Introduction
The cultivated tomato plant, Solanum lycopersicum L., is susceptible to various pests and diseases that can reduce its productive capacity. Among the main pests are the tomato leaf miner Phthorimaea (Tuta) absoluta Meyrick (Lepidoptera: Gelechiidae) and mites from the Tetranychidae family, such as the two-spotted spider mite Tetranychus urticae Koch.
Phthorimaea absoluta is one of the most devastating pests of cultivated tomatoes in the tropical and subtropical regions, causing significant economic losses (Rakha et al., Reference Rakha, Zekeya, Sevgan, Musembi, Ramasamy and Hanson2017b; Shahini et al., Reference Shahini, Bërxolli and Kokojka2021). P. absoluta larvae preferentially feed on young plant tissues, thereby negatively affecting the overall plant growth and yield (El-Aassar et al., Reference El-Aassar, Soliman and Abd Elaal2015). The two-spotted spider mite is another important pest that primarily feeds on the abaxial surface of the leaf, producing a stippled appearance that progresses into necrotic spots, causing the foliage to turn yellowish or greyish, eventually desiccating and losing its biological function (Brust and Gotoh, Reference Brust and Gotoh2018).
Controlling these pests is complex because of their large population sizes and high reproductive potential, which facilitates the rapid development of resistance to synthetic acaricides and insecticides (Han et al., Reference Han, Zhang, Ye, Wang, Wang and Gao2024; Jiang et al., Reference Jiang, Yang, Zhang, Chen, Hu, Chen and Zhang2023; Kewedar et al., Reference Kewedar, Chen, Moural, Lo, Umbel, Forrence and Zhu2025; Prasannakumar et al., Reference Prasannakumar, Jyothi, Prasadbabu, Ramkumar, Asokan, Saroja and Sridhar2023). Complementary strategies for chemical control, such as the use of resistant cultivars, are essential for effective pest management. In tomato breeding programs, emphasis has been placed on selecting the morphological and chemical traits of leaves and other plant structures that confer pest resistance (de Almeida et al., Reference Almeida, de Resende, Hata, Oliveira and Neto2023; Mutschler et al., Reference Mutschler, Smith and Shimizu2023b; Terenciano et al., Reference Terenciano, da Silva, Bastos, Fernandes, Dias and de Sena Fernandes2025).
Solanum pennellii Correll is a wild relative of cultivated tomato, notable for its high acylsugar content, and is a valuable source of genes conferring resistance to arthropod pests (Lucini et al., Reference Lucini, Faria, Rohde, Resende and de Oliveira2015; Vendemiatti et al., Reference Vendemiatti, Nowack, Peres, Benedito and Schenck2025), which can be introgressed into commercial tomato cultivars (De Oliveira et al., Reference De Oliveira, de Resende, Maluf, Lucini, de Lima Filho, de Lima and Nardi2018). Acylsugar content can be used in breeding programs as an efficient indirect selection marker of pest resistance (Maluf et al., Reference Maluf, Inoue, Ferreira, Gomes, Castro and Cardoso2007; Marinke et al., Reference Marinke, Hata, Gomes, Shimizu and de Resende2025; Resende et al., Reference Resende, Silva, Maluf, Resende, Zeist and Gabriel2020). Acylsugars directly affect pest behaviour by reducing oviposition and mobility (De Lima Filho et al., Reference De Lima Filho, Resende, de Oliveira, Nardi, Silva, Rech, Oliveira, Ventura and Ribeiro Silva2022), which is negatively correlated with acylsugar levels in tomato leaflets (Marinke et al., Reference Marinke, de Resende, Hata, Dias, de Oliveira, Ventura and de Lima Filho2022; Resende et al., Reference Resende, Maluf, Cardoso, Faria, Gonçalves and Nascimento2008). Moreover, acylsugars are sticky exudates secreted by glandular trichomes, which are typically more densely distributed on the abaxial leaf surface, where they can trap adult females and ultimately cause mortality (Kartowikromo et al., Reference Kartowikromo, Pizzo, Rutz, Love, Simmons, Ojeda and Hamid2024; Marinke et al., Reference Marinke, de Resende, Hata, Dias, de Oliveira, Ventura and de Lima Filho2022). Therefore, acylsugars confer resistance primarily through antixenosis, deterring oviposition and/or antibiosis, and adversely affecting the biology of pests at various developmental stages, thereby preventing their successful development (Marinke et al., Reference Marinke, de Resende, Hata, Dias, de Oliveira, Ventura and de Lima Filho2022; Rakha et al., Reference Rakha, Bouba, Ramasamy, Regnard and Hanson2017a, Reference Rakha, Zekeya, Sevgan, Musembi, Ramasamy and Hanson2017b; Smeda et al., Reference Smeda, Smith and Mutschler2023).
