Introduction
The green peach aphid Myzus persicae (Sulzer) is a highly polyphagous insect that can utilize a variety of weeds and crops belonging to more than 40 plant families (Blackman and Eastop, Reference Blackman and Eastop2000, Reference Blackman, Eastop, van Emden and Harrington2017). The ability of M. persicae to use several host plants with different chemical defences stems from genetically inherited mechanisms that facilitate the response to plant allelochemicals, which often results in interclonal variation in fitness (e.g., reproduction rates, longevity and feeding behaviour) between host plant species (Vorburger et al., Reference Vorburger, Lancaster and Sunnucks2003; Turcotte et al., Reference Turcotte, Reznick and Hare2011; Cabrera-Brandt et al., Reference Cabrera-Brandt, Verdugo, Ramírez, Lacroze, Sauge and Figueroa2015). Myzus persicae nicotianae is a specialized subspecies adapted to tobacco (Nicotiana tabacum L.) (Margaritopoulos et al., Reference Margaritopoulos, Tsitsipis, Zintzaras and Blackman2000; Blackman and Eastop, Reference Blackman, Eastop, van Emden and Harrington2017). Unlike Myzus persicae persicae, this tobacco-adapted subspecies can overcome the physical and chemical defences of tobacco plants via morphological, genetic, transcriptomic and metabolic traits (Blackman et al., Reference Blackman, Malarky, Margaritopoulos and Tsitsipis2007; Margaritopoulos et al., Reference Margaritopoulos, Skouras, Nikolaidou, Manolikaki, Maritsa, Tsamandani, Kanavaki, Bacandritsos, Zarpas and Tsitsipis2007b; Cabrera-Brandt et al., Reference Cabrera-Brandt, Fuentes-Contreras and Figueroa2010; Vučetić et al., Reference Vučetić, Petrović-Obradović and Stanisavljević2010; Bass et al., Reference Bass, Zimmer, Riveron, Wilding, Wondji, Kaussmann, Field, Williamson and Nauen2013; Peng et al., Reference Peng, Pan, Gao, Xi, Zhang, Ma, Wu, Zhang and Shang2016; Pan et al., Reference Pan, Peng, Xu, Zeng, Tian, Song and Shang2019a, Reference Pan, Xu, Zeng, Liu and Shang2019b; Singh et al., Reference Singh, Troczka, Duarte, Balabanidou, Trissi, Paladino, Nguyen, Zimmer, Papapostolou, Randall, Luke, Marec, Mazzoni, Williamson, Hayward, Nauen, Vontas and Bass2020).
Various groups of insecticides have been used to control M. persicae in several crops (Foster et al., Reference Foster, Devine, Devonshire, van Emden and Harrington2017), which has resulted in strong selection for insecticide resistance (Bass et al., Reference Bass, Puinean, Zimmer, Denholm, Field, Foster, Gutbrod, Nauen, Slater and Williamson2014). The tobacco aphid M. persicae nicotianae has developed several insecticide resistance mechanisms, including (i) elevated carboxylesterase levels, which confer resistance to organophosphates, carbamates and pyrethroids (Margaritopoulos et al., Reference Margaritopoulos, Skouras, Nikolaidou, Manolikaki, Maritsa, Tsamandani, Kanavaki, Bacandritsos, Zarpas and Tsitsipis2007b; Kati et al., Reference Kati, Mandrioli, Skouras, Malloch, Voudouris, Venturelli, Manicardi, Tsitsipis, Fenton and Margaritopoulos2014); (ii) modification of acetylcholinesterase (MACE), which confers resistance to dimethyl carbamates (Margaritopoulos et al., Reference Margaritopoulos, Skouras, Nikolaidou, Manolikaki, Maritsa, Tsamandani, Kanavaki, Bacandritsos, Zarpas and Tsitsipis2007b); (iii) kdr (L1014F) and super-kdr (M918T) mutations in the voltage-gated sodium channel, which confers resistance to pyrethroids and dichlorodiphenyltrichlorethane (DDT) (Margaritopoulos