Introduction

The tomato moth, Tuta absoluta (Meyrick) (Lepidoptera:Gelechiidae), is a devastating insect pest affecting tomato production in Iran and many other countries (Baniameri & Cheraghian 2011, Gharekhani & Salek-Ebrahimi 2014). The larvae produce galleries in the leaves, stems, terminal buds and fruits. The main damage is caused through larvae feeding on the parenchyma between the epidermal layers of the leaves, reducing the photosynthetic capacity of the plant with subsequent reduction of the yield (Desneux et al 2010, 2011). Under heavy infestation, the yield loss between 80–100% is common (Gebremariamd 2015).

In many agronomic and vegetable cropping systems, the primary strategy employed to control this pest involves the use of chemical insecticides. However, it has serious problems such as destruction of natural enemy populations (Campbell et al 1991), build-up of insecticide residues on tomato fruits (Walgenbach et al 1991) and in the environment, and, especially, evolution of T. absoluta resistance to many of the active ingredients available on the market (Siqueira et al 2000a, b, Lietti et al 2005, Silva et al 2011, Campos et al 2015, Roditakis et al 2015).

One of the important control methods for sustainable management of T. absoluta to minimize development of pesticide resistance is the use of resistant host plants. The use of resistant plants can be a useful component of an integrated pest management (IPM) system that could affect pest population density, herbivore damage, and decrease pesticide applications in agricultural ecosystems. In addition, in many cases, even partial resistant cultivars are useful to enhance the effects of beneficial natural enemies (Hare & Andreadis 1983, Bong et al 1991, Cogni et al 2002, Kaplan 2007, Kaplan & Thaler 2010). Plant resistance to a pest can be caused by antixenosis, a mechanism employed by the host plants, deters the insects from oviposition, feeding, seeking shelter, and colonization (Oyetunji et al 2014); antibiosis, which has a direct influence on the life history of a pest (Ofomata et al 2000, Li et al 2004); and tolerance, the plant’s capacity of keeping its production under attack for herbivore insect (Vargas 1970, Stowe et al 2000, Stevens et al 2008). Many researchers have investigated the resistance of host plants to tomato moth (Gilardón et al 2001a, Leite et al 2001, Suinaga et al 2004, Silva 2009, Gharekhani & Salek-Ebrahimi 2014). In this study, we present data on susceptibility of 11 tomato genotypes to T. absoluta. The data obtained from these experiments are used to understand the mechanism of population build-up of this pest on different tomato genotypes to develop a comprehensive pest management program for tomato.

Material and Methods

The experiment was performed in a field in the Faculty of Agriculture, Persian Gulf University, Bushehr province, Borazjan region, Bondarooz (Southern Iran) (29°12′54.1″ N, 51°13′57.1″ E, and elev. 99 m) from February 2014 to June 2015. Eleven cultivars of tomato were used in this study, including five cultivars Ps-6515, Berlina, Poolad, Petoprid-5, Zaman from FALAT Co., Iran; two cultivars Matin and Golsar from GOLSAM Co., Iran; and four cultivars Sandokan-F1, Golshan-616, Sadeen-95 and Sadeen-21 from BEHTA Co., Iran. The tomato seeds were planted in plastic transplant trays containing peat moss soil and perlite on November 2014. With the appearance of the first true leaves, the seedlings were transplanted into the main field. The evaluation of the resistance was performed in a randomized block design with three replications. A total of 20 plants per cultivar were planted in each replicate plot (plot area = 24 m2) in two 150-cm spaced rows. The space between plants in each row was 50 cm, and the space between plots was 150 cm. The cultivars were exposed to the natural infestation by indigenous population of tomato leaf miner in the field.

Leaflet damage, leaf damage, and overall plant damage caused by the insect were evaluated at days 20, 40, and 60 after infestation. Five plants were randomly selected in each plot and marked notes and the different characteristics on these plants were measured. Leaflet and leaf damage was evaluated based on the percentage of leaflet or leaf area affected by T. absoluta. In this case, three leaves were selected from the upper third of each of the five selected plants and the damaged area of each whole leaf and its leaflets was recorded. The overall plant damage estimates were also performed for each of the five selected plants. In addition, the number of mines per leaf, holes on the stem, and holes per fruit were assessed at the last sampling date. The number of mines per leaf was counted on three leaves selected from the upper third of each of the five randomly selected plants. The number of holes on the stem and the number of holes on the fruit were evaluated by performing a direct counting of these features throughout the stem and five fruits from each of these five selected plants, respectively. In addition, three leaves were selected from the upper third of each of the five randomly selected plants for counting the density of total trichomes and type VI glandular trichomes on the leaves. Then, three leaflets separated from each leaf and trichomes were counted using a stereomicroscope (×40) on three 2-cm2 regions of each leaflet. The total number of trichomes and type VI glandular trichomes were also counted on different 2-cm2 sections of the stem.

