INTRODUCTION
Tomato (Solanum lycopersicum L.) is widely cultivated in the tropical and subtropical regions of the world. Brazil is the eighth largest tomato producer worldwide, and the occurrence of phytosanitary problems is a limiting factor for its production (Toloy et al., 2018). Among the most frequent pests, the whitefly (Bemisia tabaci biotype B and B. argentifolii Bellows & Perring [Hemiptera: Aleyrodidae]) is considered as an important insect pest; it causes direct damage due to loss of the photosynthetically active area, and leads to yield losses (Desneux et al., 2011). It also causes indirect damage because it is a Geminivirus virus vector, which is responsible for irreversible physiological disorders (Inoue‑Nagata et al., 2016).
Pesticides are mostly used for pest control in tomato. Insecticides are systematically applied by spraying, generally two or three applications a week in the warmer seasons. Several problems can occur such as residue accumulation in fruits, worker poisoning, environmental pollution, and increased production costs (Silva et al., 2009; 2013).
The development of pest‑ and arthropod‑resistant genotypes through crop breeding programs is a viable alternative to solve problems arising from the indiscriminate use of agrochemicals. Programs developed in Brazil have pursued the introgression strategy of insect‑resistant alleles, which are present in wild tomato species and elite inbred lines (Silva et al., 2013; Andrade et al., 2017; Vosmam et al., 2018). Pest resistance in these wild species is mediated by allelochemicals, which are generally associated with glandular foliar trichomes (Resende et al., 2006; Maluf et al., 2007; Andrade et al., 2018; Silva et al., 2018).
Tomato genotypes with high allelochemical levels were obtained from controlled interspecific hybridization between cultivated tomato and wild species. For example, S. pennellii Correll produce the acylsugar (AS) allelochemical (Resende et al., 2006) and S. habrochaites S. Knapp & D.M. Spooner var. hirsutum produces zingiberene (ZGB), a sesquiterpene hydrocarbon (Maluf et al., 2001; Freitas et al., 2002). These allelochemicals provide resistance to whitefly (B. tabaci biotype B) (Hemiptera: Aleyrodidae) (Resende et al., 2009; Silva et al., 2009; Maluf et al., 2010; Neiva et al., 2013, Andrade et al., 2017), South American tomato pinworm (Tuta absoluta) (Lepidoptera: Gelechiidae) (Goncalves Neto et al., 2010; Oliveira et al., 2012), mites (Tetranychus urticae and T. evansi) (Acari: Tetranychidae) (Maluf et al., 2007), leafminer (Liriomyza trifolii) (Silva et al., 2018), and aphid (Myzus persicae) (Silva et al., 2013).
The Mi gene produces resistance to nematodes Meloidogyne spp. in tomato (Smith, 1994). Furthermore, the effectiveness of this gene to provide resistance to other pests has also been reported. Kaloshian et al. (1995) and Rossi et al. (1998) reported tomato resistance to aphid Macrosiphum euphorbiae (Thomas) (Hemiptera: Aphididae) associated with the presence of the Mi gene. Nombela et al. (2003) and Rodríguez‑Álvarez et al. (2017) found that the Mi-1.2 gene, or another gene linked to it, gives resistance to B. tabaci in tomato genotypes (S. lycopersicum). Resistance to B. tabaci in tomato mediated by the Mi gene, or another gene linked to it, was also observed, and the influence of AS content was tested. However, there are no reports comparing resistance levels against whitefly in tomatoes bearing ZGB associated with AS and the Mi gene.
The aim of this study was to quantify the resistance to whitefly (B. tabaci biotype B) of tomato genotypes with different foliar allelochemical contents (AS and/or ZGB) associated or not associated with the Mi gene.
MATERIALS AND METHODS
The experiment was carried out in a greenhouse at the Horticultural Experimental Station‑HortiAgroSementes Ltda., Palmital Farm in the municipality of Ijaci (21º14’16” S; 45º08’00” W; 918 m a.s.l.), Minas Gerais, Brazil and in the Horticulture Sector of the Universidade Federal de Lavras (UFLA) in the municipality of Lavras (21º14’43” S; 45º59’59” W; 918 m a.s.l.), Minas Gerais, Brazil.
Fifteen tomato genotypes belonging to the tomato breeding program of the UFLA, which exhibited different AS, ZGB, and Mi gene levels were evaluated (Table 1). The Santa Clara, TOM‑695, TOM‑556, and TOM‑584 inbred lines were used as susceptible controls, whereas the S. habrochaites var. hirsutum PI‑127826 accession was used as the resistant control. All 15 tomato genotypes were previously selected. Genotypes with high AS and ZGB foliar contents were selected based on their concentration in the leaflets by the colorimetric method proposed by Freitas et al. (2002) and Resende et al. (2006). Genotypes bearing the Mi allele were selected in nematode resistance assays.
Seeds of all genotypes were sown in polystyrene trays with 128 cells and containing commercial substrate (Tropstrato HA Hortaliças, Vida Verde, Mogi Mirim, São Paulo, Brazil). Afterward, plants were transplanted 40 d after sowing in plastic pots (500 mL) containing a mixture of soil, commercial substrate, sand, lime, and NPK.
