Box 2: What’s up with the microbes?
Microbe can affect host phenotypes: 1. by affecting important plant functional traits(Friesen et al. 2011), 2. acting as an extended phenotype (i.e. acquiring resources otherwise not available, Koskella & Bergelson 2020) or 3. by serving as an environmental cue (Metcalfet al. 2019). However, understanding the characteristics/identity of the microbes underlying these effects can be complicated. Demonstrating reciprocal evolutionary change in microbes is complicated when in many cases fitness is hard to quantify and relevant traits are unknown, although new methods are making this seem possible(Burghardtet al. 2018). Documenting differences in community composition is more tractable given the explosion of sequencing-based technologies (Hugerth & Andersson 2017). However, detecting changes in microbial richness or shifts in community composition does not exclude the possibility of simultaneous (and relevant) genetic change to microbial communities or microbial plasticity. Additionally, microbial diversity may not be a relevant reflection of microbial function. At this point, understanding and differentiating ecological from evolutionary effects on microbial communities is probably not possible given current technologies but will be an important research goal of the future. Given the issues with understanding the microbial end of the interaction, researchers more commonly explore the population biology of the plants. For example, in one study(Johnson et al. 2010) highlighted in the main text, reciprocal transplants of Andropogon gerardiigenotypes collected from areas that either had limiting phosphorus or limiting nitrogen demonstrated that local mycorrhizal communities were locally adapted to local nutrient conditions, such that home communities of AMF provided optimal benefits to plants when transplanted into home environments. AMF also had higher fitness when matched with their sympatric plant genotype. What is unclear from this study is which microbial players were responsible for the observed effect. Some options include: 1) AMF genotype(s) that were particularly good at accessing nutrients locally, 2) AMF species that were particularly good at accessing nutrients locally, or 3) synergistic effects of multiple AMF mutualists, such that the effects on plants could not be predicted based on the presence or abundance of a particular species or genotype.