Susceptibility of Drosophila suzukii larvae to the combined administration of the entomopathogens Bacillus thuringiensis and Steinernema carpocapsae

Non-native pests are often responsible for serious crop damage. Since Drosophila suzukii has invaded North America and Europe, the global production of soft, thin-skinned fruits has suffered severe losses. The control of this dipteran by pesticides, although commonly used, is not recommended because of the negative impact on the environment and human health. A possible alternative is the use of bio-insecticides, including Bacillus thuringiensis and entomopathogenic nematodes, such as Steinernema carpocapsae. These biological control agents have a fair effectiveness when used individually on D. suzukii, but both have limits related to different environmental, methodological, and physiological factors. In this work, we tested various concentrations of B. thuringiensis and S. carpocapsae to evaluate their efficacy on D. suzukii larvae, when administered individually or in combination by using agar traps. In the combined trials, we added the nematodes after 16 h or concurrently to the bacteria, and assessed larvae lethality from 16 to 48 h. The assays demonstrated a higher efficacy of the combined administration, both time-shifted and concurrent; the obtained data also showed a relevant decrease of the time needed to kill the larvae. Particularly, the maximum mortality rate, corresponding to 79% already at 16 h, was observed with the highest concentrations (0.564 µg/mL of B. thuringiensis and 8 × 102 IJs of S. carpocapsae) in the concurrent trials. This study, conducted by laboratory tests under controlled conditions, is a good starting point to develop a further application step through field studies for the control of D. suzukii.

physiology of the larvae up to septicemia and death 13,14 . The lethality and effectiveness of EPN is closely related to its interaction with the target and the possibility of reproduce itself undisturbed inside the host's body 4,15,16 . In general, both Bt and EPN have weaknesses in their efficacy, such as possible drying caused by rapid climatic variations, and, in particular for Bt, UV sensitivity and a narrow range of specificity 17,18 ; therefore, a reduction in the time lapse between administration and relevant mortality of the target insect population, would help to preserve the integrity of the crops affected by the infestation.
The fly Drosophila suzukii (Matsumura, 1931) (Diptera: Drosophilidae), also known as spotted wing drosophila, is an invasive pest endemic to South East Asia, which has recently invaded western countries 19 , threatening both European and American fruit industry. D. suzukii is one of the most important emerging pests of soft-skinned fruit such as raspberries, blackberries, strawberries, blueberries, cherries, grapes, and peaches 20 . The adult female lays her eggs into the fruit through the ovipositor by perforating the skin, and the larvae develop by feeding on the fruit flesh 21 .
The current control of D. suzukii is mainly based on pesticides, but treatments carried out on the orchards, in addition to the known negative impact on the environment, are also troublesome due to the different regulatory framework between producer countries and international export markets 22 . It is therefore advisable to implement and improve biological control techniques. Various bio-insecticides have been tested against Drosophila spp., such as parasitoids, entomopathogenic fungi, Bt and EPN such as Steinernema carpocapsae (Sc), some of them have exhibited a moderate efficiency in limiting the expansion of this pest [23][24][25][26] .
In laboratory assays, the application of Bt or EPN produces a relevant mortality of the target subjects in times ranging from 48 to 72 h after administration 25,27 . Conversely, in field treatments the times are longer by far 28 .
In this work we investigated the effects of Bt subsp. kurstaki and Sc, administered individually or in combination, on D. suzukii larvae. The individual administration of these bio-insecticides could allow to eliminate a large part of the D. suzukii population, but the time required to obtain the result is such that damage to crops is relevant. For this reason, we designed laboratory tests that, using both Bt and Sc in the same administration (concurrent or time-shifted), would allow to evaluate the combined effects of these bio-insecticides aimed to a possible reduction of the time needed to eradicate the D. suzukii population. We carried out a controlled experimental investigation using administration techniques that were as much compatible as possible with field tests, to ensure that the obtained data reflected natural behavior of target insect and entomopathogens. For this purpose, we used a breeding and feeding system for D. suzukii that allowed the larvae to feed properly by ingesting Bt and supported the EPN locomotion for the active search of the host.

