DNA screening of Drosophila suzukii predators in berry field orchards shows new predatory taxonomical groups

Drosophila suzukii (spotted wing drosophila, SWD) is a pandemic quarantine pest that attacks mostly red fruits. The high number of life cycles per year, its ability to rapidly invade and spread across new habitats, and highly polyphagous nature, makes this a particularly aggressive invasive species, for which efficient control methods are currently lacking. The use of native natural predators is particularly promising to anchor sustainable and efficient measures to control SWD. While several field studies have suggested the presence of potential predatory species in infested orchards, only a few confirmed the presence of SWD DNA in predators’ gut content. Here, we use a DNA-based approach to identify SWD predators among the arthropod diversity in South Europe, by examining the gut content of potential predator specimens collected in SWD-infested berry fields in North Portugal. These specimens were morphologically identified to the family/order, and their gut content was screened for the presence of SWD DNA using PCR. New SWD predatory taxonomical groups were identified, as Opiliones and Hemerobiidae, in addition to known SWD predators, such as Hemerobiidae, Chrysopidae, Miridae, Carabidae, Formicidae and Araneae. Additionally, the presence of a spider family, Uloboridae, in the orchards was recorded for the first time, posing this family as another SWD-candidate predator. This study sets important bases to further investigate the potential large-scale use of some of these confirmed predator taxa for SWD control in South Europe.


Introduction
The spotted wing drosophila (SWD), Drosophila suzukii (Matsumura), is a pandemic and highly invasive pest that recently arrived to Europe, with first reports in 2010 occurring in Italy and France (data from the European Plant Protection Organization (EPPO) [1]). SWD integrates since 2011 the A2 List of pests recommended for regulation as quarantine pests [2]. Being extremely polyphagous, this pest is able to develop in a wide range of cultivated and wild The aim of this study was to identify potential SWD predators among the arthropod diversity in North Portugal, using as case study berry fields, and identify field SWD predation. The results obtained should represent an important first step to the management of the pest by supporting populations of natural enemies present in the region and further contribute to the general knowledge of SWD predators in other environments.  40.838972,-8.359917; blueberry) (temperature ranged from �20-30˚C during collection time) (Fig 1). All orchards were developed and maintained under organic farming for 1-5 years upon arthropod collection time; time average size of the orchards was 2 ha, and an average of 6 years of age. All producers used homemade or commercial SWD traps to monitor the pest presence in the orchard; in L2, mass trapping was employed with a D. suzukii attractive bait. Additionally, the producers used mainly Spinosad as a SWD control agent, following all the legal procedures and recommendations. L1, L3 and L4 had wild vegetation between rows, and only L2 and L5 used mulching films-L2 along the blueberry bush rows, and L5 with the complete orchard's soil covered. In L4 and L5 nets were used, but while in L4 the nets were only placed above the blackberry trees in order to provide protection against UV-rays, in L5 the net covered the entirety of the orchard as protection against birds. The orchards' surrounding wild vegetation contained mainly oaks, pine trees, eucalyptus and olive trees; in some cases, there were also vineyards, cherry trees or other fruit trees (L2 and L4). Representative photos of each location are supplied in S1 Fig. The presence of D. suzukii in the orchard was confirmed by collecting the insect in arthropod samplings and/or in traps used by producers.

