Reconsidering the Lessons Learned from the 1970 Southern Corn Leaf Blight Epidemic.

The southern corn leaf blight epidemic of 1970 caused estimated losses of ~16% for the US corn crop, equivalent to ~$8 billion in current terms. The epidemic was caused by the prevalence of Texas male sterile cytoplasm (cms-T), used to produce most of the hybrid corn seed planted that year, combined with the emergence of a novel race of the fungus Cochliobolus heterostrophus that was exquisitely virulent on cms-T corn. Remarkably, the epidemic lasted just a single year. This episode has often been portrayed in the literature and textbooks over the last 50 years as a catastrophic mistake perpetrated by corn breeders and seed companies of the time who did not understand or account for the dangers of crop genetic uniformity. In this perspective article, we aim to present an alternative interpretation of these events. First, we contend that, rather than being caused by a grievous error on the part of the corn breeding and seed industry, this epidemic was a particularly unfortunate, unusual and unlucky consequence of a technological advancement intended to improve the efficiency of corn seed production for America's farmers. Second, we tell the story of the resolution of the epidemic as an example of timely, meticulous applied research in the public sector for the public good.

Reduction in Common Rust Severity Conferred by the Rp1D Gene in Sweet Corn Hybrids Infected by Mixtures of Rp1D-Virulent and Avirulent Puccinia sorghi.

The Rp1D gene confers a hypersensitive, chlorotic-fleck, resistant reaction to Puccinia sorghi, the casual agent of common rust of corn. About 40% of commercial sweet corn hybrids carry the Rp1D gene. Sine 1999, Rp1D-virulent (D-virulent) isolates of P. sorghi have occurred regularly in populations of P. sorghi in North America. Observations from sweet corn hybrid nurseries and other trials indicate that the frequency of D-virulent isolates affects severity of rust on Rp1D hybrids; however, the frequency of D-virulence at which the Rp1D gene is rendered completely ineffective is not known. The objective of this study was to assess whether common rust severity is reduced by the Rp1D gene in sweet corn hybrids infected by mixtures of D-virulent and Rp1D-avirulent (avirulent) P. sorghi. Forty pairs of Rp1D-resistant and susceptible (rp1d) versions of sweet corn hybrids from six different commercial breeding programs were evaluated in 2003 and 2004 in trials inoculated with one of five different ratios of avirulent:D-virulent inocula: 100:0, 90:10, 80:20, 60:40, or 0:100. When D-virulent P. sorghi was 100% of initial inoculum, common rust was equally severe on Rp1D and rp1d versions of the same hybrid. Thus, the Rp1D gene did not confer partial or residual resistance in these trials. When initial inocula consisted of 40% or less D-virulent P. sorghi, rust was significantly less severe on Rp1D versions than on rp1d versions of the same hybrids. Relationships between rust severity on Rp1D and rp1d versions of hybrids were explained by linear regressions in all trials. Slope coefficients (i.e., rust severity on Rp1D hybrids as a proportion of that on rp1d hybrids) were related to the percentage of D-virulent P. sorghi in the initial inoculum and were 0.21, 0.29, 0.51, 0.64, and 0.93 in 2003 and 0.25, 0.50, 0.67, 0.76, and 1.0 in 2004 for trials inoculated with 0, 10, 20, 40, and 100% D-virulent P. sorghi, respectively. Thus, the Rp1D gene may convey levels of control in proportion to the frequency of virulence in mixed populations of D-virulent and avirulent P. sorghi when the frequency of virulent isolates is less than 40%.

Resistance to Puccinia polysora in Maize Accessions.

