TY - JOUR
T1 - Spatial interpolation of vertisol physico-chemical properties through ordinary kriging in south-eastern Zimbabwe
AU - Chitiga, Muketiwa
AU - Soropa, Gabriel
AU - Dube, Trevor
AU - Sengera, Progress
N1 - © 2023 Southern African Plant & Soil Sciences Committee.
PY - 2023/3/13
Y1 - 2023/3/13
N2 - A decline in sugarcane yields from Block G (∼400 ha) of GreenFuel’s Chisumbanje Estate in south-eastern Zimbabwe prompted a study on spatial variability of the soil physico-chemical parameters in the 2017–2018 farming season. Using a systematic gridding schema, 123 geo-located soil samples from 16 sub-blocks were extracted from a depth of 0.3 m and analysed for various parameters. The results showed that pH was slightly acidic to alkaline (6.13–7.95) and had the lowest variability (CV ±15%). Low (N, P, Fe and Zn) and adequate (K and Cu) levels of mineral nutrients were also observed. However, some nutrients were highly variable, with P, Fe and Zn revealing the largest variability (CV ±35%). Semivariogram analysis showed a weak (clay, soil organic matter and N), moderate (pH, P, Mn and Zn) or strong (K, Fe and Cu) spatial correlation between the measured and predicted values. To optimise soil fertility and reduce variations, it is recommended that the estate employ variable-rate fertiliser application technologies, nutrient-monitoring systems, and precision satellite-based guidance for machinery. The study generated soil fertility maps using ordinary kriging; this will serve as a future reference for soil test-based fertiliser recommendations, soil fertility monitoring, and identifying spatial variability on the estate.
AB - A decline in sugarcane yields from Block G (∼400 ha) of GreenFuel’s Chisumbanje Estate in south-eastern Zimbabwe prompted a study on spatial variability of the soil physico-chemical parameters in the 2017–2018 farming season. Using a systematic gridding schema, 123 geo-located soil samples from 16 sub-blocks were extracted from a depth of 0.3 m and analysed for various parameters. The results showed that pH was slightly acidic to alkaline (6.13–7.95) and had the lowest variability (CV ±15%). Low (N, P, Fe and Zn) and adequate (K and Cu) levels of mineral nutrients were also observed. However, some nutrients were highly variable, with P, Fe and Zn revealing the largest variability (CV ±35%). Semivariogram analysis showed a weak (clay, soil organic matter and N), moderate (pH, P, Mn and Zn) or strong (K, Fe and Cu) spatial correlation between the measured and predicted values. To optimise soil fertility and reduce variations, it is recommended that the estate employ variable-rate fertiliser application technologies, nutrient-monitoring systems, and precision satellite-based guidance for machinery. The study generated soil fertility maps using ordinary kriging; this will serve as a future reference for soil test-based fertiliser recommendations, soil fertility monitoring, and identifying spatial variability on the estate.
KW - Chisumbanje Estate
KW - geostatistical interpolation
KW - physico-chemical parameters
KW - semivariogram
KW - soil fertility map
KW - soil pH
UR - http://www.scopus.com/inward/record.url?scp=85150599915&partnerID=8YFLogxK
U2 - 10.1080/02571862.2023.2168077
DO - 10.1080/02571862.2023.2168077
M3 - Review article
AN - SCOPUS:85150599915
SN - 0257-1862
VL - 40
SP - 46
EP - 57
JO - South African Journal of Plant and Soil
JF - South African Journal of Plant and Soil
IS - 1
ER -