Estimation of site index for Scots pine and Norway spruce stands in South Finland using site properties

1993

28 S.

zahlr. Lit. Ang.

Bandaufführung

2628

37354

Soil sampling and site index estimation of 415 Scots pine and 441 Norway spruce stands in southern Finland were carried out during 1980-85. Site index values were biased in relation to the stand age, especially in the spruce stands. This bias, negatively correlated with stand age, was removed to guarantee that sites with similar site properties had similiar site indices regardless of the stand age. Site index was correlated with site type, stoniness, topographic position, slope, humus layer and mineral soil thickness and soil texture. Effective temperature sum, slope, a variable for shallow soils (<30 cm), soil fine fraction (d<63 m) content and organic matter content of humus layer explained 37 % of the site index for the pine stands and effective temperature sum, slope, gravel content of the mineral soil and organic matter content of the humus layer, variables for clay contents of over 10 % and fine fraction contents of under 15 % and humus layer thickness explained 30 % of the site index for the spruce stands. The best chemical soil variables explaining site index in the pine stands were the mineral soil calcium content (kg/ha) and total nitrogen concentration (g/kg OM) of the humus layer. The best chemical predictors for the spruce stands wee nitrogen concentrations (g/kg OM) of the humus layer and the mineral soil. Variables describing soil potassium entered to the chemically oriented site index models with negative regression coefficients. Site index could be estimated better for the pine stands (R hoch 2 = 0.61) than for the spruce stands (R hoch 2 = 0.36) without any laboratory variables. These models contained effective temperature sum, site types, soil thickness, average stone rod penetration and thickness of the humus layer in the pine stands and the sum of effective temperature, site types and variables depicting high clay contents (>10 %) and low fine fraction contents (<15 %) in the spruce stands. Prediction errors of these models were 2.2 m for the pine and 2.8 m for the spruce stands. Site index estimates based on only site type and stoniness or paludification stage of a site, gave poorer results than the models using more site and soil variables. The present precision of site quality evaluation could thus be improved using additional field or laboratory characteristics.