The 0 to 20-cm surface layer of a sandy loam soil was sampled in early autumn from plots where straw had either been removed or incorporated annually for 22 years. Denitrification in whole soils, 1–2-mm wet-stable aggregates, clay and silt size fractions was determined by acetylene blocking during anaerobic incubation with excess nitrate. Thus available organic matter was the limiting factor. Samples were exposed to one or two freeze/thaw cycles, or used unfrozen. K2SO4-extractable carbon (C) was determined before and after CHCI, fumigation.

Freeze/thaw increased denitrification in whole soils and in aggregates. In aggregates and in whole soil without straw the increase in denitrification was similar following two freeze/thaw cycles, and well above the amount that could be fed by extractable soil C. In whole soils with straw addition, an extra denitrification increase occurred at first thaw only. This straw-induced denitrification surplus was matched by a decline in soil microbial biomass. For other samples and treatments, the freeze/thaw released C from additional organic matter sources.

The availability of C in clay for denitrification was twice that of silt-associated C. Straw disposal generally had no effect on the bioavailability of particle-bound C. In contrast to whole soils and aggregates, the availability of organic matter in clay and silt after one freeze/thaw cycle was only half that observed from unfrozen samples.

The effect of freeze/thaw on whole soils and aggregates may be to release organic matter available for denitrification by killing the microbial biomass and by disintegrating aggregates. However, the impact of freeze/thaw on completely dispersed samples such as clay and silt may be to promote the formation of granular structures (micro-aggregation) in which organic matter may become less accessible to denitrifiers.