Single and combined effects of Drosophila suzukii and Drosophila melanogaster on sour rot development in viticulture

Sour rot is a disease complex that causes serious damage in viticulture. The common vinegar fly Drosophila melanogaster (Diptera: Drosophilidae) is associated with sour rot in overripe or otherwise damaged grapes. Drosophila suzukii (Diptera: Drosophilidae) is an invasive species, which is suspected to induce sour rot in previously undamaged grapes due to the flies' ability to infest healthy, undamaged soft fruits with its serrated ovipositor. As a consequence, infection of healthy grapes by D. suzukii may facilitate the colonization by D. melanogaster. We investigated the single and combined effects of D. suzukii and D. melanogaster on sour rot development by measuring volatile acidity under near‐natural conditions in the vineyard, along with laboratory experiments under controlled climate. In 2017, the combined field and laboratory experiments suggested that the presence of D. suzukii and D. melanogaster increased the volatile acidity levels at a similar rate. In 2018, the field experiments showed an only marginal increase in sour rot development in treatments with both Drosophila species. Under more favourable laboratory conditions, the presence of D. suzukii, but not D. melanogaster triggered sour rot emergence. A facilitating effect of D. suzukii infestation for D. melanogaster was not detectable. These findings suggest that D. suzukii does in fact have the potential to trigger sour rot, but will probably rarely do so under field conditions in the vineyard, at least in the studied region. Instead, our study showed that D. melanogaster can have a similar impact on sour rot development as D. suzukii, emphasizing the need of comparative studies.

Vinegar flies (Drosophila spp.) are known to trigger sour rot development (Barata, Malfeito-Ferreira, et al., 2012;Barata, Santos, et al., 2012;Hall, Loeb, Cadle-Davidson, et al., 2018) and recent research showed that vinegar fly regulation with the help of insecticides can reduce disease severity . Different scenarios of Drosophila spp. contributing to sour rot development are proposed. First, the flies are able to vector the yeasts and bacteria associated with the disease (Barata, Santos, et al., 2012;Ioriatti et al., 2018). More importantly, a recent study with axenic flies demonstrated that flies also play a non-microbial role, presumably through the developing larvae, which themselves trigger the decomposition of berries and thereby reinforce sour rot development (Hall, Loeb, Cadle-Davidson, et al., 2018).
However, the common vinegar fly, Drosophila melanogaster (Diptera: Drosophilidae) is able to oviposit only in overripe, decaying or in other way damaged fruit (e.g., cracked grapes after rain or insect feeding). In contrast, Drosophila suzukii Matsumara (Diptera: Drosophilidae), an invasive vinegar fly species from Asia, introduced to North America and Europe since 2008, is able to infest healthy, undamaged fruit with its serrated ovipositor (Asplen et al., 2015;Atallah, Teixeira, Salazar, Zaragoza, & Kopp, 2014;Hamby & Becher, 2016;Ørsted & Ørsted, 2019;Schetelig et al., 2018;Walsh et al., 2011). It is a major pest of cherries, raspberries and blueberries, but can also infest some, primarily soft-skinned, grape varieties (Entling, Anslinger, Jarausch, Michl, & Hoffmann, 2019;Ioriatti et al., 2015;Kehrli, Linder, Cahenzli, & Daniel, 2017;Lee et al., 2011;Shrader, Burrack, & Pfeiffer, 2019). In viticulture, it is still debated whether D. suzukii is able to induce sour rot disease. Rombaut et al. (2017) confirmed this ability in laboratory experiments. Moreover, they hypothesize that D. melanogaster in turn is able to oviposit into berries damaged by D. suzukii, thus reinforcing disease severity. Thus, oviposition by D. suzukii may facilitate D. melanogaster access, leading to a larger damage by the combination of both species than would be expected by their single effects. However, experiments were performed with sterilized berries, which were afterwards dipped in sour rot extract and placed in cages in the laboratory. Thus, these results are not necessarily transferable to the field situation.
The aim of this study was to investigate the single and com- We expected to find higher levels of sour rot damage in treatments with vinegar flies than in the control. As D. suzukii is able to attack also undamaged grapes, we hypothesized that the treatments comprising this species would be infested heavier than the treatment with only D. melanogaster. If D. melanogaster is able to use the pre-damaged grapes from D. suzukii oviposition, we finally expected the combined treatment (with both species) to have even higher volatile acidity levels than the treatment with only D. suzukii.
The climatic chamber was set to 23°C, 75% relative humidity and a photoperiod of L16:D8h. used in the experiment were between 1 and 60 days old, reflecting the variable age that can be found in natural populations. As sex determination in D. melanogaster is relatively time-consuming and would stress the experimental animals, we decided to add the individuals of both species without sex determination, but in sufficient numbers to ensure egg-laying. Retrospective sampling from our cultures revealed a mean number of 26 females per 50 individuals, with no significant difference in the sex ratio or its variability between species (95% confidence interval: 20-32 females per sample). Both species are able to lay at least five eggs per female per day (Emiljanowicz, Ryan, Langille, & Newman, 2014)  In addition to the field experiments, we performed two laboratory repetitions starting at 29.8.2018 and 4.9.2018. Therefore, we randomly collected grape clusters from the same vineyard where the field experiment took place. In order to control the condition of the experimental grape clusters, we removed any damaged berries from the grape clusters prior to the experiments. We individually weighed the grape clusters, placed one grape cluster per sample in a plastic box with a gauze lid (11.5 × 15.5 × 12.5 cm) and added flies according to the treatment. We chose the same experimental set up as in the field experiments with four treatments repeated 10 times and distributed them randomly in the climatic chamber with 23°C, 75% relative humidity and a photoperiod of L16:D8h. After 2 weeks, we determined volatile acidity of the grape clusters at 12.9.2018 and 29.9.2018, respectively.

