Reaping the Potential of Wild Cajanus Species through Pre-Breeding for Improving Resistance to Pod Borer, Helicoverpa armigera, in Cultivated Pigeonpea (Cajanus cajan (L.) Millsp.)

Simple Summary Pigeonpea is an important legume crop that is severely affected by various insect pests, especially pod borer. Because there are low resistance levels to pod borers in the cultivated pigeonpea gene pool, it is necessary to introduce resistance-related traits from its wild relatives. In this study, we conducted a series of crosses to introduce traits related to pod borer resistance from two wild relatives of pigeonpea into two popular cultivated varieties. We generated populations from four different crosses and screened these populations for traits related to pod borer resistance: i.e., low levels of insect damage, high concentrations of insect-deterring compounds in the seeds, and the presence of trichomes on the leaves. The most promising lines were tested across seasons and locations. Ultimately, we identified 21 lines with excellent traits related to pod borer resistance. These lines will be useful for breeding new insect-resistant pigeonpea cultivars. The availability of such cultivars will reduce the use of pesticides to control pests on pigeonpea crop. Abstract Pod borer (Helicoverpa armigera) causes the highest yield losses in pigeonpea, followed by pod fly (Melanagromyza obtusa). High levels of resistance to pod borer are not available in the cultivated genepool. Several accessions of wild Cajanus species with strong resistance, and different resistance mechanisms (antixenosis and antibiosis) to pod borer have been identified. These accessions can be utilized to improve the pod borer resistance of cultivated pigeonpea. Using pod borer resistant Cajanus scarabaeoides and Cajanus acutifolius as pollen donors and popular pigeonpea varieties as recipients, pre-breeding populations were developed following simple- and complex-cross approaches. Preliminary evaluation of four backcross populations consisting of >2300 introgression lines (ILs) under un-sprayed field conditions resulted in identifying 156 ILs with low visual damage rating scores (5.0–6.0) and low pod borer damage (<50%). Precise re-screening of these ILs over different locations and years resulted in the identification of 21 ILs having improved resistance to pod borer. Because these ILs were derived from wild Cajanus species, they may contain different alleles for different resistance components to pod borer. Hence, these ILs are ready-to-use novel and diverse sources of pod borer resistance that can be utilized for improving the pod borer resistance of cultivated pigeonpea.


Introduction
Pigeonpea (Cajanus cajan (L.) Millsp.) is an often-cross-pollinated diploid (2n = 2x = 22) grain legume crop that is cultivated across an area of about 6.09 m ha with 5.01 m t production and 0.8 t ha −1 productivity in the tropical and subtropical regions of Asia and Africa [1]. Although this crop has multiple uses, it is primarily cultivated for its protein-rich seeds. India is the largest producer of pigeonpea, accounting for more than 93% of global production. It is commonly known as arhar, red gram, or tur in India, and is considered as the second most important pulse crop after chickpea. Insect pests continue to be the major biotic constraints to grain legume production, especially the pod borer complex, which causes estimated annual losses of more than USD 2 billion in the semi-arid tropics, despite the application of insecticides costing at least USD 500 million annually [2]. The most damaging pest to pigeonpea crops worldwide is pod borer, Helicoverpa armigera (Hübner), which causes the maximum yield losses (25-70%), followed by pod fly, Melanagromyza obtusa Malloch (10% losses), spotted borer, Maruca vitrata Fabricius (5-25% losses), and pod-sucking bug, Clavigralla gibbosa Spinola (10-30% losses) [3,4]. The frequent occurrence of pod borer often results in complete crop failure. The economic losses due to biotic factors have been estimated to be USD 8.48 billion. Pod borer alone may cause losses of more than USD 300 million annually, whereas yield losses caused by pod fly have been estimated at USD 256 million annually [5]. A wide range of insecticides are used to control pod borer under field conditions, but the indiscriminate use of pesticides has led to pesticide resistance, resurgence of pests, and secondary outbreaks of minor pests. The development of resistant cultivars by exploiting host plant resistance is the most effective and eco-friendly solution for the sustainable management of insect pests, including pod borers. Unfortunately, high levels of resistance to pod borer are not available in the cultivated genepool. Therefore, it is necessary to exploit new and diverse sources of resistance.
Crop wild relatives, especially Cajanus scarabaeoides, Cajanus acutifolius, and Cajanus platycarpus have been identified as potential sources of resistance to pod borer [6][7][8][9]. Different biochemical and morphological mechanisms conferring resistance to the pod borer complex, including antixenosis (oviposition non-preference by insects), antibiosis, and trichomes, have been identified in these wild Cajanus species. Therefore, they represent potential sources of resistance genes for introgression into the cultigens. The present investigation was carried out to exploit the potential of two wild Cajanus species, C. acutifolius, and C. scarabaeoides, to improve resistance to pod borer, H. armigera in the popular pigeonpea cultivars, ICPL 87119 (Asha) and ICP 8863 (Maruti). We used simple and complex cross approaches to generate introgression lines (ILs), and then identified those with high levels of resistance to the pod borer for further use in pigeonpea breeding programs.

