Susceptibility of Duponchelia fovealis Zeller (Lepidoptera: Crambidae) to Soil-Borne Entomopathogenic Fungi

Duponchelia fovealis (Lepidoptera: Crambidae) is an invasive species that has had a large impact on strawberry crops in Brazil. Pesticides have had limited effectiveness and the use of biological control agents to improve its management is the most appropriate approach. The aim of this study was to evaluate the pathogenicity and virulence of entomopathogenic fungi—isolated from soil—against Duponchelia fovealis larvae under laboratory and greenhouse conditions. Pathogenicity screenings were performed for twenty isolates from Beauveria bassiana, Beauveria caledonica, Isaria javanica, Metarhizium anisopliae, and Lecanicillium sp. against third instar larvae of D. fovealis at the concentration of 109 conidia·mL−1. Lethal concentration (LC50) and lethal time (LT50) were determined for the most pathogenic isolates and for one commercial mycoinsecticide. Mortality rates varied from 10 to 89%. The isolates B. bassiana Bea1, Bea110, Bea111 and I. javanica Isa340 were the most pathogenic. The most virulent isolates were B. bassiana Bea111 and I. javanica Isa340 with LC50 values of 2.33 × 106 and 9.69 × 105 conidia·mL−1, respectively. Under greenhouse conditions, the efficacy of LC50 of the isolates I. javanica Isa340 and B. bassiana Bea111 were 45% and 52%, respectively. Our results indicate that these isolates are strong candidates for application in the control of D. fovealis. This study is the first evaluation of soil-borne entomopathogenic fungi against D. fovealis.


Introduction
Duponchelia fovealis Zeller (Lepidoptera: Crambidae) is a native species from the marshlands of southern Europe and has become one of the most destructive pests of greenhouse and strawberry crops in several countries, including Portugal [1], Italy [2], Turkey [3] and Brazil [4]. The larvae cause economic damage by feeding on the leaves, flowers, and crown, which results in wilting, collapse, and death of the plants and consequently reducing crop quality and yield [4]. Due to the need for daily picking of strawberry fruit and with the increasing awareness of problems associated with pesticide use, development of alternative control strategies have become the main goal in the integrated pest management (IPM) of pests such as D. fovealis [5][6][7]. Biological control is one of the most promising options [4,6,8], although there have been few studies on its application against D. fovealis. Entomopathogenic fungi (EPF) have attributes that make them strong candidates for use against this pest [9]. They are widespread in terrestrial environments, especially in the soil, which is considered a reservoir [10], and are important natural regulators of insect populations through infection by contact, consequently increasing the potential for epizootics and mortality rates in the pest population [11]. Fungi of the genera Beauveria, Metarhizium, and Isaria are widely known as biological control agents [12,13] and they are safe for farmers, consumers, and the environment [14][15][16]. Thus, preventive methods through integrated pest management can be used as an alternative to the chemical control of D. fovealis.
Traits such as virulence and pathogenicity are considered important properties of entomopathogens used in pest control [17]. Although entomopathogenic fungi have been widely used against a number of insect pests, no data are available on the efficacy of fungi against D. fovealis. This study aims to contribute to the implementation of an improved IPM program for this pest, with the first report of tested EPF isolated from native forest and crop systems soils against D. fovealis larvae under laboratory and greenhouse conditions.

Fungal Isolates and Insects
EPFs were isolated from soil samples collected between January and June 2015 from different geographical sites in the state of Parana, Brazil ( Table 1). The isolates are deposited at the Paraná State Microbiological Collection Network-TAXONline (CMRP), Department of Basic Pathology, Federal University of Paraná, Curitiba, Brazil. Tests were conducted on third instar larvae of D. fovealis obtained from a stock colony reared on an artificial diet and maintained under controlled conditions of 25 ± 2 • C, 70 ± 10% relative humidity (RH), and a 14-h photophase [18].

Production of Fungal Conidial Suspensions
Each isolate (Table 1) was cultured on Sabouraud dextrose agar (SDA) medium and incubated at 25 • C in complete darkness for 15 days until sporulation was abundant. Conidia from each isolate were harvested under sterile flow conditions using a sterile steel spatula to scrape them into 30 mL glass tubes with 20 mL of 0.3% (v/v) Tween ® 80 solution. Conidial suspensions of each isolate were vortexed for 15 min at 120 rpm and then filtered, using sterile gauze (10 × 15 cm), into new 30 mL glass tubes. Suspensions were prepared independently for each assay and concentrations were adjusted by serial dilution to achieve a comparable number of conidia per milliliter in an aqueous 0.3% Tween ® 80 solution. Spore suspensions were used immediately after preparation and concentrations were adjusted using a hemocytometer. Spore viability was determined by direct counting [19,20]. Viability was then determined using a microscope (400× magnification) by visualization of 200 spores; spores that presented germinative tube growth were considered as viable. This procedure was adopted for all isolates and viability ranged from 90 to 95%.