Tomato cultivars resistant to the two-spotted spider mite and tomato leaf miner, if available, may provide growers with a more practical and cost-effective strategy for pest management (Rakha et al., Reference Rakha, Zekeya, Sevgan, Musembi, Ramasamy and Hanson2017b). Although resistance to either T. urticae or P. absoluta has been studied, few reports have simultaneously evaluated multiple pests in advanced interspecific hybrids. Therefore, the objective of this study was to evaluate the resistance of advanced tomato genotypes (hybrids) derived from interspecific crosses between S. lycopersicum and S. pennellii LA716 to infestation by T. urticae and P. absoluta and to identify the defence mechanisms involved.
Materials and methods
Experimental site
The plants were grown in a greenhouse at the Experimental Farm (FAZESC), and evaluations were conducted in the Horticulture and Entomology Laboratories at the State University of Londrina (23° 20' 24” S, 51° 12' 36” W; 544 m altitude) during the years 2022–2023. According to the Köppen classification, the regional climate is characterised as humid subtropical, with an average annual temperature of 22°C and average annual precipitation of 1,800 mm. During the summer, temperatures exceed 30°C, with rainfall ranges from 500 to 600 mm. The soil in the area is classified as a Red Latosol with clayey texture.
Plant material
Twenty-five genotypes were evaluated, including 14 F1 hybrids obtained from crosses among nine advanced lines (derived from interspecific crosses between S. lycopersicum and S. pennellii LA716 [a wild accession with high acylsugar content and resistance to arthropod pests]), the commercial hybrid Leblon F1 (S. lycopersicum) as a commercial control (Sakata Seed Sudamerica – Hortaliças | Leblon, 2024), and the wild accession S. pennellii LA716 used as a resistance standard (table 1).
Table 1. Tomato genotypes evaluated for resistance to the two-spotted spider mite and tomato pinworm
NCSU: North Carolina State University; Ind: indeterminate; Det: determinate.
Crosses and generation of F1 hybrids
Seeds of the lines TOM-808, TOM-810, TOM-667, TOM-615, TOM-761, TOM-757, TOM-717, BPX-443E-05-06-101-03, and BPX-443E-03-02-113-02 were sown in trays with commercial substrate (Carolina Soil®) and maintained in a greenhouse under standard conditions (∼25 °C; 60 ± 10% RH) and were irrigated via micro-sprinklers.
Planting beds were prepared using a bed shaper, with dimensions of 1.2 m width and 0.20 m height. Irrigation was performed using a drip system. The beds were covered with milky-coloured polyethylene mulch (25 microns thick). Basal and top fertilisation followed soil analysis recommendations,. Supplemental fertilisation was performed via fertigation, supplying macro- and micronutrients every 10 d in alternating cycles. Seedlings were transplanted at a spacing of 1.0 m between rows and 0.5 m between plants when they developed three to five true leaves (approximately 20 days after sowing).
Plants were trained using a two-stem system. Axillary shoots were removed when they reached 3–5 cm in length, and the stems were tied using plastic twine every week. Irrigation was performed twice daily for 5 min each. When the plants reached 2.10 m in height, apical pruning (the removal of terminal shoots) was performed. Disease control was achieved through preventive and curative applications of fungicides (Absoluto 720® SC at 200 mL per 100 L of water, Totalit® SC at 125 mL per 100 L of water, and Midas BR® WG at 160 g per 100 L of water). Weed control between the planting rows was performed periodically via manual removal and hoeing.
Crosses were initiated approximately two months after transplantation, coinciding with the emergence of the first floral buds. In the pollen-recipient genotypes, emasculation was performed by manually removing the anther cone from the unopened flowers to prevent self-pollination. Pollen was collected from donor genotypes using freshly opened flowers that had been detached from the plants. Using a handheld vibrator, pollen was extracted and collected in a dark-coloured cardstock container. Pollination was performed on the same day after pollen collection and flower emasculation, preferably in the morning. Each pollinated flower was marked with a piece of coloured wool tied to the peduncle, with the colour corresponding to the pollen donor genotype.
The fruits were harvested weekly starting at full maturity. The harvested fruits were placed in labelled plastic bags and stored at room temperature in the Horticulture Laboratory at the State University of Londrina. The hybrid seeds were then extracted from these fruits. After natural drying, the seeds were stored in paper envelopes in a cold chamber until they were sown. The F1 seeds from each cross were sown in 128-cell polystyrene trays filled with substrate, following the same management practices previously used for seedling production. The hybrid seeds were not subjected to any pre-planting treatment.
Acylsugar content quantification
Sampling was conducted according to Resende et al. (Reference Resende, Maluf, Cardoso, Nelson and Faria2002) using plants approximately 40–50 d after transplanting. Leaf samples were collected from the upper third of the plant, with six leaf discs collected per plant, totalling 4.21 cm2 of leaf area, using a 3/8” diameter hole punch. The samples (discs) were placed in test tubes containing 1 mL dichloromethane (CH2Cl2) and agitated in a vortex mixer for 30 s. Subsequently, the leaflets were removed to allow solvent evaporation, and 0.5 mL of 0.1 N sodium hydroxide dissolved in methanol was added, which was then evaporated. The residue was maintained at 100°C, and methanol was added three times at 2-min intervals to ensure a complete reaction. After complete evaporation of methanol, the residue was dissolved in 0.4 mL of water.
