https://orcid.org/0000-0001-6835-679X
Figures
Abstract
Thrips are a serious pest in many crops. In onion cultivation, Thrips tabaci is the most important, but not the only thrips species causing damage. We investigated which thrips species affects onion and related species worldwide, how much genetic variation there is within T. tabaci populations, and how this evolves. Furthermore, we determined the reproductive mode and the correlation between the genetic and geographic distances. Thrips samples from infested onions or related species were obtained from 14 different locations worldwide. Species and haplotypes were determined through DNA barcoding with the mitochondrial Cytochrome Oxidase subunit I (COI) gene. Thrips tabaci was the most commonly observed species, but Scirtothrips dorsalis, Thrips palmi, Frankliniella intonsa, Frankliniella occidentalis and Frankliniella tenuicornis were also found, especially at the beginning of the growing seasons and depending on the location. The Nei’s genetic distance within T. tabaci was less than 5% and the haplotypes were clustered into two phylogenetic groups, each linked to a specific mode of reproduction, thelytokous or arrhenotokous. Thelytokous thrips were more common and more widely distributed than arrhenotokous thrips. A high percentage of heteroplasmy was detected in the arrhenotokous group. Heteroplasmic thrips were only found in populations where thelytokous and arrhenotokous were present in sympatry. Some T. tabaci haplotypes were present in high frequency at several sampled locations. No correlation was found between the genetic and geographic distances, which points to anthropic activities spreading thrips haplotypes throughout the world.
Citation: Porta B, Vosman B, Visser RGF, Galván GA, Scholten OE (2023) Genetic diversity of thrips populations on Allium species around the world. PLoS ONE 18(8): e0289984. https://doi.org/10.1371/journal.pone.0289984
Editor: Muhammad Faisal Shahzad, Gomal University, PAKISTAN
Received: March 3, 2023; Accepted: July 28, 2023; Published: August 17, 2023
Copyright: © 2023 Porta et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting information files. The thrips COI sequences found were submitted to the NCBI database under accessions numbers OQ603129 - OQ603255.
Funding: BP & GG, fellowship POS_EXT_2017_1_146409, ANII (Agencia Nacional de Investigación e Innovación) - Uruguay, https://www.anii.org.uy. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. OS & BV, TU18080, Topsector Horticulture and Starting Materials, the Netherlands, https://www.topsectortu.nl/en, this funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript, together with Bejo Zaden B.V., https://www.bejo.com, East-West International B.V., https://www.eastwestseed.com, Limagrain, https://www.limagrain.nl, Nunhems Netherlands B.V., https://www.nunhems.com, Takii & Company Ltd., https://www.takiiseed.com. These funders collected thrips from their experimental fields and delivered them to BP for analysis, they accepted with publication of the manuscript. RV, Plant Breeding Wageningen University & Research gave support of one year to BP because of Covid related delay in research. This funder 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
Thrips are small insects of around 1 mm long that belong to the order Thysanoptera. This order contains 6,200 species [1], all with fringed wings and asymmetrical mouthparts. Most species feed on plants by puncturing the cells and sucking up their content, while just a few species are predators of other insects or mites [2]. A small number of thrips species are beneficial as pollinators. Nevertheless, several thrips, especially the species that show high adaptability and a polyphagous nature, are serious pests in commercially important crops [2].
Thrips tabaci Lindeman (Thysanoptera: Thripidae), commonly known as onion thrips [3], is polyphagous and occurs worldwide. It infests a wide range of plant species including vegetables, fruits, and ornamentals [4]. Thrips tabaci is the major pest in onions worldwide [4]. It feeds directly on onion foliage, diminishing the photosynthetic capacity which can cause a more than 50% yield loss in commercial bulb production [5, 6]. Thrips tabaci has a high reproductive rate [4], shows cryptic behavior, is resistant to several insecticides [7, 8], and acts as a vector of plant pathogens [9, 10], including tospoviruses such as Iris Yellow Spot Virus (IYSV) which can devastate an onion crop in a matter of weeks [6, 11]. All these characteristics make it a high-impact pest that is difficult to control [12].
Thrips tabaci comprises three phylogenetic groups associated with different host plants or reproductive modes [13]. One group is mainly found on tobacco plants [14]. The other two groups are mainly associated with onion and leek and correspond to the major reproductive modes, arrhenotokous and thelytokous [14, 15]. The arrhenotokous and thelytokous lineages differ in ecological characteristics like host preference [16] and pyrethroid resistance [7].
In the case of thelytokous thrips, reproduction is asexual by parthenogenesis in which the embryos develop into female offspring without fertilization. Arrhenotokous thrips can reproduce both sexually and asexually, producing female offspring from fertilized eggs [17], and haploid males from unfertilized eggs [15]. In rare cases, deuterotokous thrips that develop from unfertilized eggs in either male or female offspring via parthenogenesis have been reported [18]. A recent study established that deuterotoky in T. tabaci corresponds to an irregular mode of reproduction of the arrhenotokous thrips on tobacco [19]. The most common modes of reproduction, thelytoky, and arrhenotoky, are mainly mutually exclusive [20], but crosses between individuals belonging to these reproductive modes have been reported [21]. Such crosses may give thrips the ability to adapt more rapidly to changing environments, cope with pesticide pressures, and successfully infect various plant species [22].
Thrips tabaci originates from the Mediterranean region where onion is the main host plant [23]. From there, it has spread all over the world, colonizing habitats from sea level up to 2000 m [3]. How T. tabaci was distributed over the world is unknown. In a study with Chinese T. tabaci populations, no correlation was observed between genetic and geographic distances [24]. This may point to the migration of individual thrips, possibly through different routes, followed by adaptation to the new environment. Molecular data indicated that the three different phylogenetic groups in T. tabaci are not evenly distributed around the world [16].
Genetic diversity has different hierarchical levels, between and within species. The within-species variation can be subdivided into intra and inter-population diversity. How the variation is distributed over populations is highly influenced by reproductive mode, gene flow, and selection [25]. Genetic diversity within populations is determined by the presence of different genotypes, their relative frequencies, and the genetic distance between them.
Different genotypes of thrips can be discriminated by sequencing (part of) the Cytochrome Oxidase subunit I (COI) gene [13, 17, 24, 26]. The COI gene is part of the maternally inherited mitochondrial DNA and is commonly used for DNA barcoding of animals, including insects, to identify species. The mitochondrial genome is haploid and the genotypes distinguished are called haplotypes. Previous studies on genetic diversity in the Thrips genus showed that the COI gene is ideal for the simultaneous detection of intra- and interspecific genetic diversity, as the range of diversity within and between species does not overlap; between thrips species about 20% of genetic diversity is detected in the COI gene sequence, while within T. tabaci the genetic diversity is below 5% [26].
