Journal of Science: Vol. 13 | Article 124 Cock

Batrachedra nuciferae, an -feeding associated with , Cocos nucifera, and palmiste, Roystonea oleracea, in Trinidad, West Indies Downloaded from https://academic.oup.com/jinsectscience/article/13/1/124/1752783 by guest on 29 September 2021

Matthew J. W. Cock

CABI E-UK, Bakeham Lane, Egham, TW20 9TY, UK

Abstract In 2006, nuciferae Hodges (: ) was the first phytopha- gous insect to be reported from of coconut, Cocos nucifera L. (: ), in Trinidad, West Indies. At that time, it was suggested to be an introduced species contributing to decreasing coconut yields on the island and potentially a threat to other palms. In this preliminary study, inflorescences of coconut, seven indigenous palms, and six exotic orna- mental palms were surveyed in several areas of Trinidad. Caterpillars of more than 10 species of Lepidoptera were found and reared through to the adult stage. Batrachedra nuciferae was posi- tively identified. It was concluded that the caterpillars of B. nuciferae feed on pollen in the male flowers of coconut and palmiste or royal palm, Roystonea oleracea (Jacquin) O.F. Cook. There was no evidence that B. nuciferae bred on any of the other palms surveyed, but it is not conclu- sive that they do not do so. A parasitoid, Apanteles (sensu lato) sp. (Hymenoptera: Braconidae), of B. nuciferae was reared. On available information, B. nuciferae is more likely to be an indige- nous species that has hitherto been overlooked than an introduced species. In view of what is known about damage-yield relationships and biological control agents, B. nuciferae is unlikely to cause yield losses to coconut, so control measures are not justified.

Keywords: Arecaceae, Batrachedridae, caterpillar, Lepidoptera, palms, parasitoid, pollen Correspondence: [email protected] Editor: Inon Scharf was editor of this paper. Received: 3 April 2012 Accepted: 20 January 2013 Published: 5 November 2013 Copyright: This is an open access paper. We use the Creative Commons Attribution 3.0 license that permits unrestricted use, provided that the paper is properly attributed. ISSN: 1536-2442 | Vol. 13, Number 124 Cite this paper as: Cock MJW. 2013. Batrachedra nuciferae, an inflorescence-feeding moth associated with coconut, Cocos nucifera, and palmiste, Roystonea oleracea, in Trinidad, West Indies. Journal of Insect Science 13:124. Available online: http://www.insectscience.org/13.124

Journal of Insect Science | http://www.insectscience.org 1

Journal of Insect Science: Vol. 13 | Article 124 Cock Introduction descriptions of the caterpillar, pupa, and co- coon. Batrachedra nuciferae Hodges (Lepidoptera: Batrachedridae) is a newly reported pest of Since 1940, there was no published work on coconut, Cocos nucifera L. (Arecales: the coconut moth, apart from Hodges’ (1966) Arecaceae), on the island of Trinidad description of B. nuciferae, until the work of apparently introduced at some time prior to S. Sanchéz-Soto this century. The moth was 2006. Since B. nuciferae was already known the subject of his research thesis (Sánchez So- as a pest of coconut and other palms in , to 2004) and publications on the distribution concerns have been raised locally about its (Sánchez-Soto and Nakano 2002, 2004a), Downloaded from https://academic.oup.com/jinsectscience/article/13/1/124/1752783 by guest on 29 September 2021 potential impact on and indigenous morphology (Sánchez-Soto and Nakano palms in Trinidad and Tobago, and the threat 2004b), and biology (Sánchez-Soto and it may present in the wider Caribbean. Nakano 2008). The egg, caterpillar (including chaetotaxy), pupa, and adult are illustrated History, distribution, and biology of Bac- with line drawings in both Sánchez Soto trachedra nuciferae (2004) and Sánchez-Soto and Nakano Bactrachedra nuciferae, which is referred to (2004b). Sánchez-Soto and Nakano (2008) as the coconut moth in Trinidad and Tobago reported that in captivity, development to (MALMR 2008), was first recognised by adult takes 22 days, and there is an average Bondar (1940a, b), who described its biology pre-oviposition period of 2.6 days, followed in Bahia State, Brazil, under the incorrect by 11.3 days of oviposition during which an name of B. perobtusa Meyrick. It was only in average of 31.5 eggs are laid. Hodges’ (1966) revision of the American Ba- trachedra that the species was first described, Damage caused by Batrachedra nuciferae to based on Bondar’s material reared from coco- coconut nut. In 1998, B. nuciferae was recorded from In contrast to Bondar’s (1940a, 1940b) obser- coconut in Venezuela (Arnal et al. 1998), and vations that the caterpillars feed only on in 2006 it was reported from Trinidad as a coconut pollen, Arnal et al. (1998) reared B. new pest of coconut (MALMR 2006, 2008). nuciferae on male and female flowers and stated that it feeds on both. However, this was Bondar (1940a, b) studied the associ- a brief description of a rearing method on in- ated with coconut inflorescence in Bahia, florescences to confirm the presence of the Brazil. He reported that the caterpillars of B. moth in several parts of Venezuela, and it is nuciferae rest in the male flowers of coconut, not clear if careful observations were made. where they feed on pollen, and they are also Carneiro et al. (2004) provided a summary of common in the flowers of several other palms: earlier observations and reported that during coronata (= Cocos coronata), S. the dry season all male flowers may be de- vagans (= C. vagans), S. schizophylla (= C. stroyed. They state that in the Município de schizophylla), Attalea funifera, and A. pias- Parnaíba, Piauí, Northeast Brazil, the caterpil- sabossu. Bondar considered that the damage lars feed on both male and female flowers and to male flowers reduced the probability of fer- include a photograph of damaged female tilization of female flowers and hence could flowers, which certainly look in poor condi- adversely affect nut production, but presented tion, although exactly what form the damage no evidence for this conclusion. He gives brief takes is not clear, whether general abrasion

