http://orcid.org/0000-0003-3657-8421
Figures
Abstract
Hypogeococcus pungens, a mealybug native of southern South America, is devastating native cacti in Puerto Rico and threatening cactus diversity in the Caribbean, and potentially in Central and North America. The taxonomic status of H. pungens is controversial since it has been reported feeding not only on Cactaceae but also on other plant families throughout its distribution range. However, in Australia, where the species had been exported from Argentina to control weedy American cacti, it was never found on host plants other than Cactaceae. These conflicting pieces of evidence not only cast doubt on the species identity that invaded Puerto Rico, but also have a negative impact on the search for natural enemies to be used in biological control programs against this pest. Here we present reproductive incompatibility and phylogenetic evidences that give support to the hypothesis that H. pungens is a species complex in which divergence appears to be driven by the host plants. The nuclear EF1α and 18S and the mitochondrial COI genes were used as markers to evaluate the phylogenetic relationships among H. pungens populations collected in Argentina, Australia and Puerto Rico feeding on Cactaceae and/or Amaranthaceae. Additionally, we conducted reciprocal crosses between mealybugs from both hosts. Species delimitation analysis revealed two well-supported putative species within H. pungens, one including mealybugs feeding on Amaranthaceae (H. pungens sensu stricto), and a new undescribed species using Cactaceae as hosts. Additionally, we found asymmetric reproductive incompatibility between these putative species suggesting recent reproductive isolation. The Bayesian species delimitation also suggested that the Australian mealybug population may derive from another undescribed species. Overall, the patterns of genetic differentiation may be interpreted as the result of recent speciation events prompted by host plant shifts. Finally, the finding of a single haplotype in the Puerto Rico population suggests only one invasive event. We still need to identify the geographical origin of the pest in order to enable the use of biological control to reduce the threat to cacti diversity in the Caribbean.
Citation: Poveda-Martínez D, Aguirre MB, Logarzo G, Calderón L, de la Colina A, Hight S, et al. (2019) Untangling the Hypogeococcus pungens species complex (Hemiptera: Pseudococcidae) for Argentina, Australia, and Puerto Rico based on host plant associations and genetic evidence. PLoS ONE 14(7): e0220366. https://doi.org/10.1371/journal.pone.0220366
Editor: Tzen-Yuh Chiang, National Cheng Kung University, TAIWAN
Received: April 15, 2019; Accepted: July 15, 2019; Published: July 25, 2019
This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.
Data Availability: All sequences were deposited in GenBank under accession numbers MN013440 - MN013743.
Funding: This study was supported by the USDA-APHIS Invasive Species Coordination Program from 2014 to 2016 (APH-HQ-16-0181), and USDA-APHIS Farm Bill funding for fiscal year 2017 (Farm Bill FY 17: 6.0385.01). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The mealybug Hypogeococcus pungens Granara de Willink (Hemiptera: Pseudococcidae), commonly called the Harrisia cactus mealybug (HCM), is devastating the native columnar, globular and semi-epiphytic cactus species of Puerto Rico, and threatening cacti throughout the Caribbean, and Central and North America including Mexico [1,2]. H. pungens, one of the members of the genus Hypogeococcus Rau which is native to the New World, was first recorded on cacti in Puerto Rico in 2005; by 2010 the insect had covered an area of 1400 km2 and is continuing to spread on the main island [3]. Taking into account that H. pungens is found at low densities in its native range where its natural enemies are present [4], biological control is the most appropriate management option to protect wild populations of cacti from this mealybug in large natural areas. Chemical control, however, may be feasible for small areas like nurseries and commercial cultivations [5].
H. pungens is native to South America, mainly Argentina, Bolivia, Brazil, Paraguay, Peru, and Uruguay [5]. The mealybug was originally described from collections on Alternanthera pungens Kunth (Amaranthaceae) in Argentina [6]. Outside of its native range, the mealybug has been reported in Australia, South Africa, United States of America, several European countries, and in the Caribbean [7–16]. The Australian and South African Hypogeococcus populations were deliberately introduced from Argentina for the biological control of several weedy cacti [15,17–18]. The host range of H. pungens sensu lato is constrained to members of the Amaranthaceae, Cactaceae, Portulacaceae and Euphorbiaceae [5, 19–20].
However, a word of caution is needed regarding the accuracy of insect-host-plant associations and geographical distribution because there is a certain degree of confusion about the taxonomic status of H. pungens. H. pungens has been misidentified as Hypogeococcus festerianus Lizer & Trelles (a valid species restricted to Cactaceae) in several publications [8, 21–25]. McFadyen & Tomley [22,23] and Tomley & McFadyen [8] reported the successful biological control of invasive cactus with the “Harrisia cactus mealybug” then identified as H. festerianus [18]. However, the identification of this mealybug was conducted by Williams [4] before the description of H. pungens. In fact, the species introduced into Australia was subsequently identified as H. pungens by Williams & Granara de Willink [1]. The paradox is that McFadyen [8] reported that the species introduced into Australia was only found in the field on Cactaceae host plants, in spite of the availability of members of Amaranthaceae and Portulacaceae families that are common in the wild. In addition, the Hypogeococcus species introduced into Australia did not develop in any of these families in the laboratory. In biological control publications, Australian researchers continue using H. festerianus instead of H. pungens because of the polyphagous behavior of the latter [8].
A recent study evaluating aspects of the biology of H. pungens revealed marked differences between mealybugs collected on Amaranthaceae (hereafter H. pungens sensu stricto) and on Cactaceae, giving strong support to the idea that H. pungens is, in fact, a species complex [26]. Even though the progeny of gravid H. pungens collected on Amaranthaceae in nature was able to complete development on cactus hosts, it could not produce a viable second generation on Cactaceae [26]. Moreover, differences in fecundity, female body length, mode of reproduction, rate of development, and host range were also observed between H. pungens sensu stricto and the mealybugs introduced into Australia [4]. In this sense, it is worth noting that H. pungens sensu stricto can reproduce by both sexual reproduction and parthenogenesis by deuterotoky, whereas cactus breeding bugs only reproduce sexually [26].
In 2010, as part of a biological control program against this cactus pest in Puerto Rico (USDA, ARS 2014), the Fundación para el Estudio de Especies Invasivas (FuEDEI) started surveying for H. pungens sensu lato in Argentina on host plants of Alternanthera spp., other Amaranthaceae species, and species of native cacti. The survey had two goals, to search for the most promising biological control agents, and to identify the population of Hypogeococcus spp. that matched the Puerto Rico cactus pest. Surveys began in Argentina because H. pungens was originally described from specimens collected in this country. Also, Argentina was the country with the most known Hypogeococcus mealybug species that infested Cactaceae: H. pungens as well as H. festerianus and Hypogeococcus spinosus Ferris both restricted to Cactaceae.
