SOME FACTORS INFLUENCING FRUIT-SET IN SUNDROP APRICOT

Two apricot (Prunus armeniaca L.) cultivars, 'Sundrop' and 'Moorpark' were found to be cross-compatible with pollen of 32 genetically divers e cultivars. The flowers of 'Valleygold', 'CluthaGold', and 'Moorpark' remained receptive to pollination for up to 8 days after emasculation at the balloon stage. Floral receptivity declined in 'Moorpark', 'Sun drop', and 'CluthaGold' (but not 'Valleygold') over the 8 days. 'Sundr op' had the lowest receptivity and 'Moorpark' the highest. Percentage set of 'Sundrop' flowers varied between properties and years in Centra l Otago, New Zealand, with significantly lower sets being recorded in 1992 and 1993 than in 1991. In these three seasons, a 2 ha block of 'S undrop', with no pollenisers within it, showed a decrease in fruit set as the distance from the nearest polleniser on the edge of the block increased; fruit set was unacceptably low beyond the first four or fiv e rows. Bee density did not decline across the block in the same manne r. The fruit were smaller on trees adjacent to the polleniser but were generally of similar storage quality to those from 10 rows away. The addition of extra beehives in 1992 and 1993 increased the bee numbers per tree throughout the block, but not fruit set. In comparison, anoth er 1.7 ha block of 'Sundrop' on the same property with just 3% of the trees planted at random in the cultivar 'Trevatt', produced a better s et ii each of the 3 years. Over 8 years, correlations were found betwe en percentage fruit set and the meal temperature between 5 and 90% blo om; the number of days when the maximum temperature exceeded 18 degree s C between 5 and 90% bloom; and the meal temperature over the 3 weeks from 90% bloom. The roles of in-hive pollen exchange and self-pollina tion in 'Sundrop' are discussed. Planting of pollenisers within the bl ock remains the most reliable method of increasing fruit set in large blocks of 'Sundrop'.