臺灣溪頭孟宗竹林碳循環特性

Soil Carbon Cycling of a Moso Bamboo Forest in Xitou, Central Taiwan

林孟穎(Meng-Ying Lin)

指導教授 ： 久米朋宣

國立臺灣大學/生物資源暨農學院/森林環境暨資源學研究所/碩士(2014年)

https://doi.org/10.6342/NTU.2014.02632

摘要

近十幾年來在東亞地區發現竹林逐漸擴散的現象，會造成當地碳循環的改變，土壤呼吸是一個判斷森林生態系是否為碳源或碳庫的重要依據。了解森林生態系中土壤呼吸的空間異質性不僅能夠深入探討二氧化碳的變化，更能獲得比較適當的樣本數以推估土壤呼吸的年量，並且得知其如何影響環境氣候。
本研究的目的是希望了解分布在台灣中部山區雲霧林帶的孟宗竹林(Phyllostachys pubescens)其土壤碳循環特性。因此本研究檢測在空間變異中土壤呼吸的特性以及影響的因子，進而應用空間變異的特性來推估土壤呼吸的年量，並且利用土壤呼吸年量、地上部和地下部碳蓄積的測量來檢視孟宗竹林的碳循環過程。
試驗地在臺灣大學實驗林的溪頭營林區中一片孟宗竹純林中，設置一個435平方公尺的樣區進行研究，並在樣區中設置了28個樣點已量測土壤呼吸的空間變異特性，以及觀察相關的環境因子，包含土壤溫度、土壤水分、土壤孔隙率、土壤碳氮比以及林分結構與土壤呼吸間的關係。在空間變異的部份則是利用了地理統計中的半變異數來分析其特性。並測量竹林中各碳庫的碳蓄積量，包含土壤、根、地下莖、枯枝落葉和地上部竹林來了解孟宗竹林的碳循環特性。
土壤呼吸全年範圍為0.92 to 5.10 μmol CO2 m-2 s-1，最高發生於六月最低發生於一月。時間變異的控制因子主要受到土壤溫度的影響(R2= 0.873)。空間變異較大值主要發生於六月至八月間，九月和三月間的空間變異情形趨緩，而在空間自相關的範圍(range)參數為2.95- 16.96公尺,此範圍和熱帶雨林的結果相近。較大值一樣發生於六、七、八月，說明了較溫暖的季節會導致較高的土壤呼吸值以及比較大的範圍參數。本研究中發現控制土壤呼吸空間變異的因子為土壤水分和範圍間的林分結構(3.5- 8 m)，這2個環境因子和土壤呼吸呈現顯著負相關不論是在每個月分的尺度或是年間的尺度亦然。
本研究中土壤呼吸年量為3.19 μmol CO2 m-2 s-1 (=1208.02 g C m-2 yr-1)。這個結果高於相似氣候帶的針葉林與天然闊葉林。地上部和地下部的碳蓄積分別為36.37和111.75 Mg C ha-1。淨初生產力為(12.48 Mg C ha-1 yr-1)，高於鄰近的柳杉林(9.6 Mg C ha-1 yr-1)、台灣杉林(8.7 Mg C ha-1 yr-1)以及香杉林(2.95- 5.11 Mg C ha-1 yr-1)。並且此研究的淨生態系生產力(NEP) (= 7.17 Mg C ha-1 yr-1)也高於鄰近柳杉林，這結果不僅說明孟宗竹林扮演了可觀的碳庫角色更指出竹林於地下部的碳循環過程在森林生態系中扮演了重要而且獨特的特性。