The COI gene can also be used to distinguish the different T. tabaci reproductive modes (or lineages), which are impossible to discriminate morphologically, either by using the COI sequence directly or by using reproductive-mode specific primers [15]. The mode-specific primers produce fragments of a specific length for each reproductive mode. In T. tabaci the presence of heteroplasmic individuals has been reported. Heteroplasmic thrips contain mitochondria with copies of the COI gene from different lineages [27], and in these cases the reproductive mode cannot be determined conclusively using specific primers; however, the reproductive mode may be deducted from their position in the phylogenetic tree or determined by a progeny test from virgin females [13].
Although T. tabaci is the most important pest in onion [5], other thrips species have been found on onion as well [28]. It is important to know which thrips species affect onion, their genetic diversity and distribution around the world, as well as the seasonal dynamics in the populations and the reproductive mode to develop improved control strategies to be implemented in future Integrated Pest Management programs [29]. This research aimed to 1) identify thrips species in fields with onions or onion-related species at 14 locations around the world; 2) determine genetic variation and mode of reproduction within T. tabaci, and relate these with the haplotypes reported in the NCBI sequence database; 3) compare the genetic variation within and between sampled thrips per location; 4) analyze the geographical distribution of the T. tabaci haplotypes as well as the correlation between the genetic and geographic distances of populations; and 5) analyze the seasonal variation in the composition of thrips populations at three locations.
Materials and methods
Thrips samples
Thrips samples were collected from onion or onion-related species present in experimental fields aimed to identify thrips resistance at 14 locations around the world (Table 1). Adult thrips or larvae (n>60) were randomly collected from different plants and stored in an Eppendorf tube containing 95% ethanol. Thrips were collected once (referred to as location) or at two or three different times (referred to as location-time) during the season (Table 1). Thrips samples were stored at room temperature prior to DNA extraction.
DNA extraction
Thirty-three individual thrips (adults or larvae) were randomly chosen from each sample. In one case (Konan 08.08.2019) only 16 adult thrips were collected. DNA was extracted from individual thrips using the DNeasy® Blood & Tissue Kit (QIAGEN, Valencia, CA), according to the suppliers’ recommendations.
Amplification of the COI fragment and sequencing
DNA from individual thrips was used as the template for PCR amplification, using the universal COI primer pair, MTD7.2F ATTAGGAGCHCCHGAYATAGCATT and MTD9.2R CAGGCAAGATTAAAATATAAACTTCTG [30]. Twelve μl PCR solution was prepared according to the manual of the Qiagen Kit (6 μl Multiplex PCR kit solution, 0.25 μl for each 10 nM primer MTD7.2F and MTD9.2R, 3,5 μl of thrips DNA and 2 μl ddH2O). The PCR amplification protocol used was 95°C for 15 min, then 35 cycles of 95°C for 1min, 50°C for 1 min, and 72°C for 1 min, followed by a final extension at 72°C for 10 min. The PCR products were examined on a 1.5% agarose gel and sequenced in forward and reverse directions using a 3500 ABI Sequencer (Applied Biosystems, USA). The forward and reverse sequences from each band were used to build the consensus sequence with the package SeqMan Pro of the DNASTAR LASERGENE 17 program.
PCR reaction to determine the reproductive mode of T. tabaci
The T. tabaci reproductive mode-specific PCR reaction was performed using three PCR primers, one universal (TCOR-ATTGCGTAAATTATTCCTAAAAGTCCA) and two reproduction-mode-specific primers: 1) mtCOA-SSP TCOS-AACAGCTATTCTCCTTCTTTATCTC, amplifying a fragment of 261 nucleotides in arrhenotokous reproducing thrips, and, 2) mtCOT-SSP TCOC-GAACAGTATATCCACCTTTATCAACG, amplifying a fragment of 451 nucleotides in case of thelytokous reproduction [15]. In the case of heteroplasmy, both fragments amplify at the same time [27].
DNA from individual thrips of the different T. tabaci haplotypes was used as the template for PCR amplification. Twelve μl of PCR solution was prepared according to the manual of the Qiagen Kit (6 μl Multiplex PCR kit, 0.25 μl for each strain-specific 10 μM primer mtCOA-SSP TCOS and mtCOT-SSP TCOC, 0.5 μl of the 10 μM primer TCOR, 3.5 μl of thrips DNA and 1.5 μl ddH2O). The PCR amplification protocol used was 95°C for 15 min, 35 cycles of 98°C for 10 s, 60°C for 1 min, and 68°C for 1 min, followed by a final extension at 72°C for 10 min. The size of PCR products was determined by 1.5% agarose gel electrophoresis, with a 100 bp DNA Ladder (Thermo Fisher Scientific) as a reference, and stained with RedSafe™. All the reproductive mode-specific DNA fragments obtained were purified and sequenced as described above. All the consensus sequences from the different T. tabaci haplotypes were aligned using CLUSTALW [31] in MEGA 7 [32] and grouped by the reproductive mode according to the size of the fragment (thelytokous and arrehenotokous, 451, and 261 nucleotides, respectively).
Thrips tabaci reproductive mode estimation and distribution worldwide
The molecular determination of the reproductive mode [15] is not conclusive in the case of heteroplasmy [27]. Furthermore, it was also not possible to determine the reproductive mode via a progeny test because of the destructive methodology used for DNA extraction. Therefore, the reproductive mode of heteroplasmic individuals was derived from their position in the phylogenetic tree. A phylogenetic analysis was performed, including all T. tabaci haplotypes collected in our research plus the ones reported in the NCBI database for which the reproductive mode was indicated in the passport data or published papers. One F. occidentalis haplotype from NCBI was used as an outgroup. The congruence between the reproductive mode derived from mode-specific primers and the position in the phylogenetic tree was confirmed for non-heteroplasmic individuals. The frequency of thelytokous and arrhenotokous T. tabaci per location was calculated. A Chi-square test was carried out to determine if the thelytokous and arrhenotokous T. tabaci were evenly distributed over locations.
Data processing and analyses
Database generation, COI sequences alignment, and primer trimming.
A COI gene database was generated containing all the consensus sequences obtained from the individual thrips sequenced. All consensus sequences were ordered and grouped in MEGA 7 by location or location-time for those locations that were sampled more than once (Konan, Pulpi, and Sansai) (Table 1). Finally, all consensus sequences were aligned using CLUSALW [31] in MEGA 7 [32] after they had been trimmed from forward and reverse primer sequences.
Thrips haplotypes identification, frequency, and phylogenetic analysis.
Different haplotypes were identified with the Pop Art program [33] and blasted to the NCBI database to identify the species they belong to. To analyze the relationship among the different haplotypes identified, a Neighbor-Joining tree was constructed with the Mega Align Pro package of the DNASTAR program. All thrips haplotype sequences detected have been submitted to the NCBI database, under accession numbers OQ603129-OQ603255.
Location or location-time thrips haplotype constitution and genetic distances.
To depict which haplotypes were present at the sampled locations or locations-time, a stacked bar chart was made with InfoStat [34] using the haplotypes frequencies per location or location-time as input.
