Introduction

The diverse community of symbiotic microorganisms comprising an organism’s microbiome is recognized as a crucial component of its physiology (Rawls et al. 2006; Li et al. 2008; Bordenstein & Theis 2015) and behavior (Vuong et al. 2017). The microbial community that occupies an organism’s gastrointestinal tract (i.e., the gut microbiome) has been shown to be an especially important determinant of host fitness as it can regulate resource acquisition (Zhu et al. 2011), life history (Gould et al. 2018), and disease resistance (Lathrop et al. 2011). As such, the diversity and composition of the gut microbiome is emerging as an important aspect of the ‘holobiont’, or host-microbiome system (sensuZilber-Rosenberg & Rosenberg (2008)), that can itself be under strong natural selection. Determining the function of the gut microbiome (i.e., its effect on host physiology, health, and ultimately, evolutionary fitness) and the factors that shape diversity and composition of this community are two highly sought-after goals (Benson et al. 2010; Spor et al. 2011). While there are some accounts of conserved, core, microbiomes for particular species (Roeselers et al. 2011), determining the primary controls on microbiome composition is difficult because in nature gut microbiomes vary widely across individuals (Turnbaugh & Gordon 2009), space (Smith et al. 2015), and time (Nayak 2010). We use an unprecedented five-decade experiment to study the environmental and host-driven controls on gut microbiome and their potential to shape adaptive differentiation in the wild.
Bacteria enter the gut vertically from mother to offspring (Milaniet al. 2015; Beemelmanns et al. 2019) and horizontally through contact with the environment and food (Nayak 2010). Introduced taxa may be transient if they pass through the gut or die, or they can establish and become part of a self-perpetuating population. Thus, a gut community is determined by the availability of microbial colonizers in an environment, and filtering of those colonizers by host traits or genetics, and conceivably, could assemble in parallel with similar selection pressures and similar environments.
Host diet influences the gut microbiome through both colonization, as food carries microorganisms into the gut, as well as filtering, since diet can drive evolution of digestive traits. Studies at large taxonomic scales suggest diet is a stronger control on microbiome structure in the latter scenario, when it affects traits. For instance, the gut microbiomes of pandas are more similar to closely-related carnivores than distantly-related herbivores, presumably because their gut morphology evolved from a carnivorous ancestor (Karasov & Douglas 2013). The gut microbiome may be an especially strong modulator of host fitness in circumstances when there is a mismatch between traits and function – that is, host physiology does not allow for digestion of certain resources abundant in a new environment (e.g. cellulose in termites and plant toxins in mammalian herbivores), but the gut microbiome does (Abe et al. 2000; Kohl et al. 2014).
This mismatch among environment, host traits, diet, and microbiome is likely a consistent feature of host-microbiome co-evolution in the wild, but impossible to capture in most systems, which do not control for genetic background or examine populations that have already diverged. Studies can control for genetic background by rearing populations in controlled environment, but this excludes the complex way that drivers interact in nature (e.g. behavior and habitat interact to determine diet). Our study achieves an optimal middle-ground to understand the role of gut microbiomes in local adaptation: a ‘natural laboratory’ where replicate wild populations of known origin have undergone parallel phenotypic evolution to novel environments (Reznick et al. 1990, 1997; Gordon et al. 2015). In this system, populations of guppies (Poecilia reticulata ) originating from high predation (HP) Trinidadian streams were introduced to guppy-free low predation (LP) sites six times between 1957 and 2009 (Haskins, unpublished data, Endler 1980; Reznick & Bryga 1987; Travis et al. 2014). We capitalized on this time series of guppy transplant experiments conducted over the last five decades to understand the determinants of guppy gut microbiome structure and function.
Microbiome structure is likely to be shaped by several factors as high predation (HP) guppies evolve to low predation (LP) environments, several of which have been documented in native HP and LP ecotypes. First, LP environments are at higher elevations and tend to have less sunlight due to higher canopy cover and lower primary productivity (Zandonà et al. 2017). Thus, transplanted guppies may be exposed to a different pool of microorganisms to colonize their gut microbiome (Nayak 2010). As these environmental differences alter the types of resources available, guppies in LP environments eventually adopt a more omnivorous, lower-quality diet than HP guppies, which eat more nutrient-rich macroinvertebrates abundant in these sites (Zandonàet al. 2011). Finally, these differences in diet likely underlie observed differences in guppy gut length: LP populations have longer guts than HP populations (Zandonà et al. 2015), presumably to maximize nutrient absorption and energy extraction from lower-quality food types in lower-resource LP environments (Kapoor et al.1976).
Taking advantage of this ‘chronosequence’ of host evolution, we hypothesize that gut microbiota tracks the evolution of fish traits such as gut morphology, such that both will change linearly with time as populations inhabit a new predation regime. Previous work shows that gut microbiomes diverge across ecotypes, but not necessarily in parallel, perhaps due to stream-specific effects (Sullam et al. 2015). However, in this same study, gut microbiomes from HP and LP fish remained distinct even after being fed 10 weeks of the same diet in the lab, suggesting that host traits, genetic background, or high vertical transmission also controls the microbiome. If our hypothesis is supported, guppy gut microbiome of recent introductions (introduced to new LP environments 5-6 years prior to sampling) will be compositionally more similar to HP ecotypes, while gut communities sample from older introductions (30-60 years prior to sampling) will be more similar to LP native microbiomes because they have had more time to evolve predictable LP traits such as longer gut length.