Journal of Insect Science | http://www.insectscience.org 2

Journal of Insect Science: Vol. 13 | Article 124 Cock and fouling, surface feeding, or boring into the Objectives of this study flower/nutlet. Although the observations of A preliminary survey of the incidence of Lep- Carneiro et al. (2004) were based on reliably idoptera in the flowers of coconuts and other identified adult , no detailed observa- selected palms in Trinidad was carried out. tions seem to have been made, and the This survey was to generate information to possibility of other insects being involved in confirm the presence of B. nuciferae in Trini- causing the damage was not considered. Thus, dad and to provide background to assess the in spite of these two reports, detailed observa- threat to coconut and other palms and the need tions of damage to female flowers positively for prevention of spread and management of linked to caterpillars of B. nuciferae are not the moth. Downloaded from https://academic.oup.com/jinsectscience/article/13/1/124/1752783 by guest on 29 September 2021 available. Materials and Methods The pollination system of most palms, includ- ing coconut, is unknown or inadequately Field collections understood, and the impact of inflorescence All field work was carried out in October feeding insects on fruit production is also 2011. Before making surveys, methods were poorly understood, even in the case of coco- tested on coconut palms in Curepe (a subur- nut, an important crop (Moore 2001). It is ban area between Port of Spain and Arima), generally thought that damage to male flowers Centeno (mixed agricultural land south of alone is unlikely to be significant to the extent Arima), and Waller Field (a mixed plantation of limiting fertilisation, especially where in- east of Arima), which produced material for sects are implicated in the pollination identification and confirmed the biology and mechanism. For pollen availability to limit damage symptoms of B. nuciferae on coconut. fertilisation, it is likely that all or most male Three sites were prioritised for survey in Trin- flowers would need to be prevented from pro- idad: Nariva Swamp, Aripo Savannah, and the ducing pollen over large areas over an Royal Botanic Gardens, Port of Spain. Nariva extended period, and this level of damage has Swamp is an important area in not been reported. Trinidad. Samples were taken in the Bush Bush Island nature reserve and at Kernahan, a Damage to female flowers could lead to direct small community in the southern part of the losses of harvestable nuts. However, this is swamp. Aripo Savannah is also a protected not well understood either (Taylor 1930; Cor- area, notable for its unusual savannah-like bett 1931; Cock et al 1987; Waterhouse and vegetation. The indigenous moriche palm Norris 1987). Most female flowers are ferti- Mauritia flexuosa L.f. is an important compo- lised, but many abort naturally or are lost as nent of the flora. The Royal Botanic Gardens early nutfall. There are mechanisms whereby grow a selection of indigenous and introduced coconut palms through nutfall can reduce the palms. number of nuts brought to maturity in re- sponse to poor conditions, lack of nutrients, At each site, the procedure was as follows. A damage to the fronds, etc. In principle, this search was made on foot or from a vehicle to also means that there is scope for the palms to locate flowering palms in the survey area. The compensate for early damage to female flow- palms were identified from Comeau et al. ers by insects. (2003) and local knowledge. The stage of the inflorescence (pollen release or soon after be-

Journal of Insect Science | http://www.insectscience.org 3

Journal of Insect Science: Vol. 13 | Article 124 Cock ing preferred) and its accessibility by ladder or ground as gently as possible, and then cut up climbing was assessed. Where no material to and placed in the sample bag. In one case, a sample could be located, some old inflores- Euterpe precatoria Martius was felled to ob- cences were selected, as they would show tain a sample. The samples were labelled and indications of any damage by Lepidoptera. placed in a chest cool box for transport back Suitable palms were then sampled, either from to the laboratory. When samples exceeded the the ground for short palms, using a 5 m ladder capacity of the cool box, they were kept in the for some, and a bamboo ladder and climber shade and then taken to the laboratory in an for M. flexuosa at Aripo Savannah. Each sam- air-conditioned vehicle. At each site, pictures ple was a whole inflorescence and was placed were taken of the palm, its inflorescence and Downloaded from https://academic.oup.com/jinsectscience/article/13/1/124/1752783 by guest on 29 September 2021 in prepared sample bags (cotton cloth bags, other diagnostic features, and any other fea- measuring 112 cm deep x 77 cm wide, with a tures considered worth recording. Where there drawstring closure). Where possible, the inflo- was any doubt about the identity of the palm, rescence was cut from the palm with specimens of the , inflorescence, and minimum vibration (which might cause male fruit were taken to facilitate later identifica- flowers to drop) using secateurs or long han- tion. A GPS was used to record the position, dled clippers, and the detached inflorescence elevation, and time of each sample. The palms was placed directly into a sample bag. This and sample inflorescences obtained are given process was not always practical due to the in Table 1. difficulty of access or the toughness of the rachis, so in some cases a sample bag was Laboratory methods placed over the whole inflorescence before On return to the laboratory, the sample bags cutting the stem with a machete. In the case of containing inflorescence samples were placed the samples from M. flexuosa and mau- in an air-conditioned room overnight and pro- ritiiformis (Karsten) Grisebach and Wendland, cessed the next day. For each sample, the the inflorescences were too big for the sample following steps were followed. The stage of bags and so they were cut and lowered to the flowering was recorded, and a photograph was

Table 1. Collections of palm inflorescences made in Trinidad, 8–18 October 2011, and the adult Bactrachedra nuciferae and other moths reared from them until 22 December 2011. All sites are in lowland areas, at altitudes of less than 50m.

