Evaluation of Blast Thresholds for Seed-Borne Rice Blast in Arkansas

Transcription

1 PEST MANAGEMENT: DISEASES Evaluation of Blast Thresholds for Seed-Borne Rice Blast in Arkansas D.O. TeBeest, P. Schroud, and M. Ditmore ABSTRACT Rice blast, caused by the fungus Magnaporthe grisea, is a serious disease of rice in Arkansas, having previously reached epidemic proportions in Arkansas on susceptible cultivars. It has been established that artificially-infected seeds placed on the soil surface initiated significant levels of disease in the field. In 2002, seeds naturally infected by blast at levels between 1% and 3% were planted and correlated with the rice blast in the plants. The research reported here was conducted in small plots to determine if planting seeds naturally-infected at levels between 0.2% and 4% could result in significant levels of rice blast. Field experiments consisted of planting seeds of two cultivars that differ in the level of resistance to this disease from seedlots with known levels of blast infection ranging from 0% to 4%. In addition, seeds from these same seedlots were treated with azoxystrobin to determine if seed treatment with a fungicide reduced or eliminated development of the disease by affecting the presence of M. grisea on seeds. The results of field tests conducted in 2003 support our previous studies that indicate that planting naturally infected seed leads to significant levels of rice blast disease in the cultivars tested. In 2003, rice blast first appeared in plots of cultivar M201 in early July and reached 20% by 3 September while in plots of a more resistant cultivar Wells, blast appeared in early July, also, but remained at levels below 1% through 3 September. Results of tests conducted in 2003 with naturally-infected seeds treated with azoxystrobin suggest that fungicides reduce the presence of blast on seeds and may delay or reduce the impact of rice blast on yield of some of the more resistant cultivars. 231

2 AAES Research Series 517 INTRODUCTION Rice blast is an economically-important disease of rice world-wide and it is also an important disease in Arkansas on many high-yielding, but blast-susceptible cultivars (Lamey, 1970; Long et al., 2001). Previous work (Long et al, 2001) conducted in commercial fields and experimental plots illustrated several important points about rice blast in Arkansas, namely: 1) that rice blast developed in a fairly consistent pattern across years and cultivars, 2) that blast can be found randomly distributed throughout many commercial fields early in the growing season, 3) that seed artificially infested by M. grisea can initiate rice blast epidemics if placed on the soil surface, and 4) that the spread of rice blast from heavily infected areas within a field is often limited in fields with the most seriously blasted areas (neck blast) of the field at harvest coinciding with the areas most severely blasted before internode elongation. Recently, it was reported that infected seed was the likely primary source of inoculum for rice blast epidemics in California, although it was concluded that infected seeds were not the most likely source of inoculum for subsequent epidemics (Greer and Webster, 2001). Previous work (Lamey, 1970; Manandhar et al., 1998) also suggested that infection of seedlings after planting infested seed beneath the soil surface was always low, and covering with soil inhibited the infection of seedlings grown from infested seed. However, Shetty et al. (1985) showed that planting seeds of finger millet infested by M. grisea clearly resulted in epidemics in small plots in the field. Further, they showed that the amount of disease was correlated to the amount of blast found on seeds. Similar field experiments have not been reported for rice blast. Mayee (1974) also showed that planting a small number of apparently healthy seeds in highly controlled environments and conditions resulted in infection of some seedlings 21 days after planting. This suggested that the role of infected seeds in development of the disease cannot be completely ruled out. Recently, several fungicides have been evaluated for effectiveness as blast control measures when applied to seed and plants in the field (Hayakaka et al., 2002; Serghat et al., 2002). These and similar studies have evaluated the effect of fungicides on seed germination, plant vigor, and the incidence of blast over time in the field, but did not test the effectiveness of fungicides on the incidence of blast on treated seeds following treatment. The objectives of this research were: 1) to determine if planting naturally-infected seeds at different levels of infection resulted in the subsequent development of rice blast in plots; 2) whether this level of infection was correlated to the level of infection observed in the field; and 3) to determine if planting seeds infected by M. grisea treated with azoxystrobin reduced or eliminated infection of rice seed by the fungus. PROCEDURES Twenty plots of the rice cultivar M201 and 12 plots of the rice cultivar Wells were planted on 24 May 2003 with seeds harvested from field plots in 2002 and infected by rice blast. The levels of seed infestation for the seeds harvested in

