SAMPLE VARIATION AND RESOURCE-ALLOCATION FOR ERGOT ALKALOID CHARACTERIZATION IN ENDOPHYTE-INFECTED TALL FESCUE

Ergot alkaloid concentrations vary among genotypes of endophyte-infect ed (E+) tall fescue (Festuca arundinacea Schreb.). Because the ergot a lkaloids are responsible for detrimental health disorders in grazing l ivestock, but the endophyte Neotyphodium coenophialum (Morgan-Jones an d Gams) Glenn, Bacon and Hanlin is mutualistic, development of E+ tall fescue populations low in ergot alkaloid concentration may be a susta inable non-toxic solution for livestock producers, Current techniques for screening ergot alkaloids in tall fescue are labor and equipment i ntensive and time consuming, The objective of this research was to det ermine the number of replications and locations needed to detect speci fied differences among treatments varying in alkaloid concentration. S ix E+ tall fescue genotypes were harvested twice at two held locations and one greenhouse location in 1 yr, and the number of replications a nd environments required to show specific treatment differences for al kaloid concentration calculated within harvest dates. Ergot alkaloid c oncentrations for the genotypes ranged from 488 mg kg(-1) to 2504 mg k g(-1) in the spring, and 219 mg kg(-1) to 986 mg kg(-1) in the summer. Genotypic variances were proportional to the mean; hence, data were s ubjected to a log(10) transformation. Treatments harvested twice with differences of 50% of the experimental mean were detectable (P = 0.05) in experiments containing two replications grown at four environments , or two replications grown at three environments and harvested once w hen nontransformed or transformed data were used, respectively. Use of transformed data will be necessary to keep the number of observations per experimental unit manageable.