Nei’s genetic diversity was calculated as the number of base substitutions averaged over all pairwise comparisons in the thrips COI gene fragment by location or location-time using MEGA 7 [32]. Within a sample, Nei’s genetic diversity represents the average evolutionary divergence across the thrips sequence pairs among the 33 thrips analyzed per location or location-time. For samples with several species, the genetic diversity per population was calculated only for T. tabaci using the COI sequence.
To calculate the Nei’s genetic distance between T. tabaci populations the 545 T. tabaci sequences were grouped by location or location-time in MEGA 7 and the number of base substitutions per site was averaged over all sequence pairs between samples [32]. The genetic diversity within samples as well as the genetic distance between them were analyzed using the Maximum Composite Likelihood model [35] implemented in MEGA 7.
Correlation between genetic distances and geographic distances
The correlation between the genetic and geographic distances, of the T. tabaci populations at the different locations, was calculated using the matrix of Nei genetic distances and the matrix of geographic distances among locations. For locations with more than one sampling time, each sample was considered a population, and the geographic distance between them a zero. The matrix of Nei genetic distances for the T. tabaci populations was calculated in MEGA 7. The matrix of geographic distances, among the different sampling locations, was calculated using the Diva Gis program [36]. The correlation between the genetic and geographical distance was calculated through a Mantel test conducted in Genalex 6.5 [37]. The geographic distance matrix was used as X and the genetic distance matrix as Y, and 999 permutations were applied.
Retrieving and processing T. tabaci COI sequences from NCBI
All COI sequences of T. tabaci available in the NCBI database on November 25, 2022 were retrieved (https://www.ncbi.nlm.nih.gov/nuccore/?term=thrips+tabaci+coi). This resulted in a total of 504 sequences. Sequences were trimmed to obtain a fragment between positions 341 and 775 of the mitochondrial COI gene, that coincides with the fragment of our study. Sequences that missed more than five nucleotides at the beginning or end of the sequence compared to our COI gene fragment, were discarded. The T. tabaci haplotypes obtained in the present study were named H followed by a number, or HU in the case a haplotype was unique in our dataset (represented by one individual thrips only). A Neighbor-Joining tree [38] was constructed in MEGA 7 [32] based on the analysis of 170 sequences of T. tabaci and one of F. occcidentalis (accession JQ082479.1 from the NCBI database) as an outgroup (S1 Table). In the NJ tree, only different sequences are presented, whereas the identical ones are reported in a separate table. In the analysis, the option “pairwise deletion of the missing data” was used, since some NCBI sequences had up to five missing nucleotides due to the incomplete coverage (S1 Table). The genetic distances were computed using the Maximum Composite Likelihood method [35] and units are presented as the number of base substitutions per site.
Results
Diversity of thrips species in onion and related Allium species
Five hundred and eighty-six individual thrips from 14 locations were sequenced for the COI region. All data can be found in S2 Table. From all thrips analyzed, 545 (93%) were T. tabaci. Other species found were Scirtothrips dorsalis, Thrips palmi, Frankliniella intonsa, Frankliniella occidentalis and Frankliniella tenuicornis (Fig 1). Thrips tabaci was present at all locations, while the other species were observed only at four of the 14 locations (Table 2). In Sansai (Thailand), four thrips species were found, T. tabaci, S. dorsalis, T. palmi and F. intonsa. In Pulpi (Spain) and Krabbendam (The Netherlands) two species were found, T. tabaci and F. occidentalis, whereas in De Kwakel (The Netherlands), one individual of F. tenuicornis was found next to T. tabaci. Thrips species different from T. tabaci were found in Allium spp. fields only, while in the onion fields all the thrips were T. tabaci.
The tree is based on 434 nucleotides of which 184 were polymorphic and 52 of them were polymorphic among T. tabaci haplotypes. Groups 1 and 2 contain T. tabaci haplotypes. The upper right chart shows the Pairwise Nei’s genetic distances among all T. tabaci haplotypes and between the T. tabaci haplotypes and F. occidentales, F. tenuicornis, F. intonsa, T. palmi and S. dorsalis haplotypes.
The locations Pulpi and Sansai were sampled two and three times, respectively, and showed other species besides T. tabaci in the first or second sampling, while later in the season (last sampling) only T. tabaci was found (Fig 2; S3 Table). In Sansai, at the beginning of the season, the dominant species was S. dorsalis (85%) (Fig 2a, S3 Table). After 21 days, at the second sampling, the frequency of S. dorsalis had decreased to 3%, while T. tabaci had increased to 61%, and also F. intonsa (10%) and T. palmi (26%) were found. In the third sample, taken 20 days later, only T. tabaci was found (Fig 2b, S3 Table).
A) Locations or location-time in which thrips haplotypes other than T. tabaci were found. The white parts of the circles correspond to T. tabaci. Other species and their haplotypes are indicated by different colors, see legend below the circles: F.t. F. tenuicornis, F.o. F. occidentalis, S.d. S. dorsalis, F.i. F. intonsa, T. p. T. palmi, B) Thrips tabaci haplotype composition at each location or location-time. The Y-axes show the frequencies of each haplotype present at the sampled location or location-time. Each haplotype is represented by a different color according to the color legend below the graph. Codes with an H represent the haplotypes with more than one individual thrips, and with HU the unique haplotypes.
Genetic diversity within Thrips tabaci
Among the 545 T. tabaci sequenced, 58 COI haplotypes were found (Table 3). The mean Nei’s genetic distance among T. tabaci haplotypes was 0.023. The mean genetic distance between the T. tabaci haplotypes and those of other species was 0.193 (Fig 1). All T. tabaci haplotypes were grouped in one clade consisting of two groups, while the haplotypes of the other species were grouped into species-specific clades (Fig 1).
The most common haplotype of T. tabaci was H1 (28.4%), it was present at 10 of the 14 locations. In Tulancingo de Bravo, haplotype H1 was the only haplotype found. In Huron, H1 was the most important with a genetic diversity close to zero (Table 4, Fig 2b, S3 Table). The second most common haplotype was H7 (8%), present at 4 of the 14 locations (Eyragues, Pulpi, Berurim, and Sansai) followed by H4 and H28, both at a frequency of 6%. H4 was only found at the two locations in The Netherlands and in Berurim, while H28 was found in Asia, South America, and Europe (Fig 2b, S3 Table). H8 is another haplotype among the most common ones (5.1%). Haplotype H8 was only present in Konan at all three sampling times, as the most common haplotype. In addition, 19 unique (represented by only one thrips) haplotypes (HU40 to HU58) were identified considering all locations and time points together, representing 3.5% of all T. tabaci (Table 3). The largest genetic diversity among T. tabaci was observed in Konan, where 15 haplotypes were identified over three sampling times (Table 4).
Changes in T. tabaci genetic diversity within locations over time
The three locations sampled more than once, Pulpi, Sansai, and Konan, showed changes in the genetic composition of T. tabaci over time, with the least diversity at the end of the season (Table 4, Fig 2b).