Acylsugars can exist as acylglucoses or acylsucroses; thus, to hydrolyse sucrose into glucose, 0.1 mL of 0.04 N hydrochloric acid was added, and the mixture was heated to a boil for 5 min. After cooling, 0.5 mL of the Sommogy & Nelson reagent (Reagent A + Reagent B in a 25:1 ratio) was added. The tubes were boiled for 10 min and then cooled under running water. Next, arsenomolybdate (0.5 mL) was added, and the mixture was vortexed for 15 s. The absorbance was measured at 540 nm using a spectrophotometer (Nelson, Reference Nelson1944). Acylsugar concentrations were directly proportional to the absorbance values.
Mass rearing of T. urticae
The stock colony of T. urticae was established on common bean plants (Phaseolus vulgaris L.) (fig. 1A), under controlled conditions (23–30 °C), with plants renewed every 10 days.
Figure 1. Rearing of Tetranychus urticae on potted common bean plants (Phaseolus vulgaris L.) maintained inside an insect rearing cage (A); eggs and mites on the abaxial surface of bean leaves (B); bioassay setup (C and D).
No-choice bioassay of T. urticae (two-spotted spider mite)
A completely randomised experimental design with nine replicates was used. Leaf discs (30 mm in diameter) cut using a circular biopsy punch were prepared from fully expanded leaflets collected from the middle third of 40- to 50-day-old plants of each genotype. The discs were carefully placed with forceps, with the abaxial side facing upward, on sponges covered with a layer of paper towels saturated with distilled water (to prevent mite escape). These were arranged within arenas consisting of Gerbox® plastic boxes (11 cm × 11 cm × 3.5 cm), with nine leaf discs positioned per arena (fig. 1C).
Six 10-day-old T. urticae adult females were collected and transferred to leaf discs using a stereomicroscope (Olympus Microscope Package SZ51 Binocular, Tokyo, Japan) and a fine-bristled brush. The mites were maintained for 24 h in a climate-controlled BOD-type chamber (25 ±2°C; 70 ±10% RH; 12-h photoperiod). After the 24-h period, the number of live mites and nymphs was recorded, and the surviving adult females were removed, leaving only the eggs. Eggs were counted and maintained to assess hatching rates at 96 and 120 h after female removal.
Bioassay with the tomato leafminer (P. absoluta)
For the P. absoluta bioassay, seedlings with three to five true leaves were transplanted. Plants were grown in 7 dm3 pots filled with a 1:1 mixture of surface soil (amended according to soil chemical analysis) and a commercial substrate based on stabilised pine bark. Forty grams of 04-14-08 (NPK) fertilizer formulation was added to each pot. The pots were maintained in a greenhouse (25°C, 60 ±10% RH), with plants staked for support and irrigated daily, following the management practices previously described. At 30 to 40 days after sowing, the pots were transferred to another greenhouse, where a high infestation of P. absoluta was present on tomato plants of the Pietra/Sakata® variety, allowing for the conduct of the assays on the genotypes under evaluation.
Assessments were conducted weekly, starting when the plants reached the pre-flowering stage (30–40 days after transplanting). Evaluations were conducted 14, 21, 28, and 35 days after infestation to measure oviposition and plant damage.
Oviposition
Leaflets from the upper third of the plant were used for the assessment, specifically to count the number of eggs on the abaxial surface. The leaflets were collected, placed in paper bags, and transported to the Entomology Laboratory of the State University of Londrina. The counts were performed under a stereomicroscope (Olympus Microscope Package SZ51 Binocular) within an area of 2 cm2,avoiding the central vein.
Damage severity on the plants, types of lesions, and percentage of damaged leaflets
Aboveground damage was assessed 14, 21, 28, and 35 days after infestation (DAI) using a rating scale proposed by Labory et al. (Reference Labory, Santa-Cecília, Maluf, Cardoso, Bearzotti and Souza1999). The evaluators randomly selected the plants. Scores showing high standard deviations were excluded during data weighting for mean estimation. Plant damage severity (whole-plant visual assessment), types of lesions on leaflets, and damaged leaflets: percentage of damage (whole-leaf visual assessment) were evaluated.
For scoring the tomato leaf miner bioassay, analyses were performed using both the arithmetic mean and weighted mean of the scores given by the four previously trained evaluators.
Experimental design and data analysis
The experimental design consisted of a randomised block design with three replicates, each plot containing eight plants, for the mite bioassay, and a completely randomised design with four replicates, with each experimental unit consisting of one plant, for the tomato leafminer bioassay.