Arthropod collection and identification
Arthropods were collected in three different rows of the orchard: in the first row of bushes in the orchards' periphery; in the middle of the orchard; and in a row closer to wild vegetation, a water source and/or shade. The wild vegetation present in the last sampled row was characterized by large non-cultivated trees, usually surrounding a small water stream, which contributed to the presence of shade in the nearby rows. On the contrary, the first row in the orchards' periphery did not have any shade or abundant wild vegetation in the surrounding area.
For the arthropod collection, two different methods were used: suction (with a Dietrick Vacuum insect net (D-Vac) machine (Rincon-Vitova Insectaries, Inc., Ventura, CA, US, model 122)), during �3min continuously walking along the row (orchards L1-L5), in which arthropods from the canopy were sucked into a meshed bag [17]; and pitfall traps, in which cups with 50% ethanol were inserted on the soil near the shrubs for 24h, for the collection of soil arthropods [33] (L2-L5) (Fig 2). Arthropods collected using D-Vac were etherized on the field by embedding a cotton ball in ether and placing it on a plastic bag with the meshed bag inside during the transportation to the laboratory; they were preserved at 4˚C in 15 ml tubes with 96% ethanol until identification. The pitfall contents were transferred to 50 ml tubes and kept at 4˚C until arthropod identification.
Collected arthropods were sorted into morphospecies and identified to the order or family level following Triplehorn et al. [34]. Moreover, in accordance with the same authors, arthropods were assigned to feeding guilds and grouped accordingly into four trophic categories: detritivores, phytophages, omnivores and predators. Spider family identification was based on Nentwig et al. [35] and specimens were further grouped following Cardoso et al. [36]. Thus, families of Araneae were grouped in 7 functional groups based on their foraging strategy (type of web and hunting method), prey range (stenophagous or euryphagous), vertical stratification (ground or foliage) and circadian activity (diurnal or nocturnal): (1) sheet web weavers (ShW), (2) space web weavers (SpW), (3) orb web weavers (OrW), (4) specialists (Sp), (5) ambush hunters (AH), (6) ground hunters (GH), and (7) other hunters (OH). After identification, arthropods belonging to predator and omnivores trophic groups were preserved in 96% ethanol, according to place of collection and family/order, and were kept at 4˚C until DNA extraction.

Screening of field predation
Arthropods identified as potential predators were tested for field predation of D. suzukii. Potential predators were washed with a 1.5% bleach solution (v/v, in milliQ H 2 O) to remove any traces of D. suzukii DNA from their body surface. Screening of field predation was assessed by molecular gut-content analysis according to Wolf et al. [10]. DNA of potential predators was extracted with GeneMatrix Tissue Purification Kit (EURX Sp.zo.o., Poland) following the manufacturer's protocol for insect DNA extraction. Arthropods (up to 50mg) were homogenized with PBS (pH = 7,6). The arthropod's DNA was extracted as a pool within the same family, location and sampling method, except for Araneae families' DNA, which was extracted according to functional group and sampling method. DNA quality and quantity were analyzed with μDrop™ (ThermoFischer Scientific, USA) on a microplate spectrophotometer Multiskan Go (ThermoFischer Scientific, USA). Predation was confirmed by PCR amplification with a primer specific for the identification of D. suzukii [10]. PCR amplification was employed twice in order to confirm predation vs. no predation results. The reaction was performed with the highest amount of DNA possible for the enzyme (DFS Taq MasterMix, Bioron Life Science, Germany) in order to avoid false negatives due to lack of DNA quantity in the reaction. The reaction was performed in a T100™ thermocycler (Bio-Rad Laboratories, USA) and the conditions were 94˚C for 2 min, followed by 35 cycles of 94˚C for 10 sec, 48˚C for 20 sec, 72˚C for 15 sec, and a final extension at 72˚C for 3 min. A positive control consisting of DNA extracted from a single D. suzukii was used (previously phenotypically identified according to the EPPO identification key [37]). PCR results were visualized in an 1.4% agarose gel and results were considered positive if a band appeared in the 179bp region.

Data analysis
The number of arthropods identified as potential SWD predators was expressed as a percentage of the number of all arthropods collected. Differences in species richness and diversity were analysed with ANOVA. For the analysis of row diversity, the total number of potential natural enemies both from D-VAC and pitfall sampling locations was used based on each row location within the orchard. All graphs and statistical analysis were made using GraphPad Prism 8 (GraphPad Software, USA).

