Leaf-mediated response of soybean genotypes to infestation by whitefly, Bemisia tabaci (Gennadius)

Abstract

The whitefly, Bemisia tabaci (Gennadius) is documented as a major pest on soybean. It was reported that whitefly response towards its hosts and their cultivars varies, and is mediated through various host-related factors. Considering the significance of leaf morphological characteristics in influencing the host–whitefly responses, the present investigation was conducted in screen-house conditions to study the prevailing variations in leaf morphological characteristics of soybean genotypes and their role in governing the adult whitefly attractiveness and oviposition preference. In the multiple-choice test, the whitefly population (eggs, nymphs and adults) was found to be lowest in moderately resistant genotypes (SL 1028 and SL 1074) compared to highly susceptible (DS 3105) and susceptible genotypes (SL 688, SL 958 and SL 1113). The foliar trichomes were characterized using scanning electron microscopy and leaf area, leaf lamina thickness and leaf shape data were acquired using standard procedures. To determine the factors involved in the resistance/susceptible responses towards whitefly, Pearson correlation was applied between the morphological characteristics and the whitefly population. The results show that the leaf area, trichome density, trichome length and trichome angle showed a significant positive correlation with the whitefly population, whereas leaf lamina thickness was negatively correlated. Thus, for developing whitefly-resistant germplasm, breeders should choose genotypes having narrow and thick leaves with sparse, short and flat trichomes.

Introduction

Soybean is an important oil seed crop being cultivated in five agroclimatic zones of India. It is a unique crop among other legumes because of its nutritional and ecological importance such as 40% protein, 20% oil, fixes the soil nitrogen and requires less water for cultivation (Agarwal et al., 2013). In North-Western India, the soybean was recommended for cultivation to gradually decrease the predominant rice–wheat cropping system and for availing the numerous advantages (Ram et al., 2013). In several Indian states including Punjab, mung yellow vein mosaic virus disease was reported as one of the major biotic threats to soybean cultivation and is transmitted by the whitefly, Bemisia tabaci (Hemiptera: Aleyrodidae) (Rani et al., 2016). Apart from imposing indirect damage, the whitefly was documented to be a serious pest on kharif season (July–October) pulses such as mung bean, black gram and soybean and incurs a significant decrease in crop yield (Chhabra and Kooner, 1998). In Indonesia, 80% yield losses were documented in some susceptible soybean cultivars with the damage imposed by whitefly (Inayati and Marwoto, 2012). Globally, whitefly outbreaks were documented in the major soybean-producing countries such as Brazil, USA, Indonesia and Turkey (Gulluoglu et al., 2010; Vieira et al., 2011; Sulistyo and Inayati, 2016). The feeding by nymphs and adult whiteflies removes the plant sap in larger amounts and influences the physiological process of the host plants with the various toxins and effectors injected into the sap. The indirect damage also occurs with the development of sooty mould (Capnodium spp.) on the honeydew excreted by whitefly, which further decreases the photosynthesis efficiency (Firdaus, 2012). Although many farmers are primarily dependent on insecticidal usage to manage the whitefly menace, various associated disadvantages that promote sustainable whitefly management such as use of resistant variety were suggested to be one of the best alternatives (do Valle et al., 2012; Firdaus, 2012; Sulistyo and Inayati, 2016; Baldin et al., 2017). Moreover, the development of insect resistance in plants also mitigates the insect vectors and influences the virus transmission rates, by preventing feeding, development and population size (do Valle et al., 2012; Firdaus, 2012).

In the evolution process, to protect from the damage imposed by insect herbivores, various adaptations developed by the plants were conserved in their morphological, biochemical, physiological, physio-chemical and molecular traits (Stout, 2014). In association with the prevailing plant chemical composition, the morphological-based factors that affect the feeding, oviposition and host plant selection process of the insect pests can be considered potential resistance factors and were named constitutive resistance characters (Fraenkel, 1959; Stout, 2014; Baldin et al., 2017). The differences prevailed in the morphological characteristics of the host plants influence the behaviour, feeding, shelter, growth, development and population dynamics of insects, accordingly influencing host resistance/susceptibility response against insect herbivores (Firdaus, 2012; Hasanuzzaman et al., 2016; Baldin et al., 2017; Lutfi et al., 2019; Zhou et al., 2020). To determine the insect-feeding responses in genotypes or varieties, greater importance was given to prevailing variation in the trichome density, and it is often used as a basis in plant breeding programmes aimed at insect resistance (Lam and Pedigo, 2001; Zhou et al., 2020). In the host acceptance phase, foliar trichomes are the first structural traits that get in contact with the insect herbivores (Lambert et al., 1995; Peter et al., 1995; Zhou et al., 2020). The characteristics of trichomes such as density, length, angle of insertion, arrangement and types in various crops were reported to influence the whitefly population (Butter and Vir, 1989; Firdaus et al., 2011, 2012; Taggar and Gill, 2012; Hasanuzzaman et al., 2016). The trichome density was reported to be responsible for the differences documented among the cotton cultivars towards populations of whitefly (Butler et al., 1991), whereas in cucurbits, trichome length and spatial arrangement have influenced the population rather than density (Kishaba et al., 1992). Among the plant traits imparting resistance, leaf thickness is recognized as important as it can influence the leaf tissues penetration process of sucking insects and increases the mandibular abrase in biting and chewing herbivores (Schoonhoven et al., 2005).