Pathogenicity Bioassay
Strawberry leaves were disinfected superficially [21] and their petioles were wrapped in cotton and placed in a 20 mL glass tube with autoclaved sterile distilled water and then positioned in acrylic boxes (11 × 11 × 3.5 cm). The experimental design was completely randomized. Each treatment was conducted with five replications. Twenty-four third instar D. fovealis larvae were transferred to each box, so that each box was considered a replication, for a total of 120 larvae tested per treatment. A Sagyma SW776 airbrush (10 lb pol −1 ) was used to spray 1 mL of a conidial suspension at a concentration of 10 9 conidia·mL −1 and Tween ® 80 at 0.3% over the leaves and larvae. The control treatment received 1 mL of sterile distilled water with 0.3% (v/v) Tween ® 80. After spraying, boxes were kept under controlled conditions (25 ± 2 • C, 70 ± 10% RH, and 14-h photophase). Mortality was assessed after 7 days. Moribund individuals or those that did not respond to touch with a paintbrush were considered dead. These specimens were transferred to Petri dishes with moistened filter paper until fungal extrusion to check for postmortem sporulation. The experiment was carried out for 5 weeks.

Virulence Bioassay and Determination of LC 50
Fungal strains that caused high mortality rates of third instar larvae D. fovealis were selected to determine lethal concentration (LC 50 ) and lethal time (LT 50 ). Five concentrations were tested: 1 × 10 4 ; 1 × 10 5 ; 1 × 10 6 ; 1 × 10 7 , 1 × 10 8 conidia·mL −1 . For each concentration, 120 larvae were used, divided into five replicates. The control treatment consisted of sterile distilled water added 0.3% (v/v) Tween ® 80. A commercially available biopesticide (Bovemax ® ) was used for comparison, containing B. bassiana CG716 strain as the active ingredient. Spraying was conducted as described for the pathogenicity bioassay and the samples were monitored daily for 7 days. Mortality assessments were also conducted as described above.

Greenhouse Assay
The fungi that presented the lowest LC 50 (between 1 × 10 5 and 1 × 10 6 conidia·mL −1 ) in the laboratory bioassays were selected for the greenhouse assay. Healthy 3-month-old strawberry seedlings of the 'Albion' variety were transferred into 500 mL containers with the Plantmax ® potting mix. The plants were kept under natural light and watered twice a day. During the assay, minimum, maximum, and 24-h mean temperatures in the greenhouse were recorded daily with a digital thermometer (Incoterme). The experimental design was completely randomized. Each treatment was tested on 20 seedlings, each considered a replicate. The plant leaves were sprayed with 33 mL of fungal suspension of the selected isolates using a CO 2 precision sprayer (Teejet XR11008VS, Spraying Systems Company, Wheaton, IL, USA). This volume corresponds to 200 L ha −1 with a density of 60,000 plants, as recommended by manufacturers of most commercial biopesticides. The control was sprayed with sterile distilled water containing Tween ® 80 at 0.3% (v/v). Spraying was carried out from the bottom upwards to ensure coverage of all surfaces. Each plant was then infested with 10 third instar larvae of D. fovealis and isolated in cages covered by voile fabric. To prevent insects from escaping through the soil, they were isolated by a transparent laminated plastic disc (30 cm) covering the entire surface of the pot with a small central opening for the plant stem. Mortality was assessed after 7 days.

Statistical Analyses
In the pathogenicity test, mean mortality rates (%) were used to calculate fungal efficiency according to Abbott's equation [22] and the treatments were compared using Tukey's test (p < 0.05). Estimated LC 50 and LT 50 were calculated by Probit analysis using PoloPlus software [23]. For the greenhouse assay, the mean number of insects per strawberry plant was analyzed with ANOVA and Tukey's test (p < 0.05). When differences were significant compared to the control group, efficacy was calculated according to Henderson and Tilton's formula [24].

Pathogenicity Bioassay
Mortality of D. fovealis larvae varied between 10% and 89% ( Figure 1). Of the 20 fungi tested (Table 1) the entire surface of the pot with a small central opening for the plant stem. Mortality was assessed after 7 days.

Statistical Analyses
In the pathogenicity test, mean mortality rates (%) were used to calculate fungal efficiency according to Abbott's equation [22] and the treatments were compared using Tukey's test (p < 0.05). Estimated LC50 and LT50 were calculated by Probit analysis using PoloPlus software [23]. For the greenhouse assay, the mean number of insects per strawberry plant was analyzed with ANOVA and Tukey's test (p < 0.05). When differences were significant compared to the control group, efficacy was calculated according to Henderson and Tilton's formula [24].