關鍵字

土壤呼吸 ； 空間變異 ； 控制因子 ； 孟宗竹林 ； 碳循環 ； 淨初級生產力 ； 淨生態系生產力

並列摘要

The expansion of bamboo forest to surrounding ecosystems in eastern Asian countries such as Taiwan can alter the local carbon balance. Soil respiration is a key path to determine whether forest ecosystem is carbon sink or carbon source. Spatial heterogeneity of soil CO2 efflux (Rs) in forested ecosystems is essential not only for understanding CO2 dynamics but also for suitable sampling design to estimate annual Rs and the response to environmental changes. The aim of this study is to understand the characteristic of soil carbon cycle in a Moso bamboo (Phyllostachys pubescens) forest, situated in a montane cloud forest zone of central Taiwan. To this aim, this study examined spatial variations in Rs and their control factors, and estimated annual Rs with the spatial representatives. Also, using the annual Rs and carbon stock measurements including above and belowground, this study estimated carbon cycles in the moso bamboo forest.
We selected a 435 m2 plot in a pure Moso bamboo forest in National Taiwan University Forest, central Taiwan. The 28 soil CO2 efflux measurement locations were set in the plot. We observed the several factors including soil temperature, soil water content, soil porosity, root mass, soil C/N ratio and stand structure to examine controlling factors regulating spatial variations in Rs. Also, the spatial variation of Rs was analyzed by using semivariance. As well, to understand the characteristic of carbon cycle in Moso bamboo forest, we estimated soil carbon stock, root carbon stock, rhizome carbon stock, litter fall carbon stock and aboveground bamboo carbon stocks. Rs ranged from 0.92 to 5.10 μmol CO2 m-2 s-1, which was highest in June and lowest in January. The temporal changes in Rs averaged over 28 locations were strongly impacted by soil temperature (R2= 0.873). Higher spatial variation was in June to August, and less in September to March. The range of autocorrelation in soil CO2 efflux was 2.95 to 16.96 m, and was comparable with that of tropical rainforests. The larger range of autocorrelation in Rs was appeared up in June, July, and August, which indicated that warm season would lead to higher Rs and larger range of autocorrelation. We found the controlling factors to regulate the spatial variation in Rs were soil water content (SWC) and stand structure within 3.5 to 8 m around the measurement locations. These two controlling factors were significantly correlated with the Rs negatively whether it was in monthly and annual time scale.
The annual mean Rs with the consideration of Rs spatial variation characteristics was 3.19 μmol CO2 m-2 s-1 (=1208.02 g C m-2 yr-1) from May 2013 to April 2014. Rs in the bamboo forest was much higher than coniferous and natural broadleaf forest. Aboveground and belowground carbon stock was 36.37 and 111.75 Mg C ha-1, respectively. This study found that NPP in this study (12.48 Mg C ha-1 yr-1) was higher than cedar forests (9.6 Mg C ha-1 yr-1), Taiwania forest (8.7 Mg C ha-1 yr-1), and China fir forest (2.95- 5.11 Mg C ha-1 yr-1) in central Taiwan. Further, NEP in this site was estimated as 7.17 Mg C ha-1 yr-1, probably larger than the cedar forest. Therefore, the NEP informed that the Moso bamboo forest was a strong carbon sink in terrestrial forest ecosystem. These results indicated that bamboo forest is a specific forest than others and the belowground processes play an essential role of forest carbon cycle.

並列關鍵字

Soil CO2 efflux ； spatial variation ； control factors ； Moso bamboo forest ； NPP ； NEP

參考文獻

Chen, T.H., Liu, C.P. and Chung, H.Y. (2011) Growth and biomass of Moso bamboo in FengHuang mountain, Nantou county. Quarterly Jounal of Chinese Forestry, 44 (1), 19-28. (in Chinese)

Chiou, C.R., Chen, T.H., Lin, Y., Yang, Y. and Lin, S. (2009) Distribution and Change Analysis of Bamboo Forest in Northern Taiwan. Quarterly Journal of Chinese Forestry, 42 (1), 89-105.

Lin, Y., Wang, C. and Wu, S. (2011) Analyzing carbon conversion factors of four species of Taiwanese bamboo. Taiwan Journal of Forest Science, 26 (4), 341-355.

Liu, S.C. and Kao, Y.P., 1998. Generalized biomass equations for moso bamboo and plantations. Bulletin of Taiwan Forestry Research Institute, 3 (1), 393-406. (in Chinese)

Sun, C.H. and Yen, T.M. 2011 Using allometric equation to predict bamboo biomass based on combined data of makino bamboo (Phyllostachys makinoi) and moso bamboo (Phyllostachys pubescens). Quarterly Journal of Forest Research, 33 (3), 1-8. (in Chinese)