3 B.R. Wells Rice Research Studies 2003 that were used in 2003 were estimated by visual microscopic examination of seeds incubated on moistened blotter paper as described by Agarwahl et al. (1989). The levels of infection for seed lots were determined as an average of the number of seeds exhibiting sporulation by M. grisea in two replications of 180 seeds. Treatments for cultivar M201 consisted of planting seeds from seedlots estimated to be 0%, 0.2%, 0.8%, 1.0%, and 4% infected by blast and for Wells treatments consisted of planting seeds from seedlots estimated to be infected at 0.0%. 0.25%, and 0.5% seed infection. The levels of seed infection described above were obtained by mixing non-infected seeds with infected seeds in the appropriate amounts to obtain the desired levels of seed infection. Previously, we had planted seeds ranging from 0% to 3% infection (TeBeest and Ditmore, 2003). All treatments were planted in a randomized complete block design with four replications. All plots were planted with 100 g of seed in plots nine rows by 20 ft. All plots of Wells and M201 were separated in all directions with plots containing the blast-resistant cultivar, Ahrent. All plots were visually inspected for seeds remaining on the surface three and seven days after planting and again after emergence. Seeds could not be effectively removed by hand from the soil surface of the plots due to continuous bird predation of sprouted seedlings. Infested seeds of each cultivar at each level of infection were treated with a fungicide to determine if the level of infection of seed could be reduced. Fungicide treatments were done using the plastic bag method. Azoxystrobin was used based on previous reports of its effectiveness (TeBeest and Ditmore, 2002). In this treatment, 0.8 ml of azoxystrobin 100FS was added to 7.2 ml of sterile water and the total mixture added to 400 g of seed for each cultivar and infection level as described above. After 2 minutes of mixing, the treated seeds were air dried in the laboratory as described above. Untreated controls for each infection level for each cultivar consisted of adding 8.0 ml of sterile distilled water to 400 g of seed in a plastic bag. The seeds were thoroughly mixed by hand for two minutes then placed on paper towels on trays to dry. After drying for 48 to 72 hours, 100 g samples of seed were placed in small paper bags, one for each plot. Azoxystrobin 100FS was provided as a gift of Syngenta, Inc. Rice blast was monitored visually within all plots beginning 14 days after emergence and ending at harvest. On 18 September, 25 to 35 head samples (including the flag leaf, collar, neck, and head) were collected from each plot, combined and placed in paper bags, one bag per plot, and taken to the laboratory for additional study. In the laboratory, the incidence of flag leaf, collar rot, neck blast, and panicle blast were determined by visual inspection of each head according to plot. Data on the incidence of rice blast were converted to the percentage of the total number of flag leaves, collars, neck, and heads in the sample from each plot that were infected by the blast fungus. Data presented here are the averages for four plots for each treatment and cultivar. All plots were harvested on 3 October

