Abstract
Natural plant populations often exhibit marked differences in gene
expression patterns that can reflect heterogeneity in selective
pressures. Analyzing gene expression as a quantitative trait provides a
unique opportunity to evaluate the underlying genomic basis of a
plethora of traits and their interactions in driving adaptive evolution.
We investigated patterns and processes driving expression
differentiation under conditions mimicking future climates by combining
common garden experiments with transcriptome-wide datasets obtained from
hybrid populations of Pinus strobiformis and P. flexilis .
We found strong signals of genotype-environment interactions (GEI) at
the individual transcript and the co-expression module levels suggesting
a marked influence of drought related variables on adaptive evolution.
Overall, survival was positively associated with P. flexilisancestry, but it exhibited an environment-specific pattern.
Co-expression modules exhibiting strong associations with survival and
genomic ancestry were representative of similar functional categories
across both gardens. Using network topology measures, putatively
adaptive garden-specific expression traits were pleiotropic and belonged
to modules exhibiting high population differentiation yet low
preservation across gardens. Overall, our study suggests the presence of
substantial genetic variation underlying univariate and multivariate
traits in novel climates that may enable populations of long-lived
forest trees to respond to rapid shifts in climatic conditions in early
seedling stages when mortality tends to be the highest. Our finding of
pleiotropic trait architectures underlying adaptive traits, however,
implies rapid adaptive responses to changing selection pressures depend
on whether trait covariances align with the direction of change in
selection pressures.