Search

We have identified 25 MGR-fingerprint lineages of the rice blast fungus in southern India in samplings from 1991–95; some lineages appear to be geographically limited to certain states. In artificial inoculation assays, each lineage has a definable virulence spectrum on rice cultivars that is limited by avirulence shared by all lineage members. We are testing a lineage-exclusion resistance breeding strategy, i.e., combining resistances that, in complement, exclude all lineages in the pathogen population, in contrast to the conventional strategy that combines resistances to exclude detected pathotypes. In two-semester trials in a blast nursery at Pattambi, Kerala we evaluated field performance of rice lines constructed to pyramid two dominant blast resistance genes from the near-isogenic lines C101A51 (bearing Pi-2(t)) and C101LAC (bearing Pi-1(t)); previous inoculation assays indicated that no lineage in Kerala or adjoining states was compatible with Pi-2(t) and that Pi-1(t) was effective against most lineages. All pyramids remained uninfected during trials while many cultivars, including C101LAC but not C101A51, were heavily infected. Our pathogenicity survey results indicate, however, that the durability of these pyramids is better tested in other states (Tamil Nadu, Maharastra and Orissa), where we have detected isolates differentially compatible with both resistance genes. Additional pyramids of C101A51 and C103TTP (allegedly bearing a version of the Pi-1 (t) gene) were constructed. Using isolates from both India and Colombia, we demonstrated that the gene from C103TTP (renamed as Pilb(t)) was expressed as a recessive phenotype. We are currently working on introgressing commercial cultivars used in various regions of India with available lineage-excluding resistances.

By MGR-DNA fingerprinting analyses, we identified 29 lineages of Magnaporthe grisea among rice blast pathogen population in Southern India. The population structure of M. grisea in different states of India was also characterized by a Pot2-based rep-PCR method. The latter method identified 1 to 22 bands in the fingerprints of rice-infecting and less than 10 bands in the fingerprints of non rice-infecting strains of M. grisea which were in the size range of 0.4 to 23 kb. Lineage-exclusion assays showed that a combination of Pi-1 and Pi-2 genes for blast-resistance can exclude all the lineages identified. This virulence characteristic of M grisea remained stable while its lineage-composition showed drastic changes over the period 1993–97 in rice-growing areas of Kerala where annual blast epidemics is common. Therefore, gene pyramids of Pi-1/Pi-1b+Pi-2 were constructed in the CO39 (blast-susceptible) genomic background and such pyramids were resistant to blast in hot-spot locations. The CO39 pyramid now serves as the R-donor for introgressing the Pi-1+Pi-2 genes for blast resistance into cultivars such as Jyothi and IR50 which are high-yielding but blast-prone to make them durably resistant.

As part of efforts to develop alternate strategies for rice disease management, more than 400 strains of bacteria isolated from IRRI rice fields were screened in the laboratory for antagonism towards Pyricularia oryzae and Rhizoctonia solani, the fungal pathogens of blast and sheath blight respectively. On the basis of growth-inhibition in plate assays, 9 bacterial strains which included 3 strains of Pseudomonas fluorescens, 5 strains of Bacillus spp and a strain of Enterobacter were identified. Diameter of inhibition zones ranged from 20–40 mm against P. oryzae and from 20–31 mm against R. solani. These bacterial strains also suppressed sheath blight (Sh-B) in detached rice leaves. They were further evaluated in field tests for suppression of blast and Sh-B when applied as seed treatments or as seed treatments plus sprays. In three field experiments conducted in the IRRI farm, bacterial treatments afforded significant Sh-B control and performed better than validamycin, the fungicide routinely used for Sh-B control. Suppression of Sh-B was more pronounced in direct-seeded rice than in tranplanted rice. Bacterial treatments reduced leaf blast severities by 50–73% in var-IR50 and 34–80% in var C-22 in lowland experiment and 47–57% reduction in var. UPLRi-5 in an upland experiment. Further investigations with a Pseudomonas strain, 7–14 have shown that it produces 2 fluorescent antiblast antibiotics (AB1 and AB2) in culture which at 1.0 ppm inhibit conidial germination of P. oryzae and afford 90–92% reduction in leaf blast development in treated rice (var. IR 50) seedlings. Therefore, biological control of blast by Pseudomonas is antibiotic-mediated.