4 AAES Research Series 517 RESULTS AND DISCUSSION In 2003, the first lesions of rice blast on leaves were found on all three cultivars on 30 June, approximately 5 weeks after planting. The incidence of disease was very small since only one or two lesions were found in some plots at that time. The disease became much more evident within the plots by 17 July. Data summarizing the incidence of rice blast in the field plots and in head samples collected in 2003 are shown in Tables 1 and 2. Rice blast was found also in plots of M201 grown from untreated seeds on 17 July, levels estimated to be approximately 2% (Table 1). However, blast increased rapidly on M201 and by 3 September the incidence of rice blast on leaves ranged between 10% and 20% and appeared to be increasing very rapidly compared to previous estimates. The incidence of rice blast on the head samples collected from these plots shows that rice blast had infected between 7.2% and 31.2%, 28.4% and 71.9%, 29.4% and 57.1%, and 23.7% and 58.2% of the flag leaves, collars, necks, and panicles, respectively. In general, there appeared to be higher levels of disease in plots grown from seed samples that were the most heavily infected by blast. We suspect also that differences between treatments were partially obscured by interplot interference caused by airborne spores. Rice blast was also found on the susceptible cultivar M201 grown from infected seeds treated with azoxystrobin (Table 1). However, the incidence of infection on M201 in plots grown from treated seeds was lower than in plots grown from untreated seeds. In contrast to results from growing untreated seeds, the incidence of rice blast on leaves in plots grown from treated seeds ranged from 1.0% to 5.5% on 3 September although blast appeared to be increasing. In the head samples collected from these plots on 18 September, the incidence of rice blast ranged from 3.0% to 18%, from 24.4% to 63.3%, from 12.6% to 60%, and from 24.5% to 50.6% on flag leaves, collars, necks, and panicles, respectively. The overall levels of rice blast on plants grown from treated seeds were higher on M201 than on Wells. Rice blast remained at very low levels in plots of Wells grown from infected seeds treated with azoxystrobin throughout the growing season (Table 2). In all of these treatments, the average level of infection of 200 leaves remained near zero from 17 July through 3 September. However, rice blast was found in head samples collected on 18 September and the incidence of disease ranged from 5.1% to 10.5%, from 10.1% to 10.7%, from 3.0% to 6.5%, and from 5.4% to 15.1% for flag leaves, collars, necks, and panicles, respectively. These levels of infection were significantly lower than levels found in the untreated treatments at all sample times. Rice blast remained at very low levels in plots grown from untreated seeds of Wells throughout the season ending on 3 September, although it appeared to be increasing in some plots (Table 2). This observation is supported by the analysis of head samples used to determine the incidence of rice blast on flag leaves, collars, necks, and panicles that were collected on 18 September from the untreated seed. In these samples, the incidence of blast ranged from 11.6% to 19.3%, from 10.1% to 16.7%, 234

5 B.R. Wells Rice Research Studies 2003 from 6.8% to 8.8%, and from 10% to 19.4% for flag leaves, collars, necks, and panicles, respectively, across all treatments. As reported in 2002, the field data shows that significant levels of rice blast were found in replicated plots of two cultivars in which infected seed were planted. As in 2002, it is unlikely that blast originated from an external inoculum since nearby rice plots planted with non-infected seeds of the same cultivars did not have significant levels of the disease. Rice blast appeared in control plots planted with seed lots in which blast was not detected in our seed assays. However, the lowest limit of detection was only 1 seed in 360 seeds (0.28%). This level of seed infection may have been too high to assure that blast was not present in the sample and it may be too high to significantly reduce the incidence of disease. Shetty et al. (1985) estimated that one infected seed of pearl millet in ten thousand (0.01%) was sufficient to cause an epidemic of blast on pearl millet. Increased precision in estimating the amount of blast in infested seedlots is required before thresholds can be established for rice blast in Arkansas although the threshold may be higher for more resistant cultivars such as Wells than for cultivars such as susceptible as M201. Lamey (1970) suggested that planting infected rice seed beneath the soil surface inhibited infection of rice by the rice blast fungus. Our data from 2002 and 2003 suggests that significant levels of disease are possible as a result of planting infected seed using drill planters. Our data suggest that it is very difficult to assure that mechanical planting prevents seeds from coming to the surface of soils. The physical process of loading drill planters and bird predation may cause seeds to be left on the surface of the soil. The incidences of disease in our field plots in 2003 were much lower than those found in field plots of the same two cultivars planted in 2002 (TeBeest and Ditmore, 2003). In 2002, the estimated levels of infection of seeds that were used to plant the tests were higher, ranging from 1% to 3% for both M201 and Wells. In 2002, the incidence of head samples of Wells taken just before harvest were much higher in comparison to the samples collected in 2003 from plots planted with seeds with levels of infection ranging from 0% to 0.5%. These differences may be attributed to differences in the level of seed infection and to environmental differences during the growing seasons of 2002 and Smaller differences were observed for the more susceptible cultivar M201 even with much lower levels of seed infection. Additional work that includes a broader range and lower levels of seed infection will be required to better describe the impact of seed infection on the occurrence and severity of rice blast. SIGNIFICANCE OF FINDINGS The work reported here was done to corroborate our previous work that showed that rice blast could develop as a result of planting naturally infected seed. We also conducted the work to determine if fungicides could effectively reduce the inci- 235

