TRACING RECENT VEGETATION CHANGES IN AUSTRALIA’S GRAZED WOODLANDS BY d13C ANALYSES OF SOIL ORGANIC MATTER

The process of “woody thickening” describes the invasion or proliferation of trees into grasslands and savannas. Such changes have been documented worldwide and are thought to be encouraged by European-style land management practices such as domestic cattle grazing and fire suppression. The question whether a change from savannas to a “thickened” system may serve as a carbon sink or source is still highly debated, particularly with respect to carbon storage as soil organic matter. While “thickening” can be a naturally-occurring process, e.g. caused by climate change, it is the human-induced “thickening” that is of importance with respect to the Kyoto protocol and carbon accounting. Therefore, it is fundamental to establish first, whether a) these systems have indeed changed from grassland to woodland and b) that this change occurred recently and coincides with modern management practices. Human-induced thickening is believed to be very widespread in the dry, grazed woodlands of eastern Australia. Our aim was to determine at selected sites whether a change from grassy (C4) to woody (C3) vegetation had occurred since the initiation of European agricultural practices (last 50-150 years) and how this change had affected soil carbon levels. At a site where anecdotal evidence suggested invasion of woody vegetation (Acacia sp.), samples were taken along a transect, spanning from grassland to transition zone to thickened woodland. We utilized d¹³C and ¹⁴C data of bulk and size-separated soil organic matter to establish the time course of vegetation change, the change in soil carbon stocks and the change in carbon allocation. Our results indicate a strong shift from grass (C4)- to wood (C3)-derived carbon in the thickened woodland. Furthermore, most of the “young”, C3-derived carbon is allocated in the particulate organic carbon (POC) fraction (2000-53mm). Along with greater above-ground biomass, there is also an apparent increase in soil carbon stocks in the thickened compared with the grassy site; however, this apparent increase in soil carbon lies within the sampling error and more samples are necessary to establish a statistical significance. Our results show that thickening can be confidentially traced using the d¹³C technique and that potential exists for increased soil carbon storage in thickened woodlands.