Variation in soil carbon in pine plantations and implications for monitoring soil carbon stocks in relation to land-use change and forest site management in New Zealand

Various methods were assessed to determine the impact of forest harvesting on mineral soil carbon (C) in Pinus radiata stands at Puruki catchment in the central North Island of New Zealand. Previous work at this site reported a reduction in soil carbon of 3.6 Mg ha(-1) in the top 0.1 m over one 23-year rotation of pine on improved pasture [Beets, P.N., Oliver, G.R., Clinton, P.W., 2002. Soil carbon protection in podocarp/hardwood forest, and the effects of conversion to pasture and exotic pine forest. J. Environ. Pollut. 116, S63-S73]. In this study, by sampling before and after harvesting, a further reduction of 3.1 Mg ha(-1) (significant at P = 0.05) in mineral soil carbon stocks to 0.1 nt depth was found to be associated with harvesting disturbance. However, less intensive sampling was unable to detect any significant harvesting effects in depths down to 2 m. Variation between plots was large, and it was estimated that to detect a 10% change in carbon content with 95% confidence 19 permanent plots would be required for 0-0.1 m depth and 40 permanent plots required to detect changes to I m depth. Soil C content to I m depth in Puruki Rua subcatchment was estimated to average between 143 and 164 Mg ha(-1). In two paired-site studies comparing pasture with second rotation pine, the difference between land-uses in mineral soil C content to 0.1 m depth was 19.6 Mg ha(-1) in volcanic soil at Kaingaroa and 8.5 Mg ha(-1) in high clay activity soil at Ngaumu. Significant differences in mineral soil C between pine and pasture was found with cumulative depth down to 2 m, well below the rooting depth of pasture, which suggests that historic factors overshadowed influences of the current land-use and can account for the difference in soil carbon. At Kaingaroa, variability in soil carbon content was high at all depths measured, and to detect a 10% difference in soil carbon content with 95% confidence approximately 60 pits measured to I m depth would be required. (C) 2004 Elsevier B.V. All rights reserved.