et al., Reference Margaritopoulos, Skouras, Nikolaidou, Manolikaki, Maritsa, Tsamandani, Kanavaki, Bacandritsos, Zarpas and Tsitsipis2007b); (iv) a mutation in the nicotinic acetylcholine receptor β subunit (nAChRβ) associated with resistance to neonicotinoid insecticides (Bass et al., Reference Bass, Puinean, Andrews, Culter, Daniels, Elias, Laura Paul, Crossthwaite, Denholm, Field, Foster, Lind, Williamson and Slater2011; Voudouris et al., Reference Voudouris, Kati, Sadikoglou, Williamson, Skouras, Dimotsiou, Georgiou, Fenton, Skavdis and Margaritopoulos2016) and (v) a metabolic mechanism based on a cytochrome P450 monooxygenase, which confers resistance to neonicotinoids (Puinean et al., Reference Puinean, Foster, Oliphant, Denholm, Field, Millar, Williamson and Bass2010; Voudouris et al., Reference Voudouris, Kati, Sadikoglou, Williamson, Skouras, Dimotsiou, Georgiou, Fenton, Skavdis and Margaritopoulos2016). Furthermore, transcriptomic mechanisms involved in the aphid detoxification response that suppress the metabolic effects of the insecticide or nicotine have also been described in M. persicae nicotianae (Cabrera-Brandt et al., Reference Cabrera-Brandt, Silva, Le Trionnaire, Tagu and Figueroa2014; Peng et al., Reference Peng, Pan, Gao, Xi, Zhang, Ma, Wu, Zhang and Shang2016; Pan et al., Reference Pan, Peng, Xu, Zeng, Tian, Song and Shang2019a, Reference Pan, Xu, Zeng, Liu and Shang2019b; Singh et al., Reference Singh, Troczka, Duarte, Balabanidou, Trissi, Paladino, Nguyen, Zimmer, Papapostolou, Randall, Luke, Marec, Mazzoni, Williamson, Hayward, Nauen, Vontas and Bass2020).
In Chile, populations of M. persicae nicotianae were previously reported to be composed of a single predominantly red genotype on tobacco crops (Fuentes-Contreras et al., Reference Fuentes-Contreras, Figueroa, Reyes, Briones and Niemeyer2004), a genotype that was later found to also be widespread in tobacco fields in the USA, Brazil and Argentina (Zepeda-Paulo et al., Reference Zepeda-Paulo, Simon, Ramírez, Fuentes-Contreras, Margaritopoulos, Wilson, Sorenson, Briones, Azevedo, Ohashi, Lacroix, Glais and Figueroa2010). Such ecologically successful aphid genotypes are considered ‘superclones’ (Vorburger et al., Reference Vorburger, Lancaster and Sunnucks2003), which are frequently obligate parthenogenetic asexual lineages (Figueroa et al., Reference Figueroa, Fuentes-Contreras, Molina-Montenegro and Ramírez2018). This red tobacco aphid ‘superclone’ exhibited moderate levels of carboxylesterase activity (R1) with no kdr (L1014F), super-kdr (M918T) or MACE mutations (Fuentes-Contreras et al., Reference Fuentes-Contreras, Figueroa, Reyes, Briones and Niemeyer2004; Cabrera-Brandt et al., Reference Cabrera-Brandt, Silva, Le Trionnaire, Tagu and Figueroa2014). However, in the last decade, aphid colonies in tobacco fields in Chile have begun to display colour variations (e.g., green, yellow and red). This colour polymorphism has been observed throughout the geographic range of tobacco plantations in Chile, and the colour of aphids varies between plants, as each plant hosts a single colony (unpublished data). Similar patterns have been previously described in Greece (Margaritopoulos et al., Reference Margaritopoulos, Tsitsipis, Zintzaras and Blackman2000, Reference Margaritopoulos, Tsitsipis, Goundoudaki and Blackman2002; Zitoudi et al., Reference Zitoudi, Margaritopoulos, Mamuris and Tsitsipis2001; Poupoulidou et al., Reference Poupoulidou, Margaritopoulos, Kephalogianni, Zarpas