Research studies on soybean to impart resistance against whitefly were conducted in Brazil (Vieira et al., 2011; do Valle et al., 2012; Baldin et al., 2017), Indonesia (Inayati and Marwoto, 2012; Sulistyo and Inayati, 2016), Turkey (Gulluoglu et al., 2010) and the USA (Lambert et al., 1995). Whereas in the Indian scenario, among the major ongoing research works being conducted on soybean germplasm improvement programmes, the incorporation of resistance against the mung bean yellow vein mosaic virus is prominent. But relevant work pertaining to mitigating the whitefly direct damage is lacking. Therefore, the main objective of this study is to understand the possible interactions between whitefly and soybean dependent on the foliar-based morphological characteristics. Further, to determine the role of morphological characteristics in imparting resistance/susceptibility to B. tabaci, correlation and regression studies between the morphological traits and the whitefly population were undertaken.

Materials and methods

Plant materials

The experiments were conducted in screen-house conditions (kharif 2018 and 2019) of Entomological Research Farm, Punjab Agricultural University (PAU), Ludhiana. Genotypes ‘DS 3105’, ‘SL 688’, ‘SL 1113’, ‘PS 1347’, ‘PS 1572’, ‘SL 958’, ‘SL 1028’ and ‘SL 1074’ were selected. The seed materials were procured from the principal soybean breeder, pulse section, Department of Genetics and Plant Breeding, PAU. Each genotype was sown with a minimum of five seeds in five earthen pots (10 × 15 cm) with three replications per treatment. The genotypes were maintained inside the screen house conditions and were grown based on the soybean package of practices given by PAU (Anonymous, 2018).

Screening of resistance

In preliminary studies, the screening technique with the genotype ‘SL 958’ was conducted to determine all the damage grades that can incur at the earliest growth stage and with the lowest whitefly population (standardized in another experiment, data not represented here). Accordingly, the genotype response towards the adult whiteflies was evaluated at the V₃ growth stage by releasing 125 adult whiteflies and were categorized into different groups using WRI developed by Taggar et al. (2013) (Supplementary Table 1).

Whitefly culture and inoculation

Brinjal (Solanum melongela) and cotton (Gossypium herbaceum) plants were used for the whitefly culture multiplication. These host plants were grown in earthen pots (15 × 20 cm) and maintained inside the screen house conditions to keep them free from other insect pests. Nutrient and irrigation requirements were supplied at appropriate times to support the good growth of the host plants and were placed in the whitefly culture at their active vegetative stage. Initially, to begin the insect culture, using the aspirator, numerous adult whiteflies (>3000 pairs) were collected from the surrounding cotton fields and were released over the host plants. The whitefly population was allowed to develop on these hosts for some generations till the desired number of insects were available for conducting the experiment and newer host plants were placed to support the continuous multiplication when plants were found to succumb to injury.

Multiple choice test

Once genotypes have attained the V₃ growth stage, they were arranged in a completely randomized design. Adult whiteflies (125 pairs/plants) were aspirated and released at random places to allow free movement amongst the genotypes. To record the number of eggs, nymphs, red-eyed nymphs and adults in each treatment, three plants were randomly selected and tagged. The observations were recorded at weekly intervals for a total period of 5 weeks from the fully developed trifoliate leaves representing the upper, middle and lower canopies. The observations on adult whiteflies settled per trifoliate were visually counted by gently turning the leaves and the number of eggs (1 cm² region), nymphs and red-eyed nymphs (2 cm² area) were recorded using a binocular stereo microscope (40×) and magnifying lens (10×). The nymphal and adult population represented was pooled data of two seasons (kharif 2018 and 2019), as seasonal differences were non-significant. The average (N = 3) was calculated for the data recorded on the adult and nymphal population, as there were no significant differences observed among the genotypes with respect to canopy.

