O-7
Serological quantification of Fusarium graminearum in barley
Presenter: Nicholas S. Hill, University of Georgia, Athens, GA
Co-Authors: P. Schwarz and S. Neate, North Dakota State University, Fargo, ND; B. Cooper, Busch Agricultural Research Institute, Ft. Collins, CO; L. Dahleen, USDA-ARS-CRU, Fargo, ND; and R. Horsley, North Dakota State University, Fargo, ND

A simple and accurate method of quantifying the presence of Fusarium graminearum limits our ability to understand disease epidemiology, develop control measures, or improve resistance via breeding or biotechnology. Breeding based upon visual scores of head blight incidence have greater heritability than selecting low deoxynivalenol (DON), but visual scoring is not easily standardized. Head blight scores (mean r = 0.21, std. dev. = 0.24) and DON (mean r = 0.46, std. dev. = 0.22) were poorly correlated among locations in the 2003 NABSEN field trials. Both visual scoring and DON analysis are time-consuming, making neither amenable to multiple analysis for inter- or intra-experiment quality control assessments. Thus a better method of Fusarium quantification is needed to measure head blight that can be used as quality assurance methods for end users. An ELISA test for Fusarium in small grains and corn was developed. Experimental error associated with visual scoring, DON analysis, and ELISA evaluation of Fusarium presence was evaluated in three replicated field experiments grown in Osnabrock, Langdon, and Casselton, ND. Each were visually scored for FHB, analyzed for DON by GC-EC, and analyzed for F. graminearum by species-specific indirect ELISA. Coefficients of variation and a correlation matrix were calculated for each response variable. We obtained 89 samples from grain elevators in North Dakota and correlated FHB, DON, and ELISA data with one another. In another study, we developed a method to sequentially extract DON and Fusarium antigens to quantify both within individual seeds, thus eliminating variability to sampling. Mean coefficients of variability for the ELISA values were lower than FHB or DON in the field studies. Mean correlation coefficients among the field experiments were greater for ELISA vs. DON than for ELISA vs. FHB or DON vs. FHB. Similarly, correlations coefficients between ELISA and DON were greater than ELISA vs. FHB or DON vs. FHB for the grain elevator samples. Analysis of individual seeds for DON and ELISA did not improve goodness of fit among samples. Thus, we grew 3 isolates of F. graminearum in 4 different media with 10 replications. Mycelial growth was less on Shenk Hildebrandt medium than other media, but other media did not differ from one another. Fusarium antigen per gram of mycelium was not different among media or isolates. DON was highly variable within each medium, ranging from 0 to 26 ppm. Subsequent to these studies we have analyzed as many as 1000 barley samples (in duplicate) in one day using ELISA with an R sq. = 0.97 for the repeated measures. Consistency, speed, and ease of analyses make ELISA a superior method for quantifying F. graminearum.

Dr. Hill received his Ph.D in agronomy from Kansas State University. He has been a faculty member of the Dept. of Crop and Soil Sciences at the University of Georgia since 1986. He specializes in development of diagnostic procedures (NIR, HPLC, and immunochemistry) for plant disease organisms, endophytes, and mycotoxins as well mycotoxin metabolism and transport in livestock. He is a fellow in the American Society of Agronomy, member of the Crop Science Society of America, and Association of Seed Analysts. He is founding partner and president of Agrinostics Ltd. Co. (www.agrinostics.com), a company dedicated to immunochemical test kit development and contract research.