and Tsitsipis2006; Blackman et al., Reference Blackman, Malarky, Margaritopoulos and Tsitsipis2007), Italy (Margaritopoulos et al., Reference Margaritopoulos, Blackman, Tsitsipis and Sannino2003), Japan (Shigehara and Takada, Reference Shigehara and Takada2003, Reference Shigehara and Takada2004; Margaritopoulos et al., Reference Margaritopoulos, Malarky, Tsitsipis and Blackman2007a, Reference Margaritopoulos, Skouras, Nikolaidou, Manolikaki, Maritsa, Tsamandani, Kanavaki, Bacandritsos, Zarpas and Tsitsipis2007b) and the USA (Harlow and Lampert, Reference Harlow and Lampert1990; Clements et al., Reference Clements, Sorenson, Wiegmann, Neese and Roe2000a, Reference Clements, Wiegmann, Sorenson, Smith, Neese and Roe2000b; Srigiriraju et al., Reference Srigiriraju, Semtner, Anderson and Bloomquist2009), where higher genotype diversity of tobacco aphids has been observed (Zepeda-Paulo et al., Reference Zepeda-Paulo, Simon, Ramírez, Fuentes-Contreras, Margaritopoulos, Wilson, Sorenson, Briones, Azevedo, Ohashi, Lacroix, Glais and Figueroa2010). The colour polymorphisms can be explained by the presence of new M. persicae nicotianae genotypes resulting from sexual reproduction events, as reported for M. persicae in Chile on peaches (Rubiano-Rodríguez et al., Reference Rubiano-Rodríguez, Fuentes-Contreras, Figueroa, Margaritopoulos, Briones and Ramírez2014, Reference Rubiano-Rodríguez, Fuentes-Contreras and Ramírez2019; Rubio-Meléndez et al., Reference Rubio-Meléndez, Sepúlveda and Ramírez2018) or by new introduction events from neighbouring countries. These putative new genotypes could possess insecticide resistance mutations that affect tobacco aphid control in tobacco fields in Chile. The aim of this study was to evaluate whether phenotypic colour variation is related to the presence of new M. persicae nicotianae genotypes in Chile. This study provides an updated assessment of the insecticide resistance status of Chilean tobacco aphid populations by evaluating the presence previously described insecticide resistance mutations in M. persicae nicotianae.
Materials and methods
Sampling programme
A total of 32 tobacco fields were sampled from February–March in 2015 and 2016 along a 400-km latitudinal gradient that spans the entire range of the tobacco-growing area in Chile. Tobacco fields (table 1) were divided into three areas of tobacco production: the North comprising five fields with 28 samples (2015) from the northern Valparaíso Region; the Centre comprising 12 fields (2015) and four fields (2016) totalling 121 samples from the southern O'Higgins Region to the centre of the Maule Region and the South comprising six fields (2015) and five fields (2016) totalling 82 samples from the eastern Ñuble Region. A total of 231 aphid samples were obtained (table 1). Aphids were collected from the two most commonly grown tobacco types in Chile (Burley and Virginia). The minimum distance between points from which aphids were collected was 30 m. In each field, each aphid sample was collected from a single colony per tobacco plant, gently lifted from the plant with a paintbrush, placed in 1.7-ml tubes filled with absolute ethanol and stored at 4°C for later analysis.
Details of M. persicae nicotianae sampling from tobacco fields in central Chile
The sampled tobacco fields with their coordinates, number of samples, year of collection and areas of tobacco production: north (N), centre (C) and south (S). MPNsrg: single red genotype of M. persicae nicotianae.