The normality of data was assessed with Shapiro-Wilk’s test (Proc Univariate, SAS Institute (2003), Cary, NC, USA). Data which needed to be normalized were transformed before being analyzed. The percentage data were subjected to arcsin square root transformation; however, no count data transformation was performed before analysis because there was no evidence of non-normality within these data. Analysis of variance was performed using the general linear model (GLM) procedure in the SAS software and means were compared using Duncan’s multiple range test. Total damage index for each cultivar was presented as thesum of the indices gained for different evaluated traits. Index for each trait was calculated by dividing the lowest recorded number in a certain cultivar to the greatest recorded number for that trait in all cultivars. Pearson’s correlations (5% significance) were used to evaluate the relationships between traits. Data were then subjected to stepwise regression with overall plant damage at 60 day as the dependent variable. Cultivar comparison and selection was accomplished by cluster analysis according to Ward’s method using SAS software.

Results and Discussion

According to data of leaflet damage and leaf damage, and overall plant damage, we found that the cultivars Berlina, Golsar, Poolad, and Zaman more effectively avoided damage caused by T. absoluta in the three evaluation periods (Table 1). In contrast, Matin, Petopride5, SandocanF1, and PS6515 showed significantly greater damage rates than other evaluated cultivars (Table 1). There were also significant differences in the number of mines on the leaf (F = 3.33, df = 10, 20, P = 0.0107), number of holes on the stem (F = 3.37, df = 10, 20, P = 0.0110), and fruit (F = 9.10, df = 10, 20, P < 0.0001) caused by T. absoluta among the tomato cultivars (Table 2). The lowest number of mines on the leaf and holes on the stem were observed in the cultivars Berlina, Zaman, Golsar, Poolad, and Sadeen21. The cultivars Berlina, Zaman, Golsar, and Poolad also had the lowest number of holes on the fruit (Table 2).

Table 1 The mean percent of leaflet damage, leaf damage, and overall plant damage in tomato cultivars submitted to infestation of Tuta absoluta in different evaluation periods.
Full size table
Table 2 The mean number of mines on the leaf, number of holes on the stem, and number of holes on the fruit of tomato cultivars submitted to infestation of Tuta absoluta.
Full size table

Total damage index for evaluated tomato cultivars has been presented in Table 3. Based on the results, the cultivars Berlina, Golsar, Zaman, Poolad, and Sadeen21 with the lowest total damage index (1.75, 2.05, 2.15, 2.23, and 2.76, respectively) were the most resistant cultivars against T. absoluta. The greatest total damage index was obtained for the cultivars Petopride5, PS6515, Matin, and SandocanF1 (4.76, 4.6, 4.47 and 4.24, respectively) sustained less damage from pest (Table 3).

Table 3 Estimates of damage indices for different plant parts at day 60 after infestation and total damage index for tomato cultivars submitted to infestation of Tuta absoluta.
Full size table

Genetic variability is one of the characteristics of the germplasm bank subsamples that gives higher or lower susceptibility to pest insects (Fernandes et al 2012). So, observed differences between the levels of damages caused by T. absoluta on different tomato cultivars in the present study may have occurred because of genetic variability among them. Resende et al (2006), Gonçalves et al (2008), Oliveira et al (2009), Gonçalves Neto et al. (2010), Maciel et al (2011), and Gharekhani & Salek-Ebrahimi (2014) have also observed resistance to T. absoluta as non-preference and antibiosis in some evaluated tomato cultivars.

The genetic diversity of tomato cultivars may display inappropriate morphophysiological features to oviposition of T. absoluta adults and/or restrict the larvae feeding (Sobreira et al 2009). Trichome density is the most important structural feature of plants known to confer resistance to insect pests (Sharma et al 2009, He et al 2011). In the present study, significant differences were observed in the total number of trichomes on the leaf (F = 21.47, df = 10, 20, P < 0.0001), number of type VI glandular trichomes on the leaf (F = 46.58, df = 10, 20, P < 0.0001), and number of type VI glandular trichomes on the stem (F = 7.54, df = 10, 20, P < 0.0001) among the cultivars (Table 4). However, differences in the total number of trichomes on the stem were not significant (P ˃ 0.05) (Table 4). The cultivars with more total trichome density on the leaf were Sadeen95, Berlina, and Golsar. The most density of type VI glandular trichomes on the leaf was observed in Sadeen95 and Golsar (Table 4). Golsar, Sadeen95, Zaman, Petopride5, PS6515, and Golshan616 had the greatest number of type VI glandular trichomes on the stem. The lowest total number of trichomes and also type VI glandular trichomes on the leaf was observed in Petopride5. The cultivars Matin, Poolad, Berlina, and Sadeen21 had the lower number of type VI glandular trichomes on the stem (Table 4).