A number of potential sources of general and specific resistance to southern corn rust were identified from 1,890 plant introduction accessions that were screened for reactions to Puccinia polysora race 9. Resistance appeared to differ among four accessions on which uredinia were not observed in initial screenings. Resistance to P. polysora in PI 186215 (Argentine inbred 2-687) was a chlorotic fleck, hypersensitive reaction that was conditioned by a single, dominant gene that was allelic with or very closely linked to the Rpp9 gene based on tests of allelism. All but 3 of 2,357 testcross progeny, (inbred 2-687 × Rpp9) × PS were resistant. Resistance in Ames 19016 (Va59) was effective in F1 progeny and appeared to be dominant and simply inherited; however, this resistance appeared to be a slow-rusting or incomplete resistance that was effective in adult plants but not in young seedlings. Severity of southern rust was less than 10% on resistant progeny from crosses with Va59 compared with severity exceeding 70% on susceptible progeny. Resistance in plant introduction (PI) 186209 (Venezuelan flint) and NSL 75976 (IA DS61) were not effective in F1 hybrid combination and, thus, probably have limited value in commercial maize. Resistance in PI 186209 may be conditioned by a single recessive gene and resistance in NSL 75976 may be co-dominant.

Production of huitlacoche, Ustilago maydis: timing inoculation and controlling pollination.

Huitlacoche is the name given to young, fleshy, edible galls that form when ears of Zea mays are infected by Ustilago maydis. Huitlacoche is processed and sold fresh at markets in Mexico. Interest has increased recently in producing U. maydis as a specialty mushroom in the United States. Silk-channel inoculation methods developed to evaluate common smut resistance in maize can be used to produce huitlacoche commercially. This research assessed the effects of time of inoculation and preventing pollination on the severity of ear galls and yield of huitlacoche produced by inoculating silks with U. maydis. Yield of huitlacoche and severity of ear galls were closely related, as was evident from highly significant linear or curvilinear regressions. Severity and yield were greatest when ears were inoculated 4-8 d after the mid-silk growth stage. Ear galls were 5-15% more severe and yield of huitlacoche was 18-150% greater on ears that were not pollinated, compared to those that were pollinated. Maximum yield of huitlacoche was 131 g ear(-1) from unpollinated male-sterile field corn inoculated 6 d after the mid-silk growth stage and 92 g ear(-1) from detasseled sweet corn inoculated 6 d after mid-silk. About 25% of the total weight of ears consisted of marketable huitlacoche when yields were highest. Quality of huitlacoche was not affected by time of inoculation or pollination treatments, but quality of huitlacoche harvested 12-14 d after inoculation was unacceptable primarily due to lack of teliospores, which affected color and flavor.

Levels of Stewart's Wilt Resistance Necessary to Prevent Reductions in Yield of Sweet Corn Hybrids.

Stewart's wilt reactions and yield of a total of 69 sweet corn hybrids were evaluated in trials in 1999 and 2000 in order to determine the level of Stewart's wilt resistance necessary to prevent reduction in yield of sweet corn hybrids. Plants at the 2- to 3-leaf stage were inoculated with Erwinia stewartii using the pinprick method. Stewart's wilt symptoms were rated from 1 to 9, and incidence of systemic infection was determined as a percentage for each experimental unit. Primary ears were harvested about 21 days after midsilk, and yield was measured as ear weight and number of marketable ears. Percent yield was calculated for each hybrid by dividing yield from inoculated treatments by yield from noninoculated treatments and multiplying by 100. Hybrid means for Stewart's wilt ratings in inoculated treatments ranged from 2.0 to 7.3. The relationships between percent yield and Stewart's wilt severity ratings were described best by curvilinear regressions, whereas percent yield decreased linearly with the incidence of systemic infection in 2000. Ear weights of hybrids with ratings below 3 or 3.5 were estimated from the regression equations to be within 95 or 92%, respectively, of those from noninoculated treatments of the same hybrid. The number of marketable ears from hybrids with ratings of 3 or below was estimated from the regression equations to be within 90% of those from noninoculated treatments of the same hybrid. A level of resistance that resulted in Stewart's wilt ratings below 3 or 3.5 corresponded to nonsystemic infection of most plants (i.e., incidence of systemic infection below 5 or 10%, respectively). The adverse effects of Stewart's wilt on ear weight and marketability appeared to be minor for sweet corn hybrids with levels of resistance that prevented or minimized systemic infection.

