No evidence of enemy release in pathogen and microbial communities of common wasps (Vespula vulgaris) in their native and introduced range

Abstract

When invasive species move to new environments they typically experience population bottlenecks that limit the probability that pathogens and parasites are also moved. The invasive species may thus be released from biotic interactions that can be a major source of density-dependent mortality, referred to as enemy release. We examined for evidence of enemy release in populations of the common wasp (Vespula vulgaris), which attains high densities and represents a major threat to biodiversity in its invaded range. Mass spectrometry proteomic methods were used to compare the microbial communities in wasp populations in the native (Belgium and England) and invaded range (Argentina and New Zealand). We found no evidence of enemy release, as the number of microbial taxa was similar in both the introduced and native range. However, some evidence of distinctiveness in the microbial communities was observed between countries. The pathogens observed were similar to a variety of taxa observed in honey bees. These taxa included Nosema, Paenibacillus, and Yersina spp. Genomic methods confirmed a diversity of Nosema spp., Actinobacteria, and the Deformed wing and Kashmir bee viruses. We also analysed published records of bacteria, viruses, nematodes and fungi from both V. vulgaris and the related invader V. germanica. Thirty-three different microorganism taxa have been associated with wasps including Kashmir bee virus and entomophagous fungi such as Aspergillus flavus. There was no evidence that the presence or absence of these microorganisms was dependent on region of wasp samples (i.e. their native or invaded range). Given the similarity of the wasp pathogen fauna to that from honey bees, the lack of enemy release in wasp populations is probably related to spill-over or spill-back from bees and other social insects. Social insects appear to form a reservoir of generalist parasites and pathogens, which makes the management of wasp and bee disease difficult.

Fig 1. Sample locations for common wasps from the native (England and Belgium) and invaded range (New Zealand and Argentina).

Twenty adult V. vulgaris worker wasps were collected from each of the four countries. In some cases, multiple wasps were collected in the same area, but never from the same nest. For Argentina, the restricted sampling area represents the latitudinal limits of their distribution at the time of sampling in 2013. Common wasps are distributed throughout New Zealand.

Fig 2. The number of microbial taxa observed from the previously published literature and proteomics methods.

(A) The number of microbial taxa observed in published studies examining V. germanica and V. vulgaris. The numbers at the top of the bars represent the number of published studies, e.g. there were nine published papers examining fungal in wasps from the invaded range. Inset is a graph showing the non-significant relationship (p≥0.218) between the number of taxa found and the number of studies for each microbial group. (B) Results from our proteomics survey of microbes associated with wasps from the native and invaded range. No viruses were observed in the proteomics analysis. The “other” category is from peptides indicating the presence of taxa including amoeba (Acanthamoeba sp.), a protozoan (Babesia sp.), and tapeworm (Taenia sp.).

Fig 3. Microbial communities in wasp samples from the four countries.

(A) A Venn diagram showing the overlap and distinctiveness of microbial taxa from common wasps in the native (England and Belgium) and invaded range (New Zealand and Argentina). A total of 131 peptides from distinct microbial taxa were observed. Of these 131 microbial taxa, 39 taxa were shared between all countries, but different countries had between 9–14 distinct taxa. (B) Rarefaction curves showing the similarity of microbial taxa accumulation with increasing peptides sampled.

Fig 4. Maximum Composite Likelihood tree for putative 16S Nosema sequences from Vespula vulgaris sampled (bold) and the best matching sequences on GenBank, together with their accession numbers and sample collection locations (where available).

The tree was based on 2000 bootstraps of a general time-reversible model with gamma distribution and invariant sites parameters (GTR + G(0.69) +I(0.0); lnL −928.456) in MEGA6. The estimates of levels of support shown below the nodes are bootstrap values greater than 50%. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site.

Fig 5. Maximum Composite Likelihood tree for putative Actinobacteria sequences from Vespula vulgaris sampled (bold) and the best matching sequences on GenBank, together with their accession numbers and host species.

The tree was based on 2000 bootstraps of a general time-reversible model with gamma distribution and invariant sites parameters; lnL −559.13) in MEGA6. The estimates of levels of support shown below the nodes are bootstrap values greater than 50%. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. Often GenBank sequences were equally well matched to the sequences from V. vulgaris and those displayed on the tree are not exhaustive (e.g. the Ireland sample matched equally well to multiple Arthrobacter sp.).

Fig 6. Maximum Composite Likelihood trees.

(A) Putative Deformed wing virus (DWV) sequences. (B) Kashmir bee virus (KBV) sequences from Vespula vulgaris and Apis mellifera sampled (bold) and the best matching sequences on GenBank, together with their accession numbers and host species. The trees were based on 2000 bootstraps of a general time-reversible model with gamma distribution and invariant sites parameters (GTR + G(0.48) +I(200); lnL −550.04) for DWV and a general time-reversible model (GTR; lnL −195.50) for KBV in MEGA6. The estimates of levels of support shown below the nodes are bootstrap values greater than 50%. The trees are drawn to scale, with branch lengths measured in the number of substitutions per site.