Rates of soil respiration (CO2 efflux) were measured for a year in a mature Eucalyptus pauciflora forest in unfertilized and phosphorus-fertilized plots. Soil CO2 efflux showed a distinct seasonal trend, and average daily rates ranged from 124 to 574 mg CO2 m(-2) hr(-1). Temperature and moisture are the main variables that cause variation in soil CO2 efflux; hence their effects were investigated over a year so as to then differentiate the treatment effect of phosphorus (P) nutrition. Soil temperature had the greatest effect on CO2 efflux and exhibited a highly significant logarithmic relationship (r(2) = 0.81). Periods of low soil and litter moisture occurred during summer when temperatures were greater than 10 degrees C, and this resulted in depression of soil CO2 efflux. During winter, when temperatures were less than 10 degrees C, soil and litter moisture were consistently high and thus their variation had little effect on soil CO2 efflux. A multiple regression model including soil temperature, and soil and litter moisture accounted for 97% of the variance in rates of CO2 efflux, and thus can be used to predict soil CO2 efflux at this site with high accuracy. Total annual efflux of carbon from soil was estimated to be 7.11 t C ha(-1) yr(-1). The model was used to predict changes in this annual flux if temperature and moisture conditions were altered. The extent to which coefficients of the model differ among sites and forest types requires testing. Increased soil P availability resulted in a large increase in stem growth of trees but a reduction in the rate of soil CO2 efflux by approximately 8%. This reduction is suggested to be due to lower root activity resulting from reduced allocation of assimilate belowground Root activity changed when P was added to microsites within plots, and via the whole tree root system at the plot level. These relationships of belowground carbon fluxes with temperature, moisture and nutrient availability provide essential information for understanding and predicting potential changes in forest ecosystems in response to land use management or climate change.
