Effects of Rht1 haplotypes on plant height, anther retention and FHB resistance
Rht1 haplotype had a strong and significant effect on plant height in all NIL groups in greenhouse and field experiments (Table S3). The Rht1 haplotype effect on AR was significant in greenhouse experiments for NIL groups C1, CM and E4 and in field experiments for C1, CM and MI and marginally significant in the green house for MI (p = 0.052) and in field tests for E4 (p = 0.086). The impact of Rht1 haplotype on FHB incidence and AUDPC was significant in all NIL groups under field conditions. Rht1 haplotypes had a significant effect in greenhouse experiments on AUDPC and on the number of infection sites in group C1. Rht1 significantly influenced the number of infection sites in group E4 and had a weak impact in group MI (p = 0.062) under greenhouse conditions.
Comparing Rht1 haplotype groups
Plant height: Wildtype lines (B1aD1a) of group E4 had the tallest plant phenotype (≥100 cm) and were significantly taller than wildtype lines of groups C1 (90 cm) and MI (88 cm) and of groups C1 (86 cm) and CM (92 cm) in greenhouse and field experiments, respectively (Figs. 1, 2; Tables S4, S5). Plant height reduction of introgressed dwarfing alleles was significant in all four recurrent NIL groups and was highest in E4 NILs, followed by MI, and C1, and was lowest in CM (Figs. 1, 2; Tables S4, S5).
Rht-B1b and Rht-D1b dwarfing alleles led to similar decreases in height within NIL groups.
Anther retention: Average AR% of wildtype haplotypes in NIL groups C1 (31% or 33%) and CM (37% or 13%) was markedly lower than in E4 (54% or 58%) and was highest in MI (88% or 93%) in respective greenhouse or field experiments. There were significantly more anthers retained in dwarfing NILs in field and greenhouse experiments in all groups, but the difference between tall and semi-dwarfing B1b NILs was below significance in field experiments of NIL group E4 and in greenhouse experiments of NIL group MI (Figs. 1, 2). The Rht1-D1b allele had a significantly stronger effect compared to the Rht-B1b allele in NIL group C1 while in the NIL groups CM, E4 and MI no differences were observed between B1b and D1b isolines. The increase in AR in the dwarfed versus tall lines was highest in NIL group C1, followed by E4 and CM and was smallest in group MI. However, since the tall MI haplotypes had already nearly complete anther retention, there was little room for further increase.
FHB resistance traits: Generally, within NIL groups no significant differences in FHB susceptibility were found between the semi-dwarfing Rht-B1b and Rht-D1b haplotypes, although there was a trend towards higher susceptibility for D1b than for B1b carriers. Differences between tall and dwarfing lines were more pronounced under field conditions and were particularly high among NILs of groups C1 and E4 (Fig. 2). Rht1 haplotype did not affect FHB resistance level of CM and MI NILs in the greenhouse and had a relatively small impact under field conditions (Figs. 1, 2). The effect of the dwarfing alleles was low in the highly FHB resistant CM background and the dwarfed NILs remained highly resistant in greenhouse and field experiments. FHB severity in all MI NILs was high, regardless of height. The FHB severity levels in MI NILs were particularly high under field conditions. Infection of E4 NILs was relatively low in the greenhouse but high in the field.
Effects of Rht1 haplotypes in combination with FHB resistance QTL Fhb1 and Qfhs.ifa-5A on plant height, anther retention and FHB resistance
Analysis of variance revealed a significant effect of the Rht1 haplotype on plant height, AR%, FHB incidence and AUDPC in group C1 and on plant height and AUDPC in group CM. The status of the FHB resistance alleles significantly affected AR%, FHB incidence and AUDPC in group C1, and AR% and AUDPC in group CM. Significant QTL-by-Rht1 interaction was revealed for plant height in group CM and for FHB incidence and AR% in group C1 and CM (Table S6).
