Thrips are insect pests of economically important agricultural, horticultural, and forest crops. They cause damage by sucking plant sap and by transmitting several tospoviruses, ilarviruses, carmoviruses, sobemoviruses, and machlomoviruses. Accurate and timely identification is the key to successful management of thrips species. However, their small size, cryptic nature, presence of color and reproductive morphs, and intraspecies genetic variability make the identification of thrips species challenging. The use of molecular and electronic detection platforms has made thrips identification rapid, precise, sensitive, high throughput, and independent of developmental stages. Multi-locus phylogeny based on mitochondrial, nuclear, and other markers has resolved ambiguities in morphologically indistinguishable thrips species. Microsatellite, RFLP, RAPD, AFLP, and CAPS markers have helped to explain population structure, gene flow, and intraspecies heterogeneity. Recent techniques such as LAMP and RPA have been employed for sensitive and on-site identification of thrips. Artificial neural networks and high throughput diagnostics facilitate automated identification.
1. Introduction
Thrips are soft-bodied slender insects with fringed wings in the order Thysanoptera with 6337 species and 786 genera [
[1][1]. The family Phlaeothripidae is the most speciose among the thrips with about 3550 described species followed by Thripidae with more than 2100 described species
[2][3]. The phytophagous thrips cause direct damage by piercing plant tissue and imbibing sap. Thrips feeding makes the plant parts appear silvery as the empty cells are filled with air. As the damaged leaves, flowers, and fruits grow in size, they become scarred, malformed, and distorted
[4]. Economic losses due to thrips account for GBP 7–11 million annually in the United Kingdom
[5]. In addition, thrips cause indirect damage by transmitting viruses, including tospoviruses, ilarviruses, carmoviruses, sobemoviruses, and machlomoviruses. Tospoviruses are economically damaging to a wide range of food crops and ornamental species
[6][7][8], causing stunted growth, formation of chlorotic and necrotic rings, death of apical shoots, dropping of leaves, and may be lethal
[7][9][10]. Tomato spotted wilt virus (TSWV, order
Bunyavirales, and family
Tospoviridae) alone is reported to cause global economic losses of around USD 1 billion
[11]. Groundnut bud necrosis virus (GBNV) reportedly causes economic losses of about USD 89 million in Asia
[12].
Accurate identification is crucial for discriminating thrips species of quarantine concern from endemic species and in formulating effective pest management strategies. Their small size, cryptic habit, color morphs, secondary sexual characters, and genetic variants render the identification of thrips species challenging
[13][14][15][16][17]. Conventional insect taxonomy mostly relies on external morphology-based dichotomous keys for species delimitation. Several such resources are available for the identification of thrips specimens
[18][19][20][21][22][23][24][25][26][27][28]. However, species identification based on morphological characters is time-consuming as it involves processing of specimens, preparation of microscope slides, and magnification using a microscope, as well as expert morphological knowledge of the genera. Furthermore, available keys are generally limited to the adult stage, and economically important or prevalent thrips species have been illustrated. Moreover, morphological characters do not take into account the presence of cryptic species or genetic variants.
Advancements in molecular biology over the last decade offer a variety of tools for specific and accurate identification of thrips, alleviating the limitations of morphological key-based identification. Nucleic acid and protein-based techniques such as polymerase chain reaction (PCR), random amplified polymorphic DNA (RAPD), restriction fragment length polymorphism (RFLP), sequence-characterized amplified regions (SCAR) markers, quantitative PCR (qPCR), loop-mediated isothermal amplification (LAMP), and monoclonal antibodies (MAb)
[14][29][30][31][32][33][34][35][36][37] have been shown to successfully discriminate between several thrips species. Mehle and Trdan
[38] previously reviewed the limitations and advantages of traditional and early modern methods of thrips identification. Since then, many advanced molecular and automated electronic techniques have been introduced to improve efficiency or to add functionality in thrips diagnosis. Despite the limitations of traditional thrips identification and recent molecular advances, routine identification of pest thrips species continues to be based on microscopy by quarantine and agriculture departments in many countries due to its dependability. This review provides an update and comparative assessment of novel molecular, high throughput, and automated approaches for fast and precise thrips diagnosis to study thrips polymorphism and to understand thrips population structure.