Leaf-mediated morphological traits

The morphological traits in soybean genotypes were studied once the genotypes have attained the 40 days and all the observations were recorded by detaching the fully developed trifoliate leaves from the upper, middle and lower canopies. The leaf thickness was measured by placing the leaves in between the digital micrometre, leaf area was determined by the graph paper method, and leaf shapes were grouped by literature-based comparisons (Taggar and Gill, 2012). The foliar trichome characteristics (density, length and insertion angle) were imaged with the scanning electron microscope (SEM) (Model JSM 6100 JEOL) facility at Sophisticated Analytical Instrumentation Facility, Panjab University, Chandigarh. All the SEM observations were recorded as per the standardized procedure (Bozzola and Russell, 1999). The leaf-mediated morphological traits represented were pooled data of two seasons (kharif 2018 and 2019), as seasonal differences were non-significant.

Statistical analysis

The data on whitefly population and trichomes were transformed to √n + 1, prior to performing the statistical analysis. Experimental data were analysed in one-way ANOVA to compare the significance difference among the treatments (P < 0.05). Tukey's HSD test (P < 0.05) was applied to compare the differences in treatment means. Correlation and stepwise regression were conducted to ascertain the relation between the morphological characters and whitefly population (egg, nymphs and adults). All the statistical analyses were performed using IBM SPSS version 25.

Results

Adult whiteflies preference towards genotypes

Adult whitefly preference responses towards the genotypes showed significant differences (Table 1). Based on the mean observations recorded over 5 weeks, significantly lowest number of adult whiteflies were recorded in moderately resistant genotypes (SL 1074 and SL 1028) and was highest on genotype DS 3105 (highly susceptible genotype), followed by genotypes SL 688, SL 958 and SL 1113 (susceptible genotypes) and PS 1572 and PS 1347 (moderately susceptible genotypes). The ovipositing female response was significantly more towards the genotype DS 3105, as depicted by significantly a greater number of eggs, compared to other genotypes and the lowest egg number was recorded in genotypes SL 1074 and SL 1028, indicating that these were least preferred, than other genotypes. The nymphs and red-eyed nymphal population were significantly more on genotype DS 3105, followed by genotypes SL 688, SL 958 and SL 1113 (S), implying that population build-up of whitefly on the susceptible genotypes is more compared to resistant genotypes.

Table 1. B. tabaci preference and nymphal population build-up on soybean genotypes in a multiple-choice test under screen-house conditions

Numbers followed by the same letter in the same column are not significantly different (P ⩽ 0.05; Tukey's HSD test).

*Mean of three replications recorded over a period of 5 weeks; mean ± S.E.M (standard error of mean).

^# Combined mean of three canopies (N = 3) (upper, middle and lower) recorded during July 2018 and 2019.

^@ indicates middle canopy.

Leaf-mediated morphological traits

Leaf shape

The leaf shape was broadly categorized into ovate and lanceolate types (Supplementary Table 1). The lanceolate leaf genotypes were further classified into lanceolate and broadly lanceolate, based on the observable variations. Genotypes DS 3105, SL 688, SL 958 and SL 1113 had ovate leaf shapes, with leaves that were rounded and broader at the basal end, resembling a lance head shape. The lanceolate leaf shape was observed in genotypes PS 1347, PS 1572 and SL 1028, in which leaf area was reduced at the apical region with a lance-like appearance and comparatively broader at the basal part. Genotype SL 1074, grouped under a broadly lanceolate leaf shape showed to have a sharply tapered apex region and a slight border at the leaf base region.

Leaf area

The leaf area of genotypes showed significant variations (Table 2). The leaf area of various soybean genotypes differed significantly across the three canopies (upper, middle and lower), as indicated in Table 3. The mean leaf area of the different genotypes ranged from 34.41 to 61.56 cm² across the canopies. These findings highlight the diversity in leaf area among the soybean genotypes studied. The genotype PS 1347 recorded a significantly lower mean leaf surface area (34.41 cm²), which was at par with the genotype PS 1572 (35.95 cm²). The genotypes SL 688, SL 958 and SL 1113 recorded the leaf area of 56.79, 46.33 and 47.66 cm², respectively, and were at par with each other. The genotypes SL 1028 and SL 1074 recorded 40.52 and 42.18 cm² area and were at par with each other, while genotype DS 3105 recorded significantly the highest area (61.56 cm²).