Population Development
The pollen donors were C. acutifolius accession ICPW 001, which has high levels of antixenosis for ovipositing insects and antibiosis [8], and the C. scarabaeoides accession ICPW 281, which has a high density of C-type trichomes. The recipients were the two popular pigeonpea varieties ICPL 87119 and ICP 8863. Using these donors and recipients, four prebreeding populations were developed following simple and complex cross approaches at the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Patancheru, India. ICPL 87119, popularly known as 'Asha', is a high-yielding, medium-duration leading variety widely cultivated in India [10]. ICP 8863, also known as 'Maruti' [11], is a medium-duration high-yielding pigeonpea variety resistant to Fusarium wilt.
The breeding scheme used to generate the pre-breeding populations is given in Figure 1. The pigeonpea varieties ICPL 87119 and ICP 8863 were used as the female parent, and the wild species accessions were used as the pollen parent to generate F 1 hybrids. In each cross, true F 1 s were identified based on the morphological traits such as growth habit, leaf shape, stem color, flower color and streak pattern, days to first flowering, and pod traits, as well as by analysis of highly polymorphic simple sequence repeat (SSR) markers. For testing the hybridity of F 1 plants, DNA was extracted from the leaves collected after one month of germination and five highly polymorphic SSR markers (CcM0724, CcM0402, CcM0047, CcM0306, and CcM0494) were tested on each cross combination. All the true F 1 plants derived from the cross-involving C. acutifolius (ICP 8863 × ICPW 001) were semi-fertile (30-40% pollen fertility) to completely sterile (10-20% pollen fertility). In this cross, true F 1 plants were used as female parents and subsequently backcrossed with ICP 8863 to produce BC 1 F 1 seeds. In contrast, all the F 1 plants derived from the cross-involving C. scarabaeoides (ICPL 87119 × ICPW 281) were highly fertile (>80% pollen fertility) and were used as the pollen parent with ICPL 87119 to produce BC 1 F 1 seeds. This was performed to ensure identification of true BC 1 F 1 crosses from selfed F 2 s. The true BC 1 F 1 plants in both crosses were identified on the basis of morphological traits. The confirmed BC 1 F 1 plants were used for the second backcross with the respective cultivated parent to produce BC 2 F 1 seeds. True BC 2 F 1 plants were selfed twice to produce BC 2 F 3 populations in both crosses. The pre-breeding population derived from the cross ICP 8863 × ICPW 001 consisting of 1108 lines was designated as PP 1501; and that derived from the cross ICPL 87119 × ICPW 281 consisting of 288 lines was designated as PP 1505 (    With a view to combining different components governing pod borer resistance into a common genetic background, we also developed two backcross populations (four-way BC 1 F 2 ) derived from complex four-way F 1 crosses in two different genetic backgrounds, ICPL 87119 ((ICPL 87119 × ICPW 1) × (ICPL 87119 × ICPW 281)), and ICP 8863 ((ICP 8863 × ICPW 1) × (ICP 8863 × ICPW 281)). To generate a complex cross population in the ICPL 87119 background, sterile and semi-sterile F 1 plants from the ICPL 87119 × ICPW 1 cross were crossed with fertile F 1 plants from the ICPL 87119 × ICPW 281 cross. These four-way F 1 plants were further crossed with ICPL 87119 to generate BC 1 F 1 seeds, designated as 4-BC 1 F 1 . A similar approach was followed to generate 4-BC 1 F 1 seeds in the ICP 8863 background. True 4-BC 1 F 1 plants were selfed twice to produce 4-BC 1 F 3 populations from both crosses. The pre-breeding population derived from the cross ((ICPL 87119 × ICPW 1) × (ICPL 87119 × ICPW 281)) consisting of 533 ILs was designated as PP 1503; and that derived from the cross ((ICP 8863 × ICPW 1) × (ICP 8863 × ICPW 281)) consisting of 392 ILs was designated as PP 1504 (Table 1; Figure 1).

