Phenotypic associations at individual loci
Although 50 maternal families were initially selected for the greenhouse
experiment, only 42 had sufficient germination for measurements to be
included in analyses. Six out of the eight measured phenotypic traits
were significantly different in the drought treatment versus the wet
treatment. Height growth (GR) and shoot weight (SW) decreased, while
root length (RL), the root-shoot dry mass ratio (R2S), stomata density
on the adaxial side (SD_AD), and the number of stomatal rows on the
abaxial side (NR_AB) increased. We therefore focused on these traits
for the G2P association. We measure the association of SNPs to either
the average measurement of control treatment family for each trait or
the average change in the trait from wet to dry conditions (drought
responsiveness). Heritabilities of trait responses to drought ranged
from 0.15 to 0.65, and are discussed further in (Wu et al. 2023),
with variation in shoot growth in response to drought being particularly
variable.
More SNPs were associated with the trait drought responses (1,154) than
with the control traits (817). While control R2S had the most
associations and SW the least (Table 2), the opposite was the case for
drought responsiveness (Table 3). The number of SNPs associated with
more than one trait was low in both G2P analyses. The highest degree of
overlap was in control traits of RL and R2S (12 SNPs) and of R2S and
NR_AB (9 SNPs) (Fig. 3). The proportion of associated upstream SNPs was
similar across control traits (32-40%), but proportions of other
categories varied widely, with the proportion of missense SNPs ranging
from 8-25%. For drought response, the distribution of SNPs in all
categories differed, with the proportion of upstream being 19-34% and
the proportion of missense being 7-16% for traits other than R2S. R2S
was only associated with 6 SNPs, five upstream and one downstream.