Population genetic structure of Haplaxius crudus, the vector of palm lethal decline phytoplasmas, and planthopper diversity on coconut palms in Trinidad and Tobago (Hemiptera: Cixiidae, Delphacidae, Derbidae)

Abstract

Planthoppers are a diverse and interesting group of insects. In the tropics, there are many undiscovered species and recent efforts have uncovered many new taxa associated with palms. One species of Cixiidae, Haplaxius crudus (Van Duzee), is economically important due to its ability to transmit palm lethal decline phytoplasmas, which has caused significant economic losses to palm producers across the Caribbean and the United States. A survey was conducted in Trinidad and Tobago to assess the potential threat that lethal yellowing and lethal bronzing pose to Trinidad and Tobago by evaluating the status of Haplaxius crudus (and other putative vectors) on the islands and genetically characterize populations in the islands to determine if H. crudus was a distinct haplotype. Populations of H. crudus were sampled and analyzed to assess genetic variability. Specimens of H. crudus were homogenous for the COI gene but represent a novel haplotype of the species. All specimens of H. crudus were negative for phytoplasma. Additionally, 8 species of Derbidae were documented, including one new species, Oropuna tobagoensissp. n., 5 species of Cixiidae and one species of Delphacidae on coconut palms. These data provide a valuable baseline to aid in the development of a long-term, robust monitoring program that will allow for rapid and early detection of palm phytoplasmas should they be introduced to the islands and ultimately help prevent the establishment of the disease and economic losses to the coconut industry.

Introduction

The coconut palm (Cocos nucifera (L.)) is an economically important crop in the tropics that has been grown commercially in Trinidad since the early 1900s. Currently in Trinidad, there are approximately 2,500 hectares of production area that yields 62.9 kg ha⁻¹ (15,562.3 t) as of 2021 (FAO 2021). Given the economic importance of coconut to the national economy of Trinidad and Tobago, pest and pathogen surveillance and management is a critical component to protecting sustainable production in the region.

Historically, one of the most significant diseases in Trinidad has been red ring disease which is caused by the nematode Bursaphelenchus cocophilus (Cobb) and is transmitted primarily by the palm weevil, Rhynchophorus palmarum (L.) (Griffith 1968). This disease causes significant mortality, responsible for approximately 35% loss of young coconut palms in Trinidad and one report of 80% loss in Tobago (Brammer and Crow 2003). While this pathogen is currently the most significant threat to coconut production in Trinidad, other pathogens that cause significant mortality in the Caribbean have the potential for introduction to Trinidad, putting more strain on production. Arguably the most significant disease of coconut in the Caribbean is lethal yellowing (LY), a lethal decline caused by phytoplasmas (“Candidatus Phytoplasma palmae”) (Arellano and Oropeza 1995). Since the first documented report of LY in Montego Bay, Jamaica (Fawcett 1891), the disease has been recorded as far north as Florida, USA (Martinez and Roberts 1967), in southern Mexico (Oropeza et al. 2011) with the southern and easternmost record occurring recently in Guadeloupe (Pilet et al. 2023). The vector of LY phytoplasmas, Haplaxius crudus (Van Duzee) (Hemiptera: Cixiidae), is one of the most widespread and abundant planthopper species in the neotropics, being prevalent in all areas where LY is endemic (Humphries et al. 2021). However, H. crudus has been documented outside the range of LY and is also known from Colombia (Beltran-Aldana et al. 2020), Venezuela (Kramer 1979, Bourgoin 2024), Trinidad (Fennah 1945), and Costa Rica (Humphries et al. 2021). Given the widespread distribution of H. crudus in the Caribbean and the fact that LY (and a related strain of palm infecting phytoplasma that causes lethal bronzing—LB) has been introduced multiple times to regions without previously known infections (Harrison et al. 2008) highlight the risk that LY poses to countries and regions that have production of coconut. Because of this, monitoring for the presence of LY and the vector is critical for risk assessment and protecting the sustainability of coconut production in Trinidad and Tobago.

