards sought immediate escape. Ant foraging trails and recovered the infective larval stages from the body cavity radiate out from large A tta nests for distances of several hundred 100 days l~tc:;:. Ll!rvae will not develop in cockroaches and it metres and we imagine that worm-lizards move from nest to nest seems that only Coelosis larvae offer the correct conditions for using these chemical highways. the development of the parasite to a stage which can then infect Clearly much work remains to be done on the nature of the the worm-lizard. Ants play an essential part in transmission by interaction between A. alba and Atta cephalotes but we have removing parasite eggs to the refuse dumps. This in turn en­ been able to uncover a further piece of evidence which suggests hances the probability that they will be picked up by Coelosis I that the association is regular. A. alba is commonly infected with larvae which in turn are a prey item of A. alba. The strong ecolo­ an parasitic lungworm called Raillietiella gigliolii. By gical links in the various components of this life-cycle offset a monitoring the faeces of worm-lizards for parasite eggs we have very low parasite production. shown that the parasite produces very few eggs; a mere 100 per parasite per day. Yet over 85% of amphisbaenians are infected which suggests a high transmission efficiency. In related parasites ACKNOWLEDGMENT insects are used as intermediate hosts (this means that the parasite eggs passed out in the host faeces are eaten, usually accidentally, I thank Hans Boos, Curator of the Emperor Valley Zoo, for by an insect in which larval development occurs. Only if an in- help with all phases of this work.

An unusual prey for the Marine Toad, Buto marinus.

By Victor C. Quesnel P.O. Box 47, Port of Spain, Trinidad W.I.

IN their review of the natural history of the marine toad, Bufo that must have fallen out of the wrens' nest below the garage marinus, which in Trinidad is called "crapaud", Zug and Zug roof. The head and neck were covered in mucUs and seemed (1979) say that it "will apparently eat almost every animate ob­ damaged, and the breast had lost a lot of feathers. The bird could ject it can catch" and "although terrestrial arthropods form the not hold up its head and I doubted that it would live but I put it bulk of its diet, snails, earthworms, and small vertebrates are also out of reach of the toad and went to bed. engulfed". Grant (1946) also noted that it would eat noxious The next morning the bird was holding its head up and the biting or stinging prey. It is easy to believe that among the ver­ wrens were feeding it. The following day it disappeared into the tebrates other species of amphibian, and perhaps its own young, nearby vegetation but its calls and those of the wrens indicated would be included, as has been reported, even that the occasional that it was still being fed and must have recovered completely. lizard could be consumed. For instance, Pippet (1975) records I have seen so many animals, lizards, snakes and even birds, three skinks among the stomach contents of marine toads in swallow prey that appeared to be too big for them that I have no Papua New Guinea. It is surprising that venomous snakes could, doubt that had I not intervened the toad would eventually have even occasionally, form part of its diet, yet Richard ffrench (pers. managed to manoeuvre the bird down its throat especially since comm) has seen a marine toad eating a small coral snake, Micru­ toads can use their hands to help them get their food in. Thus, we rus psyches eire ina lis, and it is hard to imagine that birds could can now add young birds to the list of vertebrates that toads will fall prey to it, yet this is exactly what this note records. eat if they can get them.

On my return home on the night of lOth July 1983 a strange REFERENCES animal that seemed to have wings sprouting from its head leapt away from the vicinity of the kitchen door and disappeared in the GRANT, Chapman (1956) Selection between Armed and darkness. Getting a flashlight I located the animal again and Unarmed Arthropods as Food by Various Animals. 1. Ent. found it to be a marine toad, not quite fully grown, that was Zool. 40 (3), 66. attempting to eat a bird. The bird's head and neck were well PIPPET, J.R. (1975) The Marine Toad, Bufo marin us, in Papua down in the toad's throat but its wings, protruding one from each New Guinea. Papua New Guinea Agric. Jour. 26 0), 23. side of the mouth, prevented the toad from swallowing it. Want­ ZUG G.R. and P.D. Zug (1979) The Marine Toad Bufo marin us: ing to know what species. of bird it was, I rescued it and found it A Natural History Resume of Native Populations. Smith­ to be a nearly fully fledged glossy cowbird,Molothrus bonariensis, sonian Contributions to Zoology, No. 254, 58 pp.

25 Notes on curious methods of locomotion in two Iguanids from Trinidad, West Indies.

By julius O. Boos c/o Hans Boos, Emperor Valley Zoo, Port of Spain, Trinidad, W. I. and F. C. Urich Sangre Grande, Trinidad, W.I.

WHILE working in the Forest Reserve - Palo Seco area of south­ over of a common West African belief related there to lizards of west Trinidad, I observed what I believe to be two unreported the family Chamaeleontidae which Po/yclIms closely resembles methods of locomotion for the Iguanid lizards Anulis chryso/epis both in appearance and its lethargic manner. planiceps TROSCHEL 1848, and Polychrus marmoratus (UN­ After examining the specimen I walked to the edge of the NAEUS 1758). well location where the ground dropped off in an extremely AnoUs chryso/epis planiceps is believed to be the only native steep slope of bare earth at the bottom of which the jungle Trinidadian anole. The other species, A. aeneus GRAY 184,0 and started. About half way down the slope, approximately 10 metres A. trinitatis REINHARDT & LUTKIN 1863, are believed to have away, standing alone was a small tree 2 or 3 metres high but still been introduced from Grenada and St Vincent respectively below the level at which 1 was standing. There was a light breeze (Gorman & Dessauer 1965, 1966). A. extremus GARMAN blowing directly into my face. Grasping the lizard at the base of 1888 was introduced to Huevos Island off Trinidad from its tail I flung it vigorously intending that it should fall softly Barbados (Boos 1967). into one of the trees at the bottom of the slope at the jungle'S A.c. planiceps is a forest-dwelling species, more usually heard edge. To my surprise, immediately upon leaving my hand the before it is seen scuttling across forest paths or within cultivated lizard flattened its entire body, with its ventral surface becoming cocoa fields. It rarely climbs except as a last resort, relying on its concave, and held its legs widely spread. It glided forward and up­ cryptic colouration for protection. ward into the light breeze and when it was approximately 2 metres Early one morning, driving slowly while entering Forest past the small tree previously mentioned it seemed to use its Reserve, I saw a lizard run from right to left across the road. Al­ tail as a rudder and, making a perfectly executed turn, landed though I was unable to collect it, I recognised it as A.c. planiccps. on the tree. I walked down the hill, recaptured the lizard and, It started from one side of the road at a noticeably slow pace in returning to ncar my original position, again threw it towards the normal quadripedal gait but within one metre rose on to its the jungle never expecting to see it repeat its manoeuver. hind legs with head high and drawn backwards, front legs pressed Bu t it did. This time owing to the direction of the throw it against its sides and the tail held off the ground in an upward could not quite make the distance and instead landed at the base curving arc. The hind legs were spread wide apart and seemed to of the small tree. [ eventually had to walk down the hill, catch have a slow windmilling motion. It ran thus almost in slow the lizard again, and put it on the trunk of a jungle tree lower motion to the other side of the road where it disappeared into the down. Once there it climbed slowly upward and disappeared into bush. This gait was different from that of Ameil'a (a teiid) which the upper branches. I have also seen use a bipedal mode. Ameiva use it only for long The gliding behaviour and ability of this lizard can easily be distances and in open spaces to get up to high speed while demonstrated by gently tossing it above one's head when it will running, for exam pie on the sand track in the Queen's Park be seen to flatten its body immediately in the manner described Savannah in the City of Port of Spain. above. About two weeks later I saw either the same lizard or Gliding has been noted in several other lizards, Pt),cizozoon another individual close to th~ spot of the original sighting again homalocep/zalulll (Gekkonidae), Draco Fo/ans (Agamidae), use a bipedal gait to cross the road. Bipedal running is well known Ho/aspis guenlizeri (Lacertidae), and A Il 0 lis caro/inensis in the Australian Frilled Lizard, Chlamydosaurus kingii. and the (lguanidae) (Bellairs 1969). Central American Basilisk, Basiliscus basiliscllS. Polychrus marmoratus is a medium-sized Iguanid, normally found on the tops of forest trees, on bushes and the trunks of REFERENCES trees along the road edges and it is quite commonly seen on avacado trees in neglected gardens. It tends to be solitary. The BELLAIRS, Angus (1969): The Life of Reptiles Vol 1&2- females are larger than the males and these females will display Weidenfield and Nicholson, London. xii+590pp. and fight violently when they encounter each other, an example BOOS, Hans (1967): Reptiles on lluevos - The Trinidad Field of sex-reversed territorial agression which is quite rare in the Naturalists' Club Journal. 1967, 15-18. family Iguanidae. GORMAN, George C. & H. C. DESSAUER (1965): Hemoglobin In 1978 I caugh t one of these lizards on a bush at the edge of and transferrin electrophoresis and relationships of island an oil-drilling location in Forest Reserve. The men of the drill­ populations of Anolis lizards. - Science 150,1454-1455. ing crew, like most Trinidadians, were terrified of this inoffensive GORMAN, George C. & H.C. DESSAUER (1966): The relation­ and completely harmless lizard. They called it "twenty-rour ships of Ano/is of the Roque! species group (Sauria: hours", believing that the person on whom it "jumps" or whom 19uanidae) - I Electrophoretic Comparison of Blood Pro­ it bites will be dead within twenty-four hours. This is a carry- teins.- Compo Biochem. Physio!. 19,845-853. 26 A Preliminary study of Culex mosquitoes attracted to a turtle in Trinidad, W.I.

By Dave D. Chadee, Robin C. Persad, Wayne Ramdath and R. Ganesh. Ministry of Health and Environment, Insect Vector Control Division, No.3 Queen Street, St. Joseph, Trinidad, West Indies.

DURING the course of investigations into a possible outbreak of tis in the neotropics (International Catalogue of Arboviruses equine encephalitis on a horse farm located in the Cumuto­ 1975) so disease transmission from turtle to man is a distinct Wallerfield area, Trinidad, W.I., a "galap" turtie, Rhinoclelllmys possibility. However, further study on the reptilian-vector pUllctularia (Daudin) was collected. Rizinoclelllm),s punctularia relationship is suggested. is distributed in northern South America, east of the Andes We wish to thank Ms. B. Hull and Dr. E.S. Tikasingh, (Pritchard 1979). In Trinidad, it is widely distributed through­ Caribbean Epidemiology Centre for assistance. In addition, out the island with numerous collections occurring in the Aripo­ special thanks should go to Messrs. E.C. , N. Andalcio and A. Wallerfield area and Guayaguayare Forest (Underwood 1963, Guerra for laboratory and field assistance. We also thank Dr. Chadeeetai. 1983-1984). 1. Q-ans of Ru tgers University for reviewing the manuscript. Turtles have been known to harbour both eastern and western equine encephalomyelitis (EEE, WEE) viruses in Northern America (Crans and Rockel 1968, Nolan et al 1965) REFERENCES and numerous mosquitoes have been collected from turtle baits (Crans and Rockel 1968, De Foliart 1967, Haynes 1961, Haynes AITKEN T.H.G., C.B. Worth and E.S. Tikasingh (1968) - 1965). In addition, Nolan et al. (1965) and Haynes (1965) Arbovirus studies in Bush Bush Forest, Trinidad, W.1. (Sept. reported Aedes canadensis Theobald feeding on the Painted 1959 - Dec. 1964) III - Entomologic Studies - Am. J. turtle, Chr),sefll),s piela (Schneider) and on the Eastern Box Trop. Med. & Hyg. - 17 (2) 253 - 268. turtle, Terrapene carolina (L) in Maryland and Southern Illinois. BEARD J .S. (1946) -. The Natural Vegetation of Trinidad - During the present study, mosquitoes were collected from Oxford Forestry Memoirs, No 2 Clarendon Press, Oxford the woodlands bordering the horse farm during August 1983. The 152 pp. vegetation consisted of Moriche palms (Mauritia setigera (L), CHADEE D.O., A. Le Maitre, R. Ganesh and R.C. Persad (1983- cocorite palms (Maximi/iana caribaea) and fatpork (Chr,1'sobalanus 1984) Ectoparasite of a "Galap" Rhinoclemm.l'S punctularia icaco) fringing the farm in association with a heavy undergrowth (Daudin) in Trinidad, W.l. Living World ( J. Trin. Tob. of the sedge, Sc/eria bractaeda (Beard 1946). Fld. Nat. Club) -po 45. On 8 August 1983, the turtle was collected and was sub­ CRAN S W.J. and E.G. Rockel (1968) - The mosquitoes attracted sequently exposed for 24 hours in a Trinidad No. 17 trap (Davies to turtles - Mosq. News 28: 327 - 337. 1971) on the 9 August 1983. All mosquitoes collected were DA VIES J .B. (1971) A small mosquito trap for use with animal aspirated into mosquito jars and transported to the Caribbean or carbon dioxide bait - Mosq. News 31: 441 - 443. Epidemiology Centre for identification. The collections were as DE FOLIART G.R. (1967) - Aedes canadensis (Theobald) follows:- Trinidad 13 -Il Culex declarator Dyar and Knab, feeding on Blanding turtles - J. Med. Ent. 4: 31. Cumuto-Wallerfield. 10 Vli.l983. D.O. Chadee, I!f. Culex comiger HAYES R.O. (1961) - Host preferences of Cu/iseta me/anura Theobald, same data. All specimens were unengorged and were and allied mosquitoes Mosq. News 21: 179 - 189. deposited in the Insect Reference Collection at the Caribbean HA YES J. (1965) New host record of Aedes canadensis Epidemiology Centre (CAREC). Mosq. News 25: 344. The collection of Cx. corniger in the turtle-baited trap is HEINEMANN S.G., T.H.G. Aitken and J.N. Belkin (1980) interesting as this species was not previously collected from Collection records of the project "Mosquitoes of Middle reptilian hosts. Aitken et al (1968) reported the collection of two America" 14. Trinidad and Tobago (TR, TRM, TOB) - adult Cx. corniger from Bush Bush Forest, Trinidad, W.I. but not Mosq. Syst. 12: 179 284. from the traps with the two reptilian baits - Turpinambis sp. and lNTERNATIONAL CATALOGUE OFARBOVIRUSES (1975)­ AmciFa ameil'a (L). One possible reason for our collections in this U.S. Dept. - Health, Education and Welfare - Pub!. No. area may be related to the habitat preferences of Cx. corniger (CDC) 75: I - 789. and Cx. dee/arator, for existing collection records indicate that NOLAN Jr. M. P., M.A. Moussa and D.E. Hayes (9165) - Aedes these species appear to favour open to partially shaded areas. musquitoes feeding on turtles in nature - Moslj. News 25: Culex corniger has been collected attacking man at 1300 hours 218-219. in the Turure Forest, Trinidad, W.I. (Heinemann et al 1980). PRITCHARD P.C.H. (1979) Encyclopedia of Turtles - T.F.H. These preliminary observations show that Cc corniger and Cx. Publications Inc. Ltd. Gr. Britain - p. 180. declarator may be attracted to turtles though there is no record, UNDERWOOD G. (1962) - Reptiles of the Eastern Caribbean - yet, that they engorge. Culex comigcr and Cx. dec/aralor have Caribbean Affairs (New Series) No. I Dept. Extra- Mural been found naturally infected with Venezuelan equine encephali- Studies, U.W.1.. P.O.S., Trinidad 192 p.p.

