Predicting frost risks in plants
Résumé
For about ten years, I have been interested in the influence of frost constraints on plant physiology and in species adaptative responses. My approach combines different scales and disciplines (physics, bioclimatology, physiology, ecology and modeling) and integrates the underlying physiological processes (e.g. hydraulic conductivity in the vascular tissues, frost acclimation of living tissues and frost avoidance via dormancy and phenology), ultimately leading to better predict frost risks in overwintering plants. Mountain ecosystems are strongly limited by frost constraint at high and, even, lower elevation, but climate change is likely to jeopardize the current distribution. Actually, three components interact in the definition of frost risks: hazard (temperature), exposure (organ and freezing temperature) and vulnerability (temperature of resistance). Will climate change only release frost pressure by decreasing the hazard, or, more likely, will we assist to complex interactions including non-linear relationships, threshold effects and feedback loops? To which extent current species distribution will be affected by frost in the future is not easily predictable because (i) the vulnerability to frost acclimation is under environmental (photoperiod and temperature) and physiological control (synthesis of cryoprotective compounds), limiting acclimation in autumn and increasing deacclimation in spring (ii) the exposure to frost is mainly controlled by snow cover height and duration (insulating perennial parts from low temperature, freeze-thaw cycles and ice-induced dehydration). To predict and decrease frost risks under climate change, I develop generic and process-based approaches in perennial species to predict frost vulnerability, exposure and damages. It should help to select adapted ideotypes with respect to future environmental conditions, including interacting constraints (eg. prolonged warm and/or dry period followed by sudden freezing)