The data were subjected to normality and homogeneity of variance tests and then transformed using √(x + 1). Statistical analyses were performed using RStudio software (Rstudio Team, 2020). The ExpDes.pt package was used for analysis of variance and the Scott–Knott means test (p < 0.05) (Ferreira et al., Reference Ferreira, Cavalcanti and Nogueira2021). The gplots and gmodels packages (Warnes et al., Reference Warnes, Bolker, Bonebakker, Gentleman, Huber, Liaw, Lumley, Maechler, Magnusson, Moeller, Schwartz and Galili2022a, Reference Warnes, Bolker, Lumley and Johnson2022b) were used for F-test comparisons (f-value < 0.05) between the breeding program genotypes and control. The pheatmap package (Kolde, Reference Kolde2018) was used for hierarchical clustering analysis using Ward’s method (Ward, Reference Ward1963) combined with Euclidean distance, as represented in a heatmap. The ggstatsplot package (Patil, Reference Patil2024) was used to perform Spearman’s correlation analysis (p < 0.05), and the agricolae package (Mendiburu, Reference Mendiburu2021) was used to calculate the area under the damage progress curve for the evaluated variables obtained by summing the trapezoids under the curve.
Results
S. pennellii (LA716), the wild relative of cultivated tomatoes, exhibited a high acylsugar content (0.99 nm) and distinguished itself from the other genotypes evaluated (F0.042, 0.001 = 4.19; p < 0.05). However, these genotypes presented relatively high levels of this allelochemical (ranging from 0.31 to 0.64 nm).
Two-spotted spider mite (T. urticae) bioassay
In the no-choice bioassay, the lowest number of surviving mites (0.67 individuals) (F0.104, 0.009 = 12.79; p < 0.05) and the highest mortality rate of adult females (5.33 individuals) (F0.684, 0.091 = 7.51; p < 0.05) were observed in S. pennellii LA716, which differed significantly from the other treatments. Subsequently, the hybrids F1(TOM-810 × TOM-717), F1(TOM-808 × TOM-717), F1(TOM-810 × BPX-443E-03-02-113-02), F1(TOM-808 × TOM-757), F1(TOM-808 × BPX-443E-03-02-113-02), and the lines TOM-757, TOM-810, TOM-761, and TOM-667 exhibited reduced survival rates (3.17 to 4.17 individuals) and, together with TOM-717, promoted higher mite mortality rates (1.67 to 2.83 individuals) to the mites (F0.104, 0.009 = 12.79; p < 0.05). The remaining genotypes did not differ significantly from the commercial controls regarding the number of alive (F0.104, 0.009 = 12.79; p > 0.05) and dead mites (F0.684, 0.091 = 7.51; p > 0.05). Overall, mite survival was 8% higher in the hybrids (F0.104, 0.009 = 12.79; p < 0.05), whereas mortality was 22% higher in the lines (table 2) (F0.684, 0.091 = 7.51; p < 0.05).
Table 2. Acylsugar content, mean number of live and dead mites, and eggs 24 hours after the release of six Tetranychus urticae females, and number of nymphs at 96 and 120 hours after egg deposition on the surface of leaf discs from 25 tomato genotypes
* When the F-value is less than 0.05, acylsugar content—measured by absorbance at 540 nm—shows significant differences; results followed by the same letter in a column do not differ significantly according to the Scott-Knott test (p < 0.05).
Reduced oviposition rates were recorded in the F1 hybrids F1(TOM-810 × BPX-443E-05-06-101-03), F1(TOM-808 × TOM-757), F1(TOM-808 × TOM-717), F1(TOM-810 × TOM-717), F1(TOM-810 × TOM-757), F1(TOM-810 × BPX-443E-03-02-113-02), and F1(TOM-808 × BPX-443E-03-02-113-02), as well as in the lines TOM-757, TOM-810, TOM-667, BPX-443E-05-06-101-03, TOM-761, BPX-443E-03-02-113-02, TOM-717, TOM-808, and S. pennellii LA716 (ranging from 0 to 3.67 eggs per female) (F11.310, 0.288 = 39.35; p < 0.05). Conversely, the highest oviposition rates were observed in F1(TOM-808 × BPX-443E-05-06-101-03), F1(TOM-808 × TOM-615), F1(TOM-810 × TOM-761), TOM-615, and the commercial hybrid Leblon F1(17.17 to 25.50 eggs per female) (F11.310, 0.288 = 39.35; p < 0.05). On average, females oviposited 52% more eggs on hybrids than on their parental lines but 63% less than the commercial control (Leblon F1) (F11.310, 0.288 = 39.35; p < 0.05) (table 2).
Nymph emergence 96 h after the release of female mites onto leaf discs was lower in the genotypes F1(TOM-808 × TOM-757), F1(TOM-810 × TOM-717), F1(TOM-810 × TOM-615), F1(TOM-810 × BPX-443E-03-02-113-02), F1(TOM-810 × BPX-443E-05-06-101-03), F1(TOM-808 × BPX-443E-03-02-113-02), F1(TOM-808 × TOM-717), F1(TOM-810 × TOM-757), TOM-761, TOM-717, TOM-667, BPX-443E-05-06-101-03, BPX-443E-03-02-113-02, and TOM-808, with results that were statistically similar to those observed for S. pennellii LA176 (0 to 0.97 emerged nymphs) (F1.772, 0.118 = 15.02; p < 0.05). In contrast, highest nymph emergence rates (6.41 to 6.81 emerged nymphs) (F1.772, 0.118 = 15.02; p < 0.05) were observed in the commercial hybrid Leblon F1, F1 (TOM-808 × BPX-443E-05-06-101-03), and TOM-615. The hybrids exhibited a nymph emergence rate approximately 13% higher than that of the inbred lines; however, the percentage of emerged nymphs relative to the total number of eggs laid was 27% higher in the inbred lines. This suggested that the hybrids had a stronger effect on nymph emergence, which may be associated with a higher oviposition rate (table 2).