Recently, H. crudus from Florida was shown to be represented by 4 distinct haplotypes, with the majority of the population (≈95%) being represented by a dominant haplotype with 3 additional, rare haplotypes (Humphries et al. 2021). Additionally, populations from Costa Rica and Colombia were shown to be distinct haplotypes from Florida haplotypes and from each other (Humphries et al. 2021). The presence of multiple haplotypes (based on the 5' region of the COI gene) throughout the region provide a unique opportunity to develop new monitoring tools that are based on the distinct populations haplotype. By understanding the genetic structure of a local population, using these molecular data, it would be possible to determine if an adventive population from elsewhere had been introduced. In instances where data exist, it may be possible to link these populations to a source region and if not, at the very least determine if a population had been introduced. This is valuable in that populations from areas where LY or LB are endemic could be introduced (by humans or naturally) and carrying the phytoplasmas and potentially causing an outbreak of the disease.

The primary objective of this study was to survey coconuts in Trinidad and Tobago for the presence of H. crudus and analyze the genetic structure based on the cytochrome c oxidase subunit I (COI) barcoding region (5'-half), which has previously been demonstrated to distinguish populations/haplotypes of H. crudus (Humphries et al. 2021). The hypothesis for this objective is that given the biogeography of Trinidad and Tobago, H. crudus will be genetically homogenous for COI (small, island population) and that it will be genetically distinct from currently known populations. A secondary objective of the study was to catalog other planthoppers associated with coconut palms on the island. This will provide valuable data to government agencies to more rapidly implement monitoring/management strategies to reduce economic losses.

Materials and Methods

Site Selection and Sampling

Field sites were selected by anonymous growers with large-scale coconut production volunteering to participate in the survey with the collaboration of the Ministry of Agriculture (Fig. 1). A total of 6 sites in Tobago (Fig. 2) and 11 sites were surveyed in Trinidad (Fig. 3, Table 1). Selected sites were surveyed by sweeping accessible coconut palms with a standard insect net (net diameter 46.7 cm) (BioQuip). Sites were surveyed twice, once in May of 2022 and again in December of 2023. Accessible palms are defined as anywhere the coconut canopy was accessible for sweeping without the aid of climbing gear. Palms were sampled for no more than 1 h per site. All planthoppers collected were aspirated directly from the net and immediately transferred to vials with 95% ethanol. Specimens were exported under phytosanitary certificate number TT-E3P102VK3MX1KF and imported into the United States under permit number P526P-20-00214. All specimens are stored at the University of Florida’s Fort Lauderdale Research and Education Center (FLREC). Maps of sampled field sites were generated using SimpleMappr (Shorthouse 2010).

Specimen Identification and Photography

All specimens were sorted to family and subsequently morphospecies. Species identifications were accomplished by comparing to original descriptions in Fennah (1945) and available voucher specimens in the FLREC collection. Specimens were prepared for photography by placing them in a petri dish filled with hand sanitizer and then covered with 85% ethanol. Specimens were photographed using a VHX-7000 Digital Microscope (Keyence, Ft. Lauderdale, Florida, USA). Voucher specimens of each species were subsequently preserved in 95% EToH at −80 °C at FLREC.

