Antixenosis of bean genotypes to Chrysodeixis includens ( Lepidoptera : Noctuidae )

The objective of this work was to evaluate bean genotypes for resistance to soybean looper (Chrysodeixis includens). Initially, free‐choice tests were carried out with 59 genotypes, divided into three groups according to leaf color intensity (dark green, light green, and medium green), in order to evaluate oviposition preference. Subsequently, 12 genotypes with high potential for resistance were selected, as well as two susceptible commercial standards. With these genotypes, new tests were performed for oviposition in a greenhouse, besides tests for attractiveness and consumption under laboratory conditions (26±2oC, 65±10% RH, and 14 h light: 10 h dark photophase). In the no‐choice test with adults, in the greenhouse, the 'IAC Jabola', Arcelina 1, 'IAC Boreal', 'Flor de Mayo', and 'IAC Formoso' genotypes were the least oviposited, showing antixenosis‐type resistance for oviposition. In the free‐choice test with larvae, Arcelina 4, 'BRS Horizonte', 'Pérola', H96A102‐1‐1‐1‐52, 'IAC Boreal', 'IAC Harmonia', and 'IAC Formoso' were the less consumed genotypes, which indicates antixenosis to feeding. In the no‐choice test, all genotypes (except for 'IAPAR 57') expressed moderate levels of antixenosis to feeding against C. includens larvae.


Introduction
Brazil is considered to be the largest producer and consumer of common bean worldwide (Cabral et al., 2011).Bean production in Brazil in 2013/2014 was 3,511.1 million tons (Acompanhamento…, 2014).
Although Brazil is a great producer of this legume, the attack of various insect pests on bean plants compromises productivity in the field (Wendt & Carvalho, 2006).Among these insect pests, the soybean looper, Chrysodeixis includens (Walker) (Lepidoptera: Noctuidae), has been considered to be of growing importance for bean cultivation, due to damage caused during the last harvests (Baldin et al., 2014).Attack from C. includens causes characteristic holes on leaf surface, mainly in the more developed leaves.In the fourth instar, caterpillars can pierce the leaves, consuming large areas, keeping only main veins intact, which confers a characteristic lacy aspect to the attacked leaves (Sinclair et al., 1997).In soybean plants, one individual of C. includens can consume between 80 and 200 cm 2 (Bueno et al., 2011).
For soybean crop, the control of infestations of C. includens has been performed mainly through spraying with synthetic insecticides.However, the understanding that this practice contributes to the agroecosystem unbalance (Bernardi et al., 2012) encouraged studies on complementary methods of control which offer efficiency and, at the same time, are less destructive to the environment.In this sense, the use of resistant cultivars stands out as a valuable strategy against insect pests (Smith & Clement, 2012).In many cases, varietal resistance has shown significant efficiency, reducing pest populations to rates below the level of economic injury, consequently reducing the production costs (Smith, 2005).Plant resistance can be expressed through three different mechanisms: antixenosis, antibiosis, and tolerance (Smith & Clement, 2012).Antixenosis is the resistance mechanism employed by plants to deter or reduce colonization by insects.In general, insects orient themselves on plants for feeding, oviposition sites, or shelter.However, due to specific characteristics, attacked plants may not be utilizable and may inhibit the insects.In some cases, the antixenotic characteristics of a plant do not allow insects to colonize it.Sometimes the antixenosis mechanism is so effective that insects starve and die (Smith, 2005).
Although plant resistance is an important strategy for integrated pest management (IPM), until present, no study on common bean genotypes resistant to C. includens could be find.However, considering the importance of this crop to Brazilian population, and the high damage potential of soybean looper caterpillar to the culture, the selection of more resistant genotypes to insect attack becomes highly desirable.The results may help producers in the choice of less susceptible genotypes to insect attack.Additionally, these data may also be useful for improvement programs of common bean focusing on resistance to defoliating pests.
The objective of this work was to evaluate 59 bean genotypes for resistance to Chrysodeixis includens.

