Journal of The Trinidad and Tobago Field Naturalists' Club 1993~94
Natura Maxime Miranda in Minimis
Published November 1993 LIVING WORLD is published biennially by The Trinidad and Tobago Field Naturalists' Club. All rights reserved. Typesetting, design and mechanical art by Digital Graphics & Advertising Technologies Co. Ltd. 18 Fitt St., P.O.S., Trinidad, W.l.
Living World Journal of The Trinidad & Tobago Field NaturalistS' Club 1993 _ 1994 1 Editorial Contents
The 1991-1992 issue of Living World was a The Tree Boa (Corallus enydris) on Trinidad and Tobago special centenary issue in which we looked back at the By: Robert W. Henderson and Hans E.A. Boos ...... 3 history of the club and the lives of the founding Additional Notes on the Fauna Recorded for the Off-shore members. Short biographies of five of the eight Islands North-West of Trinidad founders were published with the hope that biographies By: Hans E. A. Boos and Victor C. Quesnel ...... 6 of the remaining three would follow in this issue. Our A list of Food Plants and some Implications of the Feeding hope has not been fulfilled. Missing too is the seventh Behaviour of the Forest Grasshopper, Coscineuta tliretlS installment of Matthew Cock's series on the skipper (Thunberg). in Trinidad. butterflies of Trinidad but this is now at hand and will By: Ulory D. Mc Cornie ...... 8 appear in the next issue. A Juvenile Butterfly Collector in Trinidad 1933-36 Our centenary issue contained an article by By: Nicholas Guppy ...... 14 Mr. A. T. Carr on the early history of the club. This Feeding Behaviour of the Squirrel. Sciurus granatensis was a republication of an article in The Caribbean, a Humboldt. on Cocoa Pods in Trinidad W.I. magazine published by a now-defunct organization we By: David D. Chadee and David P. Chadee ...... 15 called the Anglo-American Caribbean Commission. The Vegetation Surrounding Mud Volcanoes in Trinidad M. Nardin, the French Ambassador to our country, has By: Paul L. Comeau ...... 17 written us to point out that by the time of the article's publication the organization was called simply The Further Records of Birds on Trinidad and Tobago By: Richard ffrench ...... 28 Caribbean Commission. He has also given us a little more information on the Caracciolo family that could Tents and Harems: Alteration of Leaves by be included in any future article on Mr. Henry Foliage-Roosting Bats Caracciolo, the club's first president. By: Thomas H. Kunz and Gary F. McCracken ...... 32 Six years ago we had already begun to receive Crane Flies (Diptera: Tipulidae) in Trinidad Caves articles produced on computers instead of typewriters. By: Johanna P.E.C. Darlington and Jon K. Gelhaus ...... 38 We must expect this trend to continue. However, we Patterns in the Calling Activity of the Pauraque Nightjar. do urge authors to make life easier for the editors by Nyctidromus albicoUis leaving adequate margins and increasing the leading By: Victor C. QuesneL ...... 42 between the lines to apptoach the double spacing of Book Reviews the now old-fashioned typewriter. But most of all: keep A guide to the Birds of Trinidad & Tobago 2nd edition the papers coming - v.c.Q. By:Richard ffrench. Plates and drawings by John P. O'Neill, portraits by Don R. Eckelberry. 1991...... 47 Front Cover The Trinidad and Tobago Field Naturalist' Club The Trinidad Field Naturalists' Club was founded on rhe 10th July, 189\. Its name was changed to the present one in 1974. It was incoIp9rated by an Act of Parliament (Act No.1? of 1991), The objects of the club are to bring together persons interested in the study of natural history, the diffusion of the knowledge thereof and the conservacion of nature. Monthly lecture meetings are held at St. Mary's College on the second Thursday of the month while field excursions are held on the last Sunday of each month, except December, when no official club activities are organised. Membership is open to all persons of at least fifteen years of age, who subscribe to the objects of the club. Management Committee - 1993: President - Presently Vacant Vice~President - Haroon Husain (657·3707) Treasurer- Selwyn Gomes (624·8017) Secretary - Christopher Starr (662·9477 or 663·1334. Ext. 2(46) Assistant Secretary - Dan Jaggernauch (659·2795) Committee Members - Rosemary Hernandez (645·2132). a. Morpho menelaus - dorsal surface Clayton Hull (632-0571) Jalaludin Khan (623·5559) b. Morpho menelaus - ventral surface President, Tobago Branch - Wilfred Des Vignes (660-4508) c. Morpho deidamia - ventral surface Editorial Committee: Victor Quesnel. Hans Boos. Yasmin Comeau d. Morpho deidamia - dorsal surface AU enquiries concerning the club or its journal should be addressed to the (see p.14) Honorary Secretary, P.O. Box 642. Pon of Spain, Trinidad & Tobago, W.l.
Living World Journal of The Trinidad &. Tobago Field Naturalists' Club 1993· 1994 2 The Tree Boa (Corallus enydris) on Trinidad and Tobago Robert W. Henderson! and Hans E.A. Boos2 I. Section of Vertebrate Zoology, Milwaukee Public Museum, 800 W. Wells St., Milwaukee, WI 53233 USA. 2. Zoological Society of Trinidad and Tobago, Emperor Valley Zoo, Port- of-Spain, Trinidad.
Introduction including 41 from Trinidad-Tobago, have been examined. Corallus enydris is a slender, nocturnally active, It is now possible to describe variation and define the arboreal boa (known as "Cascabel" or "Cascabel species on Trinidad and, with less confidence, on Tobago. Dormillon" on Trinidad) and one of the most Methods geographically widespread of snakes in the Neotropics RWH examined 35 specimens of Corallus enydris (Fig.l). As currently understood, its mainland from Trinidad and six specimens from Tobago; all are distribution stretches from southern Costa Rica to housed in collections in the USA (American Museum of southern Brasil just south of the Tropic of Capricorn in Natural History, New York (AMNH); Field Museum of the state of Sao Paulo. Its insular distribution includes Natural History, Chicago (FMNH); Florida Museum of small islets off the Atlantic and Pacific shores of Panama, Natural History, Univ. of Florida, Gainesville (UF); Isla Margarita (Venezuela), Trinidad, Tobago, llha Milwaukee Public Museum (MPM); Museum of Grande (southern Brasil), and in the West Indies, St. Comparative Zoology, Harvard University (MCZ); Vincent, at least nine of the Grenadine Islands and National Museum of Natural History, Washington, DC Grenada. (USNM). Over its extensive range, CoraUus enydris exhibits Measurements and scale counts were taken by tremendous variation in morphological characters (body traditional methods (see Henderson, 1991), and colour proportions, scale characters, and colour and pattern). It and pattern characters are as described in Henderson also demonstrates considerable ecological plasticity, (1991). Specimen examination usually included opening occurring in habitat that includes rain forest, mangroves, the stomach in order to determine if prey remains were fruit orchards, and cactus-Acacia scrub. Its diet varies present. geographically: mainland snakes exhibit an ontogenetic Results shift from birds to rodents, whereas West Indian Distribution: Coral/us enydris occurs over much populations shift from Anolis lizards to rodents of Trinidad from sea level to at least 360 m (Fig.2), (Henderson 1991, 1993). although most specimens have been taken in forested Over 500 preserved Corallus enydris collected areas in the northern half of the island (e.g., the from throughout its range over the past 100+ years, Northern Range, Mt. Harris in the Central Range). It has
~ •• • Ii:-") 0 '-Q- -"- • :~. \.r. • • • • D Area alx'M! 300m • •• .Area above 600m ~ I::IArea aboYc 600m • "'" ~" o 2S • km • • •
Fig. 1. Corallus enydris from Trinidad (photo by H. E. A. Boos). Fig. 2. Map of Trinidad indicating localities at which Corallus enydris was collected. Circles indicate records for specimens preserved and examined in this study and triangles indicate localities at which tree boas were collected by Boos but the snakes were not preserved.
Li ving World Journal of The Trinidad & Tobago Field Naturalists' Club 1993 • 1994 3 been collected in habitats ranging from undisturbed rain Table 1. Summary of meristic characters in Coral/us forest to edificarian situations near Port of Spain. As enydris from Trinidad (upper series of figure) and Tobago elsewhere in its wide range, C . enydris is an edge species, (lower series of figures). usually encountered along bodies of water, road cuts, and Character n Mean+SE Range other habitat margins. Dorsal scale rows 33 41.6:!:.27 38-45 On Tobago, C. enydris has been taken at at midbody 6 42.3 :!: .56 40-44 Speyside, 1.5 km ENE of Charlottesville, ca. 5.0 km W of Ventrals 32 259.3:!: .67 252-268 Charlottesville near Hermitage, near Roxborough and at 6 255.8:!: 1.92 250-264 Hillsborough Dam. Subcaudals 31 107.0:!: .57 100-112 5 107.0:!: 1.70 101-111 Description: The largest example from Trinidad Supralabials 35 10.6:!: .12 10-12 that we have had the opportunity to measure was 1870 6 10.5 :!: .22 10-11 mm snout-vent length (SVL) (FMNH 49918 from San Infralabials 35 13.9:!: .13 12-15 Rafael); the largest specimen recorded from Tobago was 6 13.8:!: 1.17 13-16 1790 mm SVL (USNM) 228018 from Hillsborough No. of scales Dam). A litter of Trinidad C. enydris had SVLs of 415- bordering nasals 18 4.8:!: .15 4-6 452 mm and weighed 18-20 g. In contrast, a very large 4 5.5 :!: .50 5-7 female collected at Hollis Reservoir was 1857 mm SVL No. of scales and weighed 2200 g. Table 1 summarizes meristic between supraoculars 35 5.8:!: .22 3-9 characteristics of C. enydris from Trinidad and Tobago; 6 6.0:!: .68 4-8 there are no statistically significant (t-test; P>.05) Loreal rows 35 2.1 :!: 3.98 2-3 differences between the samples. 6 2.0:!: 0.0 2 On Trinidad and Tobago, the dorsal ground Subloreals 35 3.1 :!: .15 2-6 colour in adults is usually some shade of brown (copper, 6 3.0:!: .89 2-4 bronze, taupe, khaki, tan, yellow-brown, brownish yellow, Loreolabials 35 5.1 :!: .91 4-8 beige). Dorsally, juveniles or adults may (1) be patternless 6 5.0:!: .36 4-6 yellow-brown (20.6% in a sample of 34 snakes from Scales around eye 35 13.8:!: .14 12-15 Trinidad), (2) have a pattern of rhombic shapes that is 6 13.2:!: .75 12-14 usually more conspicuous posteriorly (20.6%), (3) have Dorsal body blotches 4 37.0:!: .71 36-39 the free margin of each dorsal scale edged in dark brown 2 38.5:!: 3.54 36-41 or black (55.9%), or (4) a combination of2 and 3 (2.9%). The ventral ground colour may be off-white, tan, pale Foraging and Diet: According to Mole and Urich creamy yellow, or a striking bright yellow. The venter of (1894), during the day Coral/us enydris will "lie in the the body may be (1) immaculate, or (2) patterned. If slender twigs at the furthest extremities of the thick flecks, smudges, and/or blotches of dark brown or black branches of the tree partially screened by the leaves" and are present they are denser towards the posterior (anterior they are "singularly inconspicuous." At Hollis Reservoir, portions of the venter are frequently devoid of any Boos has observed large C . enydris coiled in tight balls pattern). Subcaudal scales may be (1) immaculate, (2) using several support branches. Nocturnal observations of flecked lightly with dark brown or black, or (3) sometimes C. enydris on Trinidad indicate that they forage at heights almost completely covered with dark brown or black. comparable to tree boas in the West Indies (3-5 m above ground level; Henderson, 1993). They appear to be active Mole and Urich (1894) provided an accurate and foragers, although there is some question if extremely lyrical description of C. enydris: "The adults attain a large examples may not use an ambush strategy. Two length of 7 or 8 feet, and are sometimes of a yellowish specimens from Tobago had, respectively, an unidentified brown colour More often they are of a deep dark bird (snake 1630 mm SVL) and a rodent (Rattus? snake brown ... When the snake is in motion ... , if he can be 1420 mm SVL). The Trinidad sample of the diet is induced to move in sunlight he presents a remarkably comprised of an unidentified bird (snake 815 mm SVL), a beautiful appearance. The dull dark brown seems to todent (probably Rattus, about 25 cm; snake 852 mm change to a rich mosaic, over which shimmers a lovely SVL), ground-dwelling rodents (two Akodon ?urichi, about bluish iridescence as he wends his sinuous way along the 40 cm and 35 cm, respectively; snake 1360 mm SVL), a branches. Each scale is of a dark colour at the extremity mouse opossum (Marmosa robinsoni; snake 1510 mm furthest from its attachment to the skin, but underneath, SVL). In addition to these records, a just-captured snake where they are overlapped by the other scales, they are regurgitated an unidentified bat (Boos, unpubl.) and pale or bright yellow. The ventral scales are dark brown Urich (1933) repotted a "full-grown" mongoose (Herpestes and rich yellow, sometimes punctured with black".
living World Journal of The Trinidad &. Tobago Field Naturalists' Qub 1993 · 1994 4 auropuncrarus) in the stomach of "a full-grown specimen wise planter will take care that the Castabels [sic] found (6 ft. in length)". on his propetty are not molested", and F. W. Urich (1933), Discussion upon discovering a mongoose in the stomach of a C. Based on a suite of characters (meristic, colour enydris, asked the logical question "Could agriculturists be pattern), Henderson (1991), in a preliminary analysis, persuaded not ro slaughter these useful allies?" We would determined that C enydris from Trinidad and Tobago had like ro think that, roday, people will heed the advice of their closest affinities with populations in southern these pioneering and respected naturalists. Central America, northern Colombia (north of the Acknowledgments Cordillera Oriental de Colombia), Isla Margarita, and RWH is indebted ro the personnel involved in northern Venezuela (the Caribbean lowlands north of the providing loans of CoraUus from Trinidad and Tobago: C. Rio Orinoco and the Guiana Shield). The most common J. Cole and P. Damiani (AMNH), Alan Resetar (FMNH), colour pattern that occurs in C. enydris on Trinidad and D. Auth (UF), J. Rosado (MCZ), and A. Wynn and G. Tobago (dorsal ground colour yellowish-brown with the Zug (USNM). His field work in Trinidad was supported by free margin of each dorsal scale edged in dark brown or the Milwaukee Public Museum, the late Albert Schwartz, black) also occurs in notthern Venezuela. The diamond and American Airlines. John C. Murphy and Addison shape element present in the dorsal pattern of some Wynn were excellent sources of information regarding specimens from Trinidad and Tobago is common in C. field work in Trinidad, and Ronnie Hernandez of Simla enydris occurring in Panama, northern Colombia, and Research Station was a helpful, pleasant host. During northern Venezuela. exciting tree boa adventures on Hollis Reservoir we were The diet of C. enydris on Trinidad-Tobago is not, accompanied by K. Caesar, J. Kranz, and K. Kranz. Permits in basic composition, different from mainland ro collect and export C . enydris were issued by the Dept. of populations (it includes birds, bats, and rodents). CoraUus Wildlife on Trinidad, and the Water and Sewage enydris attains a greater size on Trinidad and Tobago than Authority on Trinidad allowed access ro Hollis Reservoir. elsewhere in its extensive range, rivaled only by References specimens from northern Venezuela (known maximum Henderson, R. W. 1988. The kaleidoscopic tree SVL 1770 mm) and Isla Margarita (known maximum boa: Corallus enydris in the West Indies. Lore 38(4):25-30. SVL 1725 mm). Correspondingly, the diet of C. enydris, Henderson, R. W. 1991. Qistribution and on Trinidad at least, includes larger prey items than preliminary interpretation of geographic variation in the elsewhere, with representatives of mammalian orders neotropical tree boa Corallus enydris: a progress report. (Carnivora, Marsupialia) that are not, ro date, represented Bull. Chicago Herperol. Soc. 26(5):105-110. in the mainland sample (n = 44). Despite geographic differences in diet, adult C. enydris throughout their wide Henderson R.W. 1993, Foraging and diet in range prey on rodents. West Indian Corallus enydris (Serpentes: Boidae). J. Herpetol. 27(1): In Grenada, tree boas are common in human disturbed habitat (Henderson, 1988; Henderson and Henderson, R.W. and Winstel, R.A. 1992. Winstel, 1992), and probably occur at higher densities in Activity patterns, temperature relationships, and habitat fruit orchards than in any other kind of habitat utilization in Corallus enydris (Serpentes: Boidae) on (Henderson, 1988; unpub1.). In agricultural areas, tree Grenada: a pilot study. Carib. J. Sci. 28(3-4): boas enhance the production of fruit by reducing the Mole, R.R. 1924. The Trinidad snakes. Proc. number of rodents that are attracted to the fruit and Zool. Soc. London 11:235-278. damage the crop. Corallus enydris occurs in disturbed Mole, R. R. and Urich, F. W. 1894. Biological habitats (including orchards) in Trinidad, and two notes upon some of the Ophidia of Trinidad, B. W. 1., with founding members of the Trinidad and Tobago Field a preliminary list of the species recorded from the Naturalists' Club were so enlightened as ro comment on island.Proc. Zool. Soc. London 1894:499-518. the potential benefit that could be derived from Urich, F. W. 1933. Snake v. mongoose. Trop. encouraging the presence of Corallus enydris in Agricult.l0(1):5. agricultural areas: R. R. Mole (1924) suggested that "A
living World Journal olThe Trinidad & Tobago Field Naturalists' Club 1993 ~ 1994 5 Additional Notes on the Fauna Recorded for the Off, shore Islands North,West of Trinidad Hans E. A. Boos [ and Victor C. Quesnel 2 1. Emperor Valley Zoo, Port of Spain, Trinidad, W.l. 2. P.O. Box 47, Port of Spain, Trinidad.
Introduction carbonaria, and reported a "coral snake" killed in a house A list of the terrestrial reptiles recorded and nearby. When asked if he had any theories about the collected from the small islands off the north-west presence of the tortoise on Monos Island, Tommy said he peninsula of Trinidad was published by Boos (1983). always accepted that they had been brought there by the Additions were noted by Boos (1989). In the latter illegal Grenadian immigrants, who, attracted to war-time publication, mention was made of records of mammals jobs in Trinidad between the years 1939 and 1950, were and amphibians for Monos Island. often landed on Monos Island where they worked as domestics in the island houses before moving to mainland Since the publication of he above, both authors Trinidad to find jobs there. have made additional trips to the islands, and Quesnel, has also uncovered field notes made several years ago. As Gaspar Grande (Gasparee) Island a result, several additional species can now be listed, and To date five lizards, four snakes and one information documented about these unique islands. crocodilian have been recorded for Gasper Grande. Boos Chacachacare Island (1983) suggested other reptiles were likely to be discovered there, listing nine more lizards as probable. In 1962 Underwood reported the collection of Now we record the confirmation of five of the "probable" Hemidactylus brooki (= palaichthus) from the La Tinta Bay lizards, one unexpected one and a snake for which area. No specimens were collected or seen in the positive identification is awaited. intervening years until a juvenile was caught by Boos on 22nd August, 1992 under some loose plyboard on the Lizards veranda of the abandoned dormitory building Mabuya mabouya mabouya cronologically used by the nuns, nurses, and coast guard A note of 4th Aug. 1968 by Quesnel reads as follows: personnel. Photographed and preserved, the specimen of "Seen at Gasparee some time in 1968, date not Hemidactylus palaichthus was lodged in the Museum of remembered". This species has not been seen on more Comparative Zoology in Harvard Massachusetts for recent tips to the island but it may simply have escaped comparison with the specimens collected on Monos notice because of its secretive habits. Island in 1977, 1979 and 1980. This identity has been Hemidactylus mabouia confirmed by the MCZ and it has been registered as One specimen was seen by Quesnel on the gun MCZ-R-I77138. emplacement site on the ridge above and to the east of Also seen on Chacachacare Island were: Pointe Baleine on 14th Oct. 1991. Thecadactylus rapicauda Gonatodes lIittatus First recorded for the island by Quesnel on 29th January, Cnemidophorus lemniscatus 1990 when fifteen were seen on various buildings, trees Gonatodes vittatus and rocks at Bellevue. Six were subsequently seen at Ishmael Samad recorded two separate sightings of the Bellevue on 14th Oct. 1991 and three eggs were collected squirrel Scirus granatensis near the old leprosarium that day at the recreation centre at the gun emplacement hospital. Squirrels have always been popular pets and it is site above Pointe Baleine. These hatched in due course possible that these are the descendants of escapees from and were indeed G. vittatus. the people who once lived here. Plica plica Monos One easily identifiable specimen was seen on a Confirmation that H. palaichthus is still extant on delapidated store room at Bellevue on 14th October, Monos was made on 22nd May, 1992 when one specimen 1991. Three other lizards that may also have been Plica was caught, identified and released in the now abandoned plica were seen nearby but none of these had the usual Villafana House in Grand Fond Bay, and by another adult dark cross-banding, though seemingly the same size and specimen collected and released on 30th November, 1992 form as the common Plica. They were all uniformly tan in a storage shed of the Siegert Bay house, about 350 coloured. Is this due to genetic drift in a small isolated metres from the first collection site. population? (Plica plica on Monos and Huevos Islands are similar to the mainland specimens). Two specimens were The watchman of the Siegert house, Tommy seen at the same locality on 9th April, 1992. Griffith, had as pets two sub-adult female Geochelone
living World Journal of The: Trinidad & Tobago Field Naturalists' Club 1993· 1994 6 Gymnophthalmus underwoodi description would suggest it is Leptophis ahaetulla rather First recorded by Quesnel at Bellevue on 18th June 1990 than Mastigodryas boddaerti, which, on these islands, tends when six were seen scurrying around in fallen dead leaves to conform ro the mainland colour pattern of light brown in a wooded area west of the house. with lighter dorsa-lateral stripes, or light green with scales Ameiva ameiva atrigularis outlined in black giving a netted or reticulated This species was not considered probable by Boos (1983). appearance. It is hoped that a specimen will come to However, a note of 5th July, 1970 by Quesnel reads: hand ro afford a positive identification. "Seen at Bellevue, Gasparee." Other notes of that same References day record the presence of Polychrus marmoratus and Boos, Hans E. A. 1983. A consideration of the Anolis aeneus at the same locality. Unlike Mabuya, terrestrial reptile fauna on some offshore islands north Ameiva is well-known and is conspicuous wherever it west of Trinidad. Living World. Jour. Trin. Tobago Field occurs. It has not been seen at Bellevue during the years Nat. Club, 1983-1984, 19-26. 1980 to the present though frequent trips there have been made. This raises the possibility that it no longer occurs Boos, Hans E. A. 1989. Additions to the on Gasparee, but a thorough search of other likely terrestrial fauna of the offshore islands north-west of localities would have ro be made before one could have Trinidad. Living World. Jour. Trin. Tobago Field Nat. confidence in such a conclusion. Club, 1989-1990, p. 9. Snakes Underwood, G. 1962. Reptiles of the Eastern Another specimen of the Ratonel Snake, Caribbean. Caribbean Affairs (New Series) No.1, Dept. Pseudoboa neuwiedii was collected by residents of the Extra- Mural Studies, U.W.I., Port of Spain, Trinidad. 192 houses on Point Baleine. Recently Fr. Anthony De pp Verteuil has noted a bright green snake which from his
living World Journal of The Trinidad &. Tobago Field Naruralists' Club 1993 . 1994 7 A list of Food Plants and some Implications of the Feeding Behaviour of the Forest Grasshopper, Coscineuta virens (Thunberg), in Trinidad. Lilory D. Me Cornie Entomologist, Ministry of Agriculture, Land and Marine Resources, Central Experiment Station, Centeno. Trinidad.