Journal of Insect Science | http://www.insectscience.org 4

Journal of Insect Science: Vol. 13 | Article 124 Cock taken of the entire inflorescence and any parts moths had emerged in the previous week from that were considered of interest, e.g., details of two of the emergence boxes, so the boxes inflorescence and damage symptoms. Part or were discarded. all of the inflorescence was examined by eye and under a binocular microscope to look for The subsamples of living material from the symptoms of Lepidoptera (presence of cater- first examination of the inflorescence were pillars or pupae, silk shelters or cocoons, either placed in individual vials or in ventilat- feeding damage, frass, or droppings). Samples ed plastic boxes. The former did not prove were taken of any living caterpillars or co- satisfactory, perhaps because ventilation was coons and either placed in individual vials inadequate, and most caterpillars thus treated Downloaded from https://academic.oup.com/jinsectscience/article/13/1/124/1752783 by guest on 29 September 2021 (30–45 mL) or in groups in containers (14 di- did not complete development. The ventilated ameter x 8.5 cm height round containers or plastic boxes were more successful, and moths 16.5 x 12.5 x 7.5 cm height rectangular box- were successfully reared in these. es). Notes were made on what was found, and observations were quantified as practical, e.g., The reared moths were placed in a freezer results per primary, secondary, or tertiary (approximately -18° C) at least overnight be- branch, number per 10 flowers, etc. fore pinning, spreading, and staging using normal techniques for very small moths (e.g., The remains of the inflorescence were placed Tagestad 1974). A selection was dissected us- in an emergence box and monitored for emer- ing standard techniques (e.g., Robinson 1976) gence of moths over more than two months. and used for identification by comparison An emergence box is a cardboard box (up to with Hodges (1966). 57 x 40 x 64 cm in size) with the seams sealed with masking tape to exclude ants and prevent

escape of insects. Two round holes are cut in one side exactly large enough to fit a clear

plastic or glass vial, which is then inserted with the open end just fitting into the hole and

the closed end projecting. The box is placed in a room lit with fluorescent lighting and kept at

ambient temperatures. Moths and most other insects are attracted to light, and so insects in

the inflorescence or insects that complete their development there are attracted into the trans-

parent vials from where they can be collected. The emergence boxes were checked several

times a day, and when moths were found in the transparent vials, the vial was carefully

removed, capped, and labelled with the collec- Figure 1. Early stages of Batrachedra nuciferae. a: small cater- tion and date of emergence. A record sheet on pillar at base of male flower, with frass speckled webbing to the left; b: larger caterpillar with trace of webbing to left of lower each box was used to record the dates of male flower; c: mature caterpillar, webbing, and frass bottom emergence and provisional division of the left; d: dead male flowers tied to inflorescence with silk; e: cat- erpillar spinning cocoon on surface of cloth field collection bag; moths into B. nuciferae and other species. On f: typical cocoon formed on inflorescence branch, speckled with 22 December, only a very small number of frass and small pieces of debris. High quality figures are available online.

Journal of Insect Science | http://www.insectscience.org 5

Journal of Insect Science: Vol. 13 | Article 124 Cock

Results

In total, 26 palms were surveyed, including 8

Cocos nucifera (Table 1). From these, 265 B. nuciferae and 953 other moths were reared Figure 2. Adult Batrachedra nuciferae: a: male (wingspan 8.3 mm); b: female (wingspan 10.2 mm); c: dead adult with wings (Table 1). held parallel to body, giving an impression of the moth in life. (Photo credit: Dwayne H. Burris). High quality figures are avail- able online. The caterpillars, feeding style, and cocoons of B. nuciferae were found to be distinctive, ena- Downloaded from https://academic.oup.com/jinsectscience/article/13/1/124/1752783 by guest on 29 September 2021

bling the early stages and damage of this

species to be separated from other Lepidop-

tera found in the palm inflorescences.

Caterpillars were characterised as white to

pink, with the head and dorsal plate on the

prothorax dark (Figure 1a–c, e), as described

by Sánchez Soto (2004) and Sánchez-Soto

and Nakano (2004b). The caterpillars con-

struct shelter-tunnels with silk webbing

among the male coconut flowers, usually in-

corporating scattered frass pellets (Figure 1a–

c). As male flowers dehisce, some are at-

tached by the silk webbing and so remain

attached to the inflorescence branch as they

Figure 3. Emergence of Batrachedra nuciferae (above) and turn brown and dry (Figure 1d). Caterpillars other moths (below) from emergence boxes. Kernahan 1: Roy- were found inside male flowers that had start- stonea oleracea (palmiste), collected 12 October 2011; Waller Field: Cocos nucifera (coconut) sample 1, collected 10 October ed to open and remained within the flowers 2011. High quality figures are available online. until the contents were consumed. No evi-

dence was seen that these caterpillars made flowers attached to the silk, making it distinc-

holes in the male flowers or damaged female tive (Figure 1f). Some that were formed in the

flowers, although more observations, especial- collection bags overnight on returning from

ly at high caterpillar densities, would be the field were formed without attachments

desirable to confirm this. (Figure 1e).

Pupation was in a double, white silk cocoon, The adult moths were as characterised by

either on a solid substrate such as the inflo- Sánchez Soto (2004) and Sánchez-Soto and rescence branch or among fallen male flowers. Nakano (2004b), with very narrow, elongate,

When infested male flowers were held in a fringed wings (Figure 2). They rest with the plastic container, the cocoon was normally head end raised above the substrate and the

formed on the base of the container, and in the wing tips resting on the substrate (see Sánchez case of a transparent container, was readily Soto 2004, Figure 7). The long narrow wings

visible through the container. The surface of are held tightly against the body so that the the cocoon was normally covered with scat- moth appears long and thin (Figure 2c). The

tered frass pellets or small fragments of combination of colour, size, markings, and

Journal of Insect Science | http://www.insectscience.org 6

Journal of Insect Science: Vol. 13 | Article 124 Cock resting position enabled adult B. nuciferae to new genera are not. The parasitoids attack the be distinguished from other Lepidoptera that caterpillars before maturity, and the mature emerged from the survey samples. Adult parasitoid larva only emerges once the cater- moths were thus provisionally identified as B. pillar has spun its cocoon and developed to nuciferae. Pinned specimens were critically the prepupa stage. The parasitoid larva then examined, and selected female specimens spins a white cocoon within the B. nuciferae were dissected and compared with Hodges cocoon and emerges at about the same time as (1996, Figure 129) to confirm the identifica- the adult moth would have. When cocoons tion. The size, shape, and surface structure of were formed on a translucent rearing contain- the signum were found to be diagnostic. This er, the parasitoid cocoons were readily visible Downloaded from https://academic.oup.com/jinsectscience/article/13/1/124/1752783 by guest on 29 September 2021 identification was confirmed by a USDA Lep- through the container and host cocoon. Apan- idoptera taxonomist at the United States teles sp. were thus obtained from the Centeno National Museum based on an examination of coconut sample, direct from the B. nuciferae the male and female genitalia and comparison cocoons. Of 16 cocoons reared from this sam- with reference material in the museum. ple, 4 were parasitized.