Evaluation of Pre-Breeding Populations for Pod Borer Complex
Four backcross pre-breeding populations, PP 1501, PP 1505, PP 1503, and PP 1504 consisting of 1108, 288, 533, and 392 ILs, respectively, were evaluated for pod borer damage under un-sprayed field conditions during the 2018 rainy season. In each population, the ILs along with susceptible (ICPL 87, ICPL 85010), moderately susceptible (ICPL 88039, ICP 7035) and moderately resistant checks (ICPL 87119, ICPL 332WR, and ENT 11) were planted in a single row plot (4 m long). At least one check was planted after every six ILs in each single row.
Visual damage to pods was scored using a rating on a scale of 1 to 9 (1.0: almost no damage, resistant; 9.0: severely damaged, highly susceptible) at the podding stage. The recovery resistance score was recorded at harvest stage on a scale of 1 to 9 (1.0: plants with <10% pod damage, showing good recovery from insect damage in the first flush, and the pod uniformly distributed throughout the plant; 9.0: plants with >80% pod damage, very poor recovery from insect damage, and <20% of the pods retained on the plant) [12]. A total of 156 ILs with recovery resistance scores of 5.0 to 6.0 were selected and their pod damage was estimated. In each of the 156 lines, 100 pods were randomly selected, and each pod was critically examined for damage caused by pod borer, pod fly, plume moth (Exelastis atmosa Wals.), and pod wasp (Tanaostigmodes cajaninae La Salle).
The selected ILs were re-evaluated using a randomized block design (RBD) in the black soil (Vertisols) precision fields under un-protected field conditions during the 2019 rainy season at Patancheru (17 • 51 N, 78 • 27 E; 545 m) and at Warangal (18 • 00 N, 79 • 59 E; 262 m) locations. Depending upon seed availability, 156 ILs at Patancheru and 136 ILs at Warangal along with susceptible (ICPL 87, ICPL 85010), moderately susceptible (ICPL 88039, and ICP 7035), and moderately resistant checks (ICPL 87119, ICPL 332WR, and ENT 11) were randomized and grown in three replications during the 2019 rainy season. The seeds were sown in triplicate on ridges 75 cm apart. There were four rows in each plot, and each row was 4 m long. The plants were thinned to 30 cm spacing between plants at 30 days after seedling emergence. Standard agronomic practices were followed for raising the crop, including application of basal fertilizer (nitrogen: phosphorus: potassium 100:60:40 kg ha −1 ) and top dressing (urea 50 ha −1 ). A fungicide (metalaxyl) spray (1.0 kg active ingredient (ai) ha −1 ) was applied to control Fusarium wilt. The same set of lines was evaluated for pod borer resistance by conducting pod assessment, pod bioassays, and analyses of biochemical traits under laboratory conditions. To confirm the results, the most promising 39 pod borer-tolerant lines were selected for re-evaluation during the 2020 rainy season following the same procedure. The pest susceptibility (%) of each line was calculated based on pod borer damage and pod borer complex damage in 39 lines over years using the formula derived from Abbott (1925) [13].
The pest susceptibility (%) was then converted to a 1 to 9 pest susceptibility rating (PSR) adopting the following scale [14]: Pods damaged by pod borer, H. armigera (big circular holes without larval exuviae on the pods); 3.
Pods damaged by pod fly (minute holes on pods); 4.
Pods damaged by pod wasp (minute holes at upper side of pod tip, empty pods, and pod length drastically shortened); 5.
Pods damaged by plume moth (two to three medium-sized circular holes on pods); 6.
The Maruca (spotted pod borer) damaged pods have small, darkened entry holes with frass-fecal matter and chewed remains of the pods around the entry holes. It also has a typical symptom with holes in pods at one end [15].
The numbers of healthy and damaged pods due to pod borer complex (pod borer, pod fly, pod wasp, and plume moth) were recorded and converted into percentage pod damage, as follows: % pod damage = ((number of damaged pods)/(total number of pods)) × 100.

