Modelling the effects of microclimate on bean seed desiccation rate and seed storage ability

International audience

Английский. Bean seed storage ability is of major interest for seed firms and depends, at least partially, on the effect of desiccation microclimate and desiccation rate on the mother plant; however, their effects are contradictory and not quantified. Therefore, our aims were (a) to predict seed desiccation rate from microclimatic variables at the pod level and (b) to measure the effects of seed desiccation rate and temperature on seed storage ability. For 4 years, beans (Phaseolus vulgaris) were sown at different dates in the field and once in a greenhouse. Pods having seeds at the beginning of desiccation stage were selected according to their colour and then dried in the field or under controlled conditions. In the field, pods were dried under both natural conditions and conditions modified either by defoliation or by water spraying. Under controlled conditions, pods were dried under a wide range of temperature and relative humidity combinations. Seed water content, temperature and relative humidity were measured throughout the desiccation phase. Storage ability was measured by a standard germination test after either 11 or 15 days of controlled deterioration. Seed desiccation rate was highly correlated to air vapour pressure deficit, which measures the combined effects of temperature and relative humidity on drying. With a logit transformation, deterioration test results showed a linear decrease in storage ability as seed desiccation rate increased for either controlled or field-dried lots. However, for controlled lots desiccated at temperatures lower than 24 degrees C, the lower the temperature, the lower the storage ability at a given desiccation rate. Moreover, the relationship between the results of the two deterioration periods differed between controlled and field-dried lots. Finally, the high correlation between seed storage ability, desiccation rate and vapour pressure deficit in the field has given new insights for building a crop decision-oriented model for optimising harvesting