Results
Evaluation of the efficacy of B. thuringiensis on D. suzukii larvae. D. suzukii larvae at early larval stage (L1) were treated with five increasing concentrations of Bt and mortality was monitored at 24 and 48 h. Infections were carried out with the agar-trap method, Bt was added to the soft agar at concentrations ranging from 0.047 to 1.128 μg/mL. The obtained data showed a significant concentration and time-dependent efficacy ( Fig. 1).
Sc lethal concentrations were 933.3 IJs/mL (LC 50 ) and 1655.9 IJs/mL (LC 90 ) at 24 h (Fig. 4a), and 641.5 IJs/mL (LC 50 ) and 1345.7 IJs/mL (LC 90 ) at 48 h (Fig. 4b). The LC 50  www.nature.com/scientificreports/ Effects of B. thuringensis on S. carpocapsae. We carried out in vitro assays to ascertain possible adverse effects of the presence of Bt on Sc viability. Bt (0.564 μg/mL) were co-incubated in agar traps with Sc (5 × 10, 4 × 10 2 and 8 × 10 2 IJs) and the percentage of dead nematodes was evaluated at 24 and 48 h (Fig. 5). The mortality of the treatments did not significantly differ from that of the relative controls (t test, p > 0.05) and not exceed about 3% and 5% at 24 (Fig. 5a) and 48 h (Fig. 5b) respectively. Thus, these results excluded a possible toxicity of Bt against Sc.
Effects of the time-shifted administration of B. thuringiensis and S. carpocapsae. We carried out combined trials with different concentrations of Bt (0.282 and 0.564 μg/mL) and numbers of Sc (4 × 10 2 and 8 × 10 2 IJs). Bt was administered at t 0 and the Sc addition was performed with a delay of 16 h (t 16 ). Figure 6 shows the effects of 0.282 μg/mL of Bt and 4 × 10 2 IJs (Fig. 6a) or 8 × 10 2 IJs (Fig. 6b) of Sc. A further combined assay was performed increasing the Bt concentration to 0.564 μg/mL and administering Sc at the same two amounts, 4 × 10 2 IJs (Fig. 6c) or 8 × 10 2 IJs (Fig. 6d).  www.nature.com/scientificreports/   . The mortality of Sc in all assays did not exceed about 3% (24 h) and 5% (48 h) and was comparable to the controls with Sc alone. These assays were carried out before the combined administration of the two bio-insecticides. C: control, Sc alone; T: treatment, Sc plus Bt. www.nature.com/scientificreports/ Two-way ANOVA results were significant for both factor treatment (F 4,120 = 144.5, p < 0.001) and factor time (F 3,120 = 109.0, p < 0.001), and for their interaction (F 12,120 = 7.7, p < 0.001). The mortality of treatments was always significantly higher than that of controls (Tukey test, p < 0.05, Table S3). The increase in the number of added nematodes, maintaining a constant concentration of Bt, did not result in a significant improvement in the lethal effect against D. suzukii larvae (Tukey test, p > 0.05, Table S3). The obtained data showed instead that, using a higher concentration of Bt, an increase of mortality occurred at 24 h post-treatment a mortality equal to or greater than 50% was observed and almost all D. suzukii larvae were dead after 48 h. From 24 h, the mortality of Bt plus Sc combination 0.282 μg/mL-4 × 10 2 IJs was significantly lower than that of the combinations 0.564 μg/ mL-4 × 10 2 IJs and 0.564 μg/mL-8 × 10 2 IJs (Tukey test, p < 0.01, Table S3). The mortality of Bt plus Sc combination 0.282 μg/mL-8 × 10 2 IJs was instead significantly lower than that of the highest combination 0.564 μg/ mL-8 × 10 2 IJs, at 24 and 32 h (Tukey test, p < 0.01, Table S3). Within all combinations, the mortality significantly increased after 24 h (Tukey test, p ≤ 0.005, Table S3), except for the combinations with the highest Bt concentration, 0.564 μg/mL-4 × 10 2 IJs and 0.564 μg/mL-8 × 10 2 IJs, between 32 and 48 h.

Effects of the concurrent administration of B. thuringiensis and S. carpocapsae. A second set
of trials was carried out by concurrent administration of the two entomopathogens. Bt (0.282 and 0.564 μg/mL) and Sc (4 × 10 2 and 8 × 10 2 IJs) were both added to the agar traps at t 0 (Fig. 7).