| Statistics
We performed all analyses using the open-source program R (R Core Team, 2017). For all experiments, we used n = 40 samples (four treatments × 10 replicates) to check for differences in volatile acidity levels between the treatments performing variance analyses followed by post hoc tests. We checked and controlled for a potential influence of the grape cluster weight by adding it as an additional depended variable in the model. However, as the effect of grape cluster weight was only significant in the 2017 data set, we did not consider grapes' weight in the 2018 data sets. We used diagnostic plots to estimate the quality of the model. If diagnostics plots were not optimal, we rank transformed the volatile acidity data (combined field and laboratory data 2017, laboratory data 2018). We checked for model robustness with permutation tests using the pgirmess package (Giraudoux, 2017).

| RE SULTS
In both repetitions of the combined field and laboratory experi- In the 2018 field experiment, 3 weeks after adding the flies, volatile acidity levels were far below the legal limit of 1.2 g volatile acidity/1 L must (Lemperle, 2007)   Unfortunately, the weather was exceptionally hot and dry during the ripening period in 2018 in our region, which is generally known to reduce reproductive success of D. suzukii Ryan, Emiljanowicz, Wilkinson, Kornya, & Newman, 2016). Indeed, field studies showed that D. suzukii is able to hide in cooler areas F I G U R E 1 Effects of Drosophila suzukii and Drosophila melanogaster on sour rot development in semi-field experiments 2017 in a Dornfelder vineyard. Grapes were exposed to flies in gauze-sleeves in the field, followed by an incubation in the laboratory. Presence of flies and grape weight increased volatile acidity in both subsequent repetitions (a) set up 30.8.2017, ANOVA: n = 40; treatment: F = 5.02, p = .005; grape weight: F = 4.60, p = .039 and (b) set up 6.9.2017, ANOVA: n = 40; treatment: F = 7.44, p = .0006; grape weight: F = 25.79, p < .0001). Different letters indicate significant differences (p < .05) in volatile acidity between the treatments at the 5% level following the Tukey post hoc test at daytime and becomes active during dusk/dawn avoiding unfavourable, potentially lethal conditions (Evans, Toews, & Sial, 2017;Jaffe & Guédot, 2019;. However, as the flies in our experiments were caged in the experimental bags, they were not able to escape and as a result presumably not able to oviposit at all. For future research it will be interesting if D. suzukii is more damaging in the field in more temperate summers. In the laboratory, D. melanogaster did not increase volatile acidity levels, diverging from the 2017 results. As we used only intact grape clusters without any cracking in this experiment, it seems that D. melanogaster is not able to reproduce and cause any damage under these conditions. However, as we did not check egg-laying and emergence rates, we can only speculate about the exact processes.

