A better characterization of the potato dormancy helps optimizing storage by 1) reducing the application of anti-sprouting products and, hence, storage costs, and 2) increasing the benefits for human health and the environment. The main objective is to develop statistical models to predict dormancy period using parameters related to growing conditions. To build those models, data were collected from field experiments managed in Switzerland during 25 years in five different locations and with 721 cultivars. The available explanatory variables were the following: cultivar, year, location, potato physiological stages, weather and agronomic data. The harvested tubers were stored at 8°C and 85%RH and the sprouting initiation was measured and used as dependent variable. Data was analyzed as follows: (1) analysis of variance to quantify the importance of explanatory variables on dormancy period; (2) creation of a descriptive model using regressions to study and quantify the effect of the explanatory variables and their interactions on the dormancy. Preliminary results showed that the variable “cultivar” was the most important one, explaining around 60% of the variation in the dormancy period followed by the variables “year” (20%) and “location” (4%) (p < 0.001). The regression analysis demonstrated that the combination of cultivar and the sum of temperatures between the emergence and the harvest explain 80% of the variability of the dormancy (R2=0.80; p < 0.005). The models underline the importance of genetic and climatic parameters to estimate the dormancy period. Our work will be instrumental to optimize the control of sprouting during potato storage.