Even in this case, two-way ANOVA results were significant for both factor treatment (F 4,120 = 221.3, p < 0.001) and factor time (F 3,120 = 63.1, p < 0.001), and for their interaction (F 12,120 = 5.5, p < 0.001). The mortality of treatments was always significantly higher than that of controls (Tukey test, p < 0.001, Table S4). The increase in the number of added nematodes resulted in a significant increase of D. suzukii larvae mortality only at 24 h for both the tested Bt concentrations and also at 16 h for the highest Bt concentration (0.564 μg/mL) (Tukey test, p < 0.05, Table S4). The mortality of Bt plus Sc combination 0.282 μg/mL-4 × 10 2 IJs was significantly lower than that of the two combinations with the highest concentration of Bt (0.564 μg/mL-4 × 10 2 IJs and 0.564 μg/mL-8 × 10 2 IJs) at all the monitored times (Tukey test, p < 0.05, Table S4), except for 0.564 μg/mL-4 × 10 2 IJs, at 32 h. Significant differences were also observed at all the monitored times for the combination 0.282 μg/mL-8 × 10 2 IJs compared to the combination 0.564 μg/mL-8 × 10 2 IJs (Tukey test, p < 0.001, Table S4). Within all Bt plus Sc combinations,

Comparison between single and combined treatments
Results of the comparison between single and combined administrations at 24 h (ANOVA, F 11,72 = 49.4, p < 0.001) and 48 h (ANOVA, F 10,66 = 33.3, p < 0.001) are shown in Fig. 8 and Table S5. Significant difference of larvae mortality between single treatment with Bt and Sc was observed only at 24 h between 0.564 μg/mL Bt concentration and 4 × 10 2 IJs of Sc (Tukey test, p = 0.0002) (Fig. 8a). Both combined trials (time-shifted and concurrent administration) revealed more efficient than trials with single bio-insecticide, particularly at high concentrations of both Bt and Sc and after 48 h (Tukey test, p < 0.05) (Fig. 8b). Significant differences (Tukey test, p < 0.01) in the efficacy of the combined use of the two entomopathogens between timeshifted and concurrent administration trials were recorded at 24 h (except for the lowest combination 0.282 μg/ mL-4 × 10 2 IJs) but not at 48 h (Fig. 8a,b). These results confirmed that the concurrent presence of the two entomopathogens does not affect their efficacy.

Discussion
Introduced through human activities, such as travels and transcontinental trade, and supported by climate change, several invasive species can settle in habitats where formerly they could not survive. The presence of allochthonous species can negatively affect native species reducing biodiversity, represent a threat to human health and, in the case of phytophagous insects, cause serious damage to the agroeconomic system 29 . Characterized by rapid spread and enormous impact, an invasion like that of D. suzukii has few precedents. This species is thus becoming a model of study in the biology of alien species and for the development of pest management techniques 30 .
The protection of crops from D. suzukii invasion is mainly conducted by means of insecticides 31 which, as known, are unselective, in many cases remain in the environment and, if persistent on fruit, are harmful to human and animal health 32 . In addition, their efficiency can be reduced by weather events such as rainfall, and www.nature.com/scientificreports/ their prolonged use may induce resistance phenomena in target insects. Moreover, the preference of D. suzukii for ripening fruit requires that any treatment be carried out close to the harvesting, which inevitably means that insecticide residues can remain on the marketed fruit 33 . Alternatively to synthetic chemicals, natural compounds that can act as repellents, toxicants by contact or ingestion, and overall deterrents, have been tested mainly against D. suzukii adults 34 . All these control procedures have positive aspects, such as low cost, sometimes good effectiveness, and ease of use for farmers, but, as mentioned, their massive use affects environment and organisms at various levels. Considering the limits of the chemical control, several studies have addressed the problem of D. suzukii management through the development and improvement of biological control methods. However, given the recent introduction of this insect in Europe and America specific projects in these areas are still limited. Until now, only the main entomopathogens commonly used in biological control, have been tested against D. suzukii 35 .