DNA Extraction, PCR Parameters, and Sequence Analysis

All species collected had total DNA extracted by excising a single leg and processed using the DNeasy Blood and Tissue Kit (Qiagen, Hilden, Germany). The quality and yield of the final eluate were assessed using a NanoDrop Lite Spectrophotometer (ThermoFisher, Waltham, Massachusetts, USA). Loci selected, amplified and analyzed in this study for collected taxa include; the 5 prime (5') (750 bp expected) and 3 prime (3') (750 bp expected) regions of the cytochrome c oxidase subunit I (COI) gene, 18S rRNA gene (1,500 bp expected), the D8 (770 bp expected), D9 and D10 (780 bp expected) expansion region of the 28S rRNA gene and histone 3 (H3) gene. Primers and primer-specific PCR parameters are presented in Table 2. PCR reactions contained 5x GoTaq Flexi Buffer, 25 mM MgCl₂, 10 mM dNTP’s, 10 mM of each primer, 10% PVP-40, and 2.5U GoTaq Flexi DNA Polymerase, 2 µl DNA template, and sterile dH₂0 to a final volume of 25 µl. Thermal cycling conditions for all loci involved were as follows: 2 min. initial denaturation at 95 °C, followed by 35 cycles of 30 s. denaturation at 95 °C, 30 s annealing and extension at 72 °C. Products were visualized on a 1.5% agarose gel stained with GelRed (Biotium). PCR products of the appropriate size were purified using the ExoSAP-IT Express PCR Product Cleanup Reagent per the manufacturer’s protocol (ThermoFisher Scientific, Waltham, MA, USA). The purified PCR product was quantified using a NanoDrop Lite Spectrophotometer (ThermoFisher Scientific, Waltham, MA, USA) and sequenced using the SeqStudio Genetic Analyzer (Applied Biosystems). Contiguous files were assembled using DNA Baser (Version 4.36) (Heracle BioSoft SRL, Pitesti, Romania). Maximum likelihood (ML) analysis for phylogenies generated for placement of taxa collected in this study were generated in IQ-TREE (Minh et al. 2020) using a concatenated matrix of COI (5' half), 18S rRNA and the D9-D10 expansion region of 28S rRNA for derbids and COI (5' and 3' halves), H3, D8 to D10 expansion regions of 28S and 18S for cixiids. The matrix was partitioned (Chernomor et al. 2016) by marker and the best-fit partitioning scheme and model was selected according to the Bayesian information criterion (BIC) score in ModelFinder (Kalyaanamoorthy et al. 2017) implemented in IQ-TREE. Substitution models selected and applied for each partition of the combined data matrix to ML analysis were (1) 18S + 28S: TIM3 + F + I + G4 and (2) COI: GTR + F + I + G4 for derbids and (1) H3: TN + F + R3, (2) COI3P + COI5P: TIM2 + F + R4 and (3) 28SD8 + 28SD9D10 + 18S: GTR + F + I + G4 for cixiids. Clade support was calculate by 1,000 replicates of Shimodaira-Hasegawa approximate likelihood ratio test (SH-aLRT; Guindon et al. 2010, Hoang et al. 2018) in ML analysis. Reliable support values are SH-aLRT ≥ 80 and UFboot ≥ 95 (Guidon et al. 2010, Minh et al. 2013). The resulting trees were viewed on FigTree v1.4. Accession numbers for all loci for species collected in this study and publicly available references are presented in Supplementary Table 1.

The 5' region of COI was selected as the marker to analyze the genetic variability of H. crudus within and among sites. The primers used and primer-specific PCR parameters are presented in Table 2 and PCR reaction concentrations are the same as listed above. PCR products of the appropriate size were purified using the ExoSAP-IT™ Express PCR Product Cleanup Reagent per the manufacturer’s protocol (ThermoFisher Scientific, Waltham, MA, USA). The purified PCR product was quantified using a NanoDrop Lite Spectrophotometer (ThermoFisher Scientific, Waltham, MA, USA) and sequenced using the SeqStudio Genetic Analyzer (Applied Biosystems). Contiguous files were assembled using DNA Baser (Version 4.36) (Heracle BioSoft SRL, Pitesti, Romania), and aligned using ClustalW as part of the package MEGA7 (Kumar et al. 2016). Sequence polymorphism (S), number of haplotypes (h), nucleotide diversity, and number of nucleotide differences (k) were determined using DnaSP v.6.12.03 (Rozas et al. 2017). A Maximum Likelihood tree with bootstrap support (1,000 pseudoreplicates) in MEGA7 was generated with resulting haplotypes and included H. crudus haplotypes presented in Humphries et al. (2021), with other species of HaplaxiusFowler, 1904 included as outgroups to establish interspecific species differences relative to intraspecific variability.