For B. tabaci biotype B infestation, whitefly rearing was established in the Horticulture Sector of the UFLA in a 12 m2 greenhouse; the chapel type protective structure (4 m × 3 m) consisted of 100 μ thick clear plastic cover and anti‑aphid netting on the sides. Adult insects were collected in ‘Santa Clara’ tomato plants and then transferred to a screened greenhouse. Approximately 50 ‘Santa Clara’ plants (which showed low ZGB and AS contents and susceptibility to whitefly) were used for oviposition and feeding the insects. These plants were also used as an infestation source and kept in the greenhouse during the experiment.
All 15 genotypes were transported 24 d after transplanting to another greenhouse with anti‑aphid netting previously infested with B. tabaci biotype B; this greenhouse was also located in the Horticulture Sector of the UFLA. The experiment was established in the greenhouse with a completely randomized design (CRD), one plant per pot, and six replicates for each of the 15 treatments.
(a)Lines homozygous for the described characteristics.
Means followed by the same letter in the columns belong to the same group according to the Scott-Knott test (p ≤ 0.05).
Six days after infestation, oviposition was evaluated by counting the number of eggs in a 2 cm2 leaf area on four leaflets of the upper third of each plant with a binocular loupe with 20X to 80X magnification (Oksn 9585, Atibaia, São Paulo, Brazil). Twenty days after the infestation of sampled leaflets, previously marked with white tape, these were evaluated for the number of last instar nymphs with a binocular loupe. Mean temperature and RH between the infestation period and nymph count ranged from 11.2 to 20.3 ºC and 52% to 100%, respectively.
After verifying the normality and homogeneity of variance by the Shapiro‑Wilk and Bartlett tests, respectively, data concerning the number of B. tabaci eggs and nymphs were transformed into (x + 0.5½) before performing ANOVA. Means were grouped by the Scott Knott test (p ≤ 0.05) and contrasts selected from genotype clusters with different allelochemical contents and bearing the Mi gene, and were calculated with the Sisvar statistical software (Ferreira, 2011).
RESULTS AND DISCUSSION
According to the F-test, significant differences were observed between tomato genotypes that were tested for all evaluated traits (Table 1). Inbred lines with high AS or ZGB contents (T5 to T15) or only carrying the Mi gene (T5) showed less preference for both oviposition and reduced number of nymphs compared with the Santa Clara, TOM‑695, TOM‑556, and TOM‑584 genotypes used as susceptible controls.
The Santa Clara, TOM‑695, TOM‑556, and TOM‑584 genotypes were allocated in clusters with the highest oviposition and number of nymphs indicated by the Scott‑Knott test (Table 1). These four genotypes are the most susceptible; they have in common the absence of the Mi-1 gene and lower contents of both allelochemicals. In contrast, inbred lines with high AS and/or ZGB contents (T6 to T15) as well as the inbred line with low allelochemical contents, but carrying the Mi-1 gene (T5), were less preferred for oviposition and reproduction when compared with the most susceptible controls (T1 to T4).
On the other hand, ‘TOM‑687’ and ‘TOM‑688’ that produced high AS contents, were allocated in clusters with significantly lower means than those of the T1 to T4 inbred lines (Tables 1 and 2, Contrast C1). These results prove the effectiveness of high foliar AS concentrations to produce resistance to whitefly, as previously described (Resende et al., 2009; Oliveira et al., 2012; Neiva et al., 2013). For ‘TOM‑687’ and ‘TOM‑688’, oviposition was reduced by 50.76% and 54.97% compared with ‘Santa Clara’, while the number of nymphs was reduced by 40.93% and 40.85%, respectively.
Inbred lines with high ZGB, low AS content, and without the Mi gene (ZGB-703 and ZGB-704) also showed a significant level of antixenosis for oviposition and fewer whitefly nymphs when compared with the susceptible controls (Santa Clara, TOM‑695, TOM‑556, and TOM‑584) (Tables 1 and 2, Contrast C2). The adverse effect of genotypes with high ZGB contents in the biological development of whitefly (B. tabaci biotype B) has been previously reported (Silva et al., 2009; Oliveira et al., 2012; Neiva et al., 2013). The presence of ZGB, although discrete, was significantly more effective than AS in reducing oviposition and number of nymphs (Contrast C10). For ‘ZGB‑703’ and ‘ZGB‑704’, oviposition was reduced by 59.39% and 63.04% as related to ‘Santa Clara’, while the number of nymphs was reduced by 43.24% and 42.26%, respectively. Effects on whitefly survival provided by AS and ZGB, simultaneously present in the TOM-778, TOM-779, and TOM-780 inbred lines, were efficient to induce less preference to oviposition and fewer nymphs when compared with inbred lines with low content of these allelochemicals (Contrast C3); they were also more efficient than genotypes containing only one type of allelochemical such as high AS content (Contrast C11) or high ZGB content (Contrast C13).
