Introduction

The global population is currently undergoing exponential growth, and it is projected that by 2050, the world's population will approach nearly 10 billion individuals (Reid and Greene 2012; Julot and Hiller 2021; United Nations 2023). This fact presents a formidable challenge for agricultural food production for a growing global population in accordance with the principles of agricultural sustainability (Abrol and Shankar 2014; Tiwari and Singh 2021; Patel et al. 2022; Singh and Rale 2022). In this context, according to the FAO, between 20 and 40% of global crop production is lost annually to pests (FAO 2019). Invasive insects cost the world economy approximately US$70 billion annually, while plant diseases cost it approximately $220 billion (FAO 2019; Julot and Hiller 2021). Consequently, pest control emerges as a primary concern in crop production, and the use of chemical pesticides has experienced a significant and alarming surge in recent decades, becoming a central element of the prevailing crop production system. To provide some perspective, 370 million kilograms of pesticides were sold within the European Union in 2018 (Jacquet et al. 2022). Despite their effectiveness, many of these chemical pesticides are linked to a plethora of adverse consequences for both human health and the environment (Reid and Greene 2012; Julot and Hiller 2021; Adeleke et al. 2022). Aligned with the European Commission's "Green Deal," numerous countries have integrated the reduction of pesticide usage as a primary objective within their public policies, with the aim of promoting sustainable agriculture (Julot and Hiller 2021; Jacquet et al. 2022).

Within this framework, the exploration of potential environmentally friendly entomopathogenic endophytic microorganisms such as entomopathogenic ascomycetes (EA) has the potential for establishing a stable and pest-free ecosystem, ultimately fostering higher and more sustainable crop productivity (Solter et al. 2017; Parewa et al. 2018; Quesada-Moraga et al. 2020; Quesada Moraga 2020). These fungi are recognized as excellent biocontrol tools to be used in IPM programs since they can infect a wide range of arthropod pests with a unique mode of action by contact through the integument, playing a key role in crop pest control (Quesada-Moraga et al. 2020). Among EA, the genera Metarhizium and Beauveria are also considered excellent examples of fungi with multifunctional lifestyles (Barelli et al. 2016) that positively impact plant growth and immunity against generalist herbivores (Gange et al. 2019) and other biotic (Gupta et al. 2022; Posada-Vergara et al. 2022, 2023; García-Espinoza et al. 2023a) and abiotic stresses (Khan et al. 2012, 2015; García-Espinoza et al. 2023b; Chaudhary et al. 2023).

The effectiveness of several endophytic EA against some of the most destructive piercing–sucking melon pests, such as Aphis gossypii Glover (Hemiptera: Aphididae), has been well documented (Resquín-Romero et al. 2016b; Garrido-Jurado et al. 2017; González-Mas et al. 2019a). In addition, the response of melon plants to EA colonization in terms of defense induction (González-Guzmán et al. 2022; García-Espinoza et al. 2023a, b), which can ultimately influence multitrophic interactions involving melon, A. gossypii and their natural enemies, predators and parasitoids, such as Chrysoperla carnea (Stephens) (Neuroptera: Chrysopidae) and Aphidius colemani (Dalman) (Hymenoptera: Braconidae), respectively (González-Mas et al. 2019a; Quesada-Moraga et al. 2022), has also been documented.

Regarding chewing pests, the cotton leaf worm Spodoptera littoralis (Boisduval) (Lepidoptera: Noctuidae) poses a major threat to agricultural crops in the Mediterranean region (Ahmed et al. 2019; Ghoneim et al. 2020). This pest shares its habitat with the koinobiont solitary parasitoid Hyposoter didymator (Thunberg) (Hymenoptera: Ichneumonidae), which plays a crucial ecological role as a native parasitoid in southern Spain. In this context, Miranda-Fuentes et al. (2020) observed in laboratory settings that H. didymator and Metarhizium brunneum Petch (Hypocreales: Clavicipitaceae) EAMa 01/58-Su strains were compatible for controlling these noctuid pests and even enhancing fungal performance due to parasitization. Similarly, under the same laboratory conditions, S. littoralis larvae feeding on M. brunneum-colonized plants did not affect the reproductive potential of the parasitoid, and both the fungus and parasitoid larvae were found to coexist within the same larval host (Miranda-Fuentes et al. 2021). We hereby provide a step forward: the evaluation of the biocontrol potential of the parasitoid H. didymator for controlling S. littoralis on melon plants inoculated with endophytic EA under greenhouse conditions. Additionally, we aimed to assess the possible added beneficial impact of fungal application on melon plant growth in this multitrophic system.