Coconut Palmiste Adults of B. nuciferae were reared from coco- One inflorescence from a relatively short nut at all localities at which coconut was palmiste, Roystonea oleracea (Jacquin) O.F. surveyed (Curepe, Centeno, Waller Field, Cook at Kernahan was included in the survey. Bush Bush, and Kernahan), but were not Each inflorescence contains both male and found at the Royal Botanic Gardens, where female flowers, and an unusual feature of the coconut could not be surveyed. The numbers genus is that the inflorescence is packed with obtained were variable (Table 1). Only 1 spec- millions of very small, fluffy, branched hairs imen was obtained from Bush Bush, but B. before the spathe opens (Henderson et al. nuciferae was common at Curepe and Kerna- 1995). Insects such as bees are implicated in han and very common at Waller Field (90 pollination of the congeneric R. regia (Zona adults from inflorescence 1). 2000). In this inflorescence, the male flowers had recently dropped and the female flowers Most adult moths emerged within 15 days of were still small. Many of the male flowers and collection of the inflorescence (Figure 3 top), the fluffy matrix were caught up in the inflo- and none took more than 23 days in any sam- rescence and the spathe, which was still in ple, whereas other species of moths continued place below the inflorescence. to emerge for two months (Figure 3 bottom). Examination of the male flowers and fluffy Ants and predatory wasps were common in matrix revealed what appeared to be Bac- the inflorescences sampled in the field. An trachedra caterpillars along with what were Apanteles (sensu lato) sp. parasitic wasp definitely Batrachedra sp. cocoons, as well as (Hymenoptera: Braconidae) was reared from caterpillars of at least two other species. Adult B. nuciferae, but was not identified. It is re- B. nuciferae started emerging on the day of ferred to as Apanteles (sensu lato) sp. because collection and continued to do so in the emer- although the genus Apanteles has been divid- gence box for two weeks, producing 57 adults. ed into a large number of genera, it is still More than 600 adults of at least 5 other spe- easily recognisable and familiar, whereas the cies of moths were reared from this collection

Journal of Insect Science | http://www.insectscience.org 7

Journal of Insect Science: Vol. 13 | Article 124 Cock over the next two months. Although large sequently from the emergence box. At the numbers of moths reared from this single in- Royal Botanic Gardens, a stand of B. major florescence, most emerged after the showed fruiting, and inflorescences similar to Batrachedra nuciferae adults (Figure 3), so those found at Bush Bush Island. An inflo- they were probably not feeding on pollen. It rescence from which the male flowers had seems likely that the fluffy matrix contained recently been shed showed no signs of Lepi- in inflorescences of R. oleracea provides a doptera feeding. Examination of field- food source that supports many caterpillars of collected, newly-dropped, male flowers in the some of these moths. A small sample of fallen laboratory showed that they seemed fresh and male flowers of R. oleracea from the Royal had pollen remaining, so they should have Downloaded from https://academic.oup.com/jinsectscience/article/13/1/124/1752783 by guest on 29 September 2021 Botanic Gardens produced just 5 adult moths been suitable food for Lepidoptera such as of 2 species, but no B. nuciferae. Batrachedra spp. However, no sign of Lepi- doptera feeding was found when 10 flowers Other palms were opened, and the remainder examined ex- Attalea maripa (Aublet) Martius is common at ternally, and no moths subsequently emerged Bush Bush, but only an unopened spathe, an from this material. old dead male inflorescence, and a long-dead female inflorescence from a fallen palm could At Bush Bush, no simplicifrons Mar- be included in the survey. The contents of the tius could be found with open male flowers, unopened spathe were unblemished, with no although 2 were found that had recently signs of insect damage. The very old female dropped their male flowers. These were exam- inflorescence showed no recent sign of Lepi- ined in the laboratory and no trace of doptera damage and was discarded after Lepidoptera or other insect damage could be examination in the laboratory on the basis that seen. any Lepidoptera still associated with it would be saprophytic. The old male inflorescence, in Examination of the branches on an inflo- which the flower parts, including the long pol- rescence of E. precatoria from which the male len-bearing (Henderson et al. 1995), flowers had dropped showed no signs of Lep- were completely dead and dry, showing a idoptera damage and specifically no sign of great deal of old feeding damage by Lepidop- male flowers attached to the inflorescence by tera based on the amount of frass and webbing silk, or the typical cocoons of Batrachedra that was observed. Careful search revealed no spp. No moths were obtained from the emer- indication of Batrachedra type cocoons on the gence box. stems or among the flower remains, and no Batrachedra spp. moths were obtained by At Aripo Savannah, 1 male moriche palm, M. emergence box, although other moths were flexuosa, was found that was short enough to reared. sample. It had several old, dead, dark brown inflorescences and several young, yellow- Most Bactris major Jacquin palms at Bush green ones that had not yet flowered. Both Bush Island were not fruiting or flowering on developmental stage inflorescences were in- 12 October 2011, but one that had just flow- cluded in the survey. The young inflorescence ered was sampled. In the laboratory, no trace showed no sign of Lepidoptera damage when of Lepidoptera feeding could be detected on inspected, although small numbers of 3 spe- this inflorescence, and nothing emerged sub- cies of moths other than B. nuciferae were