Abstract Seventy-eight species of plants in thirty-three , , . families are listed as food plants of the small forest ' c • " • . grasshopper Coscineauta lIiTens. Among them are many closely related native and exotic species, including crops, ornamentals, shade plants and weeds. The host range ...... Spoo. A T LA .. fie reflects the insects ability to adapt and respond to a changing environment. , , , Key Words: Coseineuta lIiTens; food plants; feeding , , o c [ . .. behaviour. Introduction Grasshoppers are important by virtue of the damage they cause by feeding (Chapman, 1976). The small polyphagous grasshopper Coseineuta lIiTens (Thunberg), also called the Moruga grasshopper, is an economically important agriculural pest in southeast ,., Trinidad. This paper gives a list of food plants and discusses some implications of its feeding behaviour in a changing environment. Fig I. Distribution of Coscineura ViTens (1985 - 1988) Methods o Hoppers !c, Adults Field observations were made from 1986 to 1989. o Known Oviposition site • Suspected Several reference works (Marshall, 1934; Adams, et ai., 1968; Williams, 1969; Adams, 1972; Purseglove, 1972 & area (Table 2). 1974 and Greesink, et ai., 1981) were used for plant (b) supplemental crops e.g., vegetables such as brassicas identification and other general information. Some and cucurbits; legumes and root crops such as cassava plants were identified at the National Herbarium at the (Table 2). University of the West Indies, St Augustine, Trinidad. (c) plants used as shade for plantation crops e.g., This survey is continuing with the intention of compiling immortelle and banana. a comprehensive list of food plants for the insect in its (d) ornamentals e.g., begonia, ixora, hibiscus. present range. (e) medicinal plants e.g., dite payee, bois canot, black Results sage. The insect inhabits an area which includes three major forest rserves: The Victoria Mayato Forest Reserve, (f) plants of strong cultural and religious significance e.g., calabash and datura. the San Pedro/Poole Forest Reserve and the Southern Watershed. Details of the distribution are given in Fig. I. The insect is an erratic feeder which may damage Large sections of the forests have been cleared for some plants repeatedly; other plants remain untouched agriculture, logging, settlements and industrial during some years and are damaged in others. All stages of the insect show a strong preference for dicotyledonous development. plants, feeding on at least forty-two species in twenty The area is popular for recreational activities seven plant families. Feeding may extend to closely such as hunting and fishing. There is also a Scientific related species in the same genera or several members of Reserve in the Moruga Forest. the same plant families. Both native and introduced Food Plants of Coscineuta virens species may be damaged. While there is some overlap of Feeding was observed on all the plants listed in plant preference among nymphs and adults, nymphs feed Table I. Food plants in southeast Trinidad included: on a wider range of plants. Adults have not been (a) economically important plants e.g., cocoa, coffee, observed to feed on many weeds and shrubs that are citrus and bananas, which are the main cash crops in the commonly eaten by the nymphs.
living World Journal of The Trinidad & Tobago Field Naturalists' Club 1993 . 1994 8 Table 1. A list of Food Plants of Coscineuta virens Family Scientific Name Common Name Acanthaceae Megaskepasma erythrochlamys (Lindau) Agavaceae (Liliaceae) Dracaena sp. Dracaena fragrans Gawl. White Rheo Amarantaceae Amaranthus spp. Spinach, BhagiAnacardiaceae Anacardium occidentale L. Cashew Mangifera indica L. Mango Spondias cythera (Sonn.) Pommecythere, Golden Apple Spondias mombin L. Hog Plum Arecaceae (Palmae) Cocos nucifera L. Coconut palm Bamis major (J acq.) Roseau palm Asteraceae (Compositae) Bidens pilosa L. Railway daisy Eupatorium odoratum L. Christmas bush Lactuca sativa L. Lettuce Begoniaceae Begonia spp. Bignoniaceae Crescentia cujete L. Calabash Tabebuia rosea (Bertol.) DC. Pink poui Bombacaceae Ochroma pyramidale (Cav.) Urb. Bois flot Boraginaceae Cordia collococca L. Manjack, lay lay Cordia curassavica L. Black sage Brassicaceae (Cruciferae) Brassica chinensis L. Pak choi Brassica oleracea capitata L. Cabbage Brassica oleracea verbotritis L. Cauliflower Caesalpiniaceae Cassia fructicosa Mill. Cocrico bush Mora excelsa Benth. Mora Cannaceae Canna edulis (Ker-Gawl.) Canna lily Cucurbitaceae Citrullus lanatus (Thunb.) Watermelon Cucumis sativus L. Cucumber Cucurbita moschata (Dutch.) Pumpkin Fabaceae (Papilionaceae) Cajanus cajan L. Pigeon pea Erythrina glauca Willd. Swamp immortelle Erythrina micropteryx (Poepp.) Mountain immortelle lnga venosa (Gr. ex Benth.) Wild pois doux Moghania strobilifera L. Money bush, Wild hops Vigna unguiculata (L.) Walp. Cow pea, bodi Euphorbiaceae Hevea brasiliensis Meull.-Ang. Rubber Hura crepitans L. Sandbox Manihot esculenta (Cram) Cassava Lamiaceae (Labiatae) Coleus blumei (Benth.) Joseph's Coat Lauraceae Persea americana Mill. Avocado Malvaceae Hibiscus esculentus L. Qchro Hibiscus rosa-sinensis L. Garden hibiscus Hibiscus sabdariffa L. Sorrel Maranthaceae Calathea lutea (Aubl.) G.F.W. Myer Souhari leaf Melastomataceae Miconia officinis DC. Moraceae Cecropia peltata L. Bois canot Artocarpus atilis (Park.) Fosberg Breadfruit Breadnut (chataigne) Musaceae Musa paradisiaca L. * Plantain Musa sapientum L. * Banana Musasp. Moko Strelitziaceae He/iconia bihai Balisier Heliconia spp. Myrtaceae Eugenia malaccensis L. Pomerac Psidium guajava L. Guava
Living World Journal of The Trinidad & Tobago Field Naturalists' Club 1993 - 1994 9 Table 1 continued Family Scientific Name Common Name Piperaceae Piper tuberculatum (Jacq.) Candle Bush Poaceae (Graminae) Axonopus compressus L. Gamelot grass Paspalum fasciculatum (Willd.) Bamboo grass, bull grass Zea mays L. Maize, com Rubiaceae Coffea arabim L. Coffee Hamelia patens Jacq. Ixora chinensis L. Ixora Vanqueria madagascariensis (J.E Gme!.) Chinese tamarind Rutaceae Citrus Umono (L.) Burm. E Lemon Citrus paradisi Macf. Grapefruit Citrus reticulata Blanco Ponugal, Mandarin Citrus sinensis (L.) Osbeck Sweet Orange Scrophulariaceae Capraria bif/ora L. Ditay payee, Goat weed Solanaceae Capsicum annuum L. Sweet pepper Dunalia arborescens L. Wild tobacco Datura mollis L. Datura Lycopersicon esculentum Mil!. Tomato Solanum melongena L. Melongene, egg plant Solanum s!Yamoniifolium (J acq.) Dog teeth, coco chat Sterculiaceae Theobrorna cacao L. Cocoa Tiliaceae Triumfetta lappula L. Burweed, bud bud Verbenaceae Lantana camara L. Grater wood Stachytarpheta jamaicensis (L.) Vah!. Vervine Stachytarpheta cayennensis (L.c. Rich:) Vervine Tectona grandis L. Teak Vitaceae Cissus sicyoides L. Snake vine, Blister bush.
Feeding Damage In the forests, Coscineuta virens feeds in the forest Older nymphs and adults may consume more canopy and on emergent trees at 40-50 metres. Feeding in than 200 mg of plant tissues per individual insect per day. the lower strata tends to be patchy and on areas where Feeding varied from slight in some families to very sunlight filters in. In open areas such as at roadsides and intense on several members of other orders or families where there are breaks. in the canopy, as at logging sites, e.g., Leguminosae (Fabales) - pigeon peas, bodi and wild the insect descends to feed on weeds and other hops, mora; Solanaceae - sweet pepper, melongene; herbaceous species. As hoppers march from forests and Graminae (Poaceae) - maize; Rubiaceae - coffee, ixora; cultivation, they feed on the wayside weeds e.g. mosquito Moraceae bois canot, breadfruit. Many bush, vervain, dite payee and plants in secondary monocotyledonous plants e.g., palms and grasses are successions e.g. bois canot, bois flot and heliconia. barely nibbled. Others such as maize and members of the The insect also feeds on the perennial tree and order Zingerberales - Cannaceae, Marantaceae and shrub crops that predominate in the various mixed Musaceae - may be extensively damaged. cropping systems practised by farmers in southeast Defoliation is the main form of plant damage, Trinidad. These include various combinations of crops though shoots, buds, flowers and fruits may occasionally listed in Table 2, grown among shade trees near the edge be affected. Young nymphs tend to feed on the upper of the forest on mostly small farms ranging in size from epidermis and mesophyll of younger leaves while leaving approximately 2.2. ha to 18.4 ha (average size = 4.9 ± the lower epidermis intact. Older nymphs are more 1.13 ha; n=70). Damage to some imponant crops grown general feeders which skeletonize leaves or devour them by farmers is described below: completely. Adults usually feed selectively on young Citrus - leaves, buds, flowers and fruits may be flush, though buds, flowers and shoots may be damaged severely damaged throughout the year. Leaves may from time to time, as has been seen on mango and citrus. remain on the tree more than a year after damage occurs. Complete defoliation may also occur e.g., breadfruit, Hopper damage to young plants in the dry season can chataigne. Occasionally fruits such as citrus and banana destroy such plants. Older trees may recover from damage may be eaten. to produce lower yield in subsequent years. Damage to Table 2. Distribution of crops grown by farmers in Southeast Trinidad who reported damage by Coscineura uirens to the Ministry of Food Production (1986-1988)
Crop [Common Name] % of Farmers Growing Crops in Combine % Origin of Plant· Moruga Tableland Rio Claro Cocoa 43.75 35.29 33.33 38.01 S. America Citrus 21.88 52.94 52.38 30.86 S. E. Asia Bananas & Plantains 31.25 41.18 38.10 28.00 S. E. Asia Coffee 34.88 47.05 42.86 20.08 Ethiopia Cassava 9.38 9.52 9.52 10.00 T. America Mango 12.25 5.88 7.14 Indo-Burma Pigeon Peas 6.25 5.88 9.52 5.71 Africa Pumpkin/squash 12.25 5.71 C. America Bodi 9.38 5.88 4.43 East Indies Cucumber 9.38 4.29 India Sweet Pepper 9.38 4.29 T. America Dasheen** 9.62 2.86 S. E. Asia Avocado 3.13 4.76 2.86 C. America & Mexico Maize 6.25 2.86 C. America Watermelon 6.25 2.85 Africa Melongene 6.25 2.85 Indian Breadfruit/chataigne 3.13 4.76 2.85 Polynesia Cabbage 6.25 2.86 Europe Ochro 3.13 1.43 T. America Sugar Cane** 5.88 1.43 Asia & Pacific No. oHarmers interviewed 32 17 21 70 * Reproduced from Purse glove 1972 & 1974 ** Feeding damage on these crops not verified buds and new flush can cause stunting and inhibit Mango - Selective feeding was observed in some flowering. Leaf damage in the rainy season may cause fields of mixed varieties. In one field the Blackman shedding of marble-sized fruits if they are present. variety was completely defoliated while Julie remained Defoliation in the dry season may also cause maturing untouched. In pure stand Julie fields feeding may occur fruits to drop. Direct fruit damage, though rare, may also on all the trees. Damage to leaf, shoot and bud during occur. the rainy season produces a pruning effect, to which the Coffee - Coffee leaves are often severely damaged trees respond with heavy vegetative growth. Flowering in the dry season (] anuary - June). The trees often shed can be suppressed for at least the two following seasons, any remaining leaves as a respOI1se to the combined but can be induced sooner by removing excessive shoots effects of feeding damage and water stress owing to the from the bushy crown. lack of irrigation. Young plants may succumb, but Breadfruit & Chataigne - Adults in particular established trees recover to produce new flush with the feed extensively on breadfruit and chataigne which may first rains. Yields may be lower owing to the production of produce new flush within three weeks of complete fewer inflorescence and fewer flowers per inflorescence. defoliation. Cocoa - Only slight foliage damage has been Cassava - Both adults and nymphs feed observed on cocoa; such damage does not seem to extenively on cassava. Leaf regeneration occurs within 2- adversely effect the plant. 3 weeks of defoliation in the rainy season. There is no Bananas and Plantain - These are very prone to leaf regrowth following hopper damage in the dry season extensive defoliation by hoppers. Scarring of the fingers, under the conditions that cassava is grown in Trinidad. though rare, may also occur and reduce the marketability There may be deterioration in the cooking qualities and of the crop. The Sucrier variety is apparently more marketability of the tubers following defoliation. susceptible to attack than other varieties. Farmers claim Vegetables - Many vegetables and legumes are that extensive foliage damage may result in poor yields extremely vulnerable to damage from which they hardly through the production of smaller hands with fewer recover. 100% defoliation has been observed on pigeon fingers. This may be a valid claim but it has not yet been peas, bodi, melongene and sweet peppers. verified. Defoliation of mature plants could cause Discussion abortion of developing bunches (Williams, pers. comm.) Coscineuta .nrens as a general feeder is unlike the many grasshoppers that specialize in certain host
living World Journal of The Trinidad & Tobago Field Naturalists' Club 1993 - 1994 11 categories (Mulkern, 1967: Gangwere et al ., 1989) This decomposition of eggs and eggpods and bodies of the list of food plants, which has been upgraded from the insects. In this way they probably influence biological modest lists of Laurence (1974), COPR (1982), and processes in the ecosystem. Popov, (1986), is impressive, even though food plants in The dramatic effects of feeding damage in the dty the forest have not yet been fully explored. Tables 1 and season might be closely linked to water stress, since field 2 show that the insect is capable of modifying its food irrigation is not practised. For instance, it has been shown choice to include a wide variety of native and introduced that in cassava, water stress associated with the loss of plants in its diet. The various stages of the insect feed on vegetation cover could cause deterioration in tuber different stages of the growth cycle of many plants. quality (Seesahai, pers. comm.). Further studies may show hosts among plants with future economic potential in Trinidad. Ordinarily in agro-ecosystems the practice of mixed-cropping tends to reduce many adverse conditions The feeding behaviour has been influenced by associated with monoculture, which could lead to insect modification of the natural forest through human population build up and insect damage. In multiple activity. Clearing of the forest for many purposes, cropping systems, one crop may be spared from damage by degradation to wasteland, logging and farming practices the more vigorous defoliation of an alternate host such as the establishment of plantations/orchards and (Ruthenberg, 1980). Such an advantage is lost in this shifting cultivation, serve to alter the environment, to situation where the crops cultivated, shade plants and create conditions the insect can exploit. Hunters and other associated vegetation are all hosts. In afforested loggers recall that many years ago C. virens was areas in Trinidad, exotic timber such as teak, grown in encountered only in the deep forest, mainly in areas pure stands to replace native forest trees, may also be where the natural vegetation was characterized as a important hosts. Coscineuta !lirens may well pose a serious predominantly mora association by Beard (1946). Older threat to any agricultural diversification effort in residents attribute what is apparently an expanded range southeast Trinidad which includes inceased production of of the insect to clearing of the forest. The introduction vegetables and ornamentals. There is also the real of exotic crops and other plants, many of which are possibility that with changing patterns of land use, the ;ttrctive to C. virens, to replace the natural vegetation - insect could spread to other parts of the island if it is not described by Beard (1946) - may be the most important properly managed. Clearly it is a situation where, to quote factor contributing to the present pest status of the Bacon (1978), "living things are surprisingly adaptable insect. A comparable situation has been observed in and human activity has produced new habitats for Australia, where clearing of woodlands for pasture has colonization by a species restricted in natue to a particular favoured short, tussocky grasses which provided food for habitat." the Australian Plague Locust (Chapman, 1976). There The insect has been successfully exploiting these are instances where practices such as grazing of sheep and cattle (in Australia) resulted in expansion of habitats new habitats created in the agro-ecosystems and adjacent wastelands of southeast Trinidad, by expanding its range suitable for locust development, especially in the less arid to areas where it had not been known to occur before grasslands that were previously unsuitable for the species. there was extensive clearing of the forests (Mc Cornie et Over-grazing has also produced habitats favouring the aI., 1989). It does this by taking advantage of the new development of the South African brown locust. In feeding opportunities that become available with the Indonesia, deforestation has provided a habitat for planting of tree and shrub crops in place of the native Locusta, where previously none existed (Chapman, climax forest vegetation in its expanding range. These 1976). Elsewhere, in Brazil, infestations of crops, which are grown in an environment that is similar Rhamonawcerus pictus (Bruner) have extended to regions to the forest ecosystem, provide continuous vegetation where forests have practically disappeared, to be replaced cover where the insect finds shelter over a period of by large scale farms (Skaf. 1985). several decades. Since tree crops occupy the same land for Coscineuta virens and other polyphagous many years, they constitute an abundant, uninterrupted grasshoppers, are probably of economic importance both supply of food. Their canopy provides a stable as beneficial insects and as crop pests. It has been environment and shelters undisturbed soil that is suitable suggested that chewing insects such as grasshoppers may as breeding sites. Here the insect completes its life cycle effectively prune plants, resulting in plants with increased successfully; adults mate and oviposit and eggs hatch to vigour and growth (USDA, 1977). Their grazing habit give rise to new generations. may influence plant productivity by producing organic The long-term impact of Coscineuta virens in a litter of faecal and plant origin (Mitchell and Pradt, changing environment in southeast Trinidad needs to be 1974). Organic matter may also be generated from quantified in relation to the impact on both the forest
Living World Joumal olThc Trinidad & Tobago Field Naturalists' Club 1993 . 1994 12 ecosystems and the agro-ecosystems. This might be COPR, (1982). The Grasshopper and Locust difficult to express in economic terms, given that yield Agricultural Manual. Centre for Overseas Pest Research, records so necessary for such an exercise are painfully London. lacking. In a general way, crop loss undoubtedly leads to Gangwere, S. K., Muralirangan M. C. and loss of revenue. There is loss in yield associated with loss Muralirangan M. (1989). Food Selection and Feeding in in photosynthetic leaf area, poor flower set and fruit Acridoids Contributions of the American Entomological production, fruit damage or fruit drop. There are also Institute, 25: 5. undetermined costs associated with rehabilitation of affected fields, an activity requiring time and a drain on Greesink, R., A. J. M. Leeuwenberk, C. E. already limited financial resources. Ridsdale and J. F. Veldkamp (1981).Thonner's Analytical Key to Families of Flowing Plants. Lienden The future management of this insect must Botanical Series, Vol. 5. Lienden University Press, The therefore involve: Hague. [a] an evaluation of the impact of feeding by hoppers in Laurence, G. (1974). Grasshopper Control in the dry season as compared to feeding by adults in the Trinidad and Tobago. Journal of the Agricultural Society rainy season. of Trinidad and Tobago 74: 373-378. [b] a better understanding of its feeding behaviour and the mechanism of food plant selection and Marshall R. C. (1934). Trees of Trinidad and Tobago. Trinidad and Tobago Government Printery, Port [c] pest management strategies that are apptopriate for and compatible with the new environments that have of Spain. been created. Mc Cornie, L. D., A. Ali and P. Siew (1989). Acknowledgements The Control of the Moruga Grasshopper in Trinidad Many people have helped with the field work (March 1985-May 1989). 3rd Annual Seminar on reported in this paper. I wish to express my gratitude to Agricultural Research. NIHERST, Port of Spain. Agricultural Assistants of the Ministry of Food Mitchell, J. E. and R. E. Pfadt (1974). A Role of Production who have worked on the project especially Grasshoppers in a Shortgrass Prairie Ecosystem. Mr. Ahid Ali, Mr Paramdath Siew and Miss Beverly Environmental Entomology, 3: 358-360. Ollivierre-Wilson. Also to Mr. Simon Surajballi and Mr. Mulkern, G. B. (1967). Food Selection in Ramchandar Basdeo of the Locust Survey and Control Grasshoppers. Annual Review of Entomology 12: 59-78. Team. Mrs Yasmin Comeau and Mr Winston Johnson of the National Herbarium assisted with the plant Popov, G. B. (1986). Grasshopper Survey and Control Project in Trinidad and Tobago TCP(TRI/4501E. identifications. The Editorial Committee of the Research Division and Mr. Gordon Laurence critically reviewed Unpublished Report, FAO, Rome. the manuscript. The assistance of Mrs Judy Ruiz is also Purseglove, J. W. (1972). Tropical Crops. gratefully acknowledged. Monocotyledons. Longman, England. References Purseglove, J. E. (1974). Tropical Crops. Adams, C. D. (1972). Flowering Plants of Dicotyledons. Longman, England. Jamaica. University of the West Indies, Jamaica. Rutenberg, H. (1980). Farming Systems in the Adams, C. D., 1. Kasasian and J. Seeyave Ttopics. Clarendon Press, Oxford. (1968). Common Weeds of the West Indies. University Skaf. R. (1985). Report on a Duty Trip from 15 of the West Indies, Trinidad. August-17 September, 1985. Unpublished Report, FAO, Bacon. P. (1978). Flora and Fauna of the Rome. Caribbean: An Introduction to the Ecology of the West USDA (1977). Review of Forage Losses Caused Indies. Key Caribbean Publications, Trinidad. by Rangeland Grasshoppers. Miscellaneous Publication Beard, J. S. (1946). The Natural Vegetation of No. 1348, USDA, Washington. Trinidad. Clarendon Press, Oxford. Williams, R. O. (1969). The Useful and Chapman, R. F. (1976). A Biology of Locusts. Ornamental Plants in Trinidad and Tobago. Trinidad and Edward Arnold, London. Tobago Government Printery, Port of Spain.
