Naslov
Delineation of soil texture suitability zones for soybean cultivation: a case study in continental Croatia

Autori
Radočaj, Dorijan ; Jurišić, Mladen ; Zebec, Vladimir ; Plaščak, Ivan

Izvornik
Agronomy (2073-4395) 10 (2020), 6; 823, 22

Vrsta, podvrsta i kategorija rada
Radovi u časopisima, članak, znanstveni

Ključne riječi
automation ; classification ; interpolation ; GIS ; python script ; kriging ; split-sample validation

Sažetak
Soil texture is a vital criterion in most cropland suitability analyses, so an accurate method for the delineation of soil texture suitability zones is necessary. In this study, an automated method was developed and evaluated for the delineation of these zones for soybean cultivation. A total of 255 soil samples were collected in the Continental biogeoregion of Croatia. Three methods for interpolation of clay, silt and sand soil content were evaluated using the split-sample method in five independent random repetitions. An automated algorithm for soil texture classification based on the United States Department of Agriculture (USDA) in 12 classes was performed using Python script. Suitability classes for soybean cultivation per soil texture class were determined according to previous agronomic and soybean land suitability studies. Ordinary kriging produced the highest accuracy of tested interpolation methods for clay, silt and sand. Highly suitable soil texture classes for soybean cultivation, loam and clay loam, were detected in the northern part of the study area, covering 5.73% of the study area. The analysis of classification results per interpolation method indicated a necessity of the evaluation of interpolation methods as their performance depended on the normality and stationarity of input samples.

Izvorni jezik
Engleski

Znanstvena područja
Interdisciplinarne tehničke znanosti, Poljoprivreda (agronomija), Interdisciplinarne biotehničke znanosti

POVEZANOST RADA