To screen specimens of H. crudus for the LY phytoplasma, the template used for barcoding was also tested using the protocol outlined by Bahder et al. (2017) and Cordóva et al. (2014)

Nomenclature

This paper and the nomenclatural act(s) it contains have been registered in Zoobank (www.zoobank.org), the official register of the International Commission on Zoological Nomenclature. The LSID (Life Science Identifier) number of the publication is urn:lsid:zoobank.org:act:17F61373-01AF-4310-9760-6C8CE9D39648

Results

Abundance, Distribution, and Genetic Variability of Haplaxius crudus and Phytoplasma Screening

Across all sites in Trinidad and Tobago, a total of 67 specimens of H. crudus (Fig. 4, Table 3) were collected. Three specimens of H. crudus were collected from 2 sites in Tobago, with the remaining 64 specimens being collected from 6 of the 11 sites in surveyed in Trinidad (Table 3). Amplicons of the 5' region of COI were generated and sequenced for all 67 specimens of H. crudus collected in this survey. After assembly and cleaning sequences, 529 bp were included for analysis. Across all sites where H. crudus were collected, all 67 sequences were 100% identical for the 5' region (0.0 polymorphisms (S), 0.0 nucleotide diversity (p)), resulting in a single haplotype (h = 1). Sequences were thus condensed to a single representative sequence HcrTT_1 to compare to haplotypes from other regions. The molecular phylogeny comparing HcrTT_1 to haplotypes from Florida, USA (HcrFL 1, HcrFL 2, HcrFL 3, and HcrFL 4), Texas, U.S.A. (HcrTX 1), Costa Rica (HcrCR 1) and Colombia (HcrCOL 1) demonstrated that while H. crudus in Trinidad is a homogenous population (based on COI sequences), it represents a distinct and novel haplotype that displays a normal level of deviation from other populations of H. crudus that is representative of intraspecific variability (Fig. 5). Accession numbers representing HcrTT_1 for all loci are presented in Supplementary Table 1. Accession numbers for COI data for all 67 specimens are PQ433589–PQ433655.

All 67 specimens of H. crudus that were barcoded and analyzed tested negative (No Ct) for the LY, LB, and 16SrIV-E phytoplasma.

Cixiidae Abundance and Diversity in Trinidad and Tobago

In addition to H. crudus, 3 other species of cixiid were documented from coconut palm in Trinidad; Melanoliarus castro Bahder & Bartlett 2023 (Fig. 6), M. kindli (Bourgoin, Wilson & Couturier) (Fig. 7), and Pintalia albolineata (Muir) (Fig. 8), resulting in 4 total species collected from coconut. The only species present at multiple sites was H. crudus, with other species found at a single site. After H. crudus, the most abundant species was M. castro (n = 15), followed by M. kindli (n = 7) and P. albolineata being the least abundant (n = 5) (Table 3).

Delphacidae Abundance and Diversity in Trinidad and Tobago

In Tobago, a single species of delphacid was collected from coconut palm, Peregrinus maidis (Ashmead) (The corn planthopper, Fig. 9). A single female was collected at TBG-1, TBG-2, and TBG-3, for a total of 3 specimens.