In Pulpi, 15 T. tabaci haplotypes were identified in the first sample. The three most frequent haplotypes were H7, H18, and H23 (Fig 2b, S3 Table). The second sample consisted of seven T. tabaci haplotypes of which H10 and H15 were new. Haplotypes H7 and H23 were the dominant haplotypes and their frequencies increased three and nearly two times, respectively. This resulted in a reduction in diversity, from 0.016 to 0.008 between samplings (Table 4).
In Sansai the first sample contained only five T. tabaci individuals (15%), all with a different haplotype (Fig 2b, S3 Table). In the second sample, the frequency of T. tabaci increased to 61%, of which 90% were the haplotype H1. The third sample contained T. tabaci only, again with H1 as the dominant haplotype (Fig 2b, S3 Table). Compared to the second sample, haplotypes H7 and H34 were no longer present, but five new haplotypes were observed (Fig 2b, S3 Table).
In Konan, all three samples consisted of T. tabaci only. In the first sample, seven haplotypes were identified with H8 as the dominant one (44%) followed by H1 (25%) (Fig 2b). In the second sample, 11 haplotypes were found, with H8 again as the most common (31%), followed by haplotypes H10 (21%), and H9 (12%). In the third sample, also 11 haplotypes were identified, with H19 and H32 as new haplotypes replacing H20 and HU46. H8 remained the most common haplotype (35%), followed by H9 (10%). The frequencies of other haplotypes were below 10% (Fig 2b, S3 Table).
Genetic diversity within Thrips tabaci and mode of reproduction
Within the COI gene fragment of 434 nucleotides, 52 sites were polymorphic (Table 3). Based on the NJ tree (Fig 1), the 58 T. tabaci haplotypes clustered into two main groups, except for one haplotype (HU58) which did not belong to any group. Of the 58 haplotypes, 20 had the thelytokous fragment (451 bp) and were present in group 1 of the NJ tree, 11 haplotypes had the arrhenotokous fragment (261 bp) and were present in group 2, and 26 haplotypes had both fragments. Haplotype HU58 did not amplify in the reproductive mode PCR-specific reaction. Among the haplotypes having both fragments, three belonged to group 1, and 23 to group 2 (Fig 1 and S2 Fig).
Alignment of the sequences obtained for the 26 haplotypes with double (arrhenotokous and thelytokous) bands [15] showed that these fragments overlap. The arrhenotokous fragment overlapped from position 192 to the end of the thelytokous fragment. Within the thelytokous fragment sequence, there is a T insertion at the position 212 (S4 Table). This confirms the presence of heteroplasmy in haplotypes with both fragments since the arrhenotokous specific primer sequence is not present within the thelytokous fragment sequence.
The majority (72%) of individual T. tabaci analyzed belonged to nine common and widely spread haplotypes (H1 through H5, H7, H8, H10, H28) (Table 3, S3 Table), of which six haplotypes belonged to group 1 (Fig 1) and had the thelytokous band (Table 3), while the others (H8, H10, and H28) belonged to group 2 (Fig 1) with both bands and are thus heteroplasmic (Table 3). The other 28% of individual T. tabaci analyzed belonged to the 49 fewer common haplotypes with frequencies below 0.04. Seventeen of these haplotypes belonged to Group 1 and had either the thelytokous band (14) or both bands (3), while 31 belonged to Group 2 of which 11 had only the arrhenotokous band and 20 had both bands.
Thrips tabaci geographic and genetic distances and distribution of reproductive mode
No correlation was found between the matrixes of Nei’s genetic distances for the T. tabaci populations collected by us and the geographic distances between the locations where they were collected (R2: 5E-07; p-value: 0.428) (S1 Fig). The frequency of the T. tabaci reproductive mode per location is presented in Fig 3. The Chi-square test (Χ2) indicated significant differences between the observed thelytokous and arrhenotokous T. tabaci frequencies. The null hypothesis, assuming that both reproductive modes are evenly distributed per location around the world was rejected (Χ2 (13, N = 544) = 333.05, p = 2.83 E-63).
Locations, where thrips were collected, are indicated by yellow dots. The reproductive mode is shown by pie charts.
Relationship between our T. tabaci haplotypes and NCBI accessions
A total of 112 accessions from the NCBI were aligned with our 58 T. tabaci haplotypes (S1 Table). The 112 NCBI accessions resulted in 45 different haplotypes because some of the NCBI accessions were identical for the 434 nucleotides COI gene fragment used in our study. Of these 45 haplotypes, 38 differed from the 58 haplotypes found in our study, resulting in a total of 96 different T. tabaci haplotypes. The phylogenetic tree, built with these T. tabaci haplotypes and F. occidentalis as an outgroup showed three main groups within T. tabaci (S2 Fig).
Group 1 consists of 49 haplotypes of which we collected 23 and 26 were reported in the NCBI database (S2 Fig). Among the haplotypes in Group 1, five of our haplotypes were reported earlier (see Table 5). An example is T. tabaci H1, which was collected in The Netherlands, France, USA, Mexico, Uruguay, Japan and Thailand, which is identical to 26 accessions from Pakistan collected in 2010 and 2011 and six accessions from China collected in 2011, 2018 and 2019. Group 2 in the phylogenetic tree consists of 41 haplotypes, representing 53 sequences (19 from the NCBI database plus 34 sequences from our study) (S2 Fig). Thirty-two of our haplotypes were not reported in the NCBI database (Table 5). The third Group defined in the phylogenetic tree is composed exclusively of five T. tabaci accessions from the NCBI collected in Greece in 2004 from tobacco plants.
Discussion
Thrips tabaci is the main species in onion cultivation and Allium spp, but not the only one
This research aimed to study thrips diversity in fields with onion or its related species around the world. We set out a comprehensive study in which we sequenced 33 randomly chosen individual thrips per location or location-time, allowing us to identify with a 95% certainty, all genotypes (haplotypes) present in the population with a frequency of at least 0.1 [39]. Some locations were sampled more than once during the season, making it possible to analyze changes in the thrips population over time. Thrips species identification was based on the homology between the COI gene sequences and the data reported in the NCBI database. The range of genetic distance found in the 434 bp COI gene fragment within species did not overlap with the range of genetic distance found between species meaning that the amplified COI region of the gene is a good fragment to distinguish T. tabaci from other species. The genetic distance between T. tabaci haplotypes was in all cases below 5%, which is similar to other studies that used the same COI primers set. Furthermore, the average genetic distance between T. tabaci haplotypes and the other species’ haplotypes was around 20%, also in agreement with previous studies [26, 30, 40]. So, the 434 bp COI gene fragment used in our study showed good discriminative power between species and, at the same time detected ample genetic variation within species, strengthening its suitability to detect genetic variation between and within species and populations.
