Genome scan identifies flowering-independent effects of barley HsDry2.2 locus on yield traits under water deficit

The wild barley HsDry2.2 allele confers a reproductive advantage that is independent of flowering time under drought conditions and is manifested through modifications in shoot morphology, reduced senescence, and longer grain filling.

with the initiation of the late drought treatment, 13 days before booting (Fig. S2E).

8
Pots were weighed manually before and after irrigation, keeping the well -watered 2 2 9 (WW) pots between around 60-90% of field capacity and the water limited pots (WL)

5 1
Pot experiment: Booting time, defined as the date at which the three spikes awns are 2 5 2 first visible in a pot (spikes are tagged), was daily recorded and used to score HEA.

5 3
MAT was recorded when the three first (tagged) spikes dried, which was right 2 5 4 followed by the rest of the plant drying out. GFP was calculated by subtracting HEA 2 5 5 from MAT. All aboveground biomass per pot was harvested at full grain maturity; 2 5 6 fertile spikes were counted to assess the number of spikes per plant, separated from 2 5 7 the vegetative organs (stems and leaves), and both were oven-dried (80°C or 38°C for 2 5 8 48 h for vegetative organs and spikes, respectively) and weighed to determine spike 2 5 9 dry matter and TDM. Spikes were threshed and total grain weight was determined. LogP=7.9) for GN.

5 2
Next, the GWAS model was modified to identify QTL by-environment were identified in VDW and no significant loci were identified for plant height ( Table   3 5 6 S5). In contrast, a larger number of interactive loci were associated with PGY or GN  HsDry2.2, a major locus with non-pleiotropic QxE interactions 3 6 3 One major locus that appeared as a pleiotropic QTL that regulates many traits is 3 6 4 located on the long arm of chromosome 2 and delimited by SCRI_RS_144592 and  allele effects under optimal as well as early and late water limitation, i.e. starting at 3 9 2 transplanting or at stem elongation stage (52 days after planting; 52DAP), 3 9 3 respectively. Type of traits included plant productivity, phenology, canopy structure 3 9 4 and leaf senescence.

9 5
Overall plant performance: Irrigation treatment effect was found significant for most 3 9 6 of the measured variables ( Fig. 5

0 9
The number of spike per plant was also reduced by the early but not by the late WL 4 1 0 (14%). Irrigation treatment showed no significant effect on grain number (Fig. 5 (Fig. 6A). However, the wild allele conferred an advantage in PGY 4 1 9 averaged across treatments (Fig. 6B). Under the WW treatment this advantage was 4 2 0 found most pronounced. Harvest index of the wild allele carriers was significantly 4 2 1 higher than in the cultivated allele group, especially under the Early WL (Fig. 6C).

2 2
Spike number per plant did not significantly differ between genotypes under the 4 2 3 different treatments, or averaged across treatments (Fig. 6D). The first three spikes to

3 3
Phenology: The first three spikes to head were used to determine plant phenology.

3 4
The Hs/_ group booted significantly earlier than Hv/Hv plants (

2
The wild allele was generally associated with longer and narrower leaves; Both

0 3
Interestingly, the prevalence for conditional QTL is by far more prevalent for isogenic lines for this wild allele in several genetic backgrounds and with larger plots.

1 3
It is argued that since drought survival is a trait under strong evolutionary 5 1 4 pressure, many drought survival loci would be expected to impart tolerance in crop cultivated genetic background, as in our study, we would not necessarily expect to 5 2 2 find alleles from the wild that would increase grain number and yield under stress.

2 3
Rather, the alleles that we expect to find are such that stabilize or buffer the effects of (~100days), is probably related to differences in thermal degrees days between 5 3 2 experiments ( Fig S6). Hence, in the relatively hot climate of the pot experiment HEA 5 3 3 was reduced for both alleles, therefore maintaining the differences between them yet 5 3 4 to a lesser extent in the pots than in the field. Furthermore, under these hot conditions 5 3 5 both genotypes flowered early, ruling out drought escape mechanism of one allele 5 3 6 compared to the other. Alternatively, the earlier flowering and elongated grain filling 5 3 7 of the wild allele may in part underlie its improved productivity.

3 8
Flowering-independent QxE effects of HsDry2.2 on reproductive traits 5 3 9 The most reproducible and significant QxE locus was identified on the long arm of 5 4 0 chromosome 2H (Fig. 2,3). Interestingly, this position matches the location of the on the total plant grain number, but no such interaction with flowering time (Fig. 4).

4 5
Both field trails and pot experiment shows that the wild allele had no 5 4 6 advantage over the cultivated one in vegetative or TDM production, yet , it shows 5 4 7 superiority in terms of reproductively (PGY and HI). This phenotype presented is     Table S1). Blue