Investigating the impact of source/sink manipulations on within tree variations of floral induction and fruit weight

In apple, carbon source/sink relationships are assumed to affect fruit growth and floral induction and consequently tree flowering potential. Nevertheless, the effect of local variation of carbon source/sink relationships and distances within trees on these processes remains unclear. This study aimed at investigating the effect of distances between leaves, fruits and meristems on fruit size and floral induction. Experiments were performed on adult trees of the ‘Golden’ cultivar from 2015 until 2017 in the south of France. Thirteen treatments with different kinds of local leaf or fruit removal were set up in mid-June on trees previously managed to be either in low or high crop loads. Mean fruit weight and floral induction rate were evaluated in the different parts of the tree including defructified and defoliated parts.The floral induction rate was reduced on parts of the tree that were defoliated, whatever the defoliation treatment and, therefore, the distance to remaining leaves. Moreover, no impact of defoliations on floral induction was observed on not-defoliated parts of the trees. Both together, these results show a major impact of local source/sink conditions on floral induction. Fruit weight was reduced on the parts of the trees that were defoliated when leaf removal was performed on one side of the Y-Shape trees or on half of the branches. This decrease was also observed but to a lesser extent on defoliated shoots when leaf removal was performed randomly on half the shoots. This suggests that carbon fluxes came from the neighboring shoots to sustain fruit growth of non-leafy shoots. Fruit weight increased for trees subjected to fruit removal treatments, probably due to a flux of carbon from the parts of trees without fruits. This study confirms that source-sink relationships have a great impact on floral induction and fruit growth. Moreover it gives new knowledge on the scale of plant organization at which source-sink relationships affect these processes. These experiments will be used to calibrate a functional structural model built to simulate floral induction variability within trees.