Soil carbon and inferred net primary production in high- and low-intensity grazing systems on the New England Tableland, eastern Australia

Management of grazing lands for the accumulation of soil carbon stocks ([C.sub.S]) has been proposed as an effective way to reduce net greenhouse gas emissions from agriculture. However, there are conflicting reports on the effects of grazing management on soil carbon. Most comparisons have involved some combination of no grazing, rotational grazing and set stocking. In the present study we compared two adjacent commercial grazing systems, distinguished on the basis of inputs and livestock productivity, located on New England basaltic landscapes experiencing a cool temperate climate. The high-intensity (H) system sustains an average stocking rate of 18 dry sheep equivalents (dse) [ha.sup.-1], with a turnoff rate of 9 dse [ha.sup.-1] [year.sup.-1], with high levels of investment in assets, management and fertiliser. The low-intensity (L) system, with less intensive management and half the fertiliser of the H system, sustains a stocking rate of 9 dse [ha.sup.-1], with a turnoff rate of 3 dse [ha.sup.-1] [year.sup.-1], which is slightly higher than the regional average. Pasture biomass production was inferred (back-calculated) from stocking rates and animal feed requirements using published data. From the H and L systems, seven paired landscapes from valley floor to upper hillslopes and plateaux were selected. The seventh included a forest reserve. One hundred and eighty-six undisturbed soil cores (0-0.5 m depth) were assessed for bulk density, total C and N, particulate C and a range of plant nutrients. There were few differences in [C.sub.S], soil pH and nutrient levels between H and L grazing systems. Average [C.sub.S] (0-0.3 m) in pasture soils was 103 Mg [ha.sup.-1], but was higher in the forest soil at 190 Mg [ha.sup.-1]. Regression of [C.sub.S] versus soil mass was a satisfactory method of dealing with the bias introduced by the higher soil bulk density in perennial pasture systems compared with the forest. The similarity of [C.sub.S] in H and L pasture soils was despite inferred net primary production being 1.9-3.6 Mg C [ha.sup.-1] [year.sup.-1] greater in H than L systems, implying higher rates of C turnover in the former. The global warming potential of the inferred annual emissions of C[H.sub.4] and [N.sub.2]O in the H and L systems was equivalent to approximately 19% and 13% of the cycling atmospheric-plant C[O.sub.2] carbon respectively. Additional keywords: grazing animal emissions, particulate soil carbon.
Introduction Increasing soil carbon stocks ([C.sub.S]) via changes in management of agricultural land to both increase biomass production and introduce less exploitative practices is an attractive way to reduce net [...]