The prevailing view of host-associated microorganisms and their role in insect-plant coevolution is rapidly changing. Acquisition of beneficial microbes serves not only to expand the ecological niche of the host, but can add novel weaponry in the adaptive battle between plants and the herbivores that colonize them [3, 5, 7, 8]. Furthermore, insects not only acquire microbes from host plants and the surrounding environment, but these microbial communities are intimately linked [9, 49, 50]. Disentangling the mechanisms driving variation in herbivore microbial communities and the ecological consequences for host plant specialization is therefore of great interest. In the current study, shifts were observed in the core microbiome of aphid species exhibiting varying degrees of host specialization within the milkweed family. Aphids feeding on milkweed share most bacterial taxa (Fig 1), but each species harbors unique populations that differ in strain diversity and relative abundance of core heritable symbionts, most notably Arsenophonus (Figs 1b, 3 & 5). However, broader patterns in bacterial symbiont diversity that scale with diet breadth were not observed (Figs 3 & 4). Instead, our results provide support for hypothesized effect of overlapping host plant range and common selective pressures (e.g. plant chemical defenses) leading to similarities in microbiomes across milkweed aphid species.
Host plant specificity can influence herbivore associated microbial communities on multiple levels, including causing changes in functional, taxonomic or strain level diversity or by altering the abundance of individual taxa. The microbiomes of generalists and specialist herbivores could also vary simply due to differences in the contribution of heritable versus environmentally acquired microbes, the latter being more variable and transient. However, diet may in fact generally be a poor predictor of insect bacterial community composition [51], with exceptions only in certain groups of herbivores [52]. For example, Lepidopteran larva have variable gut microbiomes that are likely shaped by the host plant they feed on, but general patterns associated with diet breadth are not observed due to the high turnover rate of these microbial communities [14]. In contrast, recent work in Costa Rican rolled-leaf beetles (Cephaloleia spp.) shows diet breadth is linked to microbiome diversity and community structure [24]. Previous research also indicates the composition of aphid microbial communities are structured by host plant [11, 15, 16, 20, 25, 27, 32, 34] and heritable bacterial symbionts are involved in expanding diet breadth [26, 28].
Our results show aphids specialized on a single plant family (Asclepiadaceae) that vary in diet breadth have similar bacterial communities on a taxonomic level, but differ in strain diversity and relative abundance of key symbionts (e.g. Arsenophonus). Horizontal transfer of facultative symbionts via host plants can occur in aphids [53, 54], which could contribute to similarities in symbiont communities (i.e. shared ASVs) across different species that overlap in host range and/or naturally co-occur on plants. Another possible explanation is exposure to similar nutritional and chemical profiles could homogenize microbiomes of herbivore species feeding on the same host plants. The species in this study are specialists of the Asclepias (milkweed) family and therefore may have similar core symbionts due to exposure to closely related host plants. Additionally, phylogenetic relatedness can generally result in closely related aphid species harboring more similar microbiota than distantly related species [21]. However, one unexpected result was the occurrence of a single strain of Buchnera in all three aphid species (i.e. ASV 1; see Fig 2) even after stringent filtering for false positives. This is the dominant strain infecting A. nerii and is found in significantly higher abundance compared to the other two species (Fig 2, Supplementary Table S6). The current wealth of research on Buchnera shows this primary endosymbiont of aphids lives intracellularly, relies purely on vertical transmission, and has exhibited co-cladogenesis with aphid hosts over millions of years. It is unlikely this strain has been transferred across species, but it is also unclear why our dataset shows higher levels of sequencing “cross-talk” between samples than previously observed (e.g. [43]). One possible explanation could be that taxonomic classification using the 16s rRNA gene is unable to provide strain-level resolution for Buchnera in some cases (i.e. multiple strains with identical 16s sequences grouped as ASV1) and therefore additional genomic information is needed to distinguish unique strains found across milkweed aphid species.
