Sorghum ergot continues to pose a significant threat to the sorghum industry since its worldwide outbreak during the mid 1990s. Identification of putative sources of resistance against this disease has only been reported recently. The present study utilised the IS8525 resistance source for studying the genetics of resistance, and QTL analysis to gain insight into the resistance mechanism of the disease.
An F5 recombinant inbred line (RIL) population was developed from the cross between a commercially accepted restorer line 31945-2-2 and IS8525. Genetics of ergot resistance was studied in 292 RIL individuals including two parental checks. A subset of I34 RIL individuals was used for mapping and QTL analysis. Two pollen traits, pollen quantity (pq) and pollen viability (pv) were included in the study, as previous reports described the mechanism of ergot resistance in sorghum as pollen based.
Two field trials were conducted to infer the genetics of ergot resistance under artificial epiphytotic conditions and study the pollen traits. Percent ergot infection (pcergot) in IS8525 was very low and in contrast, 31945-2-2 was heavily infested in all inoculation dates. The BLUP value of each RIL was determined from the sampling dates using a linear mixed model. The distribution of the trait data (BLUP) along the frequency histogram was normal, suggesting that the traits are polygenic. High heritability values (broad-sense) were obtained for pcergot and pq and moderately high for that qf pv, and coupled with high genetic variance, particularly for pcergot, indicating that ergot resistance could be manipulated through breeding Moderate to high genotypic correlations within and between year's observations (sampling dates) indicate that the genotypic performance of the lines developed from this cross would be predictors for ergot resistance across environments. The genetic correlation observed between pq and pv was moderately positive and with pcergot was moderately negative, indicating that there might be some common genetic factors controlling the traits of ergot development. The R2 value however suggests only a small part of the total variability, 11% and 9%, due to the variability of pq and pv, respectively.
A genetic linkage map of 31945-2-2 x IS8525 was constructed integrating 286 markers including 55 SSRs, 229 AFLPs and two morphological trait loci. The constructed map had 16 linkage groups (LGs), spanned a total length of 1599.1 cM with an average distance of 5.57 cM between loci along the chromosome. The 16 LGs of the present map could be aligned to 10 recently published sorghum chromosomes SBI-01 to SBI-10 on the basis of 50 shared markers between this map and two previously published maps of sorghum. A high degree of similarity was observed for the marker order and map position of the shared SSR loci indicating the suitability of the constructed map for QTL analysis and exchange of map information. One of the morphological marker loci, coleoptile colour (CC) mapped in a position similar to the plant colour gene and that of red leaf (RL) was reported for the first time. The level of SSR polymorphism was high (75%) and that of AFLP was low (16.9%) but comparable to the polymorphism level of the parents of other sorghum mapping populations. A high degree of distortion of the AFLP loci was observed. In spite of that AFLP was found to be an excellent source of genome coverage in this study as many AFLP loci were mapped along the telomeric regions of the chromosomes.
Composite interval mapping (CJM) identified as many as 11, 7 and 5 QTLs linked to molecular markers for pcergot and pq, and pv based on LOD \ 2.0. However, the most significant QTLs for pcergot were resolved on SBI-01, SBI-02, SBI-06 (2-QTLs), SBI-07 and SBI-08 on the basis of permutation based LOD > 2.95. All but one QTL on SBI-08 were imparted by IS8525, the resistant parent. These QTLs were verified in 94 additional RILs and explain 51.5% of the total phenotypic variation for pcergot indicating their possible use for marker-assisted selection. However, three of these main effects QTLs (2-QTLs on SBI-06 and one QTL on SBI-08 from 31945-2-2) were involved in epistatic interactions.
Co-location of QTLs for pcergot, pq and pv was found on SBI-02 and for pcergot and pq were found on SBI-06 and SBI-10. All other QTLs for the three traits were found unlinked, mapped either to the same chromosome but some distance apart or on separate chromosomes. This indicates that both pollen and non-pollen based mechanisms are operating for ergot resistance in this population of sorghum.
A total of 22 QTLs were reported for morpho-physiological traits, including for height characters and flowering. Some of these have been previously reported, however, new QTLs for peduncle length and flowering were identified. The overlapping of some of the identified QTLs for agronomic traits and pcergot, pq and pv were observed. This implied that the regions particularly on SBI-01, SBI-02 and SBI-07 simultaneously controlled the traits for which the QTLs overlapped.
In this study, for the first time, QTL associated with pcergot and two pollen traits were identified and the genetic relationship of the pollen traits with ergot resistance was determined. The markers identified for ergot resistance and some of the traits of agronomic importance will be valuable resource for future marker-assisted selection and improvement of these traits in sorghum breeding and development.