In contrast to many other agricultural pests, little is known about P. attenuata, including its internal microbial community. Here, the compositions and structures of the microbiota communities of P. attenuata from fields across the hemp growth regions in China were surveyed by high-throughput sequencing of 16S rDNA gene for bacteria and ITS1 DNA fragment for fungi. We found that Proteobacteria (59.57−86.88%) and Bacteroidetes (2.52−35.21%) dominated the bacterial community across eight of the nine geographic samples (HN was the exception). The observed bacterial distribution pattern showed both similarities and differences with those in other insect species (Supplementary Figure 2). For example, in sap-feeding insects, such as aphids (Gauthier et al., 2015; Zhao et al., 2016), Proteobacteria was also the most abundant bacteria phyla. In many insect species of Lepidoptera, Diptera, Coleoptera, Hymenoptera, Hemipter and mites, the most prevalent phyla were Proteobacteria and Firmicutes (Zhang et al., 2020). Species in phylum Proteobacteria are known to perform many important functions such as nitrogen fixation, metabolisms of critical nutritional components (including sugars and proteins), insecticide resistance, and protection against parasites and pathogens in fruit flies, aphids, and moths (Zhang et al., 2020). In our study, we found that the highest relative abundances of KEGG pathways prediction at “level 3” were “metabolic pathways,” “biosynthesis of secondary metabolites,” and “microbial metabolism in diverse environments” (Figure 5). Therefore, Proteobacteria species likely play an important role in the nutrition and metabolism of P. attenuata.

At the genus level, dominant bacterial genera were Rickettsia (Proteobacteria), Wolbachia (Proteobacteria), Candidatus_Brownia (Bacteroidetes), Enterobacter (Proteobacteria), Pantoea (Proteobacteria), and Rhodococcus (Actinobacteria) (Figure 2C). This observed pattern is different from those reported in several other insects. For example, in Dendroctonus beetles, the dominant bacterial genera were Erwinia, Serratia, Klebsiella, Rhodococcus, and Sphingomonas (Xu et al., 2019).

For the first time, Rickettsia and Wolbachia infections were observed in P. attenuata. Wolbachia belongs to alpha-proteobacteria and strains in this genus are common intracellular parasites in arthropods and nematodes. It’s estimated that up to 70% of all insect species may contain Wolbachia (Kozek and Rao, 2007). These endosymbionts can infect many host tissues but the dominant tissues are those involved in sexual reproduction. Indeed, Wolbachia bacteria have been found to alter host insect biology in several ways, including nutrient provision, feminization, male-killing, parthenogenesis, and sperm-egg incompatibility or cytoplasmic incompatibility (Werren et al., 2008). Cytoplasmic incompatibility refers to the inability of con-specific sperm and egg to form viable offspring and it occurs when a Wolbachia infected male mates with a female that is either uninfected or infected by a different Wolbachia strain (Yen and Barr, 1971). Cytoplasmic incompatibility reduces the reproductive success of uninfected females and promotes the spread of Wolbachia strains infecting females. While the major mode of their transmission is vertical, they can also spread horizontally between host species (Werren et al., 2008). Indeed, due to their significant roles, Wolbachia bacteria have been explored as a potential agent in the control of mosquito-borne diseases in humans because of their unique ability to invade host mosquito populations rapidly (Pan et al., 2012). Similar suggestions have been made for controlling plant pathogens and pests. For example, Wolbachia endosymbionts in the Asian citrus pest psyllid Diaphorina citri can repress the holin promoter activity in the alpha-proteobacteria “Candidatus Liberibacter asiaticus,” a citrus disease agent, by killing the bacteria and preventing the disease from spreading (Jain et al., 2017).

The endosymbiotic Wolbachia is closely related to another intracellular bacterial genus Rickettsia (Perlman et al., 2006; Duron et al., 2008). While Rickettsia are most noted for causing human diseases, including Rocky Mountain spotted fever and epidemic typhus, they are also widely distributed in insects and play diverse roles (Perlman et al., 2006; Karkouri et al., 2022). For example, Rickettsia have shown to provide nutrition for host insects, improve their ability to resist high temperature and natural enemies, and significantly enhance the resistance of whitefly to infection by the bacterial pathogen Pseudomonas syringae (Zhang et al., 2021). Similar to those of Wolbachia, Rickettsia bacteria have shown to be capable of manipulating the reproductive behavior of host arthropods, including fertility, male-killing, and parthenogenesis (Perlman et al., 2006). In our study, the dominant bacterial genus in AH1, AH2, AH3, and SD were Rickettsia, the dominant bacterial genus in DQ, HRB, and YN were Wolbachia, and the dominant bacterial genera in JL was Rickettsia and Wolbachia. At present, the detailed roles of Wolbachia and Rickettsia to their host P. attenuata are not known. If they are found to play key roles in the survival and reproduction of P. attenuata in these regions, a potential novel control strategy against hemp flea beetles can be developed by targeting these endosymbiotic bacteria within these beetles.

