The genotypes were selected based on preliminary tests of resistance of the melon plants to L. sativae, with the Goldex commercial hybrid used as the standard for susceptibility, and the CNPH 06-1047-343, CNPH 06-1047-333 and CNPH 06-1047-341 genotypes resistant through antibiosis to L. sativae (de Oliveira et al. 2021). These accessions were provided by the germplasm bank of Embrapa Hortaliças, in Brasília. Seeds from each of the melon progeny were placed in 200-cell polystyrene trays containing HS Florestal substrate (Holambra Substrates, Artur Nogueira, São Paulo, Brazil) and coconut fibre (1:1 w/w), and grown in a greenhouse (27 ± 3°C, 70 ± 10% relative humidity, and 12:12 h L:D photoperiod). Seven days after sowing, the emerging seedlings were transferred to plastic pots (7.5 cm in height × 10.5 cm in diameter; 0.4 kg capacity) containing HS Florestal substrate and sand (1:1 w/w) and kept under greenhouse conditions (27 ± 3°C, 70 ± 10% RH and a 12:12h L:D photoperiod) until two to three permanent leaves had fully developed (21 days after sowing). The plants were irrigated based on their water requirement (twice a day).
Rearing the Insects.
L. sativae larvae were collected from leaves of infested melon plants in plantations located in the district of Mossoró, Rio Grande do Norte, Brazil (04°03′53.7” S 40°53′34.0′′ W, altitude 32 m). A leafminer population was reared in the laboratory from adults that emerged from the collected larvae. The jack bean [Canavalia ensiformis (L.) (Fabaceae)] was selected as the host plant species to prevent pre-imaginal conditioning. Adult leafminers (at a sex ratio of 1:1) were transferred to wooden cages (100 × 100 × 100 cm) that were covered with voile and contained jack bean plants grown from seeds sown in 200-cell polystyrene trays and kept in a greenhouse at ambient temperature (27 ± 3°C, 70 ± 10% RH and a 12:12h L:D photoperiod) until showing sufficient leaf area to allow the leafminer to develop (10 days after sowing). The cages were kept in an insect rearing room (IRR) under controlled conditions (27 ± 2°C, 75 ± 10% RH and a 12:12h L:D photoperiod).
The behavioural response of females of L. sativae to the volatiles of melon genotypes was obtained using a four-arm acrylic olfactometer (12 cm x 12 cm) (PETTERSSON, 1970) under controlled conditions (25 ± 1°C and 70 ± 5% RH). In order to guarantee the impartiality of the bioassay, a blank test was initially carried out with the four arms of the olfactometer containing only constitutive volatiles of the Goldex commercial hybrid.
Two experiments were carried out. In the first, an olfactometer arm with constitutive melon volatiles (treatment) was compared to three of the arms containing air (control); four bioassays were conducted: (1) constitutive volatiles of the Goldex commercial hybrid versus air; (2) constitutive volatiles of the CNPH 06-1047-343 genotype versus air; (3) constitutive volatiles of the CNPH 06-1047-333 genotype versus air, and (4) constitutive volatiles of the CNPH 06-1047-341 genotype versus air.
In the second experiment, constitutive volatiles of the Goldex commercial hybrid were compared with each of the genotypes under study (CNPH 06-1047-343, CNPH 06-1047-333 and CNPH 06-1047-341). For these tests, two arms of the olfactometer with Goldex were compared to two arms containing one of the other genotypes. Three bioassays were conducted: (1) Goldex commercial hybrid versus the CNPH 06-1047-343 genotype, (2) Goldex commercial hybrid versus the CNPH 06-1047-333 genotype, and (3) Goldex commercial hybrid versus the CNPH 06-1047-341 genotype.
To supply the leafminer with the constitutive volatiles, the pots with the melon seedlings containing substrate were completely covered with aluminium foil to block the release of volatiles from the substrate. The seedlings were then wrapped in polyester bags (100%, 27 cm x 41 cm) that were completely sealed with Teflon tape. The bags were connected to the olfactometer using silicone hoses. In the olfactometry system, air is passed through an air pump with flow meters set at 250 mL/min for each arm of the olfactometer. Twenty replications were conducted for each bioassay. For each replication, one adult female was tested after fasting for 24h. The female was released through an orifice of the olfactometer in a neutral area, and exposed to the volatiles for 10 minutes. With each replication, the olfactometer was rotated 90° to reduce any positional effect. The response of the female was considered when it went beyond the neutral area and opted for one of the arms of the olfactometer. The replication was cancelled whenever the female remained in the neutral area for the first five minutes.