Although milkweed aphid microbiomes were overall similar, key differences in symbiotic partnerships could contribute to additional ecological variation (e.g. ant tending, parasitism rates, predation). Interestingly, variation in symbiont relative abundance and strain diversity contributed most to differences observed across milkweed aphid microbiomes. In particular, Arsenophonus was found in higher abundance in A. asclepiadis compared to the other two species. Arsenophonus is a notorious shape-shifting insect symbiont, known best for reproductive manipulation of its host [55, 56]. Most aphid facultative symbionts are found in much lower abundances compared to the obligate nutritional symbiont Buchnera, suggesting the unusually high abundance observed in A. asclepiadis could be linked to symbiont complementarity, as has occurred in other aphid species [57]. It is also possible Arsenophonus provides a general fitness boost, similar to what has been observed in the soybean aphid [58]. Finally, differences in symbiont populations could shape milkweed aphid-ant mutualisms, possibly via microbial induced changes to honeydew or emission of chemical compounds that mediate partner attraction. Previous work shows insect social partnerships not only uniquely influence each host's symbiotic microbiome [59], but that volatile organic compounds produced by aphid-associated microbes play a role in attracting ant mutualists [60]. Among milkweed aphids, A. asclepiadis is consistently tended by ants and benefits from enhanced protection from predators, while A. nerii is occasionally ant tended and M. asclepiadis appears to be a loner lacking ant friends [61]. Based on the current study an intriguing question arises; Does Arsenophonus mediate milkweed aphid-ant interactions and thus contribute to observed differences in ant attendance? Although Arsenophonus does not appear to influence the intensity of ant attendance in cowpea aphids [62], it is possible this symbiont has evolved a different ecological role in the case of A. asclepiadis. However, given the 16s rRNA metabarcoding approach used in this study only provides relative abundances, Arsenophonus titer levels will need to be confirmed using additional methods such as quantitative PCR in order to take the first step towards addressing potential functions, including nutritional supplementation or aphid-ant interactions. Detection of additional facultative symbionts (e.g. Serratia) also warrants further investigation into symbiotic relationships and functional roles in milkweed aphid biology and ecology.
The current study is limited in that aphids were collected from a single host plant (i.e. common milkweed) and generalist species with a host range outside the milkweed family (e.g. Myzus persicae) were not characterized due to low occurrence in the field. We sampled aphids from a single common host plant rather than multiple milkweed species in order to focus on identifying differences in core heritable symbionts (e.g. presence/absence of taxa, large shifts in relative abundance) and reduce variation introduced by environment and host plant differences. Consequently, our sampling design has limited ability to detect changes in environmentally acquired microbes and does not test for changes in microbiome composition induced by feeding on different host plant species. Imbalance in sampling across aphid species (e.g. fewer M. asclepiadis samples, Supplementary Table S1) resulting from natural variation in prevalence could also mean that some microbial variation was missed. Finally, while this study profiled only bacterial symbionts, additional microbes present in the broader aphid microbiome (e.g. fungi) may be affected by differences in host plant range. Additional studies are therefore needed to dive deeper into the role host plant species plays in shaping milkweed aphid symbiont community composition and function, especially potential links between plant defensive chemistry and microbiome assembly. In general, further research investigating the generalist-specialist gradient using herbivores that feed across multiple plant species and families is needed to clarify the extent to which diet breadth shapes microbial communities (e.g. [24]).
In summary, we did not find evidence for a gradient in bacterial community diversity associated with variation in diet breadth for milkweed specialized aphid species. Instead, our results suggest overlapping host plant range and shared hosts can result in selection for common microbes and thus highly similar microbiomes across species. However, milkweed aphids do harbor unique bacterial populations that vary in strain diversity and relative abundance of Arsenophonus, although a handful of other well-known aphid symbionts were also detected in low abundance. These findings suggest that while diet breadth may not be a major driver of divergence in overall taxonomic composition of aphid symbiont communities, factors such as strain level variation and differences in abundance offer alternative routes to generating adaptive potential. Further research is needed to determine the functional or ecological role played by milkweed aphid facultative symbionts and different co-occurring strains.