2. Landmarks in Thrips Diagnostics
The first recorded mention of thrips was a sketch by the Italian Jesuit scholar Filippo Bonanni in 1691. Two species in the genus
Physapus were described by Baron Charles De Geer in 1744, and Linnaeus mentioned a third species in 1746 and named this group as
Thrips. The insect order Thysanoptera was proposed in 1836, and 41 thrips species and 11 genera were described by Alexander Henry Haliday. The first monograph on thrips was published by Heinrich Uzel in 1895 which was comprised of all the previously published data on thrips and also described 11 new genera and 65 species with their identification keys
[39]. Thrips were described as vectors of tomato spotted wilt disease by Pittman in Australia in 1927
[40]. Since then, morphology-based classification of thrips has been advanced globally by many researchers. Thrips became a major concern in agriculture post-1980s due to epidemics of TSWV in North America, Europe, Africa, and Australia following the introduction of the highly efficient vector
Frankliniella occidentalis. The identification of thrips gained momentum post-1990s following the introduction of mitochondrial cytochrome C oxidase subunit I (COI)-based DNA barcoding. The first protein-based diagnostics were introduced when alloenzyme electrophoresis was used for the identification of thrips species
[41]. The first molecular characterization of thrips was reported in the late 1990s
[42], and monoclonal antibodies were developed for thrips identification
[29]. Molecular identification of thrips was subsequently diversified with the introduction of DNA marker assays. LAMP, multiplex PCR, quantitative PCR, recombinase polymerase amplification (RPA), and high-throughput sequencing were harnessed for thrips diagnostics post 2010s. The use of an artificial neural network system for automated thrips identification was initiated in 2008
[43]. A timeline of landmarks in thrips diagnosis is illustrated in
Figure 1. The many advancements in molecular thrips diagnostics, their applications, strengths, and weaknesses are summarized in the subsequent sections.
Figure 1. Timeline of milestones in the diagnostics of thrips species.
3. PCR-Based Identification of Thrips Using Molecular Markers
The utility of PCR to analyze small sample sizes insufficient for morphological thrips identification is advantageous for the diagnosis of immature insect stages. Molecular markers are generally highly reliable for resolving species ambiguities that are often not possible with morphology-based taxonomy. In thrips, DNA markers such as COI, COII, COIII, ribosomal RNA (rRNA), internal transcribed spacers (ITS), random amplified polymorphic DNA (RAPD), restriction fragment length polymorphism (RFLP), amplified fragment length polymorphism (AFLP), and simple sequence repeats (SSR) have been utilized successfully for species discrimination and phylogenetic analyses
[13][14][44][45][46].
3.1. COI Markers
The utilization of COI for large-scale DNA barcoding was first proposed in 2003
[47][48]. The first molecular study of thrips based on the partial sequence of COI was reported in 1998
[42]. Since then, more than 14,700 sequences of thrips COI have been deposited in the National Center for Biotechnology Information (NCBI) database to date. The availability of such a large number of reference sequences and robust universal primers
[49] has made COI a natural choice for thrips identification. The largest number of COI sequences is available for
Taeniothrips inconsequens (Uzel) (2171) followed by
Frankliniella occidentalis (Pergande) (826),
Thrips tabaci Lindeman (562),
Aptinothrips rufus (Haliday) (512),
T. palmi Karny (481), and
F. schultzei (Trybom) (391). There has been a continuing upwards trend of new thrips COI accessions available in the NCBI database over the years (
Figure 2A). A surge of COI data occurred post 2007, but numbers have plateaued in recent years.
Figure 2. Trend of sequence availability for important thrips species in the NCBI database. (A) COI and (B) rRNA-ITS sequence accessions. Sequence accessions for five important thrips species, viz., Frankliniella occidentalis, F. schultzei, Scirtothrips dorsalis, Thrips palmi, and T. tabaci deposited in the NCBI database were considered. Unverified accessions were removed, and data were processed according to the year of submission. Vertical axis shows the total number of accessions, and horizontal axis represents the year of submission. COI, mitochondrial cytochrome oxidase subunit I; rRNA, ribosomal RNA; ITS, internal transcribed spacer.
Analysis of COI sequences proved useful in identifying several thrips species infesting economically important crops in southern Africa
[50], India
[51], and Mexico
[52][53]. COI markers are often utilized to substantiate morphological character-based identification
[44].