Table 4 Average densities (number/2 cm2) of total trichomes and type VI glandular trichomes on the leaf and stem for the plants used in the bioassay.
Full size table

Pearson’s correlations revealed that the total number of trichomes on the leaf had negative and significant correlations with the overall plant damage (r = −0.42), leaf damage (r = −0.58), leaflet damage (r = −0.58), number of mines on the leaf (r = −0.40), number of holes on the stem (r = −0.40), and number of holes on the fruit (r = −0.63) (Table 5). Person’s correlation estimates between type VI glandular trichomes on the leaf with overall plant damage (r = −0.40), leaf damage (r = −0.56), leaflet damage (r = −0.58), number of mines on the leaf (r = −0.40), number of holes on the stem (r = −0.40), and number of holes on the fruit (r = −0.61) were also negative and significant (Table 5). These results suggest trichomes may have direct negative influence on both larval feeding and oviposition by insects (Handley et al 2005), result in the lowest number of larvae and consequently lower damage to leaves and plants. Gilardón et al (2001a) and Neves et al (2003) also reported significant positive correlation between the density of trichomes on the leaves and resistance to Tuta species, as well as its relation to trichomes type VI. Thus, the high density of leaf trichomes present in the cultivars Sadeen95, Berlina, and Golsar can be one of the possible causes of resistance to T. absoluta known as the antixenosis mechanism. Oliveira et al (2009) also observed that the HGB 1497 subsample of Solanum lycopersicum L. presented resistance by antixenosis to the tomato plant miner T. absoluta. The high density of trichomes on tomato leaves can be extremely important for a cultivar to avoid the presence of T. absoluta. In addition, different metabolites are secreted from trichomes on the stems and leaves of the tomato plants, which cause different resistance against T. absoluta (Gilardón et al 2001b). Compounds such as tridecan-2-one and undecan-2-one, especially secreted by type VI glandular trichomes on the tomato leaves, perform as physical and chemical barriers for insects and pathogens (Farrar & Kennedy 1991, Eigenbrode & Espelie 1995, Justus et al 2000, Picoaga et al 2003). Such features can be used in plant breeding programs aimed at resistance to pests with selections toward genes that express a higher number of trichomes. An exception was Poolad with a low number of trichomes on the leaf (Table 4), which showed high resistance to T. absoluta (Tables 1, 2 and 3). Also, Sadeen95 with the most density of its trichomes (Table 4), showed partially resistance to the pest (Tables 1 and 3). This result can be explained by the presence and role of other potential resistance factors such as allelochemicals that confer resistance to T. absoluta as shown by Leite et al (1999) and Suinaga et al (2004).

Table 5 Estimates of Pearson’s correlations among the evaluated characteristics in tomato cultivars submitted to infestation of Tuta absoluta.
Full size table

Person’s correlation estimates between leaf and leaflet damage with the number of mines on the leaf, number of holes on the stem, and fruit were positive and significant (Table 5). Overall plant damage also had positive and significant estimates of correlation with leaf damage (r = 0.76), leaflet damage (r = 0.78), number of mines on the leaf (r = 0.10), number of holes on the stem (r = 0.10), and number of holes on the fruit (r = 0.73) (Table 5). The number of holes on the fruit had positive and significant correlations with the number of mines on the leaf (r = 0.75) and number of holes on the stem (r = 0.74) (Table 5). The number of holes on the stem had positive and significant correlation with the number of mines on the leaf (r = 0.10) (Table 5).

Stepwise regression is an automated tool used in the exploratory stages of model building to identify a useful subset of predictors. The process systematically adds the most significant variable or removes the least significant variable during each step. In order to remove the effect of non-effective characteristics in the regression model on grain yield, stepwise regression was used. The results of the stepwise regression analysis are presented in Table 6. The number of mines on the leaf (x4) was the variable that best explained overall plant damage at day 60 after infestation (y) as shown by stepwise regression (Table 6). The total number of trichomes on the stem (x1) was the second variable that exerted influence on overall plant damage. Parameter estimates showed that the number of mines on the leaf had positive significant effect, while the number of trichomes on the stem negatively affected overall plant damage at day 60 (Table 6).

Table 6 Results of stepwise multiple regression analysis between overall plant damage at day 60 (y) and the evaluated characteristics.
Full size table

The results of cluster analysis separated 11 evaluated tomato cultivars in three distinctive categories including Petopride5, Matin, SandokanF1, and PS6515 as susceptible; Golshan616, Sadeen21, and Sadeen95 as partially resistant; and Berlina, Zaman, Golsar, and Poolad as resistant cultivars (Fig 1).

Fig 1
figure 1

Dendrogram of 11 tomato cultivars for six studied variables using hierarchical cluster analysis (Ward’s method).

Full size image

Conclusions

Significant differences in relative resistance of the studied tomato cultivars demonstrate that the ones have potential for use in backcrosses in processing tomato breeding programs. However, our results are preliminary and require future studies for identifying the other resistance factors, other than trichome density, associated with these cultivars. Also, additional analyses with molecular markers will be needed for indicating the probable genetic variation between these tomato cultivars.