Several serovars of Bt were tested at different concentrations on D. suzukii. Cossentine et al. 27 described the results of the assays with 22 serotypes of Bt, and among them, only few are effective against the target dipteran. In particular, the serovars thuringiensis, kurstaki, thompsoni, pakistani, and bolivia show an effectiveness of more than 75% on first stage larvae. The mortality rates, however, are related to a dietary administration of at least 10 8 spores/mL, below this concentration the insecticidal activity is not relevant. Moreover, the efficacy on adult individuals is extremely low, only the serovar thuringiensis induces a mortality of about 44%. In the literature, data on the low effects of serovar israelensis on this fly are also reported 36 . The data obtained from our trials, carried out with the serovar kurstaki, agree with Cossentine et al. 27 , demonstrating that only high concentrations of the bacterium produce significant effects on larvae in a relatively short time (24-48 h).
As regards the insecticidal activity of EPN, the literature includes several studies carried out in laboratory or field with applications on plants and soil. The data collected using different species of nematodes, such as S. carpocapsae (Sc), S. feltiae, S. kraussei, H. bacteriophora 25,[37][38][39] and the rare species Oscheius onirici 40 , are quite promising but, as observed for Bt, also for EPN high concentrations are required, and their effectiveness is very variable between laboratory and field studies. In accordance with previous results 29 , our current data indicate a good efficacy of Sc: treatments with an amount of 1.6 × 10 3 IJs (corresponding to 80 IJ/cm 2 ) resulted in a larvae mortality above 98%, 48 h post-treatment.
Since the reproduction of D. suzukii occurs by spawning in the mesocarp of the fruit, the technical limit of the application of bio-insecticides such as Bt or EPN cannot be disregarded. Larvae development is protected www.nature.com/scientificreports/ by the fruit pulp and this may be one of the causes of the low effectiveness of Bt in field applications. As known, to be effective, Bt must be ingested and should therefore reach the larva by penetrating inside the fruit; perhaps a timely spraying of the crops before oviposition could theoretically promote the penetration of microorganisms thanks to the drive of ovipositor. Based on these considerations, it is reasonable to assume that coupling Bt with EPN, which, being motile, can actively reach the targets, could significantly improve the effectiveness of D. suzukii control methods in the field. The simultaneous application of two control agents, with different modes of action, may result in additive or complementary effects. In particular, the tissue damage inflicted by Bt to the intestinal epithelium could facilitate the access of nematodes in the passage from the gut to the hemocoel (Fig. 9). When in the hemolymph, the nematode and its symbionts implement strategies of immunoevasion and immunodepression of host immune responses 4,41,42 , leading to a drastic and more rapid physiological alteration that results in the death of the target insect in a shorter time.
As previously reported, and supported by literature, it is desirable that studies on biological control of D. suzukii will be increased and possibly new strategies for the use of bio-insecticides will be tested. In this perspective, we started a project aimed at assessing whether the association of entomopathogenic organisms and microorganisms is a promising strategy, capable of making safer and faster measures to control the diffusion of insect pests. Besides this, a need to reduce Bt concentrations used in field application has arisen from recent observations on the possible insecticidal effect on non-target insect populations, caused by intensive use and consequent bioaccumulation of the bacterium in crop applications 43 . It is also known that high concentrations of Bt serovars which produce the cry1Ba1 and 1 beta-exotoxin toxins, are toxic not only to several dipterans but also to mammals 44,45 .
Even when using EPN, limitations have been highlighted: such as with Bt, they are susceptible to runoff in case of rainfall, have a significant sensitivity to drying at high temperatures, and, depending on the species, are more or less effective in certain environmental conditions as well as susceptible to the presence of pesticides 46,47 .
On the basis of these considerations and the relevant literature, and after excluding adverse effects on the nematode induced by the presence of Bt, we conducted the assays with combinations of Bt and Sc, administering them both simultaneously and time-shifted. According to our data, the mortality of D. suzukii larvae increases from < 45%, in single administration of Bt or Sc, to 65.8%, in the case of combined time-shifted administration (Bt: 0.564 μg/mL and Sc: 8 × 10 2 IJs) at 24 h. These effects are even more significant in the case of trials carried out by simultaneous administration; in this case the mortality within 24 h rises to 91.7%. At 48 h post-treatment, the data show a lethality that rises to 78.3% using the lowest concentrations of the two entomopathogens (Bt: 0.282 μg/mL and Sc: 4 × 10 2 IJs), until reaching the total mortality of the larvae with the highest concentrations (Bt: 0.564 μg/mL and Sc: 8 × 10 2 IJs).
