Gains and losses in C and N stocks of New Zealand pasture soils depend on land use

Previous re-sampling of 31 New Zealand pasture soil profiles to 1 m depth found large and significant losses of C and N over 2-3 decades. These profiles were predominantly on intensively grazed flat land. We have extended re-sampling to 83 profiles, to investigate whether changes in soil C and N stocks were related to land use. Over an average of 27 years, soils (0-30 cm) in flat dairy pastures lost 0.73 +/- 0.16 Mg C ha(-1) y(-1) and 57 +/- 16 kg N ha(-1) y(-1) but we observed no significant change in soil C or N in flat pasture grazed by "dry stock" (e.g., sheep, beef), or in grazed tussock grasslands. Grazed hill country soils (0-30 cm) gained 0.52 +/- 0.18 Mg C ha(-1) y(-1) and 66 +/- 18 kg N ha(-1) y(-1). The losses of C and N were strongly correlated, and C:N declined significantly. Further, results reported to 60 and 90 cm show that the pattern of losses and gains extend beyond the IPCC accounting depth of 30 cm. Specific causes for the soil C and N changes are unknown, but appear to be related to land use. In general, the losses under dairying correspond to systems with greater stocking rates, fertiliser inputs and removal of C and N in exported products. Gains in hill country pastures may be due to long-term recovery from erosion and disturbance following land clearance. The unexpected and contrary changes of C and N in different pasture systems (initially thought to be at steady state) demonstrates the need for global and national-scale collection of robust data investigating soil biogeochemical changes, not only for grasslands but also for other land uses. Re-sampling of soils can constrain the directions and magnitude of soil C and N change associated with land use and management to underpin C and N inventories and correctly identify mitigation options. (c) 2010 Elsevier B.V. All rights reserved.