Effects of Gene and Cytoplasm Substitutions in Pearl Millet on Leaf Blight Epidemics and Infection by Pyricularia grisea Jeffrey

Wilson, J. P., and Hanna, W. W. 1992. Effects of gene and cytoplasm substitutions in pearl millet on leaf blight epidemics and infection by Pyricularia grisea. Phytopathology 82:839-842. Derivatives of the pearl millet inbred Tift 23 with substitutions for allele. Disease ratings made early in the season in the 1991 field experiment various cytoplasms and alleles conferring morphologic or developmental did not correlate well with disease ratings made later in the season. Leaf traits were evaluated for differences in leaf blight epidemics in the field blight increased on early cultivars after anthesis. Therefore, leaf blight and their reactions to infection by Pyricularia grisea in the greenhouse. in the field must be assessed at similar growth stages. When inoculated None of the experiments indicated an effect of the B, A,, or A4 cytoplasms; with P. grisea, seedlings of some inbreds with the el allele had smaller the tr allele for the trichomeless character; or the d2 allele for dwarf lesion dimensions than inbreds without the allele, and no lesions developed stature on leaf blight progress or on infection by P. grisea. An apparent on Tift 23DA1 E. Differences between reactions of the inbreds in the field increased susceptibility in the field was associated with the el allele for and the greenhouse could be due in part to differences in susceptibility earliness. When disease progress curves were corrected for anthesis date, to other pathogens with an undetermined contribution to the leaf blight inbreds with the el allele were more resistant than inbreds without the complex. Additional keywords: alloplasmic, near-isogenic, Pennisetum glaucum, plant maturity. Pearl millet (Pennisetum glaucum (L.) R. Br.) is used as an in one of the three cytoplasms. Therefore, an evaluation of the annual summer forage crop in the southeastern United States effects of these substitutions on disease resistance or susceptibility and could potentially become an important grain crop as well. was considered useful in interpreting disease ratings of germ plasm Because of its present use, maintaining healthy foliage is important in the field and greenhouse. Several near-isogenic and alloplasmic to provide high-quality forage. Considering that rust (caused by derivatives of the inbred Tift 23B are available for experimental Puccinia substriata Ell. & Barth. var. indica Ramachar & Cumm.) purposes. A mutation-breeding program was conducted at Tifton adversely affects the yield and digestibility of pearl millet forage during the 1960s and 1970s. From the original tall, late-maturing, (15), it is likely that other foliar diseases also reduce forage yield trichomed inbred Tift 23B, plants with the el and tr alleles were or quality. selected. Tift 23B has been the recurrent parent in seven or more Several pathogens cause leaf blight of pearl millet (8). Pyribackcrosses to selections with the early, trichomeless, and dwarf cularia leaf spot, caused by Pyricularia grisea (Cke.) Sacc., was genes, and to the sources of the A, and A4 cytoplasms. first identified in the United States in 1968 (13). This pathogen The objectives of this study were to evaluate the effects of has become an important component of the complex of organisms gene and cytoplasm substitutions on leaf blight development in that cause leaf blight of pearl millet in Georgia. Infections are the field and lesion development on seedlings following inocuusually visible throughout the season, and it has been the only lation with P. grisea. leaf-blighting pathogen targeted in efforts for breeding for resistance (7). One approach used in pearl millet breeding at the USDA-ARS MATERIALS AND METHODS Forage and Turf Research Unit at Tifton, GA, includes back1988 field experiment. On 22 June 1988, 18 derived lines of crossing various genes or cytoplasms into inbreds with good cominbred Tit2 rith ae or cytop1asmic1substitutins wr bining ability. Alleles used for improvement of forage or grain inbred Tift 23 with allelic or cytoplasmic substitutions were millets include the d2 allele for dwarf stature (2), the el allele planted at the rate of 0.31 kg/ha in the field at the Coastal Plain for earliness (6), and the tr allele conferring a trichomeless charExperiment Station in a randomized complete block design with acter (3). Some of these alleles may affect the response of plants six replications. The inbreds (normally designated with the prefix to diseases. The tr allele has been associated with increased rust "Tift," but omitted here for brevity) were 23B, 23A 1, 23A 4, 23DB, susceptibility in the field (3), yet also confers a degree of resistance 23DA1, 23DA4, 23BE, 23AIE, 23A 4E, 23BS, 23AlS, 23A 4S, to smut, caused by Moesziomycespenicillariae (Bref.) Vanky (14). 