27 An Onychophoran from the summit of Mt. Aripo, Trinidad, with notes on other animals from that locality.

By V.M. St. 1. Read School of Animal Biology University College of North Wales, Bangor, Gwynedd, U.K.

INTRODUCTION the distribution of papillae on the dorsal body surface. The first two characters agree well with Epiperipatus ed­ Three species of onychophorans are known from Trinidad, wardsii (Blanchard). There are 32 pairs of feet in all five speci­ Macroperipatus torquatus (Kennel), Epiperipatus trinidadensis mens. The nephridial tubercles of the fourth and fifth pair are trinidadensis (Stuhlmann) and Epiperipatus imthurmi (Sclater), connected to the third arc of the sole by a narrow isthmus. The (Peck 1975, personal observations). Arnett (1961) states that he mandible has one accessory tooth on both the inner and outer has recently identified Epiperipatus edwardsii (Blanchard) from blade and seven denticles on the inner blade. Trinidad, but as he gives no other details and does not refer to Examination of an ecdysed skin by transmitted light shows a any specimens, this record must be considered as doubtful. distribution of papillae characteristic of E. edwardsii; the primary Early records of E. edwardsii from Trinidad e.g. Kennel (1883) papillae of the dorsal surface are of more or less equal size, are misidentifications of E. trinidadensis (Bouvier 1905). situated close together on the folds, not usually seperated by Recently, the author has found a peripatus on the summit of accessory papillae which are relatively rare and occur on the edges Mt. Aripo, Trinidad, which tentatively can be identified as E. of the folds. edwardsii. This is the most widespread American onychophoran Examination by scanning electron microscopy (fig. 1,2) how­ species, known from Brazil, Surinam, , Venezuela, ever reveals that although the basal areas of the papillae are and Panama, (Bouvier 1905, 1906, Brues 1914, 1925, similar, they vary considerably in height. Some are predominant Clark and Zetek 1946, Froelich 1968, Peck 1975). For compari­ over the others with a cylindrical or subsconical base and a well son E. trinidadensis and M. torquatus are not known outside developed apical piece and sensory bristle, others are shorter Trinidad and E. imthurmi occurs in Surinam, French Guiana, with conical bases, the apical piece reduced or absent. Generally and Colombia (Bouvier 1905, Brues 1914, 1925, Clark three of the smaller primary papillae occur between two large. 1915, Froelich 1968). This contradicts the description of E. edwardsii in Bouvier (1905) who states that there is little variation in the size of the primary COLLECTION DATA papillae in adults. In small specimens there are one or two smaller primary papillae between two large. An onychophoran was captured on 10th October 1982 on In life some of the largest primary papillae are yellowish or the summit of Mt. Aripo 940 m, the highest peak in Trinidad. white in colour and contrast strikingly with the brown body It was found among the basal leaves of a bromeliad Glomeropit­ colouration, forming six more or less continuous rows which cairnia erectiflora, about 1 m from the ground on a fallen tree. traverse the length of the animal on either side of the mid line. This large bromeliad grows abundantly in the canopy on the Yellow papillae are only present on alternate folds, although summit, but does not grow terrestrially here though it is found equally sized papillae are present on all folds. near the ground on fallen trees. The climate is cool, probably in The antennae and anterior part of the head are black and the region of 18-20 °C, and perpetually wet. Notes on other continuous with the dark brown mid line which is dilated seg­ animals found at this locality are given in the appendix. mentally into triangular shaped areas. The mid line is flanked by a lighter brown area, the remainder of the body being dark brown, (see fig 3). Though known only from preserved specimens whose DESCRIPTION colouration eventually fades, a similar pattern is present in E. edwardsii. The specimen was maintained alive for four months after which it was preserved. Its mean weight was 0.6124 g and did not FIELD IDENTIFICATION OF TRINIDAD ONYCHOPHORANS increase over this period so it was assumed to be fully grown. The specimen was female and gave birth to four offspring, two of The Aripo species is similar in appearance to E. trinidadensis which are being reared in captivity (see Table 1). and may easily be mistaken for it. It may be distinguished in the Determination of Neotropical onychophorans is notoriously field by the following criteria. difficult and depends chiefly on external characters, the number and morphology of the feet, mandible structure. and in particular 1. The greater number of feet. The 54 specimens of E. trini-

28 dadensis seen by the author had between 27 and 31 pairs of lack of good specific characters. Clark (1913) states that it is feet. None had 32 pairs and only three had 31. astonishingly easy to make mistakes in their determination. 2. The taller skin papillae. They give the body surface a rougher This is not the first time a specimen has been referred to appearance than that of E. trinidadensis. E. edwardsii also E. edwardsii with doubt. Froelich (196 8) assigns a specimen has a rather more elongate body when in motion. from Rio Guama, Bravil, with uncertainty to E. edwardsii but 3. Differences in intensity of yellow papillae. Both species have says it might be a new species. He was also unable to classify a a similar distribution of yellow papillae, but they are mostly specimen from Esperito Santo, Brazil, which he thought was pro­ not very distinct in E. trinidadensis. However two per seg­ bably conspecific with the example from the same state deter­ ment, situated in rows three and five, over each leg are much mined by Bouvier (1906) as E. edwardsii. Bouvier (1905) also brighter than the rest. In the Aripo species there is little lists specimens from Caracas and Lake Valencia, Venezuela, and variation in the intensity of these papillae, all of which are from Panarpa, which he assigns to E. edwardsii with doubt. bright. The author has examined, using scanning electron micro­ scopy, the skin of an example from Haut Sarare, Venezuela, pro­ I t may easily be distinguished from the other Trinidad bably one of the examples determined as E. edwardsii by Bouvier species which reach a larger size and lack a dorsal pattern. E. (1905). Its primary papillae are regular in size with conical bases imthurmi is a uniform brown while M. torquatus has a greater and it is certainly not conspecific with the Aripo species. A draw­ number of legs, 36 to 42 pairs, is red-brown and has a distinct ing of a primary pa:}illa from the type specimen, collected in yellow collar seperating the head from the body. French Guiana, in Bouvier (1905) is however similar to those of the Aripo specimen. This species is certainly near to E. edwardsii and E. trini­ dadensis, though is definately distinct from the latter. It is pos­ DISCUSSION sible that it should be regarded as a new species but a review of E. edwardsii is needed before a definite decision can be given. Onychophorans are usually terrestrial, living in leaf litter or holes in the soil. Until more data are available it is a matter for speculation as to whether peripatus actually live among the bromeliads and epiphytes in the canopy on Aripo, or whether TABLE 1 the specimen had accidentally found its way from the ground into this bromeliad which was in a fallen tree. Reproduction record of Aripo specimen (AI) from 10. Oct. 82 to Peripatus have nevertheless been recorded from bromeliads 15.Feb.83. previously: Macroperipatus ohausi (Bouvier) near Rid de J aniero, Brazil (Bouvier 1905); Epiperipatus biolleyi (Bouvier) at La WEIGHT Estrella, Costa Rica, altitude 2000 m, (Picado 1911). In the DATE AT WEIGHT cloud forests of Rancho Grande, Venezuela, peripatus inhabit PERIPATUS OF BIRTH ON NUMBER BIRTH detritus in the lower leaves of bromeliads and hollows high in (GRAMS) 5. NOV. 83 trees, (Beebe and Crane 1947). A2 9-17. Nov. 82 0.2993 The perpetual wetness, documented for at least two of these. A3 27.Dec.82 0.0400 localities, Aripo and Rancho Grande, would allow these animals, A4 13.Jan.83. 0.0407 0.2413 which are very prone to desiccation, and thus normally restricted AS l5.Feb.83 to the ground, to exploit the arboreal habitat. Onychophorans clim b well; M. torquatus often lives on vertical banks and on occasion has been seen 0.5 m high in trees (personal observa­ tions). A further factor may be waterlogging or flooding of the ground, making this habitat less suitable not only for. onycho­ phorans but for lizard and frog ovoposition sites as well (see appendix). Although it is close to E. edwardsii the distribution of skin papillae of the Aripo species somewhat resembles that of E. trinidadensis. However, in the latter species there is a much greater size difference between the small and the large primary papillae, the size of the former intergrading with the accessory papillae Which are much more numerous. An affinity with E. trinidadensis is also shown by the nature of the apical piece (fig. 2) which is large in size. This feature is also shared by E. edwardsii. In pro­ bably most other Epiperipatus the apical piece is short and conical. Present knowledge of E. edwardsii, and of the New World onychophora in general, is very fragmentary. As Froelich (1968) and Peck (1975) point out, this group is badly in need of review and revision. It is possible that in view of its wide distribution, E. edwardsii may prove to be more than one superficially similar species. Brues (1914) stated that it is possible that subspecific or varietal forms may be distinguished but was unable to find characters of sufficient constancy to warrent such a division. Neotropical onychophorans are notorious for their apparent Fig. 1

29 Fig. 1. (Previous pg.) Scanning electron micrograph of a sample of epidermis from mid-body. Note mid line at top of picture. Scale 200 urn. Fig. 2. Scanning electron micrograph. Large primary papilla from near mid line. scale 43 urn.

96: 205-213. FROELICH, C.G. 1968. On Some Brazillian Onychophores. Beitrage zur Neotropischen Fauna, 5(3): 160-171. KENNEL, J. von 1883. On the development of Peripatus. Nature, 29: 92-93. PECK, S.B. 1975. A review of the New World Onychophora with the description of a new cavernicolous and species from Jamaica. Psyche, 82: 314-358. PICADO, M. 1911. Sur un habitat nouveau des Peripatus. Bull. Mus. Nat. d'Hist. Nat., 17: 415-416. READ, V.M. St. J. 1983. A new locality record for the bromeliad dwelling hyHd Phyllodytes auratus (Boulenger) in Trinidad, the West Indies. Bull. Chicago Herp. Soc. 18(1): 30-31.

APPENDIX.

NOTES ON SOME OTHER ANIMALS FOUND ON THE SUM­ MIT OF MT. ARIPO Fig. 2 ACKNOWLEDGMENT In addition to the peripatus, the bromeliad-dwelling fauna in­ cluded large cockroaches, Dryadoblatta sp., chactid scorpions and The author wishes to thank the members of the Aripo expedition the frog Phyilodytes auratus. The latter species, which was pre­ John Seyjagat, Clyde Lilat and Richard Williams; also Dr. Jack viously know only from the type locality, the summit of Mt. Price of Simla Research Station for making his facilities available. El Tucuche, has already been reported by Read (1983). The This study was carried out while in receipt of a NERC student­ scorpions, which include several specimens from the summit ship. and one from a small bromeliad Vresia sp. on the ridge about two km west of the summit, were also known only from El LITERATURE CITED Tucuche. These have been sent to Julius Boos for determination. Two specimens of the montane lizard Proctoporus shrevei ARNETT, R.H. 1961. The Onychophora of Jamaica. Ent. News, were observed at night, one swimming in a pool of water trapped 72: 213-220. in the petiole of a fallen palm leaf, the other retreating into a BEEBE, W. and CRANE, J. 1947. Ecology of Rancho Grande a burrow under a fallen tree fern trunk. Fourteen eggs of this subtropical cloud forest in Northern Venezuela. Zoologica, species were found in a moss-covered bunch of aerial roots about 32(1): 43-59. 1 m from the ground. Ten of these adhered in pairs indicating BOUVIER, E. L. 1905. Monographie des Onychophores. 1, Ann. the usUal clutch size is two. Four other pairs were found, three Sci. Nat. Zool. (ser. 9); 2: 1-383. under moss on branches, and one in a bromeliad. BOUVIER, E.L. 1906. Le Peripatus edwardsi au Bresil. Bull. Soc. Eleutherodactylus eggs, almost certainly E. urichi, were Ent. France, 1906: 268. found in similar situations in the moss layer covering branches BRUES, C.T. 1914. A new Peripatus from Colombia. Bull. Mus. between one and two metres from the ground. At night males Compo Zool., 58: 375-382. of E. urichi were seen calling on the ground but not in the BRUES, C.T. 1925. Notes on Neotropical Onychophora. Psyche, branches. This frog presumably lays on the ground at lower 32: 159-165. elevations. CLARK, A.H. 1915. A note on the occurrence of Epiperipatus A Bushmaster, Lachesis muta, approx. length 1.5 m, was imthurmi (Sclater). Proc. BioI. Soc. Washington, 28: 182. caught on the ridge about 1.5 km east from where the Lalaja­ CLARK, A.H. and ZETEK, J. 1946. The Onychophores of Paria road crosses it. It is now in the possesion of Hans Boos at Panama and the Canal Zone. Proc. United States Nat. Mus., Emperor Valley Zoo, Trinidad.

30 Notes on the Life History of Catoblepia berecynthia berecynthia (Cramer) {Lepidoptera: Brassolidae)

By Julius o. Boos. c/o Hans Boos, Emperor Valley Zoo, Port of Spain, Trinidad.

ABSTRACT Subsequently, females of this were captured in secondary forest at bait mostly shortly after dawn, but some­ THE larva and pupa of Catoblep ia b. berecy nth ia (Cramer) are times just after dusk, in the Parrylands oilfield S. Trinidad, by figured, the latter for th.e first time. Significant differences be­ J.O. Boos. They were confined in a plastic container in which the tween the larva and pupa and those known in the genus Opsi­ lower half was dark green, the upper half transparent, with .ad­ phalles are discussed, as further possible diagnostic features to justable ports for ventilation. Freshly cut palm leaves were laid separate the genera. A method of obtaining ova in captivity is on the bottom and a small dish containing a mixture of over­ also discussed. ripe fruit, syrup and a small amount of white rum was initially placed in the container, but the fouled themselves in INTRODUCTION the mixture while attemp.ring to escape during their active periods at dusk and dawn. The dish was then removed and the butter­ In January 1980, a larva of Catoblepia b. berec)'nthia was flies were fed by removing them individually from the container collected on the leaves of the palm, Maximiliana caribaea with wings closed and placing them on the food. If they did not Crisebach & Wendl., on the edge of secondary jungle at Sangre feed voluntarily, their probosces were uncoiled into the liquid by Grande, Trinidad by F.C. Urich (Fig.l). The larva was cleansed careful manipulation with a pin. This usually caused feeding to of visible parasite eggs (possibly tachinid flies) adhering to its commernce immediatley. They were fed twice a day and given surface near the head and along its body, mainly at the base of water after feeding. the legs, and was raised to pupation on palm leaves. The pupa Depending on their reproductive condition upon capture, was brown and possessed a significant pair of cephalic horns they laid from two to six eggs per day over a one week period. (Fig. 2). As far as is known, this is the first record of the pupa The larvae were easily raised on a variety of palm species, and as of the genus Catoblepia. noted by Bristow (1981, p.l46, Fig. 10) were very variable" both between instars and between individuals of the same instaL All larvae possessed only three pairs of cephalic horns in all instars.

