EVALUATION OF SIMPLE METHODS FOR ESTIMATING BROAD-SENSE HERITABILITY IN STANDS OF RANDOMLY PLANTED GENOTYPES

Inexpensive estimates of broad-sense heritability (BSH) may be valuabl e in plant breeding. This research evaluated two methods for estimatin g BSH with data from stands of equidistantly spaced genotypes. Both me thods depend on the assumption that genetic and environmental contribu tions to plot variance (plot = group of contiguous plants) change at d ifferent rates as plot size changes if genetic variation is distribute d randomly. For the method proposed by V.J. Shrikhande, variances amon g plot means are computed and least-squares regression used to estimat e environmental and genetic variances and the change in a plot varianc e with changes in plot size. The other method involves the same approa ch, but uses a two-parameter model suggested by G.H. Freeman but not p reviously used to estimate BSH. Both methods produce biased BSH estima tes since genotypic and genotypic x environmental components of variat ion are inseparable, Our objectives were to: (i) develop software to c alculate variances for the methods, and (ii) compare BSH estimates gen erated using these methods with each other and with those from analysi s of variance (ANOVA) of data from families grown in the same environm ent. Data were from a perennial herb, Sphaeralcea emoryi Torr, grown i n Tucson, AZ. Shrikhande's method produced parameter estimates with la rge variances and BSH estimates that averaged 3.6 times larger than th ose from Freeman's method. Only BSH estimates from Freeman's method ag reed well with those from ANOVA for most traits. Freeman's method may be useful for rapidly and inexpensively generating BSH estimates in a variety of situations, especially when traditional genetic analysis ar e difficult.