23DBE, 23DA1 E, 23DA 4E, 23DBS, 23DAIS, and 23DA 4S. The Cytoplasms used include the fertile B cytoplasm (1) and the malesymbols D, E, and S indicate homozygous substitutions for the sterile A, (1) and A 4 (5) cytoplasms. d2, el, or tr allele, respectively. Symbols B, A,, and A4 indicate Many of the elite pearl millet inbreds developed at Tifton possess substitutions of genotypes into these cytoplasms. Plots were 4 one or more of the above-mentioned alleles, and nearly all are m long and spaced 1 m apart. Fertilizer (5-10-15 N-P-K) was applied in the row at planting at the rate of 280 kg/ha. Visual ratings for percent foliage in the plots with leaf blight (chlorosis and necrosis) were taken 9 August, 18 August, 1 September, and 13 September 1988. Ratings of leaf blight rather than This article is in the public domain and not copyrightable. It may be freely pyricularia severities were taken since it it difficult to separate reprinted with customary crediting of the source. The American Phytopaththe effects of different leaf-blighting pathogens, particularly after ological Society, 1992. leaves turn necrotic. Area under the disease progress curves Vol. 82, No. 8, 1992 839 (AUDPCs) were calculated by the agar has dried. Conidia and mycelia are scraped off the dried agar, and approximately equal quantities of each isolate are bulked AUDPC = Y[ Y(i+I) YI]/2 X [X(i+l) Xi] and stored in plastic self-sealing bags at -72 C. Isolates are bulked because no differences in pathogenicity to pearl millet among where Y1 = percent severity at time X,. AUDPCs were analyzed P. grisea isolates have been detected to date, and the mixture by least squares analysis of variance, and sums of squares were ensures against loss of pathogenicity of the inoculum from subpartitioned into replication and cultivar effects (9). Differences culturing. in AUDPCs due to cytoplasm or gene substitution were examined Before being used, packets of conidia were submerged in a by nonorthogonal, single-degree-of-freedom linear contrasts (11). 40 C water bath for 5 min. Conidia were suspended in deionized 1991 field experiment. The inbreds 23B, 23A 1, 23A 4, 23DB, water with one drop of Triton B 1956 per 100 ml. Final inoculum 23DA1, 23DA 4, 23BE, 23ALE, 23A 4E, 23DBE, 23DALE, and concentration was 4 X 104 conidia per milliliter. 23DA 4E were planted 16 May 1991 in a randomized complete Plants were inoculated when three leaves were fully expanded. block with 10 replications in the field at the Coastal Plain ExperiPots were placed in a randomized block design in an inoculation ment Station. Plot dimensions, establishment, and maintenance chamber where they were misted to dripping with inoculum. Plants in 1991 were identical to those evaluated in 1988. were kept in the chamber overnight and automatically misted Plots were visually rated for percent foliage with leaf blight with deionized water for 1 min every 30 min. After 18 h, plants on 10 June, 19 June, 2 July, 12 July, 23 July, 31 July, and 12 were returned to the greenhouse bench in a randomized block August. Ratings ceased when rust started to increase throughout design. As a result of damping-off within some of the pots, not the test to levels that would confound leaf blight ratings. Date all inbreds were represented in each block. of 50% anthesis was recorded for each plot. After flowering, plant Seven days after inoculation, lengths (L) and widths (W) of heights were measured. lesions on the third leaves of the seedlings were measured from AUDPCs were calculated as described for the 1988 experiment, each pot. In the first experiment, inbreds were planted in each To estimate leaf blight severities at similar growth stages, disease of six pots, and up to 10 lesions were measured per pot. In the severities were transformed to ln{0.01 + [Y/(1 Y)]}, where second and third experiments, inbreds were planted in each of Y = proportion of disease. A factor of 0.01 was included in the 10 pots, and up to six lesions were measured within each pot. logit transformation to include observed ratings of zero. TransAll lesions measured were bounded by healthy leaf tissue. Those formed severities were regressed against time (days after planting) that developed on leaf margins were ignored. Over the three by quadratic regression. Transformed blight severities were preexperiments, an