Plant–microbe interactions can impact plant growth and performance, and contribute to local adaptation. However, few studies have explored the impact of microbial communities from distinct native locations on plant functional traits, and less is known about how host-microbe interactions affect the quantitative genetics of plant traits. We used a recombinant inbred line (RIL) mapping population derived from upland and lowland ecotypes of the diploid C4 perennial bunch grass Panicum hallii to explore quantitative genetic responses to soil microbiomes. We show that the growth and development of ecotypes and their trait divergence depends on soil microbiomes. Moreover, we find that broad-sense H2 is modified by soil microbiomes, revealing important plant genotype-by-microbiome interactions for quantitative traits. We detected a number of quantitative trait loci (QTL) that interact with the soil microbiome. Our results highlight the importance of microbial interactions in ecotypic divergence and trait genetic architecture in C4 perennial grasses.

Panicle traits exhibit quantitative variation controlled by genes, the environment, and their interaction. In switchgrass, a perennial biofuel crop, identification of quantitative trait loci (QTL) and QTL x E interactions controlling panicle architecture could aid breeding efforts and cultivar development by impacting seed productivity. In this study, we evaluate the genetic architecture of panicle traits including panicle length, primary branching number, and secondary branching number in an outcrossing switchgrass population grown across ten field sites in the central United States. We evaluated pleiotropic relationships between panicle traits and flowering time, tiller production and biomass. We also identified environmental factors correlated with QTL x E interactions and potential candidate genes underlying panicle trait QTL in switchgrass. Overall, our multi-environment mixed QTL model detected 18 QTL for panicle traits. Twelve of the QTL exhibited consistent effects (i.e., no QTL x E), and most (4 of 6) of the effects with QTL x E exhibited condition-specific effects. Many of the QTL x E effects were associated with yearly mean temperature and photoperiod. Panicle QTL co-localized with previously identified flowering time QTL and candidate genes associated with flowering, supporting a pleiotropic model of panicle development based on shared developmental genetics and responses to environmental signals.

Understanding the environmental drivers and the genetic architecture of traits involved in local adaptation is of fundamental interest as it may give insight to ecological speciation. Unfortunately, local adaptation studies rarely consider early life stages. As a result, we know surprisingly little about the genetic architecture and evolution of seed-based life history traits especially considering adaptation to xeric and mesic habitats. Here, we evaluate the genetic architecture of seed trait divergence in a recombinant inbred population derived from xeric and mesic ecotypes of Panicum hallii. We found the majority of seed related QTL had overlapping intervals, suggesting tight linkage and possibly pleiotropy are important in seed trait evolution. Our results suggest that the evolution of these traits is non-independent and may involve tradeoffs that affect the direction of divergence. Seed size and germination rate shared two colocalized QTL, each with antagonistic additive effects. This supports the hypothesis of a functional genetic trade-off between these two traits, resulting in either large seed/high dormancy or small seed/low dormancy trait combinations. We identified several candidate genes that may explain this trade-off. Overall, our study provides insights into the factors facilitating and potentially constraining ecotypic differentiation in plants through seed-based life history traits.