Arthropod diversity
Different groups of arthropods were collected in four different blueberry orchards and one blackberry orchard (Table 1). In a total of 169 individuals identified as potential SWD predators collected with D-Vac (representing �25% of D-Vac collected arthropods), most belonged to the taxonomic group of Chrysopidae (larva and adults) followed by Araneae, Formicidae, Miridae, Coccinellidae (adults) and Hemerobiidae (adults) (Fig 3a). In the pitfall traps, from a total of 59 arthropods identified as potential predators (�64% of pitfall collected arthropods), most belonged to Formicidae, followed by Araneae, Carabidae and Opiliones (Fig 3b).
We were able to identify 14 different spider families, corresponding to 7 different functional groups (S3 Table), with families related to hunting functional groups being the most abundant in the collections (Fig 4). In all locations, adults of D. suzukii were collected with D-Vac, confirming the pest presence in the field. The highest number of D. suzukii was collected on the blackberry fields, with more than 100 individuals (results not shown). When considering the row of the orchard, most potential predators were collected in the middle rows of locations L1, L2 and L3, but in two locations the number was higher near the vegetation margin (L4) and in the periphery (L5) (Fig 5). Although no statistical differences were found between rows concerning species richness and diversity, the highest arthropod diversity was found in the middle row, with arthropods belonging to all of the taxonomic groups identified as potential SWD predators in this study.

PLOS ONE
Carabidae, Opiliones and Araneae, with no detection of D. suzukii DNA in Coccinellidae individuals (Fig 6a). Regarding the Araneae functional groups, D. suzukii DNA was detected only on 2 of the 7 functional groups, namely orb web weavers (family Araneidae and Uloboridae) and space web weavers (family Dictynidae and Theridiidae), all collected with D-Vac (Fig 6b). The most abundant taxonomic groups with positive results were Chrysopidae, Formicidae and

PLOS ONE
Araneae, however, the most abundant functional group of Araneae, Other hunters (Oxyopidae, Sparassidae, Cheiracanthiidae, Philodromidae and Salticidae), collected with D-Vac, did not predate on D. suzukii. If frequency of detection is considered, Formicidae was the only group collected in all locations, with positive results in individuals from 3 of the locations (Fig  7a), and both positive results in spiders were from groups collected with D-Vac (Fig 7b).

