Litter decomposition, soil respiration and soil chemical and biochemical properties at three contrasting sites in karri (Eucalyptus diversicolor F. Muell.) forests of south-western Australia

Abstract Litter decomposition, soil respiration and soil chemical and biochemical properties were examined at three contrasting sites in karri (Eucalyptus diversicolor F. Muell.) forest of south‐western Australia. The study sites were: a recently clearfelled area (site CF2) which had been subjected to a slash regeneration burn following clearing; a pole‐stand regrowth forest about 40 years old which had been regularly burnt by cool, prescribed fires (site RB40); and a pole‐stand regrowth forest about 40 years old which had remained unburnt for many years (site UB40). Leaf litter of uniform composition lost 40–54% of its original dry weight after decomposing for 82 weeks on the forest floor. A composite exponential model, with separate decay functions for labile and more resistant litter components, described rate of weight loss better than a simple exponential decay model. Labile components of litter were released at similar rates at the three sites. Decomposition of resistant litter components was slower (half‐life = 271 weeks) at the recently clearfelled site than at the two pole‐stand sites (half‐lives = 119 and 149 weeks). The order in which nutrients were released from decomposing litter, Na > Cl > K > Mg > S > Ca > N > P, was similar at each site. The rate of release of the more mobile elements Na, Cl, K, Mg and S, was also similar at each site. Changes in the amounts of Ca, N and P in decomposing litter differed between the three sites and the differences were related to the amounts of these nutrients in surface soil at each site. Annual soil respiration decreased in the order site CF2 = site UB40 > site RB40. Seasonal variation in respired CO 2 was partly explained by variation in soil moisture and temperature. Soil carbohydrase activity at the recently clearfelled site was significantly lower than at the two well vegetated pole‐stand sites. The differences between sites in enzyme activities were related to differences in the amounts of organic C in surface soils of the three sites. The amount of organic C in surface soil (0–15 cm) was 25–36% lower at the recently clearfelled site than at the two well vegetated pole‐Stand sites. Site disturbance during clearing, and combustion of soil organic matter by the subsequent slash regeneration burn, probably account for part of this difference. However, reduced inputs of organic matter in litterfall, slower rates of surface litter breakdown and increased rates of microbial mineralization of soil organic matter on recently clearfelled areas may also contribute substantially to depletion of soil organic C.