Materials and Methods
The work was developed in the Laboratório de Resistência de Plantas a Insetos e Plantas Inseticidas (Larespi) and in greenhouses of the Departamento de Proteção de Plantas, Universidade Estadual Paulista (Unesp), Botucatu, SP, Brazil, from 2012 to 2013.In the laboratory (26±2ºC, 65±10% RH, 14 h light: 10 h dark photophase), attractiveness and consumption assays were carried out with C. includens larvae.In greenhouse, the assays were conducted with adults.The evaluated common bean genotypes (59) (Table 1) were supplied by Instituto Agronômico (IAC, Campinas, SP, Brazil), and are part of the active germplasm bank of the institution.
A colony of C. includens was initiated from eggs that came from a laboratory colony maintained by DuPont (Paulínia, SP, Brazil).This colony was maintained in an artificial diet, according to the methodology proposed by Parra et al. (2009), with adaptations.
The plants used in the experiments were maintained in a greenhouse (3 m long x 2 m wide x 2 m high), and they were grown in polyethylene plastic pots (2.5 L) filled with a substrate of a mixture of soil (Oxisol), washed coarse sand, and organic matter (corral manure) at 1:1:1.Plant fertilization was performed according to Ambrosano et al. (1997).Other necessary cultural practices were also followed (irrigation and thinning).All assays were performed with plants in the V 4 phenological stage (Fernández et al., 1986).
A preliminary free-choice test was carried out in a greenhouse with the 59 bean genotypes, which were divided into three groups according to the color of their leaves (Table 1).Leaf color was determined according to the scale of minimum bean morphological descriptors items suggested by the Ministério da Agricultura Pecuária e Abastecimento (MAPA), used for the protection of plant varieties (Brasil, 2014).The genotypes 'Pérola' and 'IAC Formoso' were included as controls in all three groups, as these cultivars are susceptible to defoliators.
In this assay, plants were randomly placed in a circle inside metal cages (2.5x3.0x2.5 m), covered at the top with plastic sheeting and shade cloth (30%), and protected on the sides with white anti-aphid screens.
In free-choice tests, oviposition was evaluated in cages identical to those used during the preliminary test.In this case, plants from the 14 genotypes formed only one group.The methodology used for the evaluation of oviposition preference was the same described for the previous experiment, releasing two pairs of C. includens per genotype.A randomized block design was carried out with 14 treatments (genotypes) and seven replicates (one cage represented one replicate).
For the no-choice test, individual cages (30 cm diameter x 70 cm high) covered with organdy fabric were used around potted plants (V 4 stage) of the different genotypes.Inside the cages, two pairs of C. includens per genotype were released.The infestations were maintained for four days until the evaluation, according to the methodology described in the previous experiments (Campos et al., 2010).A completely randomized design was performed with 14 treatments (genotypes) and seven replicates (each individualized plant represented one replicate).
Plant attractiveness was evaluated in the laboratory (under already described conditions), in free-choice conditions for third-instar C. includens larvae, with leaf discs of the 14 selected genotypes.Therefore, two third-instar larvae per genotype were released inside a metallic circular arena (50 cm diameter x 4 cm height) containing the leaf disks (3.90 cm 2 ).Once larvae reached the third instar, they were subjected to a 12-hour-fast before the installation of the assay.Third-instar larvae were chosen for being less sensitive than first-and second-instar ones, and also for consuming significant amount of leaf area (Schlick- Souza et al., 2011).The number of larvae in the leaf disks of each genotype was counted at 15, 30, and 60 min, at 2, 3, 6 and 12 hours after release.
At the end of evaluations, the attractiveness index (AI) was calculated by using the equation AI = 2T/ (T + P) (Lin et al., 1990).'IAC Formoso' was also adopted as standard for the classification of genotypes.A randomized complete block design was used with 14 treatments (genotypes) and ten replicates (metallic arenas).