Comparing QTL-by-Rht1 haplotypic groups
C1 and CM wild type Rht1-haplotypes were consistently taller than their corresponding Rht-B1b or Rht-D1b variants, while only minor height differences were observed among NILs with contrasting QTL combinations at Fhb1 or Qfhs.ifa-5A (Figs. 3, S1, S2; Table 1). Generally, a higher proportion of anthers was retained in semi-dwarfing NILs compared to their tall variants (Figs. 3, S1, Table 1). The resistance alleles at Qfhs.ifa-5A clearly reduced AR in tall as well as semi-dwarf haplotypes. Differences were particularly high and always significant between Qfhs.ifa-5A carriers and non-carriers within Rht1 subgroups of C1 lines, while the differences were smaller in CM NILs and only significant within the tall and the semi-dwarfing Rht-D1b haplotypes (Fig S1). Fhb1 had no effect on AR. Tall NILs without the Qfhs.ifa-5A resistance allele had 10 to 30% more anthers retained than semi-dwarfing NILs with the Qfhs.ifa-5A resistance allele.
Table 1
Best linear unbiased estimators (BLUEs) across experiments of NIL group C1 and CM grouped by Rht1 haplotype and QTL combination
|
|
|
|
NIL group C1
|
NIL group CM
|
Trait
|
Rht1 haplotype
|
FHB resistance QTL
|
BLUEsa
|
BLUEsa
|
Plant height
|
B1aD1a
|
Fhb1
|
Qfhs.ifa-5A
|
83.4
|
b
|
90.2
|
b
|
|
B1aD1a
|
-
|
Qfhs.ifa-5A
|
85.0
|
b
|
93.0
|
b
|
|
B1aD1a
|
Fhb1
|
-
|
82.0
|
b
|
90.5
|
b
|
|
B1aD1a
|
-
|
-
|
82.2
|
b
|
89.0
|
b
|
|
B1bD1a
|
Fhb1
|
Qfhs.ifa-5A
|
71.6
|
a
|
79.7
|
a
|
|
B1bD1a
|
-
|
Qfhs.ifa-5A
|
70.6
|
a
|
79.7
|
a
|
|
B1bD1a
|
Fhb1
|
-
|
68.1
|
a
|
79.1
|
a
|
|
B1bD1a
|
-
|
-
|
69.4
|
a
|
77.8
|
a
|
|
B1aD1b
|
Fhb1
|
Qfhs.ifa-5A
|
67.5
|
a
|
81.6
|
a
|
|
B1aD1b
|
-
|
Qfhs.ifa-5A
|
68.1
|
a
|
82.0
|
a
|
|
B1aD1b
|
Fhb1
|
-
|
66.6
|
a
|
76.6
|
a
|
|
B1aD1b
|
-
|
-
|
67.2
|
a
|
77.5
|
a
|
Anther retention
|
B1aD1a
|
Fhb1
|
Qfhs.ifa-5A
|
27.8
|
a
|
13.4
|
a
|
|
B1aD1a
|
-
|
Qfhs.ifa-5A
|
29.8
|
a
|
10.7
|
a
|
|
B1aD1a
|
Fhb1
|
-
|
71.9
|
cde
|
27.7
|
abc
|
|
B1aD1a
|
-
|
-
|
78.0
|
def
|
22.0
|
ab
|
|
B1bD1a
|
Fhb1
|
Qfhs.ifa-5A
|
56.2
|
bc
|
26.0
|
abc
|
|
B1bD1a
|
-
|
Qfhs.ifa-5A
|
46.5
|
ab
|
25.2
|
abc
|
|
B1bD1a
|
Fhb1
|
-
|
87.1
|
eg
|
43.0
|
bd
|
|
B1bD1a
|
-
|
-
|
85.3
|
eg
|
41.3
|
bd
|
|
B1aD1b
|
Fhb1
|
Qfhs.ifa-5A
|
63.6
|
bcd
|
20.0
|
ab
|
|
B1aD1b
|
-
|
Qfhs.ifa-5A
|
60.3
|
bcd
|
15.6
|
a
|
|
B1aD1b
|
Fhb1
|
-
|
96.9
|
g
|
47.7
|
cd
|
|
B1aD1b
|
-
|
-
|
96.3
|
fg
|
56.7
|
d
|
FHB incidence
|
B1aD1a
|
Fhb1
|
Qfhs.ifa-5A
|
56.5
|
a
|
15.4
|
a
|
|
B1aD1a
|
-
|
Qfhs.ifa-5A
|
65.9
|
a
|
43.3
|
b
|
|
B1aD1a
|
Fhb1
|
-
|
90.8
|
bc
|
40.6
|
b
|
|
B1aD1a
|
-
|
-
|
93.4
|
bc
|
82.5
|
d
|
|
B1bD1a
|
Fhb1
|
Qfhs.ifa-5A
|
82.3
|
b
|
35.8
|
ab
|
|
B1bD1a
|
-
|
Qfhs.ifa-5A
|