DISCUSSION

As discussed by Bristow (1981) there has been disagreement among various authors on the validity of Stichel's (1902) separa­ tion of Catoblepia from Opsiphanes based on the shape and length of the fore- and hind-wing cells and of the wing shape. The length of the male hind wing cell is only half the length of the hind wing in Catoblepia but three-fifths in Opsiphanes, the reverse being true for the forewings. Also the forewings of Opsiphalles are larger than the hind wings and present a top-heavy aspect whereas in Catoblepia the fore and hind-wings are of almost equal size. Bristow (1981) concludes that these characters can be used to define an apparently homogeneous natural group. If the cephalic projections as seen on the pupa of Catoblepia b. berecynthia (Cramer) prove constant for other species of Catoblepia, (in particular the group Dyophthalmi with only two ocelli, like Opsiphanes, on the hindwing underside), it may prove to be an additional feature to separate Ca toblepia from Opsiphanes.

Fig. 1. (Far left) Larva of C. berec)'nthia Fig. 2. (Left) Pupa of Catoblepia berecYllthia

31 The pupae of the following Opsiphanes species have been It would be relatively simple for any student in the field to described or figured: O. cassin a fabricii (Boisduval) (Bates 1932, bait and capture females of the more common forms of Cato­ Young and Muyshondt 1975); o. cassiae cassiae (Linnaeus) blepia (of particular interest would be one of the Dyophthalmi) (Dewitz 1878); o. cassiae (as PaJ'onia gl)'ceria) (Jones 1882); and using the method described here obtain ova, larvae and pupae o. cassiae lucullus Fruhstorfer; O. b. bogotanus Distant and and so determine whether the cephalic projections on the pupa O. invirae remoliatus Fruhstorfer (Rothschild 1916); o. t. are constant for the genus Catoblepia. tamarindi Felder & Felder (Harrison 1963, Young & Muyshondt Barcant (1970) mentions the delicacy of the wing 1975). Additionally the authors have examined pupae of O. membranes of Catoblepia, a fact to which collectors of Catoble­ cassin a merianae Stichel and 0. c. cassiae from Trinidad. None has pia berecynthia can testify. It is of interest that, at least in the pupal cephalic projecti.ons of Catoblepia b. berecynthia. Trinidad, the only other brassolid that also has very delicate Bristow (1981, p .147) includes a previously unpublished wings, Eryphanis p. polyxena, also has very large cephalic pro­ painting by the Rev. A. Miles Moss which shows the larva of C. jections in the pupal stage. Prominent caphalic pupal horns are berecynthia to have three pairs of cephalic horns; the larva of also present in at least one species, cassiopc, of Selenophanes Catoblepia b. berecynthia from Trinidad also possesses three (Bristow 1982, Smart 1976). pairs of horns, as against four pairs on the two Trinidad species of Opsiphanes, O. cassina merianae and O. c. cassiae, on O. ACKNOWLEDGMENTS I cassiae from Venezuela (Dewitz 1878), O. t. tamarindi from Costa Rica (Harrison 1963, Young & Muyshondt 1975) and Thanks to Ken Preston-Mafham for the original photography, O. cassiae from Brazil (Jones 1882). It is thought that this to J. Evans for the black and white reproductions and to C.R. might be another useful feature in separating the genera, but on Bristow for bibliographic and taxonomic details and for a critical the illustrations of the four species of Opsiphanes larvae figured review of this paper. by Rothschild (1916), the number of horns varies from one pair to three pairs. These illustrations appear rather crudely done and may not be accurate, but Young (1977) includes a sketch of a head capsule of fifth instar O. quiteria quirinus which also has REFERENCES only three pairs of cephalic horns.

BARCANT, M. (1970) Butterflies of Trinidad & Tobago. London: Adult Catoblepia b. berecynthia. Female above, male below. Collins. BATES, M. (1932) Notes on the metamorphosis of the Brassolidae (Lepidoptera) Bull. Brooklyn Ent. Soc. 27: 155-163. BRISTOW, C.R. (1981). A revision of the brassoline genus Cato­ blepia (Lepidoptera: Rhopalocera). Zool. J. Linn. Soc. 72: 117-163. BRISTOW, C.R. (1982). A revision of the brassoline genus Selenophanes (Lepidoptera: Rhopalocera). Zool. J. Linn. Soc.,76: 273 -291. DEWITZ, H. (1878). Entwickelung einiger Venezuelanischer Schmetterlinge nach Beobachtungen von Gollmer. Archiv. Naturgesch.44: 1-36. HARRISON, J. O. (1963). On the biology of three banana pests in Costa Rica (Lepidoptera: Limacodidae, ). Ann. Ent. Soc. Am., 56: 87-94. JONES,E. O. (1881). Metamorphoses of lepidoptera from Santo Paulo, Brazil, in the Free Public Museum, Liverpool. Proc. Lit. Phil. Soc. Liverpool, 36 326-377. ROTHSCHILD, w. (1916). Notes on Amathusiidae. BrassoJidllP' Morphidae, etc. with descriptions of new forms. Novit. Zool. 23: 299-318. SMART, P. (1976). The Illustrated Encyclopaedia of the Butterfly World. Hamlyn, London, 275 pp. STICHEL, H. (1902). Aufteilung der Gattung Opsiphanes Westw., Beschreibung neuer Brassoliden und synonymische Notizen. Berl. Ent. Z. 46: 487-524. YOUNG, R.M. (1977). Notes on the defoliation of Coconut Palm (Cocos nucifera) by the butterfly Opsiphanes quiteria quirinus in northeastern Costa Rica. Dtsch. Ent. Z. 24: 353- 365. YOUNG, A.M. & Muyshondt, A. (I 975). Studies on the natural history of Central American Bu tterflies in the family cluster Satyridae Brassolidae - Morphidae (Lepidoptera: Nym­ phaloidea). III. Opsiphanes tamarindi and Opsiphanes cassina in Costa Rica and El Salvador. Studies Neotrop. Fauna, 10: 19-56.

32 The Skipper Butterflies (Hesperiidae) of Trinidad. Part 4: Pyrginae (second section)

B;v M.J. W. CocA, Commonwealth Institute of Biological Control, Imperial College at Silwood Park, Sunninghill, Nr. Ascot, Berks SLS 7PY, U.K.

IN earlier parts of this series I have dealt with seven of the eight 20. Proteides mercurius mercurius Fabricius 1787 species of Pyrrhopyginae (Cock 1981c, but see also Cock 1984a) Plate 8 and the 12 species of Pyrginae belonging to Evans's (1952) genera group B (Cock 1984b). Here I cover the first 35 species of Evans's This skipper occurs in the literature as P. idas Cramer, and genera group C, leaving the remaining 16 or so species (Astraptcs, was added to the Trinidad list under this name by Barcant (1970, Calliades, Autoc/zt()J7, Cabares) for a future contribution. The list no. 604). The nominate occurs from Mexico to species covered here conveniently include all the tailed skippers (Evans 1952) and six subspecies have been described from Trinidad, and a key is provided for these at the end of this from the Caribbean. Of the two specimens from Tobago in the part. BMNH, one, a male bequeathed by W.H. Sheldon, is P. m. /111'rcu­ The recognition of the subfamily Pyrginae, its division into rius, and the other a female collected in 1914 by Dr. F.W. sections I and 2, and the division of section 1 into genera groups Jackson, belongs to the subspecies P. m. angasi Godman & Salvin. B, C and D was covered in Cock (1984b). It is sufficient here to This subspecies is only otherwise known from Dominica and St repeat that group C is recognised by the third segments of the Lucia, so the record is unexpected. Furthermore I recently had palpi being parrellel in common with group D (Fig 1) and that the the opportunity to examine some of the Hesperiidae of the forewing veins 6 and 7 are approximate at their origin (Figs 2,3) University Museum, Oxford, which includes part of Dr. Jackson's whereas in group D veins 7 and 8 are approximate. Evans (1952) collection; amongst this material is a series of specimens labelled also states that for group C the wings are held erect in repose, "xii.1913 to end iv 1914, Dry S., TRINIDAD, Tobago or but this is not the case in the genus I'olytlzrix in the Trinidad Dominica, F W Jacksun" which included two P. mercurius angasi, fauna which hold their wings flat. Most species have a costal fold two zestos Geyer (see next species below) and a selec­ in the male, but this is not consistant in all genera and the pre­ tion of other species and subspecies which in combination could sence or absenc~ of this character is indicated in the species have come from only St Lucia or Dominica. In view of their accounts below. labels I conclUde that all came from Dominica. Because of this Most, but not all, of the larval foodplants of this group group of specimens I also conclude that Jackson's specimens of belonging to the family Leguminosae. All the young larvae of this P. m. angasi and H. Zcstos labelled Tobago in the BMNH also group with which I am familiar use a similar type of leaf shelter came from Dominica and Cock (1982) should be amended (Fig. 4), described under Epargy reus and su bsequen tly referred to accordingly. under that name. Ground colour dark brown; white hyaline spots F. Basal In my systematic list of Trinidad Hesperiidae (Cock 1982), I quarter UPF and UPH, thorax and head above orange brown; rec0rded 273 species [rom the island. Since then I have added six UNS head and anterior femora pale brown, almost white. UNH species and deleted one (Cock 1984a). In this part I add three mottled with white - strongest on disc. Fringe pale except dark specks to the Trinidad list (h'parg)'reus nUIl'a, pronla at the ends of the veins. The mottled hind wing underside and Urbanus 1'sma) and delete one (Cuda/ractus melull) bringing distinguishes P. mercuriu.\· from the superficially similar large the total for the island to 280, which will doubtless be added to females of the genus Hpargyreus considered below. P. 111. angasi in the future. differs from P. m. mercurius' principally in that the under­ Several new abbreviations are introduced in this part as surface markings are dull reddish brown where tbey arc white follows: UPF I UPH = uppersurface of forewing I hind wing UNF I UNH = undersurface of forewing I hindwing UPS I UNS = uppersurface / undersurface F I H = forewing I hind wing FIGURE LEGENDS

I would like to thank Drs. R.1. Vane-Wright and P.M. Ackery Fig I Dorsal view, head of catil/us for their hclp and co-operation at the British Museum (Natural Fig 2 Venation, Urhanus pronta d' History) (referred to as BMNH is this work), and Dr. G. C. Fig 3 F venation l'ro/eiJes mercurius cI McGavin who arranged the loan of material from the University Fig4 Schematic drawing of larval shelters of Epllrgyreus Museum, Oxford. type.

33 ------

L ..J Fig. 1 /~ ------;

'----...--J-J

Fig. 3

34 in the nominate subspecies. Illustrations in Barcant (1970, Fig 4, It is a dark brown species with yellow hyaline spots; the No.6, as Epargyreus exadeus), Riley (1975, Plate 21, No.3), hindwing underside has a weak brown bar and no white markings Lewis (1973, Plate 85, No. 35) etc. No costal fold; F 5f 36 mm. which distinguishes it from other members of the genus from This large species is always scarce in Trinidad, but probably Trinidad. Illustration in Riley (1975, Plate 21, No.5). The life occurs throughout the island. I have caught specimens at St. history is unknown. Augustine (at flowering legume tree, x. 1979) and Morne Bleu (at light, vii.1978). The life history is unknown in Trinidad, but Moss (1949) describes the larva in Brazil as very dark olive green 21. Epargyreus'socus chota Evans 1952 with a vermilion lateral stripe, a dark dorsal line, and a large dark Plate 1-4 brown head with red eye spots. Moss records finding the large larvae on Muellera moniliformis L. fils (Leguminosae) in shelters This species is variable in both appearance and genitalia of cut leaves, which rapidly dry up and discolour. Riley (1975) through its range (Mexico to Argentina) and the status of some of also records Derris eUptiea, Cassia spp. and other legume trees as the six SUbspecies recognised by Evans (1952) is likely to change. hostplants. M. moniUformis and the cassia trees are likely hosts r suspect that the series from Trinidad which I am treating under for this species in Trinidad. this name may contain more than one species. The type locality for the subspecies is Trinidad and its range extends only to Eastern Venezuela (Evans 1952). Epargyreus Ground colour dark brown with an overlay of dark rufous hairs at the base UPF, over much of UPH, and on the body; This genus is probably the most difficult taxonomically of hyaline spots tinged yellow, particularly those of the cell, space the family in Trinidad. The genitalia are similar and in at l~ast 12 and costa. UNS has silvery white markings on UNH, and a one species variable, while the wing markings are very similar. dusting of white scales centrally on the margin of both wings; Here I follow Evans (1952) and recognise four distinct species UNS labial palps and head pale brown. Evans (1952, p. 48) gives from Trinidad although I may have two or three more repre­ one of the diagnostic features of the subspecies as " .... spot in sented by single female specimens. Until the necessary breeding­ space 2 completely overlaps the cell spot"; in the serie:; before me experiments are done, this genus is unlikely to be satisfactorily it would be more accurate to say that the cell spot is at least half resolved and the liable to be changed. I hope to look overlapped by the spot in space 2. The illustration of this species more closely at this genus in the future and would welcome the in Lewis (1973, Plate 82, No. 48) shows a male of a subspecies opportunity to examine more specimens from Trinidad. other than chota, but does indicate the colouring of the species. The larva of E. exadeus and what I believe to be the larva of Costal fold; F d'26-28mm;~ 30 mm. E. soeus are described below under each species, but I have also This is the commonest species of the genus in Trinidad, and found a larva which I consider to belong to the genus Epargyreus can be obtained occasionally at flowers and the males on hill­ on a Dioclea sp. (not guianensis) at Waller Field (xi.1981). It was tops. I have seen specimens from Chacachacare Island (Cock 1.5 cm long and was hidden in a shelter similar to that described 1984a), Scotland Bay, Cumberland Hill, Mt. Tabor, Morne for the young larvae of E. soeus below. The head is black, shiny Bleu, Brigand Hill, Moreau and Parry lands, from all times of and slightly , with a pair of small red eye-spots; pro­ the year. thorax black above and red below; mesothorax, metathorax and I am not aware that the biology of this species has been abdomen dark translucent shiny green with yellow markings - recorded. I have observed and caught a female ovipositing on a thin dorso-lateral longitudinal stripe with narrow transverse Dioclea guianensis Benth. (iii. 1980, Parry1ands) and r have found lines above and a network of thin lines below, and a clear area larvae of an Epargyreus sp. on this vine which I believe were of just below the dorso-lateral line at the anterior of abdomenal this species but was unable to rear them. Three larvae were found segments 1-8; abdomenal segment 9 tinged orange; prolegs con­ on D. guianensis on the ridge to the North of the transmitting colorous with body; thoracic legs pinkish red. station on Cumberland Hill (x.1980). The first larval shelter, None of the Epargyreus spp. larvae which I have tried to rear found adjacent to the egg, is formed by cutting a flap from the really settled down in captivity and most died before pupation. I edge of the leaf (Fig 4); the basal cut is across one of the 1eaf­ would encourage collectors to try and relate larvae of this genus veins resulting in the distal end of the flap being flattened on the to their foodplants and to the adults and hopefully this will leaf, and the basal end being raised to an angular peak. The throw light on the relationships of the four or more very similar mature larva shelters between two leaves spun together and has a Epargyreus spp. from Trinidad considered below. reddish brown head, slightly chordate with red eye-spots; prothorax concolorous with head; mesothorax, metathorax and abdomen dark translucent green with a network of yellow lines [Epargyreus zestos Geyer 1832] giving a yellow green overall colour; spiracles inconspicuous, Plate 7 brown; a row of small, conspicuous, lateral, black spots above the spiracles on mesothorax, metathorax and abdomena1 segments This species has never been recorded from Trinidad, but 1-7; pro legs pale orange; thoracic legs red. The small larva lacks there is a specimen in the BMNH from Tobago (coll. Dr. F.W. the red eye-spots on the head, which is darker in colour. Jackson in 1914). The species has a curious distribution, being found from Florida through the Bahamas and Lesser Antilles to Barbados and Tobago (Evans 1952); five specimens from 22. Epargyreus exadeus exadeus Gramer 1779 Honduras indicate it also occurs in Central America. In light of Plate 5-6 the material in the University Museum, Oxford, discussed under Pro teides mereurius above, I conclude that Jackson's specimen The nominate subspecies occurs throughout tropical South is mis-labelled and this species does not occur in Trinidad or America but is scarce outside Colombia and Peru; a second sub­ Tobago. species E. e. cruza Evans occurs in Central America.