Living World Journal of Th~ Trinidad & Tobago Field Naturalisrs' Club 1993 . 1994 13 A Juvenile Butterfly Collector in Trinidad 1933,36 Nicholas Guppy 2la Shawfield Street, London SW3 4BA, England
I lived in Trinidad from my birth in 1925 until ventral side. Your specimen is close to the subspecies M. 1936 when I was brought to England by my mother. deiclama lecerfi Le Moult, from low orinoco, as it appears Between 1933 and 1936 our home was in St Ann's considering the dorsal side". Hans Boos (pers. comm.) Avenue, St Ann's, Port of Spain, opposite the convent, had come to the same conclusion concerning B, citing and it was from there that I did most of my early butterfly the illustration of M. deidama on p. 236 of Smart's and insect collecting. It was a short walk up this street Encyclopedia (1976). into the top of the Royal Botanic Gardens, and most of Prof. Blandin adds "To my knowledge, the the butterflies and other specimens that are in my presence of M. menelaus and M. deidama is new for collection, and which can be seen in the photograph *, science, even if not too much surprising: these two were caught on the site of the present-day Emperor species have a very large range in South America, with Valley Zoo. some geographic variations (a large - too large? - number At least two of these butterflies, both of them of subspecies has been described}." Morphos, do not appear in Bacant's (1970) Butterflies of Both butterflies should therefore be added to the Trinidad and Tobago, and are therefore of some interest list of Trinidad species. They were caught by me at the now. I visited the British Museum (Natural History) in very top of Emperor Valley; near the end of St Ann's South Kensington, London, to try to get them identified, Avenue, in early 1937 or 1936, and I can clearly and was there advised by Mr P.R. Ackerley of the remember their capture. In those days the hillsides above Entomology Department to send photos and the St. Ann's valley were still forested, and many measurements to the leading authority on the genus beautiful butterflies could be seen even along he main Morpho, Professor Patrick Blandin, of the Museum road. Coblenz House had huge sandbox trees (Hura National D'Histoire Naturelle in Paris. crepitans) in its grounds, around the trunks of which flew Prof. Blandin's reply (pers. comm.) indicates that crackers (Hamadryas ieronia) and 88s (Catagramma these two butterflies are new records for the island: maimuna), while further up the valley in cocoa A. Wingspan 135mm. This is alone at the plantations beautiful woodland species such as Pierella bottom left corner of the right-hand section of my lena could be caught. In 1937 we moved to 4 First collecting box in the photo of my .collection. Prof. Avenue, Cascade, a more open area with poor vegetation, Blandin identifies it as: "M . menelaus, not amathonte; A and I gave up collecting insects. clearcut character is the colour of the circle of ocellars on Acknowledgements: ventral side; the colour is a ocher-yellow in M. amathonte, I thank Hans Boos and P.R. Ackerley for helpful a reddish-ocher in M. menelaus, and it seems to be the comments and advice and Prof. Patrick Blandin for case from your photograph. M. menelaus is known from identifying the two butterflies. Venezuela, and Le Moult described a s.sp. onnocensis from References low Orinoco, but the characteristics of this s.sp. are not Barcant, M. 1970. Butterflies of Trinidad and evident." Tobago. Collins, London, 314 pp. B. On the other side of the box can be seen two Smart, P. 1976. The Illustrated Encyclopedia of Morphos (wingspan 132 mm), which Prof Blandin the Butterfly World. Hamlyn, London, 275 pp. describes as "M. deiclamia, as shown by the very typical
* Mr. Guppy submitted with the article two enlarged colour photographs of the collection and colour slides of dorsal & ventral surfaces of the two Morpho species that he is claiming are new records for the island. We have chosen to reproduce the photos of the two Morpho species (see front cover and p2) - Ed.
Living World Journal o(The Trinidad & Tobago Field Naturalists' Club 1993 . 1994 14 Feeding Behaviour of the Sguirrei, Sciurus granatensis Humboldt, on Cocoa Pods in Trinidad W.l. David D. Chadee! and David P. ChadeeZ 1. Insect Vector Control Division, Ministry of Health, 3 Queen Street, St Joseph, Trinidad, W.1. 2. 220 Tableland, Naparima Mayaro Road, Trinidad, W.1.
Sciurus granatensis Humboldt, the red-tailed (e.g. coconuts, berries), bird eggs and nestlings and squirrel, is the only squirrel species known to occur in plantspecies forming major components of its diet Trinidad (Boos 1986). It is distributed from north central (Chadee, pers. observations, Nitikman 1985). Costa Rica to central Ecuador and eastwards from the Table 1. Feeding behaviour of Sciurus granatensis Pacific Ocean through Venezuela north of the Orinoco over 20 days at a cocoa estate located on the River, including the islands of Margarita, Tobago and Naparima Mayaro Road, Trinidad, January to March, Trinidad (Nitikman 1985). This species occupies diverse 1992. ecological habitats ranging from tropical forest, where TIme (hr) No. of feeding visits % populations reach high densities, to crop lands, where it is observed considered a serious pest. Interest in squirrel activity was 04.00-05.00 o o stimulated by the massive destruction of cocoa pods and 05.00-06.00 o o the reduction of yield at a cocoa estate located near the 06.00-07.00 4 4.2 16 3/4 mile POSt on the Naparima Mayaro Road, Trinidad, 07.00-08.00 20 21.1 W.l. This paper reports some observations on the feeding 08.00-09.00 13 13.7 behaviour of the squirrel, Sciurus granatensis, on this 09.00-10.00 4 4.2 estate. 10.00-11.00 2 2.1 11.00-12.00 o o During January to March 1992 squirrel feeding 12.00-13.00 o o was monitored for 20 days. The number of squirrels 13.00-14.00 2 2.1 feeding on cocoa pods was noted every hour from 04.00 14.00-15.00 4 4.2 hours to 19.00 hours. Records for the number of squirrels 15.00-16.00 8 8.4 feeding are given both as the arithmetic and Williams 16.00-17.00 19 20.0 means (Haddow 1960). 17.00-18.00 14 14.7 18.00-19.00 5 5.3 Squirrel feeding occurred on all 20 days of 19.00-20.00 o o observation. Though derived from a small number of Total 95 100.0 observations (Table 1) the pattern of feeding is clear, and the arithmetic and Williams means correspond closely .. ___ WlWAMSMEAN (Fig 1). Feeding was diurnal, between 06.00 and 18.00 ARITHMETIC MEAN hours, and showed a bi-modal pattern with two well defined peaks, one at 06.00-09.00 (39% of feeding .. squirrels) and the other at 15.00-18.00 hours (38% of feeding squirrels). No squirrels were observed feeding after 19.00 hours or before 06.00 hours. The pattern, which occured every day is consistent with observations made by Vesey-FitzGerald (1963). In addition, solitary feeding was also observed, with squirrels maintaining a spatial distance of several metres and actively avoiding each other (Nitikman 1985). Worth et at. (1968) working in the Bush Bush Forest in the Nariva Swamp, Trinidad established trap lines at ground level but did not collect squirrels whereas during the present study both ground and canopy level feeding were observed. It is possible that the squirrels in the Worth et at. (1968) study were "trap shy" or the baits TIME (Hrs) used were not preferred. Figure 1. Feeding pattern of SciUTUS granac.ensis at a cocoa estate In Trinidad, S. granatensis is an omnivorious located on the Naparima Mayaro Road, Trinidad, January to feeder, with insects (e.g. bark beetles and larvae), fruits March, 1992. No squirrels observed between 19.00 and 06.00 hours.
living World Journal of The Trinid:td &. Tobago Field Naturaliru' Club 1993 . 1991 IS It is noteworthy that Vesey-FitzGerald (1936) H. Boos for their invaluable assistance during the course reported on the distructive nature of squirrels to cocoa in of this study and for critically reviewing a draft of the Trinidad and indicated that it was "a usual practice to pay manuscript. a bounty of six (6) cents for each tail" whereas in References Colombia, South America, the damage caused to Boos. H . (1986). Mammals of Trin idad and agricultural crops by squirrels was substantial and a Tobago. Occasional Paper No.1. Zoological Society of bounty of $1.50 per tail was paid (Hernandez-Camacho Trinidad and Tobago. Mimeograph. 1957). Haddow. A.J. (1960). Studies on the biting Based on the time of day when peak feeding habits and medical importance of East African activity has been observed in the field, an effective way to mosquitoes in the genus Aedes. l. Aedimorphus, observe feeding in nature would involve exposing food or Banksinella, and Dunnius. Bull. ent. Res. 50: 759-779. seeking out feeding sites between 06.00-09.00 hours. It is clear that further studies on this activity should be Hernandez-Camacho. J. (1957). Mammalia. conducted, especially as the demand for and price of lnforme pre!iminar sobre aves y mamifereos de cocoa increase on the world market. Therefore, it is quite Saantander, Colombia (J.I. Borrero H. and j. Hernandez possible that many abandoned cocoa estates may be C.,eds). An. Soc. BioI. Bogota, 7: 213-230. rehabiliated if the price of cocoa continues to increase. It Nitikman. L.Z. (1985). Mammalian Species. has been estimated that over 80 kilograms of dry cocoa is Sciurus gmnarensis. Am. Soc. Mamm., 246: 1-8. destroyed per squirrel during the cocoa season which lasts Vesey-FitzGerald. D. (1936) . Trinidad from September to March each year, thus culling of the mammals. Trop. Agric., 13: 161-165. squirrel populations may be an available option. Worth. C.B.• W.G. Downs. T.H.G. Aitken, and Acknowledgements E.8. Tikasingh (1968). Arbovirus studies in Bush Bush The authors thank Messrs. T. Samlalsingh, Brian Forest. Trinidad, w.I., September 1959-December 1964. Samlalsingh and Barry Samlalsingh for valued help in the IV. Vertebrate populations. Amer. j. Trop. Med. Hyg., 17: field. In addition, we thank Dr. j .M. Sutherland and Mr. 269-275.
Living World Journal of The Trinidad & Tobago Field Naturalists' Club 1993 - 1994 16 The Vegetation Surrounding Mud Volcanoes in Trinidad Paul L. Comeau c/o National Herbarium, U.W.I., St. Augustine, Trinidad, W.l.
Introduction Until recently, little attention has been directed towards the natural vegetation associated with mud volcanoes, either worldwide or here in Trinidad. Interest has been focused mainly on the geology of these sites which is reflected in the number of publications dealing with this aspect. In Trinidad, mud volcanoes were mentioned in the literature as early as 1813 when Lavayssee described the ones in Cedros. Kugler (1933, 1938, 1968) made valuable contributions to the science of sedimentary volcanism with particular reference to Trinidad, while Higgins and Saunders (1974) examined all known Trinidad sites both terrestrial and marine with respect to their geophysical and chemical properties. With regard to the vegetation associated with Trinidad mud volcanoes, there is an unpublished study by Sharma (1990) comparing two sites (Devil's Woodyard Figure 1. Map showing location of mud volcanoes in Trinidad {nos. 1-32}. See Table I\I for corresponding names. Adapted and Karamat) and an article by Farrell (1981) on the from Higgins and Saunders. vegetation around Moruga Bouffe. Otherwise, botanical information is scanty (Ramcharan 1978-1979, 1980 refers terrestrial sites have the same surface expression. There is to Bois Neuf, Comeau 1990, 1992 deals with Karamat great variation with respect to vent size and cone and Moruga Bouffe). development. Venting activity varies at each location Amongst the local population, mud volcanoes with occasional violent eruptions taking place (see Table represent something mysterious and even demonic as I). The historical records show that spectacular eruptions indicated by the names given to them, e.g. Devil's take place on average every 24.6 years. Woodyard, Morne Diablo. They have been called However, mud volcanoes do have a number of "solfatare", referring to sulphur (Lavayssee 1813), "salses", physical features in common. Their venting activity is referring to salt (Wall & Sawkins 1860) and "bouffe", a definitely elated to tectonic movement (Higgins & mis-spelling and mispronounciation of the French word Saunders 1974) while their position and alignment are "bouffee" which means puff, gust, blast, whiff (the word associated with anticlinal and fault features (Birchwood "bouffe" means comic). One of Trinidad's mud volcanoes, 1965). The parent bed common to all Trinidad's mud Columbus, serves as a Hindu prayer site where each year volcanoes is deep-water marine clay of Miocene age (22- around Easter the local community gathers to offer 25 mybp). This clay was depOSited on an unstable prayers for the return of the rainy season. submarine slope down which it slipped rapidly to give a The following paper attempts to shed some light low density deposit (Higgins & Saunders 1974). The on the type of natural vegetation that surrounds the mud pressure caused by gravity loading and tectronic volcanoes in south Trinidad and to establish any movement foced the low density mud to migrate to the vegetational patterns that may exist between the various surface via weak spots in the earth's crust (fault lines). sites. The mud's mobility is aided by gas and liquids embedded Background within. There are 25 confirmed active mud volcanoes in Mud volcanoes are composed of sediment Trinidad and all are located south of the Central Range (mainly clay, see Table II), saline water (with high levels (see Fig. 1). In addition, a few occur offshore as of exchangeable sodium), gas (mainly methane) and submarine vents and have occasionally broken the minor oil scum (approximately 40% of the Trinidad surface of coastal waters, e.g. Chatham Island (Arnold sites). The methane is colourless and odourless but bums 1912, Weeks 1929, Wilson & Birchwood 1965). No two with an orange flame. Three of Trinidad's oilfields have
Living World Journal o(The Trinidad & Tobago Field Naturalists' Club 1993 • 1991 17 Table 1 Violent Eruptions at Trinidad's Mud Volcanoes' Table II Textural Analysis of Mud from Selected Volcanoes in Mud Mud Volcano Year Trinidad • Volcano Site % Sand % Silt % Clay No. Devil's Woodyard 13 23 64 {ColumbUS 1906, 1966 Anglais Point 6 25 69 1 L'Envieuse 1860+, 1935 Moruga Bouffe 11 38 51 Galfa Pt. 1946+ Palo Seco 14 32 54 3 Chatham 1911,1928, 1964 • Analysis carried out by BT 300 Class, U.W.I., St. 4 Erin 1934,1940+,1989 (May)O Augustine. 5 Anglais Pt. 1900, 1906, 1960 15 Marac 1965 levels of silt, e.g. Lagon Bouffe. The mud dries quickly 17 Lagon Bouffe 1991' once it is exposed at the surface and has a tendency to 19 Devil's Woodyard 1852, 1888-9, 1906, flake when thinly deposited over older mud but forms 1942, 1969 deeper cracks when more thickly laid down. 23 Piparo 1953, 1968, 1969 The number of exotics, such as pebbles and small 24 Tabaquite 1918,1930 (November) boulders found on the surface and embedded in the mud, 27 (East Coast, varies with each site and mayor may not fingerprint the Mayaro Bay) 1797 28 (Challenger Shoal) 1958 passage of the mud through the earth's strata. Some of the exotics may have been part of the original deposit on the • Based on a total of 28 violent eruptions covering a sea floor. Examples of exotics include rounded and period of 196 years from 1797 or 141 years from 1852. angular fragments of siltstone, sandstone, baked clay, The sites in brackets are submarine. Source: Higgins and chert, lignite and laterite. Saunders (1974). + Source: Wilson and Birchwood (1965). The natural vegetation surrounding Trinidad's o Reported in the local press. mud volcanoes ranges from freshwater swamp in the p Author's observations. southwest (Islote) and east coast (Bois Neuf) to Semi evergreen Seasonal Forest (Beard 1946) in south Trinidad been developed within a thousand metres of active mud from Erin Bouffe to Marac and Evergreen Seasonal Forest volcanoes, Moruga East (near Moruga Bouffe), Moruga dominated by mora (Marshall 1934, Beard 1946) in the West (near Karamat) and Barrackport-Penal (near southeast (Moruga and Lagon Bouffe). Cultivated land Digity). Marac and Coora mud volcanoes have fairly large around the sites includes coconut plantation in the quantities of oil mixed in with the mud. Surface pH at extreme southwest (Columbus), teak forest in the the various sites tends to be alkaline. Southern Watershed Reserve (Coara, Landarf, L'Eau The clay at Trinidad's mud volcanoes is Michel, Marac) and at Brickfield (Tabaquite), sugar-cane predominantly montmorillonite (Higgins & Saunders (Digity, Devil's Woodyard) and bananas (Piparo). 197 4) the formation of which appears to be favoured by Methods the alkaline conditions. The high pH is due mainly to the Twenty mud volcano sites have been examined hydrolysis of sodium carbonate (Brady 1984). Carbonate by the author, 15 of which have been sampled in detail levels are quire substantial at mud volcano sites in (see Table 1II). The main selection criteia for in-depth Trinidad (Higgins & Saunders 1974). analysis were adequate natural vegetation surrounding Surface characteristics the site to give an indication of what the original forest Some of Trinidad's active mud volcanoes (60%) was like and accessability. Five of the 20 sites were almost display a tassik, a term originating in the Moluccas completely encroached upon by agricultural land or teak (Indonesia) for a round clearing in the forest without plantation. Within this group, some vegetational vegetation owing to the abundance of clay (Higgins & sampling was done at Piparo. The five mud volcanoes yet Saunders 1974). The tassik represents the "tip of the to be seen are all in remote areas where access is difficult. iceberg" in relation to the subsurface volume of mud. Once a site was chosen for detailed study a When present, the tassik's shape is highly variable and is vegetation survey of the peripheral forest up to 10 m from the result of frequent venting of the underlying mud the tassik, or the vent, and the tassik (if present) was which is at ambient temperature when it reaches the carried out, listing the species and noting presence, surface. With abundant clay, cones of various heights are abundance, habit and habitat preferences. Plants that formed from a few centimetres up to several metres. could not be identified in the field were collected and Cones tend to be absent when the mud contains higher later determined at the National Herbarium of Trinidad
Living World Joumal of The Trinidad & Tobago Field Naruralisu' Club 1993 - 1994 18 & Tobago with the aid of the Flora (1928) and reference collection. When identifications were complete, the species for each site were arranged initially according to Family (alphabetically) and categorized as tree, shrub, vine, climber, epiphyte, herb. The natural vegetation of mud volcanoes Three hundred and thirteen plant species representing 73 families (41 % of the families known from Trinidad) were recorded in the survey of e Trinidad's mud volcanoes. The most common 0u
~ ~ 0 u c When the species lists from the sampled mud 0 0 ",uo", oO"'u "'0 u <'" '"<",<",u< ",,,,<,,,«u« < X X>CXXX 1934), and Cordia curassavica (Black Sage). Trees and ~ shrubs make up the majority of the common plants at mud volcanoes followed by vines, climbers and + epiphytes. Only two herbaceous ground dwellers are on the list of common species, Cyperus ligularis, a Inn 1 robust sedge usually found in coastal areas, and .£ Oeceoclades maculata, a terrestrial orchid. £ Common species at mud volcanoes, that also o 0 1288 "'_00 are characteristic of the natural forest in the * -'"...... '" x x x X X x Southern Watershed Reserve, include Bravaisia ~g8 :382~ --'" ..... -N_ integerrima (jiggerwood), Hura crepitans (sandbox) and Sabal mauritiaeformis (Carat). These trees are important components of Marshall's (1934) Trichilia Brosimum (Acurel-Moussara) Association and Beard's (1946) Bravaisia Faciation which belong to the same association.
