Predicting nitrogen supply in plantation eucalypt forests

Nitrogen mineralization was measured experimentally and predicted by simulation modelling in three plantations of Eucalyptus globulus in south-western Australia where annual rainfall ranged from 600 to 1500 mm. Nitrogen mineralization was measured in the field each 4 weeks for 3 yr (2 sites) and 2.8 yr (1 site) using the in situ coring technique. Potentially available N was estimated at each sampling time using anaerobic incubations in the laboratory. In the field, soil temperature was logged each hour and soil moisture was estimated gravimetrically each month at the start and end of each field exposure period. The relationship of N mineralization to environmental variables was determined through aerobic incubations of soil samples in the laboratory at 6 temperatures (4 to 42 degrees C) and 10 soil moistures (air dry to 10 kPa soil moisture tension). A generalized model using an exponential function to describe temperature response and a logistic function to describe moisture response was fitted to the laboratory N mineralization incubation data. This function was used as a basis for predicting N mineralization in the field from measured soil moisture and temperature. Model output was scaled by tuning the simulated and observed N mineralization rates over the experimental period at each site. Finally, independent scale factors were derived by regressing the tuned scale factors against soil chemical (anaerobic N) and physical characteristics (soil moisture content at 25 kPa moisture tension) and these were used in model simulations to predict N mineralization. There was a marked seasonal pattern in monthly rates of mineralization related to the seasonally dry climate in south-western Australia. Rates were highest during late winter and spring when microbial activity was not limited by soil moisture deficits and soil temperatures were rising, and lowest in summer when the surface soils dried to below wilting point. Model simulations generally reflected the observed pattern of mineralization and mean annual rates of N mineralization at each sites were predicted within about plus or minus 20%. Crown Copyright (C) 1999 Published by Elsevier Science Ltd. All rights reserved.