Derbidae Abundance and Diversity in Trinidad and Tobago

In Tobago, 5 different species of derbid were collected from coconut palm; Cedusa hampora (Kramer) (Fig. 10), Omolicna cubana (Myers 1926) (Fig. 11), O. proximaFennah 1945 (Fig. 12), Persis gregaria (Fennah) (Fig. 13), and an undetermined species of Oropuna (Fennah). Of these, O. proxima was the most abundant (n = 17) and widespread, having been collected at 5 of the 6 sites (Table 4). The second most abundant was O. cubana (n = 14), followed by Oropuna sp. (n = 10), Persis gregaria (n = 2), and Cedusa hampora being the least abundant (n = 1) (Table 3). In Trinidad, 6 species of derbid were collected from coconut palms: Agoo argutiola (Bahder & Bartlett) (Fig. 14), Cedusa hampora, C. yowza (Kramer) (Fig. 15), Derbe uliginosa (Fennah) (Fig. 16), O. cubana (Fig. 11), and O. proxima (Fig. 12). The most abundant species was O. cubana (n = 138), followed by A. argutiola (n = 30), C. yowza (n = 9), C. hampora (n = 7), D. uliginosa (n = 5) with O. proxima being the least abundant (n = 1) (Table 4). Of these derbid species, both A. argutiola and O. cubana were widespread, being collected from 5 of the 11 sites surveyed in Trinidad (Table 4). Additionally, all specimens of A. argutiola and O. cubana were found at the same 5 sites (Table 4). Cedusa hampora was collected at 2 sites while C. yowza, D. uliginosa and O. proxima were only collected from a single site (Table 4). The undetermined species of Oropuna (Fig. 17) was found to be a new species.

Description of New Species of Oropuna

Systematics
Family Derbidae Spinola, 1839,
Subfamily Derbinae, Spinola, 1839,
Tribe Cenchreini Muir, 1913,
Genus OropunaFennah, 1952
Type Species: Oropuna minutiana (Caldwell)
Oropuna tobagoensis Bahder & Bartlett sp. n.
(Figs. 17–22)

Type Locality.

Tobago, Mt. St. George

Diagnosis.

Moderate sized (5.5-6.0 mm), body light orange, wings fuscous with apical margins red. Lateral margins of frons nearly parallel, lacking median carina. Sensory pits long lateral margins of vertex and frons. Medioventral process of pygofer rounded, slightly wider (at base) than tall. Aedeagus nearly symmetrical, complex, small, fin-like process on sides near apex, endosoma heavily serrated, running along entire dorsum and ventrally.

Description.

Color. Ground color uniform light orange, legs (especially coxae) slightly paler (Fig. 17); forewings fuscous, slightly darker in distal half, apical margin and adjoining veins red.

Structure. Moderate-sized, males ~5.5 mm with wings, 2.8 mm without and females ~6.0 mm with wings, 3.1 mm without (Table 5). Head. In dorsal view (Fig. 18A), vertex trapezoidal, disc medially concave, anterior margin truncate, posterior margin smoothly concave, ~1.3× wider at posterior margin than long at midpoint; 2 rows of large sensorial pits on either side of midline, in irregular lines with smaller pits medially and near apex; in lateral view (Fig. 18B), head profile smoothly rounded, extending slightly above and beyond eyes; in frontal view (Fig. 18C), lateral margins of frons subparallel, row of large sensorial pits on lateral margins, extending from fastigium to frontoclypeal suture, interspersed with smaller pits; median carina absent. Eyes semicircular (Fig. 18B, 18C), emarginate near antennae. Ocelli below eye (diagonally below antennae). Antennae short, scape ring-like, pedicel globular, slightly taller than wide, bearing many sensory plaques.

Thorax. Pronotum in dorsal view (Fig. 18A) at midline about half length of vertex; anterior margin convex, posterior margin angularly concave; tricarinate, lateral carinae arched laterally, becoming foliate in paradiscal region to form large lateral fossae, partially surrounding antennae. In lateral view, pronotum anteriorly declinate. Mesonotum large, approximately twice combined length of vertex and pronotum at midline, slightly wider than long; tricarinate, median carina weak, extending from anterior to posterior margin, lateral carinae sinuate, becoming obsolete near posterior margin. Forewing broad (Fig. 19), elongate-elliptical, distinct pits along veins Pcu and ScP + R; minute pits tracing most other veins; apex of clavus near wing midlength, composite vein Pcu + A1 reaching CuP before wing margin, forks of CuA and RP from ScP + RA at same level, well proximade of Pcu + A1 fusion; branching pattern RA 2-branched, RP 2-branched, MP 5-branched, CuA 2-branched. Lateral margin of hind tibiae without lateral spines, apical ornamentation variable (6-6-6, 6-6-7, or 6-7-7).