Journal of Insect Science | http://www.insectscience.org 8

Journal of Insect Science: Vol. 13 | Article 124 Cock obtained by emergence box. Some feeding from the emergence box, so it cannot be estab- damage and very small caterpillars were lished whether or not this species is a suitable found under the bracts of the very old inflo- food for a Batrachedra sp. or any of the rescence, but there were no signs of Lepidoptera found in this survey, but certainly Batrachedra spp. and none were reared from it is host to at least 1 species of Lepidoptera. the emergence box. An old inflorescence of L. spinosa from which Sabal mauritiiformis is unusual among the the male flowers had been dropped showed no indigenous palms surveyed in that the small sign of Lepidoptera feeding, webbing of male flowers are bisexual (Henderson et al. 1995). flowers, or cocoons. Downloaded from https://academic.oup.com/jinsectscience/article/13/1/124/1752783 by guest on 29 September 2021 One inflorescence that included open male flowers was sampled adjacent to the Forestry There was light webbing and frass among Division facility at Kernahan. Caterpillars of dropped male flowers in an inflorescence of an unidentified Batrachedra sp. were found pacifica, but no caterpillars or co- and reared from this inflorescence, but no oth- coons matching those of Batrachedra spp. er Lepidoptera were found or reared. were found, and 19 adults of 2 other species of moths were obtained by emergence box. The remaining palms are all introduced orna- mentals in Trinidad: Manila palm, Adonidea The male flowers of macarthu- merrillii (Beccari), rattan palm, Calamus sp., rii are not synchronised, hence there were mangrove fan palm, spinosa Rox- small numbers of scattered male flowers on burgh, Fiji fan palm, Pritchardia pacifica, the inflorescences included in the survey at hurricane palm, Ptychosperma macarthurii both Curepe and the Royal Botanic Gardens. (H. Wendland ex H.J. Veitch) H. Wendland These and some dropped flowers caught up ex Hook.f., and queen palm, Syagrus roman- between the inflorescence branches at the zoffiana (Chamisso) Glassman. Royal Botanic Gardens were sampled. There was no trace of webbing or frass on fresh or Although B. nuciferae caterpillars were found dropped male flowers, and no Lepidoptera on coconut inflorescences about 25 m away were obtained by emergence box. from the A. merrillii surveyed at Curepe, no sign of Lepidoptera feeding could be found in In the S. romanzoffiana inflorescence sam- the inflorescence of A. merrillii, and none pled, the male flowers had already mostly were obtained by emergence box. dropped. The inflorescence seemed complete- ly healthy, except that the ends of many apical The survey sample from Calamus sp. was 2 secondary branches were caught up in the old, dry, male inflorescences, comprising a apex of the spathe together with many male densely packed mass of dry dead flowers. Ex- flowers, which had started to go moldy. This amination showed some webbing and seemed ideal for attack by Lepidoptera, but caterpillar frass in most parts of the inflo- there was no trace of caterpillar damage. A rescence, and caterpillars with a dark head and small collection of dropped male flowers was dark purplish body were associated with this. also made from the base of the palm and set Two cocoons that could have been Ba- up separately. No Lepidoptera emerged from trachedra sp. were found, but nothing either sample. emerged from these and nothing emerged

Journal of Insect Science | http://www.insectscience.org 9

Journal of Insect Science: Vol. 13 | Article 124 Cock Discussion idoptera and assumed it was due to B. nucifer- ae without rearing out specimens (he appears A preliminary spot survey of the incidence of to have sent only specimens from coconut for Lepidoptera in the flowers of coconuts and identification); (2) the species of Attalea that selected other palms in Trinidad was carried Bondar observed are suitable food , but out. The results have enabled confirmation of A. maripa is not (it is sometimes placed in a the presence of B. nuciferae to be widespread separate genus to other Attalea spp.); (3) fur- in Trinidad. Information has been generated ther samples may show that A. maripa in on the feeding damage, phenology, and natu- Trinidad is a suitable food plant for B. nucif- ral enemies of B. nuciferae. Coconut has been erae. The last would be the easiest to test, Downloaded from https://academic.oup.com/jinsectscience/article/13/1/124/1752783 by guest on 29 September 2021 confirmed as a food plant, palmiste is a new especially if fresh flowering male inflores- food plant record, and several other palms cences could be sampled. It can be concluded have been provisionally shown to be unsuita- provisionally that A. maripa is not a suitable ble food plants. This information was used to food plant for B. nuciferae. The male flowers assess the threat that B. nuciferae presents to of A. maripa consist of little more than a coconut and other palms and the need for bunch of pollen-bearing stamens (Henderson management of the moth. et al. 1995), so they are very different in struc- ture from all the other palms that were Host range of B. nuciferae sampled, and this may be a factor in the ap- The results show that both coconut and R. parent unsuitability of A. maripa as a host. oleracea were suitable hosts for B. nuciferae in Trinidad. No evidence was found of Ba- The genus Syagrus is considered to be the trachedra spp. attacking B. major, B. closest relative of Cocos (Henderson et al. simplicifrons, A. maripa, M. flexuosa, E. 1995), so S. romanzoffiana, the queen palm, is precatoria, or 4 introduced palms. Sabal mau- a likely host for B. nuciferae, not least ritiiformis was attacked by another because Bondar (1940a, b) recorded B. Batrachedra sp., but not by B. nuciferae. nuciferae from the inflorescence of other However, while proof of attack was easily Syagrus spp. Hence, it would seem premature demonstrated, proof of non-attack cannot be to conclude that B. nuciferae does not attack demonstrated without more samples taken at a S. romanzoffiana, especially because this suitable stage. In the specific case of B. nucif- survey did not confirm that B. nuciferae is erae, these samples should be taken in places present in the Royal Botanic Gardens where where B. nuciferae is known to be present on the S. romanzoffiana was sampled. other nearby hosts. In addition to the suggestion above that Because Bondar (1940a, b) recorded B. nucif- structural differences in flowers of A. maripa erae feeding abundantly in the inflorescences could be a factor in host-specificity, it seems of Attalea spp. in Brazil, the indigenous A. likely that palms with synchronised flowering maripa were expected to be suitable hosts in in an inflorescence, which then drop all male Trinidad. However, in spite of the large flowers almost immediately, would not be amount of old damage on the old male inflo- suitable hosts for Batrachedra spp. specialised rescence, this did not appear to be the case. on male pollen-producing flowers. On this There are several possible explanations: (1) basis, the observations of Essig (1971) and Bondar saw similar damage due to other Lep- Ervig (1993), which indicate this pattern of

