Sorghum leaf blight and northern corn leaf blight, both caused by Exserohilum turcicum, are major diseases of sorghum and maize, respectively. Examining the genetic architecture of resistance in sorghum will lead to a better understanding of the relationship between resistance in sorghum and maize, which can ultimately enhance management options in both crops. In 2018 and 2019 we evaluated two sorghum recombinant inbred line (RIL) populations for resistance to E. turcicum. The BTx623 x IS3620C and BTx623 x SC155 populations consisted of 235 and 81 RILs, respectively. Resistance in both populations was moderately to highly heritable. We identified a total of six quantitative trait loci (QTL) across the two populations. Three QTL with small to moderate effect sizes were identified in the BTx623 x IS3620C population. Three QTL, including a large-effect QTL on chromosome three that explained 24% of the variation, were identified in the BTx623 x SC155 population. We compared the identified QTL with the position of northern corn leaf blight candidate genes and found eight candidate resistance gene orthologs that colocalize with the sorghum leaf blight QTL. There were also several nucleotide-binding leucine rich repeat encoding genes within the candidate intervals. Understanding host resistance in multiple species furthers our understanding of the Exserohilum turcicum pathosystem.

We evaluated two recombinant inbred line (RIL) populations. SC155 is more resistant than BTx623 to Colletotricum sublineolum, the causal agent of anthracnose (Patil et al. 2017). IS3620C has been noted to be genetically divergent from BTx623, and the two parents segregate for numerous agronomic traits (Burow et al. 2011; Kong et al. 2018) IS3620C is a guinea line (Hart et al., 2001), while SC155 is included in the sorghum conversion panel (Patil et al., 2017).

Both RIL populations were grown in the field as single row plots in the summer of 2018 and 2019 at the Crop Sciences Research and Education Center in Urbana, IL. Standard agronomic practices for central Illinois were employed. Populations were grown under irrigated conditions in 2018 and 2019 to favor disease development. For the BTx623 × IS3620C population, 246 and 235 lines were evaluated in 2018 and 2019, respectively. For the BTx623 × SC155 population 86 and 81 lines were evaluated in 2018 and 2019, respectively. Differences in the number of lines evaluated in each year was due to seed availability. Experiments were designed as a randomized complete block design with two replications for both populations in 2018 and 2019 using the agricolae package (De Mendiburu, 2014) in the statistical software R, version 3.6.0 (R Core Team, 2021). The respective parents for each RIL population were included in each replication.

Infested sorghum grains were used as the source of inoculum. Inoculum was prepared as described by X. Zhang et al. (2020) using the isolates 15st003, 16st001, and 15st008. These strains were isolated from diseased sorghum leaves in Illinois in 2015 and 2016. The isolates were cultured in vitro for two weeks on lactose-casein hydrolysate agar with a 12-hour light/dark photoperiod. Once all plates had growth covering at least one-third of the surface, 1000 mL of soaked, autoclaved sorghum grain was cultured in a mushroom bag with agar pieces taken directly from the culture plates. Two agar plates with at least one third coverage of fungal mycelia were placed in each mushroom bag. Bags were cultured for 2-3 weeks at room temperature with a 12-hour light/dark photoperiod. After grains were fully infested and spores were visually confirmed using a dissecting microscope, the grain was dried and stored at room temperature until needed for inoculation. Each plant was inoculated with 1.3 g of the prepared inoculum in the whorl at the V5 growth stage, or 29 days after planting in 2018 and 35 days after planting in 2019 (Abendroth et al., 2011).

Each plot was evaluated for percent disease coverage using a 0-100 scale in 5% increments, with 0% indicating no disease and 100% indicating that all the leaf area of the plants in the plot were necrotic with disease (Poland & Nelson, 2011). Both populations were rated four times starting two weeks before the earliest maturing plants started flowering. Plots were rated every 7-14 days. The area under the disease progress curve (AUDPC) was calculated for each plot using the agricolae package (De Mendiburu, 2014). Days to anthesis (DTA), defined as the day when at least 50% of the plot had visible anthers emerged, was recorded for both populations in 2019.

Using the AUDPC data for each population, mixed models were fitted in R/lme4 (Bates et al., 2015). Factors defined as follows: Y_ijk represents the measured AUDPC value from the genotype i in replicate j in environment k; μ represents the grand mean; G_i represents the random effect of genotype i; R_j(k) represents the random effect of replication i nested in environment k; E_k represents the random effect of environment k; GE_ik represents the random effect of the interaction between genotype i and environment k; and ε_ijk represents the random error associated with Y_ijk were included in the model. A best linear unbiased predictor (BLUP) was calculated for each line using the ranef function. For the BTx623 × SC155 population the genotype × environment interaction was included in the model, but it was not included in the model for the BTx623 × IS3620C RIL population. BLUPs were extracted from this model.