Terminalia. Pygofer in lateral view (Fig. 20A) narrow, irregular in shape, narrowest dorsally, expanded ventrad (constricted just past midpoint), widest near ventral margin, anterior and posterior margin smoothly sinuate; in ventral view (Fig. 20B), medioventral process dome-shaped, slightly wider (at base) than tall. Gonostyli in lateral view (Fig. 20A) club-like, widest subapically, ventral margin with subbasal constriction, smoothly arched to rounded, apex with median invagination, middorsal margin bearing large sclerotized process, hooked at apex. Aedeagus complex (Fig. 21), bilaterally symmetrical, shaft weakly upcurved; bearing pair of large, bifid processes arising at apex (A1 & A2), angled dorsad and cephalad, dorsal bifurcation (A1a & A2a) spinose, slender; ventral bifurcation (A1b & A2b) slightly shorter and more robust (Figs 21 & 22) subapical pair of processes (A3 & A4) arising laterally, flange-like, serrated; third pair of processes (A5 & A6) arising at midpoint, angled dorsad and cephalad, tubular basally, expanding and flattening at midpoint, distal half fin-like, anterior margin serrate, final pair of processes (A7 & A8) arising at base of A5 and A6, spinose, angled laterally and cephalad (Figs 21 & 22); endosoma complex, large, flattened pair of processes (E1 & E2) arising apically, angled cephalad, extending to aedeagal base, serrated and folded along dorsal margin, ventral margin with round lobe at apex, angled ventrad (Figs 21 & 22), second pair of processes (E3 & E4) arising from E1 and E2 in basal 1/4, elongate, slender, spinose, extending to midpoint of E1 and E2 (Figs 21 & 22); median process (E5) arising subapically, more robust than E3 and E4, spinose, flattened lobe on dorsal margin, extending to same point as A1 and A2. Anal segment in lateral view (Fig. 20A) exceeding aedeagus but not reaching apex of gonostyli, narrowly triangular in shape;, dorsal and ventral margins gently curved, ventral margin slight sinuate in distal 1/3, apex directed caudad, nearly runcate; in dorsal view (Fig. 20C), narrow and elongate, weakly spatulate distally; paraproct short.

Remarks.

Oropuna tobagoensissp. n. is the fifth species placed in Oropuna. The described species are distributed from southern Mexico (Chiapas; Caldwell 1944, viz. Oropuna minutiana) to Southeastern, Brazil (Stål 1862, viz. Oropuna orba (Stål). The male terminalia for most species (except O. orba, which was described from a female, Bahder et al. 2021b, fig. 8) have been at least partially described and differ markedly in the shape of the medioventral process of the pygofer, the gonostyli, and apparently the anal tube. Oropuna tobagoensissp. n. can be distinguished from congeners most readily by the dome-shaped medioventral process, compared to the broad, quadrate process of O. halo Bahder & Bartlett (Bahder et al. 2021b, fig. 4B), and O. minutiana (Caldwell 1944, plate 1, Fig. 2C), and the elongate process of O. fusca (Metcalf) (Bahder et al. 2021b, Fig. 10D). The species also appear to differ by geography and color (although comparison between fresh specimens and very old collections makes details of color comparision suspect.