Virulence Bioassay
The proportion of insects infected increased with the conidial concentrations tested in D. fovealis larvae. There was a significant mortality of D. fovealis larvae among isolates for each concentration,

Greenhouse Assay
Isolate Bea111 (B. bassiana) and Isa340 (I. javanica) significantly reduced insect populations. Larval mortality differed significantly between treatments and the control group (F 2,48 = 12.04, p < 0.05). The first isolate reduced larval infestation by 52% compared to the control group while the second isolate caused a 45% reduction in the initial population. Fungal extrusion occurred on average after 5 days (Table 3)

Discussion
Several species of entomopathogenic fungi were evaluated for their potential as pest control agents, with promising results of different pests for several crops that have already been extensively studied [25][26][27][28]. However, studies on the biological control of D. fovealis in strawberry crops are new. Our results have indicated that among the 20 surveyed isolates that are pathogenic for D. fovealis, the mortality rates varied between 10% and 89%. Such variation in mortality rates is expected when several isolates obtained from the environment are tested. A similar result was observed in a previous in vitro study using endophytic fungi against D. fovealis larvae, with mortality rates varying between 30% and 80% (10 9 conidia·mL −1 ) [9], reinforcing the argument that the use of entomopathogenic fungi may be effective against this pest. Another study that evaluated B. bassiana and M. anisopliae against larvae of Plutella xylostella L. (Lepidoptera: Plutellidae) also obtained mortality rates varying between 20% and 94% (10 8 conidia·mL −1 ) [29].
Although Beauveria caledonica (Bea110) produced over 57% mortality in the pathogenicity tests, this isolate did not present high virulence on D. fovealis, although it had a shorter LT 50 of approximately 5 days. This might be due to the ecology of this species since B. caledonica presents a more restricted host relationship with coleopterans. Beauveria caledonica is naturally found in soil and has already been recorded as being highly pathogenic to weevils with mortality rates between 70% and 90% (10 7 conidia·mL −1 ) and an average LT 50 between 6 and 8 days for Hylobius abietis [30].
In our study, Metarhizium anisopliae isolates induced mortality rates between 10% and 43%, unlike studies on biological control with isolates of M. anisopliae, which reported high mortality rates [31,32]. The low performance of M. anisopliae may be due to the selectivity of certain strains to particular hosts as targets [33]. Other studies using other Lepidoptera species obtained similar results [29].
The isolate Bea111 presented a low value of LC 50 compared with the commercial strain Bovemax ® . The LC 50 found in the present study was similar to other virulence studies using B. bassiana where an average LC 50 between 10 4 and 10 6 conidia·mL −1 was found [25,34].
Isaria javanica Isa340 presented the lowest LC 50 value among all evaluated isolates, reinforcing the great potential of this entomopathogenic genus for pest control. The genus Isaria is considered highly virulent for hosts of several orders of insects [35]. Studies on Isaria fumosorosea against larvae of Plutella xylostella (Lepidoptera: Plutellidae) [36] and Isaria poprawskii against nymphs of Homalodisca vitripennis (Hemiptera: Cicadellidae) [37] showed LC 50 close to the values observed in our study.
Most entomopathogenic fungi perform notably well under laboratory conditions and high levels of mortality can be observed for most insect targets under such controlled conditions. An unfortunate consequence of such reports is that this establishes unrealistic expectations regarding the performance of these agents in the field. Variations in temperature, humidity, and exposure to UV-irradiation are amongst some of the important abiotic factors that can significantly impact the efficacy of entomopathogenic fungi in field applications [15]. Our results under greenhouse conditions also confirmed the performance of isolates Bea111 and Isa340, which induced mortality rates between 45% and 52%. In addition, the behavior presented by larvae of D. fovealis, which lodges itself at the base of the strawberry plants [4], along with the plant structure itself, may construct a favorable habitat that contributes to the effectiveness of the pathogens [38]. Several species of entomopathogenic fungi have been evaluated as pest control agents under laboratory and greenhouse conditions or directly in the field, with promising results in different pests of several crops that were already widely studied [25][26][27][28]39]. However, this is the first report on evaluations of soil-borne entomopathogens against D. fovealis in these conditions.

Conclusions
In conclusion, our results indicate that isolates Beauveria bassiana Bea111 and Isaria javanica Isa340 are strong candidates for further evaluation in the control of Duponchelia fovealis and present good potential for the development of products as mycoinsecticides. Our results represent the first report of entomopathogenic fungi isolated from soil infecting D. fovealis, and the basis for future studies to design a biological control program in strawberry crops.