The frequency and length of stay of L. sativae in the different arms of the olfactometer were recorded using the SOLF software (Result Management System for Olfatometry Bioassays) v7.0 (Fancelli et al. 2017). At the end of five replications, the olfactometer was cleaned (with neutral detergent, distilled water and 70% alcohol) and a new melon seedling was offered as a source of odour .
Data on the frequency or number of times the L. sativae entered the different arms of the olfactometer (treatment and control) were compared using a χ² test (α = 0.05). In the first experiment, the frequency of each individual in the arms containing air (three) was compared with the expected fractions of 25% for the arm offering melon volatiles (one). In the second experiment, the frequency of each individual in the arms containing the constitutive volatiles of the Goldex commercial hybrid (two) was compared with the expected fractions of 50% for the arms offering constitutive volatiles of the other melon genotypes (two).
Data on the time L. sativae remained in the arms containing the constitutive melon volatiles were submitted to the T-test (α = 0.05), as described by Hegde et al. (2011) and Sobhy et al. (2017), for the arms containing air, mean values were obtained for the tests (Togni et al. 2010; Hegde et al. 2011). All the analyses were performed using the SAS statistical software (SAS Institute, 2019).
Extracting Volatiles from Melon Leaves.
The extraction of volatiles was performed by solidphase microextraction using the HS-SPME (Headspace Solid Phase Microextraction) method using the 1cm DVB/Car/PDMS (Divinylbenzene/Carboxen/ Polydimethylsiloxane) 50/30 fber. The fber was exposed to the vial headspace containing approximately 1g of leaf sample at 30°C for 15 min after the stabilization period conducted at 30°C for 30 min. For each melon genotypes (CNPH 06-1047-343, CNPH 06-1047-333, CNPH 06-1047-341 and Goldex), four replicates (plants) were used for sample withdrawal. Coupled GC-MS (Gas chromatography–mass spectrometry) analysis was performed on an Agilent model GC-7890B/MSD-5977 A (quadrupole) instrument with electron impact at 70 eV, HP-5MS methylpolysiloxane column (30 m x 0.25 mm x 0.25 µm, Agilent), helium carrier gas with 1.00 mL.min-1 (7.1 psi) fow and constant linear velocity of 36.3 cm.s-1, injector temperature 260°C, detector temperature 150°C, transfer line temperature 280°C. Chromatographic oven programming: initial temperature of 40°C, with a heating ramp from 7°C.min-1 to 260°C for 5 min at the end of the run. The identifcation of compounds was performed by analyzing the fragmentation patterns displayed in the mass specterwith those present in the database provided by the equipment (NIST version 2.0 of 2012-243,893 compounds), and from literature data.
Survival of Larvae and Pupae.
Six plants from the same genotype, each with three fully developed leaves (21 days after sowing), were transferred to a separate wooden cage (100 × 100 × 100 cm) covered with voile. Each cage received 24 L. sativae female adults (8 insects/plant), of up to 48 h in age (Celin et al. 2017), starved for 24 h, and previously maintained under controlled conditions for 24 h (27 ± 2°C .75 ± 10% RH and 12:12 h L:D photoperiod). The plants were kept under controlled conditions (27 ± 2ºC, 75 ± 10% RH and 12-hour photophase) until the leafminers emerged, and the number of emerged leafminers, pupae and adults was then quantified. Survival of the larvae and pupae was calculated as per Equations 1 and 2, respectively.
Larval survival (%) =100 (number of pupae/number of larvae) Eq. 1
Pupal survival (%) = 100 (number of emerged adults/number of pupae) Eq. 2
The larval and pupal viability data did not show a normal distribution, and were therefore submitted to the non-parametric Kruskal-Wallis test, followed by the post hoc DSCF (Dwass-Steel-Critchlow-Fligner) test at a significance level of 0.05 using the NPAR1WAY procedure of the SAS software (SAS Institute Inc, 2019).
Extracting Chemical Substances from the Ethanol Extract of the Melon Leaves.