F.
occidentalis,
Haplothrips spp.,
T. palmi,
T. vulgatissimus Haliday, and
T. tabaci could be discriminated based on a 413 bp fragment
[54]. A 433 bp fragment of COI was used as a genetic marker for the identification of the ten thrips species including
F. occidentalis,
Parthenothrips dracaenae (Heeger),
Anaphothrips obscurus (Müller),
T. palmi, T. tabaci,
T. angusticeps Uzel,
Echinothrips americanus Morgan,
Hercinothrips femoralis (Reuter),
H. haemorrhoidalis (Bouché), and
T. picipes (Zetterstedt)
[32]. These markers detected the immature stages as well as adults, even up to 1:120 dilutions without any cross-reactivity to other thrips species
[55][56]. In addition to morphology, COI sequences were used to substantiate the incursion of new species, such as
H. longisensibilis Xie, Mound, and Zhang in northern Brazil
[57]; and
T. parvispinus (Karny)
[58],
Podothrips erami Minaei, and
F. occidentalis in India
[59][60]. Furthermore, strain-specific COI PCR primers were shown to efficiently discriminate strains and reproductive stages of
T. tabaci [61]. COI-based identification of thrips species is simple, quick, and reliable. The thrips specimens studied by scanning electron microscopy for morphological keys can also be used as samples for sequencing COI
[62]. However, COI sequences of several thrips species demonstrated high intraspecific diversity resulting in a low barcode gap among the species. This can give rise to variation in reference sequence data, reduced efficiency of species-specific primers, and often results in inaccurate identification of species. Nuclear integration of COI fragments is also common in arthropods
[63][64][65][66] that may co-amplify or even be amplified instead of mitochondrial COI and negatively affect the molecular identification of thrips. A multi-locus phylogeny is, therefore, recommended for identifying/excluding species ambiguities in thrips.
3.2. Thrips Genetic Diversity Studies Using COI Markers
The high sequence variability of 5′-COI has been successfully utilized for identification of genetic variants, reproductive morphs, haplotypes, biotypes, ecotypes, allopatric speciation, subspecies, or cryptic species of several thrips
[67].
3.2.1. T. tabaci
The host preference and tospovirus transmission ability of
T. tabaci vary considerably between populations. Moreover,
T. tabaci follows a haplodiploid reproduction with two different reproductive forms, i.e., arrhenotoky and thelytoky
[68]. Arrhenotoky produces unfertilized eggs that develop into haploid males, while thelytoky produces diploid females. These two reproductive forms are indistinguishable by morphological keys
[69]. Zawirska (1976)
[70] proposed the presence of two biotypes of
T. tabaci. The ‘tabaci type’ prefers tobacco plants and efficiently transmits TSWV. In contrast, populations of the ‘communis type’ neither infest tobacco nor transmit TSWV. However, this hypothesis received little attention until Chatzivassiliou
[71] reinforced the concept and demonstrated the heterogeneity of
T. tabaci populations. The hypothesis that
T. tabaci is a heterogeneous taxon is supported by abundant variations in its COI region between different populations
[32][66][72][73][74]. Based on variation in the COI sequences, three haplotypes were identified at elevated frequencies, of which one represents a high-copy nuclear pseudogene and two are heteroplasmic variants of mitochondrial DNA
[66]. Clustering analyses and haplotype networking based on partial COI sequences strongly suggest three major lineages in
T. tabaci [13][75][76]. Clade ‘T’ or group C exclusively consists of haplotypes collected on tobacco plants, whereas haplotypes collected on leek form a separate clade. The leek clade can be subdivided into ‘L1′ or group A that contains all arrhenotokous strains and male specimens while all thelytokous strains are in group B or ‘L2′
[13][75]. T type arrhenotokous population has a high ability to transmit TSWV, while arrhenotokous L1 are poor transmitters and thelytokous L2 are non-transmitters
[75]. An ancient arrhenotokous strain probably differentiated into tobacco (T) and leek (L) types and then a thelytokous type (L2) that originated from the arrhenotokous leek type (L1). Brunner and colleagues described these linages as subspecies in view of genetic distinctiveness in sympatry, as T, L1, and L2 remain distinct both genetically and ecologically
[13]. Although Groups A (L1) and B (L2) are monophyletic
[75], reproductive isolation will require future in-detail comparisons. Lineages of
T. tabaci proposed by Brunner and colleagues
[13] are widely adopted to describe the divergence in global population of
T. tabaci.
All studied Australian
T. tabaci populations are in L2 lineages
[77]. Within this clade the seven populations from potato, three from onion, and four from chrysanthemum, impatiens, lucerne, and blackberry nightshade clustered as three distinct sub-groupings characterized by the source host. The
T. tabaci population from potatoes is capable of transmitting TSWV
[77]. However, New Zealand and Australian populations are different from each other (personal communication, LA Mound).