Interestingly, the bacterial species richness and community diversity of the HN samples were significantly higher than those of other populations (Figures 2A,B). Furthermore, principal co-ordinate analysis showed that the community structures were similar among the four biological replicates within the HN region (Figure 2D). Therefore, the high bacterial species richness and community diversity of the HN samples were likely due to the uniqueness of the ecological niche at this site. Previous studies have shown that the microbial community in insects reflects the environmental microbial populations which in turn impact the insect host’s biology (Xu et al., 2019). The sampling site within Hunan province is rich in organic matter and has a climate different from other eight sites. Specifically, the site has had a long history of growing industrial hemp and at the site, the summers are hot, humid, and overcast while the winters are cold, humid, windy, and often cloudy. Over the course of the year, the temperature typically varies from 8°C to 33°C and is rarely below 4°C or above 36°C. At present, how these two features might have contributed to its distinct hemp beetle bacterial community in HN remain to be investigated. Additional sampling at sites with similar climates to those analyzed here as well as their corresponding soil samples are needed to determine the relative contributions of climate and soil niche to differences in the microbiome of hemp flea beetles.

The genus Lactococcus was abundant in the JL samples (20.64%) but rare in the other populations (<0.4%). Lactococcus is a genus of lactic acid-producing bacteria where only a single product lactic acid is produced during glucose fermentation. The broad distribution of Lactococcus bacteria in the Jilin samples suggested the potential importance of this fermentative pathway for this population of hemp flea beetles. In addition, this metabolic uniqueness of the Jilin samples may reflect other host phenotypic differences such as pesticide resistance. For instance, insecticide- resistant strains of the diamondback moth Plutella xylostella contained more abundant bacteria belonging to Lactobacillales, Pseudomonadales, and Xanthomonadales but fewer Enterobacteriales than an insecticide-susceptible strain (Xia et al., 2013). The genus Pantoea was abundant in the SD population (33.21%) and was rare in other populations (<2%). Pantoea is a highly diverse genus that can cause plant diseases and human diseases but also have functions in habitat restoration and pesticide degradation (Walterson and Stavrinides, 2015). It’s possible that Pantoea in the SD populations functioned similarly to those found in plants and humans. At present, how bacteria in genera Lactococcus and Pantoea might contribute to trait differences among P. attenuata populations and among the host plant Cannabis sativa are unknown.

Aside from comparing bacterial diversity, we also investigated the mycobiota of the P. attenuata samples. The dominant fungal phyla in our samples belonged to Ascomycota and Basidiomycota (Supplementary Figure 5). This result was not surprising because the great majority of known terrestrial fungi are in Ascomycota and Basidiomycota (Stajich et al., 2009), including molds, mushrooms, rusts, smuts, and yeasts. Species in Ascomycota and Basidiomycota play important ecological and economical roles, including making essential contributions to the bio-product industry, nutrient cycling, antibiotic production, and as animal, human and plant pathogens (Pei et al., 2019). In our study, the fungal communities of all samples of hemp flea beetles were dominated by common environmental fungi in Ascomycota and Basidiomycota. This result is consistent with the hypothesis that environmental fungal community is a key determinant of the fungal community within hemp flea beetles.

Our analyses revealed that the total number of known bacterial and fungal species was greater than the total number of bacterial and fungal OTUs. This result is not surprising because the OTUs were defined based on a 97% sequence similarity cutoff for both the bacterial and fungal datasets while many currently known closely related sister species had higher sequence similarities than 97% at these two barcode loci (Edgar, 2013; Xu, 2016, 2020). Given that there are unclassified novel taxa in both the bacterial and fungal datasets, the true numbers of bacterial and fungal species in our samples of P. attenuata are likely much higher than what we found here. High coverage metagenome sequencing coupled with culture-based surveys will help generate more robust estimates of bacterial and fungal diversities in these samples (Xu, 2020). The obtained microbial cultures will also allow direct tests of their potential roles in the survival and reproduction of hemp flea beetles.

We note that even though several geography-specific microbial community clusters were found in both the bacterial and fungal datasets, such clusters could be due to differences in other factors associated with the specific geographic regions, including climate, host cultivars, soil conditions, and host pest genotypes. Indeed, the four broad geographic regions (i.e., Northeast China, East-central China, Central-south China, and Southwestern China) analyzed here are known to differ from each other in at least one of the above factors. In addition, we observed samples in several geographic regions that were not clustered according to their geographic origins. At present, the reasons for such differences are not known. However, fine-scale heterogeneity in soil microflora, hemp clones/cultivars, health status (e.g., microbial infections) of sampled host hemp plants, and the host plant microbiome within crop fields could all have contributed to microbiome differences among hemp flea beetles sampled from different parts of a field. Fine-scale sampling of diverse microbiomes from the soil, host plants, and hemp flea beetles are needed in order to test their specific effects on the microbiome structures of hemp flea beetles.