Microsatellite genotyping
All samples, including three samples corresponding to the single red genotype of M. persicae nicotianae (MPNsrg, table 1) previously described by Fuentes-Contreras et al. (Reference Fuentes-Contreras, Figueroa, Reyes, Briones and Niemeyer2004) and stored in our laboratory, were genotyped using six microsatellite loci (Myz2, Myz9, Myz25, M35, M37 and M40), which have been described and used extensively for M. persicae (Fuentes-Contreras et al., Reference Fuentes-Contreras, Figueroa, Reyes, Briones and Niemeyer2004; Margaritopoulos et al., Reference Margaritopoulos, Malarky, Tsitsipis and Blackman2007a, Reference Margaritopoulos, Skouras, Nikolaidou, Manolikaki, Maritsa, Tsamandani, Kanavaki, Bacandritsos, Zarpas and Tsitsipis2007b; Cabrera-Brandt et al., Reference Cabrera-Brandt, Fuentes-Contreras and Figueroa2010, Reference Cabrera-Brandt, Silva, Le Trionnaire, Tagu and Figueroa2014). The DNA quality and quantity were assessed using a Nanodrop (Nanodrop Technologies, Wilmington, DE, USA) spectrophotometer. Polymerase chain reactions (PCRs) were performed using the M13 universal primer (−21) labelled with fluorescent FAM or VIC at the 5′ end of the forward primer as described previously (Schuelke, Reference Schuelke2000). Each amplification was conducted in a 15-μl reaction volume containing 1× Mg2+-free reaction buffer, 25 mM MgCl2, 10 μM dNTPs, 1 μM each of the forward and reverse primers, 1 μM of the M13 primer, 0.5 U Platinum® Taq DNA polymerase (Invitrogen, Carlsbad, CA, USA) and 20 ng μl−1 of total DNA in sterile nanopure water. PCRs were carried out with the following thermal cycling programme: initial denaturation at 94°C for 5 min, followed by four touchdown cycles of 30 s of denaturation at 94°C, 30 s of annealing (62, 61, 59 and 57°C each cycle) and 45 s of elongation at 72°C. Next, 26 cycles of 30 s of denaturation at 94°C, 30 s of annealing at 55°C and 45 s of elongation at 72°C were performed. Finally, eight cycles using an annealing temperature of 53°C, followed by 10 min of elongation at 72°C completed the amplification. Positive DNA amplifications were checked following electrophoresis in a 2.0% agarose gel. Automated fragment analysis by sequencing was conducted by Macrogen Inc. (Seoul, South Korea). The allele size and configuration for each individual were obtained using GeneMarker® software (SoftGenetics, State College, PA, USA).
Identification of insecticide resistance mutations
DNA isolation was conducted to characterize the insecticide resistance mutations in each sample. This analysis was performed using the TaqMan assay in a STRATAGENE MX 3000 (Agilent Technologies, Santa Clara, CA) thermocycler. The kdr (L1014F) and super-kdr (M918T) mutations were identified according to Anstead et al. (Reference Anstead, Williamson, Eleftherianos and Denholm2004), while MACE (S431F) was identified according to Anstead et al. (Reference Anstead, Williamson and Denholm2008). Finally, the arginine to threonine (R81T) mutation in the nAChRβ was detected according to Voudouris et al. (Reference Voudouris, Williamson, Skouras, Kati, Sahinoglou and Margaritopoulos2017). The detection output was interpreted using the S (sensitive) allele and R (insensitive) allele nomenclature, according to Rubio-Meléndez et al. (Reference Rubio-Meléndez, Sepúlveda and Ramírez2018).