88.5
|
bc
|
82.3
|
d
|
|
B1bD1a
|
Fhb1
|
-
|
98.0
|
c
|
52.3
|
bc
|
|
B1bD1a
|
-
|
-
|
97.8
|
c
|
90.5
|
d
|
|
B1aD1b
|
Fhb1
|
Qfhs.ifa-5A
|
88.5
|
bc
|
51.0
|
bc
|
|
B1aD1b
|
-
|
Qfhs.ifa-5A
|
94.3
|
bc
|
73.3
|
cd
|
|
B1aD1b
|
Fhb1
|
-
|
99.0
|
c
|
71.0
|
cd
|
|
B1aD1b
|
-
|
-
|
99.0
|
c
|
92.0
|
d
|
AUDPC
|
B1aD1a
|
Fhb1
|
Qfhs.ifa-5A
|
300.7
|
a
|
40.6
|
a
|
|
B1aD1a
|
-
|
Qfhs.ifa-5A
|
406.6
|
ab
|
119.2
|
ac
|
|
B1aD1a
|
Fhb1
|
-
|
687.9
|
cd
|
95.5
|
ab
|
|
B1aD1a
|
-
|
-
|
796.8
|
cf
|
339.6
|
ac
|
|
B1bD1a
|
Fhb1
|
Qfhs.ifa-5A
|
454.1
|
ab
|
70.9
|
ab
|
|
B1bD1a
|
-
|
Qfhs.ifa-5A
|
584.3
|
bc
|
276.8
|
ac
|
|
B1bD1a
|
Fhb1
|
-
|
804.4
|
cf
|
123.7
|
ac
|
|
B1bD1a
|
-
|
-
|
896.1
|
df
|
459.6
|
bc
|
|
B1aD1b
|
Fhb1
|
Qfhs.ifa-5A
|
609.0
|
bc
|
91.6
|
ab
|
|
B1aD1b
|
-
|
Qfhs.ifa-5A
|
741.4
|
cde
|
260.3
|
ac
|
|
B1aD1b
|
Fhb1
|
-
|
915.9
|
ef
|
206.8
|
ac
|
|
B1aD1b
|
-
|
-
|
997.3
|
f
|
528.9
|
c
|
a BLUEs of groups with different letters are significantly different (Bonferroni p-value adjustment, p < 0.05)
|
The introgression of Qfhs.ifa-5A alone or in combination with the Fhb1 resistance allele into the C1 background decreased FHB incidence and AUDPC in tall and dwarf haplotypes (Figs. 3, S1). The ranking of subgroups within Rht-B1aD1a, Rht-B1bD1a or Rht-B1aD1b haplotypes was constant whereby NILs with Qfhs.ifa-5A and Fhb1 were most resistant, followed by NILs containing Qfhs.ifa-5A only, and NILs carrying Fhb1 alone were almost as susceptible as NILs without any resistance QTL. Differences between QTL subgroups were larger within the tall Rht1 lines (Fig. 3, Table 1). QTL subgroups that had both QTL introgressed were usually significantly more resistant than lines containing solely Fhb1 or no resistance QTL but were in most cases not different from NILs which contained only the Qfhs.ifa-5A resistance QTL (Fig S1). Semi-dwarfing NILs containing the Qfhs.ifa-5A QTL alone or in combination with Fhb1 were consistently slightly more resistant than tall variants without the Qfhs.ifa-5A QTL.
Rank order of QTL subgroups within Rht1 haplotypic groups of CM NILs was consistent, whereby CM NILs having both QTL combined performed best, second were Fhb1, closely followed by Qfhs.ifa-5A NILs, and lines without a resistance QTL were most susceptible (Figs. 3, S1, Table 1). Resistance to FHB incidence was significantly better in lines with both QTL combined compared to NILs without a resistance QTL, and NILs containing solely Fhb1 performed equally or better than NILs with Qfhs.ifa-5A only or no QTL (Fig S1). AUDPC remained generally low and was only significantly different in Rht-D1b haplotypes between NILs with both resistance QTL and NILs with no resistance QTL.