We found that the main species of thrips that affects onion and its related species is Thrips tabaci, which is in agreement with previous studies [5, 29, 41, 42]. Other thrips species such as Thrips palmi, Scirtothrips dorsalis, Frankliniella tenuicornis, Frankliniella intonsa, and Frankliniella occidentalis, were present at specific locations. The presence of F. occidentalis has previously been reported in onions in Tanzania [43] and Georgia [28]. In Georgia, the tobacco thrips, Frankliniella fusca, was the most important species in onion [28]. However, the number of incidents with T. tabaci in onions in Georgia is increasing [28], which might be explained by the fact that T. tabaci outcompetes F. fusca on onion [44]. The occurrence of F. intonsa in A. cepa and A. fistulosum has also been reported in Poland [29]. Thrips palmi and S. dorsalis are commonly reported in melon and pepper, respectively [45] and both thrips species occur occasionally in Allium spp. [46, 47]. Finally, in our study, we report for the first time F. tenuicornis in Allium. Until now, F. tenuicornis was reported as specific for wild and cultivated Gramineae [48].
Thrips species other than T. tabaci were found at certain locations only. Which thrips species are found is likely to be determined by the climatical conditions and centers of origin of the species. Sansai is located in the tropics, which may explain the presence of thrips species like S. dorsalis and T. palmi which are typical for these regions [49, 50]. In general, thrips species diversity is larger in the tropics and neotropics than in temperate regions [51]. The presence of pepper as a former crop in Sansai may explain the presence of S. dorsalis, commonly known as chili thrips. Pepper is the main host of S. dorsalis [52], but its host range includes a wide diversity of plant taxa from different families that enable S. dorsalis to remain in an area during intercropping [53]. The presence of F. intonsa at Sansai might be associated with the presence of Allium fistulosum in the Allium spp field since it has been reported as the common host plant for F. intonsa [29]. Apart from that, the Allium field in Sansai was surrounded by several (crop) species, including banana, bamboo, and herbs such as marigold and weeds, from which thrips may have moved into the field. In the case of De Kwakel, the field was surrounded by wild Gramineae grasses which are the typical host of F. tenuicornis [48], which might explain its presence. As a consequence, a crop should be considered as a unit integrated with the adjacent and previous vegetation [54, 55] and not as an isolated island.
Our results also show that the change in species composition at certain locations can be fast and dynamic, with T. tabaci as the unique species present at the end of the season. Phenomena that may be explained by a combination of factors, including a higher competition capacity [44], higher reproductive rate [56], and migration of the other species [57].
In the sampled onion fields only T. tabaci was found. In fields planted with Allium spp., other thrips species were found as well, although not all Allium spp. fields sampled had thrips species other than T. tabaci. Whether the more diverse set of host plants in the Allium spp. fields has played a role in shaping the thrips communities and populations remains to be established. Our setup does not allow us to draw any firm conclusions on this.
Thrips tabaci diversity
Groups and reproductive systemc.
The T. tabaci haplotypes obtained from onion and related species clustered into two phylogenetic groups. The haplotypes within group 1 (Fig 1) may be classified as thelytokous since all of them grouped in the same cluster with the haplotypes previously reported as thelytokous in the NCBI (S2 Fig, Table 5 and S1 Table) and they showed the PCR fragment which is specific for the thelytokous mode of reproduction (Table 3). The haplotypes within group 2 may be classified as arrhenotokous since they grouped with the arrhenotokous haplotypes reported in the NCBI and they amplified the PCR fragment specific for the arrhenotokous mode of reproduction. However, both groups also contained haplotypes that amplified both specific fragments, a phenomenon known as heteroplasmy [27]. Heteroplasmy was more common within group 2 (68% against 13% in group 1). As a consequence, the PCR reactions [15] were not conclusive in determining the reproductive mode. We assume that using the phylogenetic position (group 1 or 2) may be the best way to determine the reproductive mode in these cases, but more progeny tests are needed to conclusively establish that all heteroplasmic individuals in a certain group indeed display the expected mode of reproduction.
We found all the heteroplasmic haplotypes (except HU48) in populations that contained individuals of both phylogenetic groups (sympatric distribution), suggesting that they are the result of crossings between members of these groups. Such crossings have already been reported [20, 21]. The high percentage of heteroplasmy in group 2 (arrhenotokous) might be explained by the frequent paternal leakage [58] during sexual reproduction among arrhenotokous haplotypes, while within group 1 there is no occurrence of that since the haplotypes reproduce parthenogenetically. In both groups heteroplasmy in successive generations would be maintained via the mother line [27], if selection and genetic drift do not act against it [59].
Novel variation seems to be generated. We found clear examples of mutations. One individual at Huron (haplotype HU40) is only one base pair different from H1, and therefore a likely descendant of that haplotype. Also, recombination between different COI haplotypes that coexist in heteroplasmic individuals may be a source of novel variants, a mechanism likely to occur in group 2, where the percentage of heteroplasmy is high. We found 51 T. tabaci haplotypes not reported before in the NCBI database of which the majority (32) belongs to the group 2 (arrhenotokous lineage).
The low frequency of most arrhenotokous haplotypes is in agreement with their limited presence in the NCBI database. Many arrhenotokous haplotypes might have been undetected in former studies due to limited sampling efforts [39]. Since we sampled quite extensively, more arrhenotokous haplotypes present in low frequency were detected. Another explanation may be the narrow host range of arrhenotokous thrips compared to thelytokous [16]. The absence or limited presence of suitable hosts for arrhenotokous thrips during certain periods of the year may lead to bottlenecks in their populations reducing the haplotype frequencies.
In our study, six out of the nine more frequent haplotypes were thelytokous and showed a global distribution, which might be expected from their generalist nature [16]. Of the less frequent 49 haplotypes, 31 belong to group 2. These haplotypes may represent the less adapted or recently emerged haplotypes, as most of them are restricted to a specific location and are arrhenotokous.
Thrips tabaci is reported in more than 120 countries, but only for 29 countries, there is molecular data related to reproductive mode. Thelytokous thrips were found in all 29 countries while the arrhenotokous were reported in 16 countries only [16]. We found arrhenotokous T. tabaci for the first time in Argentina, Uruguay, France, Spain, and Thailand, and we confirmed their presence in Israel [17], Japan [13], and The Netherlands [17].
Thelytokous and arrhenotokous T. tabaci were not evenly distributed around the world. The thelytokous thrips (group 1) were present at all sampled locations (Fig 3), and five populations were composed solely of thelytokous haplotypes. Only a few arrhenotokous haplotypes showed a relatively high frequency in certain populations. The uneven distribution of both reproductive modes around the world might be explained by the availability of arrhenotokous preferred hosts [16] throughout locations and years.
Thrips tabaci, population diversity and dynamics over time.