3S The colouring is similar to that of E. socus chota, but the larva did not pupate until 25 days after it was collected, but did hyaline spots are nearer white in colour than the yellow tinted not grow much during this time and never really settled down in ones of E. socus. The underside markings are variable and cannot the rearing container; the adult emerged after 15 days. Moss be used to separate the two which can howe..ver be distinguished (1949) describes and illustrates the larva from Brazil which by the fact that the spot in space 2 broadly overlaps that of the he found on a "wild bean (Phaseolus sp.)" and a small forest cell. Costal fold, F J 27 mm, ~ 29 mm. tree doubtfully identified as Lecythis paraensis Hub. He describes I have captured just two specimens of this species: both at the larva as " .... a brilliant greenish yellow with a row of small flowers of Austroeupatorium inulifolium (L.) King & Robinson black spots supporting the median line, and another nine oval along the Textel Morne Blue Road at c. 2100 ft in the Northern spots, much larger, marking the sides. The second segment (i.e. Range (x. 1979). Three males in the BMNH collection were prothorax) and the claspers are blood red. The head is large and caught in "St. Ann's" (i-iii. 1932, A. Hall), "St. Georges" (x. bifurcated, dark brown and glossy and with big and prominent I 1891, C.W. Ellacombe) and in the "Northern Mountains" (xii. red eye-spots." Evans has confirmed Moss's identification on 1938 - i. 1939, A. Hall). I have reared one female of what I specimens in the BMNH, and these two descriptions are basically believe to be this species (xii.1981, Curepe), but the speci­ compatible. men is small and the spot in space 2 does not overlap the cell spot, yet the wing proportions and colouring match those of this species. The hostplant was a Caesalpiniaceae tree, perhaps a 22a. Epargyreus nutra Evans 1952 Cassia sp" with multipinnate leaves whose leaflets measured Plate 9-10 about 3 cm. The larva had spun a pair of terminal leaflets together and its weight caused them to droop. The mature This species is described by Evans (1952) from Colombia larva measures 3.5 cm; the head is black, rugose and slightly and Venezuela, and this is the first record from Trinidad. ch'ordate in shape, covered in fine, short, white hairs and with This species is similar in colour and markings to the last two, a pair of red eye spots; prothorax black above and red below, but the F hyaline spots are nearer yellow than white in colour; with a black lateral spot; mesothorax, metathorax and abdomen the spot in space 2 UPF does not overlap the cell spot and is dark translucent green with yellow markings dorso-lateral and nearly triangular in shape. Evans (1952) describes the UNS of the

Plates 1-12. 1, Epargyreus socus chota cluPS (Brigand Hill, 17. iv. x. 1979, MJWC); 7, E. zestos :f UNS (Tobago, FW Jackson, 1982, MJWC); 2, E. socus chota dUNS (Brigand Hill, 2l.ix. specimen in BMNH); 8, Prateides mercurius mercurius 2. UNS 1979, MJWC); 3, E. socus chota!f- UPS (Chacachacare Island, 7.i. (Morne Bleu, at light, 26. vii. 1978, MJWC); 9, Epargyreus nutra 1982, MJWC); 4, E. socus chota Si: UNS (Parrylands, ovipositing cluNS (La1aja Ridge, in cop.,28. v. 1979, MJWC); 10, E. nutra on Dioclea guianensis, 26.iii. 1980, MJWC); 5, E. exadeus exadeus .

36 Life history and foodplant unknown. more common. There is one other subspecies, punctus Bell & Comstock, from the Lesser Antilles. Kaye (1940) and hence Barcant (1970) refer to this species as Acolastus amYlltas 23. Epargyreus clal'icornis tellda Evans 1955 Fabricius, a of P. leo. Plate 11-12 Ground colour dark brown; white hyaline spots; strong l.llue flush UPF base. UNH ground colour restricted to two discal and Although Evans (1952) treated this species as E. c. orizaba one marginal bands; tornus very dark brown; space 1 A pale Scudder, in the corrections (Evans 1955, p. 476) the type of brown; space 1 B brassy; space 1 C and the margin yellow brown; orizaba is reported to be one of the subspecies of E. socus, and the rest of UNH purplish mauve; the black spot at the base of the replacement name tenda put forward for this subspecies. space 7 much reduced compared to 1'. leo. UNF dark brown E. claJ'ieornis is divided into two principal subspecies: tcnda on disc; blue flush at end cell; margin and apical quarter with from Central America and northern South America and clavi­ purple flush; marginal and submarginal ochreous bands. The cornis from the rest of South America to . Cock (1982) UNH colouring and the size of the black spot at the base of recorded this species as new to Trinidad. space 7 should serve to distinguish the two species of Poly­ The colouring is similar to that of the last three, but the gOllus. Riley (1975, Plate 21, No.9) illustrates the subspecies ground colour is a slightly darker brown. The reduced silver punctus from Dominica. No costal fold; F cf22-23mm;~ 23mm. markings on the underside of the hind wing, particularly in the Of the 19 specimens of this species from Trinidaa in the male, are distinctive. The female is very similar to that of E. BMNH, eight are from St. Anns, four from Arima, one from the mitra, and perhaps not reliably identifiablc from illustrations. Northern Range and the remainder do not specify a locality. Costal fold; F 032 mm;~ 35 mm. The three specimens which I have caught are also from the A scarce species in Trinidad; Cock (1982) lists six captures, Northern Range (Morne Catharine, St. Anns Peak, Mt. Tabor), half of which are from Parrylands and there are two more female$ so there are no records from Central or South Trinidad for this in the University Museum, Oxford, collected by R.W. species. It can be found on shady paths, similar to those used by Farmborough (Palo Seco-Siparia Road, i. 1917; Fyzabad, i. Phanus marshallii Kirby (Cock 1984b), but is considerably less 1918). I observed and caught a female ovipositing on Dioclea common. It rests under leaves with its wings closed. The life guianensis, but no larvae were reared. Beyond this observation, history in Trinidad is unknown, but in Brazil Moss (1949) re­ the life history is unknown. cords the food plants as Mudlera moniliformis (cf. Prate ides mercurius). He describes the full grown larva as " .... half green, half yellow, the transparency of the skin giving it a patchy look. 24. Polygollus leo leo Gme1in 1790 The dorsal area is enclosed by light stripes. The head is rather Plate 13 flat, light red and glossy, darker behind and with black eye­ spots in the upper portion. " This is a widespread species occurring from Florida to Argentina, but generally not common. Evans (1952) recognises three further subspecies from the Caribbean. 26. Chioides catillus catillus Cramer 1779 Ground colour brown; white hyaline spots. UNH brown, Plate 22 basal area and two discal bands pale mauve; a black spot at the base of space 7. The body is brown and the head below pale This widespread and common species is recorded from the brown. Illustrations in Lewis (1973, Plate 85, No. 21; as P. lividus southern USA to Argentina. It is variable over its range, and Hubner), Riley (1975, Plate 21, No.8; subsp.lividus from Haiti) Evans (1952) recognises eight subspecies. The nominate sub­ and Brown & Heineman (1972, Plate IX, No.4; subsp. hagar from species occurs throughout South America (to Argentina) except Jamaica). No costal fold; F rJ 20 mm; Evans (1952, p. 53) gives and the north of Peru. Although Chioides zilpa zilpa 25mm; Trinidad specimens may be smaller than normal. Butler occurs in Venezuela, I believe Kaye's record of this species The single specimen of this species from Trinidad in the from Trinidad (Kaye 1921, No. 454) to be in error. BMNH is from St. Anns (i-iii. 1932, A. Hall), while the only two It is based on a capture at Morne Diable by Sir. N. Lamont, specimens I have seen were caught along the lower part of the but there are no specimens in Lamont's collection housed track to Morne Catharine (v.1982). Apparently scarce in the at the University of the West Indies, and I have seen no island, it may prove to be restricted to the north-west. Wolcott specimens from Trinidad. (1951) records the foodplants of the subspecies P. leo lividus in Ground colour brown; yellow hyaline spots; UNH mottled Puerto Rico as Lonchocarpus sp., Derris eliptiea and Piscidia brown; UNF with dark brown triangle on costa at apex adjoining piscipula (all Leguminosae) and describes the larva as having a the apical spots, which is distinctive for the genus and will serve flat chordate head, yellow with a large dark spot on each side of to identify C. catillus from the other tailed skippers in Trinidad. the dorsal cleft. The head of the earlier instars Wolcott records The female has larger hyaline spots than the male, but other­ as black. Riley (1975) suggests the body is green with yellow wise the sexes are similar. Illustrations in Barcant (I970, p. 86, markings. Other recorded foodplants include Piseidia piscipula No. 3;~, Riley (1975, Plate 21, No. 10; c!subsp. churchi Bell and Jamaica Dogwood, P. er)'thrilla, in Florida, and LOlleho­ & Comstock from Jamaica), Lewis (1972, Plate 81, No. 38) carps sp. in Cuba (Brown & Heineman 1972). Williams (1931) and Brown & Heineman (1972, Plate IX.; No.8; cf subsp. churchi considers three species of LOllchocarpus to be indigenous in from Jamaica). Weak costal fold; tail cf 15-20mm;~ 21-22mm; Trinidad, but none of Piscidia (= lchthyomethia). F cf26-27mm;!f. 27 mm. This species occurs throughout Trinidad and Tobago at all 25. Polygollus manueli manueli Bell & Comstock 1948 times of the year. Usual habitats include gardens, rough ground Plate 14-15 and roadsides, and the adults feed freely at flowers such as Eupatorium (s.1.) spp. and Bidens pi/osa L. . Like most of the This species has J similar distribution to the last, but is rather group its resting position is with the wings closed. The food-

37 plant for the larva in Trinidad is kudzoo, Pucraria phase%ides I have specimens from San Miguel Valley and Parrylands and (Roxb.) Benth., and perhaps other legume vines. The eggs are there are specimens from St Anns in the BMNH. This species laid singly in the middle of the leaf lamina on the uppersurface or appears to be associated with at least semi-natural forest; one undersurface. The young larva hides in leaf shelters similar to specimen was taken at flowers of Chromolaena odorata. I those described for Eparg)'reus spp. The mature larva has a consider it an uncommon species, although the 11 specimens in slightly indented vertex to its large, rounded, dark brown head; the BMNH suggest it may have been more common in the past. on the face a pair of large black eye-spots with an orange patch Moss (1949) records Tabebuia pyramidata DC (Bignoniaceae) as below each. Pro thorax brown with a shiny black, narrow dorsal the foodplant of A. aurunce. It is possible that this record refers plate. Mesothorax, meta thorax and abdomen dark green speckled to A. caelus below, as Moss's series in the BMNH was mixed with yellow dorsally and white laterally; a darkened dorsal line (W.H. Evans footbote in Moss (1949) p. 42). The larva in Brazil and a narrow orange dorso-Iateral one, widened at abdomenal is described by Moss as having a red chordate head with black segment 8. Legs of prothorax black, remaining thoracic legs and eye-spots; above the level of the spiracles yellow with alternate prolegs orange brown. Small larvae have black heads and medium thin and thick black bands (one of each segment); below the sized ones develop orange eye-spots. A first instar collected 17 spiracles, legs and body extremities red; pupa bone coloured with February 1982 emerged as an adult on 9 April. Riley (1975) la teral streaks. describes the larva of this species from Mexico as having the head with the vertex scarlet, merging into red below, with a Y shaped 29. AgUila coelus Stoll 1781 black, frontal mark and feeding on Tephrosia sp. Plate 19-20

