The emerging contribution of social wasps to grape rot disease ecology

Dispersal capability experiment 1: characterization of sour microbial complex in adult P. dominulus

As part of a separate study, we characterized the total bacterial and fungal community of adult P. dominulus (A Madden, 2017, unpublished data). To determine if uninoculated P. dominulus harbor the sour rot microbial complex we looked at 49 adults from this study that were from five populations in eastern Massachusetts. Briefly, we homogenized the adult bodies and extracted DNA using a modified manufacturer’s protocol (Fierer et al., 2008). We amplified the bacterial 16S and fungal ITS1 DNA using 515f and 805r (covering the v4 region of the 16S) and the ITS1-F and ITS2 amplification primers (respectively). The primers contained illumina adapters and the reverse primers included unique 12 base pair barcodes allowing for multiplex sequencing. Amplification was carried out using the cycling parameters of 94 °C for 5 min, with 40 cycles of 94 °C for 45 s, 50 °C for 30 s, and 72 °C for 90 s, followed by a final extension at 72 °C for 10 min (for each targeted amplicon) (Barberán et al., 2015). All samples were run in triplicate with a no template control. Successful amplicons were cleaned using the Mobio Ultra-Clean™ PCR clean-up kit. DNA concentration was assessed fluorescently, using the Invitrogen Quant-IT PicoGreen^® dsDNA assay kit (Carlsbad, USA). DNA concentrations were normalized across samples and pooled (individually for bacteria and fungi). Sequencing was performed in multiplex on two runs of a MiSeq using 2 × 150 bp chemistry (separate runs for 16S and ITS amplicons). Sequences (paired end reads for bacteria and reverse reads only for fungi) were processed using the UPARSE pipeline (Edgar, 2013). We used the RDP classifier for taxonomy assignments, with methods and reference databases described elsewhere (Barberán et al., 2015). Non-target sequences (e.g., mitochondria and chloroplasts) were removed prior to downstream processing (see Madden et al., In Prep for more details of the generation and processing of amplicon sequences). To estimate the relative abundances of the sour rot microorganisms and to control for differences in sequencing depth, sequences were subsampled at 1990 sequences per wasp for bacteria and 50 for fungi. As this reduced the number of sequences and samples analyzed, we additionally used all sequences from samples when determining if a sour rot microorganism was detected in a wasp. Because of this unequal sampling, detection of any sour rot microorganism should be interpreted as presence-only, not presence-absence data. The bacterial and fungal taxa associated with sour rot were identified from literature reports (Barata, Malfeito-Ferreira & Loureiro, 2012a; Barata et al., 2012; Nally et al., 2013). These included acetic acid bacteria (family: Acetobacteraceae), and Ascomycota yeasts (class: Saccharomycetes). A group of saprophytic filamentous fungi (Aspergillus spp., Penicillium spp., Botrytis spp., Cladosporium spp., and Alternaria spp.) were additionally analyzed as these taxa are frequently associated with the development of sour rot symptoms (Rooney-Latham et al., 2008; Barata, Malfeito-Ferreira & Loureiro, 2012a; Barata, Malfeito-Ferreira & Loureiro, 2012b).

Dispersal capability experiment 2: microorganisms shed by P. dominulus when foraging

To determine if wasps could disperse viable microorganisms when foraging we made a sterile arena where we could detect the abundance of microorganisms dispersed to plants by visiting wasps. This arena included sterilized pseudoflowers that contained sugar water to motivate foraging (Fig. S4). After allowing foragers to visit the pseudoflowers during a 24-hour period, the pseudoflower petals were imprinted on microbial growth media to detect viable microorganisms transferred. Following incubation we counted the number of microbial colony forming units that grew on paired forager and forager-excluded pseudoflowers (Fig. S5). We replicated this assay 11 times with 11 unique nests and associated foragers. We compared the number of dispersed microorganisms in each trial using a paired t-test (See Fig. S4 for more details on methods and assay validation).

Inoculated wasps facilitated rot in injured berries

Wasp foraging led to a greater incidence and severity of rot symptoms in injured grape berries when the wasps were previously inoculated with a simplified sour rot microbial community. Injured grapes in enclosures with these wasps had greater disease scores than those injured berries exposed to air (Mann–Whitney U, W = 434.5, p-value = 0.02, n = 36) (Fig. 1), with more extensive symptoms of rotting, browning, and visual fungal growth (Fig. S6). Wasps are attracted to many of the volatile chemicals produced by fermenting microorganisms (Landolt et al., 1999; Davis, Boundy-Mills & Landolt, 2012). In accordance with this we often observed the wasps feeding on the most diseased of the six berries in each enclosure. This suggests that the observed effect of disease facilitation by wasps is a conservative estimate, as we are likely averaging these disease scores over those berries that had and had not come into contact with wasps.