Finally, although Zimmermann et al. [5] suggested that the most likely dispersion of Hypogeococcus was via nursery trade and wind dispersal, we could not rule out that HCM populations of Puerto Rico arrived from Australia or South Africa, where “H. pungens” specimens had been introduced from Argentina.
In classical biological control programs, the correct identification of the target species and its natural enemies is a key issue [27]. The kind of biological control selected to manage the Puerto Rico pest strongly depends on the precise identification of the mealybug species pest. Here we address three basic questions using molecular markers and reciprocal cross-breeding experiments: 1) Is H. pungens a complex of closely related species, that would explain the differences observed in the host-range breadth? 2) Is the identity of the mealybug introduced into Australia for control of cactus weeds H. pungens, H. festerianus, or a new undescribed species? and finally, 3) Which is the source population of the Puerto Rico mealybug cactus pest?
Materials and methods
Samples analyzed
Field surveys of Hypogeococcus populations, considered to be H. pungens, were conducted in Argentina, Australia and Puerto Rico between 2010 and 2016. Our study did not involve any endangered or protected insect species. For the mealybugs collected from Argentine, we obtained the permission from the land owners and no specific permits were required in Australia. However, for the mealybugs collected from Puerto Rico, we obtained the permission from the U.S Fish & Wildlife Service (Permit No. 41522-16-003), and Department of Natural and Environmental Resources (Permit No. OV-1617-15). Mealybugs were collected on Amaranthaceae and/or Cactaceae, following the host range reported for H. pungens [1, 19–20]. 10–50 plants of each potential host were inspected for mealybugs. If mealybugs were present, 5–20 adult specimens (depending on abundance) were preserved in absolute ethanol for genetic survey. Our study encompassed sites across 14 Argentine provinces (Salta, Jujuy, Tucumán, La Rioja, Catamarca, Córdoba, Mendoza, Santiago del Estero, Formosa, Chaco, San Juan, San Luis, Entre Ríos, and Corrientes). Since H. pungens was misidentified as H. festerianus in the past, we included 10 H. festerianus females collected on Cereus aethiops Haworth (Cactaceae) in Mendoza Province, as well as 10 specimens collected on cacti in Australia whose taxonomic status is still unclear [8]. Fifteen adult females, collected in Puerto Rico and tentatively identified as HCM (5 ♀ from Cabo Rojo Municipality, 5 ♀ from Guánica Municipality, and 5 ♀ from Caja de Muertos Island, Ponce Municipality) were included in the study with the aim of identifying the source populations of the cactus pest (S1 Table). The vine mealybug, Planococcus ficus (Signoret) (Hemiptera: Pseudococcidae), was used as an outgroup in the phylogenetic studies. S1 Table shows the geographic coordinates of each sampling site along with other relevant information (host plant family, host plant species, and sample size) (Fig 1).
The map shows the locations sampled from the native (Argentina) and non-native (Puerto Rico) range of the mealybug H. pungens sensu lato, and the co-generic species H. festerianus.
DNA extraction, PCR amplification and sequencing
Genomic DNA extractions were performed using the whole female body with the Qiagen DNeasy Blood & Tissue Kit following manufacturer’s instructions (Qiagen, Inc., Valencia, California, USA). For the samples collected in Argentina and Australia, we obtained information on three different genes, the mitochondrial Cytochrome Oxidase subunit I (COI), and two nuclear genes: translation elongation factor (EF1α) and 18S Ribosomal RNA. For the samples from Puerto Rico, we only obtained the COI gene because it was much more informative than the other markers. The RNA 18S gene was amplified using primers 18S-2880 (5’-CTGGTTGATCCTGCCAGTAG-3) and 18S-B (5’-CCGCGGCTGCTGGCACCAGA-3’) [28]; EF1α using primers M51.6 (5’-CARGACGTATACAAAATCGG-3’) and rcM53-2 (5’-CAATGTGRGCIGTGTGGCA-3’) [29]; and COI with primers C1-J-2183 Jerry (5’-CAACATTTATTTTGATTTTTTGG-3’) [30] and C1-N-2568 BEN3R (5’-GCWACWACRTAATAKGTATCATG-3’) [31]. The PCR reaction components for all amplifications were 12.5 μL of Taq polymerase in the appropriate buffer (PCR Master Mix-Promega, Madison, WI, USA), 1 μL of each primer [5 mM], 1 μL of DNA template and 12.5 μL of nuclease free water to complete a total volume of 25 μL. The PCR cycling protocol for COI and 18S was as follows: an initial denaturing step at 98°C for 1 min, followed by 35 cycles of 98°C for 30 sec, 56°C for 40 sec, and 72°C for 1 min, with a final extension at 72°C for 10 min. For EF1α the PCR protocol was 94°C for 4 min, followed by 35 cycles of 94°C for 1 min, 49°C for 1 min, and 72°C for 1.5 min, with a final extension at 72°C for 4 min. PCR products were checked on a 1% agarose gel, and purified using the QIAquick PCR Purification Kit (Qiagen, Inc., Valencia, California, USA). Finally, both strands of each fragment were sequenced using Sanger technology in Macrogen Inc. sequencing service (Seoul, South Korea). All sequences were used in blastn searches to discard sequences belonging to mealybug parasitoids [32].
Data analysis
Both strands of all sequences were assembled into contigs using CodonCode Aligner v8.0.2 (CodonCode Corporation, Centerville, Massachusetts, USA), and multiple sequence alignment for each gene was performed using ClustalW algorithm in MEGA v. 7 [33]. The saturation substitution index (Iss) was estimated for each gene to evaluate if the Iss was significantly lower than the critical value of Iss (Iss.c). This test is necessary to ascertain whether sequences are phylogenetically informative [34]. Iss and the proportions of conserved and polymorphic sites were computed using DAMBE v.5.5.1 [35].
Phylogenetic analyses
We explored the relationships among H. pungens sensu lato populations and the relationship between H. pungens and its co-generic H. festerianus by means of a phylogenetic analysis using Bayesian inference (BI) and Maximum Likelihood approaches. In both cases, we utilized a concatenated data set using a partition scheme with three blocks (18S, EF1α and COI) as input. To exclude redundancies, the data set was trimmed using a single individual per haplotype. The General Time Reversible + Gamma parameter (GTR+Γ) model was adopted for each partition, as determined by means of Bayesian information criterion as suggested in Jmodeltest v.2.1.3 [36]. The related mealybug Planococcus ficus was used as outgroup. The BI was implemented in MrBayes v3.2.6 [37] and the ML analysis in RAxML-HPC2 v.8.2.10 [38] in CIPRES [39]. One thousand bootstrap replicates were run to evaluate clade stability. Two independent runs were performed with four Markov Chain Monte Carlo (MCMC) chains for each run for 50 million generations, sampling every 1000th generation. The first 25% of samples were removed as burn-in, and stability and sufficient mixing of parameters (ESS>200) was checked using Tracer v.1.6 [40]. Independent ML and BI analyses of each gene were performed and are available in supplementary material (S1 Fig). Finally, we estimated genetic divergence using uncorrected p-distances and the average number of pairwise differences between and within clades obtained in the phylogenetic analysis, using COI sequences, and including the Puerto Rico cactus pest.
