Molecular Identification of Mealybug Species (Hemiptera: Pseudococcidae) Affecting Theobroma cacao for Improved Pest Management

: Theobroma cacao is affected by viruses on every continent where the crop is cultivated, with the best-known ones belonging to the Badnavirus genus. Badnaviruses are transmitted by several species of Pseudococcidae, a large, taxonomically diverse group of insects collectively known as mealy-bugs. Effective management of mealybugs depends on accurate identification of species present, as even closely related species have distinct life cycles and are vulnerable to different biological control organisms. This study compares the usefulness of the COI, ITS2, and 28S markers using the primer pairs (MFCO1/MRCO1, ITS2-M-F/ITS2-M-R, D10F/D10R, and D2F/D2R) to identify mealybugs associated with cacao plants in North America. All markers were informative for Pseudococcus com-stocki (n=4) and Maconellicoccus hirsutus (n=8), but only CO1 provided unambiguous identification for Pseudococcus jackbeardsleyi (n=11). Primer pair D2F/D2R is not recommended for mealybug identification, as it frequently yielded sequences of Anagyrus sp ., an Encyrtid parasitoid wasp commonly used for biocontrol. This study describes molecular diagnostic protocols for identifying cacao-asso-ciated mealybugs and detecting the presence of certain parasitoids. This information is essential for selecting the most effective interventions as part of an integrated pest


Introduction
Mealybugs (Hemiptera: Pseudococcidae) are phloem feeders that use long, slender mouthparts to uptake plant fluids [1], which reduces the vigor of host plants. They can feed on all plant parts, and severe infestations cause defoliation and, eventually, plant death. Some species inject plant toxins during feeding causing twisted/stunted growth [2]. The damage generated varies among taxa and is determined by their reproductive potential, temperature tolerance, preferred feeding locations, the existence of effective control strategies, and their ability to transmit viruses [3].
On cacao, the primary economic impact of mealybugs is their ability to transmit viruses [4]. Viruses have been identified on every continent where cacao is grown commercially, most of which belong to the Badnavirus genus and are transmitted by several mealybug species [5,6]. Due to the high diversity of Pseudococcidae, each region has different species composition, and morphological differentiation of closely related species is challenging for non-specialists. On cacao, mealybug populations are composed of multiple species [4,7,8]. In West Africa, Pseudococcus njalensis (Laing) is the main vector, due to its  abundance in the area [9], while Planococcus citri (Risso) is the most significant vector of cacao virus in Trinidad [4].
Accurate taxonomic identification is a major barrier in research and management of mealybugs due to the specificity of commonly used biological control organisms [10]. Molecular approaches have been used successfully to identify mealybugs from France, Egypt, South Korea, South Africa, and Japan [11][12][13]. In addition to reducing the reliance on delicate morphological features, sequence-based identification also allows for the detection of cryptic species [14]. Genetic sequences from these projects have been deposited in Gen-Bank, providing valuable reference material needed for the implementation of routine molecular identification for insects.
The purpose of this study was to develop molecular tools to determine the species composition of mealybug populations. This was done by evaluating the ability of three genetic markers and four primer pairs to identify the species affecting T. cacao (Linnaeus) in Florida. This information is essential for selecting the most effective management interventions.

Insect sampling and DNA extraction
In January 2021, female mealybugs were collected from pods, stems, leaves and flowers of four randomly selected Theobroma cacao trees in a greenhouse in Miami, FL. Up to 5 specimens were collected from each tissue type (pods, stems, leaves and flowers), and stored in 70% ethanol at 23°C (±1°C) for three to four weeks before processing.
DNA was extracted from individual specimens using the Qiagen DNeasy Blood and Tissue Kit with a shortened, 10min, lysis step, as described in Albo et al. (2019) [15]. The final resuspension step was done with 50 µ L of elution buffer for adults, and 30µ L for the smaller nymphs. A Qubit 4 Fluorometer and the 1x dsDNA High Sensitivity Assay Kit (Life Technologies Corp., Carlsbad, CA, USA) were used to quantify DNA.
Amplification was visualized on a 1% (w/v) agarose gel at 150V (for 35 minutes). PCR products were purified with Qiagen PCR Purification Kit (Hilden, Germany) and sent to Eurofins for Sanger sequencing. Table 1. Genes, primer sequences, and amplicon sizes for the markers targeted in this study.

