Respiration of downed logs in pine and oak forests in the Qinling Mountains, China

Approximately 70% of the carbon (C) stored in logs is released into the atmosphere, representing an important source of CO2 lost from terrestrial ecosystems. Log respiration (Rlog) has gained attention as a core issue in global C cycle research. In forest ecosystems that contain many logs, the Rlog flux can convert forests from C sinks into C sources; thus, Rlog should be considered in relevant research to avoid underestimating the CO2 losses in the forest C cycle. Limited information is available regarding Rlog from natural forests, and many uncertainties remain about the magnitude of Rlog. In our study, Rlog was measured in situ by infrared gas analysis in Pinus armandi and Quercus aliena var. acuteserrata forests in the Qinling Mountains, China. The objectives of this study were (1) to reveal the seasonal variation patterns of Rlog; (2) to systematically analyze the relationships between Rlog and various factors, including the tree species, decay class, temperature, water content, and chemical composition; and (3) to estimate the annual Rlog flux in P. armandi and Q. aliena var. acuteserrata forests in the Qinling Mountains, China. This study presents a full year time series of Rlog measurements for 30 logs (3 replicate logs × 5 decay classes × 2 tree species). The Rlog measurements were repeated 468 times for each log from May 2014 to April 2015. The log temperature (Tlog), air temperature (TA), soil temperature (TS) at a depth of 10 cm, and log water content (Wlog) were measured simultaneously with Rlog. Moreover, the log density (Dlog) and chemical composition (C, nitrogen (N), phosphorus (P), potassium (K), calcium (Ca), and magnesium (Mg)) were determined. Our results showed significant seasonal variation in Rlog for both species, which corresponded to variations in Tlog during the study period. The annual mean Rlog of Q. aliena var. acuteserrata (1.69 ± 1.60 μmol m−2·s-1) was higher than that of P. armandi (1.55 ± 1.43 μmol m−2·s−1), but the difference was not significant (P = 0.61). The decay classes, Tlog, Wlog, and the N, P, Ca, and Mg concentrations were positively correlated with Rlog. Moreover, the K concentration was negatively correlated with Rlog, and the C concentration in logs was not correlated with Rlog. The total annual Rlog flux did not differ significantly between the P. armandi (67.25 ± 7.28 g C·m−2·y−1) and Q. aliena var. acuteserrata (74.69 ± 9.31 g C·m−2·y−1) forests (P = 0.26). These results provide insight into the factors responsible for seasonal changes in Rlog and can improve estimates of the annual Rlog flux in natural forests.