Species delimitation analyses and divergence time
To evaluate whether H. pungens represented a unique species or a species complex, we used two types of species delimitation analyses: i) a multi-locus Bayesian approach [41] and ii) two single-locus approaches: Automatic Barcode Gap Discovery (ABGD) distance-based method [42] and multi-rate Poisson tree processes (mPTP) ML tree-based method [43]. In the multi-locus Bayesian analysis, we considered the well-supported clades identified in the phylogenetic analyses, combined with the results of experimental studies showing potential postzygotic reproductive barriers between mealybugs collected on Amaranthaceae and Cactaceae from Argentina, and the incapability of specimens collected on Cactaceae to develop and survive on Amaranthaceae and vice versa [26]. We employed the set of species trees estimated with BEAST v1.7.5 [44], including H. pungens, H. festerianus and the outgroup P. ficus, using the sequences of the three genes. This multi-locus method is based on incomplete lineage sorting as the main source of inconsistency between gene trees and species trees and assumed free recombination between genes and no recombination within genes. In this case, terminal taxa were defined a priori considering two alternative hypotheses. The first (hypothesis A) considered H. pungens as a complex of two species, one consisting of populations feeding on Amaranthaceae and the other of populations using Cactaceae. In the second scenario (hypothesis B), H. pungens is a complex of three species, namely, H. pungens sensu stricto including Amaranthaceae dwelling populations, and two new undescribed species both specialized in the use of Cactaceae: Hypogeococcus sp. 1 and Hypogeococcus sp. 2 encompassing the populations from Argentina and the specimens collected in Australia, respectively. Both hypotheses were tested by means of a Bayesian Markov Chain Monte Carlo analysis running 50 million generations and sampled every 1000th generations with a Yule Process as a tree prior and a strict clock model estimated according to COI partition. Convergence was verified with Tracer 1.6 [40] and the first 20% of sampled trees were discarded as burn-in using TreeAnnotator v1.7.5 [44]. Species trees resulting from the Bayesian analysis were visualized on DensiTree v2.1 [45]. We estimated divergence time between the entities identified by the species delimitation analyses, using a substitution rate of 0.017 substitutions per site per million years estimated for insects [46]. A posterior probability value ≥ 0.95 was considered as strong support for a speciation event [47].
In the single-locus analyses, ABGD and mPTP, we used the mitochondrial COI gene from all populations sampled in Argentina and Puerto Rico. We included the latter to determine which of the groups identified using these approaches was more related to the Puerto Rico cactus pest. The ABGD analysis was run in the ABGD website server (http://wwwabi.snv.jussieu.fr/public/abgd/abgdweb.html). This analysis uses the barcode gap, which is the gap observed when divergence among individuals of the same species is smaller than divergence among individuals from different species, to automatically find groups that might correspond to different potential species [42,48]. The range of prior intraspecific divergence (P), estimated with uncorrelated p-distances, was configured from 0.001 to 0.1 in 10 Steps, using a gap width of 1.5. [49–50]. The mTPT analysis was performed in mptp v0.2.3 and takes into account the distinct values of intraspecific divergence caused by the substitutions that should confer more credibility in the speciation events [43]. This method assumes that the number of substitutions between species is significantly higher than within species [51]. A ML tree reconstructed with RAxML using the same parameters used in the phylogenetic analysis (substitution rate model, number of bootstrap replicates, and outgroup) was used as input. Identical sequences were removed to avoid incorrect likelihood estimations. The support of the delimited species was assessed using two independent MCMC chains for 10 million generations each, discarding the first 25% as burn-in.
Reproductive compatibility
We evaluated reproductive compatibility between samples collected on the two main clades identified in phylogenetic analyses (see Results section for details). To this end, we conducted both reciprocal crosses between mealybugs collected in Amaranthaceae (Alternanthera pungens; Trancas, Tucumán province) and Cactaceae (Cleistocactus baumannii; El Portezuelo, Catamarca province) as follows: (♀Amaranthaceae host x ♂Cactaceae host) and (♀Cactaceae host x ♂Amaranthaceae host) and the respective controls (♀Amaranthaceae host x ♂ Amaranthaceae host) and (♀Cactaceae host x ♂ Cactaceae host). The response variable measured was the number of nymphs that reached adulthood per female (viable offspring). The absence of viable offspring in the crosses was assumed as reproductive incompatibility between mealybugs from different hosts. Since females of the Amaranthaceae clade can reproduce by parthenogenesis [26], we used offspring sex ratio as an indicator of parthenogenetic reproduction. A Chi-square (χ2) test was used to determine if the sex ratio deviated from 1:1 in the offspring (similar number of males and females). On the other hand, we considered that nymphs were sexually produced when sperm was observed in females spermathecae. To this end, ten espermathecae were inspected in Amaranthaceae clade females (five control and five treatment), and ten in Cactaceae clade females (five control and five in treatment).
Results
Sequence analysis
A total of 116 individuals were sampled for the present study from which a total of 304 sequences were generated. The basic statistics of sequence variation (number of segregating sites, number of haplotypes) and number of sequences per gene are detailed in S2 Table. A concatenated dataset of 1415 bp was used in the analyses described below. Excluding the gaps between sequences, we obtained 1393 bp of which 1247 bp (89.51%) were conserved positions and 146 bp (10.48%) were polymorphic (all sequences were deposited in GenBank under accession numbers MN013440—MN013743). The saturation substitution analysis revealed no saturation of phylogenetic signal in the set of sequences generated, according to the comparison between the Iss and the Iss critical (Iss.c) (S2 Table).
Phylogenetic analyses
Two main clades were observed in the BI and ML trees with high Posterior Probability (PP = 0.92) and Bootstrap Support (BS = 89%) (Fig 2) using the concatenated dataset (independent BI and ML trees for each gene are available in S1 Fig). The first clade includes all haplotypes of mealybugs feeding on Amaranthaceae (Amaranthaceae host clade), while the second clade contained all haplotypes of individuals collected on cactus hosts (Cactaceae host clade). The unique haplotype found in the Australian samples appears as the sister group of the Cactaceae host clade with high PP (1) and BS (70%). As for H. festerianus, we recovered two different haplotypes that formed independent clades with high PP (0.75) and intermediate BS (60%) (Fig 2). The results from the bayesian and ML analyses with each separate gene were moderately concordant with the results obtained with the concatenated data set (S1 Fig).
