Insects
Anthonomus grandis grandis adults were obtained from a colony maintained at Embrapa Genetic Resources and Biotechnology, in Brasília, DF, Brazil (15° 46' 46" S, 47° 55' 46 W). Insects were maintained in plastic Petri dishes (90 x 15 mm) in controlled environmental chambers (25 ± 2°C, 65 ± 10% RH, 14:10 L:D photoperiod) and reared in an artificial diet based on a mixture of agar, yeast, wheat germ, soy protein, glucose, ascorbic acid and sorbic acid, nipagin, embryo cottonseed flour (Pharmamedia®, Traders Protein,USA), mixture of Wesson salts, Vanderzant vitamin and water (Da Silva et al. 2022). The newly emerged adults were separated by sex (Suh et al. 2020), transferred to Petri dishes (20 insects /Petri dish). And feed with an artificial diet.
Plants
Gossypium hirsutum L. (genotype Delta Opal obtained by MDM-Maeda DeltaPine Monsanto, in Uberlândia – MG) were grown individually in 1.5 L pots filled with soil and organic substrate (in proportion of 1:1). Plants were grown in a greenhouse under controlled conditions temperature 27 ± 1ºC and photoperiod 14:10 LD. Cotton plants used in experiments were 12 weeks old at the reproductive stage (presence of squares).
Concentration- and time-mortality bioassay
Time-mortality bioassays were carried out to record the median survival time for the boll weevil adults using a range of exposure times, and each subjected to a range of insecticide concentrations. The concentrations used initially were based on the maximum concentration recommended for application in the field. For malathion, the concentrations evaluated were: 1000, 500, 400, 300, 200 and 100 g (active ingredient a.i) a.i./L (emulsifiable concentrate, FMC, Uberaba, MG, Brazil). For beta-cyfluthrin, concentrations evaluated were: 125, 62.5, 50, 37.5, 25 and 12.5 g a.i./L (concentrated suspension, Bayer CropScience, Belford Roxo, RJ, Brazil). Each insecticide was diluted in distilled water and 1mL of solution was applied to a transparent glass surface (90 x 15 mm) (Yuan and Chambers 1998) homogeneously. The control treatment was carried out without the use of insecticides (only distilled water), used to assess natural mortality. After application, the water was allowed to evaporate keeping the plates at ambient temperature for 24 hours. The top of the Petri dish was brushed with petroleum jelly to prevent the insects from escaping.
For each insecticide and concentration in each time-mortality bioassay, two replicates with 30 insects each were used. Each replicate consisted of a glass Petri dish. The design used was completely randomized. Mortality assessments were made at discrete (and independent) time intervals for each insecticide and concentration. The data thus obtained were submitted to time-mortality analysis using probit (PROBIT procedure; SAS, SAS Institute, Cary, NC, USA). The endpoint estimates of median lethal time (LT50) were subsequently regressed against insecticide concentration using the curve-fitting procedure of TableCurve 2D (Systat Software, San Jose, CA, USA). Model selection was based on parsimony, high F-values and steep R2 increase with model complexity, indicating that the complexity chosen for the model significantly improved the explanatory capacity of the data, justifying the inclusion of the parameters in the model.
Behavioral bioassay
Adult boll weevils were exposed to 100 g a.i. / L of malathion, per 1 min, and to beta-cyfluthrin, concentration 25 g i.a./L per 1.0 min. Each Petri dish received 1 mL of insecticidal suspension at the desired concentration (Yuan and Chambers 1998). The insecticidal residue was dried in fresh air for 24 hours, after which 30 adult insects were placed into each Petri dish, with three replicates for each insecticide. The top of the Petri dish was brushed with odorless talc to prevent insects from escaping. A control treatment (with distilled water only) was used to determine the natural mortality. Insect mortality was assessed until all insects were dead. Insects were recognized as dead when they were unable to walk coordinately, even when touched with a fine hair brush. These concentrations and exposure times were established in order to allow maximum exposure with reduced mortality of individuals, providing an adequate number of surviving individuals for the experiments and quantification of the variables with maximum accuracy and precision (Oliveira-Marra et al. 2019)
Pyrethroid Insecticide can cause a marked initial knock-down effect on insects, which can remain paralyzed and, later, return to normal movements (Casida and Durkin 2013). In order to ensure the correct observation of the possible effects of exposure to insecticides after their full recovery, the responsiveness to the grandlure synthetic pheromone was assessed after 24 h and 96 h of exposure to insecticides in tests performed on an olfactometer.