The data at 24 h indicate a clear concentration-dependence of Bt, which at the highest concentrations is extremely effective in all the combined tests. Sc seems to be mainly influenced by the incubation time with the larvae. This could be explained by the different mechanism of action and intake of the two bio-insecticides: Bt is ingested rapidly by the larva during feeding, while Sc must actively seek and penetrate the host; thus, both time www.nature.com/scientificreports/ and concentration could contribute to a greater effectiveness of the nematodes. The administration of Bt at t 0 in both combined trials is justified by the specific action at the intestinal level of this bacterial strain 48 since the damage and possible lesions at the level of the intestinal epithelium can promote the rapid passage of Sc with its symbionts to the hemocoelic cavity (Fig. 9). Moreover, the altered physiology of the larva, induced by Bt, makes the insect more permissive to the action of the nematode. Conversely, anticipating the administration of Sc would not support the action of Bt, as the debilitated larvae may not actively feed, therefore not ingesting an enough number of bacteria. Our data agree with the literature describing the increased efficacy of the use of entomopathogen combinations. Bt plus EPN combination was used against other insect species, such as the Coleoptera Cyclocephala hirta and Cyclocephala pasadenae 49 , the Lepidoptera Spodoptera exigua, Autographa gamma and Plutella xylostella 50,51 , and the Diptera Tipula paludosa 52 and Culex pipiens 53 .
The results of the application of different combined formulations, such as entomopathogens with pesticides, or with products of natural origin, suggest that their concurrent effects improve the effectiveness compared to single uses [54][55][56] . Also, the combination of entomopathogens with pesticides could reduce the needed amount of the chosen pesticide. This kind of trials always requires checking for possible adverse effects of the chemical on the viability of organisms or microorganisms coupled in the formulation and during the administration 57-59 . Our results obtained by the simultaneous administration of Bt and Sc to D. suzukii larvae show a significant increase in the efficacy and a substantial reduction in the time of killing of the larvae. A reduction of the time lapse between administration and effects is particularly important for the control of D. suzukii: since the larvae of this fly prefer fruit close to ripening, so just before harvesting and sale, the speed of neutralization is essential to avoid excessive damage to the harvest that would affect its marketing. Moreover, the combined use of Bt and Sc may allow to reduce the amount of both the bio-insecticides applied in the field, limiting possible ecological issues and maybe improving the cost-benefit ratio.
The trials performed in the laboratory are obviously not transposable in an immediate intervention in the field but provide an essential basis for the possible improvement of methods and technologies that can be used in the field. This is particularly stimulating when the test results, which prelude possible trials on crops, are unequivocally encouraging. However, it is important to consider that the variability of the protocols used by the investigators in laboratory, semi-field and field tests, and the consequent lack of standardized protocols in the experimentation, often make difficult a correct data comparison. Besides this, the great individual variability in the physiology of the organisms used, as both bio-insecticides and target organisms, should be accounted for.
To our knowledge, this is the first work carried out on D. suzukii that assesses the possible combination of B. thuringiensis and S. carpocapsae. The data obtained represent a good starting point to improve the methods of control of the spread of this pest and can stimulate further studies aimed at an extensive screening of the possible combined administrations, fundamental for a future conscious and safe use of biological control methods. Target insect and bio-insecticides. D. suzukii larvae native to Catalonia (NE Spain) were reared on a specific diet in laboratory 60 and maintained in a climate chamber at 25 °C, 45% relative humidity (RH), with a 12:12 h photoperiod. For each assay, only healthy larvae at the first stage (L1) of growth were used.

Chemicals and instruments.
