Symbiotic Microorganisms and Their Different Association Types in Aquatic and Semiaquatic Bugs

Abstract

True bugs (Hemiptera, suborder Heteroptera) constitute the largest suborder of nonholometabolous insects and occupy a wide range of habitats various from terrestrial to semiaquatic to aquatic niches. The transition and occupation of these diverse habitats impose various challenges to true bugs, including access to oxygen for the aquatic species and plant defense for the terrestrial phytophagans. Although numerous studies have demonstrated that microorganisms can provide multiple benefits to terrestrial host insects, a systematic study with comprehensive higher taxa sampling that represents aquatic and semiaquatic habitats is still lacking. To explore the role of symbiotic microorganisms in true bug adaptations, 204 samples belonging to all seven infraorders of Heteroptera were investigated, representing approximately 85% of its superfamilies and almost all known habitats. The symbiotic microbial communities of these insects were analyzed based on the full-length amplicons of the bacterial 16S rRNA gene and fungal ITS region. Bacterial communities varied among hosts inhabiting terrestrial, semiaquatic, and aquatic habitats, while fungal communities were more related to the geographical distribution of the hosts. Interestingly, co-occurrence networks showed that species inhabiting similar habitats shared symbiotic microorganism association types. Moreover, functional prediction analyses showed that the symbiotic bacterial community of aquatic species displayed richer amino acid and lipid metabolism pathways, while plant-feeding true bugs benefited more from the symbiont-provided xenobiotics biodegradation pathway. These results deepened the recognition that symbiotic microorganisms were likely to help heteropterans occupy diverse ecological habitats and provided a reference framework for further studies on how microorganisms affect host insects living in various habitats. IMPORTANCE Symbiotic bacteria and fungi generally colonize insects and provide various benefits for hosts. Although numerous studies have investigated symbionts in terrestrial plant-feeding insects, explorations of symbiotic bacterial and fungal communities in aquatic and semiaquatic insects are rare. In this study, the symbiotic microorganisms of 204 aquatic, semiaquatic, and terrestrial true bugs were explored. This comprehensive taxon sampling covers ~85% of the superfamilies of true bugs and most insect habitats. Analyses of the diversity of symbionts demonstrated that the symbiotic microbial diversities of true bugs were mainly affected by host habitats. Co-occurrence networks showed that true bugs inhabiting similar habitats shared symbiotic microbial association types. These correlations between symbionts and hosts together with the functions of bacterial communities indicated that symbiotic microbial communities may help true bugs adapt to (semi)aquatic habitats.

Keywords: Heteroptera; aquatic/semiaquatic bugs; co-occurrence network; diverse habitats; symbiotic microorganism.

FIG 1

Composition of symbiotic bacterial and fungal communities in true bugs. Relative abundance plots of bacterial (A) and fungal (B) phyla of 204 and 71 samples are shown, respectively. The host infraorders or superfamilies are shown at the bottom of each plot and are colored according to the infraorders or to the infraorders to which they belong. (C) Representative images of true bugs inhabiting four typical habitats.

FIG 2

The alpha and beta diversity of symbiotic microbial communities. The samples are grouped according to geographic position and infraorder. (A to C) are the diversity results of bacterial communities. Faith’s PD indices of symbiotic bacterial communities are mainly affected by hosts and habitats (A) but are not correlated with geographical positions (B). (C) The PCoA plot of bacterial communities based on unweighted UniFrac distance. (D to F) are the diversity results of fungal communities. Faith’s PD index of symbiotic fungal communities is mainly affected by geographical positions (E) but is not correlated with hosts (D). (F) The PCoA plot of symbiotic fungal communities based on unweighted UniFrac distance.

FIG 3

Phylogenetic relationships of true bugs and the corresponding heatmap of the abundance of symbiotic microbial communities at the order level. Branches were collapsed, and only the infraorder and superfamily names were provided. Internal tree nodes are labeled with colored dots. The heatmap on the left represents the 14 most abundant bacterial orders. The heatmap on the right represents the 14 most abundant fungal orders. For the unidentified orders, their lowest-level classifications are given instead.

FIG 4

Co-occurrence networks of the symbiotic microbial community. For each infraorder, the co-occurrence network of bacterial communities (left column), the co-occurrence network of fungal communities (middle column), and the co-occurrence network of bacterial and fungal communities (right column) are shown. No symbiotic network of Enicocephalomorpha or fungal network of Dipsocoromorpha was analyzed because of the small sample sizes. Only correlations with |r| > 0.8 and P < 0.0001 are shown. Each node represents an ASV. The nodes are colored according to phylum (left and middle columns) or kingdom (right column). The node size is proportional to the weighted degree of ASVs. Edge thickness is proportional to the weight of correlation.

FIG 5

Heatmap of functionally predicted pathways of symbiotic bacterial communities. For each infraorder, the proportions of pathways were compared with those of all other samples. The color represents the difference in mean proportions between a specific infraorder and the remaining samples. Only the pathways with P < 0.001 (Welch’s t test) are shown in the heatmap.