Variation in soil respiration under the tree canopy in a temperate mixed forest,central China,under different soil water conditions

The forest canopy cover can directly and indirectly affect soil conditions and hence soil carbon emission through soil respiration. Little is known, however, on the effects of canopy cover on soil respiration under the canopy of different tree species and soil water conditions. We have examined the variation in soil respiration at different soil water conditions (dry <10 %, wet >20 %, v/v) under different tree canopy covers in comparison with the canopy interspace in a temperate coniferous (Pinus armandii Franch) and broadleaved (Quercus aliena var. acuteserrata) mixed forest in central China. The results show that soil respiration measured under tree canopy cover varied with canopy size and soil water content. Soil respiration under small-sized canopies of P. armandii (PS) was higher than that under large-sized (PL) canopies, but the difference was only significant under the dry soil condition. However, soil respiration under large-sized canopies of Q. aliena (QL) was significantly greater than that under small-sized (QS) canopies under both dry and wet soil conditions. The difference in soil respiration between differently sized canopies of Q. aliena (33.5–35.8 %) was significantly greater than that between differently sized canopies of P. armandii (2.4–8.1 %). Differences in soil respiration between inter-plant gaps and under QS canopies in both the dry and wet soil conditions were significant. Significant increases in soil respiration (9.7–32.2 %) during the transition from dry to wet conditions were found regardless of canopy size, but the increase of soil respiration was significantly lower under P. armandii canopies (9.7–17.7 %) than under Q. aliena canopies (25.9–31.5 %). Our findings that the canopy cover of different tree species influences soil respiration under different soil moisture conditions could provide useful information for parameterizing and/or calibrating carbon flux models, especially for spatially explicit carbon models.