Posterior Probability and Bootstrap Support values are shown on each node inferred by BI and ML in a concatenated data set of COI, 18S and EF1. Both results are taken to represent the phylogenetic reconstruction of the populations of H. pungens using different hosts and the co-generic species H. festerianus. The vertical bars represent the candidate species inferred by each of the methods used in the study (see below).
Species delimitation analyses
Pairwise genetic differences between clades defined in the ML tree, measured with uncorrected p-distances, ranged from 1.4% to 3.7%, while differentiation within clades varied from 0% in Australian Population and the Puerto Rico cacti pest, to 0.67% in the Amaranthaceae host clade. The barcode gap assumes that divergence among individuals of the same species should be smaller than between species (Table 1). Our results showed that divergence between the Amaranthaceae and Cactaceae host clades was 3.0%, a value that is about five times greater than within clades (Table 1). Also, divergence estimates between either the Amaranthaceae host clade (3.2%) or the Cactaceae host clade (1.4%) with Australian mealybugs were also greater than estimates within clades. In turn, genetic divergence between H. festerianus and the Amaranthaceae host clade was 2.1%; whereas divergence between the Cactaceae host clade and H. festerianus was 3.4%. Finally, divergence estimates between the Puerto Rico cactus pest and the Cactaceae host clade and the Amaranthaceae host clade, both considered as H. pungens, was 3.5% and 1.6%, respectively (Table 1).
The three methods of species delimitation, the ABGD, the mTPT, and the Bayesian analyses, suggested that H. pungens was, in fact, a species complex. On one hand, the ABGD and the mTPT approaches detected two candidate species within H. pungens. The first, not only included populations grouped in the Amaranthaceae host clade, but also the Puerto Rico cactus pest. The second species candidate contained the Cactaceae host clade. Finally, H. festerianus formed an independent group in both ABGD and mTPT analyses (S2 and S3 Figs).
Furthermore, the results of the Bayesian analysis gave support to both hypotheses (A and B) of speciation events. Under Hypothesis A, consistent with the results obtained with the ABGD and the mTPT analyses, H. pungens was a complex that included at least two species: H. pungens sensu stricto (the Amaranthaceae host clade) and Hypogeococcus sp. 1 (the Cactaceae host clade) (Posterior Probability = 0.995; S4 Fig). The time calibrated analysis suggested that the speciation event separating these inferred species occurred nearly 35,000 years ago (confidence interval 95% HDP = 13,200–88,000), and that both shared their last common ancestor with H. festerianus approximately 65,000 years ago (confidence interval 95% HDP = 30,800–116,600). The Bayesian analysis also gave support to Hypotheses B (PP > 0.95), indicating that H. pungens was a complex of three species. The first split occurred nearly 29,000 years ago (confidence interval 95% HDP = 10,912–49,104) separating the Amaranthaceae and the Cactaceae host clades (PP = 1). The second speciation event took place approximately 10,000 years ago (confidence interval 95% HDP = 2,081–16,368) and separated two species among cactus feeding mealybugs, Hypogeococcus sp. 1, encompassing the populations collected in Argentina (Cactaceae host clade in Fig 2) and Hypogeococcus sp. 2 involving the Australian samples (PP = 0.95) (Fig 2 and S4 Fig). Finally, H. festerianus appeared to be closely related to the H. pungens species complex, from which it diverged about 44,000 years ago (confidence interval 95% HDP = 21,824–70,928). For both hypotheses, the total of 37,500 trees generated in Bayesian analyses before burn-in and plotted in DensiTree showed congruent topologies (S4 Fig).
Reproductive compatibility
The results of reproductive compatibility experiments suggested asymmetrical incompatibility. Control crosses of mealybugs from both hosts produced viable offspring (71.2% of 208 nymphs and 81.1% of 164 nymphs of the Amaranthaceae and Cactaceae host clades, respectively, reached the adult stage). In the inter-clade reciprocal crosses, 65.4% of 191 nymphs produced by Amaranthaceae clade females crossed with Cactaceae clade males reached adulthood (Table 2). In contrast, none of the 84 nymphs produced by Cactaceae clade females crossed with Amaranthaceae males reached the adult stage. In fact, none of the offspring produced in the latter cross developed beyond the second instar.
When females were dissected, we detected the presence of sperm in different proportions between treatments and controls (Table 2). In the hybrid crosses, females from Amaranthaceae host x males from Cactaceae host, we found sperm in the spermathecae in 3/5 dissected females; while in the control (♀Amaranthaceae x ♂Amaranthaceae) we observed that 4/5 females presented sperm in their spermathecae. In the reciprocal crosses, females from Cactaceae host x males from Amaranthaceae host, we found sperm in only 1 out of 5 females analyzed, while in Cactaceae host controls, 3/5 females had sperm in their spermathecae.
In both treatment and control crosses involving Amarathaceae host females we detected a biased sex-ratio. Sex ratio in the controls was 2:1 ♀:♂ (range 1:1–4:1 ♀:♂) and in the female from Amaranthaceae host x male from Cactaceae host, crosses were 1.7:1 ♀:♂ (range 1:1–6:0 ♀:♂). In contrast, we did not detect departures from the expected 1:1 sex ratio in crosses involving females from Cactaceae host (range 1:1–2:1 ♀:♂).
Discussion
The results of the present study indicated that mealybugs feeding on Amaranthaceae and Cactaceae collected in Argentina, on Cactaceae in Australia, and the Puerto Rico cactus pest, all referred to as H. pungens in the literature, are part of a species complex. Moreover, the genetic divergence in the H. pungens complex seemed to be strongly driven by the host plants. In this sense, H. pungens sensu stricto (the Amaranthaceae host clade in Fig 2, and H. pungens in S2–S4 Figs) appeared as a well-delimited species associated exclusively with the plants of the Amaranthaceae family in Argentina. However, the Puerto Rico cactus pest also appeared as part of the Amaranthaceae host clade, though genetically differentiated from all the populations from Argentina feeding on Amaranthaceae. The other members of the complex, which appeared to be ancestral of the mealybugs feeding on Amaranthaceae hosts, were specifically associated with Cactaceae (Cactaceae host clade in Fig 2, S2–S4 Figs).
Thus, we may ask whether the degree of genetic divergence among the clades feeding on different host plants could be interpreted as the result of speciation prompted by ecological interactions between mealybugs and their hosts. Such interactions might be an essential factor facilitating ecological speciation. Both types of hosts, Cactaceae and Amaranthaceae, coexist in some localities of northern Argentina (Fig 1), suggesting that feeding on alternative host plants was not only an effective isolating barrier, but also that specialization to particular hosts may have contributed to diversification in Hypogeococcus mealybugs.