Data analysis
The clonal heterogeneity indices (adapted from Simpson's index and Hill's Simpson reciprocal index) and clonal evenness index (Simpson's evenness index) were calculated using GENCLONE 2.0 software (Arnaud-Haond and Belkhir, Reference Arnaud-Haond and Belkhir2007). The observed heterozygosity (H o), expected heterozygosity (H e) and inbreeding coefficient (FIS) (log likelihood ratio statistic with dememorization number = 10,000, iterations per batch = 10,000 and batches = 100) were calculated according to Brookfield (1996) using the GENEPOP package version 1.2 (Raymond and Rousset, Reference Raymond and Rousset1995; see http://genepop.curtin.edu.au/). The inbreeding coefficient over all loci was calculated in the same software (log likelihood ratio statistic with dememorization number = 10,000, iterations per batch = 10,000 and batches = 100).
The genetic distance matrix between genotypes was evaluated using POPULATIONS 1.2.32 software, and the resulting tree was built with FIGTREE 1.4.3 software.
The proportion of genetic differentiation among areas due to molecular differences was assessed for the three areas (considering all copies and one single copy per multilocus genotype (MLG)). A hierarchical analysis of molecular variance (AMOVA) was conducted, including variation among and within areas and samples from the same source. Hence, F-statistics were calculated according to AMOVA on GenAlEx v6.4 (Peakall and Smouse, Reference Peakall and Smouse2006) to assess the genetic differentiation for microsatellite loci (which assumes a stepwise mutation model) and with the Codom-Microsat genetic distance using 999 permutations (Excoffier et al., Reference Excoffier, Smouse and Quattro1992; Peakall et al., Reference Peakall, Smouse and Huff1995).
A Bayesian clustering analysis conducted with single copies of each genotype was performed in STRUCTURE version 2 software 3 (Pritchard et al., Reference Pritchard, Stephens and Donnelly2000) using admixture ancestry and correlated allele frequency models. The number of clusters (K) was varied from 1 to 10, and the analysis was repeated five times. Each repetition consisted of a burn-in period of 60,000 iterations and 600,000 Markov chain Monte Carlo iterations. The online program STRUCTURE HARVESTER (Earl and von Holdt, Reference Earl and von Holdt2012) was used to calculate the most likely number of genetic clusters (K) using the Evanno method (Evanno et al., Reference Evanno, Regnaut and Goudet2005).
Results
Genetic diversity
Genotyping revealed the presence of 16 MLGs in the entire sample of M. persicae nicotianae from Chilean tobacco fields, which were named MPN1 to MPN16 (table 2). The single red genotype previously reported in Chile was not detected (table 2). The most frequent genotypes were MPN15, MPN5 and MPN12 (fig. 1). The population genetic parameters of M. persicae nicotianae are shown in table 3; there are no signs of heterozygote deficiency (F is negative), and heterozygosity was near expectation.
Overall frequencies of each of the 16 genotypes of M. persicae nicotianae detected in Chile.
Allele size of six loci evaluated for aphid body colour and insecticide resistance mutations for all new genotypes and the single red genotype (MPNsrg) of M. persicae nicotianae in Chile
Population genetic parameters of M. persicae nicotianae
Areas of tobacco production, north (N), centre (C) and south (S); H o, observed heterozygosity; H e: expected heterozygosity; FIS: inbreeding coefficient; H″: Shannon index; D*: adapted Simpson's index; ED*: Simpson's evenness index; MLGs: number of multilocus genotypes.
The Goldstein genetic distance was not closely related among genotypes or between the newly found genotypes and the single red genotype previously described (fig. 2).
Goldstein genetic distance between 17 M. persicae nicotianae genotypes in Chile, including the single red genotype (MPNsrg).
Differentiation among populations
The hierarchical AMOVA revealed no genetic differentiation among areas for analyses performed with all copies (FST = −0.006, P = 1.000, table 4) or with a single copy per MLG (FST = −0.021, P = 0.609, table 4). Low but significant genetic differentiation was observed within samples of M. persicae nicotianae (4.0 and 1.0% for all and single-copy MLGs, respectively). Finally, a large and significant percentage of variation (96 and 99% for all and single-copy MLGs, respectively) was observed among samples of the three areas studied (table 4).
