thought to be the main pollinators of pawpaw, and many of these insects have been observed each spring in the pawpaw orchards at KSU. Low pollinator activity was observed on cool cloudy spring days during the study. This is likely because this population was located at a higher elevation in the orchard than any other group in this study, and therefore did not sustain as much frost damage, so more flowers and developing clusters survived the frost.
Pawpaw differs from many other tree fruits in that it produces flowers for an extended period of time, and fruit is also harvested over a long period of time, usually around 2 weeks per tree. Staggered bloom period in the spring also extends harvest period. Each fruit cluster develops from an individual flower, and fruit within a cluster often ripen at different times. It would be desirable for commercial pawpaw growers to identify and develop genotypes that have a more concentrated harvest time. Length of flowering time did not consistently affect the length of harvest in pawpaw. Trees with a longer flowering time were found to have a longer harvest interval in only one year. However, trees with an earlier flowering peak did generally have an earlier harvest peak.
It is possible that the long flowering period of pawpaw is a mechanism to ensure greater fruit set. By spreading flowering and fruit set out over a longer period of time, perhaps the entire crop would not be wiped out by a frost event. However, length of flowering in pawpaw was found to lead to a greater fruit set only in 2003.
Pawpaw yields are notoriously low (Peterson 1991). Pomper and Layne (2005) reported yields that averaged 6.4, 2.0, and 3.7 kg per tree for ‘PA-Golden’, ‘Wilson’, and ‘Sunflower’, respectively, for trees in their 5 year. In another study, yields per tree in the 7th year were: 4.4 kg for ‘Sunflower’, 2.3 kg for 8-20, and 2.2 kg for ‘PA-Golden’ (Pomper et al. 2003d). Yields among populations varied; however, the highest yielding tree in 2004 produced 28.2 kg per tree, demonstrating that some genotypes may have potential as future cultivars or in breeding efforts. The fruit weights observed (>180 g) and tendency toward few or single fruit per cluster, which is desirable to growers for easier harvest of intact fruit, of some selections were more desirable than that of some currently available commercial cultivars.
It was hypothesized that southern pawpaw accessions would flower earlier than populations from more northern areas, since more northerly pawpaws may have adapted