During the later days of the experiment, the enclosures with rotten berries had the distinct smell of vinegar—the hallmark indicator of sour rot infections in vineyards (Steel, Blackman & Schmidtke, 2013). This smell is caused by the production of acetic acid and gluconic acid by the acetic acid bacteria of the sour rot complex. Despite these observations, injured grapes in the presence of wasps did not have significantly higher acetic acid concentrations than those in the absence of wasps (Mann–Whitney U, W = 203.5, p-value > 0.05, n = 24) (Fig. 1). While wasps did not significantly elevate average grape acetic acid concentrations, the highest acetic acid concentrations were associated with grapes in the presence of wasps (Fig. 1). Thus, grapes showing sour rot disease symptoms were associated with the presence of foraging wasps.

It is likely that the discrepancy in disease rot severity and acetic concentration arises from a secondary infection of some of the sour rot infected grapes, rather than the lack of a sour rot infection. Filamentous fungi and spores were observed on many of the injured grapes with wasps and in a few of the injured grapes without wasps (Fig. 1). Wasp foraging increased the incidence of this disease symptom on injured grapes (Mann–Whitney U, W = 4, p-value = 0.02, n = 6). This was somewhat surprising, as none of the microorganisms fed to these wasps result in such growth. Sequence analysis of isolates from these berries revealed that the fungus was a black mold, Aspergillus niger (Table S1). As. niger is a common pathogen of grape berries and a frequent associate of sour rot disease (UC IPM Pest Management Guidelines: Grape, 2014), where it metabolizes many of the organic acids in grapes, including acetic acid (Halliwell & Walker, 1952). Beyond assisting with berry decomposition, As. niger can produce the mycotoxin Ochratoxin A (Tjamos et al., 2004), which can contaminate resulting wine. The presence of this fungus could explain the low levels of acetic acid we observed at the conclusion of the assay, as this fungus could be metabolizing the acetic acid produced by the sour rot infection (see below). As. niger is often found within surface sterilized grape berries such as the ones we used in this experiment (Passamani et al., 2012), thus it was likely that the original berries selected for this experiment were already harboring As. niger within them. The increase in the presence of a black mold (Aspergillus rot) in the presence of wasps suggests that, like fruit flies (Barata, Malfeito-Ferreira & Loureiro, 2012b), wasps are facilitating berry disease by making injured host tissue more susceptible to infection.

Sour rot inoculated wasps did not facilitate rot in uninjured berries

Fruit flies facilitate sour rot, but are only capable of causing the disease in previously injured grapes (Barata et al., 2012). In our foraging experiment, P. dominulus similarly did not have an effect on grape disease severity or incidence unless the grapes were previously injured (Mann–Whitney U, W = 4, p-value = 0.02, n = 6, Mann–Whitney U, W = 18, p-value >0.05, n = 6) (Fig. 1, Fig. S6). This was somewhat surprising, as these wasps are known to cause initial damage to various fruits including grapes (V. vinifera) and sweet cherries (Prunus avium) (Cranshaw, Larsen & Zimmerman, 2011). A possible explanation for the difference between our results and field observations could be explained by the wasps’ inability to break through the skin of certain thick-skinned grape cultivars such as those used in this experiment (Galvan, Koch & Hutchison, 2008).

Uninoculated wasps facilitated rot in injured berries

The results of our second foraging assay revealed that P. dominulus harboring only its native microbial community facilitated grape sour rot. As in the first experiment, wasp foraging led to more severely diseased grapes (Mann–Whitney U, W = 9, p-value = 0.03, controls: n = 10, wasp treatment: n = 6) (Fig. 2). In this experiment, grape berries in the presence of these wasps had higher acetic acid concentrations, indicative of sour rot (Mann–Whitney U, W = 2.5, p-value <0.01, controls: n = 10, wasp treatment: n = 6) (Fig. 2). To the best of our knowledge, this is the first study to show that wasps carrying their native microbial community can facilitate sour rot in the absence of other vectors.

Unlike in the first foraging experiment, we detected no symptoms of As. niger infection in these grapes. The lack of these symptoms and the higher acetic acid concentration relative to controls are consistent with our hypothesis that the As. niger infection may have obscured some of the sour rot symptoms in the first experiment.

Uninoculated wasps are associated with sour rot microorganisms

The sour rot microbial complex includes yeasts, filamentous fungi, and acetic acid bacteria, but the acetic acid bacteria are responsible for the worst of the disease symptoms (Barata et al., 2012). All inoculated wasps we surveyed contained acetic acid bacteria, including Gluconobacter spp., the genus associated with acetic acid production in sour rot infections (Barata et al., 2012). Nearly half (52%) of these wasps also carried sour rot associated yeasts, the second component of the sour rot community. Saprotrophic filamentous fungi such as Aspergillus spp, Penicillium spp., Alternaria spp., Cladosporium spp., and Botrytis spp. may be associates of the sour rot complex, or they may be secondary infections associated with sour rot (Barata, Malfeito-Ferreira & Loureiro, 2012a). Of the wasps we sampled, 88% carried at least one of these fungi. Of particular note, many of these wasps contained sequences from Aspergillus spp. and Botrytis spp. which are also associated with a black mold (Aspergillus rot) and grey mold (respectively) (Barata, Malfeito-Ferreira & Loureiro, 2012a).

