Main content area

Conversion of forest to agriculture in Amazonia with the chop-and-mulch method: Does it improve the soil carbon stock?

Perrin, Anne-Sophie, Fujisaki, Kenji, Petitjean, Caroline, Sarrazin, Max, Godet, Mathieu, Garric, Bernard, Horth, Jean-Claude, Balbino, Luiz Carlos, Filho, Austrelino Silveira, de Almeida Machado, Pedro Luiz Oliveira, Brossard, Michel
Agriculture, ecosystems & environment 2014 v.184 pp. 101-114
Urochloa ruziziensis, agroecosystems, biomass, carbon, carbon sinks, corn, crop rotation, cropland, deforestation, direct seeding, discing, forests, grasses, grassland soils, grasslands, land management, landforms, mulches, no-tillage, plant establishment, soil compaction, soybeans, Amazonia, French Guiana
Fire-free forest conversion with organic inputs as an alternative to slash-and-burn could improve agro-ecosystem sustainability. We assessed soil carbon mass changes in a sandy–clayey and well-drained soil in French Guiana after forest clearing by the chop-and-mulch method and crop establishment. At the experimental site of Combi, native forest was cut down in October 2008; woody biomass was chopped and incorporated into the top 20cm of soil. After about one year of legume and grass cover, three forms of land management were compared: grassland (Urochloa ruziziensis), maize/soybean crop rotation with disk tillage and in direct seeding without tillage. There were four replicates. We measured 14.16kgm−2 of carbon in 2mm-sieved soil down to 2m depth for the initial forest. Forest clearing did not induce significant soil compaction; neither did any specific agricultural practice. In converted soils, C stocks were measured in the 0–30cm layer after each crop for three years. Carbon mass changes for soil fractions <2mm (soil C stock) and >2mm (soil C pool) in the 0–5, 5–10, 10–20 and 20–30cm soil layers were assessed on an equivalent soil mass basis. One year and 1.5 years after deforestation, higher C stocks (+0.64 to 1.16kgCm−2yr−1) and C pools (+0.52 to 0.90kgCm−2yr−1) were measured in converted soils, compared to those of the forest into the top 30cm of soil. However, the masses of carbon in these converted soils declined later. The highest rates of carbon decrease were measured between 1.5 and 2 years after forest conversion in the <2mm soil fraction, from 0.46kgCm−2yr−1 (in grassland soils) to 0.71kgCm−2yr−1 (in cropland under no tillage). The carbon pool declined during the third year at rates of 0.41kgCm−2yr−1 (cropland under disk tillage) to 0.76kgCm−2yr−1 (grassland soils). Three years after forest conversion, C masses in the top 30cm of soils for grassland showed similar values than for forest. In comparison, the carbon stock in cropped soils managed under no tillage in direct seeding (without mulch) was significantly 17% and 16% lower than in forest and grassland soils, respectively. None of the studied agricultural practices succeeded in accumulating carbon from the chopped forest biomass.