INFLUENCE OF BODY-SIZE OF SOIL FAUNA ON LITTER DECOMPOSITION AND N-15UPTAKE BY POPLAR IN A POT TRIAL

We combined the advantages of litter-basket and microcosm techniques t o study the decomposition of N-15 tagged Douglas-fir litter and uptake of labelled N by poplar plants. This was done in microcosms in the gr eenhouse under artificial summer and winter conditions. The microcosms contained litter baskets that excluded certain soil fauna based on bo dy size. Litter baskets of varying mesh-size containing N-15-enriched defaunated litter were embedded in a simulated coniferous forest floor of unlabelled organic and mineral soil layers. Four litter-basket tre atments were established: 1, 5 mm mesh allowing entry of all fauna, bu t macropredators were excluded; 2, 5 mm mesh, but macropredators were included; 3, 0.5 mm mesh with entrance for mesofauna only; and 4, 48 m u m mesh allowing only the microfauna to enter the baskets. After 40 w eeks, the development of faunal populations outside the litter baskets was similar in each treatment, except for the absence of macropredato rs and a greater numbers of millipedes in treatment 1. Fauna colonizin g the litter baskets clearly differed among treatments. No treatment d ifferences were detected in plant biomass and in the uptake of total-N by poplar. However, the N-15 content in treatment 4 plants was signif icantly lower than in others. On a simple net loss basis, significantl y larger amounts of the N and litter remained in the litter baskets wi th the 5 mm mesh than with smaller mesh (treatments 3 and 4). However, the dilution of N-15 in litter baskets indicated translocation of ext ernal organic matter into litter baskets, confirming the limitations o f conventional litter bag methods that use N content and gravimetric m easurements only. After including translocation and adjustment of N-15 content in litter baskets and surrounding pot soil, actual losses of N and mass were reversed: treatment 3 > 1 > 2 > 4, with treatment 4 be ing significantly less than others. Top predators had no apparent infl uence on any of the functional variables measured, suggesting that nut rient dynamics are mainly regulated by interactions occurring near the base of the detrital foodweb. (C) 1997 Elsevier Science Ltd.