Abstract
Alternate bearing is a common phenomenon for different perennial species. While it can have dramatic economic effects, its causes are still unclear. Alternate bearing results from irregularities in floral induction (FI) between successive years. In apple trees (Malus x domestica), large variability in bearing patterns between genotypes have been observed that are partially affected by architectural characteristics. Moreover, large within tree variability in FI also exists since FI occurs each year in a set of buds, only. Different hypotheses exist to explain between years and within tree variability in FI including the involvement of hormones (mainly gibberellins, GA), other signaling molecules (FT protein) and the carbon balance at different scales of plant organization. This thesis aimed at analyzing and modeling the physiological and architectural determinisms of the within tree and between year variability in floral induction in context of genotypic variability in apple tree. First, physiological analyses (photosynthesis, nonstructural carbohydrates and GA concentration) were performed during 2 years on adult ‘Golden Delicious’ apple trees subjected to different leaf and fruit removal practices in order to modify the leaf/fruit ratio and the distances between the sources of inhibiting and activating signals and shoot apical meristems (SAM). Results showed that effects of fruit and leaves could affect FI in SAM at short distances, only. Moreover, experimental results showed that carbohydrates appeared as not directly implicated in FI but that GA could inhibit FI in SAM. The dataset generated from this experiment was used in a modelling approach aiming at simulating the effects of leaf, fruit, and their distances to SAM on FI. The model simulated FI variability with 3D tree structures by considering the transport of inhibit and activating signal produced by fruit and leaves respectively. Signal transport was simulated considering a signal ‘attenuation’ parameter, whereas SAM fate was determined by probability functions depending on signal amounts. Model parameter estimations suggested a cumulative effect of activating and inhibiting signals on FI, with SAM being more sensitive to the inhibiting signal. Simulations results also proposed that the activating signal was transported at shorter distances than the inhibiting one. Other experiments performed on a set of genotypes issued from an apple collection, showed that the establishment of plant architecture is under the dependency of hormonal signaling (AIA and ABA) and plays a key role in the onset of biennial bearing between the different genotypes. Results also confirm the poor implication of carbohydrates on FI and showed that gibberellins and cytokinins may be implicated in FI variability between genotypes. At the end of this work, new perspectives are proposed for the identification of the signals controlling FI and for integrating genotypic variability in the modeling approach in order to simulate between-year variability in FI and the resulting bearing patterns.