Dual effects of entomopathogenic fungi on control of the pest Lobesia botrana and the pathogenic fungus Eutypella microtheca on grapevine

Background Entomopathogenic fungi (EPF) are the natural enemies of insect pests. Nevertheless, research on the use of EPF for simultaneous prevention of pest and disease agents on the same crop is limited. In this study, we explored the potential dual effects of three strains of the EPF Metarhizium anisopliae on the control of detrimental agents of Vitis vinifera L., including different developmental stages (larvae, pupae, and adult) of the insect pest Lobesia botrana and the phytopathogenic fungus Eutypella microtheca. Methods Laboratory pathogenicity trials were performed to examine the effects of the three M. anisopliae strains on the mortality rate of L. botrana. In addition, field trials were conducted to assess the biocontrol potential of one selected M. anisopliae strain on the larval stage of L. botrana. Moreover, inhibitory effects of the three EPF strains on E. microtheca growth were examined in vitro. Results All the M. anisopliae strains were highly effective, killing all stages of L. botrana as well as inhibiting the growth of E. microtheca. The in vitro mortality of larvae treated with the strains was over 75%, whereas that of treated pupae and adults was over 85%. The three EPF strains showed similar efficacy against larvae and adult stages; nevertheless, pupal mortality was observed to be strain dependent. Mortality of L. botrana larvae ranged from 64 to 91% at field conditions. Inhibition of E. microtheca growth reached 50% in comparison to the control. Conclusions Our study showed that M. anisopliae strains were highly effective in ensuring control of two different detrimental agents of V. vinifera L., providing new evidence to support the dual effects of entomopathogenic fungi.

mating disruption [21][22][23][24]. However, these methods have been ineffective, as this pest continues to proliferate in vineyards. Moreover, the widespread use of chemical products in controlling crop pests and diseases have been widely criticised because of their negative impacts on biodiversity and human health [25,26]. To the best of our knowledge, researchers are yet to develop an effective, economic, and eco-friendly method for preventing and managing L. botrana infestation and E. microtheca infection. Owing to the negative impacts of synthetic pesticides on the environment, biopesticides such as entomopathogenic fungi (EPF), have been identified as a promising alternative for environmentally friendly pest management programmes [27].
EPF are a group of environmentally safe fungi used for biological control of insects and other arthropod pests of agricultural crops. They comprise a wide range of morphologically, phylogenetically, and ecologically diverse fungal species [28]. A wide range of EPF species have been established as important biocontrol agents in many natural and artificial ecosystems [29], to defend against a large number of insect species [30]. Additionally, recent studies suggest that EPF possess inhibitory actions against some phytopathogenic fungal species [31][32][33][34]. Thus, there is growing interest in the use of these organisms in integrated pest management (IPM) programmes [35]. However, research on the use of EPF in the simultaneous prevention and management of several pest and disease agents of the same crop is limited.
Here, we examined the effectiveness of three different strains of the EPF Metarhizium anisopliae in controlling L. botrana at both laboratory and field conditions, as well as growth of E. microtheca in vitro. Therefore, the findings of this study serve as a basis for the use of EPF in pest and disease control providing new evidence for the inclusion of EPF as dual biocontrol agents in vineyards.

Lobesia botrana specimens
The effects of M. anisopliae strains against the fifth instar larvae (L 5 ), pupae (P p ), and adults (A d ) stages of L. botrana were examined. Newly emerged larvae were obtained from a breeding colony in Mendoza, Argentina (33° 01′ 52″ S, 68° 46′ 34″ W). Fifth instar larvae of L. botrana were selected for trials based on previous studies reporting that damage by L. botrana increases with larval age [39]. The larvae were reared on an artificial diet [40], which was provided ad libitum during the complete growth period. Individuals from all stages were maintained in a growth chamber under a 16-h photoperiod, at a temperature of 25 ± 5 °C, and relative humidity ranging from 30 to 50%. This procedure has been reported to be effective in producing large numbers of larvae [41].