Material and methods

Biological material and growth conditions

Fungal strains

One strain of M. brunneum (EAMa 01/58-Su) and two strains (EABb 04/01-Tip and EABb 01/33-Su) of Beauveria bassiana (Bals.) Vuill (Ascomycota: Hypocreales) from the culture collection of the Agronomy Department, University of Cordoba (Spain) were used in this work (Table 1). More details about these strains and their potential as biocontrol agents can be found in our previous works (Quesada-Moraga et al. 2006; Yousef et al. 2018; Garrido-Jurado et al. 2019; González-Mas et al. 2019b; Miranda-Fuentes et al. 2020, 2021; García-Espinoza et al. 2023a).

Table 1 Metarhizium brunneum and Beauveria bassiana strains used in this study
Full size table

Insects

The S. littoralis and H. didymator specimens used in this work came from a colony established at the insectarium of the Agricultural and Forestry Entomology Laboratory of the University of Cordoba. The growth chamber was maintained under the following conditions: 26 ± 2 °C, 70 ± 5% RH and a photoperiod of 16:8 (L:D) h (Miranda-Fuentes et al. 2020, 2021).

The pollinator Bombus terrestris L. (Hymenoptera: Apidae), used in the second experiment to obtain fruits and seeds, was acquired from a commercial stock of Koppert S.A.

Inoculum, growth conditions and plant inoculation methods

To acquire the inoculum, the procedure was carried out following the methods described in García-Espinoza et al. (2023a, b). In brief, the three strains were subcultured from their stored slant cultures on potato dextrose agar (PDA) in Petri dishes and allowed to grow for 15 days at a temperature of 25 °C in complete darkness. Subsequently, the inoculum preparation process involved scraping conidia from the Petri plates into a sterile solution of 0.1% Tween 80. To ensure homogenization of the inoculum, it underwent a 5-min sonication step. The final step involved filtration through multiple layers of cheesecloth to eliminate any residual mycelia. A hemocytometer (Malassez chamber; Blau Brand, Wertheim, Germany) was used to estimate the conidia concentration, which was finally adjusted to 1 × 108 conidia/ml by adding a sterile solution of distilled water with 0.1% Tween 80.

In all experiments, certified endophyte-free melon seeds (Cucumis melo L. cv. Galia) were employed. These seeds, which had undergone prior surface sterilization following the method outlined by (Garrido-Jurado et al. 2017), were germinated. Germination was carried out in 500 ml sterilized pots filled with universal black peat (Floragard, Germany), which was subjected to double sterilization in an autoclave at 121 °C for 30 min, with a 24-h interval, as described by González-Mas et al. (2019a). Three distinct application methods—namely, seed coating, soil drenching, and leaf spraying—were employed for each fungal strain. In the initial experiment, a completely randomized design was implemented, consisting of a total of nine treatments (comprising 3 strains and 3 application methods) and an untreated control group. There were four replicates (plants) for each treatment, as shown in SF 1A-D. In the subsequent experiment, plant inoculation was exclusively carried out via leaf spraying, as depicted in SF 1E-H. For the seed coating method, seeds were submerged in a suspension of 1 × 108 conidia/ml solution in Falcon tubes on a rotary shaker at 12 rpm for 4 h. Then, seeds were sown in pots previously prepared at 1 cm of profundity. On the other hand, soil drenching and leaf spraying were carried out when the melon plants reached the 4 true leaf stage, 30 d after seed coating treatment. For soil drenching, 5 ml of the conidial suspension was poured with a pipette onto the surface of the pot.

In the case of leaf spraying, for the first experiment (SF 1A), the entire plant was sprayed, and in the second experiment (SF 1E), only the 2 true basal leaves were sprayed (the adaxial and abaxial leaf surfaces) with 2 ml of conidia suspension using an aerograph (piston compressor of 23 l/min, 15–50 PSI and 0.3 mm nozzle diameter, Artesania Latina S.A., Madrid, Spain) (Miranda-Fuentes et al. 2021). To avoid contamination by run-off, soil and uninoculated parts of the plants were protected with aluminum foil and plastic bags, respectively. Control plants were sprayed with a sterile solution of 0.1% Tween 80. After soil drenching and leaf spraying, treated and control plants were covered with plastic bags for 48 h and maintained in the growth chamber.