The results of species delimitation analyses, using alternative algorithms implemented in different software (ABGD, mTPT, and *Beast), all revealed that H. pungens was a species complex of, at least, two species strongly associated to different host types. The literature pointing to a role of the use of alternative host plants in the rapid diversification in phytophagous insects is vast [52–55]. In this context, recent studies using molecular markers uncovered several cases of host associated genetic structure in plant feeding insects and invoked a role of host shifts in the evolution of reproductive barriers in the initial stages of speciation [52, 56–57]. For instance, studies in the phytophagous ladybird beetle, Henosepilachna diekei Jadwiszczak & Węgrzynowicz (Coleoptera: Coccinellidae), show that individuals collected in nature on distinct host plants not only exhibit extremely different host preferences and host performances, but also substantial genetic divergence [58,59]. Similarly, Ebel et al. [54] reports that the host use identification of an ecologically important relationship of the Neotropical Adelpha spp. (Lepidoptera: Nymphalidae) butterflies with a toxic host plant family, promotes the rapid diversification of the genus. Similarly, recent studies suggest the central role that host plant shifts play in the diversification of the gall midge genus Asphondylia Loew (Diptera: Cecidomyiidae) [60]. All in all, these examples suggest that the evolution of new adaptations to alternative host plants, as the result of divergent selection in alternative hosts, is the main cause of niche expansion and in the emergence of new insect species [52].
Even though there is no strict consensus on the degree of genetic divergence necessary to delimit species [61–64], the extent of genetic divergence among species estimated in our study (Table 1) was consistent with the results obtained in several orders of phytophagous insects in which cryptic species were disclosed using genetic markers [65,66]. Genetic distances among the species uncovered in our study, ranging from 1.4% to 3.7%, were consistent with values reported in other groups in which molecular approaches allowed the delimitation of new species in Hemiptera. Lee et al. [67], for instance, reported distance values greater than 2% in 1595 out of 1694 pairwise comparisons between aphid species. Other authors propose divergence values of up to 3.5% to delimit species in Hemiptera [62,68]. This limit has been calculated based on the highest intra- and interspecific genetic distances in the species complex of Bemisia tabaci (Gennadius) (Hemiptera: Aleyrodidae) [62,64,69]. However, such limit is not applicable to recent speciation events, since intra and, specially, interspecific genetic variation is frequently low [70–72], as seems to be the case in the H. pungens complex. It has been shown in other mealybugs that genetic divergence between congeneric species ranges from 2% to 19.5% [70,72]. In addition, Daane et al. [73] reported divergence estimates, using COI as genetic marker, between the sibling species Planococcus minor (Maskell) and P. citri (Risso) of 2%. Divergence values lower than 1% between well-defined species have also been reported in several species of Hemiptera [74–76].
Reproductive compatibility experiments produced evidence supporting the hypothesis that H. pungens is a species complex and that Amaranthaceae and Cactaceae host clades are different species. In effect, crosses between specimens collected on Amaranthaceae and Cactaceae revealed asymetrical postzygotic isolation; no F1 offspring could develop beyond the second nymph stage in the cross Cactaceae clade females x Amaranthaceae clade males, whereas similar proportions of F1 offspring reached adulthood in the reciprocal inter-clade crosses as compared to the respective control crosses. Even though we did not evaluate fertility in inter-clade F1 offspring that reached adulthood (Amaranthaceae females x Cactaceae males), our results are consistent with the hypotheses that the clades identified in our genetic survey may be considered different species and that adaptation to new hosts has directly or indirectly driven the evolution of isolating barriers [52,77–78]. In this context, evolutionary studies in Drosophila Fallén (Diptera: Drosophilidae) species co-distributed in northwestern Argentina, showed that the differential use of alternative cactus hosts promoted different adaptations (several fitness related traits and patterns of gene expression) that may be the consequence of the association to chemically different hosts [79]. However, testing this hypothesis in Hypogeococcus mealybugs and their hosts will require further studies involving a greater representation of the genome and transcriptomic data.
Overall, our genetic analyses indicate that the cactus feeding Hypogeococcus sp. mealybug introduced into Australia from Argentina was neither H. pungens sensu stricto nor H. festerianus. All species delimitation analyses suggested that mealybugs feeding on cacti, including the population exported to Australia, represent a new species. However, the Bayesian species delimitation analysis supported the hypothesis that the mealybugs introduced into Australia belong to a species that was not sampled in this thorough multi-year survey in its native range. In this context, it is worth mentioning that the Bayesian species delimitation method was more powerful than the distance-based ABGD and ML tree-based mTPT, which only used the mitochondrial gene. This is particularly true in recently formed species groups with low divergence (approximately 10,000 years ago according to the divergence time analysis), since ABGD is prone to false negatives, i.e. lumping less divergent clades in the same candidate species [80]. These results agree with Aguirre et al. [26] who proposed that the population introduced into Australia for biological control of the cactus Harrisia martinii (Labouret) Britton is not H. festerianus, but a new undescribed species. The present paper is expected to encourage future taxonomic work that would provide a formal description of the candidate species unveiled by our population genetic survey, denoted here as Cactaceae host clade, as well as new studies involving a wider genome exploration. Studies of reproductive isolation, such as cross-breeding experiments under quarantine conditions between Australian mealybugs and populations derived from Cactaceae host clade, are necessary to confirm the species status of the mealybugs exported to Australia. Those tasks are not only relevant but necessary for the design of biological control strategies for pest control in general, and specifically in the use of mealybugs as control agents of weedy cacti.
Finally, the detection of only one haplotype in the three localities sampled of the Puerto Rico cactus pest suggested only one invasive event. This haplotype was not found among the populations sampled in Argentina, the putative origin of the invasive stock. Samples of HCM from Puerto Rico appeared more related to H. pungens that feed on Amaranthaceae than to the cactus feeding mealybugs from Argentina, including the valid species H. festerianus and the population from Australia (Table 1). Thus, it seems possible that the Cactaceae feeding condition of the Puerto Rico pest may reflect a very recent host shift. However, identification of the geographic origin of the Puerto Rico cactus pest remains elusive. Identifying the origin of this pest is a mandatory task required to implement an effective plan for biological control to reduce the damage to cactus diversity in Puerto Rico, the adjacent Caribbean islands, and the potential expansion to the neighboring areas in the region. Therefore, new approaches should be implemented using methodologies that allow a more thorough sampling of the genome based on next generation sequencing techniques, including samples from different countries where populations of the H. pungens complex were reported.
Supporting information
S1 Fig. Maximum Likelihood and Bayesian inference trees of each gene analyzed.
https://doi.org/10.1371/journal.pone.0220366.s001
(TIF)
S2 Fig. Distance-based method Automatic Barcode Gap Discovery (ABGD).
https://doi.org/10.1371/journal.pone.0220366.s002
(TIF)
S1 Table. Geographical origin of the individuals analyzed.
https://doi.org/10.1371/journal.pone.0220366.s005
(DOCX)
S2 Table. Substitution saturation index test and sequences variation information.
https://doi.org/10.1371/journal.pone.0220366.s006
(DOCX)
Acknowledgments
We thank the associated editor Tzen-Yuh Chiang and two anonymous reviewers for comments and suggestions that helped to improve previous versions of this paper. We would like to express appreciation to Xoana Granata and Andres Sanchez Restrepo for technical assistance, Fabian Font for the taxonomic identifications of the plant hosts, Lucia Claps for the morphological identifications of mealybugs, Rachel E. McFadyen for sending the samples from Australia, and Arabella Peard for reviewing the manuscript.
