Entomopathogenic fungal strains source
B. bassiana (strain 7R), Trichoderma gamsii Samuels & Druzhin (Hypocreales: Hypocreaceae) (strain Z) and M. brunneum (strain Meta Br1) isolated from the Achaia region, Greece, were used for the in plant bioassays on pepper plants. In addition, the commercial B. bassiana GHA strain (Botanigard® 22WP, BioWorks Inc., Victor, NY) and the Mexican B. bassiana PTG4 strain (GenBank accession number KC759730.1), isolated from Periplaneta americana, were used to treat sorghum seeds, for the field experiment.
Fungi production for in plant bioassays
To achieve the fungi growth, strains from Greece were cultured on Petri dishes with Sabouraud dextrose agar (SDA) culture medium (OXOID LTD), incubated at 25 ± 2°C temperature in dark conditions. After each fungus growth (covering the dish surface), it was isolated once more time, in order to avoid contamination and to achieve pure cultures of each one. The conidia were retrieved from cultures maintained on SDA. Isolates were maintained in Petri dishes on a nutrient SDA incubated at 25 ± 1°C and were renewed every month. Fresh conidia were collected from the SDA cultures after 15 d and transferred to a 500-mL glass beaker with 100 mL sterile distilled water containing 0.05% Tergitol NP9. The conidial suspension was filtered across 4 layers of sterile cloth to remove hyphal debris and prepared by mixing the solution with a magnetic stirrer for 5 min (Dorschner et al. 1991).
In planta bioassays performed on pepper plants
In order to assess the impact by the three EPF on the M. persicae populations, in planta bioassays were settled up. M. persicae was obtained from the Plant Protection Institute of Patras, ELGO Dimitra and maintained on potted pepper plants variety Stauros Peloponnesus (Capsicum annuum L., Solanales: Solanaceae). Pepper plants were pre-germinated in 2 × 2 cm pots (one seed per pot at a depth of about 1cm) with Pindstrum plus peat substrate and then they were transplanted into one-liter pots with Pindstrum plus type peat substrate. Aphids were maintained at 25 ± 5 °C and a 14:10 h L:D regime. For this bioassay, insects were kept in a room with pepper plants throughout their developmental cycle, at a constant 25 ± 1°C temperature, 60-70% RH, and 16:8 h L:D photoperiod. Experimental pepper leaves were sprayed with 5 mL of the desired conidial concentration suspension (at 1 × 107 or 1 × 108 conidia/mL) of either B. bassiana strain 7R, T. gamsii strain Z, or M. brunneum strain Meta Br1, using a Badger 100 artist’s airbrush (Badger Air-Brush Company, Franklin Park, IL). After spraying suspensions of the selected fungus and concentration, plants were covered for 24 h with large diameter black bags to maintain high moisture on the plant surface. After these 24 h, one apterous M. persicae aphid was placed on a randomized leaf after two-hour starvation. Ten aphids were used per treatment (n=10), and each experiment (n=40) was replicated five times (n=200). Well-developed pepper leaves were used for each out of three treatments, plus one for the untreated control, in a block designed assay with four blocks. Therefore, each block consisted of three treatments plus the control, which were replicated five times, thus producing a total of 200 plants for the entire experiment. Each aphid-infected leaf was covered with an organdie 10 × 30 cage, to prevent dispersion into the experimental area. The aphid’s population was recorded on each leaf after 3, 6 and 9 d. The experiment, organized in a completely randomized design, was not analyzed as a factorial (since there was only one aphid species). Dead insects were removed and placed on 1.5% agar plates at 25 °C for an additional 2 d to detect if aerial mycelium was developed.
In order to assess how the temperature affected the EPF infection of the M. persicae exposed population and its potential dissemination, dead aphids were separated by treatment and maintained at 16L/8D photoperiod at three different temperatures: 20 °C, 25 °C, and 30 °C, under 90% RHs.
Fungi production for in field experiments
The Mexican GHA and PTG4 B. bassiana strains were growth using different culture conditions than those used for fungi production in Greece. In Mexico, the strains were activated by plating stock cultures onto potato dextrose agar (PDA, BD Difco, CDMX, Mexico) and incubated in darkness at 25 ± 2 °C for a week. To obtain a monosporic culture, a single selected colony was inoculated onto a PDA plate. This was considered the monosporic culture stock plate. Fresh conidia, collected from PDA plates that started from the monosporic stock, were incubated for 7 d in darkness at 25 ± 2 °C. A conidial suspension was prepared by scraping gently the top of the fungal cultures with a spatula and dissolving the spores and hyphae mixture in sterile distilled water. This suspension was filtered in 4 layers of sterile mesh-cloth to remove hyphal debris. Conidia were counted in a Neubauer chamber and each treatment was adjusted with distilled water to the indicated concentration.