Molecular variance AMOVA for M. persicae nicotianae aphids from samples collected in three areas of tobacco production in Chile: north (N), centre (C) and south (S)
Similarly, the Bayesian analysis conducted using a single copy per MLG revealed no genetic differentiation among areas. Analysis of the population genetic structure considering a single copy per MLG from all areas revealed three genetic clusters (K = 3) according to the Evanno method (modal value of ΔK (fig. 3a)). In the northern area of tobacco fields, the three clusters were represented in the same proportions (33.3%), while in the central area, cluster 1 (red) represented 37% of the population, cluster 2 (green) 28% and cluster 3 (blue) 35%. In the southern area, the proportions for clusters 1, 2 and 3 were 11, 18 and 71%, respectively (fig. 3b).
Bayesian assignment analysis of individuals to clusters. (a) ΔK for different numbers of clusters was K = 3. (b) Assessment of population genetic structure by Bayesian cluster analysis. The MLGs were grouped according to the areas of tobacco production (north (N), centre (C) and south (S)), and each of the MLGs was represented by vertical bars divided by colours according to their coefficients of ancestry to each cluster.
Insecticide resistance mutations
Mutations conferring resistance to insecticides were found only in MLGs (MPN7, MPN8, MPN9 and MPN13) and corresponded to the MACE mutation in the heterozygote state (RS). No MLGs were found to carry the kdr, super-kdr or nAChRβ mutations and were characterized as susceptible homozygotes (SS) (table 2).
Discussion
In 2004, the presence of the tobacco aphid M. persicae nicotianae was reported in Chile for the first time. After a population genetic survey and evaluation of insecticide resistance mechanisms, a single red genotype was found to dominate the entire population and was determined to be slightly resistant to organophosphate and carbamate and susceptible to pyrethroids (R1 level of esterase activity and kdr susceptible), despite the intense use of insecticides for its control (Fuentes-Contreras et al., Reference Fuentes-Contreras, Figueroa, Reyes, Briones and Niemeyer2004, Reference Fuentes-Contreras, Basoalto, Sandoval, Pavez, Leal, Burgos and Muñoz2007).
The predominance of this single red genotype was maintained for many years, and the mechanisms involved in its ecological success have been widely studied. The biological features of this genotype have been characterized through the study of its behaviour on tobacco (Troncoso et al., Reference Troncoso, Vargas, Tapia, Olivares-Donoso and Niemeyer2005; Vargas et al., Reference Vargas, Troncoso, Tapia, Olivares-Donoso and Niemeyer2005; Tapia et al., Reference Tapia, Silva, Ballesteros, Figueroa, Niemeyer and Ramírez2015), performance on different host plants (Olivares-Donoso et al., Reference Olivares-Donoso, Troncoso, Tapia, Aguilera-Olivares and Niemeyer2007; Tapia et al., Reference Tapia, Troncoso, Vargas, Olivares-Donoso and Niemeyer2008), route of introduction to Chile (Zepeda-Paulo et al., Reference Zepeda-Paulo, Simon, Ramírez, Fuentes-Contreras, Margaritopoulos, Wilson, Sorenson, Briones, Azevedo, Ohashi, Lacroix, Glais and Figueroa2010), the effects of individual insecticides (Fuentes-Contreras et al., Reference Fuentes-Contreras, Basoalto, Sandoval, Pavez, Leal, Burgos and Muñoz2007), transcriptomic responses for moderating the effects of insecticides (Cabrera-Brandt et al., Reference Cabrera-Brandt, Silva, Le Trionnaire, Tagu and Figueroa2014) and the properties of its allelochemical detoxification metabolism that have allowed it to overcome tobacco's chemical defences (Cabrera-Brandt et al., Reference Cabrera-Brandt, Fuentes-Contreras and Figueroa2010).
