Abstract
Several aspects of the sampling of apple fruit for mineral analysis and prediction of bitter pit incidence were studied. Variability of mineral analyses can be minimised, and precision of bitter pit prediction increased, by sampling fewer fruit per tree from a large number of trees. Less variability in Ca analysis was found by taking two opposite plugs of cortical tissue from each fruit. Fruits from the upper parts of trees had less Ca and a higher incidence of bitter pit. If a high proportion of fruit in the crop are from these sites, then a sampling scheme which excludes them may result in less accurate predictions. The relationship between fruit Ca concentration and fruit size varied with the crop load on tree tree. In heavily laden trees, the decrease in Ca concentration with increasing fruit weight was half that in lightly laden trees. The relationships between fruit size and both bitter pit incidence and Ca concentration were sufficiently strong to be useful in standardising sampling schemes for analysis and prediction.