**, ns Significant and nonsignificant at the 0.01probability level, respectively, according to the F-test.
The genotype carrying only the Mi gene (TOM‑684) exhibited moderate resistance when compared with the susceptible controls (Table 2, Contrast C4). Genotypes with high AS contents and with the Mi gene (TOM‑759 and TOM‑760) had higher resistance to oviposition and fewer whitefly nymphs than the susceptible lines (Table 2, Contrast C5). There was a reduction of 46.42%, 44.42%, and 26.18% in the number of nymphs per leaflet in lines carrying the Mi gene: TOM‑ 759 (AS + Mi gene), TOM‑760 (AS + Mi gene), and TOM‑684 (Mi gene), respectively. However, this resistance was lower than values reported by Nombela et al. (2003), who observed a reduction of 50% in the mean number of nymphs when compared with plants that did not carry the Mi resistance allele. Such differences may be justified by the relative magnitude influenced by temperature.
Inbred lines with high AS (TOM‑687 and TOM‑688), high ZGB content (ZGB‑703 and ZGB‑704), and high contents of both allelochemicals (TOM‑778, TOM‑779, and TOM‑780) showed higher resistance to oviposition and lower number of whitefly nymphs compared with the genotype TOM-684, which carries the Mi gene (Table 2, Contrasts C6, C7, and C8). Therefore, the resistance level provided by Mi is lower than AS and ZGB separately or by AS and ZGB associated in the same genotype. Inbred lines with a high AS content and carrying the Mi gene (TOM‑759 and TOM‑760) were more resistant to oviposition and exhibited fewer nymphs than the line with only the Mi gene (TOM‑684) (Tables 1 and 2, Contrast C9). On the other hand, ‘TOM‑759’ and ‘TOM‑760’ (high AS + Mi) were also more resistant to oviposition and exhibited fewer nymphs than ‘TOM‑687’ and ‘TOM‑688’, which only have high AS content (Table 2, Contrast C12).
Thus, the combined effect in oviposition due to the high AS content plus the presence of the Mi gene (high AS + Mi) is similar to the effect provided by high ZGB content. The effect of high AS + Mi as related to the number of nymphs is higher than the effect by ZGB (Table 2, Contrast 14).
The results obtained in the present study were similar to those reported by Nombela et al. (2003) and confirmed that the Mi-1 gene is involved in the partial resistance to B. tabaci biotype B in S. lycopersicum. Some studies also reported that Mi gene is efficient in controlling M. euphorbiae (Kaloshian et al., 1997; 2000; Goggin et al., 2001) and nematodes (Meloidogyne spp.) in tomato (Goggin et al., 2001; Mantelin et al., 2011; Atamian et al., 2012).
None of the genotypes were as resistant to whitefly as the wild accession PI-127826 (Table 1), indicating that the ZGB- 703 and ZGB‑704 inbred lines, both derived from ‘PI‑127826’, do not have all the genes responsible for resistance to whitefly. Therefore, these inbred lines had a resistance level similar to lines with high ZGB content, indicating that they have the same resistance mechanism.
In the present study, genotypes with high AS and ZGB contents had adverse effects on the biological development of B. tabaci biotype B. Similar results were found by Silva et al. (2009) and Neiva et al. (2013) for B. tabaci biotype B and for other pests such as Tuta absoluta and Myzus persicae (Oliveira et al., 2012; Silva et al., 2013).
Resistance levels to B. tabaci biotype B oviposition mediated by AS and ZGB in the same genotypes (TOM‑778, TOM-779, and TOM-780) were significantly higher than in genotypes with only one allelochemical in their structure (Table 2, Contrasts C10, C11, and C13). Silva et al. (2009) found different results with respect to B. tabaci biotype B in which the double heterozygote genotypes had the same behavior as heterozygotes for ZGB or AS, and there was no synergistic effect of the simultaneous presence of ZGB and AS.
These results indicate that high AS and ZGB contents and the presence of the Mi gene are associated with higher resistance to whitefly in tomato. However, resistance levels provided by the presence of the Mi gene are lower than by high AS and/or ZGB contents. High ZGB content gives a similar resistance level for nymph survival in a high AS content, but it is slightly more effective than the latter in reducing oviposition.
Genotypes with high simultaneous AS + Mi were more resistant to whitefly than genotypes with only high AS or only the Mi gene. Genotypes with high AS + ZGB content exhibited a higher resistance level than genotypes with high AS or high ZGB content. It would seem that genotypes with the simultaneous presence of high AS and ZGB contents and with the Mi gene (high AS + high ZGB + Mi) could provide even higher resistance levels. To test this hypothesis, it is necessary to obtain a genotype that has the three traits, which is not available at the present time.
CONCLUSIONS
High foliar contents of allelochemicals such as acylsugar (AS) and zingiberene (ZGB) associated with the presence of the Mi gene in tomato plants may lead to higher resistance to whitefly. The resistance to whitefly mediated by the Mi gene is lower when compared with resistance levels mediated by AS and ZGB or simultaneously mediated by AS and ZGB. The combined effect of high AS content associated with the presence of the Mi gene is greater when compared with the effect of only high ZGB content on the resistance to whitefly