While half the wasps carried the full complement of sour rot microorganisms, they carried highly variable amounts of these species (Fig. 3). In some wasps, the bacterial community was dominated by acetic acid bacteria—making up 99% of the bacterial sequences in one wasp—while in others it was barely detectable (Fig. 3). This variability is similar to that found in fruit flies, where some fruit flies harbor yeasts or acetic acid bacteria and others do not (Barata et al., 2012). This variability suggests that rather than being obligate associates of wasps, these microorganisms may be acquired from the local environment.

Uninoculated wasps transfer microorganisms to plants when foraging

Paper wasps have been implicated in the dispersal of microorganisms to plants, most notably yeast dispersal onto grape berries in vineyards (Mortimer & Polsinelli, 1999; Stefanini et al., 2012). However, few studies have provided evidence that wasps can transfer viable microorganisms when foraging. To test the hypothesis that P. dominulus disperses microorganisms while foraging, we designed a sterile arena with sterile ‘pseudoflowers’ where petals were separated from the nectar food source. This setup allowed us to test whether wasps shed viable microorganisms when performing natural foraging behaviors.

Foraging wasps dispersed significantly more microorganisms to flowers than were deposited by air, with wasps dispersing 40% more microbial propagules within 24 h than were dispersed by air alone (paired t-test, t = 7.12, df = 10, p < 0.0001) (Fig. S7). These estimates are conservative, however, as we used conditions that would select for only a portion of the transferred microbial community. In our assays, wasps were interacting with the pseudoflowers as they would interact with flowers when foraging for nectar. When wasps feed on grape berries, they are further interacting with the host tissue by walking on it, masticating pieces of it, and defecating on it (Video S1). In this assay we only looked at the effect of walking on the host tissue as part of feeding. Even with these conservativemethods, we found that wasps disperse microorganisms to host grapes and that these microorganisms are viable regardless of wasp cuticular antimicrobial secretions (Turillazzi et al., 2006).

Insights into the role of P. dominulus in grape rot ecology

Sour rot, like many rot infections, is a disease that manifests when there are the appropriate environmental conditions, a susceptible host, and the presence of a pathogen. Fruit flies are known to facilitate sour rot by increasing host susceptibility in damaged grapes, and by potentially increasing the pathogen presence by vectoring these microorganisms from their body to grapes (Barata et al., 2012). We found that foraging P. dominulus facilitated sour rot and an associated Aspergillus rot disease in grapes. The increase in disease symptoms, particularly of Aspergillus rot in surface sterilized grape berries suggests that wasps are facilitating the disease by making the host more susceptible to infection (Barata et al., 2012). The mechanism for this facilitation is currently unknown, but may relate to mechanical disruption of the berry tissue that elicits a hypersensitive response (Lattanzio, Lattanzio & Cardinali, 2006). However, given that we have shown these wasps can disperse microorganisms when foraging, that they are capable of causing sour rot in surface sterilized berries, and that all of the wasps we surveyed contained part or all of the microbial community associated with sour rot, it is likely that these wasps are also vectoring these rot microorganisms. Some of these microorganisms may directly cause the symptoms we saw (high acetic acid concentrations), or create environments that foster secondary infections (e.g. Aspergillus rot). Our results are remarkably consistent with those for fruit flies, suggesting P. dominulus may have a similar ecological role to D. melanogaster in sour rot ecology.

Our foraging assays were conducted in the lab to assure that no other insect vector was influencing disease symptoms. They further featured surface sterilized berries to better understand the insect origin of pathogenic microorganisms. These represent different conditions than experienced in the field, where berries host a plethora of microbial species, including those associated with various rots (Barata, Malfeito-Ferreira & Loureiro, 2012a; Gilbert, Van der Lelie & Zarraonaindia, 2014). While this likely makes the results of our foraging experiments highly conservative, future work will be needed to understand the relationship P. dominulus has with dispersing and facilitating sour rot in the context of vineyards. Such studies should include other wasps found in vineyards, as the relationship between the sour rot microbial community and grapes likely extends beyond P. dominulus to other insects with similar ecological roles. Vineyards contain many vespids which feed on grape berries (Mussen & Rust, 2012). Some of these wasps reach much higher abundances in vineyards than P. dominulus and have stronger mandibles capable of damaging even thick-skinned fruit (Alford, 2014).

Carnivorous wasps are rarely considered as disease vectors; however, the absence of data should not indicate the absence of such a role. The paper wasp Polybia rejecta facilitates fungal disease when it forages on Red-Eyed Tree Frog, Agalychnis callidryas, egg masses (Hughey et al., 2012). The German wasp, Vespula germanica, has been implicated in causing bacterial infections of udders when it forages on cows for protein (Yeruham, Schwimmer & Brami, 2002). It is therefore likely that wasps represent an important and relatively unexplored member in the multitrophic interactions of certain microbial infections.