Based on our results the diversity of T. tabaci within locations is reduced towards the end of the season, suggesting that some genotypes (haplotypes) have higher fitness compared to others. For example, H8 in Konan (Japan), H1 in Sansai (Thailand), and H7 in Pulpi (Spain) were the haplotypes most frequent across the season at these locations. In the case of H8 a positive interaction between haplotype and location might be expected since this haplotype is exclusively present in Konan. For the haplotypes H1 and H7 an overall higher fitness might be inferred since they were found at several locations. It will be interesting to see if the most frequent haplotypes across the season at certain locations are maintained across different years.
Different levels of genetic diversity are reported for T. tabaci populations from different locations or years [24, 41, 43, 60]. Based on our findings it is clear that the diversity of T. tabaci populations not only varies within a season but also among locations. The absence or nearly zero genetic diversity in Tulancingo de Bravo (Mexico) and Huron (USA) is not commonly reported and might point to a relatively recent invasion of T. tabaci or a poor adaptation of other haplotypes. On the other hand, highly diverse T. tabaci populations were also detected, distant from each other and composed of several haplotypes from different phylogenetic groups.
Genetic diversity in pests is expected to be high in the regions that coincide with the center of origin of the pest, and lower in places that were invaded recently [49]. High diversity at locations far away from the presumed center of origin points at invasions that took place a long time ago, resulting in well-adapted haplotypes from both phylogenetic groups. We found the most diverse T. tabaci populations both close to the center of origin [3] (Pulpi) as well as far away in Konan, Progreso, and Salto Grande (Table 4, Fig 2b). This result may also indicate that the center of origin of T. tabaci is larger than originally thought [3], or that some locations could be considered as secondary centers of diversification, former introduction sites in which adaptation and generation of new diversity took place.
The distribution pattern of the haplotypes throughout the world and their frequencies at different locations explain the lack of correlation between the geographical and genetic distances. This lack of correlation has also been reported in a study within China with T. tabaci populations [24] and F. occidentalis populations [61]. The fact that distant populations are genetically similar might be explained by the spread of the haplotypes by people [61]. Plant trading and agriculture have already been suggested as key activities in the spread of other insect pests such as Bemisia tabaci [62]. Human activities could result in high gene flow and low genetic variability between locations at a large scale [24]. The finding that nine T. tabaci haplotypes (H1, H28, H5, H19, H10, H3, H36, H23, and H27) are present both in Uruguay (Salto Grande, North, and Progreso, South) and Konan (Japan) can most probably be explained as the result of commercial trade of garlic and/or onion as a cause of the gene flow between these distant locations.
Conclusions
Thrips tabaci was the main species present in Allium fields. Nevertheless, other thrips species were also found in small numbers. Thrips other than T. tabaci were found at specific locations with diverse surroundings, mainly in the tropics and particularly frequent at the beginning of the growing season. In T. tabaci the genetic distance among haplotypes was below 5% and two different phylogenetic groups were found, which can be linked to the reproductive modes thelytokous (group 1) and arrhenotokous (group 2). Within group 2 there was a remarkable presence of heteroplasmic individuals. The thelytokous thrips were more frequent and worldwide distributed, while arrhenotokous thrips have a more restricted distribution and were always found in sympatry with the thelytokous thrips. The level of genetic diversity within populations differed among locations. Some T. tabaci haplotypes were present in a high number at different locations worldwide. No correlation was found between the genetic and geographic distances, suggesting that human activities may have spread thrips haplotypes throughout the world.
Supporting information
S1 Fig. Geographical versus genetic distance of the sampled thrips populations.
Nei’s genetic distances between the T. tabaci populations sampled at 14 different locations and the geographic distances in kilometres between the locations. There was no correlation between the Nei’s genetic distances among T. tabaci populations at the different locations and its geographical distances (Mantel test p value: 0.428).
https://doi.org/10.1371/journal.pone.0289984.s001
(TIF)
S2 Fig. Neighbor-Joining tree showing the relationships between the 58 T. tabaci haplotypes of the present study (represented as H or HU and the number) and 38 other T. tabaci haplotypes previously reported in the NCBI database (represented by accession number) and 1 sequence from NCBI database of F. occidentalis as an outgroup.
The COI gene fragment used consisted of 434 nucleotide positions. Branch lengths is based on the number of base substitutions per site. The average genetic distance among all T. tabaci sequences in the phylogenetic tree was 0.027. Next to the accession or haplotype name, the number of times a specific fragment was found among the 112 NCBI accessions numbers and our 58 haplotypes is shown between brackets, and further detailed in Table 5. Group 1 is regarded as thelytokous thrips based on the passport data from the NCBI database or former papers [13, 14, 24], accessions reported as thelytokous are colored in green, accessions or haplotypes identical to accessions reported with thelytokous reproductive system are within a green rectangle (Table 5 and S1 Table). Group 2 is regarded as arrhenotokous thrips, the accessions with this reproductive mode are colored in orange, accessions or haplotypes identical to accessions reported as arrhenotokous are within an orange rectangle (Table 5 and S1 Table). Group 3 represents the tobacco group [14].
https://doi.org/10.1371/journal.pone.0289984.s002
(TIF)
S1 Table. Passport data and sequences of the 112 T tabaci and 1 F. occidentalis NCBI accessions used to build the neighbor joining tree together with the sequences of the 58 haplotypes detected in our study.
In the last two columns the total sites and coverage of the NCBI accessions are depicted in comparison with the haplotypes found in our study after all sequences aligned and trimmed before and after position 1 and 434 respectively of the fragment amplified in our study.
https://doi.org/10.1371/journal.pone.0289984.s003
(XLSX)
S2 Table. Sequences of the 434 bp fragment of the COI gene of 586 thrips sampled at 14 locations worldwide.
For each individual thrips the haplotype and species are presented. The latter is based on a BLAST against the sequences in the NCBI database.
https://doi.org/10.1371/journal.pone.0289984.s004
(XLSX)
S3 Table. Thrips haplotypes frequencies per location or location time.
a) Thrips haplotypes frequencies considering all the thrips sequenced per sample. b) Thrips tabaci haplotype frequencies per sample considering only the T. tabaci in the samples that were originally composed by multiple thrips species. In all cases the frequencies are colored by conditional formatting from red to white in descendent order respectively.
https://doi.org/10.1371/journal.pone.0289984.s005
(XLSX)
S4 Table. COI gene fragments generated in the reproductive mode specific PCR reaction for thrips that showed double bands.
The different COI fragments are grouped per reproductive type.
https://doi.org/10.1371/journal.pone.0289984.s006
(XLSX)
Acknowledgments
The breeding companies Bejo Zaden B.V., East-West International B.V., Limagrain, Nunhems Netherlands B.V., Takii & Company Ltd. are acknowledged for maintaining experimental fields with onion and wild relatives and the collection of thrips samples in those fields all over the world. B. Porta acknowledge the support received from Facultad de Agronomía–Universidad de la República and from ANII (Uruguay). We also thank M.C. Bellacomo, who collected the samples in Bahía Blanca, and E. Vicente, V. Galván and B. Ghelfi who sampled the thrips population in Salto Grande.
References
- 1.