D. suzukii susceptibility to entomopathogens was assayed by administration of Bt, subspecies kurstaki (strain EG 2348), and Sc (strain B14) collected from urban garden soil in Barcelona (NE Spain) and kindly provided by Prof. Garcia Del Pino (Universitat Autònoma de Barcelona). Sc, supplied at the third IJ stage, were dehydrated and mixed with inert material, then stored at 4 °C to keep parasites in cryptobiosis. Before assays, 2-3 gr of the nematodes were dissolved in dechlorinated water and purified from the inert material. The Sc suspension was layered on a discontinuous sucrose gradient (75-50-25%) and centrifuged at 1000×g for 10 min at room temperature; nematodes were recovered at the sucrose 25-50% interface, then washed several times in dechlorinated sterile tap water to remove contaminants, and finally resuspended in water before trials. The number of nematodes used for the trials and vitality control were assessed under a stereomicroscope. Bt was grown in culture media, spores and crystals were isolated as described by Cossentine et al. 27 ; their presence was monitored by phase contrast microscopy. Finally, spores and crystals were freeze-dried and, before use, resuspended in filtered sterile tap water. D. suzukii susceptibility to B. thuringiensis and S. carpocapsae. As a method of infestation in laboratory trials, we used a Petri agar trap to assess the lethality of the entomopathogens against D. suzukii larvae. The agar trap consists of a Petri dish filled with a thin layer of soft agar plus saccharose substrate. The thin layer of agar is transparent enough to allow an easy and accurate evaluation of larvae morphology and vitality. The substrate for the infection was composed of 0.75% agar and 5% sugar in sterile tap water; the solution was heated to 100 °C and then poured (4 mL) into Petri dishes (Ø 5 cm). For Bt trials, when the agar solution was cooled to 37 °C, a suspension containing Bt spores and crystals was added. For Sc assays, Sc suspension (1 mL) was added to the trap when agar was solidified. Twenty D. suzukii larvae were transferred to each agar trap, and incubated D. suzukii larvae susceptibility to the combined administration of the entomopathogens. Before assessing the effects of the combined administration of Bt and Sc on D. suzukii, we verified a possible harmful effect of Bt formulation on the viability of Sc. Thus, co-incubation assays of the two entomopathogens in agar traps were performed in the absence of the target insect. Briefly, 5 × 10, 4 × 10 2 and 8 × 10 2 IJs of Sc were incubated, at 25 °C, in Petri dishes containing 0.564 μg/mL of Bt in agar. The mortality of Sc was evaluated under a stereomicroscope at 24 and 48 h, controls were performed by incubating Sc without Bt.
Combined trials were carried out using Bt at concentrations of 0.282 or 0.564 µg/mL, and 4 × 10 2 or 8 × 10 2 IJs of Sc. The entomopathogens were administered together at time zero (t 0 ), i.e., concurrent administration, or Sc was administered 16 h after the incubation with Bt, i.e., time-shifted administration. The agar traps, containing 20 D. suzukii larvae, were kept at 25 °C, 45% RH, in a climatic chamber. The mortality of the larvae was evaluated at 16, 24, 32 and 48 h, post-incubation. All controls were carried out by incubating D. suzukii larvae without bio-insecticides.
Statistical analysis. For each concentration of Bt and amount of Sc or combination between the two bioinsecticides, as for controls, through the time intervals analyzed (24 and 48 h for single bio-insecticide treatments, and 16, 24, 32, and 48 h for combined administrations), we calculated D. suzukii larvae mortality as average ± standard deviation of six replicates. Lethal concentrations (LC 50 and LC 90 ) of Bt and Sc on D. suzukii larvae were calculated at 24 and 48 h by Probit analysis with 95% confidence intervals.
For each trial (with single bio-insecticide or with a combination of both bio-insecticides), we applied the two-way analysis of variance (ANOVA) considering treatment and time as variables and the percentages of insect mortality as observations. Before the ANOVA, the percentage data were arcsin-transformed. After the ANOVA we used the Tukey HSD post-hoc test to detect possible significant (p < 0.05) differences of insect mortality between different treatments at the same time and between times within the same treatment. Moreover, we applied the one-way ANOVA followed by the Tukey HSD post-hoc test to detect possible significant (p < 0.05) differences of D. suzukii larvae mortality between different single administrations of Bt or Sc and the combined (time-shifted or concurrent) administrations of both bio-insecticides, at 24 and 48 h.
For the assays on the possible effects of Bt on Sc, t test was performed to detect significant (p < 0.05) difference between each treatment and the related control at 24 and 48 h.
The statistical analyses were performed using XLSTAT 2014 and PAST 3.09 software.
Ethics declarations. All applicable international, national and/or institutional guidelines for the care and use of animals were followed.

Data availability
All data are available upon reasonable request to the corresponding author. www.nature.com/scientificreports/