T. tabaci males are known to occur in New Zealand, but no male of
T. tabaci has ever been diagnosed from Australia. All the haplotypes infesting onion and cabbage in the state of New York are L2
[78]. The L1 lineage is also present in New York
[79]. L1 and L2 lineages are also recorded in Italy
[80]. The occasional presence of arrhenotokous populations in L2 clade may be remnants of the ancestral thrips. The presences of tetraploid females and deuterotoky are evident in
T. tabaci that may also produce haploid males
[17][81].
3.2.2. T. palmi
T. palmi populations can be divided into two distinct clades based on a high intraspecific variation of COI region
[82][83][84][85]. However, in our recent study
[86], haplotype data based on COI sequences identified 29 haplotypes of
T. palmi globally that can be divided into three major clades. The most common haplotype (
n = 121) is shared among populations from India, Pakistan, Japan, Thailand, Dominican Republic, China, United States, and Taiwan. All Indonesian specimens form a separate haplotype
[86]. These haplotypes represent three molecular operational taxonomic units (MOTUs). MOTU refers to clusters of haplotypes that are grouped by DNA sequence similarity of a specific marker gene. Sequences are clustered according to their similarity, and MOTUs are defined based on the similarity threshold. High genetic distances between these MOTUs in
T. palmi may indicate the presence of cryptic species that are morphologically indistinguishable
[87].
3.3. COII Markers
Apart from COI, COII is often used for the phylogenetic analysis of insects
[88]. A considerable amount of sequence information is available for COII gene of thrips. A total of 305 accessions of COII from different thrips species are available in NCBI, of which 180 sequences account for
A. rufus followed by
T. palmi (36) and
F. occidentalis (30). COII sequences can be used for substantiating COI-based phylogeny of thrips
[82].
3.4. COIII Markers
A total of 82 COIII sequences are available in NCBI for the order Thysanoptera of which the highest number is for
T. palmi (29). COIII-based phylogeny of
T. palmi shows a high interspecific distance with no within-species divergence
[82]. COIII-based markers appear suitable for the identification of thrips at the genus level but not at the species level.
3.5. rRNA-ITS
In insects, the spacer DNA between 18S and 5.8S RNA genes is known as ITS1, while ITS2 separates genes encoding 5.8S and 28S
[89]. Due to unequal crossing-over and gene conversion, nuclear rRNA genes undergo rapid concerted evolution by repairing mismatches among recombining chromosomes. This promotes intragenomic homogeneity of the repeat units and maintains intragenomic uniformity. Moreover, ITS is easy to detect from small quantities of DNA as it is present in high copy numbers. ITS offers additional advantages for species-level identification in thrips due to larger interspecific distances than for COI
[82][84][85][86][87]. To date, 3885 rRNA-ITS sequences of thrips can be accessed in NCBI. Most such sequences have been obtained for
F. occidentalis (409) followed by
F. schultzei (217),
A. rufus (137),
S. aurantii (131), and
S. dorsalis (111).
Figure 2B shows the year-by-year submission rate of rRNA-ITS data for economically important thrips species.
Ribosomal RNA-ITS markers have been used for detection of
S. dorsalis,
T. palmi,
F. tritici (Fitch),
F. intonsa (Trybom),
F. cephalica (Crawford),
H. cahirensis (Trybom),
Dendrothrips eremicola Priesner,
Kakothrips pisivorus (Westwood),
Hydatothrips kassimianus (Priesener), and
Ceratothripoides claratris (Shumsher)
[46][90][91][92][93][94]. ITS has been used as a natural choice of nuclear marker to substantiate mitochondrial marker-based identification and allow multi-locus phylogenetic analyses of
F. occidentalis,
F. intonsa,
F. fusca (Hinds),
T. tabaci, and
Megalurothrips distalis (Karny)
[72][95][96]. ITS sequences are also useful in distinguishing cryptic species of
F. schultzei [97],
S. dorsalis [98][99],
T. tabaci [100][101], and
S. aurantii [102]. However, a high variation among taxa indicates that ITS2 may not be appropriate for assessing intraspecific variation of
T. tabaci populations
[72]. The variation of ITS copies within individuals is also known for some insects
[103][104][105]. Analysis of ITS2 data of
S. aurantii indicates the presence of multiple non-identical copies of spacer sequences
[102]. Differences in PCR-amplified product size and the inability to generate a reliable alignment of sequences due to the presence of indels may confound the ITS-based identification of some thrips species
[106].
