LAUSR

Abstract Forecasting the biomass allocation among source and sinks organs is crucial to better understand how grapevines control the distribution of acquired resources and has a great meaning in term of making decisions about agricultural practices in vineyards. Modelling plant growth and development is one of prediction approaches that play this role when it concerns growth rates in response to variation in environmental conditions. This study was aimed to model the dynamics of current year’s above‐ground biomass in grapevine. Furthermore, the development of a relatively simple growth modelling framework aimed at the derivation of cardinal air temperatures for growth in grapevine. Trials were carried out over three growing seasons in field conditions with four grapevine cultivars. To compare the differences of growth-allocation models among cultivars, the non-linear extra-sums-of-squares method was used. Using measurements of mean daily air temperature and dry mass increments a beta-function model was fitted to the data and used to estimate cardinal air temperatures. Shoot growth and biomass allocation differed significantly among cultivars. The application of the non-linear extra-sums-of-squares procedure demonstrated to be a feasible way of growth models comparison to statistically assess significant differences among grapevine cultivars. The results of this study highlight parameters most involved in the phenotypic variability of shoot growth. Variations among cultivars result from environmental and genetic factors. Results do not support the use of common coefficient estimates for all cultivars within a specific location. The temperature response functions obtained, confirm the initial working hypothesis that because the varieties may have either different temperature optima or different thresholds that a unifying model cannot be achieved. These results suggest that some caution should be taken when incorporating shoot development and biomass partitioning coefficients in a growth model. Use of common coefficients estimates for all cultivars for dynamic modelling approaches, in fact, may result in a poor representation of the data early or late during the course of the season. The described approach can be used to account for complex variation in seasonal growth patterns and provides insight into how well a cultivar may be matched to a particular site.