Sampling and Laboratory Errors in Forest Soil Analysis

Abstract

In connection with the soil survey of the national forest inventory plots in 1986–89, two sets of independent samples were collected on 14 sample plots in different parts of Finland. The organic and mineral soil (0–5, 5–20, and 20–40 cm) layers were sampled. All the samples were analyzed twice in the laboratory for exchangeable cations (BaCl₂ extraction) and total nitrogen, and total elemental concentrations of the organic layer by dry digestion (550°C). The relative error tended to increase as the concentrations decreased, especially when the concentrations were close to the detection limit of the analytical equipment. High relative errors at low concentrations led to error distributions skewed to the right, justifying the use of a median instead of an arithmetic mean of these errors. The precision of the laboratory analyses appeared to be better than that of the field sampling, especially in the case of the mineral soil layers. On average, the relative errors in both the laboratory and the field tended to increase towards deeper soil layers.

Keywords:

• Forest soil
• Soil sampling
• Precision
• Soil analysis

Acknowledgments