Molecular Phylogeny of Planthoppers Collected

For all cixiid species and Peregrinus maidis, all loci were successfully amplified and sequenced (Supplementary Table S1). For derbids, only the 5’ region COI, 18S and the D9-D10 of 28S were successfully amplified and sequenced (Supplementary Table S1). The molecular phylogeny for the Derbidae demonstrated that the species found in this study resolved within their respective genera based on available sequence data (Fig. 23). Omolicna cubana and O. proxima resolved adjacent to each other with high clade support as sister to the remaining sampled Omolicna species, with maximum clade support (SH-aLRT = 100; UFBoot = 100) for the monophyly of Omolicna based on current markers and available taxa (Fig. 23). Persis gregaria was found included in a clade of species of Persis (non-monophyletic) + Neocenchrea heidemanni with maximum clade support (SH-aLRT = 100; UFBoot = 100) based on the loci analyzed and available taxa (Fig. 23). Agoo argutiola resolved within the genus Agoo with maximum clade support (SH-aLRT = 100; UFBoot = 100) (Fig. 23). Both species of Cedusa (C. hampora and C. yowza) were found as sister taxa with strong clade support (SH-aLRT = 94.1; UFBoot = 99) and within the clade of Cedusa species with high clade support (SH-aLRT = 93.2; UFBoot = 97) based on the loci analyzed and with available taxa (Fig. 23). The only species where congeners were unavailable for comparison was D. uliginosa, which resolved with high clade support (SH-aLRT = 94.1; UFBoot = 95) adjacent to the genus Tico (Fig. 23). There was also high clade support (SH-aLRT = 99.4; UFBoot = 100) for placement of the new species, O. tobagoensissp. n. within the genus Oropuna, resolving adjacent to O. halo (Fig. 23).

The molecular phylogeny for the Cixiidae and Delphacidae showed resolution of the collected species in this study in their respective genera (Fig. 24). There was maximum clade support (SH-aLRT = 100; UFBoot = 100) for the monophyly of the genus Haplaxius based on loci analyzed and available taxa (Fig. 24). Melanoliarus castro and M. kindli both resolved within the genus MelanoliarusFennah 1945 with maximum clade support (SH-aLRT = 100; UFBoot = 100) for monophyly of the genus (Fig. 24). Pintalia albolineata revolved adjacent to P. hannae with strong bootstrap support (SH-aLRT = 99.9; UFBoot = 100).

Discussion

The findings of this survey represent the first systematic survey of planthoppers on coconut palms in Trinidad and Tobago. The Cixiidae collected as part of this survey were less diverse and less abundant than the derbidae. However, the analysis of the local population of H. crudus is an essential component for developing a long-term monitoring program for LY. The widespread distribution of H. crudus and relatively high abundance in Trinidad is important because it is a vector of both LY (Howard and Thomas 1980) and lethal bronzing (LB), a closely related strain of palm lethal decline phytoplasma (Mou et al. 2022). Should either of these diseases become introduced in Trinidad and Tobago, the presence of the vector could allow for rapid spread and significant economic losses to the coconut industry. That no specimens of H. crudus were found to be carrying phytoplasma is encouraging. However, the proximity of Trinidad and Tobago to Caribbean nations with documented cases highlights the need for the results of this survey to be used to guide a robust monitoring program to ensure LY is not introduced, or if it is, can be identified quickly enough so that effective management and quarantine procedures can be implemented to prevent significant escalation of the disease that would undoubtedly cause severe losses in coconut production regions. The homogenous makeup of H. crudus based on COI is fascinating from a scientific standpoint in that it reveals that Trinidad appears to have an indigenous population, with currently no evidence of immigrant populations. Haplaxius crudus in Jamaica and Antigua have identical makeup as those in Florida (Humphries et al. 2021) and with new appearances of LY and LB in Florida in the 1930s and 2006, respectively, it is evident populations can move and become established, causing disease outbreaks. Cataloging the haplotypes of H. crudus, while interesting from the evolutionary perspective, also has potential for utility in monitoring programs where populations of H. crudus are consistently monitored for the presence of foreign haplotypes, and should one be identified from a region with disease, rapid testing/palm removal could be implemented to prevent outbreaks.

