Investigations into forest soils face the problem of the high level of spatial variability that is an inherent property of all forest soils.In order to investigate the effect of changes in residue management practices...Investigations into forest soils face the problem of the high level of spatial variability that is an inherent property of all forest soils.In order to investigate the effect of changes in residue management practices on soil properties in hoop pine(Araucaria cunninghamii Aiton ex A.Cunn.)plantations of subtropical Australia it was important to understand the intensity of sampling effort required to overcome the spatial variability induced by those changes.Harvest residues were formed into windrows to prevent nitrogen(N)losses through volatilisation and erosion that had previously occurred as a result of pile and burn operations.We selected second rotation(2R)hoop pine sites where the windrows(10-15 m apart)had been formed 1,2 and 3 years prior to sampling in order to examine the spatial variability in soil carbon(C) and N and in potential mineralisable N(PMN)in the areas beneath and between(inter-)the windrows.We examined the implications of soil variability on the number of samples required to detect differences in means for specific soil properties, at different ages and at specified levels of accuracy.Sample size needed to accurately reflect differences between means was not affected by the position where the samples were taken relative to the windrows but differed according to the parameter to be sampled.The relative soil sampling size required for detecting differences between means of a soil property in the inter-windrow and beneath-windrow positions was highly dependent on the soil property assessed and the acceptable relative sampling error.An alternative strategy for soil sampling should be considered,if the estimated sample size exceeds 50 replications.The possible solution to this problem is collection of composite soil samples allowing a substantial reduction in the number of samples required for chemical analysis without loss in the precision of the mean estimates for a particular soil property.展开更多
基金Project supported by a scholarship grant from the Cooperative Research Centre for Sustainable Production Forestry,Australia.
文摘Investigations into forest soils face the problem of the high level of spatial variability that is an inherent property of all forest soils.In order to investigate the effect of changes in residue management practices on soil properties in hoop pine(Araucaria cunninghamii Aiton ex A.Cunn.)plantations of subtropical Australia it was important to understand the intensity of sampling effort required to overcome the spatial variability induced by those changes.Harvest residues were formed into windrows to prevent nitrogen(N)losses through volatilisation and erosion that had previously occurred as a result of pile and burn operations.We selected second rotation(2R)hoop pine sites where the windrows(10-15 m apart)had been formed 1,2 and 3 years prior to sampling in order to examine the spatial variability in soil carbon(C) and N and in potential mineralisable N(PMN)in the areas beneath and between(inter-)the windrows.We examined the implications of soil variability on the number of samples required to detect differences in means for specific soil properties, at different ages and at specified levels of accuracy.Sample size needed to accurately reflect differences between means was not affected by the position where the samples were taken relative to the windrows but differed according to the parameter to be sampled.The relative soil sampling size required for detecting differences between means of a soil property in the inter-windrow and beneath-windrow positions was highly dependent on the soil property assessed and the acceptable relative sampling error.An alternative strategy for soil sampling should be considered,if the estimated sample size exceeds 50 replications.The possible solution to this problem is collection of composite soil samples allowing a substantial reduction in the number of samples required for chemical analysis without loss in the precision of the mean estimates for a particular soil property.