References
- 1.
Williams DJ, de Willink MC. Mealybugs of Central and South America. CAB International; 1992.
- 2. German-Ramirez E, Kairo MT, Stocks I, Haseeb M, Serra CA. New record of Hypogeococcus pungens (Hemiptera: Pseudococcidae) in the Dominican Republic with comments on specific characters. Florida Entomol. 2014;97(1): 320–322.
- 3. Segarra-Carmona AE, Ramírez-Lluch A, Cabrera-Asencio I, Jiménez-López AN. First report of a new invasive mealybug, the Harrisia cactus mealybug Hypogeococcus pungens (Hemiptera: Pseudococcidae). J. Agr. Univ. Puerto Rico. 2010;94(1/2): 183–187.
- 4. McFadyen R. The cactus mealybug Hypogeococcus festerianus (Hemiptera: Pseudococcidae) an agent for the biological control of Eriocereus martinii (Cactaceae) in Australia. Entomaphaga. 1979;24: 281–287.
- 5. Zimmermann H, Pérez M, Cuen S, Mandujano M, Golubov J. The South American mealybug that threatens North American cacti. Cactus Succulent J. 2010;82(3): 105–108.
- 6. Granara de Willink M. Nueva especie de Hypogeococcus Rau de Tucumán, República Argentina (Homoptera: Pseudococcidae). Neotrópica. 1981;27: 61–65.
- 7.
McFadyen R, Tomley A. Preliminary indications of success in the biological control of the Harrisia cactus (Eriocereus martinii Lab.) in Queensland. In: Del Fosse ES (ed). Proceedings of the First Conference of the Australian Weed Science Society, National Science Centre, Parkville, Victoria, Australia. 1978; 108–112.
- 8. Tomley A, McFadyen R. Biological control of Harrisia cactus, Eriocereus martinii, in central Queensland by the mealybug, Hypogeococcus festerianus, nine years after release. In: Proceedings of VI International Symposium of Biological Control of Weeds. Agriculture Canada, Vancouver. 1984; 843–847
- 9. Hosking J, McFadyen R, Murray N. Distribution and biological control of cactus species in eastern Australia. Plant Protect. 1988;3: 115–123.
- 10. Moran V, Zimmermann H. Biological control of cactus weeds of minor importance in South Africa. Agric. Ecosyst. Environ. 1991;37: 37–55
- 11. Halbert S. Entomology section. Triology. 1996;35: 2–4.
- 12. Klein H. Biological control of three cactaceous weeds, Pereskia aculeata Miller, Harrisia martinii (Labouret) Britton and Rose and Cereus jamacaru De Candole in South Africa. Afr. Entomol. Memoir. 1999;1: 3–14.
- 13.
Mazzeo G, Suma P, Russo A. Scale insects on succulent plants in Southern Italy. In: Branco M, Franco JC, Hodgson CJ, editors. Proceedings of the XI International Symposium on Scale Insect Studies, Oeiras, Portugal, 24–27 September 2007. ISA Press, Lisbon, Portugal. 2008; 149–152.
- 14.
Hodges A, Hodges G. Hypogeococcus pungens Granara de Willink (Insecta: Hemiptera: Pseudococcidae), a mealybug. EENY459. IFAS Extension, University of Florida. Available online: http://edis.ifas.ufl.edu/pdffiles/IN/IN82700.pdf (accessed: sep 15, 2018). 2009.
- 15. Zimmermann H, Pérez S, Cuen M. La amenaza de los piojos harinosos Hypogeococcus pungens e Hypogeococcus festerianus (Hemiptera: Pseudococcidae) a las cactáceas mexicanas y del Caribe. Cactáceas Y Suculentas Mexicanas. 2010;55: 4–17.
- 16. Beltrà A, Soto A. New records of mealybugs (Hemiptera: Pseudococcidae) from Spain. Phytoparasitica. 2011;39(4): 385–387.
- 17.
Julien M, Griffiths M. Biological control of weeds. A world catalogue of agent and their target weeds, 4th edn. CAB Publishing, Wallingford; 1999.
- 18.
McFadyen R E. Harrisia (Eriocereus) martinii (Labour.) Britton—Harrisia cactus Acanthocereus tetragonus (L.) Hummelink—sword pear. 2012. pp. 274–281 In: Julien M, McFadyen R, Cullen J, editors. Biological control of weeds in Australia. CSIRO Publishing, Collingwood, Australia. 2012. pp. 274–281.
- 19.
Ben-Dov Y. A systematic catalogue of the mealybugs of the world (Insecta: Homoptera: Coccoidea: Pseudococcidae and Putoidae) with data on geographical distribution, host plants, biology and economic importance. Intercept Limited; 1994.
- 20. Claps LE, de Haro ME. Coccoidea (Insecta: Hemiptera) associated with Cactaceae in Argentina. J. Prof. Assoc. Cactus Develop. 2001;4: 77–83.
- 21. Williams DJ. Two cactus-feeding mealybugs from Argentina (Homoptera, Coccoidea, Pseudococcidae). Bull. Entomol. Res. 1973;62(4): 565–570.
- 22.
McFadyen RE, Tomley AJ. Biological control of Harrisia cactus, Eriocereus martinii, in Queensland by the mealy bug, Hypogeococcus festerianus. In: Proceedings of the 5th International Symposium on Biological Control of Weeds. 1981 (pp. 589–59) Commonwealth Scientific and Industrial Research Organization. 1981; 589–594.
- 23. McFadyen RE. Biological control of weeds. Annu. Rev. Entomol. 1988;43(1): 369–393.
- 24. Hamon AB. A cactus mealybug, Hypogeococcus festerianus (Lizer y Trelles), in Florida (Homoptera: Coccoidea: Pseudococcidae). Entomology Circular, Division of Plant Industry, Florida Department of Agriculture and Consumer Services. 1984; 263.
- 25. Suss L, Trematerra P. Hypogeococcus festerianus (Lizer y Trelles), nocivo alle Cactaceae ornamentali in Leguria. Informatore Fitopatologia. 1986;10: 43–46.
- 26. Aguirre MB, Diaz-Soltero H, Claps LE, Saracho Bottero A, Triapitsyn S, Hasson E, et al. Studies on the biology of Hypogeococcus pungens (sensu stricto) (Hemiptera: Pseudococcidae) in Argentina to aid the identification of the mealybug pest of Cactaceae in Puerto Rico. J. Insect Sci. 2016 Jun 20. pmid:27324585
- 27. Hoelmer KA, Kirk AA. Selecting arthropod biological control agents against arthropod pests: Can the science be improved to decrease the risk of releasing ineffective agents? Biol. Control. 2005;34(3): 255–264.