Phytopathogenic fungal strain
The phytopathogenic fungal strain was obtained from a vine culture (31° 76′ 10″ S, 68° 57′ 94″ W) located in San Juan Province, Argentina. Fourteen plants displaying clear symptoms of eutypiosis, including chlorotic and small leaves as well as short, rough internodes, stunted shoots, and dark brown wood with hard wedgeshaped consistency [42], were selected from the field. The infected plants were uprooted, and infected tissues were excised and taken to the laboratory. We collected 2-5 subsamples from each symptomatic wood fragment.
To remove external non-parasitic microorganisms, the surface of each subsample was sterilised as follows: 30 s in 70% ethanol, 2 min in 3.5% NaOCl, and 30 s in 70% ethanol [43]. The subsamples were then cultured in Petri dishes containing malt extract agar medium (MEA) and on plates containing potato glucose agar (PGA) (Britania ® ). Both MEA and PGA culture media were supplemented with 100 mg/mL of streptomycin sulphate, 50 mg/mL of chlortetracycline HCl, and 5 mg/mL of dichloran to prevent bacterial and yeast contamination [44]. Plates were immediately cultured in the dark in a growth chamber at 25 °C and inspected daily over a period of 4 weeks until fresh mycelia were observed. Subsequently, successive isolations were performed to derive pure cultures of the fungal strain. The identification of the phytopathogenic fungus was based first on morphological observations, followed by molecular identification [45]. Genomic DNA was extracted from the mycelium growing on liquid Peptone Malt (MP) medium at 21 days using a DNeasy UltraCkean Microbial Kit (Qiagen ® , Germany), following the manufacturer's protocol. Then, the extracted DNA was purified, and the internal transcribed spacer (ITS) region of the complete rDNA (ITS1, ITS2 and 5.8S) was amplified by PCR. The PCR products were sent to Macrogen Inc © (Seoul, Korea) for purification and sequencing. The strain CC58 was identified as E. microtheca (SINAVIMO #9287) and preserved in the Centro de Investigación y Extensión Forestal Andino Patagónico (CIEPFAP) mycological collection, Chubut, Argentina.

EPF control on L. botrana in vitro
The evaluation of EPF efficacy in vitro and in vivo can be achieved using different application techniques, such as dipping insects in a spore suspension [46,47], using topical micro-applications [48][49][50][51], or spraying EPF formulations directly to the insects or plants [52,53]. In this study, EPF pathogenicity across different stages of L. botrana was tested using a micro-application technique, which ensures a higher level of precision, accuracy, and reliability of the pest susceptibility [54], and is useful to compare insect susceptibility to EPF among different developmental stages. A laboratory experiment was performed to examine the effects of three EPF strains on the mortality rate of L. botrana using 240 fifth instar larvae (L 5 ), 240 pupae (P P ), and 240 adults (A d ) of L. botrana using a topical micro-application technique [50]. EPF were applied to un-cocooned pupae [6,55,56]. The experiment comprised four groups: (1) insects treated with EPF strain CEP413, (2) insects treated with EPF strain CEP589, (3) insects treated with EPF strain CEP591, and (4) control group (insects treated with an EPF-free solution). Treatment groups were topically treated [50,54] with a 1 µL drop (1 × 10 6 c/mL) of the respective strain using a Hamilton ® micro dispenser (7000 Series Syringes ® ). The control group was treated with 1 µL of distilled water. Mortality rates were measured using a subset of 20 individuals per replicate. After fungal application, the L 5 , P P , and A d of L. botrana were carefully transferred to sterile 90-mm Petri dishes lined with sterile filter paper (Whatman ® ). Immediately, dishes were incubated in growth chambers at 27 °C. Considering the high competition for food among larvae [57], an artificial diet [40] was offered ad libitum during the study period. Adults were carefully inoculated immediately after pupal eclosion and transferred to plastic trays (20 cm × 10 cm × 15 cm). The adults were nourished during the experimental period with a 5% ascorbic acid solution provided using a plastic syringe [41]. Since mortality over long time periods is not representative of field conditions, the mortality rates of L 5 , P p , and A d of L. botrana were measured up to the 7th day of the experiment. Dead European grapevine moths were carefully removed from the Petri dish to avoid horizontal transmission. To prevent adults from escaping from the trays, dead individuals were not removed until the end of the experiment. Abbott's equation [58] was used to obtain the corrected mortality (CM).

