* The plant simulated structure is a framework able to host functional and specific processes, allowing a wide range of applications

* Girth growth (secondary growth) can be estimated.

* The approach shows genericity

* The plant structure is explicit hence 3D representation enabling their applications: visualization, interaction with pests, mechanics, pruning simulation, etc.

* Field measurements, calibration and validation are heavy and difficult to carry out.

* FSPM show high computing and memory costs in order to simulate the parallel development growth of organs

* FSPM show high computing and memory costs involved in estimating biomass production, especially related to light interception computation, known to be a nonlinear problem (ray tracing)

* FSPM show high computing costs for biomass transport and partitioning in the plant architecture

* Such models usually involve many parameters, preventing inverse method definition and optimisation.

* Stochastic behaviour cannot usually be considered; mean values and variance of production are difficult to assess, requiring sets of numerical experiments

* Model extension from single plant to plant population is not clear and adds a high level of computation complexity