Flower buds of two sweet cherry (Prunus avium L.), 12 sour cherry (Prunus cerasus L.) and one ground cherry (P. fruticosa Pall.) were collected monthly from Aug. 1990 to Mar. 1991, and subjected to freeze tests to determine the level of cold hardiness. LT50 values (temperatures at which 50% of the flower buds were killed) summed over all months were significantly correlated (r = 0.6844, P ≤ 0.01) to the flower bud low temperature exotherms (LTEs). Correlation of LTEs to LT50 values was highest, r = 0.85, P ≤ 0.01 for the acclimation and midwinter period, November to February collections. During this period the average LT50 occurred before and within 2.5 °C of the LTE, indicating tissue injury before the LTE occurrence. During deacclimation, represented by the March collection, the LT50 began within 2.0 °C, on average, of the LTE, and in 11 of 12 cultivars and seedlings preceded the LTE. In March, the correlation of LTEs to LT50 values was less, r = 0.69, P ≤ 0.05, indicating possible changes flower bud deep supercooling. LTE values were selected as a measure of flower bud hardiness in sour cherry. Exotherms were not detected in the flower buds of all germplasm tested on all evaluation dates, but were the best means of separating selections. While LTE analyses expressed significant differences in November, December, and March at P ≤ 0.01, the LT50 analyses expressed differences only in December and February at P ≤ 0.05. The relationship between ambient temperatures and floral tissue hardiness indicated that November and March are two critical times for flower bud injury. November injury would occur in years when sudden low temperatures occur without sufficient pre-exposure to freezing temperatures. March injury would occur in years when sudden freezing temperatures follow warm days. This type of injury would be most pronounced in southern genotypes. Spring freeze injury could be significantly reduced by the selection of cultivars and seedlings that have delayed deacclimation. Exotherm occurrence and bud volume were correlated (r = 0.95, P ≤ 0.05). In January, when exotherms were least prevalent, they were generally present only in the five cultivars and seedlings with large bud volumes. The LTEs in midwinter, occurred within 3 °C of the reported average annual minimum temperature for the northern range of Prunus commercial production (Zone 6). The results of the principal component analysis of flower bud LTEs indicated that other selection criteria as flowering time might have played a more significant role in the hardiness range of sour cherry than simply geographic origin. The first principal component (PC1), which accounted for 77% of the total variance was used to separate among cultivars and seedlings. Selections at the positive end of PC1 had flower buds that were more cold susceptible than selections at the negative end of PC. This concurs with other research showing that flower bud hardiness is related more to commercial range (i.e., the range of commercial production) than to geographic distribution.
Isozyme and Morphological Variation in a Cornus florida L. Provenance Plantation Representing Geographically Diverse Populations
