To validate one or the other hypothesis, we developed predictive models fed by the data provided by the experiment and by data founding in literature.  

Predictions based on a model

The aim of the modeling is to predict the channel of the signal based on the response time observed in the experiment. 
To achieve this objective, we employed the CRootBox model to create a realistic root and mycorrhizae architecture that we used to check if the signal could be transmitted using the mycorrhizae as a vector of information transport and to characterize the production sites of signal molecules. Then, we created a model of passive diffusion through a homogeneous matrix simulating the soil in order to estimate the time necessary to transmit a chemical signal from a plant A to a plant B. By comparing experimental results with those obtained by modeling, we want to determine the type of signal that would have been used to communicate between plants.