Application of regression-kriging and sequential Gaussian simulation for the delineation of forest areas potentially suitable for liming in the Jizera Mountains region, Czech Republic

Abstract

Acidification due to acid atmospheric deposition affected strongly soils of many forested areas with long-term consequences. Liming is often used for amelioration, however, detailed assessment of soil and stand conditions is necessary to avoid possible undesirable effects. The aim of this paper is to apply and compare two spatial modelling methods to delineate probability that the criteria for potential liming in the Jizera Mountains region strongly impacted by acidification are met and to assess the uncertainty.Soil characteristics from the last ten years were evaluated. Specific stands were excluded from consideration (peats, too skeletic stands, waterlogged stands, protected areas). Soils have low pH, low base saturation, and low content of base cations, so that most soil criteria required for potential liming approval are met. Only the ratios C/N and C/P in organic horizons fluctuate around thresholds of 20 and 250, respectively, and their distribution is thus crucial for decision making on liming. Two methods were tested: i) indicator regression-kriging with elevation as a covariate to map the probability that both criteria are met at once, ii) sequential Gaussian simulation to map the probabilities separately and then combining them by multiplying the probabilities. While both methods yielded similar maps of probabilities that the criteria are met, the second approach provided a more diverse map of error distribution, which may play an important role in the decision-making process.