As for the last, this species is monotypic and can be found 27. Aguna asander asander Hewitson 1867 from Guatemala to the Amazons. The colouring is also similar Plate 18 to that of A. aurunce. Costal fold; UNH white band width! 0.75 mm,~ 1.75 mm; F 620-21 mm'.!f.22-23 mm. Apart from subspecies restricted to Jamaica and Haiti, this This species is commoner in the south of Trinidad than in species is constant throughout its range from Mexico to the north. I have records of four specimens from the north Argentina. (Arima Valley, Guanapo Valley, Blanchisseuse - Paria Bay track) The uppersurface resembles that of an Hparg)'reus sp.: and 11 from the south (Moreau, Rock, Siparia, Parry lands, ground colour brown; hyaline spots yellow. UNS brown with a Trinity Hills). This may explain why A. coelus is less well repre­ white band running from mid costa UNH to near termen, clearly sented in the BMNH (3 specimens) than A. aurullce (11 speci­ defined on basal side, but indistinct on distal side. Body brown mens) as relatively few of the BMNH collection from Trinidad are except anterior femora and head below which are near white. A from the south. My captures are principally from the dry season distinct species owing to the broad white band UNH. Illustrations (one in September, one in October, nine in December to in Lewis (1972, Plate 80, No. 13; UNS), Riley (1975, Plate 21, February, and one in April). Five of my captures were at flowers No. 13; subsp. haitiensis Mabille & Boullet) and Brown & l of Chromolaella odorata. I would regard this as a moderately Heineman (1972, Plate IX, No.5; subsp. jasper Evans). Costal common species. Life history and foodplant unknown. fold; F l26 mm. /d~hough Kaye (1921) reports this species as being "fairly 30. Typ/zedallus undulatus Hewitson 1867 COlll11l on", I have seen only three specimens from the island: Plate 23 two collected by Sir N. Lamont at Palmiste and one I caught at flowers of Chronwlaena odorata (L.) King & Robinson This species has a wide range, from Mexico to Argentina, (Compositae) in Parry lands (i. 1981). The life history is un­ with no subspecies recognised. Kaye (1921) records it from known, but the foodplant is thought to be a species of Trinidad on the basis of a record by Crowfoot (1893). Sheldon Leguminosae. (1938) records the capture of a specimen at Scarborough, Tobago, but there are no specimens from Tobago in the BMNH (to which most or all Sheldon's Tobago material was be­ 28. Aguna aurunce Hewitson 1867 queathed). I can confirm this to be a Tobago species, having cap­ Plate 16-17 tured a specimen at Crown Point (ix.1982). A small undistinguished tailed skipper, ground colour brown; Aguna aurullce ranges from Guatemala to the Amazons as white hyaline spots. UNF brown with pale brown patch over a monotypic species. Kaye (1921, 1940) does not include this end cell and margin; UNH with dark markings and distinctive species, but he does include the similar species A. COe/liS (below) striations on the unmarked areas. The markings, particularly the thus: "Two specimens in the National Collection taken by W.E. striations, of the UNH and the alignment of the spots of the Broadway. A single specimen from St Ann's ValJey (W.J. Kaye.) cell and space 2 to form a bar are distinctive for this species. ChancelJor Road not common Nov. 1920 (W.J .K)" (Kaye 1921, Illustration in Lewis (1973, Phite 87, No.4 7). No costal fold; p. 123). Broadway's specimens in the BMNH are A. aunlllC<', so tail Jl2 mm; F Sf. 20 mm. clearly the two species were mixed together by Kaye. Apparently a scarce species in Trinidad, T. undulatus is pro­ Ground colour brown; hyaline spots yellowish; UPF base, bably easily cvcrloo!,ed. It is attracted to flowers (e.g. UPH basal half to termen, UPS thorax and head with overlay Chromolaella odorata), but can probably be obtained most easily of metallic green hairs (scales on body). UNH with white band around the foodplant or by rearing the larva. I have seen speci­ from 2/3 on costa to vein I B (1 mm in J, 2 mm in .2); an in­ mens from St Augustine (larvae colI. xi.I981), Piarco (i.1982) conspicuous white streak in space IA; basal 1/4 l\NH and and Siparia (xii. 1981; J. & F. Preston). The resting position is extreme base UNF flushed green. Thorax. abdomen below pale with the wings closed. The larval foodplant of T. ulldulatus in hrown; head below white. Costal fold; tail J3 mm:-¥ 2.5 mm; F J Trinidad is Cassia (Se/lna) obtliSifo/ius L. (= C. tora auct. nec L.) 24mm;¥ 24 mm. (Caesalpiniaceae) (Cock & Evans 1984), although it may also use

38 C. alata L., C. reticulata or C. occidentalis L. which Moss (1949) weakly differentiated dark band, cell spot and dark tails. The records as less preferred foodplants. The mature larva forms a eight spots referred to in the name assumeably include those of shelter by spinning together the terminalleaftlets which, because space 12 and the costa which are joined together. It can be of the larva's weight, hang down and are quite conspicuous. The recognised by the arrangement of the UPF spots. Illustration head is black, chordate, rugose and shiny; thorax and abdomen in Riley (1975, Plate 21, No. 16). Costal fold; tail t 7 mm, dull, translucent green; spiracles yellow, surrounded by yellow­ ~ 9 mm; F c!20 mm, .5f 23 mm. orange joining up to form a lateral stripe; legs concolorous. This is not a common species in Trinidad or Tobago. There The pupa is stubby; pale brown with dark speckles and prominant are two males from Tobago in the BMNH (one from Rox­ black spiracles. Moss (1949) describes the larva from Brazil as borough), but none from Trinidad. I have seen only three female having a dark brown head with yellow eye-spots. specimens from Trinidad: two taken at flowers of black sage,

Plates 13-23. 13, leo leo ~ UNS (St Anns, Liii.l932, A UNS (Parrylands, 2.ii.1980, MJWC); 20, A. caelus ~ UNS Hall, specimen in BMNH); 14, P. manueli manueli c! UPS (Mt. (Blanchisseuse-Paria Bay track, 22.i.1980, MJWC); 21, Tabor, 25.1.1981, MJWC); 15,P. manueli manueli5f- UNS (Morne Chrysoplectrum perniciosus perniciosus d'ups (Morne Catharine, Catharine, 18.ix.1982, MJWC); 16, Aguna aurunce cf UNS(St. 28.i.1980, MJWC); 22, Chioides catillus catillus if UNS (Morne Georges, x. 1981, CW Ellacombe, specimen in BMNH); 17, A. Catharine, 21.v.1982, MJWC); 23, Typhedanus undulatus

31. Po lythrix octomaculata octomaculata Sepp 1848 bone-coloured"with a broken side-line of reddish streaks and there Plate 24 are angular projections near the head". Steinhauser (1975) records this species feeding on Myroxylon balsamum v. pereirae (Royle) Baill. (as Toluifera perreirae) (Leguminosae) in El This and the following four species belong to the quite Salvador. distinctive genus Po lythrix. They differ from the other tailed skippers in their larvae and by the fact that the wings are held flat at rest, enabling the genus to be recognised immediately in 32. Polythrix roma Evans 1952 the field. P. octomaculata is a widespread species ranging from Plate 25 Mexico to Argentina the subspecies, decussata Ivlenetries, is res­ tricted to Haiti. This species is quite restricted in its distribution, being found Ground colour brown, slightly tawny; hyaline spots yellow only from the Amazon to Venezuela. It was described from Para tinted; UPF two indistinct dark spots in space IB; UNH with and the holotype is a specimen collected by A. Miles Moss whose Cordia curassavica (Jacq.) R. & S. (Boraginaceae) in Nariva rearing records are unique for South American Hesperiidae. Swamp (Cock 1982a) and one from Piarco (at light, vii.I979). Earlier works, e.g. Kaye (1921, No. 455), Moss (1949) refer to Moss (1949) records the larvae of this species feeding on several this species as P. asine Hewitson, a Central American species. Leguminosae including Muellera mon ili/orm is: 'the larvae have Ground colour brown, paler on disc UPH, dorsum UNF and extremely flat heart-shaped heads. They are nearly white and de­ UNH; hYaline spots white. disc 1 UPF with a dark spot to­ corated with numerous tiny longitudinal streaks of grey, the wards the base and two spots (the upper slightly hyaline, the dOlsal area being scarcely divided from the remainder by a faint lower dark) in space 1 B under the spot in space 2; the inner line. The heads are Naples yellow with looplike pattern in pale margins of the spots in spaces 2 and 3 point the costa between mauve or light red followed by a dark line. The pupae are light the costal and apical spots, rather than at the apical spots as

39 stated by Evans (1952, p. 68). Also Evans states that there are usually has four apical spots, while P. auginus has three, but this three apical spots but in the two female specimens before me is an unreliable character as I have a 5f P. caunus before me with there is an arc of 6 spots in spaces 4 to 9. The BMNH series shows three apical spots. The best characters for distinguishing the two this to be variable, ranging from the condition in some Amazon species are the white bar in space 2 of P. auginus which is reduced specimens with dark spots in spaces 4 and 5 and none in space 9 to a spot on vein lB for P. caunus and the uniform disc UNH of to the extreme shown by the Trinidad females. UPH and UNH P. auginus, as opposed to the end of cell spot and macular sub­ with two macular dark bands, the outer edged with pale brown in marginal band of P. caunus. Illustration in Lewis (1973, Plate 85, spaces 1 C and 2 UNH; tails dark. Polythrix roma can be identified No. 24; UPS). Costal fold; hair tuft vein 1 B UNH in d; tail J in Trinidad by the arrangement of the F spots. Costal fold; tail d l3mm; F d19mm,~ 20 mm. I 7 mm,~ 12 mm; F d 18 mm,~ 20-2lmm. This species is scarcer than the last, and the few records of This is another scarce species in Trinidad. There are two which I am aware are from the Northern Range. In view of my males in the BMNH and I have caught two females (viii. 1 980; ix. capture of a male on the summit of El Tucuche(vi.l979) it is 1982) and been given a third (i.l979, W. de Voogd) from Curepe. possible that they defend hilltop territories; I have caught only There is a fourth in Sir Norman Lamont's collection from one female and that was in secondary forest at about 1,000 ft Palmiste (iii. 1930), while Kaye ( 1921) records two more taken on M t. Tabor (i.l982). As with the last, I have not seen this at Palmiste by Lamont. Moss (1949) states that this species feeds species at flowers and the life history and foodplant are un­ on the same foodplants asP. octomaculata, and the larva can only known. be distinguished from that of P. octomaculata by a pair of eye-spots on the head.

35. Polythrix metallescens Mabille 1888 33. Polythrix caunus Herrich-Schaffer 1869 Plate 28 Plate 26

There are no recognised subspecies of P. caunus which is This species is found from Honduras to the Amazons, and found from Mexico to Paraguay but is not common in collec­ is scarce in collections. Above, it resembles P. caul1us and P. tions. Kaye does not record this species from Trinidad (Kaye auginus (above) in colouring, although the wing shape is dif­ 1921, 1940), but doubtless he confused it with the next, P. ferent; labial palps below green, otherwise body as P. caul1us; auginus, as A. Hall took both in St. Ann's Valley (i-iii. 1932) UNH plain brown with a conspicuous white band from vein 1 B and Kaye records these specimens as P. auginus. to 2/3 on vein 6. Costal fold; conspicuous brown hair tuft from Ground colour dark brown; hyaline spots white; basal 1/3 near base vein lB UNF in 6; tail ¥. 11 mm; F ~ 22 mm. UPF and basal 1/2 to dorsum UPH with paler brown hairs with a This is a rare species in Trinidad. Kaye (1940) lists a capture slight metallic green tint; similar green tint UNS even less marked; in Santa Cruz Valley (iii.1929, Huntingdon); there is a male in head above green, below white; thorax and abdomen above the BMNH labelled "Trinidad / A.G.B. Russell" and I have cap­ ground colour overlaid with metallic green hairs; thorax below tured a female in Parry lands (ix.198l). Life history described by with mixed green and brown hairs; abdomen below pale brown. Moss (1949) as similar to that of P. octomacu latus, utilizing the UNH with dark iridistinct end of cell spot and macular sub­ same foodplants and in addition Lonchocarpus sp. and marginal band; basal 1/3 of spaces 1Band 1 C bare (and hence Machaerium floribundum Benth. pale) associated with hair tuft on vein 1 B in d; white spot at 2/3 on vein lB; tails dark, almost black at tip. Close to the next species under which differences are considered. Costal fold; hair tuft vein lB UNH in J, tail (17-9 mm, -¥ 13 mm; F c!19mm,-¥ 20 36. Chrysoplectrum pemiciosus pemiciosus Herrich-Schaffer 1869 mm. Plate 21 This species is widespread but not common in forested situa­ tions in Trinidad to at least 2,000 ft. It is my impression that males defend territories which may be sunlit patches in the forest The nominate subspecies perniciDsus occurs from Panama to (Nariva Swamp, Andrew's Trace, Las Lappas Trace) or forested south Brazil, and the subspecies epicincia Butler & Druce is found hill tops (Mt. Tabor). I have seen no specimens at flowers, and in Central America. Kaye (1940, p. 566) records this species from caught only one female for six males. Life history and foodplant Trinidad on the basis of a specimen collected near Arima (i-iii. apparently unknown. 1932, A. Hall), but the spec·imen is not in the Btv1NH. Recent captures confirm this as a Trinidad species. Ground colour very dark brown; hyaline spots white; the 34. Polythrix auginus Hewitson 1867 male to a slight degree, and the female to a negligible degree, Plate 27 have a green metallic tint to the UPS and UNS, which is only visible at certain angles, otherwise unmarked except for a This species is closely related to the last, but has a narrower narrow white band UNH near margin in spaces 1 C and 2. The range (Nicaragua to South Brazil) and is, if anything, less males have a curious bronze comb-like double row of short spines common through this range. Kaye (1940) confused this species on the hind tarsi which is distinctive for the genus, and from with P. caunus and both occur in Trinidad. which it derives its name. There is some variation in the extent of P. auginus is close to P. caul1us in colour and markings, but the hyaline markings: one male before me has reduced markings differs as follows: UNF base of costa strongly green; green at with no spot in space I B and only a dot in space 3; the other has base UNH stronger; UNH uniform pale brown with broad dark the spot in space 3 well developed and a dot in space I B, while brown margin and tails; pale area at base of spaces 1 Band 1 C the female has the spot in space 1 B well developed. A distinctive less pronounced; white bar across space 2 at 2/3; P. caunus species owing to the markings (or lack of them) and the comb

40 on the hind femora of the male. Costal fold; F 0 20.mm; ~ 21 within the Trinidad fauna are: spots in spaces 4 and 5 UPF mm. (more developed in ~); UNH discal bands continuous, the Generally scarce in Trinidad. I have taken three males on ou ter reaching the apex; UNH margin dark to space 7 .. Illustra­ Morne Catharine at c. 1,500 ft., and a female at West Moreau tions in Lewis (1972, Plate 88, No.5, cf) and Riley (1975, Plate in the south of Trinidad; S. Alston-Smith has taken it on the 21, No. 14); the illustration of this species in Barcant (1970, Fig. Arima-Blanchisseuse Road. The males defend territories of sunlit 4, No.1) is not U proteus r! as labelled, and although not ade­ clearings in the forest, usually returning to a favourite perch after quate for reliable identification most closely resembles U. esta seeing off intruders and are wary and difficult to catch. They rest Evans (No. 41b below). Costal fold; tail 6 12mm, ¥ 15 mm; in sunshine with their wings spread or partially spread. S. Alston­ F d'23mm, ~ 25 mm. Smith has reared this species in Trinidad. This is a common and widespread species in Trinidad and Tobago, frequenting rough, open ground at all elevations. It is common at flowers and I have captured a male feeding at a 37. Co do tractus carlos arguta Evans 1952 fresh bird dropping (ix. 1979, Curepe). Although I have not Plate 29 reared this species I have found the larvae in Epargyreus type leaf shelters quite commonly on bodie beans at Aranguez Evans (1952) described this species, recognising three sub­ and Macoya Gardens. It is of little importance as a pest, and is species: carlos Evans from Mexico to Colombia, rowena Evans usually kept in check by parasitoids, of which Trichogramma from Colombia and Venezuela, and arguta restricted to Trinidad. sp(p). attacking the eggs and Apanteles sp. attacking the larvae Ground colour brown; hyaline spots yellow. UNF end of cell, are the most commonly mentioned. It probably uses a number of base of spaces 4 and 5 and UNH ground colour pale mauve other legume vines as hostplants. In Puerto Rico Wolcott (1951) brown. UNH with three dark brown bands brown margin and records it attacking beans, cowpeas, Meibomia purpurea and white bar in space 1 C between outer band and margin. Dis­ Stigmatophylum lingulatum, and illustrates the adult, larva and tinguished from the rather similar Urbanus dorantes by the more pupa. Kendall (1976) records the larvae feeding on Desmodium contrasting UNH markings, the white bar UNH and the presence neomexicanum Gray (Leguminosae) in Mexico. of a hyaline spot in space 5 UPF. No costal fold; F t 22 mm, ¥ 25 mm. The type series of this species in the BMNH: 2!, 2 5f. Northern 40. Urbanus belli Hayward 1935 Mts. xii. 193 8-i.1939, A. Hall and a d' from Caparo Collected by Plate 47 Dr. RendalL As this species closely resembles U. dorantes, I (and probably other collectors) have overlooked it. Certainly while pre­ paring this account, I was pleasantly surprised to pick out a pre­ Evans (1952) considered Urban us viterboana Ehrmann to viterboana viously unrecognised female C. c. argu ta from amongst my series have three SUbspecies: from Mexico to the Andes, of U. dorantes (vi.l979, Brigand Hill). Specimens which appear alva Evans from Mexico to Argentina and belli from to be U. dorantes but are taken in forest situations should be and Argentina. Steinhauser (1981) combines belli and alva closely examined. I do not know the resting position of this (belli is the senior name) and considers belli and viterboana to be species, but it probably rests with it wings above its head, as does distinct species. Kaye (1921, 1940) did not recognise this or the U. dorantes. Life history and foodplant unknown. next five species which have been lumped together under U. proteus in Trinidad collections. Colouring as for U proteus. Distinguishing features of U. [38. Co do tractus melon Godman & Salvin 1893] deleted belli include: no spots UPF in spaces 4 and 5 (not strictly true as Plate 31 a female before me has a dot in space 4); UNH the basal band broken into spots, the outer band indented slightly at the veins Apart from a single specimen labelled "Trinidad" in the and stopping at vein 6. Costal fold; tail 69-12 mm, !f.14mm; F 6 BMNH, this species is restricted to Central America (Mexico 21-22mm,~ 23mm. to Nicaragua). I strongly suspect this to be a mislabelled speci­ This species occurs in similar situations to the last, but also men and suggest C. melon be dropped from the Trinidad list. extends into forests, and is even commoner. In addition to Trinidad and Tobago it has been recorded from Chacachacare Island (Cock 1981b). Because of past confusion with U proteus 39. Urbanus proteus proteus Linnaeus 1758 the life history has not been recognised; undoubtedly the larval Plate 44-46 foodplants are Leguminosae, probably vines.