Procedure for Recording the Larval Count
At podding stage, three plants were randomly selected, and tagged with distinct color labels in all the three replications. The observation was recorded on the number of H. armigera larvae at podding stage on each tagged plant in each replication.

Helicoverpa Armigera Culture
The neonates of H. armigera used in bioassays were obtained from a laboratory-reared culture at ICRISAT, Patancheru, India. The H. armigera larvae were reared individually in the laboratory on a chickpea-based artificial diet [16] under the following conditions: 27 ± 2 • C, 65% to 75% relative humidity, and a 16 h:8 h (L/D) photoperiod.

Detached Pod Assay to Assess Antibiosis Mechanism of Resistance in Pigeonpea ILs Using
Third-Instar Larvae of Pod Borer, H. armigera The relative resistance of pigeonpea ILs was evaluated using third-instar larvae of H. armigera. Detached inflorescences with pods were cut with a surgical blade and immediately placed in a slanting direction onto 3% w/v agar-agar medium in a 250 mL plastic cup (9.0 × 6.5 cm diameter) [2]. There were three replications of each accession in an RBD. A single third-instar larva was released on pods of pigeonpea with two pods per plastic cup. The initial and final larval weights were recorded before and after a 4-day-feeding period, respectively, and the pod damage rating was determined at the end of the feeding period. The weight gained (in percentage) by the larvae was calculated as follows: weight gain (%) = ((final larval weight − initial larval weight)/initial larval weight) × 100.

Biochemical Profiling of Seeds
Biochemical parameters, i.e., total phenols and total flavonoids concentrations, were determined for the seeds of 156 ILs as well as those of resistant and susceptible checks. To determine total flavonoids and phenols concentrations, the seeds of each accession were oven-dried at 55 • C for 3 days, then powdered in a Willey mill (Thomas Willey Mills, Swedesboro, NJ, USA) and defatted using hexane solution (100 mL g −1 ). The resulting materials were used for analyses of total phenols and total flavonoid concentrations using spectrophotometric methods, as described below. Each line was analyzed with two replicates in a completely randomized design [4].

Estimation of Total Phenols
The total phenols concentration in seeds was determined using a colorimetric method [17]. A 0.5 g portion of defatted seed sample was ground with 80% (v/v) ethanol in a pestle and mortar and then the mixture was centrifuged at 10,000 rpm for 20 min. The extraction was repeated five times. The supernatant was evaporated to dryness and then dissolved in water (5 mL). The total phenols concentration was expressed in mg/g of dry weight of seeds.

Estimation of Total Flavonoids
The total flavonoids concentration in seeds was determined by vanillin reagent method [18]. A 0.5 g portion of defatted seed sample was homogenized in ethanol and the mixture was centrifuged at 10,000 rpm for 20 min. The supernatant was evaporated to dryness and then dissolved in water (5 mL). This mixture was used for estimation of the total flavonoid concentration (expressed in mg/g of dry weight of seeds).