6 AAES Research Series 517 dence of disease on seeds. The data from this work with naturally-infected rice seeds appears to support our previous work that showed that infected rice seed is a viable source of the inoculum for rice blast disease in Arkansas. Further, the fact that an effective fungicide treatment of the seeds reduced the overall level of disease supports this conclusion. The research also suggests that thresholds may be found and established for rice cultivars with some level of resistance to aid in controlling this important disease. The data also suggest that an effective fungicide, one that when applied at an effective rate can reduce the level of seed infestation levels by the blast fungus, might significantly delay the appearance and significance of the disease. ACKNOWLEDGMENTS The authors gratefully acknowledges the cooperation of the Arkansas rice producers, the support of the Arkansas Rice Research and Promotion Board, and the support and donation of fungicide materials by Syngenta, Inc. The authors are also grateful for the cooperation of Dr. R. Cartwright for his assistance with this research and preparation of this report. The author also appreciates the cooperation of Mr. Roger Eason and Mr. S. Clark for their invaluable assistance with the work at the Pine Tree Experiment Station, Colt, Ark. LITERATURE CITED Agarwahl, P.C., C.N. Mortensen, and S.B. Mathur Seed-borne diseases and seed health testing of rice. Tech. Bull. No. 3. CAB International Mycological Institute. Ferry Lane, Kew, Surrey TW9 3AF, United Kingdom. Pp Greer, C.A. and R.K. Webster Occurrence, distribution, epidemiology cultivar reaction, and management of rice blast in California. Plant Disease 85: Hayakaka, T., T. Matsuura, and T. Namai Behavior and control of Pyricularia oryzae in brown rice seed as inoculum source of rice seedling blast. Japan J. Phytopathol. 68: Lamey, H.A Pyricularia oryzae on rice seed in the United States. Pl. Dis. Rep. 54: Long, D.H., J.C. Correll, F.N. Lee, and D.O. TeBeest Rice blast epidemics initiated by infested rice grain on the soil surface. Plant Disease 85: Manandhar, H.K., H.J.L. Jorgenson, V. Smedegaard-Petersen, and S.B. Mathur Seedborne infection of rice by Pyricularia oryzae and its transmission to seedlings. Plant Disease 82: Mayee, C.D Perpetuation of Pyricularia oryzae Cav. through infected seeds in Vidarbha region. Indian Phytopathol. 27:

7 B.R. Wells Rice Research Studies 2003 Serghat, S., A. Mouria, A.O. Touhami, and A. Douira In vivo effect of some fungicides on the development of Pyricularia grisea and Helminthosporium oryzae. Phytopathol. Mediterr. 41: Shetty, H.S., A. Gopinath, and K. Rajashekar Relationships of seed-borne inoculum of Pyricularia grisea to the incidence of blast of finger millet in the field. Indian Phytopathol. 38: TeBeest, D.O. and M. Ditmore Preliminary examination of blast thresholds for seed-borne rice blast in Arkansas. In: R.J. Norman and T.H. Johnston (eds.). B.R. Wells Rice Research Studies University of Arkansas Agricultural Experiment Station Research Series 504: Fayetteville, Ark. 237