Nevertheless, the Chilean population of M. persicae nicotianae has experienced drastic changes. Despite identifying 16 new MLGs of M. persicae nicotianae, there were no signs of the single red superclone in any of the sampled Chilean tobacco fields. These new genotypes have phenotypic and genetic differences from the founder genotype, which include colour variation (green, yellow and red) and the presence of MACE mutations in four of the new genotypes, a mutation that confers resistance to dimethyl carbamates. Although pyrethroids and neonicotinoids have been intensively sprayed on tobacco fields to control the tobacco aphid during the last decade in Chile, none of the new genotypes have kdr (L1014F), super-kdr (M918T) or nAChRβ mutations (Fuentes-Contreras et al., Reference Fuentes-Contreras, Basoalto, Sandoval, Pavez, Leal, Burgos and Muñoz2007). Similar changes in the frequencies of genotypes and insecticide resistance mutations have been described for M. persicae nicotianae in tobacco fields in Greece (Margaritopoulos et al., Reference Margaritopoulos, Malarky, Tsitsipis and Blackman2007a, Reference Margaritopoulos, Skouras, Nikolaidou, Manolikaki, Maritsa, Tsamandani, Kanavaki, Bacandritsos, Zarpas and Tsitsipis2007b, Reference Margaritopoulos, Kati, Voudouris, Skouras and Tsitsipis2021; Kati et al., Reference Kati, Mandrioli, Skouras, Malloch, Voudouris, Venturelli, Manicardi, Tsitsipis, Fenton and Margaritopoulos2014).
The higher genetic diversity and detection of MACE mutations raise questions about the origin of these new genotypes. Multiple introduction events and sexual reproduction may account for an increase in the genetic diversity of aphid populations following demographic bottlenecks (Nibouche, et al., Reference Nibouche, Fartek, Mississipi, Delatte, Reynaud and Costet2014; Bebber, Reference Bebber2015; Figueroa et al., Reference Figueroa, Fuentes-Contreras, Molina-Montenegro and Ramírez2018). Introductions from different countries are likely to occur, as was shown for the red genotype by Zepeda-Paulo et al. (Reference Zepeda-Paulo, Simon, Ramírez, Fuentes-Contreras, Margaritopoulos, Wilson, Sorenson, Briones, Azevedo, Ohashi, Lacroix, Glais and Figueroa2010). Furthermore, sexual reproduction or hybridization events between M. persicae nicotianae and M. persicae persicae have been reported in Greece (Blackman et al., Reference Blackman, Malarky, Margaritopoulos and Tsitsipis2007; Margaritopoulos et al., Reference Margaritopoulos, Malarky, Tsitsipis and Blackman2007a) in areas where tobacco is cultivated near peach (Prunus persica) orchards. Indeed, M. persicae persicae reproduces sexually on peach in Chile (Rubiano-Rodríguez et al., Reference Rubiano-Rodríguez, Fuentes-Contreras, Figueroa, Margaritopoulos, Briones and Ramírez2014; Rubio-Meléndez et al., Reference Rubio-Meléndez, Sepúlveda and Ramírez2018), which makes hybridization with M. persicae nicotianae a likely occurrence. This idea is supported by the lack of genetic differentiation between aphids collected from different tobacco-growing areas in Chile, which form a single large population that might be connected by gene flow from aphid populations on peach. This is consistent with the finding that most of the alleles present in the new M. persicae nicotianae genotypes are present in M. persicae persicae collected from secondary and primary hosts. Second, MACE is present in M. persicae persicae on peach and other crops, where the dimethyl carbamate insecticide pirimicarb is used, but this chemical has not been applied to tobacco. However, kdr has been found to be present at high frequencies in aphids on peach and other crops but was previously noted to be absent from tobacco populations (Cabrera-Brandt et al., Reference Cabrera-Brandt, Silva, Le Trionnaire, Tagu and Figueroa2014; Rubio-Meléndez et al., Reference Rubio-Meléndez, Sepúlveda and Ramírez2018). Furthermore, the presence of other super-kdr mutations detected in M. persicae persicae, such as the super-kdr M918L mutation, is not associated with the presence of the kdr L1014F mutation (Fontaine et al., Reference Fontaine, Caddoux, Brazier, Bertho, Bertolla, Micoud and Roy2011; Panini et al., Reference Panini, Dradi, Marani, Butturini and Mazzoni2014: Mingeot et al., Reference Mingeot, Hautier and Jansen2021). These super-kdr M918L mutations have not yet been described in M. persicae nicotianae, even in Greece, where recent studies have established their absence (Margaritopoulos et al., Reference Margaritopoulos, Kati, Voudouris, Skouras and Tsitsipis2021).
