Sugarcane leaf scald disease is an important bacterial disease affecting sugarcane yield. Selection of disease-resistant genotypes can effectively reduce the incidence of this disease. This study aimed to explore the leaf-scald resistance of sugarcane genotypes, standardize resistance evaluation method, and provide a basis for the selection and utilization of germplasm resources of sugarcane.

Xanthomonas albilineans JG43 isolated from Guitang 46, was used as inoculum on 70 different sugarcane genotypes using the decapitation method by placing 500 mL of bacterial suspension on the surface previously cut above the apical meristem with scissors dipped in the inoculum suspension of 10⁸ CFU/mL. The disease incidence (IC) was calculated at 14, 28, 42, 56, and 70 days post-inoculation (Dpi). The disease index (DI) and the area under the disease progress curve (AUDPC) were calculated according to the disease severity of leaf scald in sugarcane. Variance, principal component, and discriminant analysis were performed using SPSS 25.0 software. Among them, a general linear model procedure (PROC) and the square sum model of type III were used to analyze the variance, with IC, DI and AUDPC as dependent variables, genotype, block and days post-inoculation as fixed factors. After the original data were processed by standardization (Z-score), principal component analysis was carried out by KOM and Bartlett sphere test. The Euclidean metric was calculated for cluster analysis using the WPGMA method of DPS 9.50 software. The discriminant analysis was performed to evaluate the clustering results according to Fisher’s criterion.

Some genotypes displayed white pencil lines at 14 dpi, then gradually expanded to the edge at 28 dpi. The leaves began yellowing or albinism from the edge to the veins at 42 dpi, then curled inward and died at 56 dpi. The severely infected plant withered and eventually died at 70 dpi. Variance analysis exhibited highly significant effects for IC, DI, and AUDPC among genotype (Gen), days post-inoculation (Dpi), and their interactions effect (Gen × Dpi) (P<0.01). Approximately 42% of the total sum of square was attributed to Dpi effect, indicating significant differences among genotypes resistance across the days post-inoculation. At 56 dpi, the disease reached a steady plateau, and the data in this period could be better divided among sugarcane genotypes. The results of discriminant and cluster analysis showed that 70 genotypes were divided into five different groups, including 15 highly resistant, 14 resistant, 15 moderate, 11 susceptible, and 15 highly susceptible genotypes.

The resistance of sugarcane genotypes to leaf scald was assessed using the decapitation method, the IC, DI and AUDPC at 56 dpi were used as the evaluation indicators. The combined method of clustering and discriminant analysis could improve the accuracy of clustering results. Fifteen genotypes of high resistance to leaf scald were assessed and used for the sugarcane breeding program in China, including Zhongzhe 9, Zhongzhe 4, Zhongzhe 2, GUC19, GUC8, Yunrui 03-103, Yunrui 05-649, Yunrui 05-182, Yunrui 05-367, Yunrui 89-159, Funong 11601, Funong 09-4059, Guitang 02-467, Guitang 08-297, ROC22.