average of 154 lesions were measured on each dicted from the regression equations for the observed heading inbred. Approximate lesion areas were calculated by AREA = date. Predicted values were back-transformed to percent severities. (L X W X 7r)/4. Lesion lengths, widths, and areas were anaDays to anthesis, heights, AUDPCs, observed final severities, lyzed by the general linear model procedure of SAS (9), and and predicted severities at anthesis were analyzed by least squares means were differentiated by Fisher's least significant difference. analysis of variance (9). Sums of squares were partitioned into replication and inbred effects. AUDPCs were compared by singleRESULTS degree-of-freedom linear contrasts. Means for observed final severities and predicted severities at anthesis were differentiated Field experiments. In both 1988 and 1991, inbred was a sigby Fischer's least significant difference. Pearson's correlation conificant source of variation (P< 0.01) for AUDPC, and differences efficients between severities at each evaluation date and AUDPCs could be discerned among inbreds (Table 1). Single-degree-ofwith days to anthesis and plant heights were calculated. freedom contrasts indicated that substitution with the el allele To identify fungi involved in the leaf blight complex, leaf resulted in greater AUDPCs (Table 2). In both years, no other samples were taken from plants in the fourth and tenth replication allele or the cytoplasms had any effect on leaf blight. on 12 July and 2 August 1991, respectively. Five leaf pieces with In 1991, differences existed among inbreds (P < 0.01) not only relatively isolated lesions were selected from different plants for AUDPC, but also for heading date, height, final disease throughout the plots of each inbred. Samples were randomly ratings, and predicted disease severities at anthesis. Mean sepataken, and a variety of lesion types were collected. Leaf samples ration of 1991 final disease severities revealed lower severities (about 1.5 X 2.5 cm ) were surface disinfected for 1 min in a on inbreds without the el allele (Table 3). However, comparison 0.5% NaOC1 solution and plated on 20% V8 juice and 1.5% NaOH of predicted severities at anthesis resulted in an opposite ranking. (V8) agar. Leaf pieces were incubated at 24 C under continuous Inbreds with the el allele had lower severities predicted at anthesis. fluorescent lighting. Fungi growing from leaf pieces were identified Leaf blight ratings made early in the 1991 season were not microscopically 3-5 days after plating and were subcultured to correlated with those made later in the season. Days to anthesis facilitate identification when necessary. Greenhouse experiments. Three experiments were conducted during 1990 and 1991 to evaluate differences in susceptibility of TABLE 1. Mean values for area under the disease progress curves for seedlings to P. grisea. In the first experiment, 12 pearl millet near-isogenic, alloplasmic pearl millets evaluated in 1988 and 1991 inbreds; 23B, 23A 1, 23A 4, 23DB, 23DA 1, 23DA4, 23BE, 23A 1E, 23A 4E, 23DBE, 23DAIE, and 23DA 4E were planted in 10-cmGene Cytoplasm diameter pots containing equal volumes of coarse building sand, substitution B A, A 4 peat moss, and perlite, amended with 8.8 g/L of agricultural 1988 gypsum and 5.2 g/L of Osmocote fertilizer (3to 4-mo release, 23 309.8 328.0 361.3 N-P-K 14-6.1-11.6, Sierra Chemical Co., Milpitas, CA). Plants 23Dx 401.2 419.9 350.8 were grown under natural lighting, and greenhouse temperatures 23EY 766.5 907.2 897.5 ranged from about 33 C during the day to 23 C during the night. 23SZ 397.6 329.9 305.9 Stands were thinned to five plants per pot. 23DE 911.4 873.8 856.0 Plants were inoculated with a bulk inoculum used for screening 1991 breeding lines for resistance. The inoculum consists of a mixture 23 843.5 904.0 986.8 of 10 or more isolates of P. grisea. Lesions are periodically selected 23D 637.8 759.8 813.5 from fieldand greenhouse-grown plants, and P. grisea is isolated. 23E 1,727.0 1,702.5 1,770.4 After determining that isolates are free of contaminants, four 23DE 1,338.0 1,124.7 1,543.4 5-mm-diameter plugs from colonies are placed on V8 agar in x D indicates homozygous substitution with the d2 allele for dwarfness. petri plates and fungi are allowed to colonize the surface under I E indicates homozygous substitution with the eI allele for earliness. conditions as described above. After about 7 days, plates are S indicates homozygous substitution with the tr allele for trichomeplaced in a draft-free room and lids are removed for 4 days until lessness.