Mound LA. Biodiversity of Thysanoptera. Insect Biodiversity: Science and Society: John Wiley and Sons; 2018. p. 483–99.
- 2. Mound LA. Thysanoptera: diversity and interactions. Annu Rev Entomol. 2005;50:247–69. pmid:15355240.
- 3.
Lewis T. Thrips, their biology, ecology and economic importance. London and New York: Academic Press; 1973. 349 p.
- 4. Gill HK, Garg H, Gill AK, Gillett-Kaufman JL, Nault BA. Onion thrips (Thysanoptera: Thripidae) biology, ecology, and management in onion production systems. Journal of Integrated Pest Management. 2015;6(1):1–9.
- 5. Diaz-Montano J, Fuchs M, Nault BA, Fail J, Shelton AM. Onion thrips (Thysanoptera: Thripidae): a global pest of increasing concern in onion. J Econ Entomol. 2011;104(1):1–13. Epub 2011/03/17. pmid:21404832.
- 6. Boateng CO, Schwartz HF, Havey MJ, Otto K. Evaluation of onion germplasm for resistance to iris yellow spot (iris yellow spot virus) and onion thrips, Thrips tabaci. Southwestern Entomologist. 2014;39(2):237–60.
- 7. Toda S, Morishita M. Identification of three point mutations on the sodium channel gene in pyrethroid-resistant Thrips tabaci (Thysanoptera: Thripidae). Journal of Economic Entomology. 2009;102(6):2296–300.
- 8. Adesanya AW, Waters TD, Lavine MD, Walsh DB, Lavine LC, Zhu F. Multiple insecticide resistance in onion thrips populations from Western USA. Pestic Biochem Physiol. 2020;165:104553. Epub 20200304. pmid:32359535.
- 9. Leach A, Hay F, Harding R, Damann KC, Nault B. Relationship between onion thrips (Thrips tabaci) and Stemphylium vesicarium in the development of Stemphylium leaf blight in onion. Annals of Applied Biology. 2019;176(1):55–64.
- 10. Grode A, Chen S, Walker ED, Szendrei Z. Onion thrips (Thysanoptera: Thripidae) feeding promotes infection by Pantoea ananatis in onion. J Econ Entomol. 2017;110(6):2301–7. pmid:29112728.
- 11. Pozzer L, Bezerra I. C., Kormelink R., Prins M., Peters D., Resende R. O., and de Avila A. C. Characterization of a tospovirus isolate of iris yellow spot virus associated with a disease in onion fields in Brazil. Plant Dis. 1999;83:345–50. pmid:30845585
- 12. Morse JG, Hoddle MS. Invasion biology of thrips. Annu Rev Entomol. 2006;51:67–89. Epub 2005/12/08. pmid:16332204.
- 13. Sogo K, Miura K, Aizawa M, Watanabe T, Stouthamer R. Genetic structure in relation to reproduction mode in Thrips tabaci (Insecta: Thysanoptera). Appl Entomol Zool. 2015;50(1):73–7.
- 14. Brunner PC, Chatzivassiliou EK, Katis NI, Frey JE. Host-associated genetic differentiation in Thrips tabaci (Insecta; Thysanoptera), as determined from mtDNA sequence data. Heredity (Edinb). 2004;93(4):364–70. Epub 2004/07/09. pmid:15241445.
- 15. Kobayashi K, Hasegawa E. Discrimination of reproductive forms of Thrips tabaci (Thysanoptera: Thripidae) by PCR with sequence specific primers. J Econ Entomol. 2012;105(2):555–9. Epub 2012/05/23. pmid:22606827.
- 16. Loredo Varela RC, Fail J. Host plant association and distribution of the onion thrips, Thrips tabaci cryptic species complex. Insects. 2022;13(3). pmid:35323598
- 17. Toda S, Murai T. Phylogenetic analysis based on mitochondrial COI gene sequences in Thrips tabaci Lindeman (Thysanoptera: Thripidae) in relation to reproductive forms and geographic distribution. Appl Entomol Zool. 2007;42(2):309–16.
- 18. Nault BA. Reproductive modes in onion thrips (Thysanoptera: Thripidae) populations from New York onion fields. Environ Entomol. 2006;35:1264–71. https://doc.rero.ch/record/291887/files/ee35-1264.pdf.
- 19. Woldemelak WA. The existence of deuterotokous reproduction mode in the Thrips tabaci (Thysanoptera: Thripidae) cryptic species complex. Journal of Horticultural Research. 2020;28(1):21–8.
- 20. Jacobson AL, Nault BA, Vargo EL, Kennedy GG. Restricted gene flow among lineages of Thrips tabaci supports genetic divergence among cryptic species groups. PLoS One. 2016;11(9):e0163882. Epub 20160930. pmid:27690317.
- 21. Li XW, Wang P, Fail J, Shelton AM. Detection of gene flow from sexual to asexual lineages in Thrips tabaci (Thysanoptera: Thripidae). PLoS One. 2015;10(9):e0138353. Epub 20150916. pmid:26375283.
- 22. Hurst LD, Peck JR. Recent advances in understanding of the evolution and maintenance of sex. Trends in Ecology & Evolution. 1996;11(2):46–52. pmid:21237760
- 23. Mound LA. Biological diversity. Thrips as crop pest. 1997:197–215.
- 24. Li X, Zhang Z, Zhang J, Huang J, Wang L, Li Y, et al. Population genetic diversity and structure of Thrips tabaci (Thysanoptera: Thripidae) on Allium hosts in China, inferred from mitochondrial COI gene sequences. J Econ Entomol. 2020;113(3):1426–35. Epub 2020/01/27. pmid:31982906.
- 25. Ellegren H, Galtier N. Determinants of genetic diversity. Nat Rev Genet. 2016;17(7):422–33. Epub 20160606. pmid:27265362.
- 26. Glover RH, Collins DW, Walsh K, Boonham N. Assessment of loci for DNA barcoding in the genus Thrips (Thysanoptera:Thripidae). Mol Ecol Resour. 2010;10(1):51–9. Epub 2010/01/01. pmid:21564990.
- 27. Gawande SJ, Anandhan S, Ingle AA, Jacobson A, Asokan R. Heteroplasmy due to coexistence of mtCOI haplotypes from different lineages of the Thrips tabaci cryptic species group. Bull Entomol Res. 2017;107(4):534–42. Epub 2017/02/01. pmid:28137324.
- 28. Sparks AN, Diffie S, Riley DG. Thrips species composition on onions in the Vidalia production region of Georgia. Journal of Entomological Science. 2011;46(1):40–5.
- 29. Olczyk M, Pobożniak M. Thrips (Thysanoptera) associated with onion (Allium cepa L.) and Welsh onion (Allium fistulosum L.). Folia Horticulturae. 2020;32(2):319–35.
- 30. Brunner PC, Fleming C, Frey JE. A molecular identification key for economically important thrips species (Thysanoptera: Thripidae) using direct sequencing and a PCR-RFLP-based approach. Agricultural and Forest Entomology. 2002;4:127–36.