There are no plants that are unique to mud volcano sites. Croton corylifolius, found at Morne Diablo Beach mud volcano, although previously known only from this area in Trinidad, occurs in other West Indian islands and Venezuela where it is found mainly in limestone. The species list for Trinidad's mud volcanoes includes several plants that
living World Journal of The Trinidad & Tobago Field Naturalisrs' Club 1993 . 1994 19 Table IV: Most Impottant Families at Mud Volcano Sites in Table V: Most Common Plants at Trinidad's Mud Volcanoes' Trinidad' No 01 No 01 No 01 I~rtance Family Species (Total: 21) Sites Percent Habit Famil~ Sites Species Freguency alue + Acanthaceae Bravaisia imegerrima II 69 T Leguminosae 15 25 56 96 Araceae Philodendron acutatum 8 50 E Bromeliaceae 18 56 88 14 Bignoniaceae Cresemia sp. 8 50 sT Gramineae/poaceae 14 25 47 86 Compositae/Asteraceae 16 18 51 85 Phryganocydia corymbosa 8 50 C Palmae/Arecaceae 14 7 44 65 Boraginaceae Cordia curassavica 14 88 S Moraceae 14 9 35 58 Bromeliaceae Gravisia aquilega 12 75 tIE Bignoniaceae 14 to 30 54 Cactaceae Hylocereus lemairei 9 56 V Cyperaceae 14 IJ 27 54 Compositae Baccharis trinervis 8 50 S Sapindaceae IJ 9 30 52 (Asteraceae) Pluchea carolinensis II 69 S Euphorbiaceae 12 II 28 51 Cyperaceae Cyperus ligularis to 63 H Orchidaceae IJ II 26 50 Ebenaceae Diospyros inconstans 12 75 sT Rubiaceae 12 II 25 48 Erythroxylaceae Erythroxylum ova tum 9 56 Polypodiaceae 8 14 26 48 S Araceae II 8 24 43 Euphorbiaceae Hura crepitans to 63 T Moraceae Ficus amazonica 9 56 T * Based on 16 sampled sites Nyctaginaceae Pisonia salicifolia 9 56 sT + Importance Value = No. of Sites + No. of Species + Frequency Orchidaceae Oeceoclades maculata 8 50 tH Palmae Bactris major 14 88 sT are considered to be rare or endangered in Trinidad and (Arecaceae) Desmoncus orthacamhos to 63 C Sabal mauritiaeformis to 63 T Tobago (see Table VI). Palo Seco mud volcano has the Sapindaceae Paullinia fuscens 9 56 C most of these species (six) followed by Marne Diablo Vitaceae Cissus sicyoides II 69 V Beach and Bois Neuf (four each). The criteria for * Based on 16 sampled sites. determining what is a threatened plant have been dealt T = tree tIE = terrestrial or epiphytic E = epiphyte with by Adams and Baksh (1981-1982). Three of the V = vine sT = small tree H :: herbaceous C = climber tH = terrestrial herb S ~ shrub species in Table VI are thought to be endemic, Philodendron fendlen (Trinidad), Philodendron krugii and Basanacantha phyllosepala (Trinidad and Tobago). Table VI Rare Plants found at Mud Volcanoes in Trinidad + No. 01 The following geographical areas have been Family Species (Total: 13) Occurences designated in south Trinidad: the Southwest, which Araceae * Philodendron fendleri I includes Columbus, Erin, Anglais Pt. and Palo Seco (all * Philodendron krugii 2 coastal or near-coastal), South Central Coastal, which Bromeliaceae Bromelia plumieri 7 Compositae/Asteraceae Pluchea quitoc I includes Coara, Marne Diablo Beach, Landorf and I.:Eau Euphorbiaceae Croton corylifolius I Michel, South Central Inland, which includes Karamat, Gramineae I Poaceae Lasiacis sorghoidea 3 Rock Dome and Marac, Southest, which includes Moruga Moraceae Ficus gomelleira 4 and Lagon Bouffe (both inland), North Central, which Ficus schumacheri 2 includes Devil's Woodyard and Piparo, and East (Bois Orchidaceae Oncidium lanceanum 1 Polygonaceae Coccoloba cruegeri I Neuf). Rubiaceae * Basanacantha phyllosepala 3 Some mud volcano plants show distinct Sapotaceae Sideroxylon quadriloculare I Tiliaceae Corchorus siliquosus I distribution patterns in relation to the mud volcanoes. The following eight plants occurred in two or more mud + Based on an unpublished Herbarium Checklist. * Endemic volcanoes in the Southwest geographical area: Conocarpus erectus (Combretaceae), Anguira umbrosa East geographical area to Bois Neuf. Only one of these (Cucurbitaceae), Hippomane mancinella (Eupharbiaceae), plants, Stigmaphyllon granadense (Malpighiaceae) has a Paspalum vaginatum (Graminaea), Hiraea reclinata mainly east coast distribution. (Malpighiaceae), Chiococca alba and Genipa americana Higgins and Saunders (1974) place Trinidad's (Rubiaceae) and Fagara pterota (Rutaceae). Picramnia mud volcanoes into three stratigraphic groups, the Nariva pentandra (Simaroubaceae) was the only plant found at Formation (Piparo and Tabaquite) which has mud dating more than one site in the South Central Coastal area. In from the Lower Miocene (22 mybp), the Karamat the Southeast geographical area the following species Formation (Karamat, Rock Dome, Marac, Central Balata occured at both Moruga and Lagon Bouffe: Mikania and Navette) which has younger mud but still dating micrantha (Compositae), Epidendrum fragrans from the Lower Miocene, and the Lower Cruse-Lengua (Orchidaceae), Coccoloba latifolia (Pol ygonaceae), Formation with mud of Mid-Miocene age (12-14 mybp). Nephrolepis biserrata and Polypodium piloselloides This last formation is divided into the following (Polypodiaceae). Isocarpha billbergiana (Compositae) was subgroups: a} Southern Anticline (Columbus, Islote, Erin, the only plant found at more than one site in the North Anglais Pt., Palo Seco, Chagonaray, Coora and Marne Central area while eighteen species were restricted in the
Living World Journal of The Trinidad & Tobago Field Naturalists' Club 1993 . 1994 20 Diablo Beach), b) Los Bajos Fault (Morne Diablo, while the high salinity at Moruga Bouffe, located nearly Landorf and L'Eau Michel). c) those occurring above five km (3 miles) inland from the south coast, attracts Cretaceous material (Moruga and Lagon Bouffe) d) species like Sesullium portulaeastrum (Aizoaceae), an Siparia-Ortoire Syncline (Digity and Devil's Woodyard) edible wild plant commonly known as Seaside Purslane. and e) the Rio Claro Boulder Bed (Cascadoux and Bois Twenty six species with coastal affinity (see Table Neuf). VII) were noted in the vegetation survey of 16 mud The only good correlations between the Higgins volcanoes. These plants form part of the vegetation and Saunders stratigraphic groups and the geographical component of every sampled site ranging from 22 percent areas presented previously are the Southeast sites of the species recorded for Columbus to three percent of (Moruga and Lagon Bouffe), which correspond to those listed for Marac. Only one plant on the list did not Subgroup 3 of the Lower Cruse-Lengua Formation, and occur at an inland location, namely Ipomoea stulonifeTa the East site (Bois Neuf), which corresponds with the Rio (Convolvulaceae) which was found at Anglais Pt. The Claro Boulder Bed. Moruga and Lagon Bouffe of the two species with the highest number of occurrences are Southeast area or Subgroup 3 have less clay in the mud Cyperus ligularis (Cyperaceae) and Diospyros ineonstans (see Table II) and very few exotics (Higgins and Saunders (Ebenaceae) which in combination are present at nearly 1974) which may have some bearing on the composition all the sample sites, the exception being Marac. A few of of plants at these sites. Of the five plants previously the coastal plants are abundant or common at several mentioned as being restricted ro the Southeast mud mud volcanoes. For example, Rhabdadenia biflora volcanoes, three are mainly epiphytic and only Coceoloba (Apocynaceae) is abundant at Karamat, Fimbristylis latifolia shows a preference for sandy or silty soil being eymosa (Cyperaceae) at Devil's Woodyard, Hippomane common in savannas in north Trinidad. Otherwise, it is maneinella (Euphorbiaceae) at Erin and Paspalum difficult to determine a relationship between lIaginatum (Gramineae) at Columbus. At this latter site, stratigraphical groups and vegetation distribution Conoearpus ereetus (Combretaceae) is ranked as being patterns. common and, together with Hippomane, has a similar A few plants found at mud volcanoes show a status at Palo Seco. Hippomane is a poisonour plant with strong preference for dry areas and are restricted in their a toxic latex (Seaforth 1988). distribution to south or southwest Trinidad and the Mud volcanoes surrounded by large tracts of northwest part of the island where the driest conc;litions natural vegetation (Columbus, Erin, Anglais Pt., Palo prevail (Marshall 1934). These include Ditaris polygama Seco, Mome Diablo Beach, Karamat, Moruga and Lagon (Euphorbiaceae), Albizia earibaea and Machaerium Bouffe, Bois Neuf) have a higher average percentage of robinifolium (Leguminosae), Hiraea reclinata coastal plants (13%) than those sites encrouched upon by (Malpighiaceae) and Guettarda famesiana (Rubiaceae). cultivated land (Coora, Landorf, L'Eau Michel, Marac, Based on Herbarium collections and field surveys, a small Rock Dome, Devil's Woodyard, Piparo) where the number of plants recorded at mud volcano sites appear to average is nine percent (Table VII). The percentage for have only a south distribution in Trinidad. Included here Piparo is disproportionately large owing to the limited are Crown eorylifobus, which was mentioned earlier, plus amount of vegetaional sampling done at the site. (This is Mikania cordifolia (Compositae), Eugenia procera also true of the weeds recorded at this site; see Table IX). (Myrtaceae), Oneidium laneeanum (Orchidaceae) and Sites surrounded by teak plantation (Coora, Landorf, Casearia deeandra (Samydaceae). Further collections and L'Eau Michel, Marac), have the fewest coastal plants, six surveys may extend their range as was the case with percent on average, whereas the coastal mud volcanoes Oncidium luridum which was believed to have been from (Erin, Anglais Pt., Mome Diablo Beach) have the most, the south only until it was recently picked up at Bois fourteen pecent on average. Neuf mud volcano and at a hillside quarry in north There is also a tendency for coastal plants at the Trinidad. mud volcanoes in south Trinidad to increase their Geochemical influences on mud volcano vegetation percentage values from east to west sites (see Table VII). Apart from the high clay content found at all the Rainfall decreases as you move from the Evergreen sites, high salinities and high alkalinity of mud volcanoes Seasonal (Mora) Forest around Lagon Bouffe to the influence the type of vegetation that is likely to become Semi-evergreen (Acurel-Moussara) Forest at Erin and established in the vicinity of the sites. Large numbers of Anglais Pt. (Marshall 1934). With this decrease in coastal plants are found at mud volcanoes, even when the precipitation there is a decrease in atmospheric flushing sites are located far from the sea. Devil's Woodyard, for which allows higher salt concentrations to persist in the example, is approximately ISkm (9.S miles) from the mudflow at the volcanoes thus a higher percentage of nearest coastline and has five coastal species present coastal plants.
Living World Journal of The Trinidad & Tobago Field Naturalists' Club 1993 - 1991 21 !t s· ~ ~ ~ f ..3 il, Table VII Coastal Plants Found at Mud Volcanoes in Trinidad ~ ~ .... ~ e ,. 0 w 0 , Q ~ ~ .<>= 't ::0 N Q -= ""~ .~ e .; Jl 0 u ~ c 0 No of ~ e ~ z t E ." = ~ 2 c .£ 0 c ~ g, :s: .~ fl' F.miIy Species (Total: 26) -a ·c ..c 8
N N Another interesting feature to emerge ftom the only the remote sites (Morne Diablo Beach, Landorf, vegetational analysis of the mud volcano sites is the large Karamat, Moruga and Lagon Bouffe) are considered. number of species present that are known to grow in Both L'Eau Michel and Marac are surrounded by limestone areas (see Table VIII). Plants like Erythroxylum teak plantation. L'Eau Michel has a buffer zone of natural ovatum (Erythroxylaceae), Chlorophora tinctoria vegetation but at Marac the teak forest borders right on (Moraceae) and Randia aculeata (Rubiaceae) are found at the tassik. Fire, which frequently occurs in teak stands many of the sampled sites. Mud volcanoes with the during the dry season, will affect the Marac site which is highest number of plants with limestone affinities are covered in oil scum. Weeds, which can tolerate this type L'Eau Michel, Palo Seco and Columbus, while none was of disturbance, will prevail. Marac also has the lowest nored at Rock Dome and one each were recorded for percent of coastal plants. As these plants normally grow Lagon Bouffe and Piparo. All the mud volcanoes occur in where fire is not a factor, they are very vulnerable to this non-limestone areas south of the Central Range where type of disturbance having no propagation modes adapted sandstone is the predominant parent rock type. It seems to frequent burning. A good example is Cyperus ligularis, a likely that the controlling factor here is the slight robust coastal sedge often found at mud volcanoes but alkalinity of the volcano mud which attracts plants that absent from Marac. also prefer the slightly alkaline environment of limestone areas. L'Eau Michel and Devil's Woodyard are subject to frequent visitor use, especially the latter, which may help Another geochemical factor that may influence to explain the large number of weeds at these sites. The the vegetation around mud volcanoes is the presence of weeds at L'Eau Michel are concentrated near the entry oil scum which occurs at about 40 percent of the sites but point onto the tassik, being scanty around the rest of the is most pronounced at Coora and Marac, especially the perimeter. Devil's Woodyard is located in a sugar cane latter where it forms a coating over the surface of the mud region and most of its natural forest border has been which is firm enough to walk on. Two species were found removed by cutting and fire. Livestock are allowed to which are almost exclusive to these two sites, Rauwolfia graze around the perimeter which influences plant growth ligusrrina (Apocynaceae), which also occurs at Moruga near the tassik (Sharma 1990). Bouffe, and Malachra fasciata (Malvaceae), which is also present at L'Eau Michel. What may be more significant Slightly more than 50 percent of the weedy are the families that are absent from these oil bearing species found at mud volcano sites belong to three sites but which are often found at other mud volcanoes, families (Compositae, Gramineae, Leguminosae). The most common genus is (Solanaceae) while namely the Asclepiadaceae, Guttiferae and Rubiaceae but Solanum and Vernonia cinerea (Compositae) especially the Cactaceae (Cacti) and Polypodiaceae Eupatorium odoratum plus Enicostema (Gentianaceae) occur at the (ferns) which may be sensitive to oil pollution. verticillatum largest number of sites. Little information is available about the effect Some weeds are known to have medicinal methane gas has on natural vegetation, but its levels are properties and have been used in folk medicine. Several high at mud volcanoes compared to other gasses being of these occur at the mud volcano sites, for example vented at the sites (Higgins and Saunders 1974). As far as Lantana camara (Verbenaceae), the young shoots of is known, methane has no adverse effect on the which are used to make a tea for the treatment of colds vegetation in the vicinity of the volcanoes but the matter and fever (Seaforth 1988). Lantana is also a poisonous warrants futther investigation. plant, especially the immature berries (Lampe and Human impact on the vegetation at mud volcanoes McCann 1985). in Trinidad. The uniqueness of mud volcano vegetation Weeds are a good indicator of disturbance in the natural environment. In this paper, a weed is defined as a As already stated, mud volcanoes in Trinidad plant that usually grows in waste places (roadsides, have no plant species that are restricted to the sites and abandoned land), pastures or cultivated ground. Weeds are found nowhere else. The plants that do grow at or form a large component of the vegetation at three mud near the active vents, however, have to tolerate rather volcanoes, L'Eau Michel, Marac and Devil's Woodyard unusual adverse physical and chemical conditions. They (see Table IX) where human disturbance is significant. are subject to burial by mud which can damage the The average percent of weeds at these sites is quite high vegetation. Recent eruptions (I 992?) at Lagon Bouffe (31%) compared to that at mud volcanoes surrounded by have killed a section of the peripheral forest. The tracts of natural forest where human impact is low (7%). eruption from two vents came from within the forest itself The average percentage becomes even lower (4%) when and not from the open tassik area. A new tassik is forming also inside the forest at Karamat mud volcano two hundred metres east of the old tassik. Here too, trees
Living World Journal of The Trinidad & Tobago Field Naturalists' Qub 1993 ~ 1994 23 ,... <' Table VIII: Species known to occur in Limestone Areas that are also found in Mud Volcanoes in Trinidad Ji' 0 ~ ., j '"Cl l;i ~ ~ '1: :::0 re, ~ ~ E ," 0 ~ 0: .!.l eX ~ ." ~ E g e :z" ~ e" ~ ~ 2 ';; ~ ,5 .£ 0 c ~ c" ~ .ll ~ ,9- ," No of '8 8 .::; « 0.. ~ :::0 ~ ;:.., :w: :::0 ~ ~" ~ 0.. ell ~ Family Specie. (Total: 20) Occurrences 2- Guzmania Iingulata x 1 ;f Bromeliaceae • Guzmania monoscachia x 1 ..., Burseraceae Bursera simaruba x x x x x 5 "c: Compositae/Asteraceae Eupatorium odoratum x x x x x x 6 !L J1' Erythroxylaceae Erythroxylum oV3tum x x x x x x x x x 9 Euphorbiaceae Croton corylifolius x 1 Gramineae/poaceae Oplismenw hirtellus x x x x x x 6 f Leguminosae ~ M • Acacia famesiana x 1 ~ Moraceae Brosimum alicastrum x 1 Z Chlorophora [inetoria x x x x x x x 7 2 Fiscus maxima x 1 .. Piperaceae Peperomia glabella x 1 a~ Rubiaceae Chiococca alba x x x 3 ~ ~ Randia aculeata x x x x x x x 7 Rutaceae Fagara prerota x x x 3 ~ Samydaceae Casearia sylvestris x 1 Sapindaceae Sapindus saponaria x x 2 Tiliaceae Corchorus siliquosus x 1 Muntingia calabura x x x 3 Verbenaceae Citharexylum fruticosum x x x x 4
No. of sp. with limestone affinities 7 5 4 8 5 6 3 9 4 2 2 1 2 1 5 Total no. of species at each site 32 57 43 68 32 46 45 80 41 58 71 95 54 15 70 limestone affiliates as (%) of total 22 9 9 12 16 13 7 11 10 3 3 1 4 7 7 * M - Mimoseae
Table IX: Weeds Found at Mud Volcanoes in Trinidad
0 E" ~ j Cl ~ ~ '1: l;i ~ '" :::0 8 ~ E 0 ~ eX ~ ~ e c e :z" ... 2 0 ~ .£ E -g ~ M " ';; ,~ No of 0 .ll ~ ~ 0 Family Species (Total: 58) ~ ~ j .E ~ :::0 ~ ;:.., :w: ~ :::0 :::0 ~" ~ £' ell Occurrences Acanthaceae Blechum pyramidatum x x x x 4 Ruellia tuberosa x x x x 4 Amaranthaceae Achyranthes indica x 1 Boraginaceae T ournefortia hirsutissima x 1 CompositaelAsteraceae Bidens cynapiifolia x 1 Brickellia diffusa x 1 Calea solidagnea x 1 Edipta alba x x 2 Emilia sonchifolia x 1 Erechthites hieracifolia x 1 Eupatorium odorarum x x x x x x 6 Synedrella nodiflora x 1 ....N Vernonia cinerea x x x x x x 6 ~ Table IX: Continued 5· ~ ~ ~ 0 e ~ Q ~ l;i ~ ~ .~ '" 't ~ ~ 8 re, 0 u ~ No of 1 eX ~ e c :z ~ ~ e ." .;; e Species (Total: 58) -&, E ~ ... e 2 Family c .!l 0 c ~ 0 i? Occurrences f c ".ll ~ ~ !. 3 (3 Jl « Leguminosae # P • Aeschynomene americana x 1 -p Alysicarpus vaginalis x 1 P Centrosema pubescens x 1 P Desmooium adscendens x 1 P Desmooium canum x 1 P Flemingia strobilifera x x x x 4 M Mimosa pudica x x 2 Malvaceae Malachra alceifolia x x x 3 Malachra fasciata x x x 3 Sida acuta x x x 3 Myrtaceae Psidium guajava x 1 Rubiaceae Borreria ocymoides x 1 Hamelia erecra x 1 HeJrtidiooia ocymifolia x 1 Scrophulariaceae Stemooia duranrifolia x 1 Solanaceae Solanum jamaicense x x x x 4 Solanum scabrum x 1 Solanum srramoniifolium x x x x 4 Solanum torvum x x x x 4 Tiliaceae Corchorus siliguosus x 1 Verbenaceae Lantana camara x 1 Stachytarpheta cayennensis x x 2 Stachytarpheta jamaicensis x x 1 No. of weedy species at each site 3 4 5 3 2 3 1 18 2 3 20 1 6 20 8 7 Total no. of species at each 32 57 43 68 32 46 45 80 41 21 58 71 95 54 15 70 N V> Weeds as a percent (%) of total 9 7 12 4 6 7 2 23 5 14 34 1 6 37 53 10 • P - Papilionatae; • M - Mimoseae are being buried and killed by the mud which flows easily purely to the chance that it was the first to arrive and the when freshly vented. one to survive. This may be the case· as well for Paspalum In addition to burial by mud, the frequent vaginatum (Gramineae) at Columbus which dominates expansion and contraction of the montmorillonite clay the dormant tassik to the exclusion of most other species. causes damage to the plant roots. When thick mud It too is not a "typical" or common mud volcano plant deposits dry quickly, large cracks occur. These will and chance, therefore, may have been a factor in disappear once the rains return. The clay at mud volcano establishment. Both these species can proliferate sites is very compact which retards root penetration. vegetatively by rhizomes which has probably contributed to their abundance at their respective sites. Once The chemical composition of the mud at the dominance is established then succession can be slowed volcanoes makes it quite toxic to vegetation when it first down. reaches the surface. The pH values obtained by Higgins and Saunders (1974) at nine sites ranged from 7.6 to 9.4 This is certainly not the case at Karamat mud (average 8.7). Sharma (1990) also did pH determinations volcano where the large dormant tassik is being quickly for Karamat and Devil's Woodyard. The former had colonized by vegetation. This rapid succession has taken values ranging from 9.0 in the centre of the tassik to 8.0 place since 1990 when Sharma did a detailed vegetation inside the forest while the latter had pH values ranging survey of the site and found little vegetation on the from 8.5 to 7.7 (8.1 at the centre of the tassik). Farrell tassik's surface. In addition to some of the early colonizers (1981) in his study of Moruga Bouffe, obtained a pH already mentioned, another creeper Sarcostemma clausum value of 8.2. These readings give an indication of the (Asclepiadaceae) and Grallisia aquilega (Bromeliaceae) alkaline conditions at various sites. It addition to high are well established around the margin. If the site remains pH, very high salinity levels which approximate sea water dormant for a prolonged period, natural reafforestation are characteristic of the mud volcanoes. In their analysis will take place with trees like Brallaisia integerrima of water samples from eight sites, Higgins and Saunders (Acanthaceae), Cresentia sp. (Bignoniaceae), Diospyros (1974) had salt concentrations ranging from 45,373 ppm inconstans (Ebenaceae), Bactris major and Sabal at Moruga Bouffe to 2,766 ppm at Pipato (average of mauritiaeformis (Palmae) forming part of the forest cover. 17,642 ppm). Farrell's (1981) analysis of a water sample Mud volcanoes, because of their nature and from Moruga Bouffe showed a composition of 0.96% salt. unpredictability, attract attention and can become These high levels will only be maintained near the active vulnerable to exploitation. This has already happened at vents, otherwise precipitation will flush out and dilute Devil's Woodyard while other sites have been impacted these high salt concentrations making the non-venting upon by plantation forestry and agricultural development areas a less toxic envitonment for plant gtowth. The (Digity and Cascadoux). This paper has attempted to increase in the percentage of costal plants at mud show plant composition and vegetational patterns at the volcanoes in southwest Trinidad (see Table VII) where sites and establish the floristic and ecological framework less rainfall occurs helps to substantiate this hypothesis. It upon which other studies can proceed. A number of should also be mentioned that ferns, which are good interesting physiological problems need to be investigated indicators of wetter conditions, are noticeably absent such as the effects of methane on plant growth, whether from the southwestern mud volcano sites. Atmospheric or not some of the terrestrial Bromeliads develop a soil flushing also may reduce the pH levels as these are dependency and the various adaptive mechanisms plants maintained by the hydrolysis of sodium carbonate. Once at mud volcanoes have developed to help them survive hydrolysis takes place and no more sodium carbonate is the toxic conditions. The answers to these problems will added, then flushing will dissipate the bicarbonate and increase our awareness and understanding of the hydroxide ions and lower the pH. uniqueness and importance of mud volcanoes in Trinidad. Once venting ceases at a site then colonization of Acknowledgements the tassik can progress rapidly. Some of the early I thank the following people for their generous colonizers will be sedges like Cypems ligularis, vines and support in the preparation of this article: Winston creepers like Rhabdadenia biflora (Apocynaceae) and Johnson for field and herbarium assistance in the shrubs such as Pluchea carolinensis (Compositae). identification of the numerous plant species, Veejay Sometimes a species will establish itself so successfully it Sharma for the use of his species lists for Devil's will dominate large sections of the dormant tassik. This is Woodyard and Karamat mud volcanoes, Yasmin Comeau the case at Devil's Woodyard where Fimbristylis cymosa for providing working space at the National Hebarium (Cyperaceae) forms a discontinuous mat over large and proof-reading the manuscript and Jacques Boulegue, sections of the surface, but it is the only site where this Aline Dia and Maryse Castrec for giving the final species has been observed, so its presence must be due stimulus to get the article written. Without the kind Living World Journal of The Trinidad & Tobago Field Naturalists' Club 1993 - 1994 26 assistance and encouragement of all these people this Kugler, H.G. (1968). Sedimentary volcanism. paper would not have been completed. A special word of Trans. IV Caribbean Geo!. Conf. Trinidad (1965), p.ll- thanks to Francis Morean for guiding me to some of the 13. sites and helping with the plant identifications at Piparo, Lampe, K.F. and McCann, M.A. (1985). AMA and to Aleisha Khabay for typing up the tables. handbook of poisonous and injurious plants. American References Medical Association, Chicago, IllinOiS, 432 pp. Adams, C.D. and Baksh, Y.S. (1981.1982). Lavayssee, Dauxion J.J. (1813). Voyage aux lies What is an endangered plant? Living World, j. Trinidad de la Trinite, Tobago et Margarita; et en diverses parties & Tobago Field Naturalists' Club, p 9-14. de Venezuela en Amerique du Sud. Paris, 2 te. 8 vo!' Arnold, R. (1912). Note on Mud island, Marshall, R.C. (1934). The physiography and appearing off Chatham Coast, 31.10.1911. Petroleum vegetation of Trinidad and Tobago - A study in plant World, March. ecology, Oxford Forest. Mem. No. 17, 56 pp. Beard, J. S. (1946). The natural vegetation of Ramchanan, E.K. (1980). Flora history of the Trinidad. Oxford Forest. Mem. No. 20, 152 pp. Nariva Swamp, Trinidad. Ph.D. Thesis, Dept. of Birchwood, K.M. (1965). Mud volcanoes in Biological Sciences, U.W.1. St. Augustine, Trinidad, 135 Trinidad. Inst. Petro!' Review. 