GEOSTATISTICAL ASSESSMENT OF THE REGIONAL DISTRIBUTION OF PHOSPHATE SORPTION CAPACITY PARAMETERS (FE-OX AND AL-OX) IN NORTHERN BELGIUM

Soil samples were analysed for oxalate extractable iron, aluminium and phosphate content (Fe-ox, Al-ox and P-ox) within a 700 km(2) area in northern Belgium, using a regular grid of 301 sampling locations with basic distances in between of 1800 m, and local densifications of 900 and 450 m. Variograms of the specific parameters indicated spatial cor relation between neighbouring points with range upto 6000 m for Fe-ox and Al-ox. The analysis data for each sampling location were geostatis tically interpolated (block-kriging). The spatial distribution of Fe-o x was closely related to the origin of the specific soil: (light) allu vial regions coincided with (extremely) high Fe-ox contents, where coa rse sandy soils coincided with very low Fe-ox contents. Regions that a re classified as loamy sand or sandy loam contained much more Al-ox th an did coarse or fine sandy soils within the area. The average kriging standard deviations obtained were 20-25% for Al-ox and 40-50% for Fe- ox. The phosphate sorption capacity (PSC) was calculated as PSC = 0.5 (Al-ox + Fe-ox), The kriged PSC estimates were displayed as maps. The average kriging standard deviation (KSD) for the PSC kriged estimates was about 25%. Within one area, where the sampling grid was denser (90 0 m), lower kriging standard deviations of about 15% were obtained for the PSC. P-ox and the degree of P saturation (DPS) showed some spatia l correlation but with low ranges (750-1000 m). A (large) nugget effec t was also encountered for these parameters, that can be explained by taking the large human factor determining P-ox and DPS into account. T hese findings can be used to provide indicative maps on regional scale indicating the sensitivity of certain areas for rapid P saturation an d resulting eutrophication problems, especially in intensive animal hu sbandry regions.