Parasitoids have important consequences for host physiological and ecological function. They influence the interaction of their hosts with their immediate surroundings, including host-microbiome associations, often shaping host-microbe evolutionary functions. The same is argued to occur in B. dorsalis, which is permissive to some parasitoids and not permissive to others. Here we demonstrate that virulent and avirulent parasitoids of B. dorsalis differentially shape the structure and functions of the gut bacteriome and mycobiome of this pest.
Similar to earlier microbiome reports from tephritids [10, 17, 27, 28], we found that, Proteobacteria and Firmicutes were the most abundant bacterial phyla in B. dorsalis larvae. This finding suggests the intimate association between B. dorsalis and these phyla, and that they could be contributing to the physiological functioning of this pest such as development and nutrient digestion among others. Earlier studies [16, 17, 19, 21, 29] investigating Proteobacteria and Firmicutes bacterial strains and their roles in the eco-physiological functions of B. dorsalis confirm this phenomenon.
In terms of species composition, Acetobacter species, Acinetobacter species, as well as Anoxybacillus species were highly abundant across all treatments. Species of the genus Anoxybacillus have been linked to the digestion of sugars, cellulose, fats, and proteins [30, 31]. On the other hand, Acetobacter thailandicus, like other acetic acid bacteria (see [32] and references therein) could also be involved in the breakdown of sugar in the guts of frugivorous insects such as B. dorsalis. Therefore, the high abundance of these bacteria across all treatments could be due to their metabolic roles in B. dorsalis larvae.
Earlier reports suggested that less diverse microbial communities are prone to colonization by pathogenic microbes via reduced niche and nutritional competition as well as suppressed immuno-competence [33]. Concordant with this, pathogenic bacteria like Serratia marcescens and S. maltophilia were highly abundant in the less diverse D. longicaudata-parasitized larval guts. Functionally, S. marcescens is a commensal symbiont with mild to no effects on its hosts. However, its proliferation and subsequent translocation to the hemocoel has been shown to be detrimental, rendering it pathogenic rather than commensalistic to its hosts [34, 35]. Indeed, a high load of S. marcescens has been reported to induce gut epithelia bloating and thinning in its hosts, which enhances its translocation into the hemocoel and interference with host immune function [34, 36, 37]. As such, this bacterium has been explored for its potential use in the control of arthropod pests and implicated in altering the vector competence of some insects of human health importance such as mosquitoes [34, 37–39].
On the other hand, S. maltophilia is a common gut bacterium in insects [4, 6, 40] associated with bacteremia in immune-suppressed and immunocompetent systems [41, 42]. We, therefore, postulate that the increased abundance of these bacteria could be a result of D. longicaudata-induced gut dysbiosis, which shifts the B. dorsalis larval gut bacteriome to a pathogen-dominated community. Our other studies have demonstrated that parasitisation by D. longicaudata downregulates anti-oxidative genes like glutathione transferases as well as cecropins and lysozyme B, genes responsible for antimicrobial defense in B. dorsalis [43]. It is, therefore, plausible that parasitisation by D. longicaudata increases the abundance of pathogenic gut microbes via suppression of antimicrobial defenses and activation of oxidative stress, interactive processes that advance its virulence against B. dorsalis. This finding raises a fascinating perspective of the ecological significance of these host-parasitoid microbe tripartite models and their implications for parasitoid virulence, a phenomenon that warrants further investigation.
Previous studies reported acquisition and/or increase in abundance of entirely new host gut microbial members after parasitisation due to transfer of microbes from the parasitoid to the host, stinging effect of the parasitoid and/or parasite-induced dysbiosis [5, 6, 11, 44]. In this study, we found a similar trend in the parasitised larval guts which comprised Weisella and Pantoea species, bacteria that were not present in the unparasitized larvae. These two bacteria are ubiquitous in the environment, and hence, might have been introduced into the host larvae during parasitisation and facilitating their colonization of the host gut. More interesting, however, was the finding that M. morganii was exclusively associated with D. longicaudata-parasitized larvae. Morganella morganii, is an opportunistic bacterium linked to pathogenicity in tephritids [45, 46]. This bacterium was also present in D. longicaudata, so it was perhaps transferred from the parasitoid female into the host larvae during parasitisation as a possible immune-suppressing arsenal in B. dorsalis.