Table 2. Leaf area and thickness of leaf lamina in different genotypes of soybean

Numbers followed by the same letter in the same column are not significantly different (P ⩽ 0.05; Tukey's HSD test).

*Mean of three replications.

^# Combined mean recorded during July 2018 and 2019.

Table 3. Leaf trichome characteristics of the different soybean genotypes

Numbers followed by the same letter in the same column are not significantly different (P ⩽ 0.05; Tukey's HSD test).

Figures in parentheses are √n + 1 transformed means.

*Mean of three replications.

#Combined mean recorded during July 2018 and 2019.

Leaf lamina thickness

The leaf thickness in the genotypes and amongst canopies (upper, middle and lower) varied significantly (Table 2). The mean leaf lamina thickness across the canopies ranged from 0.1667 to 0.2262 mm. The genotype DS 3105 recorded a significantly lower mean leaf lamina thickness (0.1667 mm), followed by SL 688 (0.1797 mm). The genotypes SL 958, SL 1113 and PS 1347, PS 1572 have recorded leaf lamina thickness of 0.1922, 0.1937 and 0.2051, 0.2071 mm, respectively, and were at par with each other. The genotypes SL 1028 and SL 1074 recorded leaf lamina thickness of 0.2211 and 0.2262, respectively, and were at par with each other. The mean leaf lamina thickness in the upper, middle and lower canopy was 0.1884, 0.1950 and 0.2135 mm, respectively and showed significant variations.

Trichome density

The trichome density (recorded per 6 mm²) significantly varied among genotypes and ranged from 6.22 to 16.54 trichomes (Table 3). The genotype SL 1028 had the lowest trichome density of 6.22 trichomes, which was at par with the SL 1074 (6.91 trichomes). The genotypes PS 1347, PS 1572 and SL 1113 recorded trichome density of 9.57, 10.62 and 10.67, respectively, and were at par with each other. Genotype SL 958 was observed to have the highest trichome density (16.54 trichomes) and was at par with the genotypes SL 688 (14.45 trichomes) and DS 3105 (13.74 trichomes).

Trichome length

The soybean genotype trichomes were straight, slender and non-glandular with straight or pointed tapering ends. The mean trichome length of genotypes varied from 197.56 to 378.67 μm and showed significant differences (Table 3). The genotype PS 1347 showed the shortest trichome length (197.56 μm), followed by SL 1074 (221.80 μm) and SL 1028 (260. 38 μm) and were statistically at par with each other. The genotypes PS 1572, SL 1113, SL 688 and SL 958 recorded trichome lengths ranging from 281.85 to 342.69 μm and the longest trichome length was recorded in genotype DS 3105 (378.67 μm).

Trichome angle

The trichome angle of genotypes showed significant differences, ranging from 45° to 76° (Table 3). Genotype PS 1347 has slightly erect trichomes with a trichome angle of 45°. The genotypes SL 1028 and SL 1074 recorded an angle of 56° and 56.3° and were at par. The genotypes PS 1572, SL 958, SL 1113 and SL 688 recorded trichome angles of 61.67°, 65.33°, 67.33° and 68.33°, respectively, and were at par with each other. In genotype DS 3105, trichome erectness was significantly higher (76°) than in other genotypes.

Correlation and regression studies

The data pertaining to the correlation between the whitefly population and leaf morphological traits were presented in Table 4. The leaf surface area showed a significant and positive correlation with eggs, nymphs and adults, demonstrating that genotypes processing more leaf surface area can support a larger number of whitefly population (adults, eggs and nymphs). However, the leaf thickness displayed a significant and negative correlation with the whitefly population; indicating that thicker leaves are not favourable towards them. The trichome density, length and angle have showed a significant positive correlation with the whitefly population, suggesting that trichomes favour the adult whiteflies preference and support the egg and nymphal population.

Table 4. Correlation between the morphological leaf characteristics of soybean genotypes and B. tabaci population

r, correlation coefficient; R ², coefficient of determination.

*Significant at 5% level of significance; **significant at 1% level of significance.

