Digital mapping of soil properties with application for sustainable intensification of agriculture in Cameroon

Although soil survey is required to guide agricultural land use planning and management in Cameroon, attention on it has been limited, thus creating gaps in soil resource inventory. However, with the availability of well compiled legacy soil data, nationwide reliable thematic soil properties maps, with high accuracy can be produced using digital soil mapping tools, and effectively used to guide decision making in agroecosystems. This study aimed at investigating the applicability of digital soil mapping (DSM) methods on legacy soil data for inventorying soil resources at 1-ha spatial resolution in Cameroon. Specific objectives were to (i) rescue, compile and harmonize the soil profile database; (ii) assess and map the distribution of some key soil properties (pH, Ca, Mg, K, SOC, N, P, CEC, Sand, Silt, Clay, and textural classes); (iii) assess the countrywide distribution of soil organic carbon stock (SOCS) down to 100 cm depth, and (iv) appraise the DSM methodology for applications in sustainable intensification of cocoa production in Cameroon. The spatial distribution of soil properties was predicted using machine learning algorithms at internationally recognized standard soil depths (0 – 5 cm, 5 – 15 cm, 15 – 30 cm, 30 – 60 cm, 60 – 100 cm, and 100 – 200 cm). The harmonized soil profile database (Camsodat 0.1) contains currently about 5250 layers which are unevenly distributed over 1600 georeferenced positions across the country. Prediction models estimated the national distribution of soil properties, with good performances (R2 up to 68%). Terrain and climate attributes were the most relevant environmental factors to predict soil properties distribution across the Country. The national distribution of SOCS was consistent with the pattern of agroecological zones. In total, about 5.67 Pg of SOCS is stored in the first 100 cm depth of soil in Cameroon, with 50% in the top 30 cm depth. Land evaluation revealed pieces of soils suitable for cocoa-production with varying levels of suitability and requiring site-specific management approaches. Generally, for cocoa production, 9%, of the area is highly suitable, 78% moderately suitable, and 13% marginal suitable. Five (5) soil properties mostly affecting cocoa productivity in decreasing order of importance are OC > Clay > Mg > pH > Silt. Although there are some gaps in soil survey across Cameroon, DSM showed a high potential to close such gaps, and the versatility in determining SOCS down to 100 cm depth and applicability in soil suitability analysis for intensification of cocoa production in Cameroon.