It has become increasingly clear that all cultivars do not respond equally in the fall to changes in photoperiod and temperature. Additionally, some "old" check cultivars produce inconsistent data from year to year and location to location. Lahontan has been particularly susceptible to this criticism. Yet, fall dormancy class (FD) is frequently assigned based on a single trial. Evaluation of fall dormancy in single environments has led to misclassification of some cultivars. To alleviate some of these problems, a "new" larger set of check cultivars was proposed in 1995 with three cultivars for each fall dormancy class. However, inclusion of these checks was largely based on data produced at a single Midwest location. We were concerned that multiple (but not the same) checks grown across a broad range of environments (latitudes and years) could result in further confusion when assigning fall dormancy class. To elucidate the magnitude of this problem, we initiated studies in 1995 to evaluate fall dormancy using four California locations differing substantially in latitude and temperature. Additionally, we wished to identify appropriate check cultivars for extremely non-dormant germplasm (Class 10 and 11). Trials at each location were conducted according to the standard protocol for evaluating fall dormancy and included all the old and new check cultivars (1). Fall regrowth scores were square root transformed to remove heterogeneity of variance creating a variable, natural plant height (NPH). We also utilized the data to determine variance components and the theoretical standard errors for NPH of a cultivar mean. Significant GxE interactions are present for NPH when all cultivars are included in an ANOVA. The G*L*Y interaction is 6x the G*Y and 50x the G*L variance. Rank changes among the checks result in some "check" cultivars ordering with cultivars representing dormancy groups as much as two classes from their proposed classification. A sub-set of the check cultivars has been identified -- Maverick, 1; 526, 2; 5246, 3; Legend, 4; Archer, 5; ABI 700, 6; Dona Ana, 7; Pierce, 8; CUF 101, 9; UC-1887, 10; and UC-1465, 11. When the NPH of this sub-set is regressed against FD the same slope is equivlivent to that of old checks and there is minimal deviation from regression (FD= 6.36(NPH) — 7.68, r²=0.992). We detected no significant GxE among these cultivars. Finally, because of the GxE among all cultivars, FD should be assigned based on two years of testing at a minimum of three locations.

Previous Page