At 120 h after the release of the female mites, the genotypes F1(TOM-808 × TOM-615), F1(TOM-810 × TOM-761), F1(TOM-808 × TOM-667), F1(TOM-808 × TOM-761), F1(TOM-810 × TOM-615), F1(TOM-808 × TOM-757), F1(TOM-808 × TOM-717), F1(TOM-810 × TOM-757), F1(TOM-810 × TOM-717), F1(TOM-810 × BPX-443E-03-02-113-02), F1(TOM-810 × BPX-443E-05-06-101-03), F1 (TOM-808 × BPX-443E-03-02-113-02), TOM-810, TOM-667, TOM-757, BPX-443E-03-02-113-02, TOM-761, BPX-443E-05-06-101-03, TOM-717, TOM-808, and S. pennellii LA176 stood out due to their reduced number of emerged nymphs (<3 nymphs) (F2.448, 0.147 = 16.62; p < 0.05). The nymph emergence rate at 120 h was approximately 24% higher in the hybrids (table 2).
The contrasts indicated that the hybrids exhibited significantly lower numbers of mites, deposited eggs, and emerged nymphs than the commercial control, Leblon F1 (p < 0.05). However, these strains did not show statistically similar results to S. pennellii LA716 for any of the variables assessed in the no-choice bioassay (p < 0.05) (table 2).
Acylsugar content was significantly negatively correlated with the number of live mites and eggs laid, and positively correlated with the number of dead mites (p < 0.05). This indicated that high levels of acylsugars reduced the survival rate of T. urticae, inhibited oviposition, and negatively affected mite viability. The number of live mites exhibited a strong, significant positive correlation with the number of eggs laid and the number of nymphs that emerged at both 96 and 120 h (p < 0.05). Conversely, the number of dead mites was negatively correlated with the number of eggs laid, number of hatched eggs at 96 and 120 h, and number of nymphs (p < 0.05) (fig. 2).
Figure 2. Correlation matrix of the results obtained from the no-choice test with Tetranychus urticae in the 25 genotypes analysed. Correlations were subjectively categorised as low (±0.18–0.30): moderate (±0.31–0.50), high (±0.51–0.75), and very high (±≥0.76).
In the heatmap, hierarchical clustering was performed using Euclidean distance and Ward’s method (Ward, Reference Ward1963), based on the mean values obtained from the no-choice test with the two-spotted spider mite. The highly resistant accession S. pennellii LA176 did not cluster with any of the evaluated genotypes, forming a separated group (first group). The second group consisted a commercial control (cv. Leblon), hybrid F1(TOM-808 × BPX-443E-05-06-101-03), and the inbred line TOM-615, which exhibited the highest values for surviving mites, oviposition, and the highest numbers of hatched eggs and nymphs (fig. 3).
Figure 3. Heatmap of hierarchical clustering (Euclidean distance using Ward’s method [1963]) of tomato genotypes. based on the results obtained from the no-choice test with the two-spotted spider mite (Tetranychus urticae).
The third largest group comprised the remaining genotypes, which were further subdivided into two smaller clusters. The first subgroup included F1(TOM-808 × BPX-443E-03-02-113-02), F1(TOM-808 × TOM-757), F1(TOM-810 × BPX-443E-03-02-113-02), F1(TOM-808 × TOM-717), F1(TOM-810 × TOM-717), TOM-667, TOM-761, TOM-810, TOM-757, TOM-717, TOM-808, and BPX-443E-03-02-113-02, which showed lower numbers of eggs, nymphs, and survival rates than the second subgroup. This second subgroup consisted of F1(TOM-810 × TOM-615), F1(TOM-810 × BPX-443E-05-06-101-03), F1(TOM-810 × TOM-667), F1(TOM-808 × TOM-667), F1(TOM-810 × TOM-761), F1(TOM-808 × TOM-761), F1(TOM-808 × TOM-615), F1(TOM-810 × TOM-757), and BPX-443E-05-06-101-03 (fig. 3).
Tomato pinworm bioassay
The area under the oviposition progress curve (AUOPC) was a quantitative parameter used in this study to assess the progression of oviposition and foliar damage on the plants. The AUOPC values were derived from the results of the four assessment points (14, 21, 28, and 35 days after infestation), allowing for the distinction between treatments that were more susceptible or resistant to P. absoluta infestation. Based on these results, the commercial control (cv. Leblon) was the most preferred for oviposition by tomato pinworms, whereas S. pennellii LA716 was not preferred. The remaining hybrid and inbred lines did not differ significantly (table 3).