One question raised by our findings relates to why the single red genotype, which was previously found to be widespread in Chile, apparently disappeared. This is a difficult question to answer given that the single red genotype was not detected in any of the tobacco fields sampled throughout Chile's tobacco-growing regions over the two consecutive seasons of this study. The last preserved samples of the single red genotype are from laboratory colonies used in Cabrera-Brandt et al. (Reference Cabrera-Brandt, Silva, Le Trionnaire, Tagu and Figueroa2014) and Tapia et al. (Reference Tapia, Silva, Ballesteros, Figueroa, Niemeyer and Ramírez2015). Furthermore, the genetic distance of the single red genotype is high with respect to the 16 new genotypes; thus, it is unlikely that the new genotypes evolved from the single red genotype because of the number of mutational steps that would have been required. Because the intrinsic rate of growth (r m) of the single red genotype is lower on tobacco than on sweet pepper, which is considered its optimal host (Nikolakakis et al., Reference Nikolakakis, Margaritopoulos and Tsitsipis2003; Tapia et al., Reference Tapia, Troncoso, Vargas, Olivares-Donoso and Niemeyer2008; Cabrera-Brandt et al., Reference Cabrera-Brandt, Fuentes-Contreras and Figueroa2010), the single red genotype might have been competitively displaced towards other secondary hosts by these new genotypes. Furthermore, global climate change has affected environmental conditions in Chile over the past decade; specifically, there has been a sustained increase in the number of hot days (over 30°C) per year (Dirección Meteorológica de Chile; www.meteochile.gob.cl). Such changes in climatic variables are known to affect aphid population dynamics in many crops (Bell et al., Reference Bell, Alderson, Izera, Kruger, Parker, Pickup, Shortall, Taylor, Verrier and Harrington2015).
Zepeda-Paulo et al. (Reference Zepeda-Paulo, Simon, Ramírez, Fuentes-Contreras, Margaritopoulos, Wilson, Sorenson, Briones, Azevedo, Ohashi, Lacroix, Glais and Figueroa2010) reported low-genetic diversity in the populations of M. persicae nicotianae in Argentina and Brazil; in Chile, the diversity was zero because of the presence of a single M. persicae nicotianae genotype in tobacco crops. Therefore, the increase in the genetic diversity of M. persicae nicotianae observed in recent times on Chilean tobacco crops may have also occurred in these neighbouring countries. Additional studies are needed to characterize insecticide resistance mechanisms and ecological traits, such as the performance and behaviour of these new genotypes on different host plants, and compare them with other M. persicae nicotianae populations to evaluate the generalizability of the patterns observed in Chile.
Acknowledgements
M.A.C.-B. was supported by a postdoctoral fellowship granted by the Millennium Nucleus Centre in Molecular Ecology and Evolutionary Applications in Agroecosystems. This study was partially funded by Iniciativa Científica Milenio NC 120027 granted to C.C.F. and E.F.-C., FONDECYT Postdoctoral 3190544 granted to M.E.R.-M. and FONDECYT 11160713-2016 granted to M.A.C.-B.
