Supportive Information Legends
Fig. S1 Panels: (a), phylum level distribution of microbiota
from natural habitats; (b), phylum level distribution for microbiota
from greenhouse samples; (c), differentially abundant phyla between soil
treatments for mock and native microbiota where black boxes around tiles
indicate a significant difference (adjusted P < 0.05) between
soil treatments and the red color indicates a log fold change favoring
the AI microbiota while blue favors CI microbiota;
Fig. S2 Phylum level differences with plants growing in their
native habitats. Black boxes around tiles indicate a significant
different between Austin Inoculated (AI) and Corpus Inoculated (CI)
environments (adjusted P value < 0.05). A red color indicates
a higher abundance in the AI environment, while a blue indicates higher
abundance in the CI environment.
Fig. S3 Mock treatment microbiota are more similar
than native treatment microbiota. The graph displays Bray-Curtis
dissimilarities comparing inoculation within mock (Mock Austin
Inoculated (MAI) vs. Mock Corpus Inoculated (MCI)) and native (Austin
Inoculated (AI) vs. Corpus Inoculated (CI)) treatments levels.
Fig. S4 Field and greenhouse derived samples host non-identical
microbiota. Panels: (a), PCoA graph displaying all samples collected in
the study; (b), phylum level differences between greenhouse and field
grown samples.
Table S1 Nutrient and mineral composition of native soils used
for microbial inoculum.
Table S2 Means + SE for the parental ecotypes, RILs, RIL range
and broad-sense heritability (H 2 + SE) for Mock
Inoculated (MI), Austin Inoculated (AI), and Corpus Inoculated (CI)
microbial treatments.
Table S3 P -values for genetic and microbial treatment
effects of root and shoot traits for the Panicum hallii parental
ecotypes across three microbial treatments (Mock Inoculates, Austin
Inoculated and Corpus Inoculated).
Table S4 Comparison of “base” and “GxE” linear mixed models
to evaluate the impact of the microbiome on the quantitative genetic
architecture of our measured traits (The Diagonal model allows
Va or Vaa to vary by treatment).
Table S5 QTL effects (main and epistatic) of for thePanicum hallii RIL population.
Table S6 Full model analysis of QTL–treatment interactions
using PROC mixed in SAS with QTL modeled on the marker nearest the QTL
peak.
Table S7 Main and epistatic effects of GxE QTL for thePanicum hallii RIL population.
Table S8 Tests of effect slices of significant TRT x Marker
interaction in the full model analysis of QTL x treatment interactions
using PROC mixed in SAS with QTL modeled on the marker nearest the QTL
peak.