Carbon storage in Inga-cacao agroforestry on degraded former coca-crop soils

Abstract

Former coca-crop lands in the Peruvian Amazon need restoration that rebuilds soil function and carbon pools. This study compared soil, litter and recovered-root carbon in degraded former coca-crop soil without agroforestry and Inga edulis-Theobroma cacao systems aged 2, 4 and 6 years. Three blocks provided 12 sampling units of 100 m². Soil carbon (0–20 cm) used organic matter and bulk density; litter and roots were sampled in 0.25 m² quadrats. Block-adjusted analysis and exact within-block tests detected condition effects for all pools after false-discovery-rate adjustment (adjusted P = 0.0069-0.0330). Total carbon was 13.06 ± 2.34 t C ha?¹ without agroforestry and 48.06 ± 3.44, 62.62 ± 3.87 and 59.77 ± 5.86 t C ha?¹ in the 2-, 4- and 6-year systems, respectively. A parallel first-order compartment model fitted the three pools and derived total carbon as their sum. The derived total curve had R² = 0.921 and a model-implied asymptote of 65.88 t C ha?¹; apparent response coefficients were 0.857 yr?¹ for soil, 0.387 yr?¹ for litter and 0.242 yr?¹ for roots. Agroforestry conditions contained more measured carbon than the non-agroforestry condition, but total carbon did not differ among agroforestry ages. The fitted coefficients describe a cross-sectional condition-age gradient rather than measured annual sequestration. Estimates exclude aboveground biomass and soil below 20 cm.