- 28. von Dohlen CD, Moran NA. Molecular phylogeny of the Homoptera: A paraphyletic taxon. J. Mol. Evol. 1995;41(2): 211–223. pmid:7666451
- 29. Cho S, Mitchell A, Regier JC, Mitter C, Poole RW, Friedlander TP, et al. A highly conserved nuclear gene for low-level phylogenetics: Elongation factor-1 alpha recovers morphology-based tree for heliothine moths. Mol. Biol. Evol. 1995;12(4): 650–656. pmid:7659020
- 30. Simon C, Frati F, Beckenbach A, Crespi B, Liu H, Flook P. Evolution, weighting, and phylogenetic utility of mitochondrial gene sequences and a compilation of conserved polymerase chain reaction primers. Ann. Entomol. Soc. Am. 1994;87(6): 651–701.
- 31.
Brady SG, Gadau J, Ward PS. Systematics of the ant genus Camponotus (Hymenoptera: Formicidae): a preliminary analysis using data from the mitochondrial gene cytochrome oxidase I. In Hymenoptera: evolution, biodiversity and biological control. Fourth International Hymenoptera Conference, held in Canberra, Australia, in January 1999. 2000 (pp. 131–139). CSIRO Publishing.
- 32. Triapitsyn SV, Aguirre MB, Logarzo GA, Hight SD, Ciomperlik MA, Rugman-Jones PF, et al. Complex of primary and secondary parasitoids (Hymenoptera: Encyrtidae and Signiphoridae) of Hypogeococcus spp. mealybugs (Hemiptera: Pseudococcidae) in the New World. Fla. Entomol. 2018;101(3):411–35.
- 33. Kumar S, Stecher G, Tamura K. MEGA7: Molecular evolutionary genetics analysis version 7.0 for bigger datasets. Mol. Biol. Evol. 2016;33(7):1870–4. pmid:27004904
- 34. Xia X, Xie Z, Salemi M, Chen L, Wang Y. An index of substitution saturation and its application. Mol. Phylogenet. Evol. 2003 1;26(1):1–7. pmid:12470932
- 35. Xia X, Xie Z. DAMBE: Software package for data analysis in molecular biology and evolution. J. Hered. 2001;92(4):371–3. pmid:11535656
- 36. Darrida D, Taboada G, Doallo R, Posada D. jModelTest 2: More models, new heuristics and parallel computing. Nat. Methods. 2012; 9: 772.
- 37. Ronquist F, Teslenko M, Van Der Mark P, Ayres DL, Darling A, Höhna S, et al. MrBayes 3.2: Efficient Bayesian phylogenetic inference and model choice across a large model space. Syst. Biol. 2012;61(3):539–42. pmid:22357727
- 38. Stamatakis A. Version R. 8: A tool for phylogenetic analysis and post-analysis of large phylogenies. 2014;30(9):1312–1313.
- 39.
Miller MA, Pfeiffer W, Schwartz T. Creating the CIPRES Science Gateway for inference of large phylogenetic trees. In 2010 gateway computing environments workshop (GCE). 2010(pp. 1–8).
- 40. Rambaut A, Suchard MA, Xie D, Drummond AJ. 2014. Tracer v1.6. [Software]. Available: http://beast.bio.ed.ac.uk/Tracer BEAST
- 41. Heled J, Drummond AJ. Bayesian inference of species trees from multilocus data. Mol. Biol. Evol. 2009;27(3):570–80. pmid:19906793
- 42. Puillandre N, Lambert A, Brouillet S, Achaz G. ABGD, Automatic Barcode Gap Discovery for primary species delimitation. Mol. Ecol. 2012;21(8):1864–77. pmid:21883587
- 43. Kapli P, Lutteropp S, Zhang J, Kobert K, Pavlidis P, Stamatakis A, et al. Multi-rate Poisson tree processes for single-locus species delimitation under maximum likelihood and Markov chain Monte Carlo. Bioinformatics. 2017;33(11):1630–8. pmid:28108445
- 44. Drummond AJ, Suchard MA, Xie D, Rambaut A. Bayesian phylogenetics with BEAUti and the BEAST 1.7. Mol. Biol. Evol. 2012;29(8):1969–73. pmid:22367748
- 45. Bouckaert R, Heled J. DensiTree 2: Seeing trees through the forest. BioRxiv. 2014:012401.
- 46. Papadopoulou A, Anastasiou I, Vogler AP. Revisiting the insect mitochondrial molecular clock: the mid-Aegean trench calibration. Mol. Biol. Evol. 2010;27(7):1659–72. pmid:20167609
- 47. Leaché AD, Fujita MK. Bayesian species delimitation in West African forest geckos (Hemidactylus fasciatus). Proc. Biol. Sci. 2010;277(1697):3071–7. pmid:20519219
- 48. Alfaro FM, Muñoz-Ramírez CP, Zúñiga-Reinoso Á, Trewick SA, Méndez MA. Phylogeography of the Chilean red cricket Cratomelus armatus (Orthoptera: Anostostomatidae) reveals high cryptic diversity in central Chile. Biol. J. Linn. Soc. 2018;123(4):712–27.
- 49. Collins RA, Boykin LM, Cruickshank RH, Armstrong KF. Barcoding's next top model: an evaluation of nucleotide substitution models for specimen identification. Methods Ecol. Evol. 2012;3(3):457–65.
- 50. Gibbs J. Bees of the genus Lasioglossum (Hymenoptera: Halictidae) from Greater Puerto Rico, West Indies. Eur. J. Taxon. 2018;(400).