For the bioassays, an olfactometer made of Y-shaped acrylic plate was used, with the central body of 7.7 x 1.3 cm and two arms of 7.0 x 1.3 cm each. The olfactometer was closed, both at the top and at the bottom, with glass plates. The air filtered through activated charcoal (4–20 mesh, Supelco) and humidified was conducted using an aquarium pump (flow of 1.0 L min− 1), through silicone hoses, to the arms of the olfactometer. In one arm, a glass syringe containing the sex pheromone grandlure (Luretape BW-10 from Biocontrole®, Emigsville, PA, USA) was connected to thte silicone hoses. In the other arm of the olfactometer, only filter air pass across sytinge and silicone hoses, constituting the control treatment. For a clean air flow, a suction pump was connected at the opposite end to the olfactometer inlet (0.6 L min− 1). Males and females, virgins, 7 days old (N = 40 for each sex) of cottonweed were individually placed on the olfactometer and their behavior observed for a period of up to 5.0 minutes. The measured variable was called “initial choice”, that is, the first arm of the olfactometer through which the insect enters, in more than half of its length. Every five bioassays the position of each treatment was inverted to avoid position bias.
The data of the choices made by the individuals in each treatment were analyzed by the chi-square test (at 5% significance) using the statistical program R 3.4.3 (R Statistical Software) and testing the null hypothesis of non-preference for between one of the treatments (50% of choice for pheromone or control).
Volatile collections
Ninedays-old males of the cotton weevil were exposed to malathion (N = 6 individuals) at 10% of the field concentration, 0.2 L pc. / ha (100 g ia / L), and beta-cyfluthrin (N = 9 individuals) at 20% of the field concentration (20 mL pc / ha), both for 1.0 min, following the methodology described previously. These concentrations and exposure times were predetermined according to the time-mortality curves described above. The control was exposed to a surface containing distilled water (N = 9 individuals). After exposure, the insects remained for 24 hours without feeding and subsequently placed in pairs in contact with cotton plants in reproductive stage, without previous attack, conditioned in a glass cylinder (10 L internal volume). In this system, aerations were carried out by a push-pull system with the air entering into the chambers by means of a compressor with an air flow of 1 L.min-1, connected to an activated charcoal filter. The air was exhausted through a vacuum pump, at a flow rate of 0.6 L.min− 1, connected to a glass tube containing an adsorbent polymer, via polytetrafluoroethylene (PTFE) connections. The volatiles were collected every 24 h, for four consecutive days, in glass tubes containing 60 mg of Porapak Q (50–80 mesh, Sigma-Aldrich, Bellefonte, PA, USA). These were eluted from the adsorbent tubes using 500 µL of the organic solvent n-hexane and concentrated to 50 µL under N2 flow. The samples obtained were stored at -20ºC until use in a gas chromatograph coupled to the flame ionization detector (CG-DIC) and CG coupled to the mass spectrometer (CG-EM).
Chemical analysis
For analysis of volatile samples, 1 µL of each solution was injected into the GC (Agilent 7890-A, non-polar column DB-5MS, 0.32 mm in diameter x 60 m in length and 0.25 µm film, Supelco, Bellefonte, PA, USA), with flame ionization detector (FID), using splitless mode and using helium as carrier gas. The initial temperature of the ramp was 50° C for 2 min, gradually increasing by 5°C.min-1 until reaching 180°C, where it was maintained for 0.1 min, followed by a second gradual increase of 10°C min − 1 until reaching 250°C, remaining at that temperature for 20 min. The detector temperature was 270°C and the injector temperature was 250°C.