The whitefly population (eggs, nymphs and adults) were regressed on foliar characteristics by applying the stepwise multiple linear regression to understand the role of leaf-mediated morphological traits in providing the resistance/susceptibility response against the whitefly. The regression equations and coefficient of determination (R ²) were presented below

$$\begin{align}{\rm Y}( {{\rm egg}} ) {\rm} &= 39 .148 + 0 .182( {{\rm LA}} ) \hbox{-} 312 .67( {{\rm LLT}} ) {\rm} + 0{\rm .508\ }( {{\rm TD}} ) {\rm} \cr &\quad + 0{\rm .046\ }( {{\rm TL}} ) {\rm} + 0{\rm .467\ }( {{\rm TA}} ) {\rm , \;\ }R^ 2{\rm} \cr & = 0 .967\end{align}$$

$$\begin{align}{\rm Y}( {{\rm nymph}} ) {\rm} & = 19 .44 + 0 .195( {{\rm LA}} ) \hbox{-} 128 .03( {{\rm LLT}} ) {\rm} + 0 .475( {{\rm TD}} ) {\rm} \cr & \quad + 0{\rm .024( }( {{\rm TL}} ) {\rm} + 0 .197( {{\rm T\ A}} ) {\rm , \;\ }R^ 2{\rm} \cr & = 0 .991\end{align}$$

$$\begin{align}{\rm Y}( {{\rm adult}} ) {\rm} & = 37 .65 + 0 .085( {{\rm LA}} ) \hbox{-} 259 .59( {{\rm LLT}} ) {\rm} + 0 .438( {{\rm TD}} ) {\rm} \cr & \quad + 0 .028( {{\rm TL}} ) {\rm} + 0 .260( {{\rm TA}} ) {\rm , \;\ }R^ 2{\rm} \cr &= 0 .982\end{align}$$

Based on the multiple linear regression equations, leaf-mediated morphological traits have largely influenced the nymphs followed by adult whiteflies. The leaf area, leaf lamina thickness, trichome density, length and angle showed 99% and 98% variation (R ² values) to the whitefly nymphal and adult stages, respectively.

Discussion

Based on this 2-year study, the whitefly population (adults, eggs and nymphs) preference towards the genotypes displayed significant variations, and during the overall experimental period, the lowest population number was recorded in genotypes SL 1028 and SL 1074, suggesting that these genotypes were highly unattractive for whitefly colonization and oviposition purposes and were significantly distinguished from other genotypes. However, genotype DS 3105 was observed to be highly attractive to the whitefly, followed by SL 688, SL 958 and SL 1113. The suitable plant selection by insect herbivores involves complex processes and this depends on the olfactory, visual, tactile and gustatory cues of the host. But the infestation process begins after the host acceptance, which is determined through the initiation of host feeding (Schoonhoven et al., 2005). In whiteflies, the appropriate host selection is an ascertaining step because among the four nymphal instars, only the first instar nymphs are mobile and the remaining three instars moult from the appropriate site selected by the first instar for feeding, till the adult emerges. Therefore, suitable host selection and egg deposition at a favourable place determines the fitness and survivability of offspring (Van Lenteren and Noldus, 1990; Firdaus et al., 2011; Firdaus et al., 2012). Foliar characteristics such as colour, shape, area, thickness, trichomes, etc., were reported to influence the selection and infestation process of whiteflies (Berlinger, 1986; Butter and Vir, 1989; Firdaus, 2012; Taggar and Gill, 2012; Hasanuzzaman et al., 2016; Lutfi et al., 2019). Thus, it could be postulated that ovipositing female whiteflies and adult whiteflies settled for feeding have preferred the host plants of highly susceptible genotypes over resistant genotypes possibly after determining their suitability. This hypothesis was supported by previously conducted studies in which a significant positive correlation between adult attractiveness and oviposition preference was documented (Silva et al., 2008; do Valle et al., 2012).

In the present study, susceptible genotypes had larger leaf surface area and thin leaf lamina compared to moderately resistant genotypes, indicating the suitability of these two characteristics towards whitefly. The positive correlative response between the whitefly population and the leaf area supports the idea that the increased area might have allowed the numerous adult whiteflies to settle for feeding and oviposition purposes, and thus supported the greater number of whitefly populations. This hypothesis is supported by cotton cultivars where having narrow leaf reported to offer more resistance to B. tabaci, than broadleaf varieties (Butler et al., 1991). Leaf thickness in soybean was reported to provide antixenosis and influences the resistance responses towards the whitefly (Sulistyo and Inayati, 2016). The leaf thickness affects the whitefly probing process and influences the nutrient-acquiring process from the leaf (Lutfi et al., 2019). Likewise, in black pepper (Firdaus et al., 2011) and soybean (Sulistyo and Inayati, 2016), genotypes resistant to whitefly were reported to have thicker leaves. The whitefly population response to thickness reportedly showed variations in some crops; for instance, a significant positive correlation between the whitefly population and leaf lamina thickness was documented in cotton (Butter and Vir, 1989), black gram (Taggar and Gill, 2012) and brinjal (Hasanuzzaman et al., 2016).

