Although the removal or addition of understory vegetation has been an important forest management practice in forest plantations, the effects of this management practice on soil respiration are unclear. The overall objective of this study was to measure and model soil respiration and its components in a mixed forest plantation with native species in south China and to assess the effects of understory species management on soil respiration and on the contribution of root respiration ( R r ) to total soil respiration ( R s ). An experiment was conducted in a plantation containing a mixture of 30 native tree species and in which understory plants had been removed or replaced by Cassia alata Linn. The four treatments were the control (Control), C. alata addition (CA), understory removal (UR) and understory removal with C. alata addition (UR<ce:hsp sp="0.25"/>+<ce:hsp sp="0.25"/>CA). Trenched subplots were used to quantify R r by comparing R s outside the 1-m 2 trenched subplots (plants and roots present) and inside the trenched subplots (plants and roots absent) in each treatment. Annual soil respiration were modeled using the values measured for R s , soil temperature and soil moisture. Our results indicate that understory removal reduced R s rate and soil moisture but increased soil temperature. Regression models revealed that soil temperature was the main factor and soil moisture was secondary. Understory manipulations and trenching increased the temperature sensitivity of R s . Annual R s for the Control, CA, UR and UR<ce:hsp sp="0.25"/>+<ce:hsp sp="0.25"/>CA treatments averaged 594, 718, 557 and 608<ce:hsp sp="0.25"/>g<ce:hsp sp="0.25"/>C<ce:hsp sp="0.25"/>m −2 <ce:hsp sp="0.25"/>yr −1 , respectively. UR decreased annual R s by 6%, but CA increased R s by about 21%. Our results also indicate that management of understory species increased the contribution of R r to R s .