Discussion
The need to develop new pest and disease management strategies is increasing. Biological control agents (BCAs) such as bacteria, fungi, parasitoids or predators [12] emerge as the most promising sustainable strategies. Here, we used a DNA-based strategy to produce a detailed catalogue of native predators of SWD in berry fields in Portugal, a critical first step to develop efficient and sustainable control measures. D-Vac was used to collect predators in plants and surroundings, while pitfall traps collected the potential predators mostly on the ground. Hemerobiidae, Chrysopidae, Miridae and Coccinellidae families were collected in the vegetation, Opiliones and Carabidae on the ground, and Formicidae and Araneae in both cases. Comparing the three sampling rows, predator abundance was higher in the middle of the orchard. Schmidt et al. [22] found a higher abundance of natural enemies in a pine margin of a SWD infested orchard, which correlated with a higher activity of SWD in the same area. The fact that most natural enemies were collected in rows in the middle of the orchard suggests a higher activity of SWD in that area, probably due to higher amount of food.
Of the collected arthropods, only those of the Coccinelidae family were negative for the presence of SWD-DNA specific sequence in the gut. Coccinellids are unspecific predators, and have been referenced as a potential BCA for other insect pests such as e.g., aphids, whiteflies, mealybugs or psyllids (reviewed by Kundoo and Khan [38]). Similar negative results for the presence of SWD in Coccinellids were previously reported in the USA and Switzerland [10,22], supporting coccinellids as unlikely SWD predators.
Families Chrysopidae, Miridae, Carabidae and Formicidae, which field individuals' gut showed here to contain SWD DNA, had been proposed in laboratory assays as SWD predators. However, laboratorial experiments are often not supported by field data. For example, Chrysoperla carnea (Stephens) was in the laboratory a predator of SWD [20,24], but this predation was never confirmed in the field [22,31]. Also, the laboratory predatory activities of the Carabidae Pterostichus mutus (Say) and Bembidion quadrimaculatum (LeConte) and the Miridae Dicyphus hesperus Knight against SWD were not confirmed in the field [13,14]. Field predation of SWD by Formicidae was previously suggested, as ants were observed not only digging and carrying SWD pupae, but also preying on SWD pupae and larvae in dropped blueberries [25,26]. Our data confirm these taxonomic groups as field SWD predators, thus being an important clarification of those previous results.
Among the taxa detected in our work, ants may be those with most potential as a BCA, considering its abundance and capacity to rapidly consume a large number of victims. The use of ant species as a BCA is not recent [53], and besides controlling pests in different agroecosystems [54], they also improve the soil quality [55]. Besides predating on D. suzukii, ants also predate other pests as C. pyri [56] and Hypothenemus hampei (Ferrari) [57].
This study identified for the first-time arthropods from the Opiliones and Hemerobiidae as SWD predators. Kamiyama et al. [31] had recovered Hemerobiidae individuals from traps with sentinel larvae, but did not confirm if the Hemerobiidae individuals were predators or predation was done by individuals from other taxa, such as Formicidae, Anthocoridae or Staphyilinidae [31]. Previous studies showed that Opiliones collected in orchards infested with SWD had negative results for the presence of SWD DNA [10,22]. Hemerobiids are predators of e.g., Pineus strobi (Hartig) (pine bark adelgid, a pine tree pest) [58], Aulacorthum solani (Kaltenbach) (foxglove aphid, in sweet peppers) [59] or Planococcus citri (Risso) (citrus mealybug) [60] but their predatory activity on SWD remains unknown. Similarly, harvestmen (Opiliones) were already identified as predators of Lobesia botrana (Denis & Schiffermüller) (grape berry moth) [61] or A. vittatum (striped cucumber beetle) [52]. Considering our results and that individuals belonging to this taxonomic group were already found in SWD infested orchards in North America and Europe, harvestmen should be considered for further predation studies on SWD, as they may be a suitable predator for the control of this pest.
Here, we also examined the presence of SWD DNA in spiders' gut content. The use of spiders as BCAs has been explored for a long time now, as they are known as generalist predators, representing a good strategy for certain agroecosystems [62]. Updated insights highlight that the effect of spiders as BCAs depends on the type of pest or existence of alternative preys and environmental conditions (as with any other generalist predator), and on the phenotype of spiders, namely its hunting strategy or behavior (functional group) [63]. Spiders are commonly identified in the field as predators of various pest species, namely A. vittatum [52], Frankliniella spp. [64], or Phyllocnistis citrella (Stainton) (citrus leafminer) [65]. In this study, SWD DNA was found in the gut content of spider families belonging to two functional groups, orb web weavers (family Araneidae and Uloboridae) and space web weavers (family Dictynidae and Theridiidae), both web-building groups. Individuals belonging to the family Araneidae and Theridiidae had already been identified as predators of SWD, by testing positive for SWD DNA [10,22], while Dictynidae individuals, although collected in infested orchards, tested negative [22]. Thus, our work is the first report of Uloboridae spiders being collected in SWD infested orchards, although further studies are needed as our study, based on functional groups, did not identify which family specifically predated on SWD, which also applies to the Dictynidae individuals.
When considering the abundance of each predator group in the berry orchards of this study, the most abundant were green lacewings, ants and spiders, suggesting that these are the best candidates for SWD predator-based control, due to their natural abundance in these fields. However, considering the spiders' functional groups, the most abundant individuals were identified as belonging to the group "other hunters", which had no positive results for SWD predation. Nonetheless, in previous studies, individuals belonging to this functional group were found positive for SWD-predation [10,22], and it is important to take into consideration that negative results for D. suzukii DNA presence in predators' gut may be influenced by a lack of predator-pest encounters at the time of sampling.
Our study did not identify anthocorids as SWD predators even though laboratory assays have focused on this group, namely Orius spp. [20,24,27,28]. Only in two instances anthocorids were considered potential SWD predators in field studies, but predation was not confirmed [26,31]. Woltz et al. [25] introduced a commercially available anthocorid species in an infested blueberry field with positive predation results, but there was no description if the species was already native to the orchard. Our results illustrated the need to conduct field studies, with the identification of native predators, in order to avoid the introduction of new species exclusively based in laboratory assays.

