engleski

Sampling density and soil phosphorus mapping on various field scales. What this means for farmers?

The objective of this work is to study the spatial variability of soil available phosphorus (AP) on agricultural field scales in Croatia using geostatistical tools. The application of this technique is important to understand the spatial pattern of soil AP and identify areas where the deficit of this nutrient is high. Several datasets were used from different areas, with different sampling density: (100) 100 x 100 m (183 samples, Anthrosols), (50) 50 x 50 m (84 m, Stagnosols), (225) 225 x 225 m (404 samples, Gleysols), (12.5) 12.5 x 15.5 m (200 samples, Stagnosols). The results showed a highest concentration of AP in the following order: 12.5 (208 mg/kg) > 225 (208 mg/kg) > 50 (198 mg/kg) > 100 (30 mg/kg). Spherical models were the best fitted for 12.5, 225 and 100 datasets) and the exponential models for 50 dataset). Interpolation accuracy was evaluated using the Root Mean Square (RMSE) index. The highest accuracy (lowest RMSE) were noted at the 100 sampling density dataset (RMSE 4.49), followed by 50 (RMSE 25.97), 12.5 (RMSE 32.66) and 225 dataset (46.08). Nevertheless, the RMSE/mean AP concentration ratio in a given field represent 13.1%, 14.9%, 15.4% and 22.1% of AP for 50, 100, 12.5 and 225 dataset, respectively. Laboratory costs for sample preparation and AP determination were 3.5 euro/ha, 8 euro/ha, 32 euro/ha and 128 euro/ha for 225, 100, 50 and 12.5 datasets. According the precision and laboratory costs, the farmer’s best choice for adoption of precision agriculture seems to be a sampling in regular grid 100 x 100 m.

kriging ; sampling density ; mapping ; farmer cost

nije evidentirano