Net uptake of CH4 was measured in intact soil cores (6.4 cm dia, 12 cm deep) collected from an arable wheat field, from three sites left uncultivated for more than 110 years following arable cropping and from a permanent grassland with different mineral N treatments subdivided into four pH levels. Soil cores were incubated in sealed 1 litre jars at 250 degrees C for 48 h with a CH4-amended atmosphere of 10 mu l 1(-1) at the start of incubation. The decrease in CH4 concentration followed first-order-kinetics and by log-transformation individual uptake rates could be calculated for each treatment. Soil from a calcareous site (PH 7.4) under deciduous woodland (Broadbalk Wilderness wooded section) oxidized CH4 6 times faster than the arable plot (pH 7.8) with the highest activity in the adjacent Broadbalk Wheat Experiment (with uptake rates of - 80 and - 13 nl CH4 1(-1) h(-1), respectively). The CH4 uptake rate was only 20% of that in the woodland in an adjacent area that had been uncultivated for the same period but kept as rough grassland by the annual removal of trees and shrubs and, since 1960, grazed during the summer by sheep. It is suggested that the continuous input of urea through animal excreta was mainly responsible for this difference. Another undisturbed woodland area with an acidic soil reaction (pH 4.1) did not oxidize any CH4. On a permanent grassland site (Park Grass Continuous Hay Experiment), the plot without N fertilization showed a distinct pH effect: CH4 consumption decreased from -67 to -35 nl CH4 1(-1) h(-1) with decreasing pH in the range 6.3-5.6 and declined to zero between pH 5.6 and 5.1. Mineral N applied annually as (NH4)(2)SO4, at either 96 or 144 kg N ha(-1) for 130 years, completely inhibited CH4 oxidation, even where lime was applied to maintain a soil pH of about 6. By contrast, the long-term application of N as NaNO3 (96 kg N ha(-1) a(-1)) caused no decline in CH4 oxidation compared to unfertilized grassland at the same pH and, in some cases, caused a small increase. Withholding NH4-N for 3 years caused no significant recovery of CH4-oxidizing activity; withholding NO3-N caused a slight decline. Thus, land use (arable, cut grassland, grazed grassland or woodland), soil pH, N fertilizer inputs and form of N (NH4 or NO3) all have marked and interacting effects on the extent to which aerobic soil acts as a sink for CH4. The mechanisms through which the factors operate are not known but some possibilities are discussed. The results have important implications for the planning of land use and agricultural practices that will maximize the extent to which aerobic soils can act as a sink for CH4.