Contrary to the salient gut dysbiosis recorded in D. longicaudata-parasitized larvae, parasitisation by P. cosyrae only slightly altered the bacteriome of B. dorsalis larvae. The disparity in the bacteriome composition of B. dorsalis larvae after parasitisation by these two wasps could be attributed to variations in host regulation strategies employed by both parasitoids. Pystallia cosyrae is unable to surmount the immune defenses of B. dorsalis [22, 23] and thus, it seems unlikely to cause a major shift in the microbial community of this pest. This finding was also supported by the network module results which depicted similar module clustering of the microbiome of the unparasitised larvae and that of those parasitized by P. cosyrae further confirming the insignificant impact of this parasitoid on B. dorsalis microbiome. Diachasmimorpha longicaudata, on the other hand, injects its symbiotic virus, DlEPV into the parasitized larvae which markedly disrupts the immune processes while concomitantly altering the structure and composition of the bacteriome of this host.
Regarding the parasitoid bacteriome, the high abundance of Anoxybacillus, Acinetobacter and Pseudomonas bacteria in D. longicaudata could be due to their contribution to the nutritional and metabolic needs of this parasitoid. In contrast, the gut bacteriome of P. cosyrae was strikingly less diverse and was dominated by the bacterium Arsenophonous nasoniae a widely distributed male killing secondary symbiont [47, 48]. Although not reported in other pine species, the association of A. nasoniae with P. cosyrae is not surprising since this symbiont has been reported in other hymenopteran parasitoids [48]. Unexpectedly, we found no association of A. nasoniae with B. dorsalis, a finding that deviates from the theory of shared microbiomes due to horizontal symbiont transmission between parasitoids and their hosts [44, 49, 50]. While unclear, it is possible that this could be a selective-association mechanism since B. dorsalis and P. cosyrae do not share evolutionary history, or that A. nasoniae is blocked from colonizing B. dorsalis as the parasitoid is encapsulated at the egg stage alongside the parasitoid venom cocktail [22, 23]. These arguments, however, warrant further investigation to unravel the evolutionary aspects, transmission mechanisms, and eco-physiological implications of harboring A. nasoniae by P. cosyrae. This will elucidate the intricate mechanisms underlying host-parasitoid interactions in tephritids and the role of bacterial symbionts in these host-parasitoid bi-trophic models.
Nonetheless, bipartite network analysis revealed occurence of generalist bacteria genera such as Pseudomonas, Enterobacter, Acinetobacter, Anoxybacillus, and Corynebacterium which were present across all B. dorsalis larvae and the parasitoids, especially D. longicaudata. This finding supports the host-parasitoid shared microbiomes theory depicting an intimate association between both parasitoids and B. dorsalis through horizontal transfer of symbionts between the two entities.
Similar to the gut bacteriome, we found a significant negative effect of parasitisation on the abundance, composition, and diversity of the mycobiome of B. dorsalis larvae, a finding that substantiates our argument of parasitoid-induced gut dysbiosis. We postulate that this negative effect of parasitisation on B. dorsalis larval gut mycobiome could be a consequence of parasitoid-induced alteration of host immune responses which inadvertently impact the gut fungal commensals. This is, especially because, fungi and parasitoids are eukaryotes and are thus, likely susceptible to similar immune defenses. Alternatively, one could explain the reduced fungal diversity as a result of the over proliferation of some gut bacteria. For example, S. maltophilia and Pantoea species, which were highly abundant in the guts of the parasitized larvae, have been shown to inhibit fungal growth [51, 52].
Very few studies have explored the fungal communities of tephritid fruit flies, and even fewer studies [53–56] have attempted to divulge the roles fungi play in these insects. Nevertheless, available literature suggests that most fungi are essential for nutrient acquisition and host development [56, 57]. For instance, Saccharomyces, the dominant fungal genus recorded in B. dorsalis larvae in this study, has been linked to improved host fitness [58, 59] and immune priming [60]. While this is the first report of Saccharomyces species in tephritids, it is not a surprising finding since Saccharomyces species like S. cerevisiae have been reported in other insect species [61–63]. Further investigation is warranted to determine its contribution to the eco-physiological functioning of B. dorsalis.