Field Experiment on sorghum
Survival of native Melanaphis sacchari populations, potentially affected by the seed treatments with EPF, was recorded at the Experimental Field Unit of the Autonomous University of Nuevo Leon-School of Agronomy, located in Marín, Nuevo León, with a geographical location of 25°52'24.0" N 100°03'03.0" W. Tested sorghum was sweet sorghum (Sorghum bicolor L.) (Moench) variety ‘Roger’, which germplasm is deposited in the UANL Plant Varieties National Catalog (registration number SOG-261-050315) (López-Sandin et al. 2021). Sorghum seeds used in this work were directly obtained from this facility. B. bassiana GHA or PTG4 strains were used at the final concentration of 1 × 106 conidia/mL in each sorghum seed treatment. Cornstarch (CS) (Unilever Manufacturera, S. de R.L. de C.V., CDMX, Mexico) was mixed with conidia for adequate attachment to the seeds. Conidia + CS (4% CS final concentration) was prepared by first dissolving CS in boiling distilled water to a pre-gelatinized state. When the sticker suspension was at room temperature, conidia were then added until a homogeneous suspension was obtained (both the CS and conidia suspension were first prepared as 2X, then diluted twice when preparing the final conidia + CS suspension); next, seeds [5,000 seeds/treatment] were added into a 150 mL volume of conidia + CS, incubated for 5 min and then air-dried at 25 °C for 24 h.
Sorghum field trials consisted in 4 treatments and 4 replicated randomized plots. Treatments were PTG4+CS treated seeds (PTG4), GHA+CS treated seeds (GHA), non-treated control seeds (CONTROL) and chemically treated control plants (CHEM) (Imidaclopid/Betaciflutrin). A treatment with only CS was not included. All seeds of each treatment were sown in the field and regular agricultural practices were followed. One irrigation and no fertilizers nor herbicides were applied.
Melanaphis sacchari survival during the sorghum crop cycle
After sorghum germination, the presence of M. sacchari was monitored every week, until detecting 50-125 or higher aphids’ population per plant in the untreated control plants. After this, only the chemical control was applied, following the Imidaclopid/Betaciflutrin application guidelines (Muralla Max®, Bayer CropScience, Mexico) using a backpack applicator and a flexible hose, to avoid movement of the insecticide to the randomized control and EPF treated plots. After two weeks, the flowering percentage was recorded as the number of panicles present in 100 plants/treatment and was reported as the flowering index [flowering percentage transformed to a decimal value (0 to 1) per treatment]. At the same time point, aphid’s population survival analysis was done, focused on their relative abundance per treatment, among 100 plants chosen randomly. To express the aphid’s relative abundance values per plant, a scale of A-F (A=1-25 aphids, B=26-50 aphids, C= 51-100 aphids, D=101-500 aphids, E= 501-1000 aphids and F= > of 1000 aphids, was used (Bauer, 2015; Bowling et al. 2015). The scale F damage percentage was calculated as index F damage [percentage of scale F aphid’s relative abundance transformed to a decimal value (0 to 1) per treatment] and was chosen as indicative of sorghum damage.
Sorghum yield as stem fresh weight per plant, juice and sugar content per plant.
After the sorghum crop cycle was ended, 10 plants per treatment/replicate (a total of 40 plants per treatment) were randomly selected and their fresh stems weight were recorded, using an analytical scale (L-EQ, Torrey, Ciudad de México, Mexico). Juice from those stems was extracted using a juice extractor (QJH-L100A, Kuala Lumpur, Malaysia) and the total volume was measured by a graduated plastic cylinder. A digital refractometer (PAL-1, ATAGO USA, Bellevue, Washington, USA) was used to determine sugar content as Brix-degree percentage (one-degree Brix corresponded to 1 g of sucrose in 100 g of sorghum juice) and with this data, the g of sugar per plant was calculated.
Statistical analysis
Prior to analysis, the M. persicae mortality values were arcsine transformed. Mortality data were then analyzed by univariate ANOVA means, using the IBM general linear model (version 25.0, SPSS Inc., Armonk, NY, USA). The significance level was set at P < 0.05. In case of significant F values, means were compared using the Bonferroni test. Kaplan–Meier analysis was also selected to determine the M. persicae population median survival time, following to the EPF exposure. Median survival time comparison was performed using one-way ANOVA (treatment as factor) (SPSS v.25.0). Flowering index (normalized by arcsine transformation) and the effect of M. sacchari presence in sorghum, reported as index F damage (arcsine transformed), were analyzed by non-parametric Kruskal-Wallis means. Comparison of means was done using Bonferroni test. To determine the seed treatments global effect on this sorghum variety production, all collected data from fresh stem weight (in g/plant), juice volume (in mL/plant) and yield (g of sugar/plant) were analyzed altogether using Principal Component Analysis (PCA) by the Spearman method settings by the XLSTAT statistical package (V.2021 for Windows, Addinsoft, NY, USA). PCA correlations were considered significant when the Bartlett’s sphericity test p value was ≤0.05. Variables with a correlation coefficient ≥ 0.6 were considered relevant.