The pots were kept in a greenhouse environment from October to December. During this period, the plants were watered three times a week, and their nutritional requirements were met using Nutrichem 60 fertilizer (N: 20, P: 20, K: 20) (Miller Chemical and Fertilizer Corporation, Hanover, Pennsylvania, USA). Fertilizer was added to the irrigation water at a rate of 1 g/l, and this procedure was carried out twice a month.

Compatibility between EA strains and H. didymator for S. littoralis control under greenhouse conditions

The compatibility between EA strains and the parasitoid H. didymator for controlling S. littoralis was investigated in two greenhouse experiments.

Experiment 1

The aim of the first experiment, as the first scenario, was to explore the impact of EA strains and inoculation methods on the fitness of S. littoralis within a multitrophic system that also involved the presence of the parasitoid. Melon plants with four well-developed true leaves were inoculated as described in "Inoculum, growth conditions and plant inoculation methods" section using the strains EAMa 01/58-Su of M. brunneum and EABb 04/01-Tip and EABb 01/33-Su of B. bassiana by three inoculation methods: seed coating, soil drenching and leaf spraying. For the leaf spraying method, the entire aerial part of the plant was sprayed (SF 1A). The treated plants were organized in groups of four per treatment within the greenhouse and enclosed in anti-aphid mesh cages. At 2 days postinoculation (DPI), 10 L3 (third-instar larvae) S. littoralis larvae were released in the second pair of true leaves (SF 1B), and larvae were confined with an organza bag for 24 h to ensure contact with the plant and the inoculum. Following this period, the textile bag was removed, and two females and four males of the parasitoid H. didymator were introduced per cage for each treatment (SF 1C). The parasitoids were then removed 24 h later, leaving only the S. littoralis larvae on the plants.

Larvae fed on the plants for 5 days. Then, the larvae were carefully removed from the plants and individually placed in methacrylate boxes. Here, they were provided with melon leaves from their respective treatments for an additional three days. Subsequently, they were fed an artificial diet until they reached the pupal stage. Throughout the entire process, starting from the introduction of L3 larvae until they reached the pupal stage, daily monitoring was conducted to record any instances of mortality.

Experiment 2

The aim of the second experiment, considered the second scenario, was to delve into the effect of endophytic colonization on S. littoralis fitness and H. didymator reproductive potential. For this assay, plants were inoculated with the three EA strains only by leaf spraying.

Plants with four well-developed true leaves were inoculated with the EAMa 01/58-Su M. brunneum strain and the EABb 04/01-Tip and EABb 01/33-Su B. bassiana strains. In this scenario, only the two basal leaves of each plant were subjected to spraying as detailed in "Inoculum, growth conditions and plant inoculation methods" section (SF 1E). The effects of the fungi were assessed on both S. littoralis larvae and the H. didymator parasitoid by introducing new L3 larvae onto the sprayed and unsprayed leaves.

At 2 DPI, 10 L3 S. littoralis larvae were released on sprayed or unsprayed leaves. For the first case, the larvae were enclosed within organza bags, following the procedures outlined in “Experiment 1” section (shown in SF 1F). In the case of larvae that fed on unsprayed leaves, the sprayed leaves were carefully isolated to prevent any contact between the larvae and the inoculum. After 24 h, the larvae were released, and four females and eight males of H. didymator were introduced (as illustrated in SF 1G).

With the aim of reaching fruit production, B. terrestris bumblebees were introduced during the flowering stage for a week to achieve pollination, and 4 specimens were introduced (4 times a week) per treatment. Once the fruits had matured, which occurred 115 days after sowing (DAS), they were removed to study the endophytic colonization both in the mesocarp and in the seeds (SF 1H).

Endophytic colonization by microbiological technique and by qPCR

The assessment and follow-up of the endophytic colonization of the three strains evaluated were carried out both by conventional microbiological techniques and by molecular detection and quantification by qPCR. Endophytic colonization monitoring was carried out from leaves collected at 2, 7, 14, 21 and 28 d after leaf spraying and soil drenching (SF 1D). Assessment of colonization by conventional methods was carried out according to Garrido-Jurado et al. (2017), González-Mas et al. (2019a, b) and Miranda-Fuentes et al. (2021). Following the same methodology, endophytic colonization was also evaluated on fruits (including mesocarp and seeds) collected at 115 DAS in the second experiment (SF 1H). Endophytic colonization was expressed and represented as a percentage, according to the number of fragments that presented growth of EA.

