To understand carbon source characteristics of soil respiration, a trench method was used to control carbon inputs (control treatment: CK, no litter treatment: NL, no root treatment: NR, and no litter and no root treatment: NLNR) in a Platycladus orientalis (L.) Franco plantation, so that soil respiration, temperatures and moisture content could be analyzed. Results indicate that different carbon inputs created no significant variance in soil temperatures and moisture content (P>0.05), whereas soil respiration was significantly influenced (P<0.01). During the measurement period, mean soil respiration of CK, NL, NR and NLNR were 3.69, 1.41, 3.11, and 1.99 μmol/m 2 /s, respectively. Soil respiration was reduced by 36.28 ±5.79%, 15.25 ±2.62% and 25.53 ±4.95% in the NL, NR and NLNR treatments, respectively. The proportions of mineral soil, litter and root respirations were 40.95%, 46.91% and 12.14%, respectively. Different carbon inputs caused no significant R 2 variance in the temperature index model or in the temperature-water combined model. These results provide guidelines for investigating Platycladus orientalis ecosystem carbon emissions, as well as carbon input and output balance, and the role of soil respiration in the carbon cycle.