19/221:164-167. pp. (unpublished). Brady, N.C. (1984). The nature and properties Ramcharan, E.K., Seeberan, G. and Cahdee, D. of soils, 9th ed. MacMillan Pub!. Co., New York. 750 pp. (1978.1979). Ecological observations on Bois Neuf. Comeau, P.L. (1990). Field trip to Karamat Mud Living World, j. Trinidad & Tobago Field Naturalists' Volcano on 4th March 1990. The Field Naturalist, Bul!. Club, p.30· 31. Trinidad & Tobago Field Naturalist Club, 3:3-4. Seaforth, C.E. (1988). Natural products in Comeau, P.L. (1992). A visit to Moruga Bouffe Caribbean folk medicine. The University of the West (31 May 1992). The Field Naturalist Club, Bull. Trinidad Indies, St. Augustine, Trinidad, 140 pp. & Tobago Field Naturalist Club, 3:3-4. Sharma, V.V. (1990). A preliminary study of the Farrell, T.F. (1981). The Moruga Bouffe. vegetation at Devil's Woodyard and Karamat mud Naturalist, SM Publications, Trinidad, 3(4):16-18. volcanoes, Trinidad. Project Report, Dept. of Plant SCience, U.W.I., St. Augustine, Trinidad, p.87 Flora of Trinidad and Tobago (1928 to the (unpublished). present). Ministry of Agriculture, Lands and Food Production. Government Printery, Port of Spain. 3 Wall G.P. and Sawkins, J.G. (1860). Report on Volumes. the geology of Trinidad; or Part 1 of the West Indian Survey. Memoirs of the Geo!. Survey, Longman, Green, Higgins, G.E. and Saunders, J.B. (1974). Mud Longmann and Roberts. volcanoes· their nature and origin. Verhand!. Naturf. Ges Basle. Bd 84 nr. 1 p.lOI-152. Weeks, W.G. (1929). Notes on a new mud volcano in the sea off the south coast of Trinidad. J. Inst. Kugler, H.G. (1933). Contributions to the Petro!' Techno!. Trinidad 15/74:385-392. knowledge of sedimentary volcanism in Trinidad. j. Inst. Petro!' Techno!. Trinidad 19/119:743-760. Wilson, C.C. and Birchwood, K.M. (1965). The Trinidad mud volcano island of 1964. Proc. Geo!. Kugler H.G. (1938). Nature and significance of Soc. London 1626:169-174. sedimentary volcanism. Sci. Petro!', Oxford Univ. Press, p.297-299. Living World Journal of The Trinidad & Tobago Field Naturalists' Club 1993 . 1994 27 Further Records of Birds on Trinidad and Tobago Richard ffrench Buscombe Noake, Chosen Hill, Hucclecote, Gloucester, GL3 2LT, U.K. Introduction. remain north of the equator during the southern summer. This paper presents data on the occurrence and There are January and February records from the status of 35 species on Trinidad and Tobago, including Caribbean coast of Colombia (Hilty & Brown 1986). The three new for Trinidad and eight new for Tobago. The only similar species, P. tenuirostris, has never been information comes from a variety of sources, including recorded in the Atlantic Ocean. records from observers sent directly to me, notes deposited 3. Northern Gannet* Sula bassana. An adult bird at the Asa Wright Nature Centre, tWO from reports was seen within 100 metres at sea just off Speyside, already published elsewhere, and in a few cases my own Tobago on 29 November 1991 by E. Kwater. The direct observations. I should like to thank all contributors observer, who is very familiar with this northern species, for their co-operation and efforts to update the local was able to see clearly the white secondaries and blue ornithological situation. gray bill, which together with the considerably larger size For some time now I have been anxious to see a and yellowish head and nape separate this species from valid system of bird recording established on Trinidad and the common local sulid, the Red-footed Booby, S. sula. Tobago, such as is standard in more developed countries. This is the first record for Tobago and the southern Until this is done, records unsubstantiated by adequate Caribbean area of this species, which breeds in northern field descriptions, uncorroborated by physical evidence Europe and off the northeast coast of N. America, (specimens or photographs) and unsupported by the wintering south to west Africa and the coasts of Florida testimony of experienced and knowledgeable observers, and the northern Gulf of Mexico. must be treated with due caution. For this reason, where 4. Anhinga. Anhinga anhinga. When I lived in the following records have not been adequately supported, Trinidad I looked assiduously but in vain for evidence of I have denoted them by an asterisk beside the species local breeding in this species. It was common at Pointe-a name. This is not to disparage in any way the integrity of Pierre from December to August, but records were sparse any observers, but in the hope of influenCing all those berween August and December, so it seemed likely that interested to present fully documented records in future. I birds returned to me mainland to breed. However, a nest look forward to the implementation of such a recording with two young was found at Pointe-a-Pierre in January scheme by any responsible organisation on Trinidad and 1991 by J.B. Saunders and photographed by Molly Tobago. Gaskin. It was in a casuarina tree beside one of the Species List reservoirs. It will be interesting to see if the cormorants, 1. Cory's Shearwater. Calonectris diomedea. About Phalocrocorax, that have become so common in that area 10 birds of this species were seen flying close to shore off follow the Anhinga's example. Manzanilla beach on 30 December 1991 by tour-leader W. 5. Jabiru Stork* ]abiru mycteria. An adult bird was Petersen along with other members of his party. These are seen at Caroni on 12 May 1990 by R. Ramlal and five the first records for the area for 30 years. other observers; the first record for Trinidad. It has 2. Soory Shearwater. Puffinus griseus. The body of already been recorded on Tobago. a recently dead bird of this species was found on 6. Scarlet Ibis. Eudocimus ruber. Recent studies by Manzanilla beach near Nariva Swamp on 12 December Bildstein (1990) have at last produced a credible and 1991 by Dr. C. Haass, who is thoroughly familiar with the non-controversial theory to account for the species. It was an all-black shearwater with weak whitish abandonment of Trinidad as a nesting ground by this stripes on the underwing. This is the first record for species after the 1960s, namely the need for freshwater Trinidad of this species, which breeds in the sub-antarctic, feeding areas within reach of foraging adults during the migrating north mostly in April-May into the Pacific and breeding cycle. While I was studying the Caroni Atlantic Oceans, returning in August-September. There swamp birds during 1958-1959 I witnessed the gradual are few records from the West Indies, and this record breakdown of the North-South Dyke that had preserved (occurring during the breeding season) is extremely the freshwater marsh to the east of the swamp. Now that interesting, presumably involving a non-breeding it seems unlikely that the dyke can ever be restored, we individual, some of which are known occasionally to can probably expect the ibis to continue breeding on the Living World Journal of The Trinidad &. Tobago Field Naturalists' Club 1993 . 1994 28 mainland, and to visit Trinidad only in the off-season, as pair at Laventille in July 1992, and a female at Speyside, at present. Tobago. The latter was seen and photographed by J. Mc In this connection, some most interesting Williams on 13 August 1992, the first record of the observations have recently been made. On 2 August 1992 species for Tobago; it was found again nearby on the next Dr. J. S. Kenny (pers. comm.) saw a large flock of ibis day. A most interesting record for this normally sedentary flying westwards along the north coast near species. Blanchisseuse. About 2 km offshore, the flock numbered 10. White-tailed Hawk. Buteo albicaudatus. at least 1000 birds, even quite possibly as many as 3000, Owing to some similarity with the much smaller judging ftom the size of the flock as it appeared to the B. brachyurus there has been some confusion over observer on shore; the birds were flying some 50 m above identification of this species, which frequents mainly the waves. In addition, David Rooks told me of a flock of lowland savanna and marsh. An adult was well seen at about 1000 ibis present during August 1991 and again in Nariva Swamp on 12 December 1990 by B. Soderstrom August 1992 at Buccoo Lagoon, Tobago. and two other observers. Previous Tobago records of ibis have been only of II. Savanna Hawk* Heterospizias meridionalis. very small numbers. Since the Trinidad population David Rooks has reported one at Lowlands Estate, Tobago moved its breeding site into the Orinoco delta, those in August 1990. This is the first record for Tobago, and birds that return to Caroni after breeding usually arrive in represents an interesting range extension, to be expected the area about August, with numbers peaking in in a species which regularly exploits cattle-rearing ranch September or October. Flocks have frequently been lands. observed over Trinidad's west coast or the Gulf of Paria, 12. Yellow-headed Caracara. Milvago chimachima. but the above sighting is the first to my knowledge for the David Rooks reports it as a regular visitor to western north coast of Trinidad. It is possible that the same birds Tobago since its discovery there in 1987. I would expect were involved in the Tobago and Blanchisseuse sightings. it to establish permanent residence on Tobago in the near In any case, they form evidence of a seasonal, post future. breeding movement along the South American coast, 13. Crested Caracara. Polyborus plancus. An adult originating probably in the Guianas or even N.E. Brazil. was found at Waller Field Agricultural Station on many If the birds came ftom the Orinoco delta in occasions in 1991/1992, being first seen by J. Ramlal and Venezuela, it is difficult to imagine why they should others on 21 April 1991. There were few records travel up the east coast of Trinidad, rather than westward previously of this rare visitor to Trinidad, which is, along the south coast. If on the other hand they moved however, abundant on Venezuela. along the coast from much farther east, they might well 14. American Kestrel. Falco sparverius . An have flown too far to the north and missed the channel immature male was seen at Grafton Estate and near south of Galeota Point. It is also inconceivable that a Crown Point airport, Tobago on 9/10 February 1991, large flock of ibis could have crossed the mountains of recorded by W.L. Murphy and others. This is the first the Northern Range between the Caroni plain and the record for Tobago. north coast, as birds of this family seldom fly higher than about 100 m above ground and habitually frequent low 15 . Sungrebe* Heliomis [utica. Sight-records of lying areas. At all events, these records add significantly this rarely seen species are occasionally reported, all from to the currently sparse details of our knowledge of the Hollis Reservoir, Valencia. The most recent is of two seasonal movements and dispersal within the region (see birds seen by R. Ramlal and others on 10 May 1990. Frederick et al. 1990). 16. Double-striped Thick-knee. Burhinus 7. Northern Shoveler. Anas clypeata. A female bistriatus. One was seen in a field at Orange Grove on 14 was seen at close quarters at the Port of Spain sewage August 1991 by Graham White. Typically, when ponds on 25 September 1990 by W.L. Murphy and others. approached it tended to run rather than fly. This is only The species is known to hunters but is quite rare locally. the second record for Trinidad of this strange-looking species. 8. Ring-necked Duck. Aythya eollaris. A pair seen at Buccoo marsh on 18/19 January 1992 constitute the 17. Ruff. Philomachus pugnax. Records of this first record of the species for Tobago; they were seen by vagrant shorebird from the Old World continue to be tour-leader D. Finch and members of his party. reported. Recent occurrences include a young bird seen at Port of Spain sewage ponds on 14 December 1990 by B. 9. Hook-billed Kite. Chondrohierax uncinatus. Soderstrom and others, and one photographed by G. More records of this rare raptor have been reported White at Buccoo Marsh, Tobago. recently, including one at Waller Field in April 1992, a Living World Joumal of The Trinidad &. Tobago Field Naturalists' Club 1993 . 1991 29 18. Pomarine Jaeger. StercoTarius pomarinus. A occurrence may well be a result of post-breeding dispersal, bird of this species was seen at Buccoo Reef on 19 January about which so little is known. 1992 by tour-leader D. Finch and others. This is the first 24. Oilbird. Steatornis caripensis. The first Tobago record of this seabird, which has been recorded quarterly bulletin for 1992 of the Club's newsletter, Field elsewhere in the region. Naturalist, contained an account by Hans Boos of a trip 19. Lesser Black-backed Gull. Larus fuscus . An to Hillsborough Dam, Tobago in September 1988. While individual seen at Turtle Beach on 19 January 1992 by D. collecting caiman by night on the reservoir, the party, Finch and others is only the second record for Tobago of comprising H. Boos, G. Gomes, J. Seyjagat and N. Leith, this species, which appears to be expanding its range. observed "foraging Oilbirds on the edge of the jungle 20. Orange-winged Parrot. Amazona amazonica. overhanging the lake". The clicking sounds used by the Two golden-yellow individuals of this species were birds in echo-location were also heard, so it seems observed at Morne Catherine by R. Neckles and R. unlikely that the identification (though made in the Barrow on 30 July 1991 , and one of them again on 9 dark) was erroneous. Oilbirds have not been found on August. Some good photographs were taken, showing Tobago before, nor have they been seen there since clearly the generally yellow plumage, pale primaries, and September 1988 (H. Boos, pers. comm.). Although it is some orange secondaries and tail feathers as in normal tempting to speculate on the possible existence of an plumage; the eyes were red. These were examples of Oilbird colony on Tobago, such a possibility is discounted xanthochroism, a phenomenon allied to albinism, very by the small amount of suitable habitat available on the rarely seen in the wild state* island, and by the complete lack of any record (or even rumour) or such an unusual species, well known to 21. Yellow-billed Cuckoo. Coccyzus americanus. neighbouring rural communities on Trinidad and In a survey of the Lowlands area of Tobago in November elsewhere. Dr. David Snow, the world authority on the 1991 David Rooks found "hundreds" of this species in the species, has commented (pers. comm.) that in his opinion coastal scrub. This was clearly an example of a the birds must have been vagrants, but might well survive spectacular migration "fall" in this passage migrant, on the island for some time. Certainly it is well worth which is rarely seen on our islands except on passage. checking for future OCcurrences. It is known that the 22. Mangrove Cuckoo. Coccyzus minor. Further species occasionally disperses, even migrates, in some recent sightings on Tobago include birds seen at Buccoo numbers from traditional sites, possibly because of marsh on 18/19 January 1992 by rour-leader D. Finch and seasonal variation in food supply or through population his party. It is likely that a small resident population pressures. The Tobago occurrence is the first instance of inhabits suitable areas of southwestern Tobago. the species undertaking a sea crossing of as much as 20 23 . Burrowing Owl. Speotyto cunicularia. A report miles. The bird is of course familiar with the marine in Morpho News Vol 1: No. II for April-June 1992, environment, as several caves on the North Coast of Newsletter of the Zoological Society of Trinidad and Trinidad are well-known. More data on this most Tobago, relates the discovery of this species in March interesting subject are urgently needed. 1992 at Point Lisas Industrial Estate. Further 25. Long-billed Starthroat. Heliomaster investigation in July 1992 showed that a pair of birds had longirostris. More evidence of local breeding, rarely excavated a burrow in an embankment on a building site. recorded, comes from an observation by R . Radix of a Photographs were taken. Unfortunately, industrial female incubating on a nest at the Asa Wright Nature activity at the site resulted in the birds abandoning their Centre on 17 April 1991. burrow. This is the first record for Trinidad of this unique 26. Rufous-shafted Woodstar. Chaetocercus owl, which ranges widely but locally from western North jourdanii. This tiny hummingbird is still seen occasionally America and Florida south to Honduras, throughout most at the Nature Centre, recently on 9 and 14 May 1991. of South America, and in the Bahamas and Hispaniola. Formerly resident in Antigua, St. Kitts and Nevis, it may 27. Variegated Flycatcher. Empidonomus varius. well have fallen victim there to the introduced This species, which is difficult to separate in the field mongoose. In Venezuela it is a fairly common resident of from the Piratic Flycatcher Legatus, has recently been the llanos and is widely distributed in the north of the identified at Simla on 12 January 1991 by tour-leader D. country from Zulia to Monagas and the Delta Amacuro, Finch and others, and by Graham and Alison White and also on Margarita Island. Presumably the Point Lisas birds myself near Rio Claro on 11 July 1991. Trinidad records originated from northeastern Venezuela; no migration in between May and October are likely to involve birds of the South American forms has been recorded, and this the southern migratory race varius, which moves north at * Edward Rooks observed and photographed a xanthochroic Orange - winged Parrot at Toco on 3 Nov. 1985 - Ed living World Journal o(The Trinidad & Tobago Field Naturalists' Club 1993 · 1994 30 Tents and Harems: Alteration of Leaves by Foliage,Roosting Bats Thomas H. Kunz 1 and Gary F. McCracken 2 1. Dept. of Biology, Boston University, Boston Mass. 02215, USA. 2. Dept. of Zoology, University of Tennessee, Knoxville, TN 37849, USA. When one thinks of "tents and harems," what bats in the act of making a tent. The closest that often comes to mind is the Arabian desert and the scientists have come to making such a discovery are harems that have for centuries been kept by wealthy observations that new, partially constructed tents have sultans. Instead, this article is about bats, those which appeared over several days, and that leaves of some live beneath leaves that they modify and use as shelters. occupied tents have undergone minor modifications In Trinidad, and elsewhere throughout the New World overnight. As in all disciplines of science, however, tropics, several species of fruit-eating bats modify leaves repeated observation of the same or similar events does of vines, epiphytes, and palms for their use as so-called provide important supporting evidence on which to "tents" (1-4). The group composition of bats that occupy formulate hypotheses and predictions, and to draw these tents strongly suggests a harem type of social reasonable inferences. organization. Roosting groups often consist of a single During certain times of the year, many of the so male and several females (and their young), although called tent-making bats roost in small groups, consisting others may roost alone or sometimes in all-female groups. of one male and several females sometimes with their Leaves modified into tents are thought to provide shelter young. These groups are referred to as "harems" and may from rain, predators, and sunlight, offering ideal remain together for several weeks or months in the same conditions for rearing young. or different tents. At other times of the year males may In 1932, Thomas Barbour, a naturalist from roost alone beneath such tents while females form groups Harvard University's Museum of Comparative Zoology, exclusively with their young. Sometimes, more than one who had been conducting research in Panama, species may alternately use the same tent, which suggests discovered that some bats modified plants by cutting the that not all bats that roost in tents participate in their veins of palms and banana leaves (5). He noted that construction. these cuts caused parts of the leaves to collapse downward The earliest observations of tent-making bats in forming a partially enclosed space under which the bats Trinidad were reported by two American zoologists, roosted. In the same year, Frank Chapman, a naturalist George Goodwin and Arthur Greenhall (II). who from the American Museum of Natural History who also conducted extensive investigations on bats in Trinidad had been working in Panama, reported observations and Tobago, in association with Greenhall's research on similar to those of Barbour (6). Chapman referred to paralytic rabies. They recorded that the yellow-eared bat, these modified leaves as tents, since they resembled and Uroderma bilobatum, roosted in small clusters of 10 or provided protection similar to man-made tents. In the more individuals "on the underside of the fan-shaped intervening years, several American ecologists, most leaves of certain palm trees," especially the carat palm, notably Robert TImm from the University of Kansas and Sabal mauritiaeformis. The large, native palm is common Anne Brooke from Boston University, made several new in coastal areas of eastern Trinidad, where it may grow to discoveries on these so-called tent-making bats (7-IO). heights up to 12-15 m. Goodwin and Greenhall noted Most of their research was conducted in the lowland that Uroderma constructed tents by making "a series of tropical forests of Costa Rica, supplemented by cuts across the pleated surface of a leaf, causing half of observations made in Peru and Ecuador. the leaf to bend at an angle to form a protected retreat." The collective research efforts of TImm, Brooke, Goodwin and Greenhall found three lactating and others have to date identified at least IS species of females, four non-breeding males, and two non-gravid bats that roost in tents constructed from the leaves of females of U. bilobatum under the fronds of a coconut over 50 species of Neotropical plants. These plants palm tree. They suggested that Heller's broad-nosed bat, include large and small understory palms, cyclanths, Vampyrops helleri, had similar habits to the yellow-eared epiphytic lianes, bananas, small saplings, and epiphytes. bat, Uroderma bilobatum, although no mention was made Other reports of tents and tent-roosting bats include of the type of tent that was made. Among other types of three species of bats from the Old World tropics (TImor, roost, Goodwin and Greenhall found two U. bilobatum Philippines, India, and Indonesia) (2). Although these roosting under the leaves of a carat palm, although they and other bats have been commonly referred to as tent made no explicit reference to "tents". In their account on making bats, no one has actually observed or reported the habits of the pigmy fruit bat, Artibeus cinereus, living World Joumal of The Trinidad & Tobago Field Naturalists' Club 1993· 1994 32 that time. I would suspect the JaQuary record to involve Discussion the resident northern race rufinus, which is distinguished Long-distance migrants. As long as regular by more indistinct streaking on underparts, but which has attention is being paid to the monitoring of unusual not been positively identified on Trinidad. species, especially by professional and highly experienced 28. Great Kiskadee. Pitangus sulphuratus . I was amateur ornithologists leading groups of eco-tourists, it is disturbed, but not surprised, to learn from D. Rooks that likely that records will continue to be made of species nesting birds of this species were recently observed at unknown or extremely rare on Trinidad and Tobago. Louis D'Or, Tobago, several miles south of the point Above are listed five seabirds, three of them new to our where the species was, most regrettably, introduced to islands; there are also three Old World species; and seven Tobago at Speyside by R. Deane and the late E. Lau over North American species, most of them rarely recorded 20 years ago. here, with two new to Tobago. The introduction of this aggressive and successful The South American connection. For historical species is likely to have a detrimental effect on the and political reasons, attention has hitherto been readily indigenous birds of Tobago. I hope that those in authority paid to the ornithological relationship between the will see to it that similar introductions are prevented in islands of Trinidad and Tobago and North America. future. Banding data for migrants has enabled connections to be 29 . Short-tailed Pygmy-Tyrant. Myiornis made and effectively analysed. Faunistically, however, there is a much closer connection between our islands ecaudarus. A second breeding record for Trinidad comes from a pair feeding two newly flying young near Rio and the South American mainland, in particular Claro on 11 July 1991, seen by myself, Graham and Venezuela. Yet little is known about many details of this connection, partly because of inadequate research Alison White. facilities on both sides of the divide. Now it may be 30. Cliff Swallow. Perrochelidon pyrrhonota. hoped that this deficiency may be remedied, both as a Further records of this northern migrant include one seen result of the greater willingness towards co-operation at Kilgwyn Lake, Tobago on 6 December 1990 by B. between our islands and Latin America, and because of Soderstrom with two friends, also one at Buccoo beach recent developments in ornithological circles in both on 7 February 1991 seen by L. Samuelson with three communities. This paper has included details on about others; these are the first records for Tobago. Also one twenty species that move between our islands and the was seen at Port of Spain sewage ponds on 15 January mainland, or in some cases between Trinidad and Tobago. 