Journal of Insect Science | http://www.insectscience.org 10

Journal of Insect Science: Vol. 13 | Article 124 Cock flowering in Bactris spp. and M. flexuosa, nuciferae caterpillars bore into male flowers mean these palms are unlikely to be suitable or attack female flowers. This matches the hosts, which was the provisional conclusion in observations of Bondar (1940a, b), Sánchez this study. Soto (2004), and Sánchez-Soto and Nakano (2008). However, the possibility that more These observations on host range should also direct damage to flowers could happen at be considered in relation to the classification higher densities of caterpillars should not be of Arecaceae (Dransfield et al. 2008). Attalea ignored. Thus, samples of heavily invested is the only genus of indigenous palm in the coconut and other palm inflorescence should same subtribe as Cocos (Attaleinae), and there be carefully examined before coming to a firm Downloaded from https://academic.oup.com/jinsectscience/article/13/1/124/1752783 by guest on 29 September 2021 was no evidence from this survey that B. nu- conclusion about this. Nevertheless, at the ciferae attacks A. maripa. Note however, that population levels found, the caterpillars of B. contrary to the observations reported here, nuciferae feed primarily, if not exclusively, on Bondar (1940a, 1940b) found Attalea spp. to pollen in male flowers and do not directly be hosts, albeit not confirmed with identified damage female flowers. specimens. There are 17 species of indigenous palms in four subtribes which are in the same Studies on coconut inflorescence damage, nut- tribe as Cocos (), but the ones sam- fall, and loss of yield (Taylor 1930; Corbett pled, Bactris spp. (Bactridinae) and Euterpe 1931; Cock et al 1987; Waterhouse and Norris precatoria (Euterpeae), were not attacked by 1987) have shown this is a complex subject B. nuciferae. However, R. oleracea, which is with conflicting results. However, most stud- in a different tribe (Roystoneae) of the same ies suggest that the loss of male flowers is subfamily as Cocos (Arecoideae), was at- unlikely to be significant, as shortfalls in male tacked. It seems likely that because of the flowers would not normally be limiting for great diversity of pollination systems in Are- pollination. On the other hand, damage to fe- caceae (Henderson 1986) linked to a male flowers is implicated in loss of yield, comparable diversity of inflorescence struc- although not conclusively. Thus, direct dam- ture and phenology, specific details such as age by B. nuciferae to male flowers would not flower structure, the duration of male flower- have a direct impact on the yield and econom- ing, and the time of retention of male flowers ics of coconut production or on nut production may be more important in defining a palm’s of indigenous palms. suitability as a host than strict phylogenetic relatedness. There is also the possibility of indirect impact. For example, by tying together male flowers Impact on nut production with silk, the microhabitat on the branches of The observations reported here support the the inflorescence and in areas where fallen view that the caterpillars of B. nuciferae feed flowers accumulate is changed, probably mak- only on male pollen, which is only briefly ing it more suitable for both Batrachedra spp. available, whereas the caterpillars of other and other inflorescence feeding Lepidoptera, Lepidoptera species feed on other parts of the including detritus feeders. Some of these other inflorescence and/or the decaying inflo- Lepidoptera will damage female flowers, as rescence. Careful examination of the palm Bondar (1940a, b), Santana (2008), and San- inflorescences where B. nuciferae caterpillars tana et al. (2009) reported for Atheloca were recognised showed no evidence that B. bondari and A. subrufella and Bondar (1940a,

Journal of Insect Science | http://www.insectscience.org 11

Journal of Insect Science: Vol. 13 | Article 124 Cock b) reported for Cadra cautella. In this situa- or two later, they would have produced more tion, the presence of B. nuciferae could lead to moths. Nevertheless, there was wide variation damage to female flowers, although the extent in the numbers of moths obtained from the and frequency to which this happens would emergence boxes, and large numbers of moths need to be investigated. were obtained in several cases, so this method does provide a time-efficient and cost- Effectiveness of methods effective way to recognise the relative density Obtaining samples was a significant challenge of populations of Lepidoptera in the different due to limited availability of flowers at the inflorescences at the time of sampling. How- selected sites in the week of the survey, and ever, this approach could be improved by (1) Downloaded from https://academic.oup.com/jinsectscience/article/13/1/124/1752783 by guest on 29 September 2021 those palms which were flowering were often standardising the stage of development of in- too tall to sample safely. The samples ob- florescence sampled, i.e., to pollen producing; tained were not necessarily at the best stage to (2) separating pollen producing flowers from look for flower feeding caterpillars, especially old dropped flowers and holding each sepa- those that feed on pollen as Batrachedra spp. rately for moth emergence; (3) establishing do. Nevertheless, because Lepidoptera feeding that B. nuciferae is present in a local area be- leaves recognisable traces, such as webbing, fore sampling other nearby palm species to frass, and characteristic cocoons, old inflores- assess host specificity. cences where the dead male flowers were still present provided useful information. Is Bactrachedra nuciferae indigenous or al- ien in Trinidad? Examination of inflorescences in the laborato- Batrachedra nuciferae was first found in Trin- ry gave fairly reliable insight into some of the idad in 2006. However, there are two possible species present and their damage, but over- interpretations of this: either it is an indige- looked eggs or very young individuals and nous species that had not previously been risked overlooking very low-density popula- noticed and identified, or it is an alien species tions. No objective assessment of the introduced some time before 2006. There is effectiveness of the emergence boxes was not definitive evidence to address this issue, made. but there are several points that should be considered. As a measure of population density, taking a sample at a single time point as done here un- The available lists of moths from Trinidad avoidably makes no allowance for eggs that (Kaye and Lamont 1927; Lamont and Callan would have been laid in the inflorescence after 1950) do not mention B. nuciferae. However, collection, nor for moths that had already the lists include very few species of the so- emerged before the sample was taken. By called families (which in- making a careful examination of the material clude Batrachedridae), and those mentioned before setting it up in the emergence box, are mostly important pests recorded through there is a qualitative control for the latter as- studies by the Department of Agriculture pect, but the former could not be controlled. (Trinidad and Tobago) entomologists and staff In some cases, Batrachedra moths did not or students of the former Imperial College of start to emerge until well after the set-up of Tropical Agriculture (St. Augustine, Trini- the emergence box, and it may well be that if dad). Batrachedra nuciferae is a small, these inflorescences had been sampled a week undistinguished moth, most unlikely to be

