Olive fruit persistence is a crucial component of yield and an important factor in estimating alternate bearing potential.
Unfortunately, measurement of fruit persistence exhibit considerable variation, with coefficients of variation greater than 100. Such a high degree of variation makes field studies on questions regarding flowering and fruiting unmanageable due to the large number of experimental units necessary. To determine the source of this variation and how it might be reduced, comparisons of flower and fruit number per node were made within branches and trees over the course of two seasons. Results show that while the largest population of flowers are most distal on the branch, the central portion of the branch contains the majority of the final fruit population and has the lowest coefficient of variation. Furthermore, variation in the number of flowers and fruits is greater between branches than between nodes or trees. The implications of these data on experimental design are discussed and a design is proposed for reducing variation and labor needs.