Resistance Genes in the rp1 Region of Maize Effective Against Puccinia sorghi Virulent on the Rp1-D Gene in North America.

Resistance in sweet corn conferred by the Rp1-D gene has controlled common rust, caused by Puccinia sorghi, in North American corn for nearly 15 years. Eleven isolates of P. sorghi virulent on corn with the Rp1-D gene were collected from Rp-resistant corn in 1999 from Wiscon-sin, Illinois, New York, and Minnesota. Isolates were increased on susceptible sweet corn. Urediniospores of nine isolates were bulked. Reactions of individual Rp genes in the rp1 region and reactions of linked combinations of Rp genes in the rp1 region (i.e., compound rust resistance genes) were evaluated against the bulked population of P. sorghi in several greenhouse trials. Reactions of individual and compound Rp genes also were evaluated against individual isolates of P. sorghi. Each trial contained at least two replicates of several lines with Rp genes and one susceptible check. Five to 10 two-leaved seedlings per line were inoculated at least twice with a suspension of urediniospores. Ten days after inoculation, rust reactions were rated:+ = sporulating uredinia, - = no sporulating uredinia, and I = chlorotic or necrotic tissue surrounding small uredinia. Four single genes, Rp1-E, Rp-G, Rp1-I, and Rp1-K, and eight compound genes, Rp1-JFC, Rp1-JC, Rp-GI, Rp-G5, Rp-GDJ, Rp-G5JD, Rp-G5JC, and Rp-GFJ, conferred resistance. Additional characterization of virulence in North American populations of P. sorghi that are avirulent against Rp1-D is necessary to determine if these genes will be as widely effective as the Rp1-D gene has been. Two subpopulations of P. sorghi were detected from the bulked population after it was sequentially cultured for at least five cycles on seedlings with Rp1-C or with Rp1-J. The subpopulation cultured on Rp1-J was avirulent on lines with Rp1-C/L/N, Rp1-B, and Rp1-M; whereas the subpopulation cultured on Rp1-C was virulent on lines with each of these genes. Both subpopulations were virulent on lines with Rp1-D.

Stewart's Wilt Reactions of an International Collection of Zea mays Germ Plasm Inoculated with Erwinia stewartii.

Maize accessions were evaluated in 1997, 1998, and 1999 to identify additional sources of Stewart's wilt resistance and to determine if reactions differed among accessions collected from various regions of the United States and throughout the world. The distributions of Stewart's wilt reactions rated from 1 (no appreciable spread of symptoms) to 9 (dead plants) were relatively similar among groups of accessions from all regions of the world except for those from the Mid-Atlantic/Ohio River Valley region of the United States, the southern United States, and the northeastern United States. The mean and median Stewart's wilt rating for 1,991 accessions evaluated in 1997 was 4. The mean Stewart's wilt rating for 245 accessions collected from the Mid-Atlantic/Ohio River Valley region was 3.1, which was significantly lower than that for accessions from all other regions. The mean rating for accessions from the southern United States was 3.7, which also was lower than mean ratings for accessions from all other regions. Ratings from trials in 1997 and 1998 were highly correlated (r = 0.87) for 292 accessions and 15 sweet corn hybrid checks evaluated in both years. Of 20 accessions rated below 2 in 1997 and 1998, seven were from Virginia, seven were from the Ohio River Valley or central Corn Belt of the United States, four were from the northern or western Corn Belt of the United States, and two were from Spain. Ratings for these accessions ranged from 1.7 to 3.1 in 1999. Ratings ranged from 2.6 to 3.7 for F1 hybrids of these accessions crossed with one of two susceptible sweet corn inbreds, CrseW30 or Crse16, which were rated 5.7 and 5.4, respectively. Based on the reactions of this collection of germ plasm, it appears that high levels of Stewart's wilt resistance are prevalent only among accessions collected from areas where the disease has been endemic for several years, whereas moderate levels of resistance can be found in accessions collected from nearly everywhere in the world.

