Summary• Substantial evidence has been published in recent years demonstrating that postphotosynthetic fractionations occur in plants, leading to 13 C-enrichment in heterotrophic (as compared with autotrophic) organs. However, less is known about the mechanism responsible for changes in these responses during plant development.• The isotopic signature of both organic matter and respired CO 2 for different organs of French bean ( Phaseolus vulgaris ) was investigated during early ontogeny, in order to identify the developmental stage at which isotopic changes occur. Isotopic analyses of metabolites and mass balance calculations helped to constrain the metabolic processes involved.• At the plant scale, apparent respiratory fractionation was constantly positive in the heterotrophic phase ( c . 1‰) and turned negative with autotrophy acquisition (down to -3.08‰). Initially very close to that of the dry seed (-26.83 ± 0.69‰), isotopic signatures of organic matter and respired CO 2 diverged (in opposite directions) in leaves and roots after onset of photosynthesis. Respired CO 2 reached values up to -20‰ in leaves and became 13 C-depleted down to -29‰ in roots.• It was concluded that isotopic differences between organs occurred subsequent to metabolic changes in the seedling during the transition from heterotrophy to autotrophy. They were especially related to respiration and respiratory fractionation.

Summary• Root respiration is a major contributor to soil CO 2 efflux, and thus an important component of ecosystem respiration. But its metabolic origin, in relation to the carbon isotope composition (δ 13 C), remains poorly understood.• Here, 13 C analysis was conducted on CO 2 and metabolites under typical conditions or under continuous darkness in French bean (Phaseolus vulgaris) roots. 13 C contents were measured either under natural abundance or following pulse-chase labeling with 13 C-enriched glucose or pyruvate, using isotope ratio mass spectrometer (IRMS) and nuclear magnetic resonance (NMR) techniques.• In contrast to leaves, no relationship was found between the respiratory quotient and the δ 13 C of respired CO 2 , which stayed constant at a low value (c. -27.5‰) under continuous darkness. With labeling experiments, it is shown that such a pattern is explained by the 13 C-depleting effect of the pentose phosphate pathway; and the involvement of the Krebs cycle fueled by either the glycolytic input or the lipid/protein recycling. The anaplerotic phosphoenolpyruvate carboxylase (PEPc) activity sustained glutamic acid (Glu) synthesis, with no net effect on respired CO 2 .• These results indicate that the root δ 13 C signal does not depend on the availability of root respiratory substrates and it is thus plausible that, unless the 13 C photosynthetic fractionation varies at the leaf level, the root δ 13 C signal hardly changes under a range of natural environmental conditions.