Seedlings of the genotypes were grown in a greenhouse for 21 days, the time necessary for the leaves to completely expand. The extracts were prepared by liquid-liquid partition in an ultrasonic bath; 50 mg of plant material, weighed after drying and ground in triplicate, was weighed in test tubes, 4Ml hexane was added to eliminate interference. The material was homogenised in a vortex for 1 minute and then placed in an ultrasonic bath for 20 minutes at a fixed power of 135W with 4 Ml of an ethanol/water solution (70:30), the tubes were then added and remained in the ultrasonic bath for 20 minutes to extract the compounds of interest. The tubes were centrifuged for 10 minutes at 3000 rpm to facilitate separation of the partition. A 1 Ml aliquot of the lower (ethanolic) phase was removed using a Pasteur pipette, filtered in a 0.20 µm PTFE filter, collected in vials and stored in an ultra-freezer (-80°C) for further analysis by ultra-high performance liquid chromatography coupled to high-resolution mass spectrometry (UPLC-QTOF-MSE).
To quantify the total phenolics in the ethanol extracts, the Folin-Ciocalteu method was used, with modifications. Each extract was solubilised in an ethanol/water mixture and added to 10-ml flasks in triplicate. A test tube containing 0.5 ml of each solubilised extract was then used.
For the blank, 0.5 ml of 10% ethanol was used. Then, 0.5 ml of Folin-Ciocalteu reagent was added to each tube, shaken in a vortex, and, after 3 minutes, 0.5 ml 20% anhydrous sodium carbonate solution (Na2CO3) was added, topping up the volume to 5 ml with 3.5 ml distilled water. Solutions of the extracts were left to react for 90 min in the absence of light.
Absorbance readings were taken using a Carry spectrophotometer at a wavelength of 725 nm, comparing the sample with the blank. The calibration curve was calculated, and for each concentration point, the calibration curve was constructed, and the straight-line equation obtained. The mean absorbance value for each extract was then determined from the mean value of the triplicates.
The data for total phenolic concentration were tested for normality by the Shapiro-Wilk test and for homoscedasticity by Levene’s test. The difference between the genotypes for total phenolic concentration was verified by analysis of variance, with multiple comparison of the mean values by Tukey’s test at 0.05, using the GLM procedure of the SAS software (SAS Institute Inc, 2013).
The chemical profiles found using the UPLC-QTOF-MSE system were analysed with the MassLynx software (v4.1); chemical compounds in the melon leaves, particularly the phenolic compounds, were then identified. The analysis was carried out using an Acquity UPLC system (Waters, USA) coupled to a Xevo Quadrupole and Time-of-Flight mass spectrometer (Q-TOF, Waters). The separations were performed in a C18 column (Waters Acquity® UPLC C18 - 150 mm × 2.1 mm, 1.7 µm). A 2 µL aliquot of phenolic extract was submitted to an exploratory gradient with the mobile phase comprising deionised water (A) and acetonitrile (B), both containing formic acid (0.1% v/v) under the following conditions: 2-95% for 15 min, flow 400 µL min-1.
The spectrometric analysis was performed in negative and positive electrospray ionisation (ESI) mode, acquired in the 110 to 1200 Da range. In negative mode, the capillary voltage was set to 2800 V, cone voltage to 50 V, source temperature to 120ºC, desolvation temperature to 350ºC, gas-cone flow to 20 Lh-1 and desolvation gas flow to 500 Lh-1. In positive mode, the parameters were as follows: capillary voltage, 3200 V; cone voltage, 35 V; source temperature, 120°C; desolvation temperature, 350°C and desolvation gas flow, 500 Lh-1. The acquisition mode was MSE, and the system was controlled using the MassLynx 4.1 software (Waters Corporation).
To analyse the various compounds identified in the genotypes under study, hierarchical cluster analysis was carried out, resulting in the visualisation of a heat map. The area of the 10 leaf compounds was visualised using the GENE-E software for extract recognition, with classification in rows and columns. The Pearson correlation distance method using full linkage was applied to cluster the retention times (rows) and measure the proximity of the samples (columns). The result illustrates a 3D dendrogram (heat map), in which the red colour represents the highest relative concentrations, light blue the intermediate concentrations, and dark blue the lowest relative concentrations.