Journal of Insect Science | http://www.insectscience.org 12

Journal of Insect Science: Vol. 13 | Article 124 Cock found by casual observation and collecting. It Thus, the balance of evidence does suggest would have to be located through a study or that B. nuciferae is more likely to be an indig- survey of coconut pests that examined the in- enous species in Trinidad than an alien sects associated with the inflorescence. species. However, none of these points are Similarly, although at least 12 other species of conclusive. To resolve this issue, a study on Lepidoptera were reared in this study from the molecular genetics of B. nuciferae in Trin- coconut or other palm inflorescences, only 1 idad (and elsewhere) would be needed. If B. had previously been recorded from Trinidad. nuciferae is a recently introduced species in There is no evidence that any entomologists Trinidad, it should have a much-reduced ge- have studied what insects affect coconut inflo- netic diversity there compared to areas where Downloaded from https://academic.oup.com/jinsectscience/article/13/1/124/1752783 by guest on 29 September 2021 rescence in Trinidad prior to 2006. it is indigenous.

It is only relatively recently that the inflo- The need for control of Bactrachedra nucif- rescence of coconut has become easily erae accessible in Trinidad, thanks to the planting At the moment, the evidence available does of dwarf hybrid varieties. When most flower- not suggest that control of B. nuciferae should ing palms were too tall for casual observation, be a priority compared to other adverse factors there would have been little incentive to ex- affecting coconut and other palms in Trinidad, amine the inflorescence for insects. Even in an e.g., red palm mite (), red ring area well-endowed with economic entomolo- disease (caused by the nematode Bursaphe- gists such as , it was only in 1982 that lenchus cocophilus), and other diseases it became known that A. subrufella was a known in the region. The limited observations damaging indigenous moth pest of coconut in this study did not suggest that B. nuciferae inflorescence (Habeck and Nickersen 1982), causes direct damage or occurred at high demonstrating how easily small Lepidoptera enough densities to do so. The observations feeding on coconut inflorescence may be also showed that the caterpillars are attacked overlooked. by an Apanteles (sensu lato) sp. parasitoid, which could be quite common. Ants were ob- If B. nuciferae were an introduced species, it served carrying away B. nuciferae caterpillars would be more likely to have no associated in the laboratory. Other natural enemies may specialist natural enemies, whereas if it were well attack the eggs or pupae, but would not indigenous it would be more likely to have have been detected in this survey. Until them. This survey showed that B. nuciferae is demonstrated otherwise, it is anticipated that parasitized by an Apanteles (sensu lato) sp. on the action of natural enemies (predators, para- coconut. Apanteles (sensu lato) spp. are usual- sitoids, and diseases) would increase when the ly specialist parasitoids, so this evidence density of B. nuciferae increased, bringing supports an indigenous origin rather than an any incipient outbreak under control. In ex- exotic one. Finally, if B. nuciferae were re- treme cases, for example if damage is cently introduced in Trinidad, its distribution observed early in the cropping cycle of newly might be restricted, whereas it seems to occur planted dwarf varieties of coconut, further ob- throughout the island (F. Hosein, personal servations should be made, but as long as the communication). inflorescences are being visited by pollinators, action to control flower pests such as Ba-

Journal of Insect Science | http://www.insectscience.org 13

Journal of Insect Science: Vol. 13 | Article 124 Cock trachedra spp. that do not attack the female International Agricultural Research, the UK's flowers are not likely to be useful. Department for International Development, and the Swiss Agency for Development and Acknowledgements Cooperation.

I thank the following for support and help References with aspects of this study: Ms. Velda Fergu- son-Dewsbury, Mr. Dwayne H. Burris, Mr. Arnal E, Clavijo J, Soto E, Ramos F. 1998. Farzan Hussein, Ms. Petal Ram, and staff of Batrachedra nuciferae Hodges, 1966 the Forestry Department and Royal Botanic (Lepidoptera: ) a new pest of Downloaded from https://academic.oup.com/jinsectscience/article/13/1/124/1752783 by guest on 29 September 2021 Gardens, Ministry of Food Production, Land coconut in Venezuela. Boletin de Entomologia and Marine Affairs of Trinidad and Tobago; Venezolana, Serie Monografias 13: 69–71. Dr. Elizabeth S Johnson, Ms. Suzy Wood, CABI; Dr. Yasmin Comeau and Mr. Winston Bondar G. 1940a. Insetos nocivos e moléstias Johnson, National Herbarium; Dr. Ron Bar- do coqueiro (Cocos nucifera) no Brasil. row, CARINET; Mr. Mike Rutherford, Tipographía Naval. University of the West Indies Zoological Mu- seum; Mr. Imran Gaffoor of Waller Field, Bondar G. 1940b. Notas entomológicas da Trinidad and Tobago; and Dr. M. Alma Solis Bahia. Revista de Entomologia, Rio de and Dr John W. Brown, United States De- Janeiro 11: 199–214. partment of Agriculture, Systematic Entomology Laboratory. Live and dead insect Carneiro JS, da Silva PHS, Nogueira CCP, de material was exported under permit from the Sousa HU. 2004. Morfologia externa, Plant Protection Organisation of Trinidad and biologia, danos e sugestões de controle da Tobago. Live material was imported to the traça-dos-frutos-docoqueiro Batrachedra UK and held in CABI’s UK quarantine facili- nuciferae. Comunicado Técnico, Ministério ty for emergence under licence from the da Agricultura Pecuária e Abasticimento Department for Environment, Food and Rural (Teresina, PI, Brazil) 159: 1–4. Affairs. The field visits to Nariva Swamp and Aripo Savannah were made under permit from Cock MJW, Gallego CN, Godfray HCJ. 1987. the Forestry Department, Trinidad and Toba- Biological control of Tirathaba rufivena in the go. This study was undertaken on behalf of . In: Ferrar P, Stechmann D-H, the Ministry of Food Production, Land and Editors. Biological control in the South Marine Affairs of Trinidad and Tobago, in Pacific. Report on an International Workshop support of their role in a United Nations Envi- held at Government Experimental Farm, ronment Programme – Global Environment Vaini, Kingdom of Tonga, 17-25 October Facility funded project “Mitigating the Treats 1985. pp. 147–157. Ministry of Agriculture, of Invasive Alien Species in the Insular Car- Fisheries & Forests of Tonga. ibbean,” project number GFL/ 2328–2713- 4A86 (PMS: GF/1030-09-03), which provided Comeau PL, Comeau YS, Johnson W. 2003. the financial support. CABI’s support in Trin- The Palm Book of Trinidad and Tobago idad and the preparation of this paper was including the Lesser Antilles. The funded by the CABI Development Fund, with International Palm Society. contributions from the Australian Centre for