- 31. Thompson JD, Higgins DG, Gibson TJ. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic acids research. 1994;22(22):4673–80. pmid:7984417
- 32. 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. Epub 20160322. pmid:27004904.
- 33. Leigh JW, Bryant D. POPART: full-feature software for haplotype network construction. Methods in Ecology and Evolution. 2015;6(9):1110–6.
- 34. Di Rienzo JA, Casanoves F, Balzarini MG, González L, Tablada M, Robledo YC. InfoStat versión 2011. Grupo InfoStat, FCA, Universidad Nacional de Córdoba, Argentina. URL http://www.infostat.com.ar. 2011 Jun;8:195–9.
- 35. Tamura K, Nei M, Kumar S. Prospects for inferring very large phylogenies by using the neighbor-joining method. Proceedings of the National Academy of Sciences. 2004;101(30):11030–5. pmid:15258291
- 36. Hijmans RJ, Guarino L, Cruz M, Rojas E. Computer tools for spatial analysis of plant genetic resources data: 1. DIVA-GIS. Plant Genetic Resources Newsletter. 2001:15–9.
- 37. Peakall R, Smouse PE. GenAlEx 6.5: genetic analysis in Excel. Population genetic software for teaching and research—an update. Bioinformatics. 2012;28(19):2537–9. Epub 20120720. pmid:22820204.
- 38. Saitou N, Nei M. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Molecular biology and evolution. 1987;4(4):406–25. pmid:3447015
- 39. Crossa J. Methodologies for estimating the sample size required for genetic conservation of outbreeding crops. Theor Appl Genet. 1989;77:153–61. pmid:24232522
- 40. Timm AE, Stiller M, Frey JE. A molecular identification key for economically important thrips species (Thysanoptera: Thripidae) in southern Africa. African Entomology. 2008;16(1):68–75.
- 41. Khan F, Roy MC, Kim Y. Thelytokous reproduction of onion thrips, Thrips tabaci Lindeman 1889, infesting Welsh onion and genetic variation among their subpopulations. Insects. 2022;13(1). Epub 20220111. pmid:35055921.
- 42. Iglesias L, Havey MJ, Nault BA. Management of onion thrips (Thrips tabaci) in organic onion production using multiple IPM tactics. Insects. 2021;12(3). Epub 20210301. pmid:33804399.
- 43. Kadirvel P, Srinivasan R, Hsu YC, Su FC, De La Pena R. Application of cytochrome oxidase I sequences for phylogenetic analysis and identification of thrips species occurring on vegetable crops. J Econ Entomol. 2013;106(1):408–18. pmid:23448058.
- 44. Srinivasan R, Guo F, Riley D, Diffie S, Gitaitis R, Sparks A, et al. Assessment of variation among Thrips tabaci populations from Georgia and Peru based on polymorphisms in mitochondrial cytochrome oxidase I and ribosomal ITS2 Sequences. Journal of Entomological Science. 2011;46(3):191–203.
- 45.
Capinera JL. Order Thysanoptera—Thrips. In: Capinera JL, editor. Handbook of vegetable pests. 2nd ed: Academic press; 2020. p. 581–600.
- 46.
Thrips palmi (melon thrips) [Internet]. CABI International CABI. 2022 [cited 6th July 2023].
- 47.
Scirtothrips dorsalis (chilli thrips) [Internet]. CABI International. 2022 [cited 6th July 2023].
- 48. Przybylska A, Fiedler Ż, Obrępalska-Stęplowska A. PCR-RFLP method to distinguish Frankliniella occidentalis, Frankliniella intonsa, Frankliniella pallida and Frankliniella tenuicornis. Journal of Plant Protection Research. 2016;56(1):60–6.
- 49. Dickey AM, Kumar V, Hoddle MS, Funderburk JE, Morgan JK, Jara-Cavieres A, et al. The Scirtothrips dorsalis species complex: endemism and invasion in a global pest. PLoS One. 2015;10(4):e0123747. Epub 20150420. pmid:25893251.
- 50. Vierbergen G. Thrips palmi: pathways and possibilities for spread. EPPO Bulletin. 2001;31:169–71.
- 51. Alves-Silva E, Del-Claro K. Thrips in the Neotropics: What do we know so far? Trends in Entomology. 2010;6:77–88.
- 52. Seal DR, Klassen W, Kumar V. Biological parameters of Scirtothrips dorsalis (Thysanoptera: Thripidae) on selected hosts. Environ Entomol. 2010;39(5):1389–98. pmid:22546433.
- 53. Mound LA, Palmer JM. Identification, distribution and host-plants of the pest species of Scirtothrips (Thysanoptera: Thripidae). Bulletin of Entomological Research. 2009;71(3):467–79.
- 54.
Ananthakrishnan T. Bio-ecology of thrips. Oak Park: Indira Publishing House; 1984. 233 p.
- 55.
Irwin ME. Agroecological niches and thrips (Thysanoptera: Thripidae) dynamics. In: Parker BL, Skinner M, Lewis T, editors. Towards Understanding Thysanoptera Gen. Tech. Rep. NE-147. Radnor, PA: US Department of Agriculture, Forest Service, Northeastern Forest Experiment Station; 1991. p. 133–47.
- 56. Wang JL W J, Zheng CY. Comparative study on some biological characteristics of Frankiniella occidentalis and Thrips tabaci. Chin J Appl Entomol. 2011;48:513–7.
- 57. Ananthakrishnan T. Bionomics of thrips. Annual Review of Entomology. 1993;38(1):71–92.
- 58. Mastrantonio V, Urbanelli S, Porretta D. Ancient hybridization and mtDNA introgression behind current paternal leakage and heteroplasmy in hybrid zones. Sci Rep. 2019;9(1):19177. Epub 20191216. pmid:31844110.
- 59. White DJ, Wolff JN, Pierson M, Gemmell NJ. Revealing the hidden complexities of mtDNA inheritance. Mol Ecol. 2008;17(23):4925–42. pmid:19120984.
- 60. Nault BA, Kain WC, Wang P. Seasonal changes in Thrips tabaci population structure in two cultivated hosts. PLoS One. 2014;9(7):e101791. Epub 20140703. pmid:24992484.
- 61. Cao LJ, Wang ZH, Gong YJ, Zhu L, Hoffmann AA, Wei SJ. Low genetic diversity but strong population structure reflects multiple introductions of western flower thrips (Thysanoptera: Thripidae) into China followed by human-mediated spread. Evol Appl. 2017;10(4):391–401. Epub 2017/03/30. pmid:28352298.
- 62. Firdaus S, Vosman B, Hidayati N, Jaya Supena ED, Visser RG, van Heusden AW. The Bemisia tabaci species complex: additions from different parts of the world. Insect Sci. 2013;20(6):723–33. Epub 20130124. pmid:23955997.