Variation Associated with Silk Channel Inoculation for Common Smut of Sweet Corn.

Efforts at breeding for resistance to common smut of maize (Zea mays), caused by Ustilago maydis, are hampered by the lack of a reliable and efficient method of inoculation. Silk channel injection is one of the most efficient methods of inoculating for ear galls but is less consistent than acceptable for accurate assessment of the response of genotypes. The objective of this study was to examine how the silk channel inoculation method can be modified to reduce variation while maintaining efficiency for large-scale field inoculations. Variation associated with inoculum concentration and variation among people inoculating were examined. Incidence and severity of symptomatic ears increased with inoculum concentration. Concentrations between 105 and 106 sporidia/ml are recommended. Variation among people was greater than variation among inoculum concentrations. Incidence and severity ratings were lower for people inexperienced at inoculating with U. maydis than for experienced people. Variation among people inoculating can be controlled by appropriate experimental design.

Effects of Silk Maturity and Pollination on Infection of Maize Ears by Ustilago maydis.

Host resistance is the most efficient method of controlling common smut of maize (Zea mays), caused by Ustilago maydis. Precise timing of ear inoculations with U. maydis relative to silk maturity and pollination may improve the ability to screen maize germ plasm for resistance. The objectives of this study were to determine the length of time maize kernels can be infected by U. maydis through silks, and to examine the effects of pollination on infection through silks. Two field studies were done in 1995, 1996, and 1997 at the University of Illinois South Farms. In the date-of-inoculation study, ears were inoculated at 2- to 3-day intervals from early silk emergence until 16 days after silk emergence. In the date-of-planting study, hybrids were planted on four dates and ears were inoculated on the same day for all planting dates. In each study, ear shoots were covered with shoot bags prior to silk emergence to prevent pollination, or ear shoots were left uncovered to allow silks to be pollinated normally. Maize ears were susceptible to infection by U. maydis from silk emergence until 8 to 14 days after silk emergence. During this period of susceptibility, incidence of ears with galls decreased as silks aged. Incidence of ears with galls on plants inoculated 7 days apart differed by as much as 70%. The period that maize ears were susceptible to infection by U. maydis was shorter and incidence of ears with galls decreased more rapidly when silks were exposed to pollen than when silks were not exposed. The silk channel method of inoculating for common smut does not appear to be practical for large-scale evaluations of numerous lines. The method is practical for evaluating a limited number of lines or for inducing ear galls for commercial production of huitlacoche (smut galls eaten at an immature stage).

Reactions of Processing Sweet Corn Hybrids to Gibberella Ear Rot.

Gibberella ear rot has increased in prevalence recently on late-season processing sweet corn grown in the Midwestern United States. Little information is available about the reactions of hybrids currently grown for processing. A total of 52 processing hybrids were evaluated in 1996 and 1997 for reactions to Gibberella zeae following one of two methods of inoculation: inserting an infested toothpick or injecting a spore suspension in the silk channel 7 days after the mid-silk growth stage. Ratings of ear rot severity ranged from 2.4 to 8 on a scale of 1 to 9. Severity of ear rot differed among hybrids, but none of the hybrids was highly resistant, nor did the hybrids display a great degree of variation for reaction to G. zeae. Severity of ear rot was consistently lower for cvs. Sprint and HMX 5372 than for the other hybrids. Ear rot ratings also were lower for cvs. GH 1834 and GH 2690 in three of four comparisons. Severity of ear rot was consistently higher for cv. GH 2628 in all comparisons and for cvs. Empire, Excalibur, GG 42, GH 1861, Jubilee, Rival, and WSS 3680 in three of four comparisons. Ranks of some hybrids changed between inoculation methods, years, or both, but most hybrids were ranked similarly in at least three of four comparisons. Coefficients of variation were approximately 5% higher for the injection method of inoculation in 1996 than for the other trials. The toothpick method of inoculation required less labor and ranked hybrids as consistently or more consistently than the injection method.