Zeljka Zgorelec

NITRATE CONCENTRATION IN DRINKING WATER FROM WELLS AT THREE DIFFERENT LOCATIONS IN NORTHWEST CROATIA

The findings confirm that nitrate concentrations in water samples from wells in 3 different locations in Northwest Croatia are significantly different. The highest average nitrate concentration of 28.7 mg L−1 NO3− is related to location Koprivnica, which is the area of intensive agricultural production. At location Krizevci wells are located at suburban area and average nitrate concentration is 26.5 mg L−1 NO3−. Compared to this two values significantly lower average concentration of 4.6 mg L−1 NO3− is determined at location Kalnik in water samples without any influence of agriculture and urbanisation.

Reducing sampling intensity in order to investigate spatial variability of soil pH, organic matter and available phosphorus using co-kriging techniques. A case study of acid soils in Eastern Croatia

ABSTRACT Soil pH, hydrolytic acidity (HA), organic matter (OM) and plant available phosphorus (AP) are factors controlling the environmental-friendly soil management in agroecosystems. These parameters are highly variable in space. The objective of this work is to study spatial variability of pH, HA, OM and AP using several interpolation methods in Eastern Croatia. A total of 1004 (0–30 cm) soil samples were collected, and several univariate and multivariate interpolation performances were tested. The results showed that soils of the study area had high HA and AP, while pH and OM were low. The variogram analysis revealed different spatial structures among studied soil properties and demonstrate a need for variable-rate management. Soil pH and OM had lower spatial variability compared to AP and HA. Ordinary kriging was the most accurate method to estimate the studied variables. The incorporation of auxiliary variables increased the precision of the estimations for HA. Soil AP and OM showed different results for spatial prediction obtained by co-kriging. Overall, the incorporation of pH as auxiliary variable increased the prediction of the models. However, more co-variates should be incorporated in further models, in order to identify with more precision areas that need to be restored.

Soil Carbon Variability in Some Hungarian and Croatian Soils

Loss of soil carbon and carbon storage were studied in different soil types in the agroecological conditions of the Pannonian plain and in the Mediterranean region. Total carbon concentration, pH and C:N ratio under pastures, meadows, forests, vineyards, gardens and crop fields were determined. Soil samples were taken in 2010 at depths from 0–3, 3–10, 0–30, 30–60 and 60–90 cm depending on the location, region, soil type and type of ecosystem. Soil pH varied between strongly acid 4.3 measured in crop field on distric Stagnosols (0–30 cm) to weakly alkaline 7.6 in a crop field on Chernozems; C:N ratio varied from 8 measured on distric Stagnosols to 39 on Cambisol calcaric; Soil carbon concentration observed in this study varied from 3.3 g C/kg at Szentgal on an agricultural crop field on Eutric Cambisol, in the deep layer (60–90 cm; L4) to 107.2 g C/kg on Regosol on karst measured in the surface layer of a Mediterranean grassland (0–3 cm; L10).

Precise Soil Sampling Method for Agricultural and Environmental Research Based on New Circular Soil Probe Apparatus

Conventional soil sampling usually implemented in Croatia considers sample weight of 2 kg per 4-5 ha area, which means that representative sample in relation to soil mass up to 30 cm depth is presented through the ratio 1:10000000. New sampling method changes the ratio to 1:625000, thus increasing amount of sampled soil 16 times with assumption that such sample better describes investigated area. Moreover, new soil sampling probe can be used for precision farming purposes where the central point of the probe ring is positioned with precision of ±1 cm and represented with 4, 8 or 16 samples taken in 50 cm radius from the center. Soil probe prototype was tested on agricultural land of 4 ha area with total number of 200 samples. To justify application of new constructed probe, this study gives results of geostatistical analysis of spatial variability in soil pH values up to 30 cm depth. Ordinary kriging was used as interpolation method. Spatial structure of soil parameter was analyzed by calculating semivariograms and approximated by exponential model. Root-mean-square error (RMSE) of prediction was used as measure for the best model evaluation. Interpolation analysis of reduced number of samples per investigated area of 4 ha resulted in deviation in RMSE of 13,2 % when 50 % less samples were randomly introduced, and 41 % with 60 samples, compared to total of 200 samples. Using new soil sampling method for soil survey and applied geostatistical tools provides a solution for quantifying spatial variability of soil properties, possibilities for commercial activities and a way to introduce variable rate technology in agricultural input application which can optimize farm profitability through improving yield, reducing input costs and minimizing input losses to the environment.