Derbidae were the most abundant and diverse group found associated with coconut. The findings here are similar to the findings by Segarra-Carmona et al. 2013 and Dollet et al. (2018) with 8 species documented. However, Dollet et al. (2018) included oil palms in the survey and one species, Cedusa yowza, was only documented from oil palm. This species was also collected in this survey from coconut, so it is likely it also could be found on coconut in Brazil. Another species collected in Trinidad that is also present in Brazil is Agoo argutiola. Given the proximity of Trinidad to Northern Brazil and Trinidad historically was contiguous with mainland South America (Arkle et al. 2017), it is not unexpected to find the same species in both regions. It is also likely that other species, both in Brazil and Trinidad, have broader geographic distributions than is indicated in the respective surveys. The discovery of a new species of planthopper, O. tobagoensissp. n., is not unexpected. Dollet et al. (2018) reported 2 new species in the corresponding survey (Agoo argutiola and A. spina Bahder & Bartlett, 2020) and recent efforts in Costa Rica and the Caribbean have documented many new species of derbid associated with coconut palms (Bahder et al. 2019, 2020, 2021a) but also a wide variety on native palms in natural ecosystems, one of which was also within the genus Oropuna (O. halo) (Bahder et al. 2021b).

These data will be valuable in a larger effort to monitor movement of H. crudus throughout the region and identify adventive populations, particularly if such a population appears in a region without documented cases of palm lethal decline phytoplasma and is determined to be from a region where the disease is endemic or naturally spreading. In addition, these data provide valuable insights into the biology and ecology of this interesting group of insects but also provide a foundation for developing novel and robust monitoring programs that can help protect sustainable coconut production in Trinidad and Tobago.

Version of Record, first published online April 30, 2025, with fixed content and layout in compliance with Art. 8.1.3.2 ICZN.

Acknowledgments

Funding for this research was provided by the European Union (agreement no. FED/2019/407372) with support from the International Trade Centre (ITC), Caribbean Agricultural Research and Development Institute (CARDI), the Caribbean Forum (CARIFORUM Secretariat), African Caribbean Pacific States (ACP), ALLIANCES FOR ACTION, Ministry of Agriculture Land and Fisheries—Trinidad and Tobago, the Division of Food Security Natural Resources, The Environment and Sustainable Development (DFSNRESD), and the Coconut Industry Board (CIB). The authors also thank Dr. Thomas Chouvenc for training and access to the digital microscope.

Author contributions

Brian Bahder (Conceptualization [equal], Data curation [equal], Formal analysis [equal], Investigation [equal], Methodology [equal], Writing—original draft [lead], Writing—review & editing [lead]), Wayne Myrie (Conceptualization [equal], Investigation [equal], Project administration [equal]), Melody Bloch (Data curation [equal], Formal analysis [equal], Investigation [equal], Methodology [equal], Writing—review & editing [equal]), Jeremy Lane (Data curation [equal], Formal analysis [equal], Investigation [equal], Methodology [equal]), Ericka Helmick (Data curation [equal], Formal analysis [equal], Methodology [equal], Writing—review & editing [equal]), Julia Parris (Conceptualization [equal], Investigation [equal], Methodology [equal]), Charles Bartlett (Data curation [equal], Formal analysis [equal], Methodology [equal], Writing—original draft [equal], Writing—review & editing [equal]), Amel Baksh (Investigation [equal], Methodology [equal], Project administration [equal]), Nadia Ramtahal-Singh (Conceptualization [equal], Investigation [equal], Methodology [equal]), Ian Mohammed (Conceptualization [equal], Data curation [equal], Investigation [equal], Methodology [equal], Project administration [equal], Writing—original draft [equal], Writing—review & editing [equal]), and Fayaz Shah (Conceptualization [equal], Project administration [lead], Supervision [lead])

Conflicts of interest. The authors declare they have no conflict of interest.

References