Evans (1952) recognises two subspecies of this common and widespread species: proteus which is found from the southern 40a. Urban us pronta Evans 1952 U.S.A. to Argentina, including Trinidad and Tobago, and Plate 48 dom ingo Scudder which is darker with reduced markings and found throughout the Caribbean to Grenada. This species is moderately common from Mexico to The colouring of this and the next six species is basically Paraguay. Having compared my specimen with the series in the similar: ground colour brown; base UPF, base and disc UPH and BMNH, I can now confirm the tentative record in the addendum body UPS with overlay of metallic green hairs; hyaline spots to Cock (1982). white; UNH with two discal bands, 2 spots in space 7, the tails In colouring this species resembles U pro teus, and in and to some degree the termen dark on pale brown ground markings is closest to U. belli. It can be distinguished principally colour. The identification of these seven species will cause some by the outer band UNH which is crossec by pale veins and start­ problems, and the reader is referred to the key at the end of this ing to break up into spots in spaces 3 - 6. Other characters include section. Steinhauser (1981) gives a detailed treatment of this the absence of spots in spaces 4 and 5 UPF (again not strictly part of the genus. The distinguishing features of U. proteus true, my female has a narrow arc-shaped spot in space 4), the

41 macular basal band UNH and a strong white bar beyond the outer UNH; margin dark to vein 3. The illustration of U. proteus in band in space lC. Costal fold; tail ~ 10mm; F Sf. 24mm. Barcant (1970, Fig. 4, No.1) may well be of this species. No Probably a rare species in Trinidad. I cannot add to my costal fold; tail ~ 10mm; F st 19mm. capture of a female in Parry lands (i. 1980). Life history and Probably a rare speCIes in Trinidad. I cannot add any larval foodplants apparently unknown. information to my capture of a Single female in Parry lands (viii. 1980) at flowers of Chromolaena odorata. Life history and larval foodplants apparently unrecognised. 41. Urbanus esmeraldus Butler 1877 Plate 50 42. Urbanus acawoios Williams 1926 This is another fairly common and widespread species in the Plate 52 Americas, being found from Mexico to Paraguay. Evans (1952, p. 89) records 1sd' ~ from Trinidad, so although Kaye must have This species seems to be less common than those of the genus seen some of these he did not distinguish them from U. proteus. considered so far and is found from Panama to the Amazons U esmeraZdus resembles U proteus in colouring, although only. There are no specimens from Trinidad in the BMNH, but the green may be more brilliant, and it is also closest to U Steinhauser (1981) and Cock (198 2) record this species from pro teus in markings: spots in space 4 and 5 forming (or partially Trinidad. forming) an arc between the apical spots and the spot in space Colouring as for U. proteus and others of this group, but the 3; UNH ground colour pale brown, the bands standing out green not as bright. The principal distinguishing character of this clearly; basal band split into two very clear spots in space 1C and species is the narrow forewing (width less than 1/2 costa), but it cell; outer band slightly broader than in other species and reaches can also be recognised by the plain cilia of the forewing, which vein 6 where it meets a dark patch at apex UNH; margin UNH are chequered in all the other species of the genus with green on dark to vein 3. Costal fold; tail! 10mm,~ ll-13mm; F d'21mm, the UPS. Other characters include: no spots in spaces 4 and 5 ~ 22-23mm. UPF; UNH basal band weakly broken into spots, outer band This is a somewhat localized and uncommon species in continuous to vein 6, no dark patch at apex; UPF markings Trinidad, but is probably overlooked amongst the other similar reduced and spaced out. Costal fold; tail d'8 mm; F !2Imm. species. I have a female from Nariva Swamp (Cock 1981a), and This seems to be a genuinely uncommon species in Trinidad; several specimens from Irois Beach taken at flowers of Lantana Steinhauser (1981) lists one specimen from Trinidad, Cock camara L. (ix.1982). There are specimens from Caparo and the (1982) lists captures from Arima-Blanchisseuse Road, milestone Northern Range in the BMNH. Kendall (1976) records the larval 10 and Caroni Swamp, and I can add one male from San Miguel foodplant as Desmodium neomexicanum (cL U proteus). valley at 500 ft. (ix. 1982). Foodplant and life history apparently unknown.

4la. Urbanus esma Evans 1955 Plate 49 43. Urbanus dorantes dorantes Stoll 1790 Plate 30 This species is found from Colombia to Peru to the only (Evans 1952), but was recorded from Trinidad for the This species is common throughout its range from the first time by Steinhauser (1981) and included in Cock (1982). southern USA to Argentina. In addition to the nominate sub­ U esma is another of the proteus group in colour and species from the mainland, Evans (1952) recognises two sub­ markings; it resembles U esmeraldus most but lacks the hyaline species from the Greater Antilles (but not in Jamaica), one from spots in spaces 4 and 5 UPF; UNH ground colour paler than throughout the Lesser Antilles including Barbados, and one from U. esmeraldus, basal band split into spots; outer band broad the Galapagos Islands. Trinidad and Tobago specimens belong to to vein 6 where it meets, but is distinctly darker than, a dark the nominate subspecies. patch at apex UNH; margin UNH broadly dark and a conspicuous Ground colour brown with no green colouring; hyaline spots pale bar in space lC between the outer band and the margin. yellow; cilia pale but dark at ends of veins; UNH mauve-brown The specimen listed by Steinhauser remains the only one with brown central and outer bands, interupted at vein 6, and from Trinidad known to me, but it could be confused with other submarginal wavy brown line; dark spot towards base of space 7; species of this appearance. Life history and larval food plant un­ tails dark. The combination of no green, chequered cilia and recognised. spaced out UPF spots is distinctive within the genus, but do not serve to distinguish U dorantes from Codotractus carlos arguta. These two species are most readily distinguished by the 4lb. Urban us esta Evans 1952 tone of the UNH markings which are much less contrasting in Plate 51 U dorantes, but also the absence of a hyaline spot in space 5 UPF. Illustrations in Lewis (1973, Plate 88, No.3; UNS) and This species seems quite common throughout its range from Riley (1975, Plate 21, No. 15, subsp. cramptoni Comstock from Mexico to Argentina. In the addendum to Cock (1982) I recorded Haiti). Costal fold; tail d' l2mm, ~ l4mm; F d'21-22mm'!f 22mm. a possible Trinidad specinlen, which I can now confirm, having This species is very common at flowers in open spaces all compared it with the BMNH series. over the island where its foodplant (Desmodium spp.) grows In colouring it is similar to U. proteus, and the specimen I and on Tobago and Chacachacare (Cock 1981b). It is less likely have is undOUbtedly smaller. It can be recognised by the follow­ to be found in forested areas (except along roadsides) and speci­ ing combination of characters: no costal fold in c!; UPF no mens from such areas should be examined carefully lest they be spots in spaces 4 and 5; UNH basal band macular; outer band Codotractus c. arguta. I found a larva of Urbanus dorantes feeding straight, continuous to vein 6 where it meets a dark patch at apex on leaves of Desmodium incanum DC (Leguminosae) by night

42 (19.00h). Nearby was a pair of leaves spun together which This is yet another widespread and common species found doubtless was the larval shelter. The larva had a large, dark brown from Mexico to Argentina. The colouring is similar to the last head, conspicuously rounded, but slightly indented at the vertex; two, but the hyaline spots are reduced and dUsky. Distinguished prothorax with a brown dorsal plate and dark legs; mesothorax, from the last twO' species by the dark hyaline markings and the and abdomen with a thin dark green dorsal line; lateral flap-like presence of a costal fold in the male, but this and the next species protuberance running just above legs paler than rest of body; are very similar. IllUstrations in Lewis (1973, Plate 88, No.6) and laterally an orange stripe on mesothorax, metathorax and Smart (1976, Page 113, No. 52). Costal fold; tail cr 12-13mm; F J abdomenal segment 1 developing into spots on abdomenal seg­ 2l-22mm. ments 2 to 8, made more conspicuous by darkening above and As for the last two species, this is widespread and common, below each; remaining thoracic legs and prolege concolorous; occurring on rough open ground throughou t Trinidad and spiracles inconspicuous, black; anal plate pale brown; body Tobago. It feeds freely at flowers and Stachytarpheta spp. are a covered with sparse, short, pale hairs. The pupa is thick and short favourite. Moss (1949) records the larva feeding on Schrankia sp. with a long, pointed cremaster; the wings and thorax dark brown; at Para, but the single specimen I have reared was feeding on abdomen pale brown with dark speckles, and a macular dorso­ Kudzoo, Pueraria phaseo[oides, and sheltered in an Epargyreus­ lateral stripe. The pupal stage lasted nine days. type flap. In the final instar the larva had a large black, rounded, pubescent head indented at the vertex, with raised striations running down the face; prothorax black dorsally, dark ventrally; 44. Urbanus teleus HUbner 1821 mesothorax, metathorax and abdomen ground colour brown­ Plate 40, 42 orange, darker ventro-1aterally; dark green dorsal stripe; abdomenal segments 2 - 8 with a small dark green spot on either A common species spread from Mexico to Argentina. Found side O'f the dorsal line and an O'range spot dorso-laterally on in Trinidad and Tobago but not elsewhere in the Caribbe<).n. anterior margin, the orange spots weakly joined up to form a Ground colour brown; hyaline spots on F white; UNH pale stripe; spiracles inconspicuous, black; thoracic legs dark; prolegs brown with dark brown central and outer bands. Very close to concO'lorous except tips orange; body covered with short, dark, the next species, under which differences are discussed. No scattered pubescence. In the penultimate ins tar the head is costal fold; F hyaline bar width! 0.5mm, ~ 0.75mm; tail tlO- similar; prothorax with orange lateral spot; mesothorax, meta­ 11mm,~ l3mm; F d'20-2lmm,!f. 22mm. thorax and abdomen ground cO' lour yellow-green (dull green This is a moderately common species, found on rough open speckled with yellow); clear dorsal line; mesothorax, metathorax ground throughout Trinidad and Tobago at all elevations. It feeds at flowers, especially Eupatorium (s.l.) spp. Moss (1949) and abdomenal segments 1 - 8 with orange spots dorso-laterally records the larva feeding on a Schrankia sp. (Leguminosae) at on anteriO'r margin, most prominent on abdomenal segment 2 - 7; Para. In contrast, Kendall (1976) observed females in Mexico prbthoracic legs dark, remaining thoracic legs and pro legs concolorous. The pupa is 25mm long, light brown with black ovipositing on Panicum maximum Jacq. (Gramineae) and bred a long series feeding the larvae on P. maximum and , spiracles; thorax speckled; abdO'men with dorsal bands of SPO'ts centrally, and dashes posteriorly, O'n each segment. Sorghum halepense (L.) Pers. He describes the young larvae as cutting and folding over a portion of leaf; mature larvae hide in the grass near the ground and pupation is in a flimsy cocoon among debris on or near the ground. The detail of Kendall's 47. Urban us procne Plotz 1881 observations throws doubt upon the record above of Moss. Plate 33, 34, 39

45. Urban us tanna Evans 1952 Plate 41,43 This is a widespread species found from the southern USA to Argentina, but less common than the last. Not recorded from Not as widespread as the last, occurring from Mexico to Trinidad by Kaye (1921, 1940) but listed by Evans (1952). Ecuador and the Guyanas, commonest in northern South Very close to the last in colouring, but the F hyaline spots are America from Colombia to Trinidad. Colouring as for the last. almost cO'mpletely obscured in the male (less so in the female) The simplest difference between U. teleus and U. tanna is the and the UNH bands stand out more against a clearer ground num ber of apical spots on F: 4 for U. teleus and 5 for U. tanna; colour. The best feature to distinguish U. procne from U. this holds good for all Trinidad specimens that I have seen, but simplicius is that in the former the two spots in space 7 UNH I would expect this to be a variable character. A better character are separated from the central band, whereas in U. simplicius is the width of the UPF hyaline bar: which in each sex is the central bands joins the outer spot in space 7. Costal fold; narrower in U. teleus (but note f u. teleus is comparable to ! tail 613mm; F cf22mm. u. tanna). No costal fold; F hyaline bar width cfO.75mm, ~ lmm; This species is less common than the last, although doubtless tail !9-l1mm,~ l4mm; F d'21mm, 5t. 23mm. confused with it in the field. The fact that three of the six speci­ I have found this species to be about equally common as the mens I have seen were from swampy areas (Nariva Swamp, Caroni last, although the preponderance of U. tanna in the BMNH (43 Swamp) is suggestive. Kendall (1966, 1976) has recorded this as opposed to 8), suggests it may be the commoner. It occurs in species feeding on grasses in Texas and Mexico. Oviposition was similar situations as the last and has similar habits. Life history noted on (L.) Pers. and the larvae fed on and larval foodplant not recognised; the two distinct foodplant several other species of grass; the behaviour is similar to that records for U. teleus above lead one to speculate that Moss's described for U. teleus above. (1949) record of Schrankia sp. may actually apply to U. tanna.

