IN SEARCH OF THE ELUSIVE ACTIVE FRACTION OF SOIL ORGANIC-MATTER - 3 SIZE-DENSITY FRACTIONATION METHODS FOR TRACING THE FATE OF HOMOGENEOUSLY C-14-LABELED PLANT MATERIALS

To improve the understanding of nutrient cycling in soil there is a ne ed for development of methods to quantify biologically-meaningful frac tions of soil organic matter which turn over in the short or medium-te rm. Homogeneously C-14-labelled shoots from ryegrass grown at ambient (350 mu l l(-1)) and elevated (700 mu l l(-1)) CO2 concentrations were added to a loamy sand and incubated for up to 200 days. Three size-de nsity methods were tested in order to elucidate the breakdown of the p lant material. One approach involved density separation in Ludox TM40 (a colloidal silica suspension) but only included soil materials >150 mu m. The other two approaches in which sodium polytungstate was used as density agent included all solid and soluble soil material. One of these involved a size separation (at 100 mu m) prior to density separa tion, while the other was performed on whole soil. Density fractionati on in a centrifuge (10,000 g) without initial size-separation substant ially reduced the recovery of freshly-added plant material in the ligh t fraction. We assume that this was partly due to the loss of air entr apped in intact tissue during centrifugation, and partly due to intera ctions between small heavy particles and the large light plant materia l. Fractionation by size and density thus seems a more powerful approa ch for separating soil organic matter fractions than fractionation bas ed on density alone. Separation of finer textured materials (<100 mu m ) by density resulted in fractions with similar specific activity, ind icating that they did not differ greatly in their turnover rates. The changes with time in the specific activity of the fine fractions indic ated that they acted as sinks for microbial products, and only contrib uted slightly to the mineralization of the freshly-added C. The solubl e carbon was consistently the most C-14-enriched fraction and containe d a substantial amount of C-14 throughout the incubation. The large, l ight fractions consisted of identifiable plant residues and were enric hed in C-14 during the 200 day incubation. Subdivision of the large fr action by density resulted in fractions with considerably different in itial enrichment, presumably due to greater airfilled porosity in less decomposed or frayed materials. Losses of ''native'' soil carbon were small, compared with the analytical uncertainties, and thus the ident ification of active ''native'' soil fractions was hampered. Difference s in the decomposition patterns between ryegrass grown at ambient and elevated CO2 concentrations, measured by CO2 respiration after 10 days , were observed with the large (>150 mu m) light Ludox fractions. At t he end of the experiment no differences between plant material grown a t ambient and elevated CO2 concentrations were detected in earlier CO2 evolution or in the different soil organic matter fractions. Minerali zation of C from previously leached plant materials was considerably e nhanced by exposure to Ludox and retarded by exposure to sodium polytu ngstate.