Geospatial distribution of soil organic carbon and soil pH within the cocoa agroecological zones of Ghana

• Published in: Geoderma 2021-03, Vol.386, p.114921, Article 114921
• Main Authors: Quaye, Amos Kojo, Doe, Eric Kofi, Attua, Emmanuel Morgan, Yiran, Gerald, Arthur, Alfred, Dogbatse, Jerome Agbesi, Konlan, Sampson, Nkroma, Yaw Dankwa, Addo, Divine
• Format: Article
• Language: English
• Subjects: Co-kriging; Cocoa districts; Cocoa soils; Coefficient of variability; Kriging; Precision agriculture; Soil fertility management; Soil health; Spatial autocorrelation
• ISSN: 0016-7061; 1872-6259
• DOI: 10.1016/j.geoderma.2020.114921

•A new spatial pattern of soil organic carbon and soil pH for the cocoa districts and agroecological zones of Ghana.•Unsuitably low cocoa soil organic and strong acidity.•A loss of soil organic carbon in a pattern that reflects historic boom-bust cycle of cocoa hot and cold spots transitions.•Homogeneity of soil organic carbon is moderate within a longer range than soil pH.•Covariates of soil pH like Fe and Al elongate the range of homogeneity in favour of soil pH. Geospatial distribution of soil organic carbon (SOC) and soil pH is an important component of soil fertility management in sustainable cocoa (Theobroma cacao L.) production. Spatial precision of cocoa soil fertility management in Ghana is rare due to paucity of spatially precise knowledge of farm soil properties. This study sought to provide spatially explicit knowledge of the underlying trends of SOC and pH in and across cocoa agroecological zones, covering 61 cocoa districts in Ghana. Soil samples collected at 20 cm depth from 739 newly established cocoa farms were analyzed using classical and geospatial statistical methods. We described the spatial pattern (clustering) of SOC and pH and predicted their spatial continuity from observed and unobserved locations. Results generally revealed unsuitably low SOC (0.39 ± 0.009 g/100 g) and strong acidity (5.35 ± 0.027). Univariately, SOC varied more continuously (58%) over a longer distance (1.424 km) than pH, which varied moderately around 55.7% over a shorter distance (0.598 km). Covariates improved the co-regionalized structure and homogeneity of the predictions. Thus, the spatial dependencies of SOC and pH were moderate but the risk of imprecision was higher for SOC than pH. This knowledge is crucial in our understanding of the geo-physicochemical phenomena controlling the spatial distribution of SOC and pH in the cocoa farms and districts across the agroecological zones. The findings underscored dwindling SOC and pH in line with historic boom-bust cycle of cocoa production hot and cold spots transitions. The findings are also indicative of the adverse implications of relying on current blanket fertilizer recommendations. Therefore, a change from the current agronomic practice of fertilizer use to one that recognizes the SOC and pH needs of specific areas, farms and cluster of districts is recommended.