Abstract

An incubation study was conducted to test the effects of decomposing clover tops (added at 0, 6·2 or 12·5 mg organic matter/g soil) and earthworm activity on the contribution of carbohydrate C to the stability of aggregates in a sandy soil. Soils incubated with and without earthworms were separated into surface-casts and bulk soil, and the amounts of water-stable aggregates >1 mm surviving slow and rapid rewetting (when air-dry) in these soil separates were determined. Organic C and acid- and water-extractable carbohydrate C concentrations were determined in the aggregates and bulk soil. The treatments of 6·2 and 12·5 mg organic matter/g soil increased the >1 mm aggregation of the bulk soil by more than 2·2- and 2·8-fold, respectively, compared with the non-amended soils. With the addition of earthworms, there were increases from 1·7- to 1·8-fold only in aggregates surviving slow rewetting. The acid- and water-extractable carbohydrate C contents of aggregates >1 mm in the bulk and surface-cast soils were generally not greater than the carbohydrate C in the bulk soil. Generally, the carbohydrate C fractions were also not increased in the more stable aggregates (rapidly rewet) compared with the weaker aggregates (slowly rewet). Carbohydrate C in bulk soil was generally (P < 0·05) correlated with the amounts of aggregates surviving each rewetting treatment (r > 0·71, P < 0·01). In contrast, greater amounts of carbohydrate in aggregates surviving slow rewetting were not correlated (r < -0·45, P > 0·05), with a greater proportion of these aggregates resisting disruption when the soils were rapidly rewet (except for acid-extractable carbohydrate C; r = -0·84, P < 0·05). These results cast doubt on the usefulness of correlations in assessing the contribution of carbohydrate C to aggregation. The amounts of carbohydrate materials in the soil appeared to have little influence on aggregation, probably because the location of bonding compounds in the soil pore matrix is more critical.