Molecular identification by qPCR was carried out as described by García-Espinoza et al. (2023a). Briefly, plant material was ground to a fine powder with a mortar and pestle in liquid nitrogen. Total DNA was isolated using a HigherPurity™ Plant DNA Purification Kit (Canvax Biotech S.L., Córdoba, Spain) according to the manufacturer’s instructions and resuspended in 100 μl of elution buffer. The concentration and quality of DNA were assessed by measuring absorbance at 260 nm and 280 nm in a NanoDrop™ 2000 (Thermo Fisher Scientific Inc.). The final concentration was homogenized until it reached 30 ng/μl. To identify and quantify the EAMa 01/58-Su M. brunneum strain, the primer pair of the nrr gene (F: TCA GGC GAT CTC GTG GTA AG, R: GGG GTG TAC TTG AGG AAT GGG) was used (Barelli et al. 2018), while in the case of the two strains of B. bassiana (EABb 04/01-Tip and EABb 01/33-Su), the ITSII rRNA gene (F: GCC GGC CCT GAA ATG G, R: GAT TCG AGG TCA ACG TTC AGA AG) pair primer was used (Bell et al. 2009). Real-time PCRs were performed in a qRT‒PCR Bio-Rad CFX Connect thermal cycler. Absolute quantification was carried out according to Bell et al. (2009) and Barelli et al. (2018). To set up the standard curves, a gradient of 1:4 from 40 ng to 0.16 pg of fungal and plant genomic DNA was used; absolute quantification was determined by comparing threshold cycle numbers against the standard curve previously generated (Bell et al. 2009; Barelli et al. 2018).

Assessment of melon growth promotion

At 8, 15, 22 and 28 DPI, measurements of plant length were recorded to assess plant growth. Subsequently, at 77 DPI (115 days after sowing), the fresh weight of both the aerial parts and roots was measured. To determine the weight of dry matter, which includes both aerial parts and roots, plant material was placed in paper envelopes and dried in an oven at 60 °C for 96 h.

Statistical analysis

Mortality data, expressed as percentages, were analyzed using a generalized linear mixed model with binomial distribution and logit link function. The significance of the treatment was analyzed with the F test and Tukey’s multiple comparisons (p < 0.05) (JMP 8.0, SAS Institute Inc.). Data on pupae weight, larval stage duration and plant growth were analyzed using analysis of variance (ANOVA) followed by a Tukey multiple range test among treatments (Statistix 9.0®, Analytical Software, Tallahassee, FL, USA). Means were compared by the HSD All-Pairwise Comparisons method. Different letters within columns or over the bars, as specified in the Fig. or Table legends, indicate significant differences (p < 0.05). The H. didymator mortality data were subjected to Kaplan–Meier survival analysis to calculate average survival time (AST) values in days and compared by the log-rank test calculated with IBM SPSS 25.0 software (SPSS Inc., Chicago, IL, USA).

The qPCR values represent the mean ± SE of four independent replicates. Fungal DNA quantification data were analyzed using one-way analysis of variance (ANOVA) followed by a Tukey test.

Results

In the first scenario, the melon plants were inoculated with the fungal strains by soil drenching, seed coating or plant spraying and then infested with noctuid larvae and parasitoids released at a 1:20 ratio. Fungal inoculation had a significant impact on larval mortality (χ2(1) = 75.99, p = 0.0001), with higher total mortality rates observed for the EAMa 01/58-Su M. brunneun strain applied by spraying (χ2(1) = 47.72, p = 0.0001), whereas the percentage of dead S. littoralis pupae, including those that showed abnormalities, was significant for all strains and application methods [EAMa 01/58-Su (χ2(3) = 58.06, p = 0.0001), EABb 04/01-Tip (χ2(3) = 103.92, p = 0.0001) and EABb 01/33-Su (χ2(3) = 39.47, p = 0.0001)] (Table 2). There were significant pupal mortality rates ranging between 16.67 and 38.89%, mainly for the soil drenching inoculation and leaf spray treatments (Table 2). Indeed, the EAMa 01/58-Su and EABb 04/01-Tip strain application methods significantly affected the total mortality of S. littoralis [EAMa 01/58-Su (χ2(2) = 163.35, p = 0.0001), EABb 04/01-Tip (χ2(2) = 29.19, p = 0.0001)] (Table 2). In this context, for the EAMa 01/58-Su strain, there were significant differences in mortality rates between seed coating and soil drenching (χ2(1) = 75.99, p = 0.0001), between seed coating and leaf spraying (χ2(1) = 163.05, p = 0.0001) and between soil drenching and leaf spraying (χ2(1) = 18.73, p = 0.0001); in the case of EABb 04/01-Tip, a significant difference was recorded between seed coating (χ2(1) = 22.73, p = 0.0001) and soil drenching (χ2(1) = 19.05, p = 0.0001) compared to leaf spray (Table 2).