In several reviews, it was reported that leaf trichome characteristics function as a ‘soft weapon’ and potential resistant trait in influencing the behaviour and performance of the insect pests on their host plants; however, they influence the density of insect pests both in positive and negative ways, i.e. attracting or repelling them (reviewed in Dalin et al., 2008). The trichomes of soybean are simple, non-glandular in nature measuring an average length of 1–3 mm and are either curly or straight (Peter et al., 1995). Foliar trichomes provide resistance to infestation either with the toxic chemical compound secretion, as seen in glandular trichomes of tomatoes; however, the resistance mechanism offered with non-glandular trichomes varies (Firdaus et al., 2012). Banded-wing whitefly, Trialeurodes abutilonea (Haldeman) (Hemiptera: Aleyrodidae) and Bemisia argentifolii (Bellows and Perring) (Hemiptera: Aleyrodidae) populations were documented to be higher in soybean genotypes having straight or erect trichomes, as compared with flat trichomes (Lambert et al., 1995). The positive responses with the non-exude and dense trichomes create favourable microclimatic conditions (high humidity, mild temperature), which supports the oviposition and nymphal development (Berlinger, 1986). Soybean genotypes having denser and long-length trichomes are highly preferred for whitefly oviposition than other types (Sulistyo and Inayati, 2006; do Valle et al., 2012; Lutfi et al., 2019). Similar responses were recorded in common beans (Silva et al., 2019), brinjal (Hasanuzzaman et al., 2016) and pepper (Firdaus et al., 2011). Denser trichomes in soybean genotypes are highly favourable to adult whiteflies as they prevent them from being blown away by the winds (Vieira et al., 2011). Higher trichome density might also act as an oviposition stimulant (Silva et al., 2019) and provide thigmotactic stimuli, which are essential for feeding and oviposition in B. tabaci (Lambert et al., 1995). In crops like common bean (Silva et al., 2019), brinjal and pepper (Firdaus et al., 2012), higher trichome density showed a positive influence on the whitefly oviposition. However, in brinjal (Ayyasamy and Baskaran, 2005) and black gram (Taggar and Gill, 2012), a negative correlation between the trichome density and whitefly oviposition was reported; these differences might be due to the involvement of the other host factors, which could affect the overall effect of trichome density on whitefly oviposition.

In the current study, it was observed that trichome length was more in susceptible genotypes than in resistance genotypes. These results are in agreement with Butter and Vir (1989) and Taggar and Gill (2012) observations, who have also documented similar results in cotton and black gram genotypes having resistant responses to whitefly. The basal portion of trichomes acts as a suitable site for inserting the egg and encourages the B. tabaci oviposition (Berlinger, 1986; Butter and Vir, 1989; Vieira et al., 2011) and this response could be an evolutionary response to gain protection from natural enemies (Inbar and Gerling, 2008). Further, in this study, trichome insertion angle was more in susceptible genotypes than resistant genotypes. This was supported by previously conducted documented observations in which oviposition response of whitefly was observed to be more in soybean genotypes having lower trichome erectness (Lambert et al., 1995) and inclined to erect trichomes (do Valle et al., 2012). In bean plants, the capture efficiency of insects was observed to be poor with the lower trichome angles, as insertion angles (<30°) are suggested to be ineffective in restricting the insect movement. The smaller angles reduce the contacting probability with the insects present on the leaf surface, but an increase in erectness significantly increases insect entrapment (Pillemer and Tingey, 1976).

The current investigation thus indicates that except leaf lamina thickness, other characteristics, i.e. leaf area and trichomes (density, length and angle), have positive effects on the whitefly population (as evident from regression equations). Thus, the current study results might be helpful for the plant breeders to select, incorporate and develop durable whitefly-resistant cultivars. Therefore, for screening the soybean breeding material against whitefly, major significance should be given to the adult and nymphal populations. Integrated plant resistance can be effectively used by understanding the factors supporting the resistance (inherent bases), and with the support of suitable traditional or modern breeding methods, necessary traits can be incorporated. In the areas where severe whitefly incidence occurs, preference should be given towards the varieties having narrow, thick and fewer trichome density with lower trichome length and angle for effective reduction of whiteflies.