Morphological Parameters
On the basis of the performance of selected ILs during the 2019 rainy season, the most promising 39 pod borer resistant ILs were selected for further analysis in a trial in the 2020 rainy season. The trichome density on the leaves of these 39 ILs was recorded by observing a minimum of three uniformly developed leaves from each accession. Each IL was analyzed with three replications. The leaf samples were immersed in an acetic acid and ethanol mixture (2:1) in stoppered 10 mL glass vials for 24 h to remove chlorophyll, and subsequently transferred into lactic acid (90% v/v) as a preservative. The calyxes and the pods were examined at 10× magnification under a stereomicroscope (Carl Zeiss, Inc., Thornwood, NY, USA) equipped with an ocular measuring grid. The number of different types of trichomes (types A, B, C, and D; Figure 2) and their density (mm 2 ) within the microscopic field were recorded [4,7].

Statistical Analysis
Data were subjected to analysis of variance (ANOVA) using GenStat software (14th Edition). The significance of differences between the genotypes was determined by F-test, while differences among treatment means were determined by the least significant difference (LSD) test at p ≤ 0.05. The mean performance of the test entries was assessed across seasons.

Statistical Analysis
Data were subjected to analysis of variance (ANOVA) using GenStat software (14th Edition). The significance of differences between the genotypes was determined by F-test, while differences among treatment means were determined by the least significant difference (LSD) test at p ≤ 0.05. The mean performance of the test entries was assessed across seasons. In PP 1501, the visual damage rating at the podding stage ranged from 6 to 9 and the recovery resistance score at the harvest stage ranged from 4 to 7. Total 243 ILs with recovery resistance scores of 4 to 6 along with the resistant and susceptible checks were selected to evaluate damage caused by the pod borer complex.
From this population, we selected the 79 ILs with the lowest levels of pod borer complex/pod borer/pod fly damage (Table 2) for further field and laboratory evaluations during the 2019 rainy season.

PP 1503
The visual damage rating at the podding stage ranged from 4 to 9 and the recovery resistance score at the harvest stage ranged from 5 to 8. Total 342 lines with recovery resistance scores of 5 to 6 along with the susceptible and resistant checks were further evaluated for pod borer complex damage.
On the basis of the lowest pod borer complex damage, we selected 41 ILs (Table 2) for further field and laboratory evaluations during the 2019 rainy season. The visual damage rating at the podding stage ranged from 6 to 9 and the recovery resistance score at the harvest stage varied from 5 to 7. Ninety-one ILs with recovery resistance scores of 5 to 6 at the harvest stage, along with checks, were further evaluated for pod borer, pod fly, and pod borer complex damage.
On the basis of the lowest levels of pod borer, pod fly, and pod borer complex damage, eight ILs were selected for re-evaluation both under field and laboratory studies during the 2019 rainy season ( Table 2).

PP 1505
The visual damage rating at the podding stage ranged from 6 to 9 and the recovery resistance score at the harvest stage ranged from 5 to 8. A total of 393 ILs with recovery resistance scores of 5 to 6 at the harvest stage, along with checks, were further evaluated for pod borer, pod fly, and pod borer complex damage.
The 28 ILs with the lowest levels of pod borer/pod fly/pod borer were selected for further evaluation in the field and the laboratory during the 2019 rainy season ( Table 2).

Phenotyping under Unprotected Field Conditions
In total, 156 ILs with checks were re-evaluated using a RBD during the 2019 rainy season. We detected significant differences among the ILs for most of the traits (p ≤ 0.001; Supplementary Table S5). The visual damage rating at the podding stage ranged from 5 to 7 and the recovery resistance score at the harvest stage ranged from 4.5 to 7.5. Ninetysix ILs with recovery resistance scores of 5 to 6 at the harvest stage, along with checks, were further evaluated to determine the extent of damage by pod borer, pod fly, and pod borer complex and other biochemical traits (Supplementary Table S5).
Besides pod borer complex damage, the number of larvae on each IL at the podding stage in each replication was recorded. The mean larval count per plant ranged from 0.11 to 2.44 (Supplementary Table S5). The lowest larval count was recorded in PP1501-1-10-  Table S5).
On the basis of the re-evaluation of 156 selected ILs for different traits (damage due to pod borer, pod borer complex, damage rating, larval weight gain (%), phenol content, flavonoid content, and larval count at the podding stage), the 39 ILs with the least pod borer damage were selected and evaluated during the 2020 rainy season to confirm their pod borer resistance.

