Dynamics of carbon fluxes with responses to vegetation, meteorological and terrain factors in the south-eastern Tibetan Plateau

Tibetan Plateau (TP) is the highest and most extensive plateau in the world and has been known as the roof of the world, and it is sensitive to climate change. The researches of CO₂ fluxes (F _C) in the TP region play a significant role in understanding regional and global carbon balance and climate change. Eddy covariance flux measurements were conducted at three sites of south-eastern TP comprising Dali (DL, cropland ecosystem), LinZhi (LZ, alpine meadow ecosystem) and Wenjiang (WJ, cropland ecosystem); amongst those DL and LZ are located in plateau region, while WJ is in plain region. Dynamics of F _C and influences of vegetation, meteorological (air temperature, photosynthetically active radiation, soil temperature and soil water content) and terrain factors (altitude) were analysed on the basis of data taken during 2008. The results showed that, in the cool sub-season (March, April, October and December), carbon sink appeared even in December with fluxes of (−0.021 to −0.05) mg CO₂ m⁻² s⁻¹ and carbon source only in October (0.03 ± 0.0048) mg CO₂ m⁻² s⁻¹ in DL and WJ site. In LZ site, carbon sink was observed in April: (−0.036 ± 0.0023) mg CO₂ m⁻² s⁻¹ and carbon sources in December and March (0.008–0.010 mg CO₂ m⁻² s⁻¹). In the hot sub-season (May–August), carbon source was observed only in May with (0.011 ± 0.0022), (0.104 ± 0.0029) and (0.036 ± 0.0017) fluxes in LZ, DL and WJ site, respectively, while carbon sinks with (−0.021 ± 0.0041), (−0.213 ± 0.0007) and (−0.110 ± 0.0015) mg CO₂ m⁻² s⁻¹ fluxes in LZ, DL, and WJ, respectively. Comparing with plain region (WJ), carbon sinks in plateau region (DL and LZ) lasted for a longer time, and the absorption sum was large and up to (–357.718 ± 0.0054) and (−371.111 ± 0.0039) g C m⁻² year⁻¹, respectively. The LZ site had the weakest carbon sink with (−178.547 ± 0.0070) g C m⁻² year⁻¹. Multivariate analysis of covariance showed that altitude (AL) as an independent factor explained 39.5 % of F _C (P < 0.026). F _C had a quadratic relationship with Normalized difference vegetation index (NDVI) (R ² ranges from 0.485 to 0.640 for three sites), an exponential relationship with soil temperature at 5-cm depth (ST ₅) at night time and a quadratic relationship with air temperature (T _a) at day time. Path analysis indicated that photosynthetically active radiation (PAR), sensible heat fluxes (H) and other factors all had direct or indirect effects on F _C in all of the three tested sites around the south-eastern TP.