Gene
Primer

Mealybug identification
The protocols developed during this study were tested on 23 mealybug specimens feeding on cacao in Florida. Sequences were edited and aligned using Geneious ® 11.1.2 (Biomatters Ltd., Auckland, New Zealand), and a subset was deposited in GenBank. The usefulness of primers pairs for mealybug identification was determined by analyzing the resulting sequences in BLASTn. Top matches were selected based on max score, and if multiple species were among the top matches for a given sequence, the first two to three entries were recorded in the results table. Specimen identification was determined based on BLASTn results of the COI sequences, because this genetic region is considered the most biologically informative.

Results
Three primer pairs (MFCO1/MRCO1, ITS2-M-F/ITS2-M-R, D10F/D10R) produced clear, single bands in all species tested, and high-quality sequences (>75%) after aligning and editing. Sequences obtained with all three primer pairs yielded consistent species matches for Pseudococcus comstocki (Kuwana) (n=4) and Maconellicoccus hirsutus (n=8), with coverage and identities ranging from 96-100% (Table 2). However, amplification with D10F/D10R was only achieved in half of the P. comstocki samples. , only CO1 sequences provided unambiguous identification. Sequences obtained with the D10F/D10R primer pair were close matches to three different species (Pseudococcus viburni (Signoret), Oracella acuta (Lobdell), and P. jackbeardsleyi) available in GenBank. In contrast, no highly similar sequences were found for ITS2 in GenBank. For each gene region, multiple alignments showed no nucleotide level variation among individuals of the same species.
Primer pair D2F/D2R is not recommended for mealybug identification, as it frequently yielded sequences of Anagyrus sp. (Howard) (Hymenoptera: Encyrtidae), parasitoids commonly used for biocontrol. Although these primers were developed for use in Hymenoptera [17], they were selected for this study because later work [8] found them to be more effective than D10F/D10R at amplifying the 28S region of Pseudococcidae. A subset of mealybug and parasitoid sequences generated in this study were deposited in Gen-Bank (Tables 3 and 4).

Discussion
Cytochrome c oxidase subunit I (COI) is considered the most informative marker for insects and most living organisms [18]. However, it has proven difficult to amplify in some groups, such as mealybugs, leading to the development of numerous different primer pairs [13,14]. The COI primers used in this study were designed by [8] and validated on taxa collected from cacao in Asia, Africa, and the Americas (Planococcus, Ferrisia, Dysmicoccus, and Pseudococcus). It amplifies a small section of the universal barcode region [18], but this fragment provided unambiguous identification in the species examined here.
Both P. jackbeardsleyi and M. hirsutus, have been reported affecting cacao in Africa and the Americas [19,20]. Neither has been tested for their ability to transmit viruses to cacao, but they are closely related to confirmed vectors, and have been detected on CSSV-infected cacao in Cote d'Ivoire [19]. Pseudococcus comstocki was one of the first confirmed vectors of a cacao virus in Trinidad [4] but no reports were found of it being used in transmission tests for other cacao viruses.
Accurate identification of species present in a population is essential for the selection of effective controls. Several parasitoids and predators are commercially available for controlling mealybugs, but they are not effective against all taxa. For example, a study investigating target species of the parasitoid Anagyrus sinope (Noyes and Menezes) found that it could only parasitize one of the five Pseudococcus species tested [10].
Anagyrus kamali (Moursi) has been effectively used to control M. hirsutus, in Egypt and the Caribbean [21,22], however, it is fairly host specific. Following the release of A. kamali in the Caribbean, Sagarra et al. (2001) conducted host range studies using nine mealybug species prevalent in the area [23]. Although A. kamali occasionally laid eggs in species other than M. hirsutus, it could not complete its life cycle on these hosts.
Parasitoid establishment and survival are additional obstacles to the successful use of biological control. The detection of Anagyrus sp. and A. kamali sequences in this study indicates that these organisms are established in the area, and additional releases would have little effect on pest populations.
The implementation of routine molecular identification for insects, relies on the availability of high-quality reference libraries against which sequences can be compared [24]. Low representation of an organism in GenBank results in ambiguous, or incorrect, identification [15]. In this study, ITS2 sequences were not considered informative for P. jackbeardsleyi, due to the absence of these sequences in GenBank. However, the ITS2 sequences generated in this study were deposited in GenBank, making this marker valuable for future research.
Although the tools presented here were developed for cacao mealybugs, the species detected are highly polyphagous and affect a wide range of geographic locations. These protocols can be used by agricultural inspectors and scientists to identify mealybug specimens and study pest populations.