Table 3. Acylsugar content (AA) and area under the damage progress curve (AUDPC) for oviposition (OV), plant damage (DP), type of leaflet damage (TD), and percentage of damaged leaflets (DF) at 14, 21, 28, and 35 days after infestation with Phthorimaea absoluta in 25 evaluated genotypes
* Indicates significance at the 5% level according to the F-test and associated probability level. Means followed by the same letter within a column do not differ significantly according to the Scott-Knott test at a 5% significance level.
Regarding plant damage intensity, the commercial control was the most susceptible to tomato pinworm. S. pennellii LA716 exhibited the highest resistance, followed by the genotypes F1(TOM-808 × TOM-717), F1(TOM-808 × TOM-667), TOM-615, TOM-757, BPX-443E-05-06-101-03, TOM-667, and TOM-810, which, although differing from the wild tomato, showed fewer foliar lesions than the other genotypes (F371.489, 160.824 = 2.31; p < 0.05) (table 3).
In terms of lesion type, S. pennellii LA716 was the most resistant, with only a few small foliar lesions (F242.553, 74.563 = 3.25; p < 0.05). The most resistant genotypes thereafter were F1(TOM-808 × TOM-667), F1(TOM-808 × TOM-717), F1(TOM-808 × BPX-443E-05-06-101-03), F1(TOM-810 × TOM-717), F1(TOM-808 × TOM-757), F1(TOM-810 × TOM-761), F1(TOM-810 × TOM-667), F1(TOM-810 × BPX-443E-05-06-101-03), TOM-615, TOM-810, TOM-757, BPX-443E-05-06-101-03, TOM-667, TOM-717, and TOM-808, all of which exhibited a low numbers of small to medium lesions (F242.553, 74.563 = 3.25; p < 0.05) (table 3).
Leblon F1, F1(TOM-808 × BPX-443E-03-02-113-02), F1(TOM-808 × TOM-615), and F1(TOM-810 × TOM-615) were the most susceptible, showing the greatest foliar loss, whereas S. pennellii LA716 had the lowest percentage of damaged leaflets (F250.242, 53.228 = 4.70; p < 0.05). Overall, F1(TOM-808 × TOM-717), F1(TOM-808 × TOM-667), TOM-615, BPX-443E-05-06-101-03, TOM-757, TOM-667, and TOM-810 were the most resistant to P. absoluta, as they simultaneously presented the lowest number of eggs and reduced foliar damage (F250.242, 53.228 = 4.70; p < 0.05) (table 3).
Contrast analysis indicated more resistance among the hybrids compared to the commercial control, although it was still lower than that observed in S. pennellii LA716 (p < 0.05). Correlations with acylsugar content were not statistically significant, but were negative, suggesting a potential deleterious effect on tomato pinworms (p > 0.05). In contrast, oviposition was positively associated with the percentage of plant damage, leaflet damage, and lesion type (p < 0.05) (table 3).
Cluster analysis based on the area under the progress curve for each assessed variable grouped genotypes into four distinct clusters. The first group, comprising F1(TOM-810 × TOM-757), F1(TOM-808 × TOM-615), F1(TOM-808 × BPX-443E-03-02-113-02), F1(TOM-808 × TOM-757), F1(TOM-810 × TOM-667), F1(TOM-808 × TOM-761), F1(TOM-810 × BPX-443E-05-06-101-03), F1(TOM-810 × TOM-615), F1(TOM-808 × BPX-443E-05-06-101-03), F1(TOM-810 × TOM-717), F1(TOM-810 × TOM-761), F1(TOM-810 × BPX-443E-03-02-113-02), TOM-761, BPX-443E-03-02-113-02, TOM-717, and TOM-808, showed intermediate performance in both oviposition rate and foliar damage (fig. 4).
Figure 4. Heatmap of hierarchical clustering (Euclidean distance using Ward’s method [1963]) of tomato genotypes. based on the analysis of the area under the damage progress curve caused by the tomato leafminer (Phthorimaea absoluta). AA: acylsugar; OV: oviposition; DP: plant damage; TP: type of damage; DF: leaflet damage.
The second group – F1 (TOM-808 × TOM-667), F1 (TOM-808 × TOM-717), TOM-810, TOM-757, TOM-667, BPX-443E-05-06-101-03, and TOM-615 – demonstrated, after S. pennellii LA716, the greatest resistance to oviposition and foliar attack. The final two clusters separated S. pennellii LA716, the wild and resistant accession to tomato pinworm, from Leblon F1 (the commercial control), which was the most susceptible, exhibiting the highest oviposition preference and feeding damage, resulting in greater foliar injury (fig. 4).
Discussion
In the current study, the number of eggs laid by the mites ranged from 0.03 to 4.25. Reduced oviposition levels may be associated with the release of acylsugars by plants attacked by mites, since these pests tend to prefer more susceptible genotypes, and the presence of this secondary metabolite may act as a repellent (Tabary et al., Reference Tabary, Navajas, Tixier and Navia2024a, Reference Tabary, Navia, Auger, Migeon, Navajas and Tixier2024b). According to Resende et al. (Reference Resende, Maluf, Cardoso, Faria, Gonçalves and Nascimento2008), high levels of acylsugars repel Tetranychus evansi infestation in tomatoes (S. lycopersicum). The reduced number of eggs laid by T. urticae is often related to elevated acylsugar content (Lucini et al., Reference Lucini, Faria, Rohde, Resende and de Oliveira2015; Rakha et al., Reference Rakha, Bouba, Ramasamy, Regnard and Hanson2017a), which may be associated with reduced mite movement in more resistant plants (De Lima Filho et al., Reference De Lima Filho, Resende, de Oliveira, Nardi, Silva, Rech, Oliveira, Ventura and Ribeiro Silva2022).
