Introduction
Major evolutionary innovations and events that span the tree of life are the result of host interactions with microorganisms. Perhaps the most dramatic example of microbe-mediated evolutionary events occurred when symbioses with the bacterial predecessors of chloroplasts and mitochondria were incorporated as part of the host cell (Sagan 1967). In more recent evolutionary history, the gut microbiome has been implicated in the rapid diversification of herbivorous mammals (Price et al.2012) and wasps (Brucker & Bordenstein 2013). In these cases, interactions with microbiota transformed the macroevolutionary trajectory of their hosts. There is accumulating evidence that microorganisms are also affecting patterns of host adaptation on microevolutionary timescales.
The human microbiome project as well as work with other vertebrates, insects, and plants has illuminated a complex feedback loop of host affecting microbial form and function and microbial form and function feeding back to affect host phenotype (Fig. 1A; Kohl et al. 2014; Mueller & Sachs 2015; Sanders et al. 2015; Weese et al.2015; Gehring et al. 2017; Moeller et al. 2019; Petipaset al. 2020a). Host genotypes enrich for specific microbiome components. For example, the mycorrhizal communities associated with Pinyon pines are almost entirely determined by pine morphotype (Gehringet al. 2017), soil microbiomes differ across Arabidopsis thaliana genotypes (Bulgarelli et al. 2012; Lundberg et al. 2012), and both mouse (Benson et al. 2010) and human (Goodrich et al. 2014) genotype shape their respective microbiomes. Reciprocally, this variation in microbiome composition can affect host phenotypes and performance. Across host organisms, the microbiome can affect nutrient acquisition (Krajmalnik-Brown et al. 2012; Newell & Douglas 2014), for example, access to phosphorus is often mediated by the unique enzymatic capabilities of arbuscular mycorrhizal fungi (Smith & Read 2008). Microbes also affect stress tolerance (Bang et al. 2018), immune phenotype (Foster et al. 2017), and pathogen susceptibility (King et al. 2016). For example, microbial symbionts of aphids provide their insect hosts with enhanced heat tolerance (Russell & Moran 2006), and Clostridiumin the human gut produces butyrate, a compound essential to host immune homeostasis (Velasquez-Manoff 2015). Often these changes are assumed to be adaptive (i.e. increasing host fitness, Kohl & Carey 2016) but this assumption is rarely tested. If these microbial effects increase host fitness then they can lead to microbe-mediated adaptation , defined as enhanced host fitness in a particular environment that is partially or entirely the result of interacting with microorganisms.
Adaptive responses can occur through local adaptation or adaptive plasticity, two non-mutually-exclusive responses to the heterogeneous selection pressures species experience in nature. Local adaptation is the result of genetic differentiation in response to local conditions and is manifest when local genotypes have higher fitness in their home habitat compared with foreign genotypes (Kawecki & Ebert 2004). Adaptive plasticity is a form of phenotypic plasticity and is manifest when the environment affects organismal traits in ways that increase fitness in that particular environment (i.e., local phenotypeshave higher fitness in their home habitat compared with foreignphenotypes ; Dudley & Schmitt 1996). Both local adaptation and phenotypic plasticity might be influenced by microbes, where in the absence of microbes you might observe low fitness and no pattern of local adaptation (Fig. 1B). We propose that microbe-mediated adaptive responses are the result of microbe-mediated local adaptationwhen local host genotypes have higher fitness than foreign genotypes because of a genotype-specific affiliation with locally important microbes (Fig. 1C), or microbe-mediated adaptive plasticitywhen local host phenotypes have higher fitness than foreign phenotypes as a result of interactions with locally important microbes (Fig. 1D).
Although microbe-mediated adaptation (including both microbe-mediated local adaptation and microbe-mediated phenotypic plasticity) may occur for many taxa (Alberdi et al. 2016; Sharpton 2018; Trevellineet al. 2019; Moeller & Sanders 2020), here we focus primarily on plants for three reasons: 1) The foundation for investigations into microbe-mediated adaptation have been laid through decades of avid interest in plant-microbe interactions. 2) Plants are tractable experimental systems amenable to classic experimental designs for testing local adaptation and adaptive plasticity, and 3) as sessile organisms, plants cannot move to escape stress and therefore may be even more dependent on microbes for adaptive responses. Additionally, while many interactions with microorganisms may be antagonistic (reducing plant fitness), here we focus on local adaptation to beneficial microorganisms as they have the potential to affect host adaptive responses, thus providing a unique avenue to adaptation. The population level consequences of antagonistic interactions have been extensively discussed elsewhere (e.g. Thompson 2005).
There are a growing number of examples showing that microbes canaffect adaptive plant responses (Chanway et al. 1989; Schultzet al. 2001; Johnson et al. 2010, 2013; Smith et al. 2012; Lau & Lennon 2012; Lankau & Nodurft 2013; Wagner et al. 2014; Pickles et al. 2015; Barrett et al. 2016; RĂșaet al. 2016; Van Nuland et al. 2016; Revillini et al. 2016; Gehring et al. 2017; Porter et al. 2020). Here we provide a framework that first identifies and defines the potential patterns and processes underlying microbe-mediated effects on adaptation (i.e., microbe-mediated local adaptation and microbe-mediated adaptive plasticity) and then propose empirical approaches to identify and differentiate between these two modes of microbe-mediated adaptation. Finally, we discuss implications and propose future research directions in the study of microbe-mediated adaptation.