Table 2 Lethal and sublethal effects of endophytic entomopathogenic fungi on S. littoralis larvae that were fed on colonized melon leaves in planta for 5 days and for 3 additional days on colonized leaf fragments
Full size table

The parasitoid was shown to be compatible with the fungal strains for S. littoralis control (χ2(1), p ≥ 0.05) for all strains and application methods, with mortality rates ranging from 11.11 to 77.78% (Table 2). Indeed, the three fungal strains led to a significant extension of the noctuid pupal development time (F3,81 = 5.09, p = 0.0029), whereas M. brunneum also caused an increase in the larval development time (F3,94 = 3.27, p = 0.0294) and a decrease in the noctuid pupal weight (F3,94 = 3.13, p = 0.0295). The lowest pupal weight was recorded for those specimens fed on plants inoculated by soil drenching and leaf spraying (Table 2).

In addition, the endophytic colonization of plants was assessed over time, with the three strains being able to colonize melon plants, whereas the intensity of colonization over time was strain and application method dependent, as shown by both microbiological and qPCR techniques (Fig. 1). In plants inoculated by leaf spraying, the presence of the EAMa 01/58-Su, EABb 04/01-Tip, and EABb 01/33-Su strains as endophytes was detected using both microbiological techniques and qPCR at all observation time points. Specifically, through microbiological techniques, EAMa 01/58-Su was detected at 2, 7, 21, and 28 DPI, EABb 04/01-Tip was detected at 2, 7, and 21 DPI, and EABb 01/33-Su was detected at 2, 7, 14, and 21 DPI (Fig. 1a–c). In contrast, qPCR analysis showed a similar prevalence of all three strains at 2, 7, and 14 DPI, which was significantly different from the levels observed at 21 and 28 DPI (p < 0.05) (Fig. 1d, e). No fungal presence was microbiologically detected in the mesocarp of fruits from inoculated plants with any of the strains used, except in the seeds from inoculated plants with the EABb 04/01-Tip strain, in which 20% of seeds presented fungal growth (SF 2). This result was confirmed by qPCR. In 33.3% of fruits from plants sprayed with EABb 04/01-Tip, a reading of 1.68 ± 0.19 pg/40 ng of total DNA per qPCR was recorded, while EABb 01/33-Su was detected in 50% of fruits from plants sprayed with this strain, showing 0.73 ± 0.05 pg/40 ng of total DNA per qPCR. However, the presence of the EAMa 01/58-Su M. brunneum strain in seeds was not detectable.

Fig. 1
figure 1

Detection of endophytic presence of EAMa 01/58-Su M. brunneum strain (a, d) and EABb 04/01-Tip (b, e) and EABb 01/33-Su (c, d) B. bassiana strains by microbiological technique (up) and by qPCR (down) in melon plants inoculated by seed coating, soil drenching, and leaves spraying. Samples were collected at 2, 7, 14, 21 and 28 days post-inoculation from untreated leaves. Endophytic colonization is expressed as a percentage of melon leaf fragments in which fungal growth was observed; molecular detection and quantification is expressed in fungal DNA picograms (pg) relative to 40 nanograms (ng) of total DNA per reaction. For qPCR quantification, bars represent the mean values of two technical replicates from each of four independent biological replicates. Letter over the bars denotes a significant difference between plants treated with each strain by seed coating, soil drenching or leaves spraying analyzed by sampling time by completely randomized ANOVA followed by a Tukey test (p < 0.05)

Full size image

In general, higher growth rates at 77 DPI (115 DAS) were observed in plants inoculated with EA, regardless of the specific strain or method of application used, with significant differences (p < 0.05) found in foliar fresh weight from all treatments except in those plants inoculated by soil drenching with the EABb 01/33-Su strain (Fig. 2a). Inoculation with the EABb 04/01-Tip strain by any of the three methods led to a significant increase in root fresh weight (Fig. 2b). Roots from plants inoculated with the EAMa 01/58-Su strain by leaf spraying were also significantly higher than those from the controls (Fig. 2b). There was a significant increase in the plant fresh weight in all inoculated plants except for the EABb 01/33-Su strain by soil drenching and leaf spraying (Fig. 2c).

