Abstract:The stem surface CO2 flux (RW) is the major component of tree autotrophic respiration, and represents tree growth activities and metabolism. The RW variability, however, has not been well quantified for the Chinese temperate forests. The RW was in situ measured for 14 major tree species coexisting in the temperate forests in northeastern China. The species included diffuse-porous species (Betula platyphylla, B. costata, Populus davidiana, Tilia amurensis, and Acer mono), ring- and semi-ring-porous species (Ulmus propinqua, Phellodendron amurense, Juglans mandshurica, Quercus mongolica, and Fraxinus mandshurica), and coniferous species (Larix gmelinii, Pinus koraiensis, Picea koraiensis, and P. sylvestris var. mongolica). The objective of this study was to (1) quantify intra- and inter-specific variations in RW for the tree species, and (2) examine changes in RW with tree diameters and seasons in order to mechanistically understand forest carbon cycling and provide solid data for developing and validating the carbon cycling model for the temperate forests. For each tree species, 12-18 trees were randomly sampled to cover as wide DBH (diameter at breast height) range in the stands as possible. A polyvinyl chloride collar (inner diameter 10.4 cm, height 5.0-6.0 cm) was cut and polished to fit the stem surface shape of each sample tree, and installed on the north side at breast height (1.3 m). The collar was attached with waterproof silicon adhesive to the stem surface that was pretreated without causing any injury of the live tissues, and remained in place throughout the measuring period. An infrared gas exchange analyzer (LI-6400 IRGA) was used to measure the RW once every month from 08:00 to 17:00 within three to four consecutive sunny days during the period from June to October 2009. Stem temperature at 1 cm depth beneath the bark (TW) was simultaneously measured with a digital thermometer. Tree species, measuring month and their interactions significantly influenced the RW (P < 0.001). The mean RW during the measuring period varied from 1.32 μmol CO2 · m-2 · s-1 for P. amurense to 3.12 μmol CO2 · m-2 · s-1 for L. gmelinii. The mean RW for the ring- and semi-ring-porous species was greater than that for the diffuse-porous species, while the mean RW for the coniferous species varied greatly. The mean RW for all species showed a unimodal seasonal pattern, with the maximum and minimum occurring in July and October, respectively. The intra-specific mean absolute (standard errors) and relative variations (coefficients of variation) varied from 0.11-0.29 μmol CO2 · m-2 · s-1 and 61%-89%, respectively. The RW tended to increase with DBH increasing for all species, but the forms and determination coefficients of the regression models were species-dependent. There were significant relationships between mean RW and DBH (P < 0.05) for all species except for U. propinqua and F. mandshurica (P > 0.05), suggesting that DBH be a simple and practical proxy for predicting and extrapolating tree- or stand-level RW. This study highlights the importance of taking the intra- and inter-specific variations in RW measurements into account in cross-comparing RW and extrapolating chamber-based RW measurements to tree- or stand-level estimates.