46. Urbanus simplicius Stoll 1790 (Urbanus athesis Hewitson 1867) Plate 32, 35 Plate 36

43 A scare species of restricted distribution, U. a thesis is BMNH and 1 have two males from Morne Catharine and a third recorded from Panama, Colombia, Venezuela and Tobago (Evans from Chacachacare Island (Cock 1981 c). Although the specimens 1952). It is similar in appearance and colouring to U. teleus, U. from Morne Catharine were taken in light forest, that from tanna and U. carmelita, but can be distinguished from the first Chacachacare was feeding on flowers of Chromolaena odorata two by the hyaline spot in space 3 UPF being out of line with the in the open near the lighthouse. The life history in Trinidad is bar formed of the other spots or absent, and from U. carmelita unknown, but Moss (1949) records this species feeding on by the absence of the white margin to the termen of the UNH Schrankia sp. and Cassia sp. in Brazil: "the larva is purplish before the cilia which that species has. Costal fold. brown, the middle area on each side of a dark median line being Although this species has not been recorded from Trinidad, marked by some six patches of light sienna and a broken design it occurs on the mainland, and there is a single female in the shaped like an X. The head is round and dark brown." BMNH from Tobago bequeathed by W.H. Sheldon. 1 suggest that this may be the "Eudamus aminias Hew " listed by Sheldon 49. Urbanus doryssus doryssus Swains on 1936) and Barcant (1970), but which I was unable to recognise Plate 38 (Cock 1982). It would be well worth watching out for this species on both islands. The nominate subspecies is found from Mexico to the Amazons, and another subspecies, albicuspis Herrich-Schaffer 48. Urban us cal'melita trebia Moschler 1878 is found from southern Brazil to Argentina. Similar in colour and Plate 37 markings to U. carmelita, apart from a broad white margin UNH apex to tail and UPH vein 3 to tail. This is the only species known This is an uncommon species listed by Evans (1952) from from Trinidad with white tails, although there are others known Nicaragua, Panama, Colombia and Trinidad for this subspecies, from the Guyanas and Venezuela. Illustrations in Barcant (1970, and from Brazil for the nominate subspecies carmelita Page 86, No.4; as EUdamus brachius, a synonym) and Lewis Herrich-Schaffer. Colouring as for U. teieus, U. (anna and U. (1973, Plate 88, No.4) Costal fold; tail cf4mm,~5mm; F t20, 0 athesis; distinguished from the first two by the spot in space 3 ~22 mm. UPF being absent or if present detached fro~ the central band (as This distinctive species is not common in Trinidad, but cap­ for U. a thesis) and the shorter tails; distinguished from U. athesis tures have been made at all altitudes in forested areas in the by the UNH termen narrowly white before the cilia. Costal fold; North (El Tucuche iii.1979; Fondes Amandes, iv.1979 S. tail ci 4-5mm; F a21-22mm. Alston-Smith) and more commonly in the South (Moreau, Morne This species appears to be restricted to the North-west of Diable, Sir N. Lamont; Trinity Hills, xii.198l; Parrylands iii. Trinidad. There is one female from St Anns (A. Hall) in the 1980). Life history and larval foodplant unknown.

Codotractus melol15f; UNS ("Trinidad", specimen in BMNH); 32, Plates 24-38. 24. Polythrix octomaculata octomaculata if UPS Urbanus simplicius If UPS C'!I est Moreau, 3l.xii. 1979, MJWC); ~Nariva Swamp, 22.v.1979, MJWC); 25, P. roma ~ UPS (Curepe, 33, U. proclle r1 UPS (Caroni Swamp, 20.ii.1982, MJWC); 34, U. lX. 1982, MJWC); 26, P. CGlInus dUNS (Morne Catharine, 24.iii. procne rf UNS (as 33); 35, U. simpiicills r! UNS (Spanish Farm, 1982, MJWC); 27, P. auginus

44 Plates 39-52. 39, U. procne ~ UPS (French GUiana, specimen in MJWC); 46, U. proteus Si- UNS (Grande Ravine, n. vi. 1979, BMNH); 40, U. teleus dUNS (Parrylands, 2.ii. 1980, MJWC); MJWC); 47, U. belli cluN's (Andrews Trace, 14.x.1980, MJWC); 41, U. tanna c! UNS (Tobago, Speyside, 15.v. 1982, MJWC); 48, U. prollta ~ UNS (Parry1ands, 13 .i.1981, MJWC); 49, U. esma 42, U. teleus!f.: UNS (Arima-Blanchisseuse Texte1 Road, 8.x .1979, Sf.. UNS (GuYana, specimen in BMNH); 50, U. esmeraldus JUNS MJWC); 43, U. tanna Sf. UNS (Curepe, 14.viii.l978, MJWC); (Marne Catharine, 18.ix.1982, MJWC); 51, U. esta 4? UNS 44, U. proteus J UPS (Curepe, feeding on bird dropping, 23. (Parrylands, 21.viii.1980, MJWC); 52, U. acawoios c! UNS (San ix. 1979, MJWC); 45, U. proteus f UPS (Cats Hill, 19.ix.1982, Miguel Valley, 3.ix. 1982, MJWC).

8 UNH central dark band clearly broken into distinct separate Key to the tailed skippers of Trinidad spots in space 1 C and cell; these spots stand out sharply against the ground colour 9 UPS with at least some green colour 2 UNH central dark band more or less unbroken across space UPS with no green colouring 14 1 C and cell; these spots do not stand out as sharply against ground colour 11 2 UNH with white band running width of wing 3 (Note this is a somewhat subjective character and should be UNH no white band 5 used in conjunction with the plates; Steinhauser (1981) interprets it differently). 3 H tail long (more than 8mm); apical spots in straight line; spot in space 2 continguous with spot in space 3 not cell 9 UPF with white hyaline markings in spaces 4 and 5 spot Polythrix metalleu'ens Urbanus esmeraldus H tail short (less than 5mm); spots in spaces 6 and 7 dis­ UPF no markings in spaces 4 and 5 10 placed from apical line; spot in space 2 contiguous with cell spot not spot in space 3 4 10 Larger (F 22mm); green on UPH extending almost to termen Urban us esma 4 HW lobed rather than tailed (2mm); white band angled Smaller (F 18mm); green on UPH little more than basal half towards dorsum; white band 2.5mm in:f' 1mm in d Urban us esta Aguila aurunce HW tailed (4mm); white band angled towards margin near 11 UNH outer band crossed by pale veins and becoming macular tail; white band 1.75mm in !f.' 0.75 in d Aguna coelus in upper part Urballus pronta 5 UPF white hyaline spot in space 1 B 7 UNH outer band with concolorous veins; the band may be UPF no white hyaline spot in space 1 B 6 indented at the veins (u. belli) but is not maculate 12

6 UNF base of costa strongly green below; white band across 12 Usually white hyaline spots in spaces 4 and 5; UNH outer space 1C UNH Polythrix auginus band straight and continuous to dark area at apex UNF base of costa negligibly green; only white spot on vein Urballus proteus 1B Poly thrix caunus Usually no white hyaline spots in spaces 4 and 5; UNH outer band indented at veins, stopping at vein 6 and no dark 7 F narrow, less than 1/2 costa; F cilia plain Urballus acawoios area at apex Urballus belli F broader, greater than half costa; F cilia chequered (dark at end of veins) 8 13 Tails white Urballus dvr)'ssus

45 Tails dark 14 COCK, M.J.W. (1981a). Lepidoptera of Nariva Swamp. Living World. Journal of the Trinidad and Tobago Field Naturalists' 14 Dark spots at 1/3 and 2/3 in space 1B UPF 15 Club, 1981-1982,21-22. No such dark spots 16 COCK, M.l W. (1981 b). Butterflies from Chacachacare Island in­ cluding three species new to Trinidad. Living World. Journal 15 Hyaline spots in spaces 4 and 5 UPF Polythrix roma of the Trinidad and Tobago Field Naturalists' Club, 1981- No hyaline spots in spaces 4 and 5 UPF 1982,25. Polythrix octomaculata COCK, M.J.W. (1981 c). The skipper butterflies (Hesperiidae) of Trinidad. Part I Introduction and PYrrhopyginae. Living 16 Spots of costa, cell, spaces 2, 1 B and sometimes 3 aligned World. Journal of the Trinidad and Tobago Field Naturalists' to form a bar OR such markings reduced to give an un- Club, 1981-1982, 52-56. marked appearance 17 COCK, M.J.W. (1982). The skipper butterflies (Hesperiidae) of Markings not arranged in such a bar, but scattered 22 Trinidad. Part II SYstematic list of the Trinidad and Tobago species. Occasional Papers of the Department of Zoology, 17 UPF white hyaline markings clear and conspicuous 18 UWI, St Augustine, No.5, 49 pp. UPF hyaline markings dusky, reduced or absent 21 COCK, M.J.W. (1984a). Lepidoptera Notes 1-6. Living World. Journal of the Trinidad and Tobago Field Naturalists' Club, 18 UPF white hyaline spot in space 3 not aligned with those of 1983-1984,35-37. 1B, 2, cell and costa to form bar, OR spot in space 2 absent COCK, M.J.W. (1984b). The skipper butterflies (Hesperiidae) of 19 Trinidad. Part 3 Pyrginae (first section). Living WorId. UPF white hyaline spot of space 3 aligned with others to Journal of the Trinidad and Tobago Field Naturalists' Club, form bar 20 1983-1984,38-42. COCK, M.J.W. and EVANS, H.C. (1984). Possibilities for 19 UNH termen narrowly white before cilia Urban us carmelita biological control of Cassia tara and C. obtusifolia. Tropical UNH brown up to cilia Urbanus athesis Pest Management, 30, 339-350. CROWFOOT, W.M. (1893). Preliminary list of Trinidad butter­ 20 Four apical spots Urbanus teleus flies. Journal of the Trinidad Field Naturalists' Club, 1, 173- Five apical spots Urbanus tanna 174. DE JONG, R. (1983). Annotated list of the Hesperiidae (Lepi­ 21 UNH spots in space 7 completely distinct from central doptera) of Surinam, with descriptions of new taxa. Tijdsch­ band Urbanus procne rift voor Entomologie, 126, 233-268. UNH central band runs into outer spot in space 7 EV ANS, W.H. (1952). A catalogue of the American Hesperiidae Urbanus simplicius in the British Museum (Natural History). Part II Pyrginae. Section I. British Museum (Natural History) Publication, 22 UPF white hyaline spot of space 1 B absent (or if present 178 pp. directly under that of space 2); UNH with dark striations EVANS, W.H. (1955). A catalogue of the American Hesperiidae Typhedanus undulatus in the British Museum (Natural History). Part IV Hesperiinae. UPF white hyaline spot present in space 1 B (under spot in British Museum (Natural History) Publication, 499 pp. space 3); UNH without dark striations, mottled instead 23 KAYE, W.J. (1921). Catalogue of the Trinidad Lepidoptera Rhophalocera. Memoirs of the Department of AgricultUre, 23 UNF with dark triangle on costa, apex of triangle just Trinidad and Tobago, 2, 163 pp. beyond spot in space 6 Chioides catillus KA YE, W.J. (1940). Additions and corrections to the recorded UNF without such triangle 24 species of Trinidad butterflies. Transactions of the Royal Entomological Society of London, 90, 55-573. 24 Male with costal fold; UNH ground colour brown; con­ KENDALL, R.O. (1966). Larval foodplants for five Texas spicuous dark spot at base of space 7 UNH; UPF with Hesperiidae. Journal of the Lepidopterist' Society, 20, 35-41. no hyaline spot in space 5 Urbanus dorantes KENDALL, R.O. (1976). Larval foodplants for thirty species of Male without costal fold; UNH ground colour pale brown; skippers_ (Lepidoptera: Hesperiidae) from Mexico. Bulletin no dark spot at base of space 7 UNH; hyaline spot in space of the Allyn Museum, 39, 9 pp. 5 UPF Co do trac tus c. arguta LEWIS, H.L. (1972). Butterflies of the World. Harrap, London, 312 pp. MOSS, A.M. (1949). Biological Notes of some Hesperiidae of the Amazon. Acta zoologica lilloana, Tucuman, 7,27-79. RILEY, N.D. (1975) A field guide to the butterflies of the West Indies. Collins, London 224 pp. SHELDON, W.G. (1936). Tobago and its butterflies. The Entomologist, 69, 1-9. REFERENCES SHELDON, W.G. (1938). Additions to the butterflies of Tobago The Entomologist, 71,29-31. BARCANT, M. (1970). Butterflies of Trinidad and Tobago. SMART, P. (1976) The illustrated encyclopedia of the butterfly Collins, London, 314 pp. world. Hamlyn, London, 275 pp. STEINHAUSER, S.R. (1975). An annotated list of the BROWN, F.M. and HEINEMAN, B. (1972). Jamaica and its Hesperiidae of El Salvador. Bulletin of the Allyn Museum, butterflies. E.W. Classey Ltd., 478 pp. 29,34 pp.

46 STEINHAUSER, S.R. (1981). A revision of the proteus group of Tobago, 1(4),200-308. genus Urban us Hubner Lepidoptera: Hesperiidae. Bulletin of WOLCOTT, C.N. (1951). Hesperiidae: the skippers. pp. 562- the Allyn Museum of Entomology, 62, 41 pp. 573 in Wolcott, C.N. The insects of Puerto Rico. Journal of WILLIAMS, R.O. (1931). Leguminosae. Flora of Trinidad and Agriculture of the University of Puerto Rico, 32, 418-748.

47 The Family Aristolochiaceae in Trinidad, with reference to its medicinal uses, its folklore, and its use as a larval foodplant by Papilionidae.

By Julius O. Boos 314 N.E. 26th Drive, Wilton Manors, Fort Lauderdale, Florida 33334 U.S.A.

ARISTOLOCHIA, a genu~ with worldwide distribution, is known and used for its alleged medicinal properties. Five species (one introduced) are found on Trinidad, one of these, A. boosii Panter 1981" being a species new to science. Other species have been re­ corded as being introduced for cultivation, i.e., A. elegans Mast., A. hians Willd., and A. odoratissfma L. I have been unable to find specimens of these in Trinidad. The following are the species found on Trinidad and Tobago, with notes on their alleged medicinal and mystical properties. The synonymy is form Pfeifer (1966). 1. GRANDIFLORA Sw., Prod. Veg. Ind. Occ. 126. 1788, non Vahl (1791), nec Gomez (1803), nec Arruda (1816). (Type: Swartz s.n.,S) - Fig. 1. A. scandens P. Browne, Civ. & nat. hist. Jamaica 329, 1756, non Mill. (1768). (ex char.) A. cordi/lora .Mutis ex H.B.K., Nov. Gen. Sp. PI. 2: 149, 1817. (Type: Mutfs s. n., not seen) A /oetens Lindley, Bot. Reg. t. 1824, 1836. (ex ic.) A. gigas Lindley, lac. cit. t. 60, 1842. (Type: Hartweg s.n., not seen) A. gigantea Hook., Bot. Mag. t. 4221,1846. (ex ic.) Howardia gigantea (Hook.) Klotzsch, ~/10natsb. Acad. Berlin 1859: 610, 1859. (quoad syn. Hook.) H. grandi/lora (Sw.) Klotzsch lac. cit. Aristolochfa grandi/lora (Sw.) B hookeri Duchr. in DC., Prod. 15 (1 ): 473, 1864. (Based on4 A. gigal1tea Hook.) Local name: "swan flower". Fig. 1. Aristolochia grandiflora Sw. Description & Distribution. This vine is common in the valleys south of the Northern Range and I have also seen it on Barcant in "Butterflies of Trinidad and Tobago" erroneously lists the road between Maracas Bay and Las Cuevas on the North it as the possible foodplant of sesostris (R & J), however, Coast. It is also reported from Mexico, Guatemela, Honduras, El I have never seen A. gralldij70ra in the southwest of the island to which area P. sessostris is confined, and the foodplant of P. Salvador, Nicaragua, Costa Rica, Panama and Cuba. it sesostris on Trinidad and Tobago is A. boosi!. Hamilton A. Tyler can be a substantial plant at times covering whole in "The Swallowtail Butterflies of N. America" (Naturegraph clumps of bamboo, cocoa trees, etc. The flowers are Publtshers, Inc., Healdsburg, CA ..95448.) lists PaI'ides photinus large, up to 25 cm across, with a pungent odor of (Doubleday) 1844, P. montezuma (Westwood) 1842, P. i. decaying meat which attracts flies and carrion beetles (Staphalinidae and Histeridae) for pollination. The seed iphidamus (Fab.) 1793 and P. arcas myiotes (Bates) 1861, as pods split at the bottom allowing the seeds to fall to the feeding on it in El Salvador. Some small larvae of sawflies ground. (Hymenoptera) and moths have been seen feeding on this plant on Trinidad. Medicinal Properties. No references can be traced to this species regarding medicinal uses except a note that it is extremely 2. ARISTOLOCHIA RUGOSA Lam., Encyc\. Meth. Bot. 1: poisonous. 252, 1783. (ex ic. cit.) - Fig. 2 Butterfly Larval Foodplant. No Papilionidae have been found A. obtusata Sw .. Prod. Veg. Ind. Occ. 126, 1788. (ex ic. cit.) using this as a larval foodplant on Trinidad and Tobago. Malcolm A. barbata Jacq., Ic. Pl. Rar. 3: t. 608, 1789. (ex ic.)

48 A. dictyantha Duchr., Ann. Sci. Nat., seL 4, 2: 40. 1854. Medicinal Properties & Folklore. This plant is used medicinal­ (Type: Vargas s.n., photo G-DC) ly as an antidote for snake bite and scorpion stings' where the A. eur),stoma Duchr., loco cit. 41. (Type: Beaupertuis S.Il., P) whole roots are soaked in alcohol together with other ingredients, Howardia obtusata (Sw.) Klotzsch, Monatsb. Acad. Berlin as an abortifacient and a remedy for excessive pain during 1859: 612, 1859. menstruation (dysmenorrhoea), and as an aid to ease the pain of H. barbata (Jacq.) Klotzsch, loco cit. 613. childbirth. It has also been reported as being used to treat indi­ H. schomburgkii Klotzsch, loco cit. (ex char.) gestion, heat, amenorrhoea, jaundice, and as a post partum Aristolochia dictyantha Duchr. p. schomburgkii (Klotzsch) depurant. It was identified to me by numerous people as being Duchr. in DC., Prod. 150): 447,1864. the "wild tref." A. rumicifolia Schomb. ex Duchr., loco cit., non Mart. Butterfly Larval Foodplant. Two of the three species of (1824) pro syn. Parides, neophilus parianua (R. & J.) and anchises cymocles Local names: "wild tref", "matte root", "mat root". (Db!.) use this as a larval foodplant, as does one of the three species of , B. polydamus (R. & J.) found on Trinidad and Description & Distribution. This plant is the most widely Tobago. distributed species on Trinidad and Tobago. I have found it throughout the two islands. It is also recorded from Antigua, 3. ARISTOLOCHIA TRILOBATA L., Sp. PI. 960,1753. (Type Guadeloupe and Martinique. It is normally a small plant, Linn. Herb. London, no. 1071.1, photo) - Fig. 3. preferring to grow up banks formed by road cutting where, under optimum conditions, the leaves can reach a length of 25 cm. A. scandens P. Browne, Civ. & nat. hist. Jamaica 329, 1756. Normally however, the leaves do not exceed 13 cm. The flowers (ex char.) are small, not exceeding 10 cm in length and 2.5 cm in width at A. trifida Lam., EncycI. Meth, Bot. 1: 251, 1783. (ex char.) the lip, and are sometimes called "dutchman's pipe". They also A. triloba Salish., Prod. 214, 1796. (Based on A. trilobata L.) smell of decaying flesh and attract flies and minute beetles for A. surinamensis Willd., Sp. PI. 4: 151,1805. (ex ic. cit.) pollination. The roots tend to be tuberous. The seeds are A macroura Gomez, Mem. Acad. Lisboa 3: Mem. dos dispersed as in A. grandij7ora. Corresp. 77, 1812. (Type: Martius inter Basil 1817, photo M) A. appendiculata Veil., Fl. Flum. 9: t. 98, 1827. (ex ic.) A. caudata Booth ex Lind!., Bot. Reg. t. 1453, 1831, non Jacq. (762), nec Parodi (1878). (ex ic.) A. macrota Duchr. in DC, Prod. 15 0): 447, 1864. (Type: Sc/zomburgk 679, photo G-DC.) A. tapetotricha Lem., Illustr. Hortic. 3: Misc. 22, 1856 (ex char.) Howardia macroura (Gomez) Klotzsch, Monatsb. Acad. Berlin 1859: 617, 1859. H. surinamensis (Willd.) Klotzsch, loc. cit. 613. H. trilobata (L.) Klotzsch, loc. cit. 617. Local names: "tref", "twef."

Description & Distribution. This plant is perhaps the most widely distributed species of Aristolochia in the area, being found throughout the West Indies, down the Gulf Coast of Mexico through Central America, Panama and into South America. I have found it only under cultivation on Trinidad and Tobago. However, it has been reported from the Manzanilla Road growing wild, but human habitation occurs in this area so it pro­ bably has gone wild from someone's garden. It can be a SUbstan­ tial plant if conditions are right. Leaves are dark green, waxy, trilobate. The flowers are reminiscent of the pitcher plants of N. America (Sarracenia), and attract flies and beetles for pollina­ tion. The seeds are dispersed as in A. grand if/ora.

-.:1~~~t'r'ft<'t.r) Fig. 3 Aristolochia trilobata L. Fig. 2. Aristolochia rugosa Lam.

49 Medicinal Properties & Folklore. This plant has a wealth of wider dispersal than the previously mentioned three species. So folklore associated with it and is venerated by some as a god. Al­ far A. boosii has been recorded only from the S.W. area of though surrounded by secrecy, so much so that it is not often Trinidad. spoken about, I have learned some, though not all, of the various uses of this plant and the rituals connected with it. Medicinally, it is the primary plant used as an abortifacient and is given by mothers to their daughters ostensibly as a "pain reliever" on or about when their periods are due. I have been told that the dose is three mature leaves boiled in a small quantity of water with a pinch of salt. This potion, which is reported to be extremely bitter, is then strained and drunk. The larvae of Battus polydamus (Papilionidae) which are commonly found feeding on this plant are collected and soaked in strong rum to be used both as a remedy for pains of any kind and to counteract the effects of potions given by women to men (locally called "stan home" or "stay home") to ensure marriage. It is also reportedly used to ease parturition, as a post-partum depurant and to treat hyper­ tension, diabetes, dysmenorrhoea, snake bite and scorpion sting. The leaves are considered to bring good luck in fishing, gambling and athletic sports if worn pinned, out of sight, inside the vest against the heart. It is believed that this plant must be planted in front of homes. An old English silver threepence piece is buried next to it together with other objects and potions. Various prayers are then said and the plant is then said to be "mounted" and now has additional powers of protection. If a person with evil intentions enters the yard the plant will warn the owner by "whistling". To maintain this good relationship, no one must be allowed to pick leaves or disturb the plant in any way. It must only be spoken to in pleasant tones; permisstion must be sought from it by the owner to pick a leaf or remove a sucker (young plant). These plants should only be given to people who appreciate their importance. Having them and treating them in the proper manner ensures good fortune and wealth to the owner. I stress that these beliefs are not confined to anyone class of person. My own specimen was obtained from an employee who kept it at White Hall, office of the Prime Minister! I was also told that proof of the plant's happiness is its flowering, and that when Fig. 4. Aristolochia boosii Panter. this happens good fortune is bound to occur for the lucky owner. A. habit B. flower dissected longitudinally, semi-digrammatic; C. gynostemium; D. fruit. Butterfly Larval Foodplant. As far as I know, Battus polydamus (R. & L), Parides neophilus parianus (R. & J .), P. anchises cymocles (Dbl.) use it as a larval food plant when it is Butterfly Larval Foodplant. This plant is used as food by the planted near their habitats. larvae of Parides sesostris (R. & J.) and Battus belus varus (R. & J .), both of whose larvae were collected and reared on it, and 4. ARISTOLOCHIA BOOSII Panter, 1981. Kew Bulletin Vol. also, I suspect, Battus lycidas (Cram). Adult females of this rare 30 (2) (1981) 231-233. Fig. 4. species were seen performing egg-laying behaviour around this plant. Since it occurs only in the S.W. Peninsula of Trinidad and is the larval foodplant of the mentioned butterfly species, their DescriPtion & Distribution. Specimens of this new plant to very local distribu tion on the island of Trinidad is accounted for. science and for Trinidad and Tobago were collected by myself at F.e. Urich reports that he saw a P. sesostris laying on this plant in Palo Seco and Gran Ravine, S, Trinidad, while following a female Los !ros forest in 1962. P. sesostris (R. & J.) in its egg laying. These specimens, which I took to our National Herberium, could not be identified locally Medicinal Properties & Folklore. No medicinal uses are re- and were taken to Kew Gardens by Dr. D. Philcox where they ported for this plant. were described as a new species (1. A. Panter, 1981) and named after the collector, (Holotype TRIN 22067 K.) It is a large vine 5. ARISTOLOCHIA ARBOREA Linden. that grows up quite tall trees and spreads across their canopies at 10-15 m. It is also found growing between and over bushes. The Description & Distribution. This species is at present known flowers are slender and upright, each abou t 7.5 cm long, and are only from two plants in the gardens near the National Herbarium borne on innoresences arising at the leaf nodes with one flower at the U.W.I., St. Augustine, Trinidad. It has been previously mis­ opening at a time, instead of singly at the basal leaf nodes as in identified as A. iongzjoiia Champ. It grows as a medium sized the other species. The seed pods are interesting. When the capsule shrub, and small purple-brown flowers are borne in cauliflorous, splits it forms a basket or "censer" which is blown or shaken by many-branched inflorescenses at the base of the upright "trunks" the wind causing the seeds to fall ou t and be blown a way from and on the exposed base of the roots. It occurs naturally in the area. Dispersal can take place only in wind, thus ensuring Mexico, Guatemala and El Salvador.

50 Medicinal Properties & Folklore. Nothing is known about the REFERENCES: properties of this plant. Its leaves when crushed do not smell like PANTER, J.A., 1981. Notes on Aristolochia from Trinidad the other species of Aristolochia in Trinidad. and the description of A. boo sii. Kew Bulletin Vol. 36 (2) 231-233. Butterfly Larval Foodplant. No butterfly is known to breed PFEIFER, H. W., 1966. Revision of the North and Central on this species in Trinidad. American Hexandrous species of Ar isto loch ia (Aristolochiacea). Annals of the Missouri Botanical Garden, Vol. 53. PHILCOX, D., 1977. Family Aristolochiaceae, Flora of Trinidad ACKNOWLEDGMENTS and Tobago Vol. II, Pt VIII pp 501-504 Ministry of Thanks to Dr. C.D. Adams, Dr D. Philcox, Miss J.A. Panter Agriculture Lands and Fisheries, Trinidad and Tobago. for their assistance in getting A. boosii identified, in fact SEAFORTH, C.E., Adams, C.D., and Sylvester, V., 1983. A described, and for providing needed literature. Hans Boos, Guide to the Medicinal Plants of Trinidad and Tobago. provided specimens upon which some of the art-work is based. Revised Version. Commonwealth Secretarial, Marlborough Susan Boos typed the draft. House, London.

51 BookReview

RICHARD FFRENCH. Birds of Trinidad and Tobago. Macmillan Caribbean, London and Basingstoke, 1986, to exclude must have posed a problem. I must confess to some 87 pp. £ 3.75. surprise at the inclusion of the Saffron Finch, a bird I have never Page proofs for this issue of Living World had already been seen, and the exclusion of the House Wren and the Turkey Vul­ corrected when I received a copy of this book for review. Rather ture but a choice had to be made and the author made his choice, than wait another 18 months or more I decided to stall the pub­ perhaps at least in part, on the availability of photographs. lication process for a few days so that a review could be included The photographs are mostly good, sometimes very good, but in this issue. a few do not do full justice to the bird and one or two, notably The book is one of a series of Macmillan books of similar for­ those of the Grey-breasted Martin and the swift, are almost use­ mat on aspects of Caribbean natural history. It deals with 83 less for the purpose of identifying the bird. The author most have species, "mostly the more common ones", of the 400 recorded known this but decided (rightly, I think) not to hold back in our islands and so does not pretend to be complete; the same publication until better ones were available. author has already given us a complete treatment. This book is All the information given about the birds I know well con­ aimed more at the "non-specialist reader" rather than the forms with my knowledge of them except in one instance. About advanced ornithologist and is a very attractive and welcome addi­ the Rufous-browed peppershrike the author says "Each bird has tion to the literature on our natural history. its own tune which it hardly alters". This is what the situation There is first of all an introduction that deals with general appears to be but the reality, revealed only to some-one who topics such as geography and ecology, the habits of birds, the actually lives where the bird is common, is much more fascinat­ description of birds, how to study birds and where to go to see ing. Each bird has a repertory of over forty tunes and within a dis­ them. There is much useful information here both for the begin­ trict each bird apparently has the same repertory. Why a pepper­ ner and the more experienced ornithologist. The author stresses shrike should need such a large collection of tunes while other the importance of using clues from habits and habitat preference birds need only two or three is a topic crying out for investi­ to assist in the identification of the more difficult species that gation. cannot be easily identified from a colour photograph. For a At this stage in the development of our country when en­ climpse at the problems of identification turn to the pictures vironmental abuses increase daily and there is an ever greater of the 3 sandpipers and the one plover to see how similar dif­ need for good conservation practices it is important that as many ferent species may appear in the field and these are only four of people as possible be made aware of this. the nine plovers and 24 sandpipers that occur in Trinidad. This book should help to stimulate that love and concern for The 83 chosen species are described under seven headings: living things from which the awarenes must spring. From this Gardens and parks; Open country and scrubland; Cultivated land standpoint it does not matter much that a couple of photographs with large trees; The forest; Swamps, marshes and reservoirs; The are not as good as the rest. It is much better to get the book pub­ coast; The air. There is a colour photograph of each species, a lished for the good it will do and I hope it will be a huge success. short description, and information about its song, food, nesting and general behaviour. The question of what to include and what Victor C. Quesnel

52