Fig. 2
figure 2

Effects of endophytic entomopathogenic fungi on shoots (a) and roots (b) fresh weight and total fresh weight (c) of melon plants at 77 DPI (115 DAS) under greenhouse conditions. Plants were inoculated with EAMa 01/58-Su M. brunneum strain and EABb 04/01-Tip and EABb 01/33-Su B. bassiana strains by seed coating, soil drenching and leaves spraying; from 2 to 6 DPI plants were infested with 10 S. littoralis larvae which posteriorly were recollected and grown under laboratory conditions. Asterisks over the bars (mean ± SE) denote significant difference between treatments and control. Data were analyzed by completely randomized ANOVA followed by a Tukey test (p < 0.05) (n = 4)

Full size image

Significant differences (p < 0.05) were also found in foliar dry weight from plants inoculated with EAMa 01/58-Su or EABb 04/01-Tip strains by soil drenching and leaf spraying (Fig. 3a). The EAMa 01/58-Su strain only increased the dry root weight of plants inoculated by leaf spraying, while the EABb 04/01-Tip strain significantly increased the root dry weight of plants inoculated by all three inoculation methods used (Fig. 3b). Total dry weight was significantly different in plants inoculated with EAMa 01/58-Su (by soil drenching and leaf spraying) and EABb 04/01-Tip (by all inoculation methods used) strains when compared to controls (Fig. 3c).

Fig. 3
figure 3

Effects of endophytic entomopathogenic fungi on shoots (a) and roots (b) dry weight and total dry weight (c) of melon plants at 77 DPI (115 DAS) under greenhouse conditions. Plants were inoculated with EAMa 01/58-Su M. brunneum strain and EABb 04/01-Tip and EABb 01/33-Su B. bassiana strains by seed coating, soil drenching and leaves spraying; from 2 to 6 DPI plants were infested with 10 S. littoralis larvae which posteriorly were recollected and grown under laboratory conditions. To obtain shoots and roots dry matter weight, the vegetal material was placed in paper envelopes and dried in a stove at 60 °C for 96 h. Asterisks over the bars (mean ± SE) denotes significant difference between treatments and control. Data were analyzed by completely randomized ANOVA followed by a Tukey test (p < 0.05) (n = 4)

Full size image

Significant differences (p < 0.05) were also found in the length of plants inoculated with EAMa 01/58-Su and EABb 04/01-Tip strains by seed coating and soil drenching. At 15 DPI (Fig. 4a, b), the plants inoculated with EABb 01/33-Su by the three inoculation methods had significantly increased lengths when compared to the control (Fig. 4c). Only plants inoculated with the EABb 04/01-Tip strain by soil drenching showed a significant increase in length at 22 DPI (Fig. 4b).

Fig. 4
figure 4

Effects of endophytic entomopathogenic fungi on plant length at 8, 15 and 22 DPI under greenhouse conditions. Plants were inoculated with EAMa 01/58-Su M. brunneum strain and EABb 04/01-Tip and EABb 01/33-Su B. bassiana strains by seed coating, soil drenching and leaves spraying; from 2 to 6 DPI plants were infested with 10 S. littoralis larvae which posteriorly were recollected and grown under laboratory conditions. Asterisks over the bars (mean ± SE) denotes significant difference between treatments and control. Data were analyzed by completely randomized ANOVA followed by a Tukey test (p < 0.05) (n = 4)

Full size image

In a second scenario, significant larval S. littoralis mortality rates of 33–35% (for EAMa 01/58-Su strain), 15–28% (for EABb 04/01-Tip strain) and 16% (for EABb 01/33-Su strain) were observed after feeding larvae on sprayed or unsprayed melon leaves [sprayed (χ2(1) = 84.44, p = 0.0001) and unsprayed leaves (χ2(1) = 69.61, p = 0.0001) for EAMa 01/58-Su strain, sprayed leaves (χ2(1) = 60.35, p = 0.0001) and unsprayed leaves (χ2(1) = 17.45, p = 0.0001) for EABb 04/01-Tip and unsprayed leaves (χ2(1) = 18.64, p = 0.0001) for EABb 01/33-Su strain] (Table 3). In this scenario, relatively low significant pupal mortalities were recorded (Table 3).