1992 by D. Finch and his party. It is perhaps not surprising that these include six diurnal 31. Red-breasted Blackbird. Srumella militaris. D. birds of prey, for members of this group are well-known Rooks reports the species is no longer to be found on for their extended territories and their wide range. Less Tobago. If this is the case, one must conclude that the predictable are the range extensions of birds of such few individuals recorded on Tobago between 1974 and diverse taxa as Srearomis, Speotyto, Burhinus, Srumella and 1980 were unable to establish a viable population. Thraupis palmarum. From time to time we must expect 32. Black-throated Blue Warbler* Dendroica records of secretive species such as Heliomis and Coccyzus caerulescens. One was seen at Aripo Heights on 22 March minor along with very small species like Chaetocercus and 1992 by R. Neckles, only the second record for Trinidad. Myiomis which might well be missed by all but the most It is rarely seen as far south. diligent (or lucky!) field worker. 33. Blackburnian Warbler. Dendroica fusca. A References male seen and photographed by R. Neckles at Mount Bildstein, K.L. (1990). Status, Conservation Catherine on 8 March 1992 was the fifth record for and Management of the Scarlet Ibis Eudocimus ruber in Trinidad. The species is a common winter resident in the Caroni Swamp, Trinidad, West Indies. BioI. Conserv. Venezuela. 54: 61-78. 34. Palm Tanager. Thraupis palmarum. This has Frederick, P.C., L. G. Morales, A. L. Spaans apparently established itself in small numbers locally on and C. L. Luthin (Eds.) (1990). The Scarlet Ibis Tobago since its first appearance there in 1982. (Eudocimus ruber). Status, Conservation and recent research. Slimbridge, u.K.: International Waterfowl and 35. Scarlet Tanager. Piranga olivacea. One was Wetlands Research Bureau. seen on Little Tobago in March 1992 by D. Rooks and others of his party; the third record for the island. Hilty, S.L. and W. L. Brown. (1986). A Guide to the Birds of Colombia. Princeton Univ. Press, Princeton, New Jersey. Living World Journal ofThc: Trinidad & Tobago Field Naturalists' Club 1993 . 1994 31 Table 1. Plants used by tent-roosting bats in Trinidad. Plant Family Plant Species Bat Species Localitiesa Source Araceae Philodendron fragrantissiumum Artibeus cinereus Kunz et al. 1993 Philodendron omatum Artibeus cinereus 1,4 Kunz et al. 1993 Philodendron simsii None observed 1,4 Kunz et al. 1993 Philodendron fendlen None observed 1 Kunzetal. 1993 Xanthosoma undipes Artibeus cinereus 2 Kunz et al. 1993 Anthunum jenmanii Artibeus cinereus Kunz et al. 1993 Mesophylla macconnellii 1 Kunz et al. 1993 Cecropiaceae Cecropia peltata Uroderma bilobatum 2 Buchanan, 1969 Heliconiaceae Heliconia sp. Artibeus cinereus 3 Kunz et al. 1993 Musaceae Musasp. not determ ined 2 Buchanan, 1969 Palmae Sabal mauritiaeformis Uroderma bilobatum 7 Goodwin & Greenhall, 1961 Uroderma bilobatum 3 Kunz et al. 1993 Artibeus jamaicensis 3 Kunz et al. 1993 Cocos nucifera Uroderma bilobatum 7 Goodwin & Greenhall, 1961 Artibeus cinereus 6 Kunz et al. 1993 Manicaria saccifera None observed 1 Kunz et al. 1993 "palm" Artibeus cinereus 7 Goodwin & Greenhall, 1961 Mauntia f/exuosa None observed 5 Kunz et al. 1993 Pres toea pubigera None observed 4 Kunz et al. 1993 Pntchardia thurstonii None observed 3 Buchanan, 1969; Kunz et al. 1993 Polygonaceae Coccoloba latifolia A rtibeus cinereus Kunz et al. 1993 aLocalities: (1) Long Stretch Scientific Reserve, (2) Arima Valley, (3) North Manzanilla, (4) Morne Bleu, (5) Aripo Savanna, (6) Manzanilla-Mayaro, (7) Royal Botanical Garden. Goodwin and Greenhall noted that this bat roosted in lesser Trinidadian fruit bat, A. jamaicensis, roosted under small colonies of a few individuals under the cut leaves of leaves, in hollow trees, and in caves, but he made no palm trees and on the underside of banana leaves. No reference to its occupancy or construction of tents. mention was made of any particular palm species or During the mid-1970's and early 1980's published whether the banana leaves had actually been cut. In reports by Robert Timm and his colleagues working in 1962, Greenhall and Goodwin reported a small group of Costa Rica, and similar observations reported by other ten McConnell's yellow-eared bats (Mesophylla investigators, suggested that some of the so-called tent macconnelli) roosting under leaves of a large forest making bats may form stable roosting groups. Intrigued by epiphyte, Anthurium jenmanii, although they made no this idea, we sought support from the National reference to its use for tent construction ( 12). Geographic Society, the Organization of American States, A few years later, Marcus Buchanan (1969), a and our respective universities to begin an investigation former resident director of Simla (the William Beebe on the roosting habits and social organization of tent Tropical Research Station), published a report entitled making bats in Trinidad. We began our field studies in the "Bats of the Arima Valley, Trinidad, W.I." (13), and noted spring of 1984, after conducting a preliminary survey of that "medium-sized bats with white facial markings were potential habitats suggested to us by Professor Julian flushed from cut and partially dried banana, Cecropia, or Kenny of the University of West Indies. Initially, we palm leaves." Although no bats apparently were captured, concentrated our efforts in the Marsh Forest and Palm Buchanan suggested that they were probably Uroderma Marsh habitats in the Aripo Savannas (Long Stretch) bilobatum, Vampyrops helleri, or Artibeus cinereus. In his Scientific Reserve (14), and at two sites along the account of U. bilobatum Buchanan stated that this bat northeastern coast. Later, observations at these sites were roosts in tents made by cutting leaf-ribs causing leaf (sic) supplemented by less frequent visits to a coastal site near to fold, especially palm (Saba! sp.), Cecropia, and banana Blanchisseuse, to Morne Bleu, and to two sites in the (Musa sp.). He suggested that Vampyrops hellen had habits Arima Valley (St. Pat's and Simla). This report similar to U. bilobatum, but occasionally toosted in summarizes some of our findings on tent-making bats in hollow trees and buildings. Buchanan reported that the Trinidad, spanning the period from March 1984 through April 1990 (Table O. Most of our observations were made living World Journal o(The Trinidad & Tobago Field NaturalistS' Club 1993 . 1994 33 Figure 1. A solitary Artibeus cinereus in the Marsh Forest of the Figure 2_ Two apical-shaped tents in Philodendron omatum in Aripo Savanna hanging from the mid-rib of a Philodendron the seasonal Palm Marsh of the Aripo Savanna Scientific simsii leaf that was modified into a simple apical tent. Reserve used both by the pigmy fruit bat, Artibeus cinereus and McConnell's yellow-eared bat, Mesophylla macconnelli. Note from mid-March through mid-September 1984 January that the basal veins are chewed, causing the lobes of the leaves to collapse. and April 1985, January and April 1987, and May 1990. In 1984, while censusing the Marsh Forest and sloshing sounds of our wellingtons, as we waded nearly Palm Forest habitats of the Aripo Savanna for tent calf-deep in water-filled trails. roosting bats, we recorded tents constructed in the leaves In upland habitats along the eastern coast, and in of at least four species of epiphytic lianes (Philodendron forest habitats that had little or no standing water, it was fragrantissimum, P. fendleri, P. ornatum, P. simsii), one possible to approach and capture bats with more success. species of epiphyte (Anthurium jenmanii), and two species Overall, we captured nearly 100 tent-roosting bats, which of palms (Manicaria saccifera and Mauritia flexuosa). We accounts for about half of those we actually observed in observed and lor captured (and released) individuals and tents. Sometimes we were able to capture only a few small groups of three species of bats (Artibeus cinereus, members of roosting groups, as some individuals eluded Uroderma bilobatum and Mesophylla macconnelli) roosting our attempts to capture them. On other occasions we in tents (Figure 1). Our observation in April of 1984 that either flushed bats prematurely as we approached a roost Mesophylla macconnelli roosted in tents was consistent or they roosted too high to be captured with our hand with a report published in that same year by Juliane nets. Koepcke (15), a German ecologist who found small In the Marsh Forest the heart-shaped leaves of groups of M. macconnelli roosting beneath cut leaves of Philodendron spp. were the ones most commonly used by Anthurium sp. and a small, understory palm (Geonoma bats for tent construction (Figure 2). Tents in sp.) in the lowland rain forest of Peru. Philodendron typically were constructed 2 or 3 m above In the Marsh Forest habitat we marked (using the ground, where they were usually protected from direct orange plastic tags) and recorded the location of each sunlight during most of the day. Most leaves that had tent, the number of modified leaves on a particular plant, been modified into tents were oriented so that the space the height of tents above the ground, the size of the directly below the bats was usually free from branches and leaves, the number of chewed veins, the shape of each vines so as to allow the bats to fly in and out of a tent tent, and the relative vertical position of each tent in the without being impeded by nearby vegetation. Typically, understory. We also recorded ambient temperature and tents were made from Philodendron leaves when bats humidity from above and below tents to characterize chewed several of the basal veins, and sometimes the roost microclimates. We recorded the presence or absence mid-rib, apptoximately one-third of the way ftom the leaf of bats, and censused the number of bats present (visually tip. This caused the lateral lobes of the leaves to dtoop or by direct capture) in most occupied tents. On several downward. If the mid-rib was chewed, the distal tip of the occasions bats either escaped our capture efforts or flew blade also drooped downward, forming a semi-enclosed, from tents before we were close enough to capture them. apically-shaped structure_ There was some variation in Captured bats were marked on the forearm with small, how heart-shaped leaves were modified by bats; some plastic, numbered bands for later identification, weighed, tents were formed when each of the lateral veins were measured, and released at the site of capture. Censusing chewed, either at an angle or parallel to the midrib. bats in the Marsh Forest was especially difficult during the rainy season because of the noise created from the Living World Journal of The Trinidad & Tobago Field Naturalists' Club 1993 . 1994 34 Figure 3. Bifid leaves of immature coconut palms (Cocos nucifera), along the Manzanilla-Mayaro coast that were modified into tents by the pigmy fruit bat. Arabeus cinereus. The Palm Marsh in the Aripo Savanna was dominated by large stands of the palm Maurita flexuosa. interspersed by expanses of sedge savanna. Mature trees of this species often reach 30 m or more and. because of Figure 4. Leaf of the palm Sabal mauritiaeformis that was their height above the ground. we could not see bats modified into an umbrella tent (near North Mamanilla). Bats from Trinidad known to conStruct and occupy this type of tent roosting in these leaves. Although we observed umbrella include Peter's tent-making bat (Uroderma bilobatum) and the shaped tents in several of the large leaves of young Lesser Trinidadian fruit bat (Arobeus jamaicensis). understory Mauritia palms, we did not observe bats roosting in them. Tents made from leaves of these adult male and one or more females (sometimes with their immature Mauritia trees often exceeded heights of 5 m young). above the ground, thus making it nearly impossible to In the relatively dry, upland areas on the capture roosting bats. The chewed areas in the crowns of northeast coast of Trinidad (near North Manzanilla), we these leaves were roughly spade-shaped. The long, distal observed numerous tents in the large palmate palm, SabaJ segments of the leaf drooped downward at an angle, mauritiaeformis (Figure 4). Several of these tents were forming a rather airy umbrella-like enclosure. What is occupied by small groups and singles of Uroderma intriguing about bats that construct tents in these large bilobatum and Artibeus jamaicensis. The composition of palmate leaves is that they are able to chew and sever the these roosting groups, which consisted of several females very tough and fibrous rachis and veins. When the tents (sometimes with young) and a single male, suggests a are made and how long it takes a bat to complete such a harem type of social organization. The umbrella-shaped tent or whether several bats participate in their tents in Sabal were made when bats created a roughly construction is unknown. heart-shaped or circular shaped cut around the areas Along the north and east coastal regions of where the petiole is attached to the blade. The distal parts Trinidad we commonly observed tents and tent-roosting of the blade collapsed downward at an angle, as in bats in the broad, bifid leaves of immature coconut palms Mauritia flexuosa, forming a semi-enclosed roosting space, (Cocos nucifera). Several of these tents were occupied by resembling a partially opened umbrella. Because the singles or small groups of Artibeus cinereus. Tents in young leaves of Sabal are less deeply dissected, as compared to coconut palm leaves and other bifid-leaved palms are Mauritia, the resultant enclosure is darker and less exposed formed when two, roughly "J -shaped" cuts are made, one to wind, rain, and direct sunlight. Judging from the tooth in each segment of the bifid blade (Figure 3). This causes marks on these and other palm leaves, it appears that the the two lobes to collapse downward, forming a pyramid veins are chewed from above and the plications are shaped enclosure. Typically, tents constructed in young chewed from below. We also observed similar types of Cocos leaves ranged from 1 to 2.5 m above the ground. In tents constructed from leaves of the palmate palm, the Marsh Forest of the Aripo Savanna, we observed a Coccothrynax barbadensis, some of which were occupied by similar rype of tent cut in the understory palm Manicaria singles and small roosting groups of Uroderma bilobatum saccifera, and these tents were at least two meters above and Artibeus jamaicensis. The shape of the cut and the the ground. Roosting groups of Artibeus cinereus ranged resultant tent architecture were similar to those observed from 2 to 6 individuals, and usually consisted of a single living World Journal o(The Trinidad & Tobago Fic:ld Naturalists' Club 1993 . 1994 35 tenuously attached to the severely burned but standing host trees, all had suffered extensive heat and/or direct fire damage. Although they were badly scorched, we did find a few of the orange tags that we had used to mark the location of tents when we began our study. Intense ground fires were still burning in the Marsh Forest when we left the island three weeks later. Obviously, we were extremely disheartened and saddened that we could not continue the research that we had begun at this site three years earlier. The fire damage to the Marsh Forest and surrounding habitats was so severe that recovery to its former condition will most likely take decades. Figure 5. Patch of wild tannia (Xanthosoma undipes) in the Arima Valley, that is used by the pigmy fruit bat (Artibeus The Marsh Forest was not the only site in cinereus) for tent· making. One of the authors (Kunz) can be Trinidad that suffered from extensive fire damage in 1987. seen trying to capture bats roosting beneath a tent. This type of This had been one of the driest years on record, and tent is formed when bats chew the basal veins of simple heart shaped leaves, causing the basal lobes of a leaf to droop scattered fires damaged crops and razed valuable forest downward. land, and destroyed many human dwellings, especially those in rural areas. Brush and forest fires throughout the in Sabal mauritiaeformis, except that the cut areas Nonhern Range had done severe damage to primary forest generally were rounded in shape. habitats, including those on El Tucuche. One of our In the moist, rocky ravines in the Arima Valley, primary study sites in North Manzanilla also suffered we observed tents and tent-roosting bats in the large extensive fire damage, where a primary stand of Sabal heart-shaped leaves of Xanthosoma undipes (Figure 5). mauritiaeformis was almost completely destroyed. The heat This succulent plant commonly grows in shaded, moist from ground fires destroyed several of the large trees by areas in what appears to be several separate plants all deeply burning the roots, many of which had already been connected by thick, ground-hugging rhizomes. Within a reduced to red-hot embers by the time of our arrival. It single patch we found upwards of 15 tents, sometimes was obvious from the severely damaged leaves that most, with 3 to 4 leaves on a "single plant" having been if not all of the tents that were present at the time of the modified into tents. These tents were similar to those we initial fire had been destroyed; presumably all of the bats observed in Philodendron sp., in which several of the basal had either abandoned the site or were killed from the veins were chewed and the lobes of the leaves dropped intense heat. downward. The leaves selected for tents in Xanthosoma In May 1990 we again returned to Trinidad and ranged in height from 1 to 2 m above the ground. surveyed the two primary study sites that had sustained Typically, the bats chose large leaves that were oriented severe fire damage in 1987. It was immediately obvious horizontally. Given the foul-smelling (and foul-tasting) that what we had judged to be extensive damage in 1987 latex-like sap that is produced when the veins of leaves was worsened by additional fires in 1988. Many of the from Xanthosoma are severed, it is unclear why bats select large trees that were still standing in 1987, and that we leaves of this plant for tent-making. The only bats that assumed might survive, were totally destroyed in 1988. we either observed or captured in or near Xanthosoma The deep, intense, slow burning ground fires in the tents were Artibeus cinereus. hummocky Marsh Forest of the Aripo Savanna had In mid-April 1987 when we returned to Trinidad caused such severe damage that short-term recovery to continue our studies on the bats and tents that we had seemed most unlikely. marked on previous visits, we learned to our utter dismay Regrettably, what remained of the Marsh Forest in that two of our primary study sites (Marsh Forest and a May of 1990 consisted of burned trunks of formerly coastal site near North Mamanilla) had suffered from enormous palms and the decaying litter left from fallen extensive fire damage. When we first arrived at the Marsh trees. Although the Marsh Forest was again becoming Forest site, it was clear to us that few if any of the tents green from the growth of rapidly invading pioneer plants, had survived the intense fire and heat that had swept the former character of the forest, including its unique through it only days before. Most of the large palms and flora and fauna, had been destroyed. It is clear that the understory trees that had supponed the epiphytic lianes destructive fires in 1987 and 1988 have radically changed (Philodendron spp.) and a large epiphyte (Anthurium) had a very significant pan of the Trinidadian landscape. Many been burned extensively and were no longer standing. of the unique plants, animals, fungi, and microbes that Among the few lianes and epiphytes that remained living World Jou rnal oi Thc Trinidad & Tobago Field Naturalists' Club 1993·1994 36 made up this forest community are today little more than architecture and tent-making behaviour among a memory. Neotropical and Paleotropical species. (in press). Belatedly, there is a hard lesson to be learned 3. Tunm, R.M. 1987. Tent construction by bats from all of this. Many, if not most of the habitats in of the genera Artibeus and Uroderma. pp. 187-212, in tropical and other regions of this planet, harbour unique, Studies in Neotropical Mammalogy: essays in Honor of fragile communities of interacting organisms, each of Philip Hershrovitz (B.D. Patterson & R.M. TImm, eds), which is intimately linked to the soil that supports it. Fieldiana: Zoology, New Series, No. 39. Inadvertent or intentional acts which lead to the burning 4. Tunm, R. M. & B. L. Clauson. 1990. A roof of even the most seemingly simple habitats, adversely over their feet. Natural History (New York), 3/90: 55-58. affect the survival of the natural communities of which bats are a part. Although wind-blown surface fires in 5. Barbour, T. 1932. A peculiar roosting habit of many parts of the world play an important ecological role bats. Quarterly Review of Biology, 7:307-312. in maintaining the character of grassland, chaparral, and 6. Chapman, F.M. 1932. A home making bat. coniferous forest communities, the deep, slow burning Natural History (New York), 32:555. fires that ravaged many parts of Trinidad in 1987 and 7. Foster, M.S. & R.M. Timm. 1976. Tent 1988, especially the Marsh Forest, have caused a loss of making by Artibeus jamaicensis (Chiroptera: biological diversity from which recovery will take Phyllostomidae) with comments on plants used by bats for decades, if not centuries. tents. Biotropica, 8:265-269. Acknowledgments 8. Choe, J.C. & R.M. Timm. 1985. Roosting We are grateful to the National Geographic site selection by Artibeus watsoni (Chiroptera: SOCiety and the Organization for American States who Phyllostomidae) in Anthurium rallenni (Araceae) in Costa were the principal sources of funding for this study. We Rica. Journal of Tropical Ecology, 1:241-247. also are grateful for the financial support of Bosron 9. Brooke, A.P. 1987. Tent construction and University and the American Philosophical Society. social organization in Vampyressa nymphaea (Chiroptera: Numerous individuals either assisted us in the field or Phyllostomidae). Journal of Tropical Ecology, 8:1-5. supported our efforts in other ways. For field assistance we thank Hugh Britten, Al Kurta, Beth Nemec, Dixie 10. Brooke, A. P. 1990. Tent lSelection, roosting Pierson, Bill Rainey, Simon Robson, Marty Fujita and ecology and social organization of the tent-making bat, Anne Brooke. We are especially grateful to Julian (Jake) Ectophylla alba, in Costa Rica. Journal of Zoology Kenny (University of West Indies) who suggested (London},221:11-19. potential study sites and offered gracious West Indian 11. Goodwin, G.G. & A.M. Greenhall. 1961. A hospitality. Edward Rooks accompanied us in the field, review of bats of Trinidad and Tobago: descriptiOns, rabies made sketches of plants and tents, and was helpful in infections, and ecology. Bulletin of the American Museum arranging for our accommodation. We also thank Ian of Natural History, 122:187-302. Lambie, Rita Iton, Edward Rooks, and Francis Morean of 12. Goodwin, G.G & A.M. Greenhall. 1962. the Asa Wright Nature Centre for their support and Two new bats from Trinidad, with comments on the status assistance during our stay at Simla. Carol W. James and of the genus Mesophylla. American Museum Novitates, Beesham Ramdial (Wildlife Section, Forestry Division, 2080:1-18. Min. of Agriculture, Land and Marine Resources) were instrumental in arranging for collecting permits. Thomas 13. Buchanan, G.M. 1969. Bats of the Arima Croat, Richard Howard, and Robert Read kindly Valley, Trinidad, W.l. Asa Wright Nature Centre, identified or verified voucher specimens of plants which Trinidad, West Indies. the bats modified into tents. April Allgaier, and Ruth 14Anonymous. 1982. Management and Utzurrum kindly read and made helpful comments on an development plan for the Aripo Savannas Scientific earlier version of this paper. Reserve. Technical Document Forestry Division/OAS References Project, Port-of-Spain, Trinidad, 43 pp. 1. Kunz, T.H. 1982. Roosting ecology of bats. 15. Koepcke, J. 1984. "Blattzelte" als pp. I-50, in Ecology of bats (T.H. Kuru, ed.). New York: Schlafplatze der Fledermaus Ectophylla macconnelli. Plenum. Saugetierkundliche Mitteillungen, 31: 123-126. 