Carbon Sequestration and Soil Fertility Management in Sandy and Clayey Soils Revealed by Over Four Decades of Long-Term Field Experiments in Thailand

Long-term field experiments on the soil C cycle are essential for understanding C dynamics in agricultural soils. However, such studies are limited in the humid tropics. This study quantifies the effects of chemical fertilizer application, organic matter application, and their combinations on soil C sequestration. Data was obtained from 4 sites of the 45-year long-term field experiments in Thailand. Furthermore, a structural equation model (SEM) was employed to visualize the relationships among organic matter application, soil carbon, basic chemical properties, and cassava yield.

Compared to the control without any application, soil carbon sequestration was 2.0 ± 2.1 and 2.8 ± 2.0 Mg C ha⁻¹ (0.2 m depth) for chemical fertilizer and crop residue incorporation, respectively. The largest soil C sequestration occurred when chemical fertilizers were combined with organic matter application. Specifically, when chemical fertilizer was combined with crop residue incorporation or compost application, soil C sequestration reached 5.6 ± 3.1 and 10.1 ± 6.5 Mg C ha⁻¹ (0.2 m depth), respectively. These findings underscore the importance of C contributions from crop biomass and direct C inputs from organic matter.

SEM showed that the effects of chemical fertilizer and organic matter application on soil C concentration in clayey soils were predominantly observed in the 0–0.2 m and 0.2–0.4 m surface layers. Conversely, treatment effects were significant in sandy soils at all depths up to 1.0 m. The increase in soil C in sandy soils also significantly improved basal soil fertility, such as soil pH, available phosphorus, and exchangeable potassium, resulting in higher cassava yields. In contrast, no significant relationship was found between soil C concentration and cassava yield in clayey soils.

Currently, soil C dynamics models for agricultural lands in low-latitude regions, such as Southeast Asia, are primarily based on databases from high-latitude areas (e.g., the RothC model). The findings from this study are expected to contribute to developing tropical-specific C dynamics models for agricultural lands in low-latitude regions. Furthermore, the standard set by the Intergovernmental Panel on Climate Change (IPCC) for calculating soil carbon sequestration (30 cm from the surface) may be insufficient for sandy soils, highlighting the importance of evaluating carbon sequestration at deeper soil layers.