3.6. Other Marker Genes Used for Thrips Identification
In addition to mitochondrial and rRNA-ITS, other genes including histone H3, elongation factor (EF) 1-α, and cytoskeleton maker α -tubulin have also been used in thrips phylogenetic studies. Histone H3 efficiently determined gene flow from an arrhenotokous form of
T. tabaci to thelytokous form by confirming the passage of the arrhenotokous male-originated histone H3 gene allele to the F
2 generation
[107]. Histone H3 combined with COI and 28s rRNA reveals that asexuality in
A. stylifer Trybom and
A. karnyi John has a genetic basis, while it is governed by endosymbionts in
A. rufus [108]. Histone H3, α-tubulin, and EF1-α support concatenated phylogenetic analysis of thrips species together with commonly used mitochondrial and nuclear markers
[109]. A combined rRNA and EF1α tree is well suited for differentiating
Scritothrips lineage from
Frankliniella [45]. Histone H3 and EF1α are also useful for substantiating intra-population and inter-population genetic diversity in sexual and asexual populations of
A. rufus [110]. However, phylogenetic analysis of
T. palmi, T. nigropilosus Uzel, and
T. flavus Schrank using histone H3 shows clear overlaps of interspecific and intraspecific distances without a barcode gap
[82]. This suggests that histone H3 may be a useful marker for the identification of discrete genera rather than for species-level studies across a genus
[86].
3.7. SSR/Microsatellite Markers
Microsatellites or simple sequence repeats (SSRs) are repetitive DNA motifs composed of 1–6 bp in both coding and non-coding regions of the genome. SSRs are preferred markers for population genetics studies because of their high polymorphism and abundance, co-dominance, high allelic diversity, and ease of detection by PCR
[111][112]. SSRs provide demographic information on founder events, invasion history, local adaptation, allelic fixation index (FST), population size, and gene flow of insect pests.
Microsatellite markers helped to study the migration pattern of
D. minowai Priesner,
F. occidentalis, and
T. palmi [113][114][115].
D. minowai probably originated from multiple regions and gradually separated into two groups. High migration rates indicate gene flow from northeast to southwest China
[114]. Populations of
T. palmi that invaded early show relatively high genetic diversity compared to recently emerged populations. The analysis suggests limited ongoing dispersal and geographical isolation of populations by distance. Greenhouses may play a crucial role in the expansion of
T. palmi distribution to new areas
[115]. The genetic diversity of
F. occidentalis populations in China, USA, and Kenya indicates a relatively low level of gene flow
[116][117][118]. However, a considerable genetic divergence exists in
F. occidentalis populations between host plant species that suggests low gene flow and possible development of biotypes
[118]. An expressed sequence tag (EST) database of
F. occidentalis has also proven helpful in the development EST-SSRs
[119][120][121]. Similarly, six and eleven polymorphic SSR loci have been identified from an enriched genomic library in order to gain better insights into the genetic makeup and migration pattern of
S. perseae and
T. hawaiiensis [122][123]. More recently, high-throughput sequencing has been successfully utilized to identify SSRs in
F. occidentalis and
T. palmi [115][124][125]. Species-specific markers can also be designed for SSR-based identification of thrips species.
3.8. RFLP Markers
RFLP technique distinguishes individuals based on size differences of restriction fragments of an amplified DNA region generated by a specific or multiple sets of restriction endonucleases. RFLP has been successfully employed to diagnose different species and reproductive and color morphs of thrips as detailed below.