For qualitative analysis, selected samples were injected into the CG-MS (Agilent 5975-MSD) equipped with a quadrupole analyzer, in a DB-5MS apolar column (0.25 mm in diameter x 30 m in length, with 0.25 µm film, Supelco, Bellefonte, PA, USA), with electron impact ionization (70 -eV, temperature 200°C) and injector in splitless mode. Helium was used as the carrier gas. The data were collected and analyzed with the ChemStation software. The identification of volatile compounds was performed by comparing the fragmentation pattern of the sample components with that of cataloged data in spectral libraries (NIST 2008).
For the quantitative analysis the ratio between the four pheromone components of the cotton weevil was obtained using the area obtained from the chromatography analysis (GC-FID). Pheromone components: [cis-2-isopropenyl-1-methylcyclobutane-ethanol (grandlure I), cis-2-(3,3- dimethyl)-cyclohexylidene-ethanol (grandlure II), cis-3,3-dimethylcyclohexylidene- acetaldehyde (grandlure III) and trans-3,3-dimethylcyclohexylidene-acetaldehyde (grandlure IV)]. The proportions between the components of the pheromone obtained were compared using Kruskal-Wallis (P < 0.05). The presence or absence of the four main components release by the insects inside each chamber was observed during the evaluation periods (0–24, 24–48, 48–72 and 72–96 h). The same plants and insects were used along the experimental period. The influence of the treatments on the pheromone release were tested using athe generalized linear model of repeated measures (GLM) with binomial error distribution. The mean quantities of pheromone production from insects submitted to different treatments were compared through contrast analysis. The determination of the proportion of individuals capable of releasing pheromone by treatment, as well as their confidence interval, were performed using a generalized linear model (GLM) with binomial error distribution. All analyzes were performed using the statistical program R 3.4.3 (R Statistical Software).
Fertility table bioassay
Newly emerged males and females were sexed and kept in a Petri dish containing the diet for seven days, until their complete sexual maturation (Showler et al. 2004). After this period, the insects were exposed to dry insecticide residue, as described above and at the desired concentration and exposure time (Guimarães et al. 2008). The insecticide malathion (100 g a.i./L) was used for 1.0 min and beta-cyfluthrin (25 g a.i./L) for 1.0 min. The control was treated with sterile distilled water (Yuan and Chambers 2008).
The couples stayed together for 48 hours to ensure mating. After this period, the couples were separated, and the females were observed and evaluated until their death. Each female constituted an experimental unit, totaling 15 experimental units per treatment. The tests were carried out in a completely randomized design, which were placed in an air-conditioned room, at a temperature of 26 ± 2 ° C, RH, 70 ± 5% and a photophase of 12 hours. The treatments consisted of: 1) non-exposed male and female; 2) exposed male; 3) exposed female; 4) exposed male and female. The number of eggs deposited by mated females was checked daily and they were placed in a Petri dish containing an artificial diet to determine the viability of the eggs. The pre-oviposition period (time established between the emergence of the female and the beginning of the laying), total and average fertility (number of eggs / female), the oviposition period (laying period) and the longevity of females in each treatment were determined. From the data on survival and oviposition, fertility tables were drawn up (Garcia et al. 2006). Subsequently, the average number of eggs per female (mx) was calculated on each oviposition date (x), considering the total number of females, the accumulated female survival index (lx) during the oviposition period, and the number of descendants who reached age x. These values constituted the columns of the life tables (Teodoro et al. 2014).
Based on the information condensed in the fertility table, the following parameters were estimated for each treatment (Chi et al. 2020): average generation time (days), which represents the average time between laying a generation and the posture of the next generation; net reproduction rate (Ro), which is the estimate of the average number of offspring generated per female over the oviposition period and which will reach the next generation; and intrinsic growth rate (rm), which is the factor related to the population growth rate (daily production of progeny per adult female).
The estimates of the parameters of the fertility life table were performed using the jackknife method, using a protocol developed for the R statistical program (Maia et al. 2014). The means were compared by the Tukey HSD test to P < 0,05.