Leaf disk consumption by C. includens third-instar larvae was evaluated in a free-choice test.Thus, two third-instar larvae per genotype were released inside metallic circular arenas (the same used for the attractiveness), containing the leaf disks (3.90 cm 2 ) of the 14 genotypes.The arenas were observed until some leaf disks of the different genotypes in one of these areas were consumed by nearly 90%, when the test has ended.The remaining disks were taken in identifiable paper bags, which were then placed in an oven at 60 °C, for one day, in order to obtain the dry mass (Boiça Junior et al., 2013).Together with the disks remaining at the end of the experiments, another ten discs (controls) of different genotypes were also dried and served as an aliquot for determining the consumed dry mass (Boiça Junior et al., 2013).Dry mass of the leaf disks was weighted in an analytical balance with 0.0001 g precision, model Marte AY 220 (Shimadzu Corporation, Kyoto, Japan).A randomized complete block design was used, with 14 treatments (genotypes) and ten replicates (metallic arenas).
Leaf consumption was also evaluated in no-choice tests.Therefore, two third-instar larvae were individually placed on leaf disks of the different genotypes inside Petri dishes (8x2 cm).Evaluations followed the same criteria of the free-choice test.A completely randomized design was carried out with 14 treatments (genotypes) and ten replicates (Petri dishes).Before initializing the two types of test, third-instar larvae were subjected to a 12-hour-fast.
Data were subjected to analysis of variance by F-test, and checked for normality using the Shapiro-Wilk test, and for homogeneity using the Levene test.When treatment effects were significant at 5% probability, means were separated using Scott-Knott and Fisher Least Significant Difference (LSD) tests, using the procedure Proc Mixed of SAS software (SAS Institute Inc., Cary, NC, USA).
Analyzing the oviposition behavior of C. includens on genotypes of different groups (variable coloration of the leaves), it was observed that adult insects showed lower oviposition preference (plant's greater deterrence) for genotypes with dark green leaves (group 1) and medium (group 2) in relation to those with lighter green leaves (group 3).Although no studies were found on the host selection by adults of C. inludens on common bean, differently from the present study, Mercader et al. ( 2007) reported a higher oviposition preference by Papilio glaucus L. (Lepidoptera: Papilionidae) on dark green leaves of Fraxinus americana L., Liriodendron tulipifera L., and Prunus serotina Ehrh., than for light-green or yellowish-green leaves.Color intensity of the vegetable substrate is one of the extrinsic factors positively or negatively affecting the selection of host plants by phytophagous insects (Mercader & Scriber, 2007).This could explain the behavior observed for C. includens in the present work.
In addition, it was cannot disregard the potential action of undesirable volatiles (allomones) in different genotypes, which may affect host selection by insects.According to Cunningham & Zalucki (2014), the interaction between visual and olfactory cues may be important in evoking oviposition responses.In the present work, chemical characterization of volatiles present in the plants of the different genotypes was not carried out.However, Bruce et al. (2005) and Benda et al. (2011) reported that several species of Lepidoptera are able to recognize specific kairomones in preferred hosts.A highly polyphagous insect may be capable of classifying odors in plants, such that the majority of odors are responded to as good, poor, and non-hosts (Cunningham, 2012).Although chemical analysis of volatiles has not been the object of this study, in the future, the characterization of these less infested genotypes would be desirable, in order to identify the mechanisms involved.
Kidd & Orr ( 2001) evaluated the oviposition of C. includens on soybean and kudzu (Pueraria montana Lour.), in free-choice and no-choice tests, and observed a higher number of eggs on soybean leaves (68.5 and 570.9, respectively), compared to kudzu leaves (44.0 and 325.7, respectively).Independent of the host, the authors reported a higher number of eggs obtained in no-choice tests.In the present research, this kind of behavior was also observed for C. includens on 'BRS Horizonte ' (316.86),'IAPAR 57' (263.57),and 'Pérola' (209.43), which showed higher numbers of eggs in the no-choice tests.In free-choice test, adult lepidopterans do not respond to all host plants equally, and the females may show greater preference for certain plants; in no-choice test, when the insect is maintained isolated on a host plant, it may also attack the plant that would normally not be preferred, causing considerable damage (Cunningham & Zalucki, 2014).