In addition to influencing behaviour, a high acylsugar content may exert an antibiotic effect on mites. Gomes et al. (Reference Gomes, Cardoso, Resende, Thomasi, Soares and Ferreira2017) observed that extracts from the wild species S. pennellii, which is rich in acylsugars, have ovicidal effects and impair nymph emergence in T. urticae by increasing the egg incubation period and reducing hatching rates, indicating an antibiotic effect. Ovicidal activity was observed in the tested genotypes, as most of them did not reach 100% nymph emergence. The highest number of emerged nymphs was recorded after 120 h, suggesting either delayed development or embryo non-viability (table 2).
Adult T. urticae survival was affected by the acylsugar content of the evaluated genotypes, with only 11% of the females surviving on S. pennellii LA716 in the no-choice bioassay (table 2). In another study, the cultivar Redenção (a susceptibility standard) (S. lycopersicum) exhibited a significantly higher mean number of live mites (T. urticae) than S. pennellii (LA716), with values of 4.7 and 0.6, respectively, representing an 87% difference (Lucini et al., Reference Lucini, Faria, Rohde, Resende and de Oliveira2015). In an experiment by De Lima Filho et al. (Reference De Lima Filho, Resende, de Oliveira, Nardi, Silva, Rech, Oliveira, Ventura and Ribeiro Silva2022), the survival rate was calculated as the mean survival rate of one mite on S. pennellii and five mites on S. lycopersicum (cv. Redenção).
Similarly to their effects on mites, acylsugars also negatively impact the life cycle of the tomato pinworm, and at high concentrations, they interfere with the biological behaviour of the insect, including the inhibition of oviposition (Dias et al., Reference Dias, Corte, Resende, Zeffa, Resende, Zanin and Lima Filho2021), as observed in the genotypes evaluated. According to Dias et al. (Reference Dias, de Resende, Zeist, Gabriel, Santos and Vilela2019), a negative correlation between acylsugar content and the number of P. absoluta eggs and larvae suggests that insects prefer plants with lower levels of this allelochemical in their leaflets, supporting the role of acylsugars in plant defence via antixenosis. Furthermore, high population densities of the pest and the absence of a susceptible host may promote the migration of these insects to genotypes with higher acylsugar content, a situation commonly observed in confined environments such as greenhouses.
The evaluations indicated the presence of antixenotic resistance in S. pennellii LA716, and, to a lesser extent, in F1(TOM-808 × TOM-757), F1(TOM-810 × TOM-667), F1(TOM-808 × TOM-667), and F1(TOM-808 × TOM-717), as well as in the inbred lines TOM-667, TOM-808, TOM-810, TOM-615, and TOM-757, which showed reduced attractiveness for oviposition by tomato pinworms . The hybrid F1(TOM-808 × TOM-667) showed lower scores for foliar damage, comparable to those of S. pennellii LA716, indicating that feeding was negatively affected in these genotypes (table 3). A plant that is less preferred for oviposition is one with a lower likelihood of supporting the establishment of the tomato pinworm larval stage, thereby reducing the chances of initial infestation. The constitutive emission of specific volatile deterrents is lower in domesticated tomato compared to its wild relatives, which likely influences oviposition preferences (Paudel et al., Reference Paudel, Lin, Foolad, Ali, Rajotte and Felton2019).
Higher acylsugar content confers increased resistance to feeding damage caused by tomato pinworms, as observed in S. pennellii LA716 in the assessments of foliar damage percentage and damage type (table 3). The cultivar Redenção, commonly used as a susceptibility standard in bioassays, was completely destroyed 21 days after infestation with P. absoluta, whereas S. pennellii, which is rich in acylsugars, exhibited only minor lesions and less severe damage types (Dias et al., Reference Dias, de Resende, Zeist, Gabriel, Santos and Vilela2019). In another study, S. pennellii was compared with 76 experimental lines and proved to be significantly more resistant, with an average of 1.6 eggs laid and mean scores for leaflet lesions, attacked leaflets, and overall plant damage of 0.48, 0.46, and 0.46, respectively (Resende et al., Reference Resende, Silva, Maluf, Resende, Zeist and Gabriel2020), consistent with the results obtained in the current study.