EPF control on L. botrana in the field
Efficacy of EPF on the control of the larval stage of L. botrana was tested in the field. Based on the corrected mortality (CM) observed in the laboratory experiments, we selected one M. anisopliae strain (CEP591) for field trials. The trial was repeated across different seasons over 1 year (September 2018-March 2019): (S 1 ) during spring, when plants displayed well developed inflorescences, (S 2 ) during early summer, when plants had undeveloped bunches, and (S 3 ) during late summer, which comprised the vine veraison stage. Trials were carried out in an experimental field ("Las Mellizas") located in San Juan, Argentina (31° 58′ 58″ S; 68° 26′ 28″ W), which is planted with V. vinifera L. cv. Merlot since 2011. The field trial was performed within an approximate area of 4.63 ha. No chemical or biological insecticides were added to vines 45 days before trials or during trials. To avoid pupation during trials, we performed the field experiments with newly emerged L 3 -L 4 larvae of L. botrana. Trials consisted of two treatments (EPF+ and EPF−), with six replicates each treatment. Micro-application technique was used. For the EPF+ treatment, 20 larvae were topically treated with 1 µL drop (1 × 10 6 c/mL) of CEP591. The control group (EPF−) were treated with 1 µL of sterile distilled water. After larvae were topically treated, they were immediately brought to the experimental field and carefully placed on selected grapevine bunches in groups of 20 treated individuals per bunch. The grapevine bunches were then immediately covered with a fabric mesh of 500 × 500 μm of pore diameter. Fabric bands were then attached from the top of the mesh to the rachis of each bunch to enclose the experimental unit. This helped to maintain a closed system, preventing the entry or exit of insects into the experimental unit. Treatments were spatially separated by at least 20 m. After 7 days, bunches were carefully removed from plants and transferred to the laboratory in closed plastic containers to determine larvae mortality. Environmental temperature (T°) and relative humidity (RH) during the three seasons were obtained from an automatic weather station (Davis Instruments ® -Mod. Vantage pro-2).

Eutypella microtheca growth assessment
We examined the inhibitory rate of the three strains of EPF (CEP413, CEP589, and CEP591) on the growth of E. microtheca strain (CC58). CC58 was used in this study because it was the most abundant phytopathogenic strain affecting grapevine plants at the study site (see above for detailed geolocation information). CEP413, CEP589, and CEP591 were co-cultured with CC58 in separate Petri dishes containing potato glucose agar (PGA) media (Britania ® ). To achieve this, a 5-mm mycelial disc, obtained from the edges of a 10-days-old culture of CC58, was placed at the centre of a dish containing 20 mL of PGA. Immediately, four discs of one EPF strain (from 10-daysold cultures) of the same diameter as those of the CC58 strain were placed carefully at four sites at approximately 3 cm from the centre of the Petri dish [59]. The Petri dishes were inverted to prevent conidia from falling onto the agar medium and then incubated in the dark at 28 ± 2 °C. The growth of the E. microtheca colony was measured every 96 h on two perpendicular axes over the following 20 days under a stereomicroscope (Lancet Instruments ® , Model ZTX-30Y-C2, China) using a digital calliper. The control was determined by measuring the radial growth of E. microtheca strain growing in isolation (agar discs without EPF) on separate Petri dishes (indicative of the potential growth of the E. microtheca strain). The percentage of growth inhibition (GI) was calculated with the following formula: GI (%) = ((A − B)/A) × 100, where A represents the radial growth (mm) of the control treatment, and B represents the radial growth (mm) of the pathogen with the EPF setup [60]. Three replicate plates were prepared for each phytopathogenic/EPF strain combination and for the control treatment.

