Soil organic carbon in density fractions of tropical soils under forest – pasture – secondary forest land use changes

Authors

  • S. Paul,

    Corresponding author
    1. aBuesgen Institute, Soil Science of Tropical and Subtropical Ecosystems, Georg-August University of Goettingen, Buesgenweg 2, 37077 Goettingen, and bBuesgen Institute, Soil Science of Temperate and Boreal Ecosystems, Georg-August University of Goettingen, Buesgenweg 2, 37077 Goettingen, Germany
      S. Paul. E-mail: spaul1@gwdg.de
    Search for more papers by this author
  • a E. Veldkamp,

    1. aBuesgen Institute, Soil Science of Tropical and Subtropical Ecosystems, Georg-August University of Goettingen, Buesgenweg 2, 37077 Goettingen, and bBuesgen Institute, Soil Science of Temperate and Boreal Ecosystems, Georg-August University of Goettingen, Buesgenweg 2, 37077 Goettingen, Germany
    Search for more papers by this author
  • and a H. FLESSA b

    1. aBuesgen Institute, Soil Science of Tropical and Subtropical Ecosystems, Georg-August University of Goettingen, Buesgenweg 2, 37077 Goettingen, and bBuesgen Institute, Soil Science of Temperate and Boreal Ecosystems, Georg-August University of Goettingen, Buesgenweg 2, 37077 Goettingen, Germany
    Search for more papers by this author

S. Paul. E-mail: spaul1@gwdg.de

Summary

Our knowledge of effects of land use changes and soil types on the storage and stability of different soil organic carbon (SOC) fractions in the tropics is limited. We analysed the effect of land use (natural forest, pasture, secondary forest) on SOC storage (depth 0–0.1 m) in density fractions of soils developed on marine Tertiary sediments and on volcanic ashes in the humid tropics of northwest Ecuador. The origin of organic carbon stored in free light (< 1.6 g cm−3) fractions, and in two light fractions (LF) occluded within aggregates of different stability, was determined by means of δ13C natural abundance. Light occluded organic matter was isolated in a first step after aggregate disruption by shaking aggregates with glass pearls (occluded I LF) and in a subsequent step by manual destruction of the most stable microaggregates that survived the first step (occluded II LF). SOC storage in LFs was greater in volcanic ash soils (7.6 ± 0.6 Mg C ha−1) than in sedimentary soils (4.3 ± 0.3 Mg C ha−1). The contribution of the LFs to SOC storage was greater in natural forest (19.2 ± 1.2%) and secondary forest (16.6 ± 1.0%) than in pasture soils (12.8 ± 1.0%), independent of soil parent material. The amount of SOC stored in the occluded I LF material increased with increasing silt + clay content (sedimentary soils, r = 0.73; volcanic ash soils, r = 0.58) and aggregation (sedimentary soils, r = 0.52; volcanic ash soils, r = 0.45). SOC associated with occluded I LF, had the smallest proportion of new, pasture-derived carbon, indicating the stabilizing effect of aggregation. Fast turnover of the occluded II LF material, which was separated from highly stable microaggregates, strongly suggested that this fraction is important in the initial process of aggregate formation. No pasture-derived carbon could be detected in any density fractions of volcanic ash soils under secondary forest, indicating fast turnover of these fractions in tropical volcanic ash soils.

Ancillary