ITS-RFLP technique has been used to identify important thrips species such as
F. bispinosa (Morgan),
Pezothrips kellyanus (Bagnall),
S. citri (Moulton),
S. dorsalis,
T. tabaci,
T. nigropilosus,
F. occidentalis F. intonsa,
F. pallida (Uzel),
F. tenuicornis (Uzel), and
A. obscurus without any cross-reactivity
[126][127][128]. The restriction pattern with
AluI and
Sau3AI allows unambiguous detection of many thrips species including
F. occidentalis,
T. palmi,
T. tabaci,
T. angusticeps,
Parthenothrips dracaenae (Heeger),
A. obscurus,
E. americanus,
H. femoralis,
H. haemorrhoidalis, and
T. picipes [32]. Thrips species, viz.,
E. americanus,
F. occidentalis,
F. tenuicornis,
Helionothrips aino (Ishida),
H. spinosus Wilson,
H. haemorrhoidalis,
H. femoralis,
Limothrips cerealium Haliday,
L. denticornis (Haliday),
Moundothrips apterygus Wilson,
P. dracaenae,
Pseudanaphothrips achaetus (Bagnall),
Rhipiphorothrips cruentatus Hood,
Selenothrips rubrocinctus (Giard),
Sigmothrips aotearoana Ward,
Suocerathrips linguis Mound and Marullo,
T. nigropilosus,
T. physapus Linnaeus, and
T. tabaci can be discriminated from each other based on characteristic banding patterns of RFLP
[129]. Another RFLP protocol developed by Toda and Komazaki
[33] has allowed identification of nine species of thrips from Japanese fruit trees. A similar approach has been used by Rugman-Jones and colleagues for identifying seven species of
Scirtothrips [35]. The color morphs of
F. schultzei can also be diagnosed based on ITS-RFLP
[97]. COI-based RFLP can efficiently discriminate two reproductive morphs, arrhenotokous and thelytokous, of
T. tabaci upon the digestion of a 490 bp COI amplicon with EcoO109I
[130].
The major limitation of PCR-RFLP is the requirement for the specific restriction of endonucleases and the difficulty in identifying specific variations when several SNPs are targeted simultaneously. This limitation may be overcome by mixing two endonucleases in a single reaction. However, double and triple digests in RFLP add higher costs in post-PCR analysis
[131].
3.9. RAPD Markers
The RAPD technique uses random primers in PCR for rapid analysis of polymorphisms in genomic DNA
[132]. In the case of thrips, RAPD markers were used for the first time by Klein and Gafni
[133] to discriminate three morphotypes of
T. tabaci. Intraspecific genetic variants of
T. tabaci,
T. palmi, and
F. intonsa were also identified using RAPD
[134][135][136]. Using RAPD analysis, the Hungarian thrips population has been divided into two groups, Aeolothripidae (
A. intermedius Bagnall) and Thripidae (
F. intonsa,
K. robustus (Uzel),
Odontothrips confusus Priesner,
T. dilatatus Uzel, and
T. tabaci)
[31]. RAPD markers were also used to assess the population structure and inter-population and intra-population variability of
Gynaikothrips uzeli (Zimmermann)
[137].
RAPD is often used to complement RFLP and ISSR markers. For example, an analysis based on an RFLP marker followed by RAPD has enabled rapid discrimination of early larval stages of
F. occidentalis and
F. intonsa [138]. A combination of RAPD and ISSR markers used to characterize thrips populations in India indicates that RAPD markers are more informative than ISSR markers
[139]. The main drawback associated with RAPD is its dominant nature that reduces the information provided by each locus. This loss in information can be compensated by using a larger number of markers
[140]. RAPD may also generate unstable and variable amplicons due to the low annealing temperature of the short primers used in PCR.
3.10. AFLP Markers
To overcome the limitations of RAPD and RFLP, amplified fragment length polymorphism (AFLP) has been adopted
[141]. AFLP yields a large number of marker loci with an average of 50–100 amplicons per primer pair per sample. Moreover, AFLP is highly reproducible and co-dominant; however, dominant AFLPs are also amplified sometimes
[142]. AFLP can also be applied to cDNA and used to study differential gene expression in insects
[143].
AFLP markers have been found useful for studying genetic polymorphisms and relationships of
T. tabaci and
F. occidentalis. A few unique bands specific to each species may also be helpful in developing species-specific diagnostics
[144]. A high level of polymorphism among
F. occidentalis populations has been detected using AFLP markers that suggest that the population from the Netherlands may have migrated to Beijing, China
[145]. AFLP also helps to understand the host-related polymorphisms in
F. occidentalis populations; a thrips laboratory culture was the most distant from other populations in this analysis
[146].
Despite being more informative than RAPD and RFLP, AFLP requires more time to complete and uses radioactively labeled primers. The amplified products need to be resolved in polyacrylamide sequencing gels or by using automated genotyping equipment for scoring. Sometimes, the presence of microsatellites in the AFLP loci can make the scoring difficult
[142].