Statistical analyses
Effects of CEP413, CEP589, and CEP591 on mortality rate of larvae (L 5 ), pupae (P p ), and adults (A d ) of L. botrana in vitro were analysed using a one-way ANOVA (independent factor: EPF strains, response variable: corrected mortality (CM)). Efficacy exhibited by CEP591 over L. botrana in the field was analysed using a one-way ANOVA (independent factor: seasons, response variable: CM). Antagonistic effects of EPF on E. microtheca growth were also analysed using a one-way ANOVA (independent factor: EPF strains, response variable: pathogen inhibition growth (%). A post hoc Fisher's LSD test was performed to analyse differences among treatments. All statistical analyses were performed using Infostat ® statistical software [61].

Biological control of EPF on L. botrana in vitro
All M. anisopliae strains tested in this study were able to infect L. botrana at any developmental stage, including larvae (L 5 ), pupae (P p ), and adults (A d ) (Fig. 1a-c,  respectively). No significant differences were observed among EPF strain efficacy for L 5 (F = 0.22, p = 0.8097) and A d (F = 0.12, p = 0.8905). Nevertheless, for P p stage significant differences in EPF strain efficacy were observed (F = 33.25, p = 0.0006). CEP591 was observed to be more effective (99.9%) than CEP589 (81.6%) and CEP413 (79.98%) in controlling the pupal stage of L. botrana (Fig. 2). Overall, the mortality of the control group was lower than 15% in all cases.

Biological control of EPF on L. botrana in the field
At field conditions, larvae of L. botrana were susceptible to EPF infection across all seasons. Nevertheless, L. botrana mortality was significantly affected by the season (F = 6.92, p = 0.0074). Overall, larvae of L. botrana were more susceptible to CEP591 at early spring (91%) and early summer (81.5%) than in late summer (Table 1). In late summer, CM drastically decreased to 64.9%. Mortality of control larvae was relatively constant across seasons (F = 1.11, p = 0.3551), which rounded 5%.

Inhibitory effects of EPF on E. microtheca
The growth of E. microtheca was significantly affected by the EPF strains (F = 151.49, p < 0.0001). There was a significant decrease in the growth of E. microtheca treated with CEP413, CEP589, and CEP591 compared with that of the control group (Fig. 3). The inhibitory percentage of the strain CEP413 (65.2% ± 12.24) was significantly higher than those of CEP591 (55.07% ± 10.11) and

