We present the first species comparison of spatio-temporal variations of respired δ13CO2 (δ13Cres) and its putative substrate (water-soluble organic matter, WSOM) of leaves and roots along the plant axis and over the diurnal course using a rapid in-tube incubation technique. Pronounced spatial differences in δ13Cres (up to 10.2‰) between the most enriched sun leaves at the top canopy and the most depleted roots tips were found along the plant axis. Additionally, a diurnal enrichment up to 15.9‰ of leaf respiration above WSOM was found. Furthermore, we found very rapid post-illumination changes in δ13Cres, which generally exhibited a 2 to 5‰ decrease within 30 min of darkness. Interestingly, the magnitude of this decrease exhibited a diurnal cycle. Compound specific labelling with positional labelled pyruvate and theoretical calculations were used to explore different hypotheses for the observed variations in δ13Cres i) changes in signature and pool-size of the putative respiratory substrate, ii) apparent fractionation in the dark respiratory pathways, iii) potential effects of a transient decarboxylation of an enriched malate pool and, iv) Rayleigh fractionation processes of enzymatic reactions in the respiratory pathways on δ13Cres. The marked dynamics in leaf respired δ13Cres both on the spatial and the diurnal scale were independent from the putative respiratory substrates. Positional labelling experiments showed that the observed diurnal enrichment is due to an increase in C flux through pyruvate dehydrogenase PDH over the light period, probably into the secondary metabolism (e.g. isoprene or aromatic compounds), relative to a constant Krebs cycle (KC) activity. In contrast to foliage respiration none of the investigated species displayed distinct diurnal pattern in δ13Cres of roots. However, neither shifts in PDH and KC activity nor decarboxylation of an enriched malate pool can fully explain the diurnal δ13Cres enrichment or post-illumination changes. A theoretical approach indicates that a Rayleigh fractionation process against 13C of any respiratory enzyme may explain the observed pattern. Further research on the fractionation mechanisms is required, where compound specific labelling provides an efficient method to identify the processes involved in δ13C variation of dark respiration.