8 AAES Research Series 517 Table 1. The incidence of rice blast on leaves during the growing season and incidence of blast on flag leaves, collars, necks, and panicls collected near harvest on cultivar M201 resulting from planting infected or treated seeds at different levels of seed infection at the Pine Tree Branch Station, Colt, Ark., in Seed Vegetative y Hard dough Incidence near harvest y Treatment infected z 17 July 20 Aug 3 Sept Flag Collar Neck Panicle (%) (%) Untreated b x 6.8 b 17.5 cd 14.7 ab 34.6 a 29.4 ab 23.7 a b 10.0 b 17.5 cd 30.8 b 42.4 a 32.8 b 38.1 ab b 8.8 b 10.0 bc 11.3 a 38.6 a 34.5 bc 40.1 ab b 8.8 b 20.0 d 31.2 b 28.4 a 57.1 d 58.2 b a 8.8 b 15.0 cd 7.2 a 71.9 b 52.4 cd 56.9 b Treated a 0.5 a 5.5 ab 16.4 ab 24.4 a 12.6 a 24.5 a a 1.0 a 1.3 a 3.0 a 25.0 a 21.2 ab 25.9 a a 0.8 a 1.0 a 7.5 a 37.2 a 39.9 ab 47.8 b a 1.8 a 1.0 a 18.0 ab 63.3 b 60.0 d 50.6 b a 1.5 a 3.3 ab 5.3 ab 40.0 a 28.5 ab 40.2 ab z The percentage of seeds infected was determined as an average of two replications of 180 seeds taken from infected seedlots. Seeds were incubated on moistened filter paper at 22ºC under 12 hr lighting for four days. The controls (0%) consisted of seeds from a seedlot in which blast was not detected. Individual levels of infection were prepared by diluting the infected seedlot with the control seedlot in the appropriate amount. To treat seeds with azoxystrobin, 0.8 ml of azoxystrobin 100FS was added to 7.2 ml of water and applied to 400 g of seeds in plastic bags. y The incidence of rice blast on 17 July and 20 August 2002 was determined by estimating the number of leaves from 100 to 200 leaves per plot exhibiting typical blast lesions. The incidence of rice blast near harvest was determined by collecting 25 to 35 heads (including the flag leaf) and determining the number of flag leaves, collars, necks, and panicles from within each sample that were infected by blast. All data are the averages of four replications of each treatment. x Means within a column followed by the same letter are not significantly different using the Least Significant Difference Procedure (P=0.05). 238

9 B.R. Wells Rice Research Studies 2003 Table 2. The incidence of rice blast on leaves during the growing season and incidence of blast on flag leaves, collars, necks, and panicles collected near harvest on cultivar Wells resulting from planting infected or treated seeds at different levels of seed infection at the Pine Tree Branch Station, Colt, Ark., in Seed Vegetative y Hard dough Incidence near harvest y Treatment infected z 17 July 20 Aug 3 Sept Flag Collar Neck Panicle (%) (%) Untreated a 0.0 a 0.0 a 11.6 ab 16.7 a 8.8 a 10.9 ab b 0.25 ab 0.2 ab 13.5 ab 12.6 a 6.8 a 19.4 b a 0.5 b 0.5 b 19.3 b 10.1 a 8.1 a 10.0 ab Treated a 0.0 a 0.0 a 10.5 ab 10.4 a 6.5 a 5.4 a a 0.0 a 0.0 a 5.1 a 10.7 a 3.0 a 9.9 ab a 0.0 a 0.0 a 10.1 ab 10.1 a 3.0 a 15.1 ab z The percentage of seeds infected was determined as an average of two replications of 180 seeds taken from infected seedlots. Seeds were incubated on moistened filter paper at 22ºC under 12 hr lighting for four days. The controls (0%) consisted of seeds from a seedlot in which blast was not detected. Individual levels of infection were prepared by diluting the infected seedlot with the control seedlot in the appropriate amount. To treat seeds with azoxystrobin, 0.8 ml of azoxystrobin 100FS was added to 7.2 ml of water and applied to 400 g of seeds in plastic bags. y The incidence of rice blast on 17 July and 20 August 2002 was determined by estimating the number of leaves from 100 to 200 leaves per plot exhibiting typical blast lesions. The incidence of rice blast near harvest was determined by collecting 25 to 35 heads (including the flag leaf) and determining the number of flag leaves, collars, necks, and panicles from within each sample that were infected by blast. All data are the averages of four replications of each treatment. x Means within a column followed by the same letter are not significantly different using the Least Significant Difference Procedure (P=0.05). 239