Table 3 Lethal and sublethal effects of endophytic entomopathogenic fungi on S. littoralis larvae that were fed on sprayed or unsprayed melon leaves in planta for 5 days and for 3 additional days fed on colonized leaf fragments
Full size table

The reproductive potential of the parasitoid H. didymator, as indicated by its parasitization rates, varied from 38.89% to 89.44%, being significantly influenced by the fungal strain and application method when larvae fed on both sprayed and unsprayed leaves [(χ2(2) = 54.14, p = 0.0001) for EAMa 01/58-Su, (χ2(2) = 32.53, p = 0.0001) for EABb 04/01-Tip and (χ2(2) = 13.02, p = 0.0015) for EABb 01/33-Tip strain] (Table 3). It must be highlighted that the EAMa 01/58-Su strain, which caused higher larval mortality ratios, also led to the lower reproductive potential of the parasitoid (Table 3). Nonetheless, no significant differences were detected in the total mortality rates of the control larvae and the larvae exposed to melon leaves challenged by EAMa 01/58-Su (χ2(2) = 4.25, p = 0.1196), EABb 04/01-Tip (χ2(2) = 4.76, p = 0.0925) or EABb 01/33-Tip (χ2(2) = 1.30, p = 0.5217) strains, regardless of whether the S. littoralis larvae fed on sprayed or unsprayed melon leaves (Table 3).

Finally, there were sublethal effects observed in some life parameters of the F1 parasitoid generation. Specifically, pupal development time showed a significant elongation in parasitoids that developed on S. littoralis larvae fed melon plants inoculated with the EAMa 01/58-Su strain. The development time was 12.71 days in the sprayed leaf treatment (F1,44 = 6.42, p = 0.0150) and 13.44 days in the unsprayed leaf treatment (F1,39 = 13.61, p = 0.0007) compared to the control group, which showed 11.71 days for this parameter (Table 4). Likewise, in these treatments, a significant elongation of the preimaginal development time was also observed, showing 18.36 d for the control and 19.46 d and 20.38 d for those parasitoids developed on S. littoralis larvae fed sprayed (F1,40 = 9.25, p = 0.0042) and unsprayed leaves (F1,35 = 16.47, p = 0.0003), respectively (Table 4). The average survival time (AST) of H. didymator F1 adults that developed on S. littoralis larvae fed on plants inoculated with EAMa 01/58-Su was 28.91 d (sprayed leaves) and 27.75 d (unsprayed leaves), which was significantly different from the control in specimens that developed on S. littoralis larvae that fed on sprayed leaves (F1,40 = 9.39, p = 0.0039) (Table 4).

Table 4 Sublethal effects on F1 H. didymator that developed on S. littoralis larvae after being fed on sprayed and unsprayed leaves of melon plants
Full size table

Discussion

The multifunctionality of endophytic entomopathogenic ascomycetes extends their possible use beyond pest control and paves the way for new tools and applications in IPM and crop production in protected crops (Quesada Moraga 2020). Among them, an IPM strategy based upon the combined use of a natural enemy with endophytic entomopathogenic fungi either applied directly to target the pest or indirectly targeting the crop (Quesada-Moraga et al. 2022). While our previous work sheds light on the compatibility of the H. didymator – entomopathogenic fungus system when the fungal biocontrol agent targets S. littoralis larvae (Miranda-Fuentes et al. 2020, 2021), the possible multitrophic impact of the fungus as an endophyte on parasitoids and even on crop growth under real pest control greenhouse scenarios remained unknown.

The strains EAMa 01/58-Su, EABb 04/01-Tip, and EABb 01/33-Su successfully colonized the melon plants, and their endophytic presence was identified through both microbiological techniques and qPCR. The highest amount of fungal DNA was recorded in plants that were inoculated via leaf spraying. Interestingly, the B. bassiana EABb 04/01-Tip strain was reisolated from 20% of F1 seeds, which supports previous work demonstrating the vertical transmission of this fungal strain (Quesada-Moraga et al. 2014). The first scenario designed in the present study reveals the compatibility of the parasitoid with the three fungal strains when they target pest larvae by colonizing the plant. Likewise, the effect of the endophytic EA strains, even if lower than when they were directly sprayed onto the pest larvae (Miranda-Fuentes et al. 2020), led to significant larval mortality and anomalous pupation that strengthened the combined effect of the parasitoid and the fungus. Plant factors related to the endophytic EA-induced systemic defense responses in melon upon priming through the leaves, seeds or roots could be the possible cause of the observed fungal-related mortality rates and sublethal developmental effects (García-Espinoza et al. 2023a), which is further supported by the lack of fungal outgrowth from the cadavers of S. littoralis larvae feeding on EA-colonized melon leaves (Miranda-Fuentes et al. 2021).

