MAppleT

MappleT simulates the growth and branching processes of apple tree with Markovian models estimated on previously collected data set (Fuji cv). At axis scale, growth units (GU) are represented by four states (“long”, “medium”, “short” and “floral”) which succession is modeled by a Markov chain. The branching of long and medium GU is organized by zones, each one being characterized by laterals it contains, their succession being modeled with a Hidden Semi-Markov Chain which parameters depend on the parent shoot length. Branch bending is supported by a biomechanical model, where the torques imposed by gravity and tropism are applied to each internode, recursively extended to the axis; the simulation of cambial growth is supported by the pipe model. The individual leaf area and internode length development follow a logistic growth curve over a period of 12 and 10 days, respectively, with final dimensions indexed on the branching zone.

This model was used for calculating light interception efficiency (Da Silva et al. 2014) and to design ideotypes (Picheny et al. 2017, Migault et al. 2017).