| D ISCUSS I ON
Interestingly, the combined effect of D. suzukii and D. melanogaster was not higher than the effect of the single treatments. Rombaut et al. (2017) hypothesized that grape damage induced by D. suzukii facilitates D. melanogaster oviposition leading to a synergistic induction of sour rot. As we added twice as many flies, we expected at least twice as much volatile acidity development. Moreover, if development is possible because of pre-damaged grapes, sour rot damage should be more extensive in D. melanogaster than in D. suzukii, as its oviposition rate is up to ten times higher than in D. suzukii (Asplen et al., 2015;Emiljanowicz et al., 2014). We can imagine different explanations for the lack of such facilitation. Firstly, the injuries produced by the D. suzukii egg-laying may be too small to provide access for D. melanogaster. In addition, D. melanogaster could also avoid oviposition in grapes already occupied by D. suzukii in order to prevent competition. However, predatory cannibalism seem to be a functional behaviour in D. melanogaster larvae and next-generation experiments on raspberries showed that D. melanogaster emergence was not affected by previous egg-laying of D. suzukii, which is in contrast to what we would have expected (Shaw, Brain, Wijnen, & Fountain, 2018;Vijendravarma, Narasimha, & Kawecki, 2013). Another explanation could be that the time may be too short for synergisms to develop, as D. suzukii had to oviposit before D. melanogaster could benefit from the resulting injuries to the grape skin. In this regard, Rombaut et al. (2017)  Different letters indicate significant differences (p < .05) in volatile acidity between the treatments at the 5% level following the Tukey post hoc test appearance in Germany in 2011, and it is still unclear to what degree D. suzukii was contributing to the severe sour rot development, and in how far it was merely benefitting from already high grape damage and humid weather conditions (C. Hoffmann, JKI Siebeldingen, unpublished data). Thus, the risk of D. suzukii for viticulture appears minor, at least in our region. Instead, our study shows that the common vinegar fly D. melanogaster is able to induce sour rot as well, but in contrast to D. suzukii, it is dependent on grapes' physical conditions. Drosophila melanogaster is a well-known amplifier of this disease and winegrowers keep an eye on it at all times (Hall, Loeb, Cadle-Davidson, et al., 2018;. Thus, it requires further investigation to determine under which circumstances D. suzukii can really increase grape damage beyond the damage levels that D. melanogaster can cause. Moreover, we think it is important to keep investigating the combined effect of D. suzukii and D. melanogaster. We could not detect a synergistic induction of sour rot, but we suspect that the experiment was too short for a combined effect to develop. Future research should investigate if this synergistic induction may be possible over longer time periods.

ACK N OWLED G EM ENTS
We thank Thomas Gramm for managing the experimental vineyard, Sonja Anslinger for her help with field and laboratory work and Gertraud Michl for her assistance with must analysis. We are grateful to Florian Schwander and his group for the possibility to use their analytical facilities and Ulrike Braun for technical support with barcoding. We thank Theresa Pennington for language improvement and Martin Entling for statistic advice and valuable comments on earlier versions of this manuscript.

CO N FLI C T O F I NTE R E S T
The authors declare that there is no conflict of interest.

AUTH O R CO NTR I B UTI O N
WE and CH designed research. WE conducted the experiments, analysed data and wrote the manuscript. WE and CH edited and approved the manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
When accepted, the data that support the findings of this study will be openly available at https ://www.opena grar.de/recei ve/opena