Résumé
The concept of soil nutrient availability is still widely viewed within the framework of crop yield responses to fertilizer applications as the intermediary variable linking the rate of application of a single nutrient to the absorption of this nutrient by plants according to the pioneer work of Boussingault (1855), von Liebig (1855), and Mitscherlich (1924). For interpreting the huge variability of crop yield responses to increasing fertilizer applications, agronomists and soil scientists have focused on soil nutrient dynamics in order to estimate the quantity of each nutrient available for plant uptake. This linear approach considering “available nutrient in soil” as an external factor to which plants respond does not correspond to the reality for three main reasons: (i) the root absorption capacity is deeply feed-back controlled by the plant growth capacity itself and, therefore, does not depend univocally on soil nutrient availability; (ii) interactions among different nutrients in soils and plants imply that the availability of one nutrient for plants depends of the availability of others, requiring a more integrated approach; and (iii) the plant itself influences nutrient dynamic processes in soils through interactions with microbial communities in its rhizosphere. Consequently, soil nutrient availability cannot be only considered as a property of the external medium to which plants adapt, but also, and more importantly, as resulting of the functioning of the whole plant-soil-living organisms ecosystem. This review paper proposes an integrated and hypothesis-based vision of plant mineral nutrition based on several recent findings: (i) the corroboration and verification of hypotheses of regulation of plant nutrient uptake at the whole plant level by recent advances in the molecular physiology of plant nutrition, (ii) the physiological basis for interactions among different plant nutrients, and (iii) the increasing evidence of plant-soil interactions at the rhizosphere level.