Phenotyping for Pod Borer Complex Damage under Unprotected Field Conditions
We re-evaluated the 136 selected ILs, along with checks, using a RBD during the 2019 rainy season. We detected significant differences among the ILs for most of the traits (p ≤ 0.001). The recovery resistance score at the harvest stage ranged from 4 to 8. Thirty-five ILs with recovery resistance scores of 4 to 6, along with the checks, were further evaluated to determine pod borer, pod fly, and pod borer complex damage (Supplementary Table S7).

Identification of Promising Pod Borer Resistant ILs
On the basis of the evaluation of the 39 most promising bod borer resistant ILs during the 2020 rainy season and the performance of these 39 ILs in the 2019 rainy season, we identified the most promising ILs showing lower pod damage than that of the respective cultivated recurrent parents and the pod borer resistant checks over the tested years. Based on this criterion, 21 ILs from three populations (11 from PP 1501, two from PP 1503, and eight from PP 1505) with PSR based on pod borer complex and pod borer (H. armigera) damage ranging from 3.0 to 4.0 were identified (Table 5; Figure 5). The performance of these ILs across seasons in terms of the levels of pod borer, pod fly, and pod borer complex damage and their biochemical and morphological traits was compared with that of their respective recurrent parents ( Table 5). Most of these ILs were also better than the existing pod borer resistant checks. On the basis of these comparisons, the ILs with specific traits related to resistance were identified as candidates for use in pigeonpea improvement programs. These 21 ILs, on an average, took 98-139 days to 50% flowering and 141-180 days to maturity at Warangal. Table 6 lists the best ILs with specific traits related to resistance for ready use in pigeonpea breeding.  Table 6. Details of the most promising pod borer-tolerant introgression lines that could be utilized in pigeonpea improvement programs.