2. Kuru, T.H., M.S. Fujita, A.P. Brooke & G.F. McCracken. 1993. Bats and tents: convergence in tent Living Wodd Journal of The Trinidad &. Tobago Held NaturaJi.sts' aub 1993 . 1m 37 Crane Flies (Diptera: Tipulidae) in Trinidad Caves Johanna P.E.C. Darlington 1 and Jon K. Gelhaus2 1. Zoology Department, University of the West Indies, St. Augustine, Trinidad, West Indies. 2. Dep. of Entomology, Academy of Natural Sciences, 1900 Benjamin Franklyn Parkway, Philadelphia, PA 19103-1195, U.S.A. Abstract Table 1. List of species of crane flies recorded from Seven species of crane flies were collected in Trinidad (compiled mostly from Alexander and three Trinidad caves (Oropouche, Aripo Main and Aripo Alexander 1970). Soho). Four of them represent new generic records for Order DIPTERA Trinidad (Cryptolabis, Gnophomyia, Helius, Teucholabis) , Sub-order Nematocera and one species, Erioptera (Mesocyphona) troglodyta Family TIpulidae Edwards, is recorded for he first time since its description in 1918. Species in the genera Crypwlabis and Helius Sub-family TIpulinae appear to be true troglobites, with the remaining crane fly Brachypremna dispeUens (Walker) 1861 species probably troglophiles or trogloxenes. Megistocera longipennis (Macquart) 1838 Introduction Tripula (Microtipula) trinidadensis (Alexander) 1912 Crane flies (family Tipulidae) are among the Tipula (Microtipula) trinitatis Alexander 1941 more primitive of the Diptera (two-winged or true flies). They have long slender bodies and wings, and often have Sub-family Limoniinae a slow, blundering flight. Most characteristic are their Tribe Limoniini elongated legs (giving rise to their common name of Limonia (Geranomyia) lycaon Alexander 1952 "daddy-long-legs") which are usually held outstretched Limonia (Geranomyia) plumbeipleura (Alexander) 1916 when in flight. Tribe Eriopterini Crane flies are the largest family of the true flies Erioptera (Mesocyphona) troglodyta Edwards 1918 with over 3000 species described from Latin America and the Caribbean alone (Alexander and Alexander 1970). Erioptera (Mesocyphona) withycombei Alexander 1929 However, the crane flies of Trinidad and Tobago are Gonomyia (Lipophleps) extensa Alexander 1914 almost unknown. Only ten species have been recorded Neognophomyia trinitatis Alexander 1927 from Trinidad (Table 1) and none from Tobago, although 75-100 species or moremight be expected to inhabit the friends; the identifications of the cane flies and life islands. Adult crane flies typically favour moist shady history infonnation were provided by the second author sites, including the mouths of caves. The larvae may be (JKG). We thank Margot Livingston for the excellent terrestrial, living in soil, leaf litter, rotting wood, etc. or line drawings of crane fly wing venation. aquatic, living in swamps, ponds or the edges and bottoms Methods of streams (Alexander and Byers, 1981; Byers, 1981). The immature stages of less than a dozen species of Samples were taken in six caves (Oropouche, Neotropical crane flies are described (Gelhaus and Young, Aripo Main, Aripo Soho, Guanapo, Tamana Main and 1991) and very little is known about their habits. Tamana Dry caves) with a Hausherr portable light trap powered by a 6V powerpack or by four torch batteries. One of the few species previously known from The trap was suspended at a height between two and five Trinidad was collected in Oropouche Cave by F. W. Urich feet above the floor and run for approximately one hour. and described as Erioptera troglodyta by F. W. Edwards The manufacturer's original gauze cage was used for one (1918). However, not a single crane fly was found among sample (Aripo Main Cave near the entrance) but proved the tens of thousands of flies that were collected in most unsatisfactory because some flies were small enough Tamana caves in the Central Range during the 1960s and to wriggle through the mesh, and many others became 1980s (JPECD, unpubl.). Samples of flies collected in entangled and were damaged. In all the other samples the several caves in the Northern Range in 1989-1991 did flies were collected in a dish of water (with a dash of include small numbers of crane flies. Prelininary results detergent) suspended beneath the trap (Jennings and are presented in this paper. Darlington, 1990). The water was later poured off The samples were collected by the first author through plankton mesh and the catch was preserved in (JPECD), with the help of many field natualists and other 70% alcohol. This method is good for rubust species but is not recommended for delicate flies like crane flies. Living World Journal of The Trinidad & Tobago Field Naturalists' Club 1993 . 1994 38 Table 2. Summary of distribution data for crane flies collected in Trinidad caves in 1989-1991. Sub-family Limonoiinae Cave Location OR AM AD AS Tribe Limoniinae Helius sp. x x Umonia (Rhipidia) sp.A X Umonia(Rhipidia) sp.B X Tribe Eriopterini Crypwlabis sp. X Erioptera (Mesocyphona) troglodyta Edwards X X Gnophomyia sp. X X Teucholabis (Teucholabis) sp. X Location Key: OR = Oropouche orCumaca Cave (1989)· AM = Aripo Main Cave (entrance) (1989,1991) AD = Aripo Main Cave (deep) (1990) • AS = Aripo Soho Cave (1990,1991) Many of the species belong to large Neotopical habitat represented here. The second species, represented genera that are in need of revision and lack keys for by one male and two females, was rather more robust and identification. With one exception, these cave species had distinctive dark blotches on the wings (Fig. 2). This can be identified only to genus at the present time, but was identified as Teucholabis (Teucholabis) sp. The most probably represent undescribed species. All the subgenus contains about 175 species in the Neotropics specimens are deposited for reference in the collection of . but none has been previously recorded from Trinidad. the Academy of Natural Sciences of Philadelphia. The larvae of the North American species occur under Results the bark of recently fallen logs (Alexander, 1920; Rogers, All of the six caves sampled yielded large 1933). numbers of flies but crane flies were found in only three Aripo Main Cave of them, the Oropouche, Aripo Main and Aripo Soho The Aripo Main Cave, the largest known cave in Caves. All three caves are in forested valleys on the south Trinidad, is also a linear cave but with the stream passage side of the Northern Range. They are solution caves in entering it instead of emerging from it. The outer part of limestone, two of them (Oropouche and Aripo Main) the cave is normally dry, but a stream wells up through containing streams while the third (Aripo Soho) is dry. the .floor some 1500 ft in from the mouth. The cave was All three caves contain nests of the oilbird Steatornis carefully surveyed by A.E. Gunther and others who found their greatest practicable penetration to be 2800 ft (860 caripensis (Humbolt) and floor deposits conSisting of bird m) from, and 475 ft (150 m) below, the cave mouth guano and regurgitated palm seeds mixed with some bat (Gunther, 1940). guano (although the deep sample site in Aripo Main In the twilight zone near the mouth is a large, Cave has only bat guano). high-roofed chamber with a small number of oilbirds Oropouche or Cumaca Cave nesting high up on the walls, and large numbers of bats, This is a linear cave, large by Trinidad standards, about Glossophaga s. soricina, roosting under an overhang on one 200 m long and with an emergent steam running the full side. The steep, rock-piled floor is covered in places by length of it. The outer part is occupied by about two deep beds of mixed bat and bird guano and the usual hundred nesting oilbirds and small numbers of bats. regurgitated seeds. A light trap run in this zone on 18 Guano and regurgitated palm seeds cover the rocky floor June 1989 caught a great number and variety of flies and wherever it is not washed clean by the stream. small non-social wasps, including two species of crane On 7 Oct. 1989 a light trap was set up at the flies. One is a striking species of He/ius with each antenna innermost edge of the part occupied by the oilbirds, in consisting of only three segments; typically there ae 16 complete darkness, beside the stream and near to a large segments in other species of Helius. Five males were bank of dry guano. Rather small numbers of insects were caught (Fig. 3). An additional five males were collected caught, but the catch included two species of crane flies. at a trap set on 24 February 1991. There are about 45 One was easily recognisable as Erioptera troglodyta described Neotropical species of Helius, but none Edwards, of which four males and five females were previously recorded from Trinidad. Larvae of the few obtained (Fig. 1). Larvae of other species in this genus are species known are semi-aquatic, in decaying vegetation or semi-aquatic, found in sediments along streams and sediments of marshes (Alexander 1920). No such habitats similar sites (Alexander 1920; Gelhaus, unpubl.), a seemed to be available in this rather dry part of the cave, 39 4 I I l • Figs. 1-3. Wing venation of adult Tipulidae. 1, Erioptera Figs. 4-6. Wing venation of adult Tipulidae. 4, Gnophomyia (Mesocyphona) troglodyta (34x); 2, Teucholabis (Teucholabis) sp. (20x); 5, Cryptolabis sp. (26x); 6, Limonia (Rhipidia) sp. B sp. (20x); 3, Helius sp. (22x) (! 7x) but the floor of hugh tumbled rocks may well have connects to a 16 m high vertical shaft that opens to concealed inaccessible damp pockets of mud underneath. daylight by a small hole only about 1.5 m across and flush with the forest floor (forming a natural pitfall trap, and a The second species found here in both 1989 and very good reason not to blunder around in the forest at 1991 is Gnophomyia sp. (Fig. 4). represented by five night). By way of a low passage it also connects specimens of each sex; this is also a new record for downwards into a small but high-roofed dark chamber Trinidad, although there are 91 known species in the where several pairs of oilbirds nest on ledges. There is no Neotropics. Larvae of the few known species are found running or standing water anywhere in the cave, but under bark of moist decaying wood or decaying flower seepage water keeps the floor of the dark chamber moist. bracts (Alexander 1920; Rogers, 1927). a possible but Light traps run in this chamber on 22 April 1990 and 3 minor habitat here since only a small amount of vegetable March 1991 each caught three species of crane flies. Two debris falls into the cave. of them were the same species as those caught near the mouth of the Aripo Main Cave, namely Helius sp. (three The 1991 trap also yielded a single female of males in 1990; seventeen males and one female in 1991) Erioptera troglodyta, the first record for this species away and Gnophomyia sp. (one female each trap). The from the type locality cave. remaining flies were of two different spicies in the Below the entrance chamber the cave narrows subgenus Limonia (Rhipidia), species A represented by one male in 1990, and species B by a female specimen in 1991 and plunges steeply down over rocky floors and two near (Fig. 6). The larvae are reported to live in decaying vertical drops requiring ropes and/or ladders. This part of vegetation (Alexander 1920), in this case presumably the the cave is dry and clean, with no bat roosts or guano bird guano and seed debris. There are about 110 species beds. Where the stream rises through the cave floor there described in the subgenus Rhipidia in the Neotropics but are clusters of bats, Anoura g. geoffroyi (Ordway 1953), none from Trinidad. roosting on the walls and ceiling, and small Discussion accumulations of guano on the floor. This part of the Animals found in caves are divided into three cave is very humid. The light-trap here on 25 March 1990 categories. Trogloxenes (= cave strangers) enter caves caught two males and five females of a small, pale, only occasionally or accidentally. Troglophiles (= cave delicate species of Cryptolabis (Fig. 5). The subgenus lovers) seek out caves and may pass their whole lives in includes nearly fifty species, none recorded from Trinidad, them, but they also live in the surrounding area. and only one from Venezuela. Known larvae are fully Troglobites (= animals confined to caves) have become aquatic, living in bottom sediments of rapid streams adapted to living in caves and do not live outside them. (Hynes 1963), a habitat well represented here. The results of this simple survey cannot clearly Aripo Soho Cave distinguish which category the crane flies belong to, The exact position of this cave is uncertain, but it is particularly since no juvenile stages were found nor were probably within a mile or two of Aripo Main Cave and adults collected in the surrounding forest. The somewhat lower. It is a non-linear cave consisting of a characteristics of the sampling sites do give clues to long domed chamber that has collapsed at one end probable categories. The entrance chamber of Aripo Main providing easy access down a steep soil slope. The whole of this large chamber is twilit, and only a few bats roost in Cave was the only site sampled that was not completely its darkest comers. At its inner end the entrance chamber dark, nor was the chamber deep within the cave. Even so, Living World Journal o(The Trinidad & Tobago Field Na[Uralists' Club 1993 . 1994 40 two of the species of crane flies caught there, Erioptera are best preserved in envelopes in the field and then troglodyta and Gnophomyia sp., were also found in the mounted on their left side on points; alternatively, they Aripo Soho Cave or Oropouche Cave. Thus these species can be preserved in alcohol. Larvae can be collected in a are at least troglophiles, preferring (at least as adults) to great variety of habitats, and reared to the adult stage for inhabit caves. identification. Anyone willing to specialise in the group The continued presence of Erioptera tToglodyta in can be assured of many new records, and will probably Oropouche Cave more than seventy years after it was first find a good number of new species as well. collected there also seems to indicate a stable association References with this cave, although no other species in this subgenus Alexander, C.P. 1920. The crane-flies of New have been recorded from caves. Despite its specific name, York. Part 11. Biology and Phylogeny. Cornell University suggesting an obligate association with caves, this may be Agricultural Experiment Station. Memoirs 38: 691-1133. a troglophile rather than a troglobite, as the population Alexander, C.P. 1961. A new cave-inhabiting sampled is not isolated from the cave entrance. crane fly from Malaya (Diptera: Tipulidae). Pacific Insects Two species are probably toglobites and deserve 3(1): 27-29. closer study. Cryptolabis sp. was collected only from the Alexander, C.P. & Alexander, M.M. 1970. deep site in Aripo Main Cave. This site, with its bat roost Family Tipulidae, fascicle 4:1-259. In: N. Papavero (ed.), and subterranean stream, is isolated by a long stretch of Catalogue of the Diptera of the Americas south of the United States. Museu de Zoologia, Universidade Sao dry, clean passage from the cave entrance, preventing easy Paulo. access from comparable habitats outside. However, the flies showed no obvious morphological adaptation to cave Alexander, C.P. & Byers, G.W. 1981. Tipulidae, pp. life, such as reduction in eye size or paler body coloration, 153-190. In: J.E McAlpine et aI., eds., Manual of Nearctic Diptera, Vol. 1. Research Branch Agriculture when compared to other species of the genus which do Canada, Monograph 27, 674 pp. nor live in caves such as C. paradoxa and C. pallida. No other species of this genus have ever been reported from Byers, G.W. 1981. Tipulidae, pp. 231-241. In: Hurlbert, S.H. et aI., eds., Aquatic Biota of Tropical caves. South America, Part 1. San Diego State University, Unlike Cryptolabis, Helius sp., a probable California. troglobite, does show morphological modification when Edwards, F.W. 1918. Two new Diptera from compared to other Neotropical species in the genus, Trinidad. Annals Magazine Narural History (9) 1:424- namely, a strong reduction in antennal segmentation; this 425. is possibly related to cave life. In addition, HeUus sp. was Gelhaus, J .K. & Young C.W. 1991. The repeatedly found in both the Aripo Main Cave and the immature instars and biology of the crane fly genus more isolated Soho Cave. Other species of Helius have Brachypremna Osten Sacken (Diptera: Tipulidae). been recorded from caves in Southeast Asia (Alexander Proceedings of the Entomological Society of Washington 1961), and a species in Peru was found aggregated in dark 93(3):613-621. crevices along a dry stream bed (R. Bouchard, personal Graham, R.E. 1966. Crane-flies (Trichoceridae, communication), although these species show no Tipulidae) in California caves. Cave Notes 8:41-48. reduction in the antennal structure. The two North Gunther, A.E. 1940. The Aripo Caves. Trinidad American species of Helius are not known from caves. Guardian, 26 May 1940, 17. (Published anonymously). Of some interest is the total lack of the genus Hynes, C.D. 1963. Description of the immature Trentepohlia from these caves. Species of this genus are stages of Cryptolabis magnistyla Alexander (Diptera: commonly found in dark crevices and holes in Peru and Tipulidae). Pan-Pacific Entomologist 39:255-260. Ecuador (Gelhaus, unpubl.), where adults may group in & large numbers, all flying in tight circles, with only the Jennings, G.M. Darlington, J.P.E.C., 1990. outstretched white tarsi of the legs visible in the darkness. Flies in Tamana Cave. U.W.l. BioSpectrum 2:4-8. The group is widespread in the tropics and probably Ordway, E. 1953. The big Aripo Cave of occurs in Trinidad. Possibly the adults are not attracted to Trinidad. The Column 1(3). 2. light and so would not have been sampled by the trap Rogers, J.S. 1927. The immature stages of used. Gnophomyia jacobsoni Alexander (Dipt.). Supplementa Crane flies are common and widespread insects in Entomologica 16:77-80. Trinidad and Tobago. They are so poorly known solely Rogers, J.S. 1933. The ecological distribution of because no-one has ever seriously collected them. The the crane-flies of northern Florida. Ecological adults are easy enough to catch, but can be fragile. Adults Monographs 3:1-74. living World Journal of The Trinidad & Tobago Field Naturalists' Club 1993 . 1994 41 Patterns in the Calling Activity of the Pauraque Nightjar, Nyctidromus albicollis By Victor C. Quesnel P.O. Box 47, Port of Spain, Trinidad W.l. The Pauraque, Nyctidromus albicollis, is a very Table 1 Dates for the end and start of calling activity 1985 - vocal night jar with a varied repertory of sounds. I have 1992; call 1 only. The day of the lunar cycle is given in brackets, day 1 being new moon. heard five distinctly different calls that are common and two others that are not common. The common ones are: No. days Last Call First Call between dates 1. A loud, far-carrying whistle that I now render as to 6.V1.85 (019) 21.X.85 (08) 138 wee-oo; 2. a soft uh-wup, uh-wup, uh-wup, uh-wup-per-ah 10.V1.86 (04) 15.lX.86 (012) 97 that is variable in the number of preliminary uh-wups and 21.Vl.87 (025) 9.VIII.87 (016) 49 may have none; 3. a soft but sharp wut or wit, usually 27.V1.88 (014) 20.VIII.88 (09) 54 repeated two or three times; 4. a drumming sound, thun 22.VI.89 (020) 15.lX.89 (016) 85 thun-thun-thun ... (or thack-thack-thack. .. ) that is soft, low 6.V1.90 (014) 25.lX.90 (07) 111 and sustained for many seconds and 5. a "growl". I have < 12.VI.91 (01) 26.lX.91 (019) 106 once heard a sound I would render as waa-oo and again 21.V.92 (020) 7.lX.92 (05) 109 once a low, soft drawn-out woooo. As is to be expected, the sounds are used in other work I was doing. If I realized that calling had different situations. The first call seems to be a stopp~d without my being sure of the time, I took the declaration of territorial ownership, but additional uses time of the stoppage as the time half way between are discussed later. The second call seems to be used for noticing the absence of calls and the time at the last communication between members of a pair or family check of the watch. Any error here is a systematic error group. The third call seems to be an alarm. The fourth affecting all observations to a greater or lesser degree and call accompanies a ritualized display that I interpret as does not affect the conclusions that can be drawn. courtship. The fifth call I have heard only mixed with Only the period from sunset to midnight is other calls and its use is not at all clear. covered. During periods of moonlight, calling goes on This paper mainly concerns the first call. after midnight, and probably all night during the few days Methods around full moon, at least during the months February - In December 1983 I began a study of the May. Sporadic observations were made in the period from frequency of calls 1 and 2 by counting the number of calls midnight to sunrise and the methods used in treating in a five-minute period at 1 hr, 1 1/2 hr, 2 hr, 3 hr and 4 these observations are given when they are dealt with in hr after sunset on as many days as possible in each the next section. month. The study lasted until the end of June 1984 when All the observations were made at Haven Hill calling had ceased. The results were plotted but gave no Farm on Leotaud Trace near Talparo. clear picture, so the method was altered for the two Results subsequent seasons 1984-1985 and 1985-1986. Annual Pattern: On at least two nights per week, as I sat at my With the onset of the rainy season calling ceases desk working, I would listen for the calls of the Pauraque. (Table I). Calling begins again in August with the wup I could hear up to four different birds calling at the same per-ah call but the to-wee-oo call may be delayed several time and I would record the time when a bout of calling weeks (Table I). In all years from 1985 to 1992 calling has began and when it ceased. Both calls 1 and 2 were noted, ceased in late Mayor June. The resumption of the to-wee- but whereas the first was often repeated continually for 00 call, however, is spread out over a much longer period, many minutes the second was always isolated or repeated from August to October. This may be attributed to the only once or twice. In the analysis only call 1 was used. fact that the start of calling may be only a single call The start of the bout of calling was always easy to which may not be repeated for another month. Perhaps notice, and once calling had begun, I would check my where the beginning is late I may have missed an earlier watch every 10 - 15 min. However, the end was not call, but weather may be an influence as well. It should be always easy to notice because the calls sometimes got noted that whereas the cessation of calling is spread over softer towards the end of a long bout, and also because many days of the lunar cycle the resumption of calling has my attention would be distracted from the calls by the . always fallen between Day 5 and Day 19 of the lunar living World Joomal of' The Trinidad &. Tobago Field NaNralillU' Club 1993 . 1994 42 cycle i.e. from near first quarter ro just after full moon " (Table l). In the study from Dec. 1983 to June 1984 when " counts were made in the specified 5-min intervals (see .• Methods) the only pattern that emerged on analysis was an annual pattern (Table II). Fom a low frequency in Dec. 1983 calls increased up to March 1984 and thereafter • declined. For both the other seasons the number of minutes § ~ . spent in calling and the number of minutes spent in ; observation per month were determined (Table III). In ,~ . 