Because the widely recognized high complexity inherent to rhizosphere systems imposes many methodological limitations, rhizosphere studies have often been performed in laboratories under highly artificial conditions. The relevance of the results from these studies to the real interactions taking place in field ecosystems has long been an open issue. Addressing this issue has been one of my research objectives since the 1980s. For instance, discerning rhizosphere carbon fluxes have been methodologically challenging. Carbon dioxide released by a system of living roots and soil has two main origins: (1) rhizosphere respiration (the sum of root respiration and rhizo-microbial respiration) of plant-derived carbon, and (2) microbial respiration of original soil C. Since the 1960s, continuous 14C labeling or pulse labeling in laboratories has been used in studies of the two sources of CO2. However, because of safety concerns with radioactive materials, 14C-labeling is rarely used in experiments longer than one or two months. Later, non-radioactive 13C-enriched CO2 has been used in place of 14CO2. However, the extremely high cost of the 13C-enriched CO2 makes it unsuitable for large experiments or for continuous labeling. In order to overcome some of the difficulties mentioned above, I have developed a natural 13C tracer method for such tasks. This natural 13C natural tracer method takes the advantage of the difference in the 13C:12C isotope ratio (often reported in d13C values) between plants with the C3 photosynthetic pathway and plants with the C4 pathway and the isotopic difference between soil organic carbon produced by the two kinds of plants. This natural 13C tracer method has several advantages over existing labeling methods, one of which is the elimination of disturbance caused by labeling. This method has been the most valuable tool for my rhizosphere studies in the past decade.
Although this natural tracer method is safe and inexpensive, it requires a large difference in 13C natural abundance between the soil-derived C and the root-derived C, which restricts its use to two kinds of plant-soil couplings: C3 plants grow in soils developed under C4 plant-dominated vegetation ("C4 soils"); or C4 plants grow in soils developed under C3 plant-dominated vegetation ("C3 soils"). This method, therefore, cannot be used with most natural plant-soil couplings, which are typically C3 plants in C3 soils or C4 plants in C4 soils. In order to apply this stable isotope method to natural plant-soil couplings, I have developed a continuous labeling system (Cheng and Dijkstra 2007) using commercially available, inexpensive CO2 produced from natural gas (13C-depleted, having δ13C values from -40‰ to -55‰). This continuous 13C-labeling system allows us to separate rhizosphere C fluxes that would otherwise be difficult or impossible to measure both in the greenhouse and in the growth chamber, even in field ecosystems in the immediate future. It also avoids the use of radioactive 14C or expensive 13C-enriched CO2 like in most existing labeling methods, and clearly has more applications in other areas of research. The potential of this continuous labeling approach is yet to be realized.