|Rice Yellow Mottle Virus sobemovirus (RYMV) is highly variable and several resistance breaking strains have been identified. This makes breeding for durable resistant very complicated since a variety proven to be resistance to the dominant strain in one location may succumb in another location where a different strain exists. One strategy to improve the durability of RYMV is to pyramid genes for both complete and partial resistance. However, the appropriate sources of resistance to RYMV need to be identified and characterised. In this study, the mode of inheritance of resistance to RYMV strain in Uganda was determined for nine (9) parental lines that had been recently introduced into the rice breeding program. Five of these have been reported to be resistant to RYMV and include upland NERICA varieties 8, 11, 12, 13 and an O. Sativa lowland variety, Gigante, while three lowland NERICA’s, 4, 6, 10 and IR64 (an O. Sativa lowland variety), are susceptible. RYMV inoculums was collected from three main rice growing areas in Uganda that represent “hot-spots” for RYMV disease. These were Iganga in Eastern Uganda, Lira in Northern Uganda, and Luweero in Central Uganda. When the isolates were tested on RYMV susceptible cultivar (IR64), preliminary results indicated that the isolate from Iganga (in Eastern Uganda) was most virulent, and was thus used in this study. The several isolates that were collected are presently being characterised.
Four patterns of reaction to RYMV were observed among the cultivars; (a) four highly susceptible lowland cultivars (N-4 lwl, N-6 lwl, N-10 lwl and IR64 lwl) showed severe symptoms, (b) one susceptible upland cultivar (N-12) had moderate symptoms, (c) three moderately resistant upland cultivars (N-8 upl, N-11 upl and N-13 upl) had mild symptoms, and (d) a resistant lowland cultivar (Gigante lwl) had very mild RYMV symptoms. A full diallel mating design was used to generate F1 and F2 progeny. Segregating F2 progeny were planted in both field and screen house in an α-lattice design, and inoculated by standard mechanical procedures. GCA effects were significant (P≤0.001), and were larger than SCA effects. Both the field and screen house experiments expressed strong genotypic differences. GCA, SCA and reciprocal effects were significant in both experiments (P≤0.001). The narrow sense coefficient of genetic determination based on genotype means was relatively high in both screen house (76%) and field (84%) experiments. These results suggest that additive gene effects were more important than non-additive effects, and that selecting as early as F2 or F3 generations can be effective. However, the fact that reciprocal effects showed a consistent tendency of F2 progeny to reflect the resistance level of the female parent indicates the need for careful choice of male and female parents in hybridisation programs to achieve RYMV resistance in the offspring. The Chisquare analysis of phenotypic evaluation confirmed the 3S:1R ratio of a single recessive gene in the cross IR64 (susceptible) x Gigante (resistant) and a two-gene ratio of 13S:3R in the cross N-4 lwl (susceptible) x N-11 upl (resistant). Additionally, chromosomal segments reported to be linked to a resistant gene for RYMV were screened using appropriate SSR markers. Of the 17
SSR primer pairs used, 7 (41%) were polymorphic between RYMV resistant (N-11 upl) and susceptible (N-4 lwl) parental genotypes. Results from three markers chosen for evaluation matched the expected 1:2:1 ratio of co-dominant markers for the F2 segregating population. Significant coefficient of determination (R2) values for SSR markers RM 17 (0.28***) and RM 252 (0.21***) suggested a close association with phenotypic scores for RYMV resistance. Information generated through this study will provide guidance for breeding for resistance to RYMV in Uganda.