Plants possess defence mechanisms against pests, including physical barriers or chemical defences, which can be exploited in integrated pest management (IPM) programs (de Almeida et al., Reference Almeida, de Resende, Hata, Oliveira and Neto2023; Marinke et al., Reference Marinke, Hata, Gomes, Shimizu and de Resende2025; Tabary et al., Reference Tabary, Navajas, Tixier and Navia2024a). Species within the Lycopersicon group exhibit both non-glandular (types II, III, V, and VIII) and glandular trichomes (types I, IV, VI, and VII) (Almeida et al., Reference Almeida, de Resende, Hata, Oliveira and Neto2023; Luckwill, Reference Luckwill1943). Non-glandular trichomes act mechanically, impeding pest movement and access to plant tissues, whereas glandular trichomes produce sticky and/or toxic exudates capable of entrapping and potentially killing arthropods (Kartowikromo et al., Reference Kartowikromo, Pizzo, Rutz, Love, Simmons, Ojeda and Hamid2024; Marinke et al., Reference Marinke, de Resende, Hata, Dias, de Oliveira, Ventura and de Lima Filho2022).
Solanum pennellii, a wild relative of the cultivated tomato (S. lycopersicum), produces allelochemicals in its glandular trichomes, known as acylsugars, which significantly hinder feeding and reproduction in many arthropod pests (Lybrand et al., Reference Lybrand, Anthony, Jones and Last2020). Therefore, this species has been widely used in interspecific crosses to develop pest-resistant lines (Dias et al., Reference Dias, de Resende, Zeist, Gabriel, Santos and Vilela2019; Marinke et al., Reference Marinke, Hata, Gomes, Shimizu and de Resende2025), particularly because of its high heritability of resistance. This enables for a combination of high levels of arthropod resistance with desirable commercial traits in improved lines (Resende et al., Reference Resende, Silva, Maluf, Resende, Zeist and Gabriel2020). Acylsugars are secreted as droplets at the tips of trichomes, allowing any arthropod arriving on the leaf surface to come into direct contact with the allelochemical, which is continuously produced and replenished when the droplets are removed by touch or mechanical force (Mutschler et al., Reference Mutschler, Smith and Shimizu2023b).
Plant breeding represents a key strategy in IPM programs as it is fully compatible with other control methods. Biological and chemical products are typically used in combination; however, their use may be constrained by limitations or antagonistic and incompatible interactions (Papari et al.,, Reference Papari, Dousti, Fallahzadeh, Haddi, Desneux and Saghaei2024; Sehat-Niaki et al., Reference Sehat-Niaki, Zahedi Golpayegani, Torabi, Saboori, Amiri-Besheli and Fathipour2025b). In addition, certain chemical pesticides can irritate predatory mites such as Neoseiulus californicus (McGregor) (Acari: Phytoseiidae) and Phytoseiulus persimilis Athias-Henriot (Acari: Phytoseiidae), altering their behaviour and reducing the effectiveness of T. urticae control (Sehat-Niaki et al., Reference Sehat-Niaki, Zahedi Golpayegani, Torabi, Amiri-Besheli and Saboori2025a, Reference Sehat-Niaki, Zahedi Golpayegani, Torabi, Saboori, Amiri-Besheli and Fathipour2025b). When incorporated into IPM programs, resistant genotypes can reduce the excessive use of chemical pesticides, lower pesticide residues in fruits, and delay the development of resistance to synthetic acaricides and insecticides (Han et al., Reference Han, Zhang, Ye, Wang, Wang and Gao2024; Kewedar et al., Reference Kewedar, Chen, Moural, Lo, Umbel, Forrence and Zhu2025; Munir et al., Reference Munir, Azeem, Zaman and Haq2024; Prasannakumar et al., Reference Prasannakumar, Jyothi, Prasadbabu, Ramkumar, Asokan, Saroja and Sridhar2023).
The results of this study highlight the potential of introgressing resistance traits from S. pennellii into advanced tomato lines and hybrids to improve resistance against T. urticae and P. absoluta. However, future research should expand beyond controlled bioassays and incorporate field trials to validate resistance under natural infestation pressure and diverse environmental conditions. Such studies should not only assess the stability of pest resistance these two specific pest species, but other pests, e.g. whitefly, Bemisia tabaci Genn. and thrips, Frankliniella occidentalis (Pergande). Also, evaluation of agronomic performance, including yield, fruit quality, and adaptability to commercial production systems should be considered.
Conclusions
F1(TOM-808 × BPX-443E-03-02-113-02), F1(TOM-810 × BPX-443E-03-02-113-02), F1(TOM-808 × TOM-717), F1(TOM-808 × TOM-757), and F1(TOM-810 × TOM-717) were the most resistant genotypes to two-spotted spider mites, promoting higher female mortality, reduced oviposition, and no nymph emergence was observed.
The hybrids that showed the greatest resistance to P. absoluta were F1(TOM-808 × TOM-667) and F1(TOM-808 × TOM-717), which were the least oviposited and exhibited the lowest foliar damage intensity and severity.
The hybrid F1(TOM-808 × TOM-717) exhibits notable resistance to both pests.
Acknowledgements
The authors thank the Coordination for the Improvement of Higher Education Personnel (CAPES) for their contribution to the research. The authors also acknowledge Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) for providing a research productivity fellowship to J.T.V.R.
Competing interests
All authors declare no conflicts of interest.