Conclusion
This study focused on the assessment of SWD field predation in infested Portuguese berry fields. The results confirmed the presence of SWD predators previously known or suspected, and importantly identified new predatory taxonomical groups, Opiliones and Hemerobiidae. This is the first study to confirm predation of SWD by individuals belonging to these families/ orders and to identify the spider family Uloboridae in SWD infested orchards, also with the potential to be a SWD predator. Our results, together with previous studies, encourage additional field studies to further understand the role of the identified predator families/orders as Biological Control Agents, not only of SWD but also other insect pests. Only that will allow determining which native species would be more adequate for pest control, as field predation by these arthropods is still poorly understood.
Supporting information S1 Fig. Representative photos of each arthropod sampling location. a) Location 1; b) Location 2; c) Location 3; d) Location 4; e) Location 5. L1, L3 and L4 had wild vegetation between rows, and only L2 and L5 used mulching films-L2 along the blueberry shrub rows, and L5 with the complete orchard's soil covered. In L4 and L5 nets were used, being that in L4 the nets were only placed above the blackberry trees in order to provide protection against UVrays, and in L5 the net covered the entirety of the orchard as protection against birds. (PNG) S1 Table. List of potential D. suzukii predators previously identified. In the Species column, when a species is "Not specified", it indicates that the predator was only identified to the family level. In the Type of study column, studies were considered as "Laboratory" when predation trials occurred under controlled conditions in the laboratory; "Field" when the predators or potential predators were captured in the field or predation was observed in the field; when predators were captured in the field but predation trials occurred in the laboratory or both in the laboratory and in the field, studies are identified with "Laboratory" and "Field". Predation was considered "Positive" or "Negative" when specific predators predated or not on D. suzukii, respectively; it was considered "Not confirmed" when the study identified potential predators and predation was observed, but it was not possible to identify which specific arthropod was the predator. DNA presence was considered "Positive" if DNA amplification occurred with SWD specific primers, "Negative" when there was no DNA amplification with specific primers to identify SWD, and "Not tested" when predation was not assessed based on SWD DNA presence. Origin refers to the country where the experiments took place, either in the laboratory or field. (DOCX) S2 Table. List of spiders considered as potential D. suzukii predators. Spider families are grouped according to its functional group into Undefined/Specialists, web-building and hunting spiders. Spiders in studies where family was not specified were classified as Spiders. In the Type of study column, studies were considered as "Field" when the predators or potential predators were captured in the field or predation was observed in the field. Predation was considered "Positive" or "Negative" when specific predators predated or not on D. suzukii, respectively; it was considered "Not confirmed" when the study identified potential predators and predation was observed, but it was not possible to identify which specific arthropod was the predator. DNA presence was considered "Positive" if DNA amplification occurred with SWD specific primers, "Negative" when there was no DNA amplification with specific primers to identify SWD, and "Not tested" when predation was not assessed based on SWD DNA presence. Origin refers to the country where the experiments took place, either in the laboratory or field. (DOCX) S3 Table. Spider families grouped in the corresponding functional group (identified according to Cardoso et al. [1]) and general group (web-building spiders, hunting spiders and other spiders). (DOCX)