Discussion
Strains of M. anisopliae were highly effective to controlling two detrimental agents of V. vinifera L. in vitro and in vivo. EPF strains showed high mortality rates in all developmental stages of L. botrana and reduced the growth of the phytopathogenic fungus E. microtheca.
Overall, the three M. anisopliae strains were highly effective to controlling L. botrana at the L 5 , P p , and A d stages. The average corrected mortality (CM) of larvae treated with the EPF strains was over 75%, while that of treated pupae and adults was over 85%. However, the CM of pupae treated with CEP591 was significantly higher than that with the other two strains, causing approximately 100% of mortality. This result is particularly important given that the pupal stage is highly resistant to fungal infection, at least in some Lepidoptera [62], Coleoptera [63], and Blattodea [64] species. Consistent with our results, previous studies have also shown the effectiveness of EPF strains in controlling immature stages of L. botrana, including larvae and pupae [32]. Similarly, recent findings [65] showed that Metarhizium and Beauveria, which were isolated from soils under vineyards in Argentina, were effective against the fifth-instar larvae of L. botrana in vitro. Additionally, several studies have shown that Metarhizium is a successful EPF genus in the biological control of some lepidopteran pests in vitro, such as Spodoptera litura, S. frugiperda, Tuta absoluta, and Plutella xylostella [66][67][68][69], suggesting that strains of this genus have the potential to be used as biopesticides against several lepidopteran species. Previous studies on the effectiveness of EPF strains against L. botrana have been particularly focused on a single developmental stage of the insect's life cycle [55,65,70,71]. In this study, we have shown the strong effectiveness of M. anisopliae strains against different developmental stages of L. botrana, including larvae, pupae, and adults in vitro. These results agree with other studies, which have demonstrated the role of EPF strains as an efficient method controlling all developmental stages of different insect pests such as Megalurothrips sjostedti [72], S. litura [68], Dermanyssus gallinae [73], and Tetranychus urticae [74]. Differences in EPF effectiveness between the larval and pupal stage of L. botrana may be associated with differences in the ecdysis time between both stages. Larvae, in contrast to pupae, may have shorter time intervals between successive ecdysis, which has been reported to be an important factor in insect resistance to fungal infection [75,76].
In the field, the strain CEP591 was shown to be effective in killing L. botrana larvae across seasons. Nevertheless, efficacy of CEP591 in controlling L. botrana in the field was dependent on the season. In late summer, it was observed that CEP591 was less pathogenic to L. botrana, causing a mortality of around 64% compared to the other two seasons, in which larval mortality ranged  Table 1 Evaluation of EPF at field conditions Corrected mortality (CM) ± SD (standard deviation) for L 3 -L 4 larvae of Lobesia botrana caused by Metarhizium anisopliae strain CEP591 at field conditions. All trials were performed in vivo (Vitis vinifera L. cv. Merlot) on an experimental field located in San Juan (Argentina) (31° 58′ 58″ S; 68° 26′ 28″ W). The trial was initiated at early spring (S 1 ) and repeated at early summer (S 2 ) and late summer (S 3 ), comprising a complete vine productive season. Letter's a, b indicates significant differences between seasons (Fisher's LSD; p < 0.05)  [77][78][79]. For example, Rodríguez et al. [77] observed that the M. anisopliae dose required to kill 50% of Varroa destructor (Acari) in vitro was considerably lower (3.8 × 10 5 c/mL) to that required under natural conditions (5 × 10 10 c/mL). Similar observations were found by Altimira et al. [55] using the EPF Beauveria pseudobassiana to control L. botrana in vitro and in vivo. This efficacy variation of EPF in vitro versus in vivo may be associated with abiotic factors, including temperature, relative humidity, and UV radiation, which are known to influence some physiological insect processes such as conidia germination [55,80,81].

EPF strain CM (%)-S 1 CM (%)-S 2 CM (%)-S
The results here showed that the three M. anisopliae strains were also effective inhibiting the E. microtheca growth when co-cultured during 20 days at 28 °C in the dark. The three tested strains significantly inhibited the growth of E. microtheca by over 50%. The strain CEP413 was, however, more effective than CEP589 and CEP591 in inhibiting the proliferation and growth of E. microtheca over time. Different EPF species, such as Beauveria bassiana, Metarhizium brunneum, M. anisopliae, Lecanicillium lecanii, and Isaria javanica [33,34,82,83] have been identified as suitable candidates for controlling several phytopathogenic fungi, including Rhizoctonia solani, Pythium myriotylum [84], Sphaerotheca fuliginea [85], and Botrytis cinerea [86]. However, to the best of our records this is the first study reporting an antagonistic effect of M. anisopliae over E. microtheca, showing that M. anisopliae strains are also appropriate candidates for use in biological control of one of the most important pathogens of V. vinifera L.

Conclusions
The present study revealed that three strains of M. anisopliae were highly effective infecting and killing different developmental stages of L. botrana in vitro and in vivo. Additionally, the same EPF fungal strains were able to inhibit the growth of the phytopathogen E. microtheca in vitro. Our results showed that M. anisopliae strains have the potential to be used in a program of integrated pest management aimed to control different detrimental agents of vines. Additional research, however, is