Based on the average attractiveness index of third-instar larvae of C. includens, the genotypes H96A102-1-1-1-52, 'Flor de Mayo' and Arcelina 4 were classified as repellents in comparison to 'IAC Formoso', which is the susceptible commercial standard (Figure 2).'IAPAR 57','IAPAR 44','Pérola','Rubi','IAC Harmonia','IAC Jabola','BRS Horizonte',and 'IAC Boreal' showed to be neutral, and the genotypes Arcelina 1 and 'IAPAR 81' were considered attractive, in comparison to the susceptible standard.The lower attractiveness of leaf disks of H96A102-1-1-1-52, 'Flor de Mayo' and Arcelina 4 indicates that these genotypes showed secondary repellent substances generally associated with antixenosis.Although there are no reports about repellency or feeding deterrence associated with these genotypes, the repellent effect of certain plants on insects may occur, due to the volatilization of chemical substances from the leaves (War et al., 2011), which negatively affects insect preference.
The lowest insect attraction to leaf disks of some genotypes was expected to correspond to the lowest leaf consumption of those disks.However, such a situation did not occur in all materials.For instance, 'Flor de Mayo' was lightly attractive (Figure 2), but revealed a high consumption index.This suggests that the biggest attraction of the insect by the host plant does not necessarily reveal to be the appropriate food for larval development (Von Mérey et al., 2013).As reported Table 2. Mean±standard error number of eggs of Chrysodeixis includens on leaves of bean genotypes, in free and no-choice tests, in greenhouse (1) .by Von Mérey et al. (2013) and Cunningham (2012), volatile organic compounds emitted by plants were attractive to larvae, but the exposure to volatiles decreased the growth rate of caterpillars, possibly due to the lower leaf consumption.This may explain the divergences found between very attractive materials and few consumed ones or the inverse of this.
Based on the results for the leaf consumption in no-choice test (Figure 3), all genotypes differed from 'IAPAR 57', which was the most consumed plant by larvae.Lower consumption of the materials is possibly due to the presence of deterrent substances to feeding.Through their sensory organ, as gustatory semiochemicals, the phytophagous insects have the ability to select a host plant which, in turn, may show an attractive or deterrent effect on insect feeding (Bruce et al., 2005).Souza et al. (2012) found lower leaf consumption by Lepidoptera larvae, possibly due to morphological and chemical factors intrinsic to them which also provide degrees of resistance to insects.
Evaluating consumption by larvae of C. includens on two soybean genotypes, Reynolds et al. (1984) found a growth reduction in insects feeding on PI 227687.These authors reported that this fact is related to the presence of deterrent substances in the plant or to the plant failing to feeding stimuli.Hwang et al. (2008) affirm that insect growth is directly correlated with nutrient input, once Lepidoptera larvae fed high nutrient food showed faster growth rates than those fed nutrient-poor food.
Because of the damage potential that C. includens shows for bean cultivation, the results obtained with the evaluated materials, in the present study, can help genetic improvement programs focusing on this pest management.In the future, more detailed assessments as to the chemical, physical, and morphological aspects of leaves of bean genotypes will be addressed aiming to identify the main causes associated with antixenosis, showed by some of the evaluated genotypes in this work.

Figure 1 .
Figure 1.Oviposition preference index (OPI), and classification of bean genotypes -group 1, dark leaves (A); group 2, medium leaves (B); and group 3, light leaves (C) -, as a function of oviposition of Chrysodeixis includens, in free choice test, in greenhouse.SE, standard error.

Figure 3 .
Figure 3. Mean±standard error of leaf consumption by third instar larvae of Chrysodeixis includens, in free and no-choice tests, under laboratory conditions (26±2ºC, 65±10% RH, 14-hour photophase).Means followed by the same letter, in the columns of the same color, are not significantly different, by Scott Knott test at 5% probability.

Table 1 .
Name, access number at BAG IAC, and characteristics (genealogy or source) of bean genotypes of dark, medium, and light color leaf.Pots containing the plants were spaced approximately 15 cm apart, in order to avoid contact between plants.Each cage was considered a replicate, following a randomized block design, with four replicates for each group.The cages were infested with two pairs of C. includens per genotype.