The present research provides strong evidence of the multifunctionality of the selected EA strains, as indicated by their compatibility with the parasitoid for S. littoralis control while benefiting plant growth (Quesada-Moraga et al. 2020; Gupta et al. 2022; Posada-Vergara et al. 2022; García-Espinoza et al. 2023a). Likewise, significant differences in both the total and shoot and root fresh and dry weight of melon plants were observed for most of the fungal strain-combination method combinations. These findings agree with previous recent work showing the importance of EA as a promoter of plant growth (Raya-Díaz et al. 2017; Sánchez-Rodríguez et al. 2018; Tall and Meyling 2018; Gonzalez-Guzman et al. 2021; Mantzoukas et al. 2022; Batool et al. 2022; Adedayo and Babalola 2023; García-Espinoza et al. 2023b) and highlight the expanding role of EA beyond its traditional function in insect pest control (Quesada Moraga 2020; Quesada-Moraga et al. 2022).

The second scenario explored in our work, with a higher ratio of parasitoids released and inoculation of the basal leaves of melon plants, was conducive to better testing the translocation of the EA strains in the plant and their possible effect on the reproductive potential of the parasitoid. The fact that larval mortality rates were similar when the larvae fed on sprayed and unsprayed leaves again suggests the existence of direct and indirect effects on the pest larvae related to the fungal colonization of the melon plant. Several studies have reported that the effects of endophytic EA on target pests can be attributed to the presence of fungal inoculum in plant tissues, as shown for B. bassiana (Jaber and Enkerli 2016; Jensen et al. 2020; Agbessenou et al. 2020; Silva et al. 2020; Gupta et al. 2022; Torkaman et al. 2023), M. brunneum (Jaber and Enkerli 2016; Gupta et al. 2022; Posada-Vergara et al. 2023), Metarhizium robertsii (Metchnikoff) Sorokin (Hypocreales: Clavicipitaceae) (Liao et al. 2017; Ahmad et al. 2020, 2022) and Lecanicillium lecanii (Zimm.) Zare & W. Gams (Hypocreales: Clavicipitaceae) (Mejía and Espinel 2022), whereas as stated before, some EA can act by inducing systemic defense responses in plants even by priming them (Rondot and Reineke 2019; Ahmad et al. 2020; Gupta et al. 2022; Posada-Vergara et al. 2022; Van Hee et al. 2023; García-Espinoza et al. 2023a). Interestingly, the second scenario also revealed the sublethal effects caused by the three EA strains on the noctuid larvae and pupae as previously reported by Resquín-Romero et al. (2016a), who found a weight reduction in S. littoralis larvae treated with B. bassiana EABb 01/33-Su and M. brunneum EAMb 09/01-Su strains at a concentration of 1 × 108 conidia/ml. Likewise, Kalvnadi et al. (2018) reported a significant reduction in pupal weight in F1 Helicoverpa armigera (Hübner) (Lepidoptera: Noctuidae) descendants treated with sublethal doses of B. bassiana. Even if there are several previous works reporting the compatibility of parasitoids with insect pests directly exposed to EA as detailed in a revision conducted by Quesada-Moraga et al. (2022), information on parasitoid behavior when parasitized insect hosts feeding on EA-colonized plants is very scarce. Indeed, Jensen et al. (2020) and Oreste et al. (2016) suggested that the inoculation of plants with endophytic EA may indeed affect beneficial insects, making a key issue in evaluating this interaction before establishing any multitrophic system for IPM. In our work, the reproductive potential of the parasitoid H. didymator was not affected by any of the EA strain and application method combinations, demonstrating in all cases a parasitization capacity similar to that previously reported in laboratory settings (Miranda-Fuentes et al. 2021). Moreover, although a slight extension in larval development time and pupation time was observed in parasitoids emerging from S. littoralis larvae that had fed on inoculated plants (through direct or indirect contact), the average survival time was not reduced.

The present research represents a significant step forward in the pursuit of sustainability in food production by fully integrating macrobials such as the parasitoid H. didymator with endophytic M. brunneum and B. bassiana within real greenhouse agriculture settings. Our research underscores the compatible use of the endophytic EAMa 01/58-Su M. brunneum strain and EABb 04/01-Tip and EABb 01/33-Su B. bassiana strains with the parasitoid H. didymator for a sustainable IPM strategy for controlling S. littoralis in greenhouse conditions that can additionally exploit their multifunctionality for melon crop production.

Author contributions

EQM, MYY and FGE conceptualized the experiments; FGE MCM, MJG and MYY performed assays and assessments; FGE, MYY and MJG analyzed the data; FGE and MYY prepared the original draft; writing, review and editing, FGE, EQM, MYY, MJG and MCM. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

The data of this study are owned by the authors and can be viewed upon request. For further information, please contact the corresponding author. Supplementary material is available in additional files in the web version of this work.