Bacterial panicle blight (BPB) is among the three most limiting rice diseases in Louisiana and the southern United States. The identity and characterization of pathogens associated with this disease was unclear. This research details studies carried out on the pathogens causing BPB on rice in Louisiana and other rice producing southern states. Bacterial strains were isolated from BPB-infected sheath, panicle, or grain samples collected from rice fields in Louisiana, Arkansas, Texas, and Mississippi. In greenhouse inoculation tests, 292 of 364 strains were pathogenic on rice seedlings or panicles. Identification of strains in the pathogen complex by growth on S-PG medium, carbon source utilization profile (Biolog), cellular fatty acid analysis, and polymerase chain reaction (PCR) methods revealed that 76 and 5% of the strains were Burkholderia glumae and B. gladioli, respectively. The other strains have not been conclusively identified. Although strains of both species produced similar symptoms on rice, B. glumae strains were generally more aggressive and caused more severe symptoms on rice than B. gladioli. Virulent strains of both species produced toxoflavin in culture. The two species had similar growth responses to temperature, and optima ranged from 38 to 40°C for B. glumae and 35 to 37°C for B. gladioli. PCR was the most sensitive and accurate method tested for identifying the bacterial pathogens to the species level. The 16S rDNA gene and 16S-23S rDNA internal transcribed spacer (ITS) region sequences of the B. glumae and B. gladioli strains from rice showed more than 99% sequence homology with published sequences. A real-time PCR system was developed to detect and quantify this pathogen from infected seed lots. Our results clearly indicate that B. glumae and B. gladioli were the major pathogens causing BPB in the southern United States.

Willyerd, K. T., Li, C., Madden, L. V., Bradley, C. A., Bergstrom, G. C., Sweets, L. E., McMullen, M., Ransom, J. K., Grybauskas, A., Osborne, L., Wegulo, S. N., Hershman, D. E., Wise, K., Bockus, W. W., Groth, D., Dill-Macky, R., Milus, E., Esker, P. D., Waxman, K. D., Adee, E. A., Ebelhar, S. E., Young, B. G., and Paul, P. A. 2012 [MR_UT]) were used in multivariate meta-analyses, and mean log response ratios across trials were estimated and transformed to estimate mean percent control ( C ) due to the management combinations relative to S_UT. All combinations led to a significant reduction in index and DON (P < 0.001). MR_TR was the most effective combination, with a C of 76% for index and 71% for DON, followed by MS_TR (71 and 58%, respectively), MR_UT (54 and 51%, respectively), S_TR (53 and 39%, respectively), and MS_UT (43 and 30%, respectively). Calculations based on the principle of treatment independence showed that the combination of fungicide application and resistance was additive in terms of percent control for index and DON. Management combinations were ranked based on percent control relative to S_UT within each trial, and nonparametric analyses were performed to determine management combination stability across environments (trials) using the Kendall coefficient of concordance (W). There was a significant concordance of management combinations for both index and DON (P < 0.001), indicating a nonrandom ranking across environments and relatively low variability in the within-environment ranking of management combinations. MR_TR had the highest mean rank (best control relative to S_UT) and was one of the most stable management combinations across environments, with low rank stability variance (0.99 for index and 0.67 for DON). MS_UT had the lowest mean rank (poorest control) but was also one of the most stable management combinations. Based on Piepho's nonparametric rank-based variance homogeneity U test, there was an interaction of management combination and environment for index (P = 0.011) but not for DON (P = 0.147), indicating that the rank ordering for index depended somewhat on environment. In conclusion, although the magnitude of percent control will likely vary among environments, integrating a single tebuconazole + prothioconazole application at anthesis with cultivar resistance will be a more effective and stable management practice for both index and DON than either approach used alone.

No abstract

Bacterial panicle blight (BPB) of rice (Oryza sativa L.) occurs when the bacterium Burkholderia glumae Kurita and Tabei infects emerging and flowering panicles, causing kernels to abort. To identify quantitative trait loci (QTLs) for BPB resistance, a population of 300 recombinant inbred lines (RILs) derived from a cross between ‘Lemont’ and ‘TeQing’ were evaluated in 2001 and 2002 in field plots spray‐inoculated with B. glumae at the time of flowering. Because this RIL population had been previously used to map QTLs for three other diseases, present use of this population allowed direct comparison between the various disease resistance QTLs. Multiple interval mapping using QTL Cartographer v2.5 putatively identified 12 BPB QTLs, three of which were statistically significant in both years and found to have epistatic effects in 2002. TeQing was the source of resistance for eight QTLs; Lemont for four. Four BPB QTLs colocated with QTLs previously identified as providing resistance to one or multiple other diseases. Three BPB QTLs were also associated with late flowering. Because late flowering panicles are subjected to cooler temperatures that are less conducive to disease development during grain fill, it is possible that the genetic effects of the heading‐related QTLs were biased. The present data could not distinguish between pleiotropy and close linkage of the BPB QTLs with the previously identified heading and disease resistance QTLs.

No abstract