Determination of sampling volumes needed for representative analysis of alpine soils

Abstract

Alpine soils are characterized by a high variability of chemical and physical parameters. The spatial inhomogeneity of a specific parcel must be regarded as a nuisance factor for representative soil sampling. The only way to reduce the effect of the variation between the samples is to combine several samples taken within an area of a few square meters. The presented approach shows how much soil material has to be taken in order to obtain representative values of soil properties such as soil density, the proportion of soil skeleton and fine earth, organic C, N, and the root content. Bulk density can be determined with an error ≤ 10% by sampling c. 6300 cm 3 of soil within an area of 6 × 4.5 m for 0-20 cm sampling depth. With the same volume, the accuracy of soil skeleton measurement is ≤ 20%. In our case study, the determination of organic C in the fine earth of the whole soil profile needs a sample volume of 15,800 cm 3 in order to reduce the error tolerance below ± 10%. In 0-20 cfm soil depth, the error tolerance of ± 10% is achieved with c. 4000 cm 3 for N and c. 6300 cm 3 for organic C. Nearly 16,000 cm 3 were needed to determine the soil root mass in the top 20 cm of a soil with an accuracy of ± 10%. In order to yield reasonable results, large sampling volumes are needed for soils in alpine areas. These volumes depend on the parameters to be investigated. Our investigation makes an initial attempt at optimizing the sampling strategy within an area of a few square meters in order to overcome this nuisance factor variation.