Implication of carbon allocation in plants for soil organic matter cycling.

Plants allocate carbon to roots, shoots, respiration and rhizodeposition. The quantitative partitioning depends on genetic and environmental factors, as well as on the plant’s phenology. The first two or three pools represent most of the carbon input to the soil, depending on the ecosystem considered. Quantifying this partitioning is of major importance as roots, shoots, and rhizodeposition are constituted of several organic compounds differing in residence time. Moreover, besides being carbon inputs, these compounds modify in different ways their surroundings and might slow down or accelerate the cycling of carbon and nutrients.

One frequently used method to quantify this partitioning is the 13C labelling of atmospheric CO2, which allows to trace organic carbon in the soil-plant system. 13C labelling of CO2 can be either continuous, or as single or multiple pulses. It can be combined with gas measurements that estimate the priming effect induced by the plant inputs. This poster aims at synthesizing the literature about experimental quantification of carbon allocation to different plant pools, and comparing the methods used. It allows to identify the plant traits that predict this allocation. Moreover, we show the preliminary results of a continuous 13C labelling experiment that compares plant traits with carbon content variations in soil columns.

Whereas shoots and roots are relatively simple to quantify, rhizodeposition is not. It therefore represents a major incertitude when considering the carbon inputs to soil. Moreover, labile exudates represent an important part of rhizodeposition that induce an increase of older soil carbon mineralisation. Several experiments show that the loss of carbon induced by rhizodeposition priming effect is bigger than the input. These results are often associated to plants with high photosynthetic activity.

A trade-off is to be found when considering 13C labelling. Continuous labelling allows a quantification of net rhizodeposition which is independent of the phenology and that integrates the whole plant growth. On the other side, pulse labelling is easier and less expensive. Regarding priming effect quantification, gas flux measurements are precise but do not integrate the whole plant growth. Mass quantification through carbon content change do, but it requires very accurate estimations of these changes.