Discussion
Pod borer (H. armigera), pod fly (M. obtusa), spotted pod borer (Maruca vitrata), and pod-sucking bug (C. gibbosa) can cause grain losses of more than 40% in pigeonpea [2]. Of these, pod borer is the most damaging pest, and it has become strongly resistant to insecticides [19]. Only low to moderate levels of resistance are present in the cultivated germplasm of pigeonpea [20], most of which show >50% pod borer complex damage. Hence, there is a need to exploit the wild relatives of pigeonpea for resistance to this pest. Crop wild relatives (CWR) are important to reintroduce genetic diversity for crop improvement [3]. Wild Cajanus species comprise a diverse genetic pool, which can be used to broaden the narrow genetic base of pigeonpea for crop improvement [21,22]. There are different mechanisms conferring resistance to pod borer in wild Cajanus accessions, including low density of glandular trichomes (A and B types), high density of non-glandular trichomes (C and D type), antixenosis (oviposition non-preference by insects), antibiosis, and high concentrations of phenols [19]. The wild Cajanus species C. acutifolius (ICPW 001) and C. scarabaeoides (ICPW 281) have been reported to be highly resistant to pod borer, H. armigera [9,23]. The pod borer resistance of ICPW 001, which is native to Australia, was reported to be due to high levels of antixenosis and antibiosis [9]. In contrast, the pod borer resistance of ICPW 281, which is native to India, is due to the high density of non-glandular trichomes [19,24]. In the present work, we focused on introgressing pod borer resistance from C. acutifolius and C. scarabaeoides into two popular pigeonpea cultivars, ICP 8863 (Maruti) and ICPL 87119 (Asha). Here, simple and complex backcross approaches followed by one to two cycles of backcrossing were used to develop four prebreeding populations. Precise evaluation for pod borer damage over different years and locations was performed. To combine different mechanisms conferring pod borer tolerance from wild Cajanus species into the common cultivated background, two complex crosses were generated using ICPW 001 and ICPW 281 as donors in the genetic backgrounds of ICP 8863 and ICPL 87119.
In this study, we evaluated four pre-breeding populations comprising more than 2300 ILs, and then re-evaluated the promising ILs across years and locations. On the basis of these evaluations, we identified 39 ILs with improved resistance to pod borer. These 39 ILs were sourced from three populations: PP 1501 (22 ILs), PP 1503 (6 ILs), and PP 1505 (11 ILs). PP 1503 and PP 1505 were derived from pod borer resistant C. scarabaeoides accession ICPW 281 as the pollen donor. The pod borer resistance of ICPW 281 is due to the high density of non-glandular trichomes. All six ILs from PP 1503 and 10 ILs from PP 1505 had high-density type C trichomes, indicating that this trait was successfully introgressed from C. scarabaeoides into the pigeonpea cultivar ICPL 87119. Trichome density is negatively associated with larval growth and survival [25][26][27]. In the present study, five ILs from PP 1503 and seven ILs from PP 1505 with a low density of type B trichomes on adaxial and abaxial leaf surfaces and a high density of type C trichomes on the adaxial leaf surface also showed comparatively improved levels of antibiosis (i.e., lower % larval weight gain as compared with those in the susceptible cultivar ICPL 87 and the cultivated recurrent parent ICPL 87119). Based on the level of pod damage (%), two ILs (PP1503-6-1-4 and PP1503-5-2-4) from PP 1503 and eight ILs (PP1505-63-2-4, PP1505-20-5-2, PP1505-36-4-1, PP1505-36-4-2, PP1505-34-3-6, PP1505-11-2-6, PP1505-13-6-3, and PP1505-11-2-5) from PP 1505 showed lower pod borer damage than that of the recipient parent ICPL 87119 both in 2019 and 2020. These ILs also exhibited resistance to pod borer at the Warangal site.
The populations PP 1503 and PP 1504 were developed to combine different mechanisms of pod borer resistance such as high flavonoids and phenols concentrations from C. acutifolius and the high density of non-glandular trichomes from C. scarabaeoides into the cultivated varieties ICP 8863 and ICPL 87119. Although some ILs in PP 1504 with higher flavonoid concentrations (2.1-3.7 mg/g) than that in the cultivated recurrent parent ICP 8863 (1.67 mg/g) were identified in the 2019 rainy season trial, all these lines had high levels of pod borer complex damage and were not selected for further evaluation. In PP1503, ILs with higher phenols (up to 5.3 mg/g) and flavonoids (12.02 mg/g) concentrations than those in the cultivated recurrent parent ICPL 87119 (2.53 mg/g phenols and 4.58 mg/g flavonoids) were identified, but most of these ILs also had high levels of pod damage due to the large number of insect species that damage pigeonpea (Helicoverpa armigera, Maruca vitrata, Etiella zincknella, Melanogromyza obtusa, and others). Only six ILs with low levels of pod damage were selected for further evaluation in 2020. Based on all the traits studied here, only one IL, PP1503-5-2-4, showing low pod damage in both the 2019 and 2020 rainy seasons, had a high total phenols concentration (4.58 mg/g), and a high density of non-glandular type C trichomes on the adaxial leaf surface. This line also showed improved antibiosis for ovipositing insects compared to the cultivated parent ICPL 87119. These results suggest that resistance to pod borers is a complex trait, and that for high concentrations of phenols and/or flavonoids, antibiosis may not be useful as a single criterion to select pod borer resistant ILs. This conclusion is consistent with the findings of other studies [9,[35][36][37]. The pod borer resistance of PP1503-6-1-4 may be attributed to the low density of type A and type B trichomes on the adaxial and abaxial leaf surfaces, high density of type C and D trichomes on the adaxial leaf surface and strong antibiosis for larval development.