Organic matter contained in soil aggregates from a tropical chronosequence: correction for sand and light fraction

A cultivation sequence of Ultisols (sandy loam) from the Amazon Basin of Peru was physically fractionated using wet sieving followed by heavy liquid separation and total organic C and N analysis. Soils were sampled to an equivalent plow depth (22.5 cm) in a cultivation chronosequence containing the following treatments: an old growth forest (75 plus years), a 13-year fallowed forest, a 3-year continuous cultivation corn-soybean rotation following a 10-year forest fallow and a 16-year intensively managed corn field. The soil was air dried and re-wetted (misted) to field capacity before shaking for 10 min and wet sieving into five size classes: <0.02, 0.02–0.05, 0.05–0.212, 0.212–1.0 and >1.0 mm. The weight of soil, total organic C and N, and percent sand was determined on each size class. Subsamples of each size class were gently stirred in sodium metatungstate (specific gravity = 1.8), left overnight and the floating material removed by aspiration. The removed material was filtered through pre-ashed glass fiber filters and analyzed for total organic C and N. The sand content of size classes ranged from 45 to 95%. Light fraction C comprised 24, 28, 14, 5 and 2% of the total C of the size classes ranging from the largest to smallest particle sizes. Total C concentrations were highest in the largest and smallest particle size classes. Within any size class, all treatments had similar C concentration except for the <0.02-mm particles where the 75-year forest had lower concentrations. When corrected for sand content (i.e., expressed on a silt plus clay basis) and with light fraction C subtracted, the organic C content was, with few exceptions, not different among size classes or treatments. However, when all treatments within a size class were averaged, the 0.020–0.053-mm class had a higher concentration of C. We applied methods that allowed correction of soil organic matter in aggregate size classes and suggest the importance of such corrections when interpreting aggregate organic matter data. Trends in organic matter concentrations were quite different when fractions were corrected for light fraction and sand content compared to when they were not corrected.