Journal of Insect Science | http://www.insectscience.org 14

Journal of Insect Science: Vol. 13 | Article 124 Cock Corbett GH. 1931. The control and economic Transactions of the American Entomological importance of the greater spike moth Society 92: 585–651. (Tirathaba rufivena Walk.). Department of Agriculture of the Straits Settlements and Kaye WJ, Lamont N. 1927. A catalogue of the Federated Malay States, Scientific Series 8: Trinidad Lepidoptera Heterocera (moths). 1–14. Memoirs of the Department of Agriculture, Trinidad and Tobago 3: 144. Dransfield J, Uhl NW, Asmussen CB, Baker WJ, Harley MM, Lewis CE. 2008. Genera Lamont N, Callan E McC. 1950. Moths new Palmarum - The Evolution and Classification to Trinidad, B.W.I. Zoologica 35: 197–207. Downloaded from https://academic.oup.com/jinsectscience/article/13/1/124/1752783 by guest on 29 September 2021 of Palms. Kew Publishing. MALMR (Ministry of Agriculture, Lands and Ervig F. 1993. Notes on the phenology and Marine Resources). 2006. 1.1.8 Coconut moth pollination of the dioecious palms Mauritia (Batrachedra nuciferae). In: Ministry of flexuosa (Calamoideae) and Aphandra natalla Agriculture, Land and Marine Resources. (Phytelephantoideae) in . In: Barthlott Annual Report 2006. pp. 86–87. Ministry of W, Naumann C, Schmidt-Loske C, Agriculture, Land and Marine Resources, Schuchmann K, Editors. -Plant Trinidad & Tobago. Interactions in Tropical Environments. pp. 7– 12. Zoologisches Forschungsinstitut und MALMR (Ministry of Agriculture, Lands and Museum Alexander Koenig. Marine Resources). 2008. “Coconut Moth” Batrachedra nuciferae - A New Pest of Essig FB. 1971. Observations on pollination Coconuts. Coconut Moth Fact Sheet # 1. in Bactris. Principes 15: 20–24. Ministry of Agriculture, Lands and Marine Resources, Trinidad & Tobago – Entomology. Habeck DH, Nickerson JC. 1982. Atheloca subrufella (Hulst), a pest of coconuts Moore D. 2001. Insects of palm flowers and (Lepidoptera: : ). Florida fruits. In: Howard FW, Moore D, Giblin- Department of Agriculture and Consumer Davis RM, Abad RG, Editors. Insects on Services, Division of Plant Industry, Palms. pp. 233–266. CABI Publishing. Entomology Circular 241: 1–2. Robinson GS. 1976. The preparation of slides Henderson A. 1986. A review of pollination of with special reference studies in the Palmae. Botanical Review 52: to the Microlepidoptera. Entomologist’s 221–259. Gazette 27: 127–132.

Henderson A, Galeano G, Bernal R. 1995. Sánchez Soto S. 2004. Morphologia, biologia Field Guide to the Palms of the Americas. e fototropisma de Batrachedra nuciferae Princeton University Press. Hodges (Lepidoptera: ). MSc thesis, Universidade de São Paulo. Hodges RW. 1966. Review of the New World species of Batrachedra with description of Sánchez Soto S, Nakano O. 2008. Biología de three new genera (Lepidoptera: ). Batrachedra nuciferae Hodges (Lepidoptera: Coleophoridae), plaga del cocotero en

Journal of Insect Science | http://www.insectscience.org 15

Journal of Insect Science: Vol. 13 | Article 124 Cock América del Sur. Boletín del Museo de Entomología de la Universidad del Valle 9: 22–26.

Sánchez-Soto S, Nakano O. 2004a. Presença de Batrachedra nuciferae Hodges na cultura da coqueira no estado de São Paulo. Revista de Agricultura 79: 155–157.

Sánchez-Soto S, Nakano O. 2004b. Estudo Downloaded from https://academic.oup.com/jinsectscience/article/13/1/124/1752783 by guest on 29 September 2021 morfológico de Batrachedra nuciferae Hodges (Lepidotera: Coleophoridae). Neotropical Entomology 33: 173–178.

Santana SWJ. 2008. Criação e bioecologia de Atheloca subrufella (Hulst) (Lepidoptera: Phycitidae). PhD Thesis, Universidade Federal Rural de Pernambuco.

Santana SWJ, Torres JB, Gondim Jr MGC, Barros R. 2009. Infestation of coconut fruits by Aceria guerreronis enhances the pest status of the coconut moth Atheloca subrufella. Annals of Applied Biology 155: 277–284.

Tagestad AD. 1974. A technique for mounting Microlepidoptera. Journal of the Kansas Entomological Society 47: 26–30.

Taylor THC. 1930. Early nutfall from coconut palms in Fiji with special reference to insects attacking the flowers. Bulletin of the Department of Agriculture, Fiji 17: 1–42.

Waterhouse DF, Norris KR. 1987. 26. Tirathaba rufivena (Walker). In: Waterhouse DF, Norris KR, Editors. Biological Control: Pacific Prospects. pp. 211–218. Inkata Press.

Zona S. 2000. 201. Arecaceae Schultz- Schultzenstein. In: Flora of North America Editorial Committee, Editors. Flora of North America North of , volume 22. pp. 95– 123. Oxford University Press.

Journal of Insect Science | http://www.insectscience.org 16