臺灣溪頭柳杉人工林之樹液流特性

Abstract

本研究探討在臺灣中部柳杉林的蒸散特性趨勢。樹液流測量為一項良好的方法用來測量森林的蒸散量，但有報告指出經過一段長時間，利用熱擴散法探針之量測法可能會因為樹木的生長、傷害或是其他水流途徑的改變減少量測的能力。因此，本研究在位於臺灣中部的臺大溪頭實驗林的柳杉人工林中實施樹液流測量，試圖探討樹液流的長期特性。為了達到目的，首先從兩個方面了解單棵樹規模樹液流的季節變化: 1)水蒸氣壓差(vapor pressure deficit)的影響。2)其他氣候因子(氣溫及土壤水分)的影響。其次，從三個面向瞭解林分規模的蒸散量: 3)樹液流是否會受到長期測量的影響而造成表現力下降。4)估算林分規模的蒸散量並了解其季節變異。5)與其他地區比較年蒸散量。研究結果顯示樹液流速會隨著VPD的增加而增加並於VPD大約0.5 kPa時漸趨穩定，且在穩定期的樹液流速會隨著季節變化。總體來說，夏天的數值是最高的，而冬天是最低的。比較一整年的變異幅度，則可發現內部的樹液流速(2-4公分)並不如外部明顯(0-2公分)。從結果來看，樹液流速會因為長期插探針的影響而稍微削弱其表現力。而樹液流除了受到VPD影響外，也和溫度及土壤水分含量有關，且溫度比起土壤水分含量對於季節性的影響可以說較為明顯。另外，我們也發現每棵樹對於季節變異反應的個體差異。林分蒸散量的部分，在約莫夏季六至九月份間及春季四到五月的蒸散量較大，相對地，在三月較少，而十一月中旬至十二月則更少，林分蒸散的季節變異幾乎與該季節的VPD高度相關。最後，雖然在差不多的氣候條件，在溪頭人工柳杉林的年蒸散量明顯比其他地區的雲霧林來得低。

This study was conducted to clarify seasonal trends of sap flow with total amount of annual transpiration rate in a Japanese Cedar plantation, Sitou NTU experimental forest, in central Taiwan. To this aim, first, this study examined seasonal changes in tree-scale sap flow in terms of two aspects: 1) the effect of VPD and 2) the effects of other meteorological elements (air temperature and soil water content). Second, this study examined stand-scale transpiration (Et) in terms of three aspects: 3) the effects of reduction in sap flow due to annual sap flow sensor installation on Et estimates 4) clarifying seasonal variation in Et (5) characterizing from the comparison with those of other literatures. The sap flow measurements were carried out at sapwood depth of 0-2 cm and 2-4 cm in 19 individuals through 2010 September to 2011 August. This study found the sap flow density mostly increased with increases in VPD and tended to the plateau when VPD exceeded about 0.5 kPa, and the sap flow density at the plateau varied with seasons. Generally, the values (sap flow density at the VPD=0.5kPa) were the highest in summer and lowest in winter. Comparing the range of variation in the relationships between VPD and sap flow density, we can see that inner sap flow (2-4 cm) is not obvious as outer one (0-2 cm) during the studied year. Results suggested that sap flow density was slightly affected by annual sap flow sensor installation. In addition, sap flow density was affected by temperature and soil water content besides vapor pressure deficit, although the significance of the affections were different among individuals. Further, it seems temperature effects on the seasonality could be more significant than that of soil moisture in this site. At the stand-scale estimates, Et was larger during summer from June to September, as well as during spring from April to May. Conversely Et was lower during March, and very low from November to December. The seasonal variations in Et were mostly related to those of VPD. Also, we clarified that the annual Et at this site was significantly lower than those in other cloud forests.