than any state except Decatur Co., Indiana (16), and a lower total yield (0.59 kg) than populations from Washington Co., Indiana, New York, or Maryland. Fruit weight was greatest in trees from Maryland (141.8 g). Yield efficiency varied greatly, and once again the West Virginia population had the poorest production, significantly lower than accessions from Washington Co., Indiana, New York, or Maryland. Accessions from Kentucky and Washington County, Indiana had a significantly longer harvest period than trees from West Virginia, Maryland, and Decatur County, Indiana. Harvest peak varied among states, with fruit from the New York accessions ripening earliest and fruit from trees from West Virginia and Kentucky ripening latest. GDDs for ripening varied among populations, with the New York population requiring the fewest GDDs, at 2492, and accessions from West Virginia requiring the highest number of GDDs, 2980.

Trees producing a higher number of flowers led to a higher number of fruit clusters produced by that tree (r=0.60, Table 5.4). An earlier flowering peak also led to an earlier harvest peak (r=0.42), and length of flowering time and length of harvest were found to be correlated (r=0.40). Trunk cross sectional area and percentage fruit set were not correlated, but a higher trunk cross sectional area was found to lead to a higher number of flowers per tree (r=0.75) higher total yield in weight per tree (r=0.44) and a higher number of fruit per tree (r=0.43). Trunk cross sectional area also did not influence the yield efficiency of the tree. Neither length of flowering time nor total number of flowers affected percentage fruit set. Length of flowering did positively affect the total number of fruit produced per tree (r=0.40) and the total yield by weight (r=0.37).