- 51. Zhang J, Kapli P, Pavlidis P, Stamatakis A. A general species delimitation method with applications to phylogenetic placements. Bioinformatics. 2013;29(22):2869–76. pmid:23990417
- 52. Forbes AA, Devine SN, Hippee AC, Tvedte ES, Ward AK, Widmayer HA, et al. Revisiting the particular role of host shifts in initiating insect speciation. Evolution. 2017;71(5):1126–37. pmid:28052326
- 53. Soudi S, Reinhold K, Engqvist L. Ecologically dependent and intrinsic genetic signatures of postzygotic isolation between sympatric host races of the leaf beetle Lochmaea capreae. Evolution. 2016;70(2):471–9. pmid:26882317
- 54. Ebel ER, DaCosta JM, Sorenson MD, Hill RI, Briscoe AD, Willmott KR, et al. Rapid diversification associated with ecological specialization in Neotropical Adelpha butterflies. Mol. Ecol. 2015;24(10):2392–405. pmid:25809206
- 55. Funk DJ, Egan SP, Nosil P. Isolation by adaptation in Neochlamisus leaf beetles: host‐related selection promotes neutral genomic divergence. Mol. Ecol. 2011;20(22):4671–82. pmid:21999361
- 56. Hernández-Vera GE, Mitrović M, Jović J, Tosevski I, Caldara R, Gassmann A, et al. Host associated genetic differentiation in a seed parasitic weevil Rhinusa antirrhini (Coleptera: Curculionidae) revealed by mitochondrial and nuclear sequence data. Mol. Ecol. 2010;19(11):2286–300. pmid:20465586
- 57. Xue HJ, Li WZ, Yang XK. Assortative mating between two sympatric closely-related specialists: inferred from molecular phylogenetic analysis and behavioral data. Sci. Rep. 2014;4:5436. pmid:24961567
- 58. Matsubayashi KW, Kahono S, Katakura H. Divergent host plant specialization as the critical driving force in speciation between populations of a phytophagous ladybird beetle. J. Evol. Biol 2011;24(7):1421–32. pmid:21545419
- 59. Matsubayashi KW, Kahono S, Katakura H. Divergent host plant preference causes assortative mating between sympatric host races of the ladybird beetle, Henosepilachna diekei. Biol. J. Linn. Soc. 2013;110(3):606–14.
- 60. Dorchin N, Joy JB, Hilke LK, Wise MJ, Abrahamson WG. Taxonomy and phylogeny of the Asphondylia species (Diptera: Cecidomyiidae) of North American goldenrods: challenging morphology, complex host associations, and cryptic speciation. Zool. J. Linn. Soc. 2015;174(2):265–304.
- 61. June Lee Yo, Hill KB. Systematic revision of the genus Psithyristria Stål (Hemiptera: Cicadidae) with seven new species and a molecular phylogeny of the genus and higher taxa. Syst. Entomol. 2010 Apr;35(2):277–305.
- 62. Dinsdale A, Cook L, Riginos C, Buckley YM, De Barro P. Refined global analysis of Bemisia tabaci (Hemiptera: Sternorrhyncha: Aleyrodoidea: Aleyrodidae) mitochondrial cytochrome oxidase 1 to identify species level genetic boundaries. Ann. Entomol. Soc. Am. 2010;103(2):196–208.
- 63. Chu D, Wan FH, Tao YL, Liu GX, Fan ZX, Bi YP. Genetic differentiation of Bemisia tabaci (Gennadius)(Hemiptera: Aleyrodidae) biotype Q based on mitochondrial DNA markers. Insect Sci. 2008;15(2):115–23.
- 64. Lee W, Park J, Lee GS, Lee S, Akimoto SI. Taxonomic status of the Bemisia tabaci complex (Hemiptera: Aleyrodidae) and reassessment of the number of its constituent species. PLoS One. 2013; 8(5): e63817. pmid:23675507
- 65. Lavinia PD, Bustos EO, Kopuchian C, Lijtmaer DA, García NC, Hebert PD, et al. Barcoding the butterflies of southern South America: Species delimitation efficacy, cryptic diversity and geographic patterns of divergence. PloS one. 2017; 12(10): e0186845 pmid:29049373
- 66. Richards NK, Mansfield S, Townsend RJ, Ferguson CM. Genetic variation within species and haplotypes of the Wiseana (Lepidoptera: Hepialidae) complex and development of non‐sequenced based identification tools to aid field studies. Pest. Manag. Sci. 2017;73(11):2334–44. pmid:28544267
- 67. Lee W, Kim H, Lim J, Choi HR, Kim Y, et al. Barcoding aphids (Hemiptera: Aphididae) of the Korean Peninsula: updating the global data set. Mol. Eco. Resour. 2011;11(1):32–7.
- 68. De Barro PJ, Liu SS, Boykin LM, Dinsdale AB. Bemisia tabaci: a statement of species status. Annu. Rev. Entomol. 2011;56:1–9. pmid:20690829
- 69. Barbosa LD, Marubayashi JM, De Marchi BR, Yuki VA, Pavan MA, Moriones E, et al. Indigenous American species of the Bemisia tabaci complex are still widespread in the Americas. Pest. Manag. Sci. 2014;70(10):1440–5. pmid:24458534
- 70. Rung A, Scheffer SJ, Evans G, Miller D. Molecular identification of two closely related species of mealybugs of the genus Planococcus (Homoptera: Pseudococcidae). Ann. Entomol. Soc. Am. 2008;101(3):525–32.
- 71. Rung A, Miller DR, Scheffer SJ. Polymerase chain reaction-restriction fragment length polymorphism method to distinguish three mealybug groups within the Planococcus citri–P. minor species complex (Hemiptera: Coccoidea: Pseudococcidae). J. Econ. Entomol. 2009;102(1):8–12. pmid:19253611
- 72. Wang XB, Zhang JT, Deng J, Zhou QS, Zhang YZ, Wu SA. DNA barcoding of mealybugs (Hemiptera: Coccoidea: Pseudococcidae) from mainland China. Ann Entomol Soc Am. 2016;109(3):438–46.
- 73. Daane KM, Middleton MC, Sforza RF, Kamps-Hughes N, Watson GW, Almeida RP, et al. Determining the geographic origin of invasive populations of the mealybug Planococcus ficus based on molecular genetic analysis. PloS one. 2018;13(3):e0193852. pmid:29565996
- 74. Coeur d’acier A, Cocuzza G, Jousselin E, Cavalieri V, Barbagallo S. Molecular phylogeny and systematic in the genus Brachycaudus (Homoptera: Aphididae): insights from a combined analysis of nuclear and mitochondrial genes. Zool. scripta. 2008;37(2):175–93.
- 75. Foottit RG, Maw HE, Von Dohlen CD, Hebert PD. Species identification of aphids (Insecta: Hemiptera: Aphididae) through DNA barcodes. Mol. Ecol. Resour 2008;8(6):1189–201. pmid:21586006
- 76. Kim H, Lee S. Molecular systematics of the genus Megoura (Hemiptera: Aphididae) using mitochondrial and nuclear DNA sequences. Mol. Cell. 2008;25(4).
- 77. Diehl SR, Bush GL. An evolutionary and applied perspective of insect biotypes. Annu. Rev. Entomol. 1984;29(1):471–504.
- 78. Drès M, Mallet J. Host races in plant–feeding insects and their importance in sympatric speciation. Philos. Trans. Royal Soc. B. 2002;357(1420):471–92.
- 79. Hasson E, De Panis D, Hurtado J, Mensch J. Host plant adaptation in cactophilic species of the Drosophila buzzatii cluster: fitness and transcriptomics. J. Hered. 2018;110(1):46–57.
- 80. Yu G, Rao D, Matsui M, Yang J. Coalescent-based delimitation outperforms distance-based methods for delineating less divergent species: the case of Kurixalus odontotarsus species group. Sci. rep. 2017;7(1):16124. pmid:29170403