1985 the peak rate of calling (minutes of calling as a ~ percent of minutes of observation) occurred in February , . whereas in 1986 the peak came in March. It is clear that • for every season the rate of calling was low when calling began, increased up to February or March and thereafter remained high until May. Calling always ceased with the arrival of the rainy season. Lunar pattern lUNERCVCLE· OAYS On the record sheets for the seasons 1984-85 and Fig. 1 The influence of moonlight on calling activity. The 1985-86 the day of the lunar cycle was written in against solid line, derived from the observations, shows that the each observation with new moon as Day 1. Then a table amount of time spent in calling between 1800 hrs and was prepared for each season's record giving for each midnight increases as the moon waxes and then drop off sharply four days after full moon. Adding a mirror image of observation in chronological order the following items: itself to this histogram in order to take account of calling the day of the lunar cycle, the number of minutes of between midnight and 0600 hrs gives a symmetrical histogram calling, and the number of minutes spent in observation. (dashed line) which is similar to that for the number of hours for which the moon is visible from 1800 hrs to 0600 hrs over From these two tables (one for each of the two seasons) the lunar cycle (dotted line). another table was prepared from the pooled results by rearranging the material ro give under each day of the Table II Annual pattern of calling - Dec. 1983 - June 1984 lunar cycle the number of minutes spent in calling and Total number of calls per month obtained by adding all counts in the S~min. observation periods (see text). the number of minutes spent in observation. The percentage of observation time spent in calling was then No.5-min. Month No. calls periods Calls/period determined for each day and the figures plotted as a histogram (Fig. 1). Dec. 433 26 16.65 Jan. 831 38 21.87 The figure shows that the percentage of time Feb. 1101 34 32.38 spent in calling increases from Day 1 to about Day 5 and Mar. 2464 39 63.18 then changes less noticeable up to Day 18, two days after Apr. 1470 24 61.25 full moon. There is then a dramatic drop to a low level May 730 28 26.07 until Day 28 when the percentage rises again. Thus, the June 462 15 30.80 pattern is not symmetrical about Day 16; there is July o 49 o Table III Annual pattern of calling activity for the two seasons 1984-85 and 1985-86 1984-85 1985-86 Min. Min. calls/obs Min Min. calls/obs Month calling observ. XI00 calling observ. XI00 Oct 49 1080 4.54 166 1890 8.78 Nov 181 1181 15.53 63 2940 2.14 Dec 372 3035 12.26 III 3630 3.06 Jan 618 4340 14.24 215 3175 6.77 Feb 877 2710 32.36 1183 3815 31.01 Mar 883 3410 25.89 1298 2640 49.17 Apr 735 2345 31.34 1474 3480 42.36 May 773 2485 31.11 1962 5225 37.55 June 82 1080 7.41 34 1255 2.71 July 0 2528 0 Living World Journal of The Trinidad & Tobago Field Naturalists' Club 1993 - 1994 43 markedly less calling in the nights after Day 16 than is visible and when it is not. These are approximate only before. and were obtained by assuming that sunrise and sunset Most probably, the reason for this lies in the fact occurred at 0600 hr and 1800 hr respectively, that the that the observations were restricted to the period new moon occurred at 0600 hr on Day 1 of the lunar between 1800 hr and midnight and that moonlight varies cycle and that the moon rose 48 min. later each day. over this period from less than 1 hr at Day 1 to 6 hr at As the table shows, there was calling in 40 out of day 16, with no moonlight at all from Day 24 to Day 30., 47 observations that were made during periods of greater Therefore, if moonlight stimulates the birds to call, a or lesser moonlight and in 17 out of 23 observations that great diminution in calling after about Day 20 can be were made when there was no moonlight. If calling expected. This is exactly what appears in Fig. 1. If results occurs regularly in the hour before sunrise as it does in were available for the period midnight to 0600 hr then the hour after sunset we should expect it on all or nearly the figure might well be symmetrical. all days. Unfortunately, there were no observations at this If moonlight were as important in the period time from Day 10 to Day 23 inclusive so that this after midnight as it is in the period before, we would probability is not tested. However, there were exactly the expect the pattern of calls in the period after midnight to same number of observations of calling in this 1 hr period be a mirror image of the pattern shown in Fig. 1 for calls (eleven) during the dark half of the lunar cycle as during before midnight. I tried to find evidence for this from my the light half. Omitting these observations and the one notes. They contain seventy observations for the period observation with no calls as possibly biasing the results midnight - 0600 hr, thirty three of which are notes of because of sunlight, we are left with calling in 6 out of 12 calling in progress at a particular time, and another two observations in the dark phase (50%) and 29 out of 35 in of no calling at a particular time. The others refer to the light phase (83%). It would seem therefore that the observations over several minutes with one as long as two results, though based on few observations, support the hours. thesis that moonlight influences the number of calls in the six hours after midnight just as it does in the six hours To treat all of these observations in a uniform before midnight. way (as though single observations at a particular time) the period after midnight was divided into six 1 hr In this period (midnight-dawn) there is no periods and any observation of continuous calling over moonlight from Day 1 to Day 8. Moonlight increases more than 1 hr or for less than one hour but spanning from 48 min on Day 9 to 6 hr on Day 16 and after 0 Z3 parts of two 1 hr periods was considered as two steadily declines to 48 min. near dawn on Day 30. This observations. The 2 hr period of calling that spanned 1 hr pattern is the mirror image of the pattern for the period and parts of two others was considered as three 1800 hr - 2400 hr. and we would expect the pattern of observations. Continuous observation where there was no calls to be a mirror image too. If we accept this and add a calling was similarly treated. Two or more periods of mirror image pattern to the pattern shown by the solid continuous calling within an hour were considered as one lines of Fig. 1 we get the symmetrical pattern of the observation and an observation of calling on the hour dashed lines which is much the same as the pattern for was referred to the period ending with that hour. All the the variation in the amount of time for which the moon observations were then entered into a table, Table IV. is visible, the dotted lines of Fig. 1. This analysis leaves The regular numerals refer to the observations with calls, out of consideration the effect of light intensity on the the bold numerals to the observations without calls. frequency of calling. In clear skies light intensity would Indicated in the table too are the periods when the moon peak at Day 16 and fall off on either side like the graph Table IV Distribution of night jar bouts of calling in the hours after midnight over the lunar month. Regular numerals refer to the number of observations when calls were heard during each hour from midnight to 0600 hr. Bold numerals refer to the number of observations with no calls. There is only one entty per hour on anyone day. Any observation on the hour was counted for the predceeding hour. The shaded areas designate the nights when the moon was not visible (See text). Day: 1 2 3 4 5 6 7 8 9 1011 12 13 14 15 16 17 18 19 20 21 ZZ Z3 24 25 26 27 28 29 30 Hr 112111111112121 1412111 1 2 1 1 1 1 3 1 4 2 1 1 1 1 1 5 1 1 1 1 1 6 1 2 3 2 1 1 1 1 21 1 2 1 4 living World Joumal of The Trinidad & Tobago Field Naturalists' Club 1993 . 1994 44 for duration of moonlight. It is reasonable to expect that it too affects calling. ....2 ~ ~ t:""-:== ______Daily Pattern ,.. , .. .. From the observations over the two , .. = ~. = '" seasons 1984-1985 and 1985-1986 those -- ," .. '" that covered the full six hours from 1800 hr ,'" .. ,. ", ~ , .. to midnight were selected and are presented ,,, , .. in Fig. 2. Several points should be noted. .. ,. " = '"~ . 1. There is no daily pattern in the ..'" '" ~" sense that calling is restricted to a certain ". '0 , ", ....~ period of the night that is the same from ". night to night; the time during which ~ '"C)14° " 300 :: ::::i: t>,~ calling occurs varies and so does the . - .~ ~ duration of calling. 302'lD:!I ,. .... Do , 2. As mentioned in an earlier study j;. (»I D" ... 15 In : (Quesnel 1990), even if no calling occurs in " . g 'm~:: the rest of the night it often occurs within , .. ,. =,. " an hour of sunset. There are 22 instances of , .. ~.= --- this in Fig. 2. ~,=, .. '" 3. On the other hand, there are :;;,'".. ~ very few cases (4) where there is no calling :;:, .. early in the evening when there is calling ::~ .. late at night. However, if the period :: ~ between midnight and 0600 hr is :;:='" considered as well, there are likely to be 01 '-..-00 --...:...;"""'------2000.,-----':-'00,..,------.,,'200----:-'''''':-:------,.,.---'00 many more nights when calling occurs late TIMEl HOURS . f 'd' h d d 1 h h Fig. 2 Variation in the duration of calling on selected nights during the two I.e. a ter ml mg t an near awn, a t aug seasons 1984 . 85 and 1985 . 86. The nights selected are those where it has not occured at sunset. These nights observation was continuous from 1800 hrs. to midnight would occur between full moon and the new moon. 10.lY.1987 and 22.II1.1992 agree well with this. From July 4. In the early part of the season calling tends ro or August little parties of birds may be seen on the roads. occur in periods of shortened duration that are less One or two of the birds may seem smaller and with fequent than those later in the season. From about shorter tails than usual. I believe that these are young February, spells of calling increase in duration. birds accompanying their parents on to the roads. At this 5. Apart from the hour after sunset and that time the numbers of birds on the roads increase (Quesnel before dawn, calling at other times is highly dependent 1990) and by October start to fall off again, presumably as on the presence of moonlight (see Fig. 1 and Fig. 2) the young birds die or disperse. I have no records of seemingly young birds after January, so by this time· 6. On some nights there was no calling although dispersal must be complete or the juveniles have become the weather was fine and the moon visible for part of the indistinguishable from their parents. Presumably, moult night. These episodes are surprising and I have no occurs during the period July - September, though I have explanation for them. For instance, the night of 29 Jan. no record of it. 1986 was fine and cool, on the edge of the period of maximum calling in February, and the moon was visible In the eight years of observation calling has from about 2200 hr, yet there was not a single call up to always ceased in late Mayor early June (Table 1). During 15 min past midnight. On the other hand, it seems clear the ensuing two months none of the various calls was that rainy weather suppresses calling. heard. On the resumption of calling, the wup-per-rah call was the first to be heard with the to-wee-oo call following Discussion several weeks later. Thereafter the rate of calling built up The annual pattern of calling is clearly related to slowly at first and then with increasing speed to the peak the breeding season. Belcher and Smooker (1936) state in February or March (Tables II and III). In Costa Rica that "the peak of the breeding time falls between the pattern seems to be much the same (Skutch 1972). February and April but G.D.S. has found fresh eggs as late as July". My records of eggs on 19.1Y.1976, 19.V1.l978, Living World Joumal of The Trinidad &. Tobago Field Naruralis[s' Club 1993· 1994 45 I first heard the drumming call on 3 March 1985 others, concluding that there is a lunar effect on many and thereafter heard it and saw the birds perform their biological processes, including human behaviour. His ritual display several times in the breeding season of 1985 studies of homicide, which he claimed clearly showed the and again five times between 5 February 1986 and 12 influence of the moon, were criticised by Abell (1979) April 1986. The display seems to be courtship for it is who did a statistical test on them and found them not performed by both male and female when they are close significant. He also criticised Lieber's hypothesis about a together. By this time the to-wee-oo calls have increased biological tide and, it is true, Lieber gives little detail in frequency over the low levels of September and concerning the mechanism by which the supposed tide October. Thus, the increase in calling coincides with the produces its effect. However, the influence of moonlight onset of courtship, and, presumably, territory selection as on the calling activity of the pauraque is not like the well. When the time for laying arrives early in the year influence dealt with by Lieber. His graphs show, where calling is at or near its peak, and continues strongly they show any clear pattern at all, two or more peaks in thereafter until the rainy season arrives in June when the lunar month. Fig. 1 above shows just one peak and is calling ceases completely. therefore quite different. Other night jars too are According to Skutch (1972) both sexes incubate influenced by moonlight (Brigham and Barclay 1992). during the day with the female alone incubating at night. My notes contain numerous references to the This means that during incubation the male is free to weather. For example, in December 1985 on 12th, 14th, patrol the terrirory and call from his accustomed stations. 15th and 16th the pauraques called for at least some However, both sexes brood the young, though Skutch minutes in the period 1800 - 1930 hours with the moon does not specifically say that both brood at night. When in Day 1 to Day 5 of the cycle. On 18th, 22nd, 27th and the male broods he is not free ro patrol the territory, but I 30th during Day 7 to Day 16 of the cycle when calling believe that the female then does this instead. It is very should have increased, there were no calls at all. On 18th difficult to observe these birds in the dim light in which and 27th the sky was altenately clear and cloudy; on 22nd they perform most of their activities, but my observations and 30th the sky was overcast and drizzly. On 26th show that the female too makes the to-wee-oo call, February 1985 on Day 8 of the lunar cycle when frequent though it is more like twee-oo, fainter and higher-pitched calling would be expected, there was one call at 2115 hr than the male's call, and resembles the male's call at the and silence before and after. The weather was rainy. On end of a long series when they become fainter and often 26th February 1986 on Day 19 of the cycle calling was in contracted to twee-oo. progress when I arrived home at 2230 hr; it went on However, the to-wee-oo call may have other continuously until 30 minutes after midnight at which functions. The conclusions that can be drawn from my time rain fell and the calling stopped abruptly. All these notes are as follows: observations, and similar ones, suggest that rainy weather suppresses calling. On the other hand, on 29th and 30th I. The male's to-wee-oo call attracts females (only January 1986 on Day 20 and Day 21 of the cycle there was his mate?). 2. The to-wee-oo call is sometimes made in not a single call all night from 1800 hr to midnight close proximity to the female while she is "drumming". 3. though both nights were clear and cool. Although the The to-wee-oo call can be made while the male is moon would not have risen until after 2100 hours and displaying to the famale. 4. The to-wee-oo call can occur 2200 hours respectively, the lack of calling on these in a sequence mixed with other calls. 5. The famale can nights is strange, and suggests that not all the factors that answer the male's to-wee-oo call by calling to-wee-oo influence calling have been recognised. herself or with a wut or by "drumming", which can go on continuously while the male is calling. One final point seems worth mentioning. It is that the occurrence of calling just after sunset and just Thus it appears that the male's too-wee-oo call is before dawn indicates the importance of light, not not simply a call to claim territory, but also attracts specifically moonlight, in stimulating calling. Also, the females and is part of courtship. Hence the intimate white wing bars and tail stripes of the male and the relationship between the annual pattern of calling behaviour that displays them to the female would hardly activity and the breeding cycle. have evolved if the birds were most active in near total The lunar pattern is, of course, determined by the darkness. It would seem therefore, that the pauraque is a lunar cycle. Simply put, the more moonlight there is, the bird superbly adapted to crepuscular conditions. more calling there is. The pattern seems clear and Acknowledgement incontrovertible but the mere existence of biological I thank Richard ffrench for reading an earlier rhythms that are influenced by the moon is controversial. version of this paper and suggesting a number of useful Lieber (1979), in his book The Lunar Effect, gives the .alterations for making it more effective. results of his own investigations and summarises many Living World Journal of The Trinidad & Tobago Field Naturalists' Club 1993 . 1994 46 References Lieber, A.L. 1979. The Lunar Effect. A Corgi Abell, G.O. 1979. Review of The Lunar Effect: book, Transworld Publishers Ltd, London. Biological Tides and Human Emotions, by Arnold L. Quesnel, V.C. 1990. Observations on the Lieber. The Skeptical Inquirer 3: 68-93. Pauraque night jar, Nyccidromus albicollis, on the roads at Belcher, C. and G. D. Smooker 1936. Birds of night. Living World, J. Trin. Tobago Field Nat. Club the colony of Trinidad and Tobago, Part III. The Ibis 13th 1989-1990,12-15. Series 66(1): 1-35. Skutch, A.F. 1972. Family Caprimulgidae, Brigham, R.M. and R.M.R. Barclay 1992. Pauraque, Nyccidromus albicollis . Studies of Tropical Lunar influences on foraging and nesting activity of American Birds (No.10). Nuttall Ornith. Club, common poorwills, Phalaenopcilus nucwllii. The Auk Cambridge, Mass., 3-21. 109(2): 315-320. Book Reviews A guide to the Birds of Trinidad & Tobago 2nd edition Richard ffrench. Plates and drawings by John P. O'Neill, portraits by Don R. &kelberry. 1991. Comstock Publishing Associates, a division of Come II University Press, Ithaca, New York. xvii + 426 pp. I This publication completely updates the first have been recast in far better detail than in the first edition of this unique and indispensable work, which was edition. published in 1973 and last revised in the early 1980's. As expected in a new edition, the information Although that edition is still useable, the new edition presentd here has been updated and expanded from that contains many improvements and a plethora of new in the first edition. The original introduction was so information. thorough that little could be added here except for a few Illustrations are the heart of a field guide. The timely updates, such as the recovery of certain species in original paintings were very good, but I feel that an Tobago. Several tables have been revised to reflect new opportunity has been lost to eliminate confusion due to information on the occurrence of migratory species in crowding. All of the colour plates have been reprinted Trin idad and Tobago. intact at the same size as in the first edition, even though The body of the book consists of detailed the new edition features a larger page size. A better use of accounts of more than 400 species of birds. Suffice it to the larger page size would have been to reposition male say that ffrench has done another masterful job of pulling female pairs and to provide more space among species, together the old and the new and has added descriptions especially on the plates of the hummingbirds, manakins, of several species new to Trinidad and Tobago. Birders and tanagers. Many plates now depict rather small images familiar with the first edition will appreciate the large of birds tightly clumped on pages with wide, unused amount of new information that ffrench has added to the margins. existing species descriptions, drawing upon firsthand John O'Neill painted one new colour plate for reports from competent birders as well as the literature this edition, illustrating such gorgeous but unrelated base. When one considers the length of time required to species as Channel-billed Toucan and Scarlet Ibis. The produce a book of this nature, it is amazing to find that it exceptional whiteness of the paper on which the plates includes records as current as February 1991, only six are printed adds to the brilliance of the colours. months before publication. All of the portraits from the first edition have Among the other improvements, ffrench has been consolidated at the center of the book, following the incorporated the latest American Ornithologists Union plates, which makes finding them much easier than names, substituting Whistling-Duck for Tree-duck, before. The portraits would have been even more Common Piping-Guan for Trinidad Piping-Guan, accessible had they been inserted in phylogenetic order Olivaceous Cormorant for Neotropic Cormorant, etc. among the plates rather than being lumped after them. Such consistency should ease the synonym problem for The present arrangement, however, still is better than users comparing species descriptions among field guides. that in the first edition. As in the first edition, 24 species The titles of some of the plates have been change to more are illustrated as line drawings. A dozen of these drawings accurately reflect their content, e.g., "Large Raptors" instead of "Hawkes and Vulture", "Medium-sized and Living World Journal of The Trinidad & Tobago Field Naruralists' Club 1993 • 1991 47 Small Raptors" instead of "Kites and Falcons", and species accounts, species that have been decimated by "Hermits and Larger Hummingbirds" and "Smaller human disturbance and hunting. He calls for additional Hummingbirds" instead of just "Hummingbirds". protection for species that, as everywhere, are being Welcome evidence of editorial generosity threatened by unrestricted clearing of land and pervades the text. The editor has been especially unenforced restrictions or bans on hunting. generous with eye-relieving white space such as blank All books described as "field guides" should be lines between table of contents items and between topics portable. This edition meets that criterion, being midway in the species accounts. Such breaks are especially helpful in size between the National Geographic SOCiety's Field in allowing the eye to locate section headings quickly. In Guide to Birds of North America and Stiles and Skutch's the same vein, almost all of the figures, maps, and A Guide to the Birds of Costa Rica. It fits comfortably photographs have been printed larger than in the first into a belt-strap book pouch. Those who balk at the edition. Moreover, the resolution of the photographs has weight of the hardcover version (suggested list price been enhanced. Finally, the type is set more tightly than US$72.50) will welcome the availability of a rugged in the first edition, which makes for easier reading. softback version (about US$35.00). Besides being technically accurate, the author has William L. Murphy performed a valuable service by pointing out, in the 7202 Mathew Street Greenbelt, MD 20770 (301-474-1880 home) II It will not hurt to have two reviews of such an think that the bird calls oo-wooo-uh-dit, repeated once or important book as this, though mine only enlarges on twice and finishing on oo-wooo}. some of the matters already raised by Murphy. Perhaps, Some minor changes of artangement may not at we in Trinidad and Tobago are fortunate in having only first be noticed, but they point to major changes taking 450 species as against the 900 species in Venezuela, for place in taxonomy. The Green Heron, Butorides lIirescens, the smaller number makes it practical to give in a is now considered to be identical to the Striated Heron, conveniently sized book a summary of everything known B. striatus, so both common names have been abandoned about the birds rather than just descriptions and notes on in favour of Green-backed Heron (Butorides striatus). The habitat and distibution. The importance of this cannot be Bright-rumped Atilla, the White-winged Becard and the overemphasized. It makes the book much more than just Black-tailed l1tyra, formerly classified in the Cotingidae, a field guide and calls attention to the gaps in our are now placed with the Tyrant Flycatchers. The knowledge as well as what is known. ffrench mentions Bicoloured Conebill, the Purple Honeycreeper, the Red this in his preface and we can thank him for taking this legged Honeycreeper, the Green Honeycreeper and the approach since we, the inhabitants of these islands, are Blue Dacnis, all formerly classified with the Barianaquit largely cut off from the stream of research running in the Coeribidae, have now been moved to the through the journals of the more advanced countries. Thraupidae with the Tanagers. Some taxonomists now ffrench has scoured the literature and brought us favour lumping all the Thraupidae with the lcteridae as up to date in this edition. Only a few examples will have well as other radical changes. Mercifully, ffrench has left to suffice. There are records of newly artived species such his classification pretty much as it was. He alludes to this as the Double-striped Thick-knee; new records of in his introduction (p 29) where he states "I agree with breeding such as those of the Yellow-headed Caracara and those other authors of major works on Neotropical the White-tailed Kite,; new information on breeding, Avifauna ..... that since classification is presently in a state such as the incubation and fledging times for the Yellow of flux, it would be premature to adopt major taxonomic hooded Blackbird; new information on feeding behaviour, changes ... "However, if the more radical changes are such as that for the Copper-rumped Hummingbird; even supported by research with the more modem techniques better descriptions of some calls such as that of the Scaled of gene analysis, and become more widely accepted, the Pigeon. ("A deep cooing in 2 syllables, croo-kuk, or in 4 third edition in 15 years time may well have a different syllables, cuck-a-loo-oo, the accent on the penultimate" is look. In the meantime, let us welcome this edition as a better than the original "a deep cooing in 2 syllables, the fine achievement which can only enhance the reputation first longer than the second, croo-kuk", though I still which the author gained with the original edition. Victor C. Quesnel (Ed.) Living World Journal of The Trinidad & Tobago Field Naturalists' Club 1993 . 1994 48