Modelling the spatial variability of the vine water status at a within field scale
Résumé
This work aims at proposing a first approach towards a spatial model of the vine water status. The goal is to test whether or not it is possible to predict the plant water status using such a model at several unsampled locations and at different time considering only one reference measurement for each prediction date. An experiment was carried out on a vineyard located in Languedoc-Roussillon region, France, during 2003 and 2004. The vineyard is a non-irrigated Syrah field of 1.2 ha. planted on calcareous soil. 49 sites of measurements (s1, s2 ,..., si : i=1,2, ..., 49) were defined on a regular grid over the whole field. For each point grid, the Predawn Leaf Water Potential (PLWP) was measured at 7 different dates in 2003 and 6 different dates in 2004. The results exhibit a significant variability of the PLWP values either in space or in time. For each year, results point out a strong linear relationship between all the values of PLWP measured on the grid. A simple model was proposed. It allows us to perform prediction of plant water potential and its spatial variability over the whole field from only one measurement performed on a reference site. Moreover, the